.
~
Microneurosurgery 1114Volumes
M.G.Ya~argil IVA CNS Tumors: Surgical Anatomy Neuropathology Neuroradiology Neuro...
118 downloads
972 Views
127MB Size
Report
This content was uploaded by our users and we assume good faith they have the permission to share this book. If you own the copyright to this book and it is wrongfully on our website, we offer a simple DMCA procedure to remove your content from our site. Start by pressing the button below!
Report copyright / DMCA form
. .
~
Microneurosurgery 1114Volumes
M.G.Ya~argil IVA CNS Tumors: Surgical Anatomy Neuropathology Neuroradiology Neurophysiology Clinical Considerations Operability Treatment Options
Thieme
...
~, ~
ProfessOrJosef Kliiigler (18-88-1963), Professor of Anatomy at the University of Basel, Switzerland, from whom 1 learned a unique process for the dissection of'the brain and particularly the white matter systems, and thereby gained an inimitabk neuroanatomical perspective óf these complex structures. Dissection of both Hippocampal formations, by Professor1.Klingler, who always maintained that the Hippocampus was the most complex structure in all of Nature, possessing qualities which are not immediately apparent from its confined anatomical structure.
Microneurosurgery in 4 Volumes
M. G. Y a~argil
1 Microsurgical Anatorny of the Basal Cisterns and Vessels of the Brain, Diagnostic Studies, General Operative Techniques and Pathological Considerations of the Intracranial Aneurysrns 11 Clinical Considerations, Surgery of the Intracranial Aneurysrns and Results 111 A A VM of the Brain, History, Ernbryology, Pathological Considerations, Hernodynarnics, Diagnostic Studies, Microsurgical Anatorny 111 B A VM of the Brain, Clinical Considerations, General and Special Operative Techniques, Surgical Results, Nonoperated Cases, Cavernous and Venous Angiornas, Neuroanesthesia
IV A CNS Turnors: Surgical Anatorny, Neuropathology, Neuroradiology, Neurophysiology, Clinical Considerations, Operability, Treatrnent Options
IVB Microsurgery of CNS Turnors
Georg Thieme Verlag Stuttgart . NewYork Thieme Medical Publishers,Inc. NewYork
IVA
CNS Turnors: Surgical Anatorny, N europathology, N euroradiology, N europhysiology, Clinical Considerations, Operability, Treatrnent Options
M. G.Ya§argil
.
Collaborators: T.E. Adamson, G. F. Cravens, R. 1.Johnson, 1.D. Reeves, :P.1.Teddy, A.Valavanis, W.Wichmann, A. M. Wild and :P.H.Young Anatomical preparations by A. Lang, U.Türe Illustrated
by :P.Roth
1126 illustrations,
58 tables
1994 Georg Thieme Verlag Stuttgart . New York Thieme Medieal Publishers, lne. NewYork
..
IV
Author's Address: M. G. Ya~argil, M. D. Professor and Chairman (emeritus) Dept. of Neurosurgery University Hospital Zurich, Switzerland
Collaborators of Volume IV A: T. E. Adamson, M. n Charlotte Neurosurgical Associates, P.A. 1010 Edgehill Road North, At East Morehead, Charlotte, North Carolina 28207-1830,USA G. F. Cravens, M. n, FACS Center for Neurological Disorders 1319 Summit Ave. Fort Worth, Texas 76102, USA R. 1.Johnson, M. n Department of Neurosurgery, Louisiana State University, School of Medicine, Medical Center, 1542Tulane Avenue, New Orleans, LA 70112-2822 A. Lang Anatomical Preparator for Biology and Medicine, Anatomical Institute, University of Zurich, Winterthurerstrasse 190, CH-8057 Zurich 1.n Reeves, M. D. Neurosurgeon, 1801Fairfield Ave., Suite 200, Shreveport, LA 71101, USA
P. Roth Scientific Artist, Neurosurgical Department, University Hospital of Zurich, CH-8091 Zurich P. 1. Teddy, DPhil, FRCS Consultant Neurosurgeon, The Department of Neurological Surgery, Oxfordshire Health Authority, The RadcIiffe Infirmary, Oxford OX2 6HE, UK U. Türe, M. D. Neurosurgeon Neurosurgical Department, Sisli-Etfal Hospital, Istanbul, Turkey A. Valavanis, M. D. Professor, Director and Chairman, Institute of Neuroradiology, University Hospital, CH-8091 Zurich W. Wichmann, M. n Institute of Neuroradiology, University Hospital, CH-8091 Zurich A. M. Wild, FRCS Neurosurgeon, Coordinator Joint European Project on Minimally Invasive Neurosurgery and Neuroendoscopy Eaton Socon Huntingdon Cambridge PE19 3PU, GB P. H. Young, M. D. Microsurgery and Brain Research Institute, P. C. 6725 Chippewa Street, Sto Louis, Missouri 63109, USA
<9 1994 Georg Thieme Verlag, RüdigerstraBe 14, D-70469 Stuttgart, Germany Thieme Medical Publishers, Inc., 381 Park Avenue South, New York, N.Y. 10016 Typesetting by primustype Hurler GmbH, D-73274 Notzingen, typeset on Linotronic 330 Printed in Germany by Appl, D-86650 Wemding
ISBN 3-13-645101-5 (GTV, Stuttgart) ISBN 0-86577-260-6 (TMP, New York) 1 2 3 456
Library of Congress Cataloging-in-Publication Data Ya~argil, Mahmut Gazi. CNS tumors : surgical anatomy, neuropathology, neuroradiology, neurophysiology : cIinical considerations, operability, treatment options / M. G. Ya~argil : collaborators, T. E. Adamson... [et aL] : anatomical preparations by A. Lang : illustrated by P. Roth. p. cm. - (Microneurosurgery : 4) IncIudes bibliographical references and index. ISBN 3-13-645101-5 (GTV, Stuttgart). - ISBN 0-86577-260-6 (TMP, New York) 1. Central nervous system- Tumors-Surgery. 2. Microsurgery. I. Adamson, T. E. II. Title. III. Series: Ya~argil, Mahmut Gazi. Microneurosurgery : 4. [DNLM: 1. Central Nervous System Neoplasms-physiopathology. 2. Central Nervous System Neoplasms-surgery. WL 358] RD663.Y37 1994 619.99' 281059-dc20 DNLM/DLC for Library of Congress 93-37101 CIP
Cover drawing by P. Roth (modified after Robert S. Gessner, Construction 1, 1942) Important Note: Medicine is an ever-changing science undergoing continual development. Research and cIinical experience are continually expanding our knowledge, in particular our knowledge of proper treatment and drug therapy. Insofar as this book mentions any dosage or application, readers may rest assured that the authors, editors and publishers have made every effort to ensure that such references are in accordance with the state of knowledge at the time of production of the book. Nevertheless this does not involve, imply, or express any guarantee or responsibility on the part of the publishers in respect of any dosage instructions and forms of application stated in the book. Every user is requested to examine carefully the manufacturers' leaflets accompanying each drug and to check, if necessary in consultation with a physician or specialist, whether the dosage schedules mentioned therein or the contraindications stated by the manufacturers differ from the statements made in the present book. Such examination is particularly important with drugs that are eitber rarely used or have been newly released on the market. Every dosage schedule or every form of application used is entirely at the user's own risk and responsibility. The authors and publishers request every user to report to the publishers any discrepancies or inaccuracies noticed.
Any reference to or mention of manufacturers or specific brand names should not be interpreted as an endorsement or advertisement for any company or producto Some of the product names, patents and registe red designs referred to in this book are in fact registe red trademarks or proprietary names even though specific reference to this fact is not always made in the textoTherefore, the appearance of a name without designation as proprietary is not to be construed as a representation by the publisher that it is in the public domain. This book, incIuding all parts thereof, is legally protected by copyright. Any use, exploitation or commercialization outside the narrow limits set by copyright legislation, without the publisher's consent, is illegal and liable to prosecution. This applies in particular to photostat reproduction, copying, mimeographing or duplication of any kind, translating, preparation of microfilms, and electronic data processing and storage.
.
Acknowledgments
Thefinal preparation of this volume began in 1985 and some of thechaptershave been written, rewritten, corrected, and changed by visiting colleagues, mainly from the UK and USA, over this periodoTheir hard work and advice is greatly appreciated. Besides this, we have worked closely over the years with the membersof the Institute of Brain Research (Prof. K. Akert), Department of Neurology (Prof. G. Baumgartner, Prof. H.-G. Wieser,Prof. T. Landis, Dr. M. Regard), Neuroradiology (Prof. A. Valavanis,Dr. W. Wichmann), Neuropathology (Prof. P. Kleihues, Prof.O. D. Wiestler, Dr. C. Moll), Neuro-Nuclear-Medicine (Prof. G. K. von Schulthess), Clinical Immunology (Prof. A. Fontana, Dr. K. Frei), Internal Medicine (Dr. G. Siegenthaler), Neuroanesthetists(Dr. M. Curcic, Dr. M. Kis, Dr. P. Mica) and last but not leastthe senior consultant (Dr. H.-G. Imhof) and the chief residents (Dr. K. von Ammon, Dr. E. Cavazos, and Dr. D. Jeanmonod) of the Department of Neurosurgery at the University Hospitalof Zurich. This whole work arose from the various discussions with these colleagues. We have also had immense help from the Neuroanatomical Departmentin Zurich; who allowed us to study the very many high-quality preparations they have on exhibition. Additionally, someof the injection preparations shown were specially made by
A. Lang. We would also like to thank the generosity of Prof. W. Zenker, Prof. P. Groscurth, and Prof. S. Kubik for several neuroanatomical figures within this book. I would also like to thank the secretaries M. Traber and M. Jent for collecting the clinical material of all the cases and preparing the manuscripts. My thanks and appreciation go to Mr. Roth for his many excellent drawings and illustrations within this volume, and throughout the whole series, as well as for collecting the neuroimaging studies used in the case presentations. My thanks also to Mrs. S. Hess for the photography of many of the figures within this volume. I would like to add a special thanks to Miss R. Frick for collecting, checking, and collating the many references and literature searches, without whose help this book could not have been completed. Finally, a special thanks to Thieme for their help, patience, and advice, in particular to Dr. h. c. G. Hauff, owner of Thieme, and to A. Menge and R. Zeller, during the preparation and production of this book. In autumn 1993
M. G. Ya~argil
Preface
The purpose of Volumes IV A .and B is to discuss the reliability of the present basic neuroscientific investigations and to present clinical observations, management strategies, microsurgical approaches, and operative results in over 3400 patients with CNS tumors. The decision-making process in the treatment of CNS tumors involves the application of knowledge from several associated disciplines. Precision in determining the diagnosis, localization, treatment options, and prognosis of each lesion requires accurate neuroanatomical, neurophysiological, neuropathological, neuroinvestigational, and neuroclinical knowledge. Those involved in the care of CNS tumor patients must be familiar with the important role each of these fields plays in the final determination of a treatment strategy. In addition, we must be aware of the remarkable recent advances being made in these areas. Neuropathologists, using immunocytochemistry and allied techniques, have be en able to offer us an increasingly precise classification of tumors. However, the understanding of fundamental tumor biology is still elusive. In addition, they have not yet provided an accurate method of anticipating the growth and growth pattern, spread, migration, or penetration of an individual tumor, the factors which are most helpful in the determination of tumor operability and intensity of treatment. Clinical observations, on the other hand, indica te that the invasive potentials of benign and malignant extrinsic and intrinsic tumors are limited by anatomical and biological barriers, at least in the initial and intermediate phases of tumor growth. This important fact indicates great changes concerning the therapeutic options (see Chapter 2, Neuropathology). During the initial period of microsurgery in Zurich (between 1967 and 1975), the diagnosis and the indications for surgery of CNS tumors were based in the classic manner on the complete history, physical examination, and adjuvant studies, including EEG, pneumoencephalography, angiography, and myelography. Within the past 17 years, significant diagnostic advances have been achieved: the visualization techniques of cr scan, functional and dynamic MRI, MR spectroscopy, and other techniques (PET, SPECT, Doppler) have proved to be great innovations in the immediate recognition of CNS tumors and peritumoral changes in both the pre- and postoperative phases. These modalities, combined with microtechniques, have greatly improved the surgical treatment of CNS tumors. Along with these advances, however, there are new problems in diagnosis, indications for treatment, and decisioRSon the treatment options of these lesions, which are presented and discussed in Neuroradiology (Chapter 3) and Operability (Chapter 5). The century-old neurophysiological concepts concerning the balance of intracranial and intraspinal volumes (parenchymal mass, blood, cerebrospinal fluid, and tumor) were based only on mechanical forces. Clinical observations, however, combined with
modern neuroimaging techniques, demand a more complex concept. This new dynamic concept should not only account for mechanical factors but must allude to the complex biochemical, biological, and immunological interactions between normal tissue and pathological tissue. An understanding of this dynamic normal tissue-Iesion imbalance may then explain the changes seen in pathological situations and their resultant consequences (see Chapter 4, Neurophysiology). Surgical anatomy is another essential issue; it is not only the summary of available knowledge gained from macroscopical, microscopical, and ultramicroscopical investigations of the anatomist or from observations and studies of the neuroradiologist, neurophysiologist, and neurosurgeon. Although three-dimensional morphological and functional anatomical models of the brain are very instructive, they rema in segregated models of reality. The individual variations of "morphology-pathomorphology" and "physiology-pathophysiology" are other dimensions, which provoke and challenge the surgeon to continuous imagination and projection of adequate dynamic concepts. Additionally, the reality of a surgical exploration involves a gross amount of still undiscovered, therefore unknown, morphological and functional facts. It would be ideal to have the assistance of the clinical neuroanatomist, neurophysiologist, and computer engineer during the whole perioperative phase. Personally, I am convinced that this will be the case in the near future in the modern, computerized operating room, where the combined morphological and functional anatomica13-D dynamic models can be examined and checked according to the needs of the given situation with respect to the desires of the surgeon. The ongoing technical developments (i. e., frameless stereotactic localization system with the help of infrared LED devices) require from us accurate anatomical knowledge and new surgical anatomical concepts. The unique construction and functions of the brain require, topographically, a more differentiated concept according to the functional anatomy and sophisticated predilection site of the lesions; to provide the reader with an adequate and systematic picture of the complex brain anatomy, the gyral convolutions, the gyral segments, the sectorial and peduncular organization of the white matter, the topography of the basal ganglia and central nuclei, the pattern of arterial and venous vascularization, pathways of neuronal migration, and pathways of neurotransmitters, I have used not only my own material of sectional specimens and those of the Anatomical Institute of the University of Zurich, but also related, essential, wellknown pictures of other authors. It is my intention to give interested colleagues my personal surgical anatomical concepts, which I need and use for my surgical decisions and actions. To facilitate the transformation of surgical anatomical images and their memorization, the figures have often been colo red as in Chapter 1, Neuroanatomy.
vii Our ability to continue to rely on the old paradigms of neuroanatomy, neuropathology, neuroradiology, and neurophysiologyis questionable. Therefore, conceptually innovative ideas relating to individual surgical anatomy, neuroimaging, neuropathology,and neurophysiology are presented here. A new definitionand approach to operability is examined, perhaps giving lifelOa new paradigm for the treatment of CNS tumors. Advancesand limitations in neuroanatomy, neuropathology, neuroimaging,and neurophysiology and their relation to clinical neurosurgery,especially regarding the operability of CNS tumors andthe treatment options, are presented and discussed in Chapter5, ClinicalConsiderations. Noninvasive explorations and pure tumorectomies have beenperformed along the transcisternal, transsulcal, and transfissuralpathwayswithout lobectomy or gyrectomy, fully respecting the brain and other vital structures. These microsurgical techniquesof "nontraumatization" of the brain or with "very minimal traumatization,"are documented in numerous cases with pre- and postoperativethree-planar MR images, at the end of Chapters 2, 3.4, and 5. 1 have been asked from visiting colleagues to write this volumeand describe the applied microsurgical techniques with perioperativesurgical-anatomical perspectives and concepts of the surgeon. Each surgical action comprises not only science, experience,knowledge, and techniques, but also artistic, philosophical,and religious attitudes from a neurosurgeon. We respect thaterasisthecohesivepowerwhichholds the multimodal potentialsof the human brain together with its continuous stimulation anddrives. The artistic masters of the Renaissance discovered the perspectiveof three-dimensionality and immediately began to study theimpactof functionality in exact measurement techniques. Up untilthe twentieth century, orientation depended on the topology ofa pictoriallyordered world of objects in which relationships are expressedby spatial metaphors. Cubism broke down these forms ofperception,dismantling the identity of the figurative into nonidenticalaspects that could be projected upon each other. Constructiveand concrete art entered into deeper dimensions, searchingtheelementsof functionality. These trends culminated il) mathematizedart and philosophy, and computerized science and tech-
niques. The masterpieces of constructive and concrete art deal scientifically with modern modular, serial and parallel processes, some relating to the philosophy of monads from G. W. Leibniz (1646-1716), whose genius had already developed an early computero The neurophysiologist introduced the concept of a modular and columnar cortical organization of the brain (Y. B. Mountcastle, 1957; D. Hubel, T. Wiesel, 1977): computational neurobiology has arrived. The convergence of science, technology, art, and philosophy is obvious. The displayed masterpieces of constructive and concrete art in this volume should be seen symbolically as a prelude to the related topics of the different chapters. It is hoped that the combination of pure art with the art of scientific illustration will enhance the interest of the reader. We should remember that the fortunate cooperation between H. Cushing and W. E. Dandy with the gifted illustrator Max BrOdel at Johns Hopkins, Baltimore, opened new perspectives for scientific medicine at the beginning of 20th century. The presented experiences, observations, ideas, perspectives, and concepts are intended to stimulate open discussions, which will subsequently further advances throughout the entire field of neurosurgery. Volume IV B is devoted to surgical strategies, tactics, and techniques and to specific types of tumors and their management. It will include detailed descriptions of both surgical approaches and the microtechnical removal of different types of CNS tumors, and is followed by a complete analysis of the operated cases of the senior author. Initially, Volumes IV A and B had been scheduled to be published at the latest in 1990. My daily clinical work, however, did not permit me to keep my promise. Only one part of the anatomy chapter, namely the "cerebral sulci," could be accomplished and was investigated by Dr. M. Ono and Dr. S. Kubik. The decision was made to publish this homogeneous part in advance as Atlas of the Cerebral Sulci (Stuttgart: Thieme, 1990) to be followed later by these volumes. The unforeseen separation of sulcal anatomy, requires therefore, frequent references to the figures in this abovementioned atlas. In autumn 1993
M. G. Ya~argil
La naissance de Palias Athénée-Minerve / Geburt der Palias Athena-Minerva, in: Michael Meier, Atlanta Fugiens, hoc est, Emblemata Nova de Secretis Naturae, Oppenheim, 1618.
The very first neurosurgeon of Greek myt~ology was Hephaestos, god of fire and surgery. The patient here is Zeus. Since giving birth was the only true mark of divinity in ancient beliefs, the first male gods to claim any sort of supremacy had to also claim the ability to give birth. Lacking any birth organ, many male gods in the different mythologies tried to give birth from various parts of their bodies. Hellenic Greeks pretended their new father Zeus gave birth to the goddess Athena from his head. The real mother of Athena was Methis. Methis means "female imagination." She was a source of the feminine art of healing and her name is related to the word "medicine." She was the mythological mother of Athena and was assimilated into the Zeus cult by the claim that Zeus impregnated her, then swallowed her after she had transformed herself into a bee, so that her "imagination principIe" became part of himself. This "female imagination principIe" must have been too heavy for Zeus' brain, for he suffered terrible headaches. Hermes called Hephaestos to perform the first neurosurgical procedure in Greek mythology, that of assisting Zeus to give birth to Athena through his head. In the background is Aphrodite with Ares (god of war). Their children are named Harmonia and the twins, Phobos (Panic) and Deimos (Fear), names which, interestingly, are related to the psychology of surgical actions.
Abbreviations Used
ACA ACTH ADH ATP AV3V AV AVM AVP BUDR CBF CNS CPP CSF CT DM DNA DSA ECF EEG EMG EPG GABA Gd-DTPA GH HSR ICA ICP ISF LH
Anterior cerebral artery Adrenocorticotropic hormone Antidiuretic hormone Adenosine triphosphate Anteroventral third ventricle Anterior ventral Arteriovenous malformation Arginine vasopressin 5-bromodeoxyuridine Cerebral blood flow Central nervous system Cerebral perfusion pressure Cerebrospinal fluid Computed tomography Double minute Deoxyribonucleic acid Digital subtraction angiography Extracellular fluid Electroencephalogram Electromyogram Electropneumogram Gamma-aminobutyric acid Gadolinium diethylenetriaminepentaacetic Growth hormone Homogeneously staining region Internal carotid artery Intracranial pressure Interstitial fluid Luteinizing hormone
acid
LHRH LTO MCA MEG MGB MR MRA MRI MTO NOR PCA PET PNET PNS PVI rCBF REM RNA SCA SON TGF TRH VA VL VMD VMN VPL VPM
Luteinizing hormone releasing hormone Lateral temporo-occipital Middle cerebral artery Magnetic encephalography Medial geniculate body Magnetic resonance Magnetic resonance angiography Magnetic resonance imaging Middle temporo-occipital Nucleolar organizer region Posterior cerebral artery Positron emission tomography Primitive neuroectodermal tumor Peripheral nervous system Peripheral vascular insufficiency Regional cerebral blood flow Rapid eye movement Ribonucleic acid Superior cerebellar artery Supraoptic nucleus Transforming growth factor Thyrotropin-releasing hormone Ventral anterior Ventrolateral Ventromedial dorsal Ventromedial nucleus Ventroposterolateral Ventroposteromedial
Tableof Contents
1
1 Anatomy TopographicAnatomy for Microsurgical Approaches loIntrinsicBrain Tumors . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Introduction.. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . HistoricalSketch of Brain Architecture .. . . . . . . . . . . . . . . . . Embryology Neuroembryogenesis Neurogenesis Myelinization . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .. Divisionsofthe Brain (Encephalon) . . . . . . . . . . . . . . . . .. . .. Cerebrum(Telencephalon) The Borders of the Telencephalon TheConceptofCerebralLobes Anatomyofthe Sulci . . . . . . . . . . . . . . . . . . . . . . . . . . . . .. Gyral Cerebral Anatomy . . . . . . . . . . . . . . . . . . . . . . . . .. TheWhite Matter of the Cerebrum . . . . . . . . . . . . . . . . .. White-Matter Sublevels and Clinical Implications Topographical Anatomy of the Lobes andGyriofthe Brain . . . . . . . . . . . . . . . . . . . . . . . . . . . . .. The Central Zone of White Matter (External and Internal Capsules) Summary InfratentorialTopographic Anatomy . . . . . . . . . . . . . . . . . . ..
2 2 3 8 8 12 13 14 14 16 16 19 20 25 39
39 65 69
81
Divisions of the CerebeIlum ... . . . . . . . . . . . . . . . . . . . .. CerebelIar Lobes and Lobules . . . . . . . . . . . . . . . . . . . . .. CerebeIlar Hemispheric Borders and Surfaces CerebeIlar Fissures and Sulci . . . . . . . . . . . . . . . . . .. The White Matter of the CerebeIlum The Predilective Location of Intrinsic CerebeIlar Tumors and Surgical Planning ... . . . . . . . . . . . . . . . . . .. Vascular Anatomy Arteries Arteries within Sulci Blood Supply of Cerebral and CerebeIlar Tumors . . . . .. Arterial Supply of Central NucIei and the Internal Capsule . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .. Arterial Supply of the Thalamus .. . . . . . . . . . . . . . . . . .. Arterial Supply ofthe Hypothalamus . . . . . . . . . . . . . . .. CerebeIlar Blood Supply Arterial Supply of the Midbrain . . . . . . . . . . . . . . . . . . . .. Arterial Supply of the Pons Arterial Supply of the MeduIla Veins . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .. The Deep White-Matter Veins of the Brain . . . . . . . . . .. Conclusions
Inlroduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .. 116
HistoricalPerspective . . . . . . . . . . . . . . . . . . . . . . . . .. TheScopeofModernNeuropathology .. .. . .. ... ... .. Epidemiologyand Pathogenesis BiologicalActivity TheNeurosurgeon's Viewpoint . . . . . . . . . . . . . . . . . . . . . . . .. GeneralConsiderations: Categorization ofCNS Tumors . . .. SpecificConsiderations Growth Pattern lntrinsicTumors
. . . . . . . . . . . . . . . . . . . . . . . . . . . . .. :
116
116 119 119 121 122 123 123
123 125
.,.. . . . . . . . .....
125
initialGrowth Pattern . . . . . . . . . . . . . . . . . . . . . . . . . . . .. Localization TumorInfiltration .. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .. PeritumoralChanges
127 129 144 146
Tumor Demarcation
146
Predilection
93 95 95 98 99 102 103 104 104 105 106 106 109 109 114
115
2 Neuropathology
Extrinsic Cranial Tumors
81 82 84 86 88
. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . ..
Brain- Tumor Interface: Adherence and Adhesiveness . . . . .. Tumor Vascularization . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .. The Numbers and Types ofTumors Location of MuIticentric Tumors Within the CNS . . . . .. Conclusions The Future The Present Cases . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .. Gyral Localization of Neocerebral Tumors . . . . . . . . . .. Tumors of the Limbic Lobe Rare Localizations Rentrolenticular Tumors Diffusely Growing Gliomas Benign Tumor Behavior . . . . . . . . . . . . . . . . . . . . . . . . . . . . . ..
147 148 149 151 151 153 153 154 154 161 171 173 174 180
MuItipleTumors DifficuIt Histology
184 188
XII
Table of Contents
193
3 Neuroradiology Historical Review Plain Radiographs . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .. Contrast Encephalography Cerebral Angiography Neuroimaging: Computed Tomography . . . . . . . . . . . . . . . . .. Neuroimaging: Magnetic Resonance Imaging . . . . . . . . . . . .. Isotope Brain Scanning . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .. Functional Neuroimaging Xenon CT . . . . . . . . . . . . .'. . . . . . . . . . . . . . . . . . . . . . . . .. Emission Computed Tomography Positron Emission Tomography Single Photon Emission Computed Tomography ... . .. Current Trends MagneticEncephalography Magnetic Resonance
. . . . . . . . . ..
Angiography
Echo-planar MRI Current Neuroimaging
194 194 194 194 194 195 196 196 196 196 196 197 197 197
with CT and MRI
. . . . . . . . . . . . . . . .. ,
MR Image Selection Contrast MRI Usefulness of MRI in Distinguishing Vascular Mass Lesions from Tumors Functional Studies . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . ., Failure ofMRIto Detail Surgical Pathophysiology . . . .. Future Correlation of MRI with Functional Imaging . .. Recent MRIImprovements . . . . . . . . . . . . . . . . . . . . . . .. The Application of Neuroimaging Capabilities to CNS Tumors . . . . . . . . , . . . , . . . . . . . . . . . . . . . , . . . . . . . . ., Lesion Morphology (Structural Changes) Lesion Location (Topography)
201 201 201 202 202 203 203
Cleavage . . . . . . . . . . . . . . . . . . . . . . . . . . . . , . . . . . . . . , .. Detailed Lesion Reconstruction Summary Conclusions
204 207 207 209
Cases
. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .. 209
197
Diagnostic Difficulties between Intrinsic and Extrinsic Tumors in the Parachiasmatic Area . . . . .. 209 Differential Diagnostic Difficulties between Meningiomas and Glioblastomas 217 Difficulties with the Origin of Intraventricular
198
Tumors
197
197
Comparisonof CT withMRI
Other Diseases Ventricular Abnormalities . . . . . . . . . . . . . . . . . . . . . . . . . . . .. Postoperative Morphological Changes . . , . . . . . . . . . . , . . . .. Functional Changes Difficulties and Limitations of Neuroimaging . . . . . .. Topographic Difficulties Peritumoral Changes
198 199 199 199 199 200 200 200 200 200
,
Difficulties with Tumors in Parapineal Areas, their Precise Location, and Tumor Type . . . . . . . . . . . . .. Difficulties with Tumor Type in the Cerebropontine Angle Difficulties with Intrinsic and Extrinsic Lesions in the Infratentorial Area Radiological Diagnostic Difficulties between Neoplastic, Infective, Degenerative, and Traumatic Disease Processes Problems Associated with Perilesional Changes .. . . , . . . . .. Edema White Matter Changes
4 Neurophysiology
Introduction
. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .. 248
Fluid System
248 250 250 251 251 251 255
. . . . . . . . . . . . . . . . .. 256
Protective Barriers
258 The Intracranial Buffering System 258 Intracranial Pressure . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .. 259 ICP Monitoring 259 Normal ICP ; 259
Elevated ICP
,
,
221 225 228
231 235 243 245
247
Physiological Subsystems . . . . . . . . . . . . . . . . . . . . . . . . . . . . .. Neural and Glial Parenchymal Systems Cerebral Cortex . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .. Symptoms and Signs of Cerebral Tumors . . . . . . . . . . . .. Localization Herniations Pathophysiology: Cerebral Edema . . .. Cerebrospinal
220
,
,
Cerebrovascular System . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .. Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .. Arterial Supply Venous Drainage Cerebral Blood Flow . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . , ..
260 262 262 263 263 263
Autoregulation Pathophysiology Thresholds Tumor Effects on the Cerebrovascular System Circumventricular Organ System The Subfornical Organ The Organum Vasculosum of the Lamina terminalis The Area Postrema
263 265 265 265 265 266 266
. . . . . . . . . . . . . . . . . . . . . . . . . .. 267
Pathophysiology Neuroendocrine System The Hypothalamus The Pituitary Gland The Pineal Gland . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .. Pathophysiology of the Pineal Gland Pathophysiology of the Neuroendocrine System as a Whole . Neurotransmitters . . . . . . . . . . . . . . . .. . . . . . . . . . . . . . . . . . .. Central Nervous and Immune Systems . . . . . . . . . . . . . . . . . .. Immune Cells . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .. Immunoregulators Pathophysiology and Future Implications
267 267 267 270
270 271 271 272 273 273 273 274
Table of Contents Pathophysiologyof CNS Disease Processes . . . . . . . . . . . . . .. 274 lnteractionsBetween Neurological Diseases andPhysiologicalSystems 274 Tumorsas Dynamic Systems 277
5 ClinicalConsiderations
Final Remarks ConcIusions Cases . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .. Effects of Chronic or Subacute Growth
- Operability
GeneralRemarks Introduction. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .. TheEraofNeuroimaging TheTraditional History and Physical Examination . . . . . . . .. TheExplosion of Neuroc1inical Technology and its ImpactonClinicalNeurosurgery TheClinicalDecision-Making Process . . . . . . . . . . . . . . . . . .. TheDecisionto Operate Symptomatology Tumor Topography - Its Impact on the Decision to Operate Recurrent Tumors . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . ..
278 278 280 301
317
318 318 318 318 318
319 320 320 321 321
Operability . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .. 321 Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .. 321 Decision-Making . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .. Perioperative Care Current Treatment Options . . . . . . . . . . . . . . . . . . . . . . .. Treatment Option ... . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . ..
XIII
321 322 323 325
Applied Microsurgery for Extrinsic and Intrinsic Tumors . .. Extrinsic Tumors . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .. Sphenopetroc1ival Meningiomas ... . . . . . . . . . . . . . . . .. Chordomas Epidermoid Optic Gliomas Hypothalamic Astrocytoma . . . . . . . . . . . . . . . . . . . . . . .. Large Insular Oligodendroglioma .. . . . . . . . . . . . . . . . .. Parietal Astrocytomas Temporo-Insular Oligodendroglioma Limbic Gliomas Intraventricular Tumors . . . . . . . . . . . . . . . . . . . . . . . . . .. Mesencephalic Tumors . . . . . . . . . . . . . . . . . . . . . . . . . . .. IV Ventricular Tumors . . . . . . . . . . . . . . . . . . . . . . . . . . .. Pontine Tumors . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .. Inoperable Tumors Final Comments
325 325 326 330 331 332 334 335 335 337 338 341
348 352 354 356 359
References
361
Index
389
I
xv
r
Ilntroduction
Hippocrates (circa 375-460 B.c.), stated "Askin peri ta nosimata dio: ofelin i mi vlaptin," which, roughly translated, means: In cases of disease there are two ways to provide: to help, or at least cause no damage. "Nil nocere" is the Latin tenant "do not harm." These age-old principIes we should all respect. Physicians are taught as students to ask: "What is wrong? How can I help?," and: "What is the likely outcome?" These questions must be engraved in our approach to each and every patient. Great advances in the natural sciences and associated techniques in the 18th and 19th centuries made possible the establishment of the field of neuroscience. Over the last one hundred years this field has developed into a large number of closeiy associated disciplines. The discipline of neurosurgery is no longer considered in isolation with regard to the treatment decisions concerning CNS neoplasms. Our successes and our failures are no longer ours to bear completely alone. By means of the accelerated deveiopments in the fields listed in Table 1, almost every recent decade has brought new diagnostic tools and therapeutic aids (Table 2).
Table 1
Neuroscience disciplines
Neuroanatomy Neuroembryology Neurophysiology Neuropathology Neuroradiology Neuronuclear medicine Neurobiochemistry Neuropharmacology Neurobioengineering
Table 2
Neurology Neuropsychology + Behavioral neurology Neuroanesthesia Neurosurgery Neurootosurgery Neuroradiotherapy Neurochemotherapy Neuroimmunology Neurogenetics
Developments in neurodiagnosis
1. Before 1900
2. 1900-1930 3. 1930-1970
4. 1970-1990
History, neurological examination, ophthalmoscope X-Ray, PEG, myelography, audiometry, examination of CSF EEG, cerebral and spinal angiography, stereoangiography, echography, radioisotope studies Ultrasound (Ooppler), CT, MRI, OSA, superselective angiography, MRA, PET, MEG, dynamic 30-MRI and MR angiography.
The pioneers of neurology and neurosurgery were forced to rely only upon careful history-taking and astute clinical observations gleaned from palpation, percussion, and auscultation, to reach a diagnosis. The first neurological diagnostic instruments
were patellar hammers, pins, and smelling bottles, followed later by ophthalmoscopes, perimetry screens, and audiovestibular tests. Even with these limited resources, reasonable estimates as to the generallocation of lesions and their pathological diagnoses could be determined by our predecessors. It was felt that a diagnosis could be reached 90% of the time following a detailed history and thorough physical examination. A further workup was felt not to significantly increase the yield (Osler). These fundamental clinical skills still remain important today. From the neuroclinical point of view, tumors within the CNS may present with one or a constellation of signs and symptoms, indicating either local tumor effects or a generalized change in intracranial pressure. These findings may suggest focal pathology or they may indicate a false localization. On the other hand, tumors may be discovered accidentally. It remains fascinating, but as yet unclear, how even very large lesions with a significant intracranial compression can be completely silent and leave the patient asymptomatic. Advances in morphological diagnosis using noninvasive instantaneous modalities such as CT scanning and MRI, have clearly revolutionized the neuroscientist's ability to precisely define CNS pathology. At the same time it is strongly advised that neuroclinical findings continue to direct and guide the "spotlight" which has been provided by advanced neuroimaging techniques and, thus, avoid the "shotgun approach" to diagnosis, namely that of over-utilization of the newest and most expensive technology. In assessing patients clinically and studying the results of a variety of investigations, we frequently come to almost instantaneous conclusions regarding the "treatability-operability" of their lesions. These critically important decisions are based on a variety of factors (see Table 3), but principally upon the topographical accessibility and the assumed nature (malignant propensity) of the tumor. It is decided whether there is great hope, an uncertain future, or little chance of survival with any meaningful quality of life. Using learned paradigms, the informed physician reaches a conclusion as to what seems to be in the best interests of the patient. A physician's assessment of an individual patient's prognosis at the time of the initial decision-making process is influenced most by the individual physician's own past experiences. When a patient's final outcome is assessed and confirms the predicted outcome, the physician's opinion regarding the correctness of his or her original treatment decisions are often reinforced (even though other treatment strategies may have resulted in a better outcome). In other words, if we avoid analyzing and looking closely at our failures, we may find ourselves judging them as successes. Despite advances in almost every sphere of neuroscientific investigation and treatment, a major deficiency still exists in the understanding of fundamental CNS tumor biology. The structural and functional relationship between tumor and normal tissue
.,.
xv (both focal and global) remains ambiguous and much work towarda comprehensive understanding of what can be expected whena particularlesion is diagnosed, needs to be undertaken. Our knowledge of fundamental brain organization and its local andgeneralizedreaction to insult is vastly incomplete. This informationis critical to the neurosurgeon's decision-making process in patients harboring CNS neoplasms. The measurement of tumoral physiological activity and other tumor-related changes haveprovided some essential diagnostic information. More comprehensive,specific, instantaneous imaging is becoming available through still other new technological advances including PET, magnetic resonance spectroscopy, and dynamic 3D-MRI. The limitsof neuroimaging technology as related to the needs of CNS tumor surgeons, however, will need to be repeatedly evaluated andmodifiedin order to finally atta in the required standard. Noninvasive and instant neuroimaging techniques have immenselyimproved diagnostic capacities, but they have not c\arified or helped to indica te fully convincing therapeutic options. On the contrary, increasingly controversial developments are observed concerning therapeutic modalities, for instance:tumor biopsy only, decompression, partial or complete removalof the tumor using various technical approaches with or without additional therapeutic options such as radiation techniques,chemo-, immuno-, thermo-, or phototherapies and, finally, no active treatment at all, apart from perhaps merely symptomaticpharmacotherapy attended by the attitude "wait and see." The situation, which is irritating for the patients and their families,for their primary-care physicians, and for the colleagues innonneurosciencefields, is not only due to the advances in neuroimaging,but mainly due to the lack of corresponding break-
ment of neurological plasticity and accommodation, and in endless variability in the outcome following pathological insults or surgery. There is little doubt, however, that the infinite variability of tumor growth rates, aggressiveness, immunological character, cytochemistry, invasiveness, and interaction with the surrounding areas of the CNS, willleave us with problems of management for many years to come. Many of these basic problems associated with tumor surgery that plagued Cushing and Dandy still exist today, for instance, our ability to preoperatively describe a tumors' location, composition, and interface with neighboring structures. It is of the utmost importance for us to understand not only the advances that have been achieved, but also the limitations that remam. In the final analysis, each Surgeon should constantly reassess his or her approach to the operability of a given case, based upon up-to-date principies and the collective worldwide experience, and it is our hope that this Book will, in some measure, help in this purpose.
Table 3
The internal computer lor decision-making
Stage
Question
1.
Nonsurgical or surgical lesion Vascular, neoplastic, inlection, or autoimmune lesion Type 01tumor - Classilication - Grading Supra- or inlratentorial: extrinsic or intrinsic or mixed
2. 3.
Preliminary diagnosis . Preliminary differential diagnosis Special differential diagnosis
throughs in biology.
As there is still no curative therapy available, each therapeuticendeavorremains, in the final analysis, a palliative action, as in thecase of anaplastic glioma as well as in some cases of extrinsic tumors such as adenomas, craniopharyngiomas, meningiomas, andneurinomas. Despite significant deficiencies in our neuropathological and neurophysiologicalunderstanding of CNS tumors, a comprehensiveapproach primarily utilizing current neuroanatomical and neuroradiologicalknowledge should be formulated. New paradigmsto better determine the operability of individual patients' tumorsshould be constructed. Two of the great nineteenth-centuryneurophysiologists and neurologists,notably John Hughlings Jacksonand David Ferrier, showed that the brain was not a single unifiedorgan but contained within it complex functionallocalizations.Thistheory is all too often forgotten. The brain is a conglomerate of many organs linked by a vast and complex network of communicationsystems. Thiscomplex organization within the brain allows for great paradoxesin clinical presentation of CNS tumors and in surgical In additionit leadsto enormous diversityin the developoutcome.
.
4. 5.
6.
General topographic diagnosis Special topographic diagnosis a) More precise localization b) Modes 01expansion (circumscribed or diffuse) Prognostic inlormation a) Vital
b) Functional
c)
7.
Possible reaction to therapeutic modalities
Final decision
Epidural, dural, subdural, subarachnoid, or mixed, or neopallial, transitional, central nuclei, intraventricular
} Natural history 01tumors Blood, CSF tests (enzyme) Possible reaction to the therapeutic modalities (surgery, radio- and chemotherapy) Potential lunctional delicits (eloquent or noneloquent areas) Radiation, chemotherapy
Which treatment, alone or multimodality therapy?
1
Max Bill Surface of a Spíral @VG Bild-Kunst,Bonn 1993
Anatomy
2
1 Anatomy
Topographic Anatomy for Microsurgical Approaches to Intrinsic Brain Tumors
Introduction
I
The surgical anatomy of the cerebral and cerebellar hemispheres and its relation to the surgery of tumors will be discussed in this chapter, and in particular, the new concepts developed (in an evolving patchwork fashion) by the senior author over the last 26 years in the operative treatment of over 3400 tumors of the central nervous system. Some of the special surgical anatomy in relation to the extrinsic group of tumors will be presented in Volume IVB. The anatomy of the central nervous system (including special considerations, such as the embryological, parenchymal, vascular,
r
functional, topographical, neuroradiological, and stereotactic aspects) with both macrosurgical and microsurgical perspectives, is extensively described in numerous textbooks and publications.
l
I
Furthermore, excellently illustrated atlases, beautifully stained gross cross-sectional references with stereoscopic slides, and idealized computer models are readily available today to further aid our understanding of the fascinating morphology of the CNS (Key and Retzius 1875, Retzius 1896, von Economo and Koskinas 1925,Elze 1932, Rose 1935, Villiger and Ludwig 1946,1951, Mettler 1948, Basset 1952, Krieg 1953, Ludwig and Klingler 1956, Clara 1959,Delmas and Pertuiset 1959, Schaltenbrand and Bailey 1959,Wolf-Heidegger 1962, ElIiott 1969, Kahle 1969, Ferner 1970, Rohen 1971, Lang 1973-1988, Waddington 1974, Stephan 1975, Szikla et al. 1977, Kieffer and Heitzmann 1979, Dejerine 1980, Pernkopf 1980, Angevine et al. 1961, Brodal 1981, Smith 1981, Tondury 1981, Koritké and Sick 1982, Creutzfeld 1983, Heines 1983, McGrath 1984, Benninghoff 1984, Fitzgerald 1985, Gray 1985, Clemente 1985, Watson 1985, Formann and Heym 1985, Jones and Peters 1986, Maillot 1986, Wilkinson 1986, Fi~ 1987, 1992, Leonhard et al. (eds.) 1987, Zenker (ed.): 1985, Barr and Kiernan 1988, de Groot and Chusid 1988, Niewenhuys et al. 1988. Ferner, Staubesand (eds.): 1973, 1975, Talairach et al. 1988, Armon de et al. 1989, Martin 1989, Williams and Warwick 1989, Lasjaunias and Berenstein 1990, Lippert and Seiderer 1990, Von Hagens et al. 1990, Carpenter 1985-1991, England and Wakely 1991). (See also Ya~argil, 1984, vol. 1. pp. 5-53-subarachnoid cisterns; pp. 54-168-intracranial arteries; vol. III A, 1987, pp. 284-319-sulci and fissures; pp. 320-6-microcirculation; pp. 327-32-venous system; pp. 338-49-cortical blood vessels; pp. 350-68-calcarine sulcus; and vol. III B, p.7, pp. 205-10, pp. 287-90.) The work ofBasset 1952), Huang (1964-1985), Stephens and Stilwell (1969), Duvernoy (1969-1988), Waddington (1974), Newton and Potts (1974), Salamon (1973), Williams and Warwick (1975), Lang (1973-1992), Seeger (1978, 1980, 1984), and Marinkovic (1985-1992) has given us, tI? a degree, the necessary topographical details we require. The elegant series of studies by Rhoton and his associates (1976-1992) describes the precise microsurgical details of various brain regions, with their corresponding vasculature, from the neurosurgeon's point of view. Nevertheless certain areas of the brain have not received the
same precise and neurosurgically relevant anatomical analysis. Important aspects of the brain, such as the surgical anatomy of the gyri and white matter, and associated variations in the leptomeningeal and cortical vasculature have only partially been elucidated (Salamon et al. 1972-74, Waddington 1974). Experts in stereotactic techniques have particularly contributed to the development of precise atlases of areas deep within the brain (Szikla et al. 1977). Similarly, computer technology has proved of immense help in displaying three-planar maps of these structures (Matsui and Hirano 1978, Heiss et al. 1985, William and Haughton 1985, Daniels et al. 1987, Gouaje and Salamon 1988, Talairach and Tournoux 1988, Courchesne et al. 1989, Hirsch et al. 1989, Press et al. 1989, Aichner et al. 1989, Schnitzlein and Murtag 1990, Steinmetz and Huang 1990, Toga 1990, Kretschmann and Weinreich 1991, Dietemann 1993). Three-dimensional magnetic resonance images will undoubtedly become available in full color and in film sequences (Levin et al. 1989, Hu X et al. 1990). Positron emission tomography (PET) and magnetic electroencephalography (MEG) are providing new insights into the localization and regional integration of cerebral functions in both normal and pathological brains and demonstrate a spectrum of variation not previously imagined. The impact of the function (local and global) of similar pathological lesions (often widely disparate) can now be studied in evolution. Developments in imaging techniques have done much to advance the convergence of (sometimes conflicting) morphological and functional brain maps. The more detailed information provided by neuroimaging techniques, the results of neurophysiological, neuropathological, and neuroimmunological investigations, and extensive surgical anatomical experience, have demonstrated that the CNS is a unique morphofunctional unit consisting of an integrated, dynamic network of several subsystems. Precise topographic and functionally defined diagnoses ofbrain lesions and adequate treatment decisions require from the neurosurgeon (as well as from interventional neuroradiologists, radiotherapists, and chemotherapists) more sophisticated knowledge of the anatomy and physiology of the CNS. Following the historical sketch of brain architecture and a short description of the neuroembryogenesis and neurogenesis, the surgical anatomy of the cerebral and cerebellar hemispheres, including the borderlines, sulci, gyri, and white-matter segments, and their arterial and venous patterns will be presented.
Topographic Anatomy for Microsurgical Approaches to Intrinsic Brain Tumors
3
Historical Sketch of Brain Architecture The160-yearhistory of research on the structure of the brain cortexhasbeenwell described by several authors (Soury 1899, Scarff 1940,Rasmussen1947, Lorente de Nó 1949, Brazier 1959, 1963, 1978, Riese and Hoff 1951, Haymaker and Baer 1953, Kolle 1954-1963,Penfield and Rasmussen 1957, Walker 1957, Magoun 1958,Zangwill 1963, Klingler 1967, McHenry 1969, Haymaker and Schiller 1970, Meyer 1971, Clarke and Dewhurst 1972, Stephan1975,Creutzfeldt 1983). The history of the discovery of cerebralgyri in particular, is eloquently presented in the monographby Clarke and Dewhurst (1972). According to Clarke and Dewhurst,"This story is of particular fascination and relevance because it not only reveals a sequence of intriguing notions, but also contributes to our understanding and appreciation of the modern view." As the anatomy of gyri is an important part of this chapter, some passages from this excellently illustrated monographwillbe presented here. "Erasistratus of Alexandria (260 B. c.) stated that the gyri werecomparable to the coils of the small intestine, but interestingly he contended that in animals the gyral complexity was directlyproportional to intelligence." The earliest pictorial representation of the convolutions is in an eleventh century manuscript, the earliest known Western illustrationof brain function (Fig. 1.1a). It is in the shape of a Celtic stone cross, sometimes found in Anglo-Saxon diagrams. Around the circ1e is written, "There are present four principal human members," which are, in clockwise sequence from 12 o'cIock, the liver, heart, testes, and brain ("cerebrum"). The latter is,in fact, a drawing of the skull facing inwards and seen from above,with the coronal, sagittal, and lambdoid sutures representedby double lines. The mental faculties inscribed on it are "fantasia" (imagination), "intellectus" (reasoning), and "memoria" (memory) (Fig. 1.1a). Figure 1.1b, showing the eyes and brain, "seems to have originatedin a treatise on ophthalmology written in the second half of the thirteenth century by a Syrian, Halifa, but actually it may date backfurther (to at least, A. D. 1000). It is, therefore, probably the oldestfigure of its kind, though the present manuscript is dated to 1560.The drawing has be en reported on many occasions but is only fully described by K. Sudhoff (1914) and 1. Hirschberg (1905)." "Guido da Vigevano (ca. 1280-ca. 1349) wrote a discourse onanatomy,illustrated by a series of interesting drawings (1345). Thereis vague patterning on the surface of the exposed brain whichmay possibly be an attempt to draw the cerebral convolutions.If so, it is the earliest portrayal of them" (Fig. 1.1 e). "As with most other organs and systems, the best brain illustrations[of the Renaissance] appeared in the De Fabrica of Vesalius (1543).Vesalius was more concerned to identify structuresother than the convolutions. He agreed, however, that they appearedlike coils of small gut or like clouds drawn by schoolboys,having no standard pattern." "Rayrnond de Vieussens of MontpeIlier (1644-1716), whose nameis still occasionally associated with the 'centrum.ovale,' publishedin 1685his Neurographia Universalis, wherein the cerebral convolutions and cortex received some attention." "Samuel Thomas Soernrnering (1755-1830) introduced our presentcIassification of the cranial nerves. The gyri are well depictedin a drawing, but they are not yet named."
.
b Thisschematic depiction of the eyes and brain, dating fromthe late thirteenth century, is probably the oldest of its kind
-
4
1 Anatomy
,."
,
Fig.1.1c-d e The history 01the anatomy 01the cerebral convolutions by Guido de Vigevano (ea. 1280-1349). There is a vague patterning on the surface 01the exposed brain, which may be an attempt to show the cerebral convolutions. It so, it is the earlist portrayal 01 them
"Felix Vicq d' Azyr (1748-1794), whose name is still associated with the mammillothalamic tract, began to differentiate gyri by grouping them into anterior, middle, posterior, and inferior and by describing the pre- and post-central convolutions. Vicq d'Azyr also described some of the lobes' constituent convolutions: the 'convolution that follows the corpus callosum.' He introduced the name 'uncus,' which means crochet (hook). "The fissure of Sylvius was the first su1custo be identified and named. The island of Reil or insula, c1early depicted by Bartholin (1641), was described in detail by Johann Christian Reil (1759-1813) in 1809. They were the first two landmarks, to which was added the fissure of L. Rolando (1931)." Rolando, who performed the first studies of the effects of electrical current on the brain of animals, named the precentral and postcentral gyri "processi verticali di mezlO." "Louis Pierre Gratiolet (1815-1865) by comparative studies distinguished primary from secondary gyri, in accordance with their chronological appearance in evolutionary sequence. Gratiolet used the terms 'frontal,' 'temporal,' 'parietal,' and 'occipital' lobes (introduced by F. Arnold in 1838), and defined their limits. He first published these in 1854, and Fig. LId shows the high standard that pictorial representation of the gyri had reached. The legend gives the convolutions names, several of which are still used. Arnold's publication has only one picture showing the brain from a basal view; the gyri are not well delineated. Arnold himself gives credit to Chaussier, but does not give any references." C. Bell (1774-1842) established the fact that the nerves of special senses could be traced from specific areas of the brain to their end organs (Bell's palsy). F. Burdach (1819) gave the names' to the gyrus cinguli, precuneus, and cuneus (tweak, pinch). J. G. F. Baillarger (1840) discovered the connection between the white and gray matter. He introduced the terms lissencephalic and gyrencephalic cortex.
d Louis-Pierre Gratiolet (1815-1865) is the most renowned 01 the French anatomists who investigated the cerebral gyri and sulci. Gratiolet used the terms "Irontal," "temporal," "parietal," and "occipital," which had been introduced by Arnold in 1838 and delined the boundaries 01these areas. This picture illustrates the beginning 01 a new conceptual era
R. Remark (1844) was the first to recognize histologically the six cortical celllayers. R. A. von Kolliker (1845) anticipated the neuron theory. Thomas H. Huxley (1861) gave the name to the superior and inferior frontal su1ci and ca1carine fissures. "The first lithograph photograph of the brain was published by Emil Huschke (1797-1858) of Jena. It appeared in 1854 and was the first step towards direct photography of the brain. Huschke contributed to gyral morphology by naming the gyrus centralis anterior and posterior and the 'fusiform' and 'lingual' convolutions." William Turner (1832-1916) of Edinburgh redefined the limits of the cerebrallobes and established the fissure of Rolando as the posterior limit of the frontallobe}n 1866. Alexander Ecker (Freiburg, Germany) in 1869 described in a small book (53 pages) all the su1ciand gyri of the telencephalon in detail, which can be called complete and valid even 124 years later (See Fig. 1.1e-g). He gave their names to the orbital, precentral, parieto-occipital, interparietal, and transverse occipital su1ci. He drew attention to the fact that a gyrus consists of three parts: the primary gyrus in a convoluted chain; the secondary gyri, separated by secondary su1ci; and the tertiary gyri, intrasu1cal extensions that cannot be seen on the surface without spreading the su1ci,for example the temporal transverse gyri of Heschl. The tertiary gyri parts of which are well visualized on MR images, should be given a gene rally accepted anatomical name-tertiary or transverse gyri. G. Retzius (1896) suggested for the secondary gyri the term "gyruli," and for the tertiary gyri the term "transitivi" or "profundi."
Topographic Anatomy for Microsurgical Approaches to Intrinsic Brain Tumors
5
f
Fig. 1.1e-g Depictions of the brain (from Ecker, Die Hirnwindungen des Menschen [1869], 2nd ed. Braunschweig 1883) e Superiorview ip Interparietal sulcus A Anteriorcentral gyrus B Posteriorcentral gyrus O Occipitallobe C Central sulcus o Transverseoccipital gyrus cm Callosomarginal sulcus F Frontallobe F1 Superior frontal gyrus f1 Superior precentral sulcus f2 Inferior precentral sulcus F2 Middle frontal gyrus F3 Inferior frontal gyrus f3 Precentral sulcus
01 P P1 P2
First occipital gyrus Parietal lobe Superior parietallobule Inferior parietallobule (P2 = supramarginal gyrus, P2' = angular gyrus) po Parieto-occipital fissure t1 Superior temporal sulcus
9 Lateralview.The gyri are precisely named cm Callosomarginal sulcus S Sylvian fissure 11 Superiorfrontalsulcus S' Horizontal branch f2 Inferiorfrontalsulcus S" Ascending branch f3 Precentralsulcus T1 Superior temporal gyrus ip Interparietalsulcus t1 First temporal sulcus 01 Superiororbital gyrus T2 Middle temporal gyrus 02 Middleorbital gyrus t2 Second temporal sulcus 03 Inferiororbitalgyrus T3 Inferior temporal gyrus
f Basal view C Callosum CC Corpora candicantia f4 Olfactory sulcus f5 Orbital sulcus H Hippocampus K Cerebral peduncle
oc po t2 t3 t4 U
Calcarine fissure Parieto-occipital fissure Middle temporal sulcus Inferior temporal sulcus Inferior occipitotemporal sulcus Uncus
6
1 Anatomy
Fig. 1.1 h The left cerebral hemisphere (Irom Retzius, Das Menschenhirn; illustration by Sigrid Anderson, Stockholm: Nordstedt, 1896). "An organ as important as the human brain deserves to be precisely illustrated in all its parts and variations" (Retzius, p. iv)
Fig. 1.1 i Medial view 01the right cerebral hemisphere (Irom Retzius, Das Menschenhirn, Stockholm: Nordstedt, 1896); illustration by Sigrid Andersonj
The study of the macroscopic anatomy of the CNS has developed on a large scale; E. Huschke (1854), T. L. Bischoff (1868), A. Richter (1887), W. His (1889), F. Marand (1891), N.1. Cunningham (1897), F. Hochstetter (1894), and G. Retzius, who in 1896,published two volumes ofbrain anatomy incIuding an unprecendented 96 plates (Fig. 1.1h, i). Once the gyri and suIci had been accurately delineated and named, the stage was set during the 1860s for the emergence of new concepts of functionallocalization. T. Meynert (1833-1892), teacher of C. von Economo (1876-1931) introduced the terms "association" and "projection" systems, which was a great step forwardin our understandingof the cortex.Meynert thus opened the
field for exploring the relationship between the cortex and other structures, such as the basal ganglia, a term that had been introduced by F. 1. Gall (cited by Brazier 1978). C. Golgi discovered, using the silver stain technique, previously invisible nerve structures and connections (1883). P. E. FIechsig, who invented the concept of myelogenesis and discovered auditive radiation, developed (1893) the theory of projection and association centers. W. Waldeyer introduced the term "neuron" in 1891, before W. His and A. Forel independentIy formulated (1887) the concept of the cellular functional unit, which was supported a few years later by Ramon y Cajal (1890).
¡ ,! !~ ~
1
Topographic Anatomy tor Microsurgical Approaches to Intrinsic Brain Tumors W.Campbell (1905), K. Brodmann (1908, 1909), O. and C. Vogt(1919,1926),C. von Economo and G. N. Koskinas (1925) publishedrenowned myelo- and cytoarchitectonic maps. Donaldson(1895)pointed out that one-third of the neocortex is on the surface,whiletwo-thirds are in deep sulcal are as (cited in Creutzfeldt1983).Campbell pointed out the important deficiency of all such"maps,"the artist's inability to show parts of the cortex concealed in the depths of sulci.
G. Elliot Smith (1919) gave meaning to the phylogenic inthesurfacetopographyof the brain through his studies changes ofthesubcorticaland cortical factors that influence the elaborationandmodificationof the convolutions of the pallium (Walker 1957). O.and C.Vogt introduced the terms isocortex and allocortex (1912).C. Sherrington (1937) introduced the term "synapse"instead of "Syndesm."
Thediscoveryof the alpha rhythm in the human brain by H. Berger(1924-1931),using an electroencephalograph, opened an entirelyneweraforthe studyof brain function,whichthen culminatedin the intracellular registration of electrical activities (Adrian1936,Adrian and Moruzzi 1939, Jung et al. 1953, Eccles 1953,Philips 1956, Jaspers 1960). Bard (1928) and Hess (1948) pioneeredepoch-making research on the hypothalamus. The intensive research work carried out during surgical explorationof human brains resulted in the creation of the term "sensoryand motor homunculus" by W. Penfield, E. Boldrey and R. Rasmussen (1937-1968). G. Moruzzi (1954) and H. W. Magoun(1958)showed that the reticular formation maintains the tone of cerebral neurons.
7
E. W. Demsey and R. S. Morrison (1943) investigated the thalamocorticothalamic circuit. H. Gastaut (1961) pointed out that the thalamic nuclei are responsible for maintaining the direction of attention. The functional role of the limbic system in relation to the corticallocalization was investigated by Papez (1937) and McLean (1949-1953). Various techniques have been used to study the connectivity of functional systems of the brain, such as the retrograde changes after axonal lesions (Walker 1957), silver-staining degenerated axons and terminals (Glees, Nauta, and Kuypers (1938-1954). Combined studies using the Golgi technique, degeneration techniques, and electron-microscopic investigations, led to a renewal offunctional microanatomical concepts (Sholl1956, Szentágothai 1970). . Hubel and Wiesel (1977) presented maps of functional anatomy, and Eccles (1969), who succeeded in 1953 in identifying the intracellular spinal motoneuron, coined the terms "excitatory" and "inhibitory" neurotransmitters. New fields of research, on the chemoarchitecture and immune network of the CNS, are very promising (for detailed information, see H6cfeld et al. 1985, Niewenhuys 1985, Coon 1958 and Roitt, 1993). It would be beyond the scope of this book to mention all the neuroscientists who have contributed essential insights to build up our present state of knowledge of the anatomy and physiology of the CNS. Fascinating discoveries on brain architecture are incessantly continuing.
8
1 Anatomy
Embryology
Neuroembryogenesis At approximately 16-17 days of age (15 mm), the ectoderm along the dorsal aspect of the midline thickens to form the neural plate (Fig.1.2). A primitive streak then develops (Henson's node), forming a neural groove and neural folds on each side. The anterior portion of the neural plate enlarges to form the brain plate, which eventually becomes the brain (for figures depicting the development of the nervous ~ystem, see Hamilton et al. 1962, Kahle 1969, Starck 1955, Rakic 1972, Rager 1985, Carpenter 1983, Tondury and Kubik 1987, Tondury and ZilIes 1987, Hinrichsen 1990). After approximately one month, the brain tube further differentiates and enlarges into the three primary brain vesicles known as: the (1) prosencephalon or forebrain, the (2) mesencephalon or midbrain, and the (3) rhombencephalon or hindbrain (Table 1.2a). The brain continues to develop over the next several weeks at a greater rate than the overall growth rate of the neural tube. This development of the "contorted" brain involves several complex, integrated processes: 1)inward bending of the brain vesicles to form recognizable flexures, 2) differential deyelopment of several areas, 3) an "over"-growth of the cerebral hemispheres, consuming deeper parts, and finally, 4) the formation of gyri and sulci within the hemispheres.
2
9
At approximately two months' gestational age, the brain has differentiated into five recognizable parts. These are the telencephalon, diencephalon, mesencephalon, metencephalon, and myelencephalon. The telencephalon enlarges to become the cerebral hemispheres, which encircle the underlying diencephalon. The hippocampal formation occupies an intermediate zone, and enlarges with the telencephalon. The cerebellum develops from a bilateral expansion of the metencephalic alar plate (rhombic lips), and comes to occupy most of the posterior fossa, overlying the posterior aspects of the pons and the medulla. The telencephalon is phylogenetically the most recently developed portion of the nervous system. Its many convolutions and infoldings point to its general anatomic complexity. The overall form of the brain is recognizable at three months. It is the continual growth and differentiation of the cerebral hemispheres thereafter that leads to the various lobe, sulci, and gyral patterns (Figs. 1.3-1.7). The accompanying figures in this section are to help formulate within our memory the embryological development of the gyral convolutions, for a better understanding of the predilection sites of CNS tumors.
í
3
,
'., :..' .." ':'.,~"
1
6
".,
4 ..K.,.'.Ú\..::::,:".,., ::.::..,...
7
~~
3
13
5
10 11
.' "'/"'''',.:.,'",
4
12
. Fig.1.2 The humanembryo(7 somitestage) (fromLeonhardt,Tóndury and Zilles, Rauber/Kopsch: Anatomie des Menschen, Stuttgart: Thieme,1987, vol. 3, p. 12, Fig.2.1) 1 Ectodermfromthe neuroplate 8 Extraembryonalcoelom 2 Amnioticcavity 9 Chorda dorsalis 3 Primitiveknot 10 Intraembryonicmesenchyme 4 Oropharyngealmembrane 11 Entoderm 5 Dottersack 12 Cloacal membrane 6 Allantois 13 Extraembryonicmesenchyme 7 Heart and pericaval cavity
I
O,5mm
Fig. 1.3 Fetus, 5.3 cm long (Hochstetter, Zagreb; from Hinrichsen, Humanembryologie, Berlin: Springer, 1990, p. 391) 1 Sylvian fissure 4 Olfactory bulb 2 Hypothalamus 5 Cerebellum 3 Stalk of eye 6 Mesencephalon
, Topographic Anatomy for Microsurgical Approaches to Intrinsic Brain Tumors
Fig.1.4 The development 01 the rhinencephalon (green) (alter Elliot andPenlield, Textbook of the Nervous System, Philadelphia: Lippincott, 1954). Elliotand Penlield state, "Primitively the rhinencephalon (or small brain) occupies all 01 the hemisphere. Nonollactory elements are depicted as a small white area on the lateral surface. With expansion 01 theseareas, however, the rhinencephalon proper comes to occupy only a relativelysmall collar around the neck 01the hemisphere"
... TI::JI(..Jl~ ,Ir"
'lo\.
9
a
b Fig. 1.5 b-c Lateral view 01 the brain 01 ahuman embryo (Irom Ono, Kubik, and Abernathey, Atlas of the Cerebral Sulci, Stuttgart: Thieme, 1990, p. 7, Fig. 1.3) b 19cm e 24 cm. The developing Sylvian lissure is quite conspicuous (arrow)
2 3
Fig.1.5a The brain 01 ahuman embryo (24 cm), seen lrom above (from Leonhardt, T6ndury and Zilles, Rauber/Kopsch: Menschen, Stuttgart: Thieme, 1 Interhemispheric sulcus 2 Central sulcus 3 Postcentral sulcus 4 Parieto-occipital
sulcus
Anatomie des
1987, vol. 3, p. 139, Fig. 7.23)
10
1 Anatomy
4
5
'.
,
13 12 11
th
"
;
hy
I
,
j
c'
\
9 8 a
\. \\ .
'
,
14 15 -.
-
d
18
,
-
-
.
19 20
16
-
17
21 22
,.'-
"075
--,1 ....
-
.- .
6 9 8
b Fig.1.6a-b
Medialview of the brain of ahuman embryo (fromLeonhardt, Tóndury and Zilles, Rauber/Kopsch: Anatomie des Menschen, Stuttgart:Thieme, 1987, vol. 3, p. 139, Fig. 7.22). a 19 cm, b 24 cm 1 Frontalpole 13 Anterior commissure 2 Occipital pole 14 Callosal sulcus 3 Callosalbody 15 Cingular sulcus. 4 Choroidaltela of the third ventricle 16 Lamina terminalis 5 Quadrigeminalplate 17 Optic nerve 6 Cerebellarvermis 18 Subparietal sulcus 7 Fourthventricle 19 Parieto-occipital sulcus 8 Medulla oblongata 20 Splenium 9 Pons 21 Calcarine sulcus 10 Infundibulum 22 Lunate sulcus 11 Optic chiasm th Thalamus 12 Laminaterminalis hy Hypothalamus
Fig.1.61 A case
of agenesis of Corpus Callosum, showing associated maldevelopment of the Cingulate gyrus, in the brain of a newborn.
Fig. 1.6c-e The development of the commissural system, the fornix, and the septum pellucidum (blue) (from Hinrichsen, Humanembryologie, Berlin: Springer, 1990, p. 393) c Fetus, 10.5 cm d Fetus, 12.5 cm e Adult
Topographic Anatomy tor Microsurgical Approaches to Intrinsic Brain Tumors
.
'. ...
Fig. 1.7
Male fetus, 48 cm (from Retzius, Das Menschenhirn, Stockholm: Nordstedt. 1896; illustration by Sigrid Anderson)
11
12
1 Anatomy
Neurogenesis
The arrival of a group of similar neurons at their final location (selective cell aggregation) triggers a complex series of
Both neuronal and glial cells are generated from precursor cells (germinal cells) located in elose proximity to the cerebral and cerebellar ventrieles. Each distinct type of neuron and its supporting glial cells are formed during a narrow period of time (a few days to two weeks) during neurogenesis. It appears that, immediately after proliferation, each neuron undergoes topographical programming as to its eventual functional status. At this point, neuronal (and associated glial) migration occurs, with ameboid-like movement of the nerve cells along the processus of special astrocytes until the final "adult" location is reached. The most complex (neocortical-associative) neuroÍ1s are produced last, as archiogenesis is repeated (Rakic 1972) (Fig. 1.8, Table 1.1). The direction of this migration (as directed by these special astrocytes) from the ventrical matrix to cortical regions may explain the growth patterns of intrinsic gliomas (from subcortical areas of origin back towards the ventriele). It may be that neurogenesis is reversed in this manner.
events, during which the functional activity of the neuron is developed. Steps in this process inelude: 1) Neuronal cytodifferentiation (the development ofaxons and dendrites, the synthesis of ion channels, and the beginning of transmitter function). 2) Axonaloutgrowth (the growth of functional similar axons along defined paths to reach intended target fields. 3) Synaptogenesis (the refinement of connections, with the dying out of perhaps 50% of the neurons and the elimination of many synapses). This process of synapse development and then elimination permits the development of definitivecircuitry within the CNS, and forms the basis for developmental and neuromorphologic plasticity.
B
~
RF
.
~
¿)}.~': .~:.. ',;~~~.: '.:: RF
RF'
2500
:Ji gap
~
RF
a Fig.1.8a, b Neurogenesis and migration. Young neurons gene rally migrate outward from the ventricular zone along the surfaces of a specializedclass of glialcells. The illustrationon the leftshows the dispositionof the radial glial in a section through the developing cerebral hemisphere,and the insets on the right show how they span the entire
b thickness of the wall of the hemisphere and provide a substrate for the migrating neurons as they move out to form the cortical plate. It is noteworthy that this migration pattern is similar to the pattern of spread for most white-matter gliomas, except that these spread in the reverse direction (Rakic 1972, Cowan 1979)
Topographic Anatomy for Microsurgical Approaches Table1.1 Histogenesis 01the neocortex (alter Kostovic 1990, p. 410)
-
HistoEmbryonal Earlyletal period, genetic period, 30-88 mm processesstage 15-23 weeks
Middle letal period, 9-23.8 cm
Lateletal Newborn period and early newborn, 25-38 weeks, 24-45 cm, 5003800 9
Proliferation
++
++
++
+-
-
Migration
+ -
+++ +
+++ +++
+ +++
+++
+
+++
+++
++
-
-+
++
++
++
-
+ +?
++ +?
+++ ? ?
+++ ? ?
Dilterenti-
ation Growth 01 alterent fibers Growth 01 elterent libers Synapse Celldeath Plasticity
-
Fig.1.9a, b The myelinization 01 intracortical libers in theneocortex and the myelogenetic organization 01the humancartex (alter Flechsig 1927). Fields 1-6: prenatal myelinizedareas; lields 17-45: postnatal myelinized areas(between the lirst and lourth postnatal months). Thenumber lar each lield corresponds to the time sequence01the myelinization process; myelinization alsocontinues alter lour months (alter Creutzleldt, CortexCerebri,Berlin: Springer, 1983, Fig. 2.11, p. 22)
to Intrinsic Brain Tumors
13
Myelinization The myelinization of central fibers and cranial nerves follows a distinct embryologic timetable, related to the activities of oligodendroglia. The fibers of most of the cranial nerves begin to acquire myelin sheets around the fourteenth week of intrauterine life. This process is delayed until the twenty-second week for the sensory part of the trigeminal nerve and the cochlear division of the vestibular-cochlear nerve. The optic nerve does not undergo myelinization until the late stages of gestation (Keene and Hewel 1931). A description of fully myelinated, partially myelinated, and unmyelinated fibers is beyond the scope of this book; for further information, see Kostavic and Rakic (1972). The myelinization patterns of intracortical fibers were comprehensively investigated by Flechsig (1927) (Fig. 1.9).
14
1 Anatomy
Divisions of the Brain (Encephalon) Although in essence a continuum, the nervous system may be divided, for convenience of study, into a number of parts, regions, or subsystems (Fig. 1.10, Table 1.2a-b). The encephalon is described as consisting of a number of regions, which are of considerable morphological and functional significance.These are-in ascending order from the spinal cordthe rhombencephalon or hindbrain, the mesencephalon or midbrain, and the prosencephalon or forebrain. The rhombencephalon includes the myelencephalon or medulla oblonga te, the metencephalon or pons and cerebellum. The prosencephalon is also subdivided into the diencephalon ("between brain"), which is the central connecting part of the forebrain (corresponding approximately to the thalamus and hypothalamus), and the telencephalon, which consists of the two so-called cerebral "hemispheres" (or cerebrum). The midbrain, pons, and medulla oblongata are collectively termed the brainstem, connecting the forebrain and spinal cord. The anatomy of the brainstem is well covered in existing texts. In order to enhance surgical localization and planning, we have applied the following concept for the cerebral and cerebellar hemispheres. The cerebrum and cerebellum are divided into the following parts: Cerebrum Neocerebral areas Archi- and paleo-cerebral areas Basal ganglia Intraventricular areas Cerebellum Neocerebellar areas Archicerebellar are as Paleocerebellar areas Central nuclei Intraventricular are as Though this may also initially seem arbitrary, it has embryological, evolutionary, physiological, and surgical merit (see Tables 1.3b, 1.3e on p. 18 and Table 1.8on p. 81).
2
3 4
5
7
6
Fig. 1.10 Divisions 01 the brain, arbitrarily distinguished lor teaching purposes (Irom Leonhardt, Tóndury and Zilles, Rauber/Kopsch: Anatomíe des Menschen, Stuttgart: Thieme, 1987, vol. 3, p. 44, Fig. 3.1) Blue: Spinal cord Violet/red/orange: Rhombencephalon Beige/yellow: Prosencephalon 5: Medulla Oblongata 1: Telencephalon 6: Cerebellum 2: Diencephalon 3: Mesencephalon 7: Spinal cord 4: Pons
Table 1.2 a Divisions of the brain Telencephalon Endbrain Prosencephalon Forebrain
Encephalon Brain
[ Diencephalon Interbrain
Mesencephalon Midbrain Metencephalon Crossbrain Rhombencephalon Hindbrain
[ Myelencephalon Medulla oblongata
Cerebrum (Telencephalon) "The cerebrum is described as consisting of two large convoluted cerebral hemispheres." In fact, the halves of the cerebrum are not hemispheral; together, they do form roughly one hemisphere. Perhaps the term was originally used in the singular, but the plural usage (however unsuitable as regards shape) is customary. Each hemisphere is roughly equal to a quarter of a sphere in shape, and contains a large, crescent-shaped lateral ventricle, continuous medially with the third ventricle in the diencephalon (William and Warwick 1975). Each hemisphere has an external layer of gray matter, the cerebral cortex, and a central core of white matter, the medullary substance, in which there are severallarge masses of gray matter, the basal ganglia or nuclei. The anatomy of the cerebral hemisphere has traditionally been difficult to analyze, as it requires an understanding of the underlying gyral and sulcal anatomy. The use of an "inside-out"
anatomical concept has simplified this task, and has proved to be of great practical surgical benefit in tumor management. Our own anatomical studies and surgical experience are synthesized here with information and knowledge drawn from the literature, in the hope of offering young neurosurgeons pragmatic ideas.
~
Topographic Anatomy for Microsurgical Approaches to Intrinsic Brain Tumors
15
Table1.2b Subdivisions of the brain
Pallium
Frontal Central Parietal Occipital Temporal
Neocortical cerebrum
}-
lobes
White matter
=]
Endbrain Telencephalon
Lateral ventricles
Insula (partially mesocortical) Claustrum (!) Allocortical and mesocortical cerebrum Caudatum
Forebrain Prosencephalon
Putamen
Basal ganglia
Third ventricle Interbrain DiencePhalo~
Brain Encephalon
Central nuclei
-{
Colliculi Brachii Aqueduct Reticular substance Nucleus of III-IV nerves Ruber nucleus, substantia nigra Lemnisci
Base Crus cerebri
corticopontine Corticobulbar (nuclei) Corticospinal
Medulla oblongata
Tegmentum Pyramis Oliva
~
Pons Fourth Ventricle --{
{
Cerebral peduncles
}-
t t
Tegmentum Base
,
Brainstem
Reticularis Nuclei V, VI, VII, VIII Fibers Reticular substance Nuclei Fibers IX, X, XI, XII
Paleocerebellum -
Neocerebellum Crossbrain
Corpus striatum
-r
Tegmentum
Medulla Myelencephalon
}
Neostriatum
===:J---
Pallidum Lentiform ncl. (paleostriatum) Amygdala (partially cortical) Thalamus Metathalamus Epithalamus Hypothalamus Subthalamus
Midbrain f" Tectum Mesencephalon
Hindbrain Rhombencephalon
===:J---
Limbic lobe
]
. Palhum
.
,
I
i
Archicerebellum -
Metencephalon
Cerebellum ---J
AOL, Centralis, Culmen (anterior cerebellar lobe)
POL,SSL,Declive,Folium (middle cerebellar lobe) ISL,L. gracilis, biventer,tonsil (posteriorcerebellar lobe) Flol
Dentate
Centralnucleus
-{
Emboliform Globiform Fastigial
}-
. nuclel
16
1 Anatomy
The Borders 01 the Telencephalon Each hemisphere presents five well-defined borders, which define the cortical surfaces. 1. The superior medial border (German: Mantelkante) is the curved upper boundary of the hemisphere. It extends from the frontal to the occipital pole and separates the convexity from the medial surface (Fig. 1.11a, b). 2. The superciliary border separates the lateral surface of the frontallobe from its orbital surface. It extends from the frontal pole to the beginning of the Sylvian fissure (Fig. 1.11a). 3. The inferior lateral border follows the outline of the base of the brain on its lateral aspect, extending from the temporal to the occipitallobe. It separates the lateral surfaces of the temporal and the occipital lobes from their basal surfaces. The downward convex anterior portion of the bordering line corresponds to the contact between the temporal lobe and the middle cranial fossa; the upward convex posterior portion
marks the contact between the occipital lobe and the tentorium cerebelli (Fig. 1.11a, b). 4. The medial orbital border separates the medial surface of the frontallobe from its orbital surface. It is the basal continuation of the dorsomedial border, forming a straight line between the frontal pole and the lamina terminalis (Fig. 1.11b). 5. The medial occipital border can only be seen when the caudal part of the brainstem and the cerebellum are removed. As a coronal section shows, this somewhat rounded border lies in the angle between the falx cerebri and the tentorium cerebelli, separating the medial surface of the occipital lobe from its obliquely situated basal surface (Fig. 1.11b). This border lies caudal to the caIcarine suIcus and extends from the occipital marks the contact between the occipital lobe and the tentorium cerebelli (Fig. 1.11a, b).
Superomedial border
Superomedial border
b Superciliary border
a Inferolateral border
The Concept 01 Cerebral Lobes Although it is well known that the subdivisions of the cerebrum are purely arbitrary and do not correspond exactIy to any known functional divisions, neuroanatomists, neurophysiologists, neuroradiologists, and neurosurgeons have traditionally distinguished four or five lobes in each cerebral hemisphere: the frontal, parietal, occipital and temporal (and limbic) (Fig. 1.2, p.7, Vol. III B). In 1938, the German anatomist Friedrich Arnold advocated relating the terminology for the lobes to the overlying parts of the calvarium: frontal, temporal, parietal and occipital (previously, the terms anterior, posterior, and inferior were used). In 1866, the Scottish anatomist William Turner established the fissure of Rolando as the posterior limit of the frontallobe.
Fig. 1.11 The borders of the cerebral hemispheres. neuraxis a Lateral aspect of the left cerebral hemisphere b Medial aspect of the left cerebral hemisphere
In reality, the precentral, postcentral, and paracentral gyri are always named as specific entities not belonging to the frontal or parietallobes (Fig. 1.12). The seven cerebrallobes are: 1 Frontallobe 2 Centrallobe (precentral, postcentral, and paracentral gyri) 3 Parietallobe 4 Occipitallobe 5 Temporallobe 6 Insular lobe (partIy mesocortical) (see Fig. 1.51e, p. 54). 7 Limbic lobe (allocortical and mesocortical)
Topographic Anatomy for Microsurgical Approaches to Intrinsic Brain Tumors
') :re- l': a
)
,,--
17
~.-
7
b
Fig.1.12 A surgiealeoneeption of the eerebrum, with divisions into sevenlobes.Notethe uneertainborderlines between the lobes a Lateralview b Medialview 1-5 Neoeortieal 1 Frontal(yellow) 2 Central(blue) 3 Parietal(beige) 4 Oceipital(dark green) 5 Temporal(Iight green) 6 Insula(ef. Fig. 1.12c), partially alloeortieal 7 Limbie(orange),alloeortiealand mesocortical
I .. (
)
t 1
t 1
\.
Usingdynamic three-dimensional neuroimaging studies, as well as dynamic temporal (time) and spatial (space) magnetoencephalographic mapping, the coming generation of neuroscientistswill probablyintroduce an entirely new terminology in brain anatomy. Until such advances are made, it seems reasonable to us tocontinueto use the concept of centrallobes and, moreover, we advocateexpanding the lobar terminology from five to seven ¡obeso The reason for this division is that the insular gyri anatomicallyandfunctionally represent a well-defined lobe; similarly, the precentral,paracentral, and postcentral gyri have neuroanatomic, neurophysiological, and neuropathological similarities, and shouldthusbe termed the centrallobe.
/' /
Fig1.12c The limbic lobe of the cerebrum, seen from the surfaee. Neocerebral: uncolored; archipaleocerebral: orange; insula: red (see also Fig. 2.12, p. 137)
18
1 Anatomy
This seven-Iobe concept of tbe cerebrum better accounts for the structural and functional similarity of tbe regions comprising each lobe. Still, tbe brain should not be imagined in purely gross anatomical (and pbysiological) parts (i. e., seven lobes per hemisphere, a central core, and a brainstem, witb an elaborate connecting fiber tract and modulating system). Ratber, it is a dynamically functioning unit, witb a multitude of harmoniously interacting subsystems tbat extend beyond any artificial "lobar" concept. Clearly, there is localization of function in populations of neurons, but much less so tban tbe pioneering direct stimulation experimentation originally suggested. Overlapping populations of neurons, the chemoarchitecture of different regions of the brain, and neuronal plasticity, all upset tbe notion of a rigid, immutable functional anatomy. As a result, anatomical systems and subsystems (with many known,. as well as unknown, connections) sbould be viewed as dynamic structures composed of groups or pools of neurons and glia. As we attempt to better understand the neuropathological and neuropbysiological aspects of CNS tumors in relation to tbeir precise topography, we find ourselves looking beyond the lobar
Table 1.3 b Neocorticaltelencephalon
(frontal, central, parietal, occipi-
tal, temporal,and insularlobes)
systems. The internal structures of the brain, for tbese purposes, must be imagined in three dimensions, and tbougbt of in relation to the surface shape and orientation of tbe individual gyri. The histological and embryological aspects of tbe cortical organization, as related to tbe different lobes, are summarized in Tables 1.3a-c.
Table 1.3 a
Histological and embryological aspects of cortical
organization (from Leonhardt, Tondury and Zilles, Rauber/Kopsch: Anatomie des Menschen, Stuttgart: Thieme, 1989, vol. 3, p. 382 Semicortex Allocortex
~
MeoocMe,
~
Paleocortex
_______ Eupaleocortex
Archiocortex
Pe"allocorte, Prolsocortex
~
Pe"paleocorte, Penarchiocortex
Isocortex
Table
Neocortex
1.3c Transitional (allocortical
cephalon
(see al so Fig. 1.54a-d
in pages
and
mesocortical)
telen-
56-58)
Allocortex Brodmann area
Primary sensorimotor idiotypical isocortical areas
Unimodal association areas
Precentral gyrus Postcentral Cuneus + MTO (calcarine) T, posterior and transverse gyrus Sensorimotor operculum P, superior F" F2 posterior, middle F3posterior (pars opercularis) O" O2, 03
4 1,2,3 (S 1) 17
T3 middle and posterior T2 middle and posterior T, middle and posterior T2' T3 posterior
20
F, middle
9 10 45 Broca
F, anterior F3middle (pars triangularis) Heteromodal
asso-
ciation areas (homotypical cortex)
iso-
41,42 43 (S 11) 5, 7 6,8 44 Broca 18, 19
21 22 37
Frontalpole
11
F, medial rostral P2middle (angular) P3inferior (supramarginal) F2anterior and middle F3pars orbitalis Insula Posterior gyri
9,10,11,12 39 40 46 11,47 Isocortical
Paleocortical areas .Olfactory bulb Retrobulbar areas Olfactory tubercle Prepiriform area (ambient gyrus) Periamygdalar area (semilunar gyrus) Septal area (Parahippocampal gyrus)
Olfactory cortices
I
Archicortical areas Dentate gyrus Ammon's horn Subiculum Uncus Indusium griseum Hippocampal rudiment
Limbic lobe
Hippocampus
Mesocortex Peripaleocortical areas Postero-orbitofrontal (agranular) Anteroinferoinsular (agranular) Periarchicerebral areas Entorhinal area Presubiculum and parasubiculum
}
Parahippocampus >
Retrosplenial,cingulate, and subcallosal areas (23, 25, 33)
Cingulate
} Proísorcortícal areas Perirhinal (35, ectorhinal (36) Agranular temporal pole (38) Dysgranular ínsula Dysgranular cortex (13 posterior, 14 posterior)
Paralimbiclobe
Topographic Anatomy tor Microsurgical Approaches to Intrinsic Brain Tumors
Anatomy of the Sulci
19
concomitant variations in sulcal anatomy. Nevertheless, certain consistent patterns in sulcal anatomy can be observed and classified.
Surgically relevant descriptions of the sulcal anatomy and its associatedvascularityhave, until recently, been lacking. The Atlas o[IheCerebralSulci (Ono et al. 1990) details many of the anatomicvariations in the sulcal patterns, and is highly recommended.The key features of normal sulcal anatomy, which can be ofpracticalhelp in microsurgical transsulcal approaches, are outlinedhere. Thereis considerable individual variation between gyral patterns,as a result of the genetic and remodelling process (see the sectionon embryogenesis above). These gyral variations result in
The classifications shown here of the large (Fig. 1.13) and short main sulci (Fig. 1.14) are reproduced from Ono et al. 1990. Only four of the large named sulci consistently have a 100% uninterrupted rate: the Sylvian fissure, the collateral sulcus, the callosal sulcus, and the parieto-occipital sulcus. The central and calcarine sulci are found to be uninterrupted in 92% of cases. The rates of uninterruption for other sulci vary from 28% to 58%.
a
a
b
b
e
e
Flg.1.13 Percentage incidence rates of large median sulci (fram Ono, Kubik,andAbernathey,Atlas of the Cerebral Sulei, Stuttgart: Thieme, 1990,p. 14, Fig. 3.1). a Lateral surface, b basar surface, e mediar surface I Sulciwitha high eontinuous (uninterrupted) rate 4 Callosal sulcus, 100% 1 Centralsulcus,92% 2 Sylvianfissure, 100% 5 Calcarine sulcus, 92% 3 Collateralsulcus, 100% 6 Parieto-occipital sulcus, 100% /1 Sulciwitha lowcontinuous(uninterrupted)rate , Superiorfrontal sulcus, 36% 11 Superior temporal sulcus, 32% 8 Inferiorfrontal sulcus, 48% 12 Occipitotemporal sulcus, 36% 9 Postcentralsulcus, 46% 13 Cingulate sulcus, 58% 10 Intraparietalsulcus, 50% Iff Regularly interruptedsulei 14 Precentral sulcus, 100% 15 Inferiortemporal sulcus, 100%
Fig.1.14 Percentage incidence rates of short, branched, and supplementarysulci(fromOno,Kubik,and Abernathey,Atlasof theCerebral Sulci, Stuttgart: Thieme, 1990, p. 15, Fig.3.2). a Lateral surface, b basal surface, e medial surface
Classification of the Sulci
Short main sulci 1 Lateral occipital sulcus, 96% 2 Olfactory sulcus, 100% 3 Rhinal sulcus, 100% 4 Superior rastral sulcus, 100% 5 Inferior rostral sulcus, 98% 6 Anterior parolfactory sulcus, 88% Sulci eomposed of several branehes 7 Orbital sulci, 100% 8 Subparietal sulcus, 100% Supplementary (free) sulei 9 Intermediate precental sulcus, 14% 10 Marginal precentral sulcus, 30% 11 Intermediate frontal sulcus, 86% 12 Medial frontal sulcus, 68%
13 Lunate sulcus, 62%
20
1 Anatomy
Types of Sulci There are four types of sulci: axial, limiting, operculated, and complete. Axial sulci develop along the long axis of rapidly growing homogeneous are as. They are 10ngitudina11yinfolded (as seen in the posterior calcarine sulcus ofthe visual cortex). The invaginations or indentations made by axial sulci in any given gyrus lead to the formation of subgyri (as seen on coronal or horizontal brain sections). The subgyral white matter, limited by these sulcal invaginations, can be termed the subgyral sector of a named gyrus. Limiting sulci develop along planes separating cortical areas, which differ in the functiQns they predominantly subserve. Formed earlier in embryonic development than the axial sulci, they are more prominent in their appearance and greater in depth. (An example is the central suIcus). An operculated sulcus is similar to a limiting sulcus in that it separates structura11y and functiona11y different are as, but the transition occurs at the lip and not the floor. Often a third area of function is present in the floor and wa11sof the sulcus. An example is the lunate suIcus (separating the striate and peristriate areas at the surface), which contains the parastriate area within its wa11s. A suIcus that is deep enough to produce an elevation in the wa11of a ventricle is ca11eda complete suIcus. There is no obvious functional significance attached to the fact that some suIci are complete and others incomplete. (An example of a complete sulcus is the collateral suIcus). The suIci are not simply invaginations with accompanying vessels. A11sulci have complex anatomical shapes within their depths. A straight side-by-side relationship does not exist. Instead, interdigitating gyral ridges and promontories characterize the sulcal depths, resulting in serpentine CSF channels betweeen the gyri. A gyrus may have fingerlike projections within its depths. These projections at the bottom of sulci may indent upon extensions descending from other proximal or even distal gyri. The depth of the sulci may range from less than 1 cm in an axial suIcus to 2-3 cm or more in a complete suIcus. With crude
Gyral Cerebral Anatomy The large increase in cortical surface area, necessitated by the evolutionary development of the neocortex in man, occurred without a proportionate increase in cranial volume. This was made possible by the invagination of over two-thirds of the cortical surface into the depths of the sulci and fissures, resulting in the convoluted shape of the gyri of the human cerebral mantle. This gyral complexity has recently assumed greater clinical importance with the advent of MR imaging. Most neurosurgeons know the basic morphology of the cerebral gyri, and can fo11owand describe (on the surface, at least), their course, extent, and connections. It is genera11yaccepted that each hemisphere has the fo11owingbroad orientations of recognizable gyri: three horizontal frontal gyri, three horizontal temporal gyri, two slanting perpendicular gyri (precentral and postcentral), four to five diagonal insular gyri, and two to three semicircular parietal occipitallobules. This information is partia11ycorrect and of some descriptive value, but it is grossly inadequate for microneurosurgical applications.
lobectomy and lobulectomy techniques, injury to extensions of normal and pathologica11y uninvolved gyri at the base of a suIcus may occur, without recognizable injury at the convexity surface. The Importance of Sulcal Anatomy The study by Ono et al. (1990), based on twenty-five human brains clearly demonstrates the complexity and the extent of individual variation in the surface gyri. It is essential for neurosUfgeons to know the large and short main suIci, but it is unrealistic and unnecessary for them to know a11the variations in detail. Furthermore, an unfortunate problem exists. Many of the sulci described, observed in anatomical material, and visualized clearly in MR images of living brains, cannot be recognized on the surgica11yexplored brain surface. The MR images seem to overrepresent in-vivo sulcal widths and underestimate gyral surfaces (a pseudoshrinkage of gyri due to the electronic effect of the MRI technique). During microsurgical exploration of the living brain surface, it is therefore extremely difficult (or often impossible) to identify with certainty the same suIci that appear so obvious on the MR images. The gyri are tightly apposed to one another in the normal human brain in vivo, and often even more tightly apposed in the hemisphere of a brain containing a tumor. The lesion may cause a spectrum of swe11ing, distortion, and displacement. In such cases, even recognizing main sulci, like the central, precentral and postcentral, parieto-occipital, calcarine, and cingulate sulci, or even the Sylvian fissure, may be difficult and sometimes impossible during surgery. Nevertheless, a knowledge of suIcal anatomy, particularly the main large and short sulci (and their variations), is very important, as it provides the microsurgeon with the primary key to a systematic understanding of the anatomy of the gyri.
Though subtle anatomic variations of gyral convolutions exist from one patient to another, careful analysis reveals patterns that permit useful generalizations to be made. Neurosurgeons are presently confronted with a situation similar to that which occurred almost sixty years ago with the advent of cerebral angiography. The variations in the arterial and venous anatomy must have seemed overwhelming at that time. Eventua11y, however, analysis and experience demonstrated that the normal anatomy fo11owed basic rules and the cerebral vascular tree was systematized. This axiom also holds true for gyral and sulcal anatomy. The fascinating history of the discovery of gyral anatomy has taken centuries to unfold and wi11continue we11into the future. The Gyral Convolutions Any attempt to draw or depict individual gyral convolutions, even in two dimensions, is frustrated by the complexities of deciding where an individual gyrus begins and ends and which "extension
Topographic Anatomy tor Microsurgical Approaches to Intrinsic Brain Tumors arm"connectionsbetween gyri belong to which gyrus. As a result, it is cIearthat this approach to the formulation of a systematic planofgyralanatomy (from the outside) is flawed. Even a meticulousexaminationof the surfaceof the brain giveslittle idea of the complexityof the extensive infolding and interconnecting structureof the gyrihidden within the sulcal depths. A study of the depthsof the sulcireveals an impressively intricate anatomy, composedof interdigitating gyral extensions, with a seemingly randomseriesof dovetailing ridges, dips, and promontories between adjacentgyri (see Fig. 1.16c, 1.17, 1.34e, 1.44b, 1.51b). In addilÍon,there are considerable variations in the gyral anatomy, not onlybetween individuals but also between hemispheres in the sameperson (see Fig. 1.37, p. 45). Ultimately, the single most constantlyidentifiablesurfacefeature is the narrow Sylvianfissure,a resultof the overlapping growth of the infolding opercula of the surroundinglobes onto the insula (Fig. 1.15). Nevertheless, certainconsistentgeneralizations can be made about gyral patterns. Al! gyriareirregularinshape,and are composedof small,undulatingsubgyri(somesubgyriare even curved in a semicircularmanner).The only apparent exception is the gyrus rectus, which does appearto be straighl. On the surface, each gyrus seems to be separated from the adjacentgyri by short or long furrows that extend down to the bottomoftheinterveningsulcus(1-3 cm in depth). ThisgyralsepFig.1.15 Gyralconvolutionsof the Jeft cerebralhemisphere 1 Superior 2 Middle 3 Inferior F Frontalgyrus O Occipitalgyrus Op Opercularpart of inferiorfrontalgyrus OrbOrbitalpartof inferiorfrontal gyrus P Parietalgyrus Pe Postcentralgyrus Pr Precentralgyrus T Temporalgyrus Tri Triangularpartof inferiorfrontal gyrus
21
aration is, however, very superficial. At a depth of 2-3 mm, we start to see multiple extension "arms" connecting the gyri. These are the short or long, small or voluminous gyri of the main gyri. These transverse gyri cross over at acute angles to the line of the sulcus (Fig. 1.16). When the depths of the sulci are examined, reciprocal transverse gyral interdigitations (like intertwined fingers) are seen, that add to the complexity of this puzzle. Due to the shorter and longer intrasulcal furrows, the inner surface of the gyri becomes undulating. The wavelike opposing surfaces of the gyri are interlocked like cogwheels. (Ono et al. 1990, Figs. 6.3a-f, 6.191,19.3). We have long been aware of the hidden gyri of the insula (within the Sylvian fissure). But within every sulcus, there are numerous short and long gyri that should be termed the transverse gyri of a main gyrus, e. g., the transverse gyri (Heschl) of the superior temporal gyrus (Fig. 1.17; see also Figs. 1.16a, e, 1.34e, 1.35, 1.42c, 1.43b, 1.44a, 1.45c, 1.46b, 1.48b, 1.49b, e, 1.51b). In fact, each gyrus has 5-10 well-shaped intrasulcal extensions (transverse gyri), hidden within the depths of the sulci. They make up over two-thirds of the cortical surface and have not, as yet, been mapped by physiological studies. Neither present-day triplanar nor even three-dimensional MR imaging can demonstrate the extensive nature of this gyral interface.
22
1 Anatomy
Fig.1.16 The left superior frontal gyrus. Note the multiple extension "arms" (arrows) connecting this gyrus to the surrounding gyri. The transverse gyri are not visible from the surface, and can only be appreciated when the surrounding sulci are fully opened
Fig. 1.17 a The left frontal lo be of a rubberized brain. The depths of the sulci and interdigitations of the gyri are well shown. Note the variations in the inferior frontal gyrus. There is a connective arm Iying superoanteriorly to the middle frontal gyrus (small arrow). Note the transverse gyri in the frontal sulci. Central sulcus (large arrow) F1 Superior frontal gyrus F2 Middle frontal gyrus Op Opercular part
Orb Orbital part Tn Tnangular part Pr Precentral gyrus
of inferiorfrontal gyrus
}
Topographic Anatomy for Microsurgical Approaches to Intrinsic Brain Tumors Fig.1.17b The right centrallobes of a rubberizedbrain with opened sulci. Note the intersulcal transversegyri and their connections 1 Precentralsulcus 2 Centralsulcus 3 Postcentralsulcus Pc Postcentralgyrus Pr Precentralgyrus
Fig.1.17e The left Sylvian fissure in a rubberizedbrain,viewed through the slightly opened sulci.The intersulcal structures can be well recognized.The middle temporal gyrus lies between thesuperiortemporal sulcus (sts) and the interruptedmiddle temporal sulcus (mts) 1 Insula 2 Opercular part of inferior frontal gyrus 3 Precentralgyrus 4 Postcentralgyrus 5 Inferiorparietallobule 6 Superiortemporal gyrus 7 Anteriorand posterior transverse gyri (Heschl) 8 Middletemporal gyrus mIs Middletemporal sulcus sts Superiortemporal sulcus
23
I 24
1 Anatomy
General Features of Hemispheric Gyri Another problem is that the gyri form a continuous convolution system. It is noteworthy that no sharp delineation exists between what may be deemed the "beginning" and what appears to be the "end" of a gyrus. Starting at a polar area (whether frontal, temporal, or occipital) and tracing the contour of any prominent gyrus, it is possible to follow that same gyrus without interruption along the whole length of the hemisphere. This gyral continuum concept (uninterrupted gyri throughout a hemisphere) is iIIustrated in Figure 1.18. This continuous gyral pattern is hemispheric in the cerebrum and bihemispheric in the cerebellum. The striking feature of the gyri on the lateral, medial, or basal surfaces of the hemispheres is the serpentine configuration of the gyral convolutions. They are fairly constant in number, size, and orientation, but vary in their connections, both on the surface and in the depths of the sulci and fissures. A consequence of this gyral variation are the irregular interruptions of suIci. The individual variations of gyri on MR images and during surgical exploration of the brain surface, have convinced us that all attempts to identify areas or structures of the brain by surface "Iandmarks" are, at best, limited.. For example, it is simply impossible to recognize and identify, with any precision, the suIci and gyri and distinguish them from each other during open surgery. Even using computer-assisted intraoperative location devices, the tight suIci significantIYalter accuracy.
As a result, we have found that, for cIinical examinations, the complexities of gyral anatomy are best conceptualized by examining the basic white-matter patterns underlying gyral developmento This results in a simplified scheme that allows for variations. It is used to enhance anatomical descriptions, so that all those involved in the care of brain tumor patients immediately recognize the are a alluded to. It should prove useful in comparative studies and treatment trials between tumor groups, surgeons' patient groups, and institutions. Most important of all, it is very useful in planning the correct approach for intrinsic neoplastic lesions of the CNS. As previously mentioned, however, the application of topographic gyral information to surgery is still problematic. Hopefully, the newly-developed frameless stereotactic localization system will provide much-needed improved accurary (Kikinis and Jolesz 1993). Conversely, the technical innovations will produce new anatomical concepts, such as the proposed scheme of gyral segment or white matter, which is based on the embryological, functional, pathological, topographic, and surgical characteristics of the white matter of the brain.
/'
Fig. 1.18 The gyral continuum concept, demonstrated in the left cerebral hemisphere. There is an uninterrupted connection, either on the surface or intrasulcally, between every gyrus within the hemisphere
Topographic Anatomy tor Microsurgical Approaches to Intrinsic Brain Tumors
TheWhiteMatterof the Cerebrum Themassivemedullary center of each cerebral hemisphere consistsof myelinated connecting fibers, other subsystem pathways, andvessels(Fig. 1.19, 1.20). Myelin is a semifluid fatty substance, whosepresence gives the characteristic color to the white matter. ofthesefibersystemscan nowadaysbe clearlyidentified on Some triplanarMR images (optic tract, corticospinal tract, anterior commissure,etc.). Our ability to think in three dimensions can be
greatly improved by studying coronal, sagittal and axial MR sections. With practice, it becomes a straightforward matter to construct a three-dimensional image of the orientation and major connecting of fiber systems within the white matter, including the relations to the central nuclei and the ventricular system. When combined with the three-dimensional gyral anatomy, this greatIy aids the surgeon in gyral, su1cal, and CSF space identification during microsurgical dissection.
/ /
Fig.1.19 The organizationand course of the projecting fibers within the cerebral white matter (after Krieg, Neuroanatomy, 2nd ed. New York: McGraw-Hill,1953, p. 659, Fig. 106). Note the following projection bundles:frontal, precentral and postcentral (pyramidal fibers), occipital(opticradiation), temporal (acoustic radiation), and temporoinsulofrontal(uncinate fasciculus)
25
26
1 Anatomy
J~
J
1 ~./
/
~--~
>
Fig. 1.20 Artistic perception 01 the organization and complexity 01the white-matter libers within the cerebrum Red Projection libers Blue Association libers Yellow Commissural libers Green Cingulate libers
Gross anatomical and pathological examination of the brain reveals a substructure beneath the lobar level which is immediately evident, but for which no common terminology exists. The suIci and fissures serve to divide the convolutions and underlying white matter into a "cauliflowerlike" substructural arrangement. (Fig. 1.21). This substructure (readily evident on MR images) can now be viewed in vivo with great cIarity. One great advantage of this scheme of cerebral anatomical organization is that individual gyri can be consistentIy and reliably identified, even when
expanded or displaced by tumors. This inside-out conceptual ization is the key to understanding the sectorial architectural organization of the cerebrum. In this section, a practical scheme is presented that emphasizes not the traditional lobar architecture of the cerebral hemispheres, but the gyral and subgyral white-matter patterns and relationships. These patterns are, in fact, a consequence of the underlying anatomical cascade, based on an organized sectorial framework. The word "cascade" may appear to be a surprising choice in
Topographic Anatomy tor Microsurgical Approaches to Intrinsic Brain Tumors ananatomicalscheme, but is chosen to indicate that the structure ofthe brain is genetically programmed and repeatedly modified duringdevelopment at a variable, multidichotomic rateo Early topographicaldevelopment is especially dynamic and subject to manyintrinsic and extrinsic modifications (see the section on embryogenesis above). The ultimate gyral configuration is dependenton the sum of total developmental "events" (pro-
27
grammed embryological stages, maldevelopments, intrauterine or neonatal insults, etc.) and of "inputs" (programmed, acquired, and learned information, etc.). The basic sectorial (peduncular) organization of the brain reflects its true "inside-out" development by neuronal migration, myelination, selective neuronal death and synaptic development (see the section on neurogenesis above ).
b Fig.1.21 Sectoral (peduncular) architectural organization, as seen in nature. a structure
Contents01the White Matter In triplanar section, the white matter appears as a uniform structure.Its voluminous mass can usually be equated with only the fibersystemwithin it. The white matter, however, contains other structures,indicated in Table 1.4. The courses of blood vessels and connectivefiber systems within the white matter have been well studiedand described. Other sub-systems (5-6 in Table 1.4 and Fig.1.25),however, are still not entirely known. In the near future, morespecific information about these pathways will hopefully be available.
Surface view 01a caulillower, and b
its sectoral (peduncular) sub-
Table 1.4 Contents of the white matter (subsystems) cephalon
of the telen-
The embryological neuroglial migration network oriented in the ventriculocortical axis within individual gyral segments 2 The vascular pattern 01white matter withinthe gyral segments a Arterial pathways: The branches 01the leptomeningeal arteries pass centripetally and supply the external lour-lifth01the white matter, while branches 01the perlorating arteries course centrilugally and supply the internal lifth01the white matter b Venous pathways: In contrast to the arterial distribution, only the external lifth01the white matter drains centrilugally to the superlicial medullary veins. The internal lour-lifths 01the white matter drain centripetally to the deep medullary veins 3 The connective libers a Association libers (hemispheric pathways) b Commissural libers (hemispheric pathways) c Projection libers (hemispheric, but crossing) - Corticothalamic, corticostriatal, thalamocortical - Corticopontine, corticobulbar, corticospinal 4 Transcerebral pathways 01the CSF 5 Neurotransmitter pathways 6 Neuroimmune pathways
28
1 Anatomy
White Matter Sublevels The white matter substancecan be subdivided into two distinct zones,the peripheral zone (gyral) and the central zone (capsular). The large peripheral zone consists of white matter located between the cortex and the periventricular matrix of the lateral ventricles. The smaller central zone consists of the projection fibers surrounding the lateral border of the lentiform nucleus (external capsule) and the V-shaped projection fiber bundles, located lateral to the lentiform nucleus and lateral to the caudate nucleus and thalamus (intemal capsule). The peripheral zone contains all of the axonsthat travel to or from the hemispheric or bihemispheric cerebral cortex, while the central zone contains only axonsthat connect the cortex to structures outside of it (corticopontine, corticobulbar, cortico-spinal, corticothalamic, and corticostriatal). While all varieties of whitematter fibers (association, commissural, and projection fibers) are found in the peripheral zone, only projection fibers are located in the central zone.As a result of its make-up (to andfrom cortical axons), the peripheral zone is composed of groups oí fibers that are segmentally related to the adjacent, overlying cortex (gyrus). Morphologically, each gyral segment resembles a cone (or pyramid), whosebaseoccupies a segment of a gyrusand whose apex is directed perpendicularly towards the underlying periventricular matrix (Fig. 1.22). Thus, the entire peripheral zone can be subdivided into numerous gyral segments. The number of gyral segments per gyrus varies between individual gyri (i.e., the gyrus rectus only has one segment, while the superior frontal gyrus has more than ten). Each gyral segment can be subdivided into sectors (Fig. 1.23, Table 1.5). Sectors 1 to 4 comprise the peripheral white matter zone, while sector 5 alone constitutes the central zone. It is noteworthy that the telencephalic vascularization pattern (arterial and venous), the embryonic neuroglia migration pathways, and the perivascular CSF drainage routes all resemble (to a striking degree) the characteristic gyral segmental white-matter pattern which courses from the cortex (sector O)through the peripheral zone (sectors 1 to 4) towards the periventricular matrix. This pattern is clearly different from that of the projection white-matter fibers, which run a similar course through sectors 1 to 4, but then leave sector 4 to enter sector 5 (the capsular sector). The term "peduncular" better describes the projection fibers, as they cascade in peduncles from the corte x to the internal capsule. (Fig. 1.24). The hemispheric association fibers and bihemispheric commissural fibers also leave sector 4, but do not enter sector 5.
b
Table
/
of white matter within gyral segments
Description
Composition
o 1
Cortical sector Subcortical sector (peduncle) Subgyral sector (peduncle) Gyral sector (peduncle)
Gray matter
3
Fig. 1.22 The pyramidal (conical) structure of gyral segments a Lateralview b Coronal view, demonstrating the gyral segments of the individual lobes e A gyral segment in relation to the lateral ventricle
Sectors
Sector
2
e
1.5
4
Peripheral white matter
Lobar sector
(peduncle) 5
Capsular
sector
(peduncle)
Central white matter (external and internal capsule)
Topographic Anatomy for Microsurgical Approaches to Intrinsic Brain Tumors F:g.1.23a
29
Artistic diagram of the white matter sectors within a gyral
segment(coronal section) O Cortex (gray) . Subcortical (brown) 2 Subgyral (Iight brown) 3 Gyral (yellow) 4 Lobar (dark brown) 5 Capsular. pyramidal
~
I
Flg.1.23b The white matter sectors and connective fibers (coronal section)
Short association
fibers
Long association
fibers
Commissural fibers Sector of gyral segments
Head of caudate nucieus
Claustrum Putamen
Thalamus
Globus pallidus Subthalamic nucieus
Anterior commissure
Corticopontine, corticobulbar, and corticospinal fibers , Substantia nigra
30
1 Anatomy Fig, 1.24 The sectorial organization 01 the projection libers (coronal section), 0-1 e: Sectors or peduncles, respectively, in relation to the projection libers O Cortical sector
1a Subcortical sector 1b 1c 1d 1e
Subgyral sector Gyral sector Lobar sector Capsular sector
(or peduncles)
11 2 3a 3b 3c 4 5
Callosal radiation (red) Hippocampus Parahippocampus (Iight green) Insula (olive) Cingulum (dark green) Central nuclei (gray) Ventricular (blue)
}
I
l) l'
White Matter Subsystems
), !(
\
The white matter is composed of six subsystems (Table 1.4). 1. The migration pathway of neuroglia follows their embryologic course from the periventricular matrix to the cortical areas (sector O) along the sectors of gyral segments 4, 3, 2, 1 (Figs. 1.8a, 1.2Se) (Kostovic and Rakic 1980, Hinrichsen 1990: 380-448). 2. The vaseularizatiou patteru of gyral segmeuts. Arterial (see M. Marin-Padilla, "Embryology," Vol. lIlA in Chapter 2 of the present work). The branches of the leptomeningeal arteries course through each gyral segment centripetally along sectors O, 1, 2, 3, down to sector 4 (and the periventricular area) (Fig. 1.23a). This results in a conical-shaped pattern of white-matter arterial supply (resembling the shape of the individual gyral segments). The branches of the perforating arteries centrifugally supply sectors 5 and 4. The external capsule receives its arterial blood supply from M2 branches, whereas the internal capsule receives blood supply from branches of the lenticulostriate arteries, from the anterior choroidal artery and from Heubner's artery (Table 1.12b, p. 103, Fig.1.2Sa). Venous. The white-matter venous drainage pattern is contrary to the pattern of the arterial supply. The venous drainage of sector 1 is outwards (centrifugal) into the superficial medullary veins, whereas sectors 2, 3, 4, and the internal capsule drain centripetally into deep medullary veins. The external capsule drains into both superficial and deep medullary veins (Fig. 1.2Sb).
3. The trauseerebral CSF- pathways. The subarachnoidal CSF circulates along the subpial-periarterial spaces centripetally from cortical sectors down to the ventricle (sectors O, 1,2,3,4). Along the subpial-perivenous spaces of transcerebral veins, CSF also circulates along gyral segments between the ventricular and subarachnoidal cisterns (Krisch et al. 1984, Weller et al. 1986, Cserr et al. 1992). Cserr et al. (1992) state: The main channel of interstitial fluid (ISF) flow appears to be the perivascular space, or Virchow-Robin space (Fig. 4.Sa), as first suggested by His (1965), over a century ago. This space extends along arteries and veins penetrating into the nervous tissue, down to the point at which arterioles and venules merge with capillaries. Recent studies (Krisch et al. 1984, Hutchings and Weller 1986) have shown that it also extends along vessels in the subarachnoid space, as first suggested by Foldi et al. (1968). Thus, perivascular spaces can be categorized according to their location as either intracerebral or subarachnoid. Because the outer sheath of the perivascular space is permeable (Zhang et al. 1990) and, further, since blood vessels penetrate to all portions of the brain, the vast network of perivascular spaces provides a pathway for bulk flow of extracellular fluid between CSF and the entire CNS, comparable in extent to the lymphatics. It is not known whether the perforating arteries and the deep medullary veins are also surrounded by a similar channel with such a subependymal CSF pathway (Figs. 1.2Sd); (see also Chapter 4).
, j "J
Topographic Anatomy for Microsurgical Approaches to Intrinsic Brain Tumors
b
31
e
e
Fig. 1.25 White-matter subsystem segmental patterns (coronal view) a Arterial blood supply from the leptomeningeal arteries centripetally, and from the perforators centrifugally b The superficial and deep venous drainage e Neuroglial migration pathways d Transcerebral CSF pathways e The association and commissural fibers f Corticopontine and corticospinal fibers 9 Thalamic-striatal radiation h Neurotransmitter pathways
32
1 Anatomy
4. The connective fibers (Table 1.6). a) Association fibers (Figs.1.25e, 1.26, 1.27). The short association fibers (arcuate or U-fibers) are located only in sector 1 (connecting parts of a single gyral segment). Fibers range into sectors 2 and 3, connecting the cortical are as within the same gyrus or neighboring gyri. The arterial supply of these fibers is from branches of leptomeningeal arteries, and the venous drainage is centrifugal to the superficial medullary veins. The long association fibers (fasciculi) course through sections O,1,3, and 4 of numerous gyral segments, and be come a well-circumscribed bundle in sector 4 that runs from one lobe to another (over long distances) to conI}ect ipsilateral gyri. These fiber bundles intersect the radiating prajection fibers at sector 4. These long fiber bundles receive arterial blood from branches of numerous leptomeningeal arteries (from the anterior, middle, and posterior cerebral arteries). The venous drainage is centripetal to the numerous deep medullary veins.
Table
1.6
Reciprocal
Connective
b) Commissural fibers. 1. Bihemispheric bundle s of anterior commissural fibers course in a semicircular fashion from sectors O, 1,2,3, and 4 of temporal gyral segments along the basal ganglia to the opposite temporal gyral sectors 4, 3, 2, 1 and O (Fig. 1.25e). The posterior (hippocampal) bundle connects the hippocampi from both sides. The anterior commissure in temporal areas receives its arterial blood supply fram the leptomeningeal arteries of A¡, MCA, and PCA, whereas the middle part receives its blood supply fram branches of the anterior communicating artery, Heubner's artery and Al branches. The hippocampal commissure receives its arterial blood supply from the posterior choroidal artery and P3 and P4branches and drains into the atrial veins. The anterior commissure venous drainage is to the superficial medullary vein in sector 1, and to the deep medullary veins in sectors 2, 3, and 4 (bilateral).
fibers of the telencephalon
associative
fibers (i"psilateral hemispheric)
Short arcuate fibers
Reciprocal connection between two ipsilateral gyri, interconnecting the primary, modality-specific areas of the 90rtex
(U-fibers)
Between the ipsilateral neighboring gyri: connective fibers between modalityspecific areas of the parasensory association cortex and multimodal sensory areas
Long arcuate fibers
Superior longitudinal fasciculus Inferior longitudinal fasciculus Superior fronto-occipital fasciculus Inferior fronto-occipital fasciculus Vertical occipital fasciculus
Connecting the ipsilateral modalityspecific parasensory association cortex and the multimodal areas of the neocerebrum in the occipital, temporal, and parietal lobes with the premotor and prefrontal cortex of the frontal lobe
Uncinate fasciculus (hook bundle)
Connection fibers between the frontal and temporallobes (neocortical, allocortical and mesocortical): fronto-orbital, anterior F2' F3and anterior T3' T2' T, and parahippocampus, and to the anterior part of the insula (under the limen insulae)
Arcuate fasciculus (Zenker 1985, p. 157)
Between middle F2' F3external capsule and middle T3-T2
Cingulum
Main medial connective fibers within the white matter of the cingulate gyrus, between the neocerebrur:1 (frontal, central, parietal, occipital, temporal) and the limbic lobe (allocerebrum)
Corpus callosum Rostrum, genu, body, splenium Forceps minor (frontal)
commissural
Anterior commissure Anterior crus Posterior crus
Allocortical connection of the contralateral parahippocampal areas via the splenium
Forceps major (occipital) Hippocampal commisure (fornical commissure) Posterior commissure
Allocortical connection to the crus fornicis (psalterium); fibers from the entorhinal and perirhinal cortices Not a real commissure. A fiber connection between the contralateral thalamic medullary stria and the pretectal nuclei
Projection fibers Partially reciprocal corticofugal and corticopetal connective fibers Ipsilateral thalamostriatal, thalamocortical and corticothalamic fibers crossing over within the brain base or brainstem Olfactory tract (traverse allocortical areas) Gustatory fibers Acoustic tract Optic tract Ipsilateral hemispheric fibers, a minor part crossing the midline along the commissural system Fornix Connective fibers between the hippocampus and the septal nuclei, the anterior nucleus of the thalamus, the head nucleus of the stria terminalis and the mamillary body Striatal Crossing over via the subcallosal system Ipsilateral telencephalic fibers, crossing over within the brainstem Neocortical
Reciprocal
Interconnecting the corresponding neocortical are as of the contralateral hemispheric lobes
fibers (contralateral
hemispheric)
Connection to the olfactory bulb Contralateral connective fibers between the allocortical areas; amygdala, parahippocampus
Allocortical and mesocortical
r
Corticodiencephalic
(thalamus,
hypothalamus) Corticomesencephalic (nucleus ruber, substantia nigra) Corticopontine (frontopontine, temporopontine, parietopontine, occipitopontine, insulopontine undiscovered) Corticobulbar (nuclear) Corticospinal Limbic lobe to diencephalon, mesencephalon and metencephalon
Topographic Anatomy tor Microsurgical Approaches to Intrinsic Brain Tumors
33
b
Fig. 1.26 Coronal view of the white-matter connective fibers within the cerebrum Blue Projection fibers Green Association fibers Red Commissural fibers 1-5 Long association fibers 1 Cingulum 2 Superior fronto-occipital fasciculus 3 Superior longitudinal fasciculus 4 Inferior fronto-occipital fasciculus 5 Inferior longitudinal fasciculus a, b Commissural fibers a Anterior commissure b Callosal fibers c Corticobulbar, corticopontine, and corticospinal fibers u U-fibers, short association fibers
@)
4
b
Fig.1.27a,b
Lateralview of the left cerebral hemisphere: association
fibersdissected by Dr. J. Klingler, Institute of Anatomy, Basle (from Lud..IgandKlingler, Atlas Humani Cerebri, Basel: Karger, 1956, Tabula 8)
1 2 3 4 5 6 7 8 9 10
Inferior occipitofrontal fasciculus Uncinate fasciculus External capsule White matter of the middle and inferior temporal gyri Inferior occipitofrontal fasciculus Superior longitudinal fasciculus Superior longitudinal fasciculus Sagittal stratum Fibers of splenium Parieto-occipital sulcus
34
1 Anatomy
2. Cal/osal[ibers are bihemispheric between the corresponding cortical areas of the seven cerebral lobes. They traverse thraugh sectors O,1, 2, 3, and 4 of the gyral segments of one side, then cross the midline and traverse the opposite sectors 4, 3, 2, 1, and Oof the gyral segments of the opposite side. The course of the callosal fibers is shown in Fig. 1.28a-b. Comprehensive information on the anatomy and physiology is given in Innocenti (1986): "The adult corpus callosum consists of myelinated and unmyelinated axons. The former comprise 43-69% of the total. There is but scanty information regarding the number of callosal axons. In an electron-micrascopic study, we have recently counted an average of 23 million axons in the ~dult cat, i. e. nearly 5 times more than the previous light-micrascopic figure. Few studies have tried to correlate the topography of the corpus callosum with that of the hemispheres." Sunderland (1990) concluded his Marchi study on the Macaque by stating, "The localization in the corpus callosum is of a very general type. Not only are the commissural fibers fram some cortical areas diffusely spread over the corpus callosum, but there is also an overlap of fibers coming from difierent areas in the same lobe, and also, apparently, fram areas in difierent lobes." This statement is still appropriate. More precisely, though, in cats and monkeys the rostrocaudal axis of the corpus
callosum corresponds roughly to that of the hemisphere, while the dorsoventral callosal axis does not seem to correspond to a mediolateral trajectory of the hemisphere. The topography ofaxons in the corpus callosum is interesting for at least two, very difierent, reasons. Sidfis et al. (1~81) demonstrated that different parts of the corpus callosum transfer the sensory and semantic attributes of a visual stimulus. It would be important to know which area-to-area connections are responsible for these two actions. On the other hand, sex-related differences in the shape of the corpus callosum have been found in human brains (De Lacoste-Utamsing and Holloway 1982). "Women seem to have a more bulbous and larger splenium than men, which may indicate difierences in the connectivity of specific cortical areas in the occipital, temporal, or parietal lobe. Along the same lines, it is interesting that left-handers seem to have a larger corpus callosum (at equal brain weight) than right-handers" (Witelson 1983). More detailed information is available in Innocenti (1986) and from recent MRI-based morphometric studies by Weiss et al. (1993). The anatomical variation of the corpus callosum as well as of the splenium and fornix has been studied using MRI. These are shown in Fig. 1.29a-c.
-
I ,,
I
\'1
"
..
k
r
-
I
I
4
JtI-
'
vr
.... 11
1
,
t. l\'),\ ..
.1
l
b a Fig. 1.28a The commissural fibers of the telencephalon, viewed from the basal side of the brain (from Niewenhuys et al., 8, Berlin: Springer, 1988, p. 366, Fig. 214)
Fig. 1.28b An anatomical dissection of the corpus callosum, performed by Dr. U. Türe. The indusium griseum has not been removed in the anterior and posterior areas of the corpus
Topographic Anatomy tor Microsurgical Approaches to Intrinsic Brain Tumors
35
Flg.1.29a-c Average measurements and variations in the callosal body,related to the analysis of 150 MR images on normal subjects over sixyearsof age, and 25 anatomical brain sections (from Iliev-Urbaniec, Kubik,and Valavanis, Corpus callosum 1988). See also Figures 1.6b,d, 1.7a-e, 1.34c,1.38a, b, 1.41a, b, 1.53a, 1.57a, b, 1.58a, b, 160d, 1.67a, b a Variationsin the corpus callosum b Variationsin the splenium e Variationsin the position of the fornix
e
b
Thecallosal arterial supply in sectors 1-3 is from branches of ¡hesegmentalleptomeningeal arteries, and in sector 4 from the perforating arteries. The arterial vascularization of the hemispheric callosal fibers involves the participation of all three cerebral arteries (ACA, MCA, and bilateral PCA), in addition to of the lCA along the anterior choroidal artery. branches
The outer layers of the anterior two-thirds of the callosal bodyreceivetheir arterial blood supply from Az branches, and the inferiorlayer of this area from perforators (even from the anteriorcommunicatingartery). The posterior third of the corpus callosum(splenium) receives its blood supply from P3 and P4 branches (and the inner layer of the same area from perforators of ¡heposterior choroidal arteries).
The venous drainage in sector one is outwards to the superficial medullary veins, and in sectors 2, 3 and 4 down to the deep medullary veins. The venous drainage of the anterior callosal body (externallayers) is outwards to the superficial veins, while the inner layers drain to the deep medullary veins (or to the septal and frontal atrial veins). There are collaterals between both venous drainage systems (which can be seen in cases of callosal AVMs). The venous drainage of the posterior part of the callosalbody is outwards (in the outer two-thirds layers) and inwards (in the internallayer) to the deep medullary vein (atrial). It is interesting that the callosal arterial branches of both hemispheres do not have collaterals along the callosal fibers that cross the callosal body, while the veins of the callosal body have multiple connections across the midline.
36
1 Anatomy
c. Projection fibers (afferent and efferent). These fibers connect the telencephalic cortex with the basal ganglia, central nudei, brainstem, and spinal cord. There are two main connective fiber systems. 1. The regional corticothalamic-thalamocortical fiber system (optic, acoustic, vestibular, somatosensory, gustatory, and olfactory), and the corticostrio-pallido-thalamocortical fiber systemoTheir fibers pass along the gyral segments 0,1,2,3, and 4, and enter into the striatum and thalamus through the internal capsule, the corticodausatral fiber through the external capsule. They return from the thalamus along the sectors 4, 3, 2,1, and the corte x (Fig. 1.25g, 1.30a, b). . 2. The corticopontine, corticobulbar, and corticospinal fibers pass through the sectors of gyral segments O, 1,2, 3, and 4, and merge in a peduncular fashion to form the external and internal capsules (sector 5, or the hemispheric pedunde). (see pages 66-68 and Fig. 1.25f, 1.31a, b.) The arterial supply of fibers in sectors 1, 2, 3 is from leptomeninge al branches, and in sector 4 from perforating branches of all the cerebral arteries (ACA, MCA, and PCA) and the anterior choroidal artery. The venous drainage is centrifugal in sector 1, centripetal in sectors 2, 3, 4.
It is dear that commissural and projection fibers course together with the vascular and other network systems (through sectors 1,2,3, and 4) within gyral segments. At the lobar level (the central zone of the white matter), the commissural and projection fibers merge into well-defined bundles that intersect each other in a semicircular fashion and leave the gyral segments. The anterior and hippocampal commissures and fornix are selectively small bundles in comparison to the enormous fiber system of the anterior, superior, and posterior parts of the callosal commissure. The supralenticular, retrolenticular, and sublenticular projection fibers of the seven lobes (frontal, central, parietal, occipital, temporal, insular, and limbic lobes) course within the radiate corona to converge into the anterior and posterior limbs and genu of the internal capsule, and then finally merge as the crus cerebri (Fig. 1.31). Within the internal capsule, crus cerebri, and cerebral pedunde, these projection fibers have a distinct topographic organization, corresponding to the telencephalic lobar and thalamic pedundes (pp. 65-68, Fig. 1.62, 1.63).
Gyri orbitales
Gyri occipitales
Fig. 1.30a Connections between the thalamic nuclei and the cerebral cortex, in horizontal section (from Niewenhuys et al., The human Central Nervous System, Berlin: Springer, 1988, p. 242, Fig. 167) 1 Gyrus cinguli 2 Corpus striatum 3 Globus pallidus 4 Nucleus anterior thalami 5 Nucleus medialis thalami 6 Nucleus ventralis anterior 7 Nucleus ventralis lateralis 8 Nucleus ventralis posterior 9 Nucleus ventralis posterior, pars parvocellularis 10 Nucleus lateralis posterior 11 Nucleus centromedianus 12 Nucleus parafascicularis 13 Pulvinar thalami, pars anterior 14 Pulvinar thalami, pars medialis 15 Pulvinar thalami, pars lateralis 16 Corpus geniculatum laterale 17 Corpus geniculatum mediale
Topographic Anatomy for Microsurgical Approaches to Intrinsic Brain Tumors
37
Superior thalamic peduncle to the premotor and motor cortex and the somatic.sensory cortex nterior
thalamic
lo the cortex
peduncle
Posterior thalamic
01
he frontallobe, nd cingulate
peduncle, to the cortex 01 the parietal, occipital and temporallobes
gyrus
!nlerior thalamic peduncle, to the ital cortex 01the entallobe, the temporal ele and the amygdaloid
nucleus
Fig. 1.30b The thalamic peduncles (radiations projected onto the brain surface), lateral view (alter England and Wakely, A color atlas 01 the brain and spinal cord, London: Wolle, 1991) Diagram to show the thalamic peduncles (radiations) projected onto the brain surlace, lateral view
b
. .1.31a, b
A dissection 01 the projection and commissural
libers,
rlormedby Dr.J. Klingler,Institute 01Anatomy, Basle (Irom Ludwig AtlasHumaniCerebri,Basel:Karger,1956) d Klingler,
1 2 3 4 5 6 7 8 9 10 11 12
8
7
Corona radiata Internal capsule Decussation of the callosal body fibers and corona radiata Impression 01the lentilorm nucleus Ollactory tract Cortex 01the parahippocampal gyrus within the amygdala Lateral cerebral lossa Anterior perlorator substance Anterior commissure Temporopontine tract alter partial removal of the optic tract Temporopontine tract The optic radiation in sagiUal projection
38
1 Anatomy
The semicircular and semispiral courses of connective fibers are indeed difficult to imagine, even though numerous publications have presented brilliant two-dimensional and three-dimensional illustrations. There are also models, three-dimensional slides in stereoscopic atlases, and three-dimensional videotapes available. Still, an inborn artistic ability is necessary to develop an instantaneous yet understandable imaginative representation of these pathways from a specific perspective. This mental image can be stimulated by repeated exercises in drawing, and particularly by hand-crafting different pathways using colored wires and plastic (Fig. 1.32).
5. The neurotransmitter pathways are currently being investigated; current knowledge is presented in related publications of neuroscientists in the field of neurochemistry (Niewenhuys et al. 1985). Basically, their course is along the associate, commissural, and projectingiiber pathways. (Fig. 1.25h). (See Chapter 4, Neurophysiology. ) 6. The pathway of the neuroimmune system follows the vascular and neuroglial pattern (gyral segmental fashion), but may enter the basal ganglia, central nuclei, and brainstem (see Chapter 4). (See Fig. 1.25c.)
Fig. 1.32 The sectorial five-tiercascading organization of the white matter Orange Commissural fibers Yellow Frontopontine fibers Red Corticospinal fibers Beige Parietopontine and temporopontine fibers Green Occipitopontine fibers
Topographic Anatomy tor Microsurgical Approaches to Intrinsic Brain Tumors
WhiteMatterSublevelsand Clinicallmplications 8asedon MR coronal views of the brain, five sublevels in white matter,according to major sectorial divisions, can be identified (Fig.1.23a-b, 1.24). Starting from the outside they are subcorticallyand proceeding inwards: subcortical, subgyral, gyral, lobar, and capsular. This subdivision of the white matter into sectors has madeour interpretation
of CT and MR images more methodical,
hascIarifiedthe way we interpret surgical pathology, and has modifiedour approach to surgery for intrinsic CNS tumors. It has been evidentto us that gliomas may begin in what we have termed the subcorticalsectors of a gyrus (see Figs. 2.2-2.4). They then grow by expandingfirst the subgyral, and then the gyral sectors, and
39
only finalIy the lobar sector. The adoption of this system has allowed us to identify the precise sublevel of white matter that is involved by any lesion. Somewhat surprisingly, it has even forced us to revise our thoughts on the growth and spread of glial tumors in their initial and intermediate phases, as these lesions do not follow the connective fiber bundles, as generally believed, but rather grow in the direction of the gyral segmental vascular supply and in the reverse direction to that of the embryological neuroglial migration (see p. 12). Accordingly, a new perspective on the topographical anatomy of each of the cerebral lobes and their gyri, based on a sectorial five-tier hierarchical organization of the white matter, is presented here.
Topographical Anatomy01the Lobes and Gyri 01 theBrain
Boundaries. The frontallobe extends on the dorsolateral surface fromthe frontal pole to the superior and inferior precentral sulcUS. withoutclear separation from the precentral gyrus. It is separatedframthetemporallobe by the Sylvianfissure,and from the insulaby the superoanterior part of the cingular sulcus. On the medial surface,itisseparatedfrom the gyruscinguli(but not from the paracentral gyrus) by the sulcus cinguli. Mediobasally, the frontallobeis separated by the superior rostral sulcus from the subcallosal part of the gyrus cinguli. Basally, the entire orbital surfacebelongsto the frontallobe (For variations of the frontal gyri andsulci,seealsoFig.8.10-8.23,11.5-11.7in Ono et al. 1990). Surfaces. The frontallobe has, in practical terms, four main surfaces (inferomedial and superomedial, superolateral, operculoinsular,and fronto-orbital) and several sulcal surfaces. Figure 1.33showsthe surfacesof the frontal lobe. The frontallobe is separatedconvenientlyin its sagittal, fronto-occipital extension, into threeparts:anterior (level of the temporal pole), middle (level of the striatum), and posterior (level of the foramen of Monro). In coronal section five pedunculi can be identified; an anterior part of the cingulate gyrus, together with superior, middle, inferior,and orbital frontal gyri (Fig. 1.69A. a-e, p. 78). The cingulate gyrus, with its supracallosal and subcallosal parts, is topographicallyincorporated into the medial part of the frontal lobe. Thecingulategyrus will be discussed later as part of the limbic lobe.
Fig. 1.33 The surface of the frontal lobe (coronal section through the striatum) a, Superomedial (interhemispheric) surface a2 Inferomedial(interhemispheric) surface b Superolateral surface c Opercular-insular (Sylvian)surface d Fronto-orbitalsurface
40
1 Anatomy Sulciof the frontallobe. The followingsulci can be identified: Medial
Superolateral
Operculoinsular Fronto-orbit
Anterior part of the callosal sulcus Anterior part of the cingular sulcus Superorostral sulcus Inferorostral sulcus Parolfactory sulcus Superior sulcus Middle sulcus Inferior sulcus U nnamed (opercular) sulcus 5- 7 sulci Olfactory sulcus H -shaped orbital sulci (cruciform Rolando)
sulcus of
The (intermediate) middle frontal sulcus, occasionally single, and often interrupted, several times separates the gyrus into two or more subdivisions (secondary gyri). Classification of frontal gyri. 1 Superior frontal gyrus a Medial part b Superolateral part 2 Middle frontal gyrus 3 Inferior frontal gyrus a Orbital part b Triangular part c Opercular part 4 Gyrus rectus 5 Orbital gyri
Fig. 1.34 Frontal lo be gyri a MRI 01the lrontallobe (horizontal section). Note the superior frontal gyri (right and left), both with numerous symmetrical gyral interdigitations b Left cerebral hemisphere (superior view). An intermediate sulcus separates the superior Irontal gyrus (F1) into two longitudinal parts. Note the undulating course 01F2, and the connection 01F1 and F2 to the precentral gyrus (Pr, arrows) F1 Superior lrontal gyrus F2 Middle lrontal gyrus Pc Postcentral gyrus Pr Precentral gyrus e Left cerebral hemisphere (medial view). Note the semicircular medial lrontal gyrus around the anterior hall 01 the cingulate gyrus (between the two white arrows). The black arrow indicates the connection between the cingulate gyrus and the medial Irontal gyrus 1 Paracentral gyrus
There are three longitudinally oriented gyri (with a serpentine cours.e) that constitute the convex surface of the frontal lobe. 1. The superior frontal gyrus (F¡ with anterior, middle, posterior segments, Brodmann areas 6, 8 and 9) is one of the principal gyri of the voluminous frontal lobe. It is made up of a substantial parasagittal part, measuring approximately 1-2 cm in width and 11 cm in length (from the superior precentral gyrus to the frontal pole). Posteriorly, it has connections to the superior part of the precentral gyrus, and anteriorly it merges (within the frontal pole area) with the middle frontal, orbital, and rectus gyri. It has two parts, one supero lateral and one media!. Approximately 10 transverse gyri form an intertwining interface with the adjacent transverse gyri of the middle frontal gyrus. The subgyral anatomy is highly variable, as demonstrated by the 12-15 small extension arms of cortex merging with neighboring gyri (Fig. 1.34a-c). The superolateral part is often divided by one long, or numerous small, sulci in two longitudinal areas.
Topographic Anatomy for Microsurgical Approaches to Intrinsic Brain Tumors Themedial surface of the superior frontal gyrus is often given aseparatename-the medial frontal gyrus. (Brodmann area 6, 8, 9.10,11,12, and 32). It commences immediately anterior to the paraolfactorysulcus, and curves upwards, anteriorly between the and inferiorrostral sulci, towards the frontal pole (see superior Fig.1.53p. 57). It then turns posteriorly around the cingulate gyrus.11lecingulate sulcus separates this gyrus from the cingulate gyrus, butthere are individualvariations with one or more small orlargeconnective arms between the medial part of the superior frontalgyrus and the anterior part of the cingulate gyrus (see Figures1.57,1.58 and 1.59). Along the cingulate sulcus, several gyrifrom the medial, frontal gyrus and the cingulate transverse gyrusinterdigitatewith each other (Fig. 1.58). Many subgyri, dividedby axial sulci, mark the medial surface of this gyrus (Fig.1.34c).The frontopolar artery courses anteriorly within the superiorrostral sulcus. 2. The middle frontal gyrus (F2 with anterior, middle, and posteriorsegments), (Brodmann are as 6, 8, 9, and 10) is a more serpentine,longitudinal convolution. The secondary sulci (middle frontalsulcus) divide the gyrus into two or more sections. The middlefrontal sulcus can be regularly recognized on coronal MRl.The F2gyrus has connective arms to F¡ and F3 on the surface.Its removal results in the appearance of a "valley ringed by The interlockingpaUern of the juxtaposed gyri is tallmountains." evident.It is separated from the inferior frontal gyrus by the inferiorfrontalsulcus,whichcontainsportions of the prefrontal arteriesthatdipinto the sulcusas they ascend over the lateral surface ofthehemisphere.This sulcus sometimes appears as an anteriorlydirected branchofthe precentralsulcus(Fig.1.34d, e). It has regularlyone,andoccasionallytwo,connectivearms to the precentral gyrus,which cause the interruption of the precentral gyrus into t\Voor three sections. Along the superior and inferior frontal sulci,the middle frontal gyrus extends several transverse gyri (see Fig.1.34e). 3. The inferior frontal gyrus (F3 with anterior, middle, and posteriorsegments, Brodmann areas 44, 45, and 47), located betweenthe inferior Sylvian fissure and the antero-superior lip of theSylvianfissure, forming the frontal operculum, overlying the anteriorhalf of the insula. The frontal operculum is divided into three parts between theanteriorandascendingrami of the Sylvianfissureand the inferior precentralsulcus: a. The orbital part (Brodmann are a 47) merges anteriorly \Viththe lateral orbital gyri, superiorly with the anterior parts of F2'andposteriorly with the triangular part. b. The triangular part (Brodmann area 45) lies between the anteriorand ascending rami of the Sylvian fissure, and merges anteriorlywith the orbital part, posteriorly with the opercular part,and also has connective arms to the middle parts of F2. The triangularpart is easy to recognize on anatomical section and on MRIpictures(Naidich 1991). c. The opercular part (Brodmann 44) lies between the ramiof the Sylvianfissure and the inferior precentral ascending andinferiorfrontal sulci, and merges anteriorly with the triangularpartand posteroinferiorly with the inferior part of the precentralgyrus.The opercular part of F3coincides with Brodmann area 44.but some authors also inc1ude area 45. In anatomical and functionalpublications, the location and extension of the Broca areais veryapproximate;the encirc1edareas on drawingsare irregular,usuallyextending into the adjacent F2and inferior part of precentral gyrus.
41
Fig. 1.34 d The left superior (F1), middle (F2), and inferior (F3) frontal gyri (anterolateral view). Note the connection between the voluminous middle frontal gyrus and the precentral gyrus (arrow) F1 Superior frontal gyrus Pc Postcentral gyrus F2 Middle frontal gyrus Pr Precentral gyrus F3 Inferior frontal gyrus
Fig.1.34e The transverse gyri of the superior frontal gyrus (F1) and inferiorfrontalgyrus (F3) are wellvisualized after removalof the middle frontal gyrus (arrows)
Broca's are a (well-known since 1861) fascinates neuroscientists as a reliable functional center. Interestingly, the individual anatomical variations of this area have not been thoroughly investigated; this part may consists of one voluminous convolution with a superficial short sulcus on the surface, or it may comprise two or even three gyri displayed parallel to the inferior part of the precentral gyrus (Fig. 1.34f-i). The variations of the inferior frontal and inferior precentral sulci, and of the anterior and ascending rami of the Sylvian fissure, are described in ano et al. (1990). As the sulcal variations result from gyral developments, the individual variability of the frontal operculum is obvious. In addition to these variations (which can be easily recognized on the surface of the gyral configuration), there are also more hidden variations within the sulcal depth, with interdigitations (transverse gyri) into the inferior frontal sulcus, into the inferior precentral gyrus, and especially into the Sylvian fissure, where five or six short or long transverse interdigitations (3-4 cm) extend lateromedially from
42
1 Anatomy
Fig.1.341 The left inferior frontal gyrus (F3). Note the gyral eontinuity betweenthe opereular part of F3 and the middle frontal gyrus (F2) and preeentralgyrus (Pr,arrows) F2 Middle frontal gyrus F3 Inferiorfrontal gyrus Op Opereular gyrus Orb Orbital gyrus Pe Posteentralgyrus Pr Preeentralgyrus Tri Triangularpart of the inferior frontal gyrus
"'
~
'":
.
.
/'
r '
'.
.
.
.
~
.
F'. ""
,
.
';"
'.
.
'>i
;'.
. .¡$'".¡
/.
~.
.;1ft"'''' "
...
'.
,
~, f .
... .
~
"
/
.
~.
"
.
.'
.
';
.
~
.
.~
.r"
-
' ..
. ""
.
..
..
.
.'
.
:} '
..
.
.
'
....
Fig. 1.349 A variation of the inferior frontal gyrus (F3). The preeentral gyrus is not interrupted F1 Superior frontal gyrus F2 Middle frontal gyrus F3 Inferior frontal gyrus Pe Posteentral gyrus Pr Preeentral gyrus
.
.
'1
,t
~.
"
ij<.
l
.
.
Fig. 1.34h A very rare variation of the middle and inferior frontal gyri F2 Middle frontal gyrus F3 Inferiorfrontal gyrus Pr Preeentralgyrus
the Sylvian fissure deep into the superior part of the cingular sulcus. (Figs.1.35, 1.48b). Two-thirds of the gyral extensions are therefore not visible from surface inspection or electrophysiological studies (Fig. 1.34i, j). (For variations, see Figs. 8.15-8.19, 11.8, 15.4-15.6 in Ono et al. 1990.)
Fig. F3 Op Orb Tri
1.34i A variation of the inferior frontal gyrus Inferior frontal gyrus Opereular gyrus Orbital gyrus Triangular part of the inferior frontal gyrus
Topographic Anatomy for Microsurgical Approaches to Intrinsic Brain Tumors
Fig.1.34j A specimen lollowing removal 01the opercular and triangularportions 01the inlerior lrontal gyrus. Note the transverse gyri 01 the middleIrontal gyrus
43
Fig.1.34k The inlerior lrontal gyrus has been completely removed. Note the depth 01the underlying insular gyri (arrow)
"Flg.1.35 The Sylvian lissure widely opened in a rubberized left cere- 1> bral hemisphere, revealing the insular surfaces 01 the lrontal, central, parietal, and temporal lobes. The transverse gyri 01 these lobes (arrows)extend into the cingular sulcus 01 the insula (el. Fig. 1.51 b, p.54)
4. The gyrus rectus (Brodmann area 11) is the medial basal par!,measuring approximately 5 mm wide and 5 cm long. This long.narrow convolution is the least variable of all the cerebral gyri. The deep olfactory sulcus, covered on the surface by the olfactory tract, divides the gyrus rectus from the orbital gyri (Fig.1.36). 5. The orbital gyri (Brodmann areas 11 and 47) are found lateralto the gyrus rectus and olfactory nerve. Small sulci in an "H-shaped"pattern divide the orbital gyri into four areas; the anterior,posterior, medial, and lateral gyri. These four gyri are
{
I
'¡"rt
~,~
similar in size and shape.They are connectedwith the superior, middle, and inferior frontal gyri around the polar area. The posterior orbital gyrus lies anterior to the anterior perforated substance, and connects via the uncinate fasciculus with the anterior insula. The basal orbital surface is concave from side to side, and rests on top of the cribriform plate of the ethmoid, the orbital surface of the frontal bone, and the two wings of the sphenoid bone. Figure 1.36a-d shows some variations in the orbital gyri. (Other variations are seen in Fig. 11.1-11.4 in Ono et al. 1990.)
44
1 Anatomy Fig. 1.36a-d Architecturalvariations of the orbital gyri and gyrus rectus. The olfactory tract (O)has been removed in a-c, but is indicated in d. Note that, in spite of the considerable variation,the orbital gyri presents a consistent pattern of four parts: superior, inferior, medial, and lateral I Inferiororbital gyrus L Lateral orbital gyrus M Medialorbital gyrus O Olfactory tract R Gyrus rectus S Superior fronto-orbitalgyrus
/
.
\
.¿ I d
Central Lobe
The precentral, postcentral, and paracentral gyri together constitute an ellipse between the frontal and parietallobes that we term the central lobe. Recent evidence lends support to the concept that such a unit has merit, being reciprocally integrated embryologically,developmentally, and functionally (Uematsu 1992). The precentral, postcentral, and paracentral gyri have always been mentioned separately in neuroanatomy, neuroradiology, neurosurgery, and neurophysiology, and are rarely mentioned as a posterior part of the frontal or anterior part of the parietallobe. The precentral and postcentral gyri have an obliquely diagonal perpendicular direction with a less serpentine outline than other gyri. Each gyrus gives off three prominent extension arms, which connect anteriorly with the superior, middle, and inferior frontal gyri, posteriorly with the parietal lobules, and superiorly with the middle frontal and paracentral gyri (Fig. 1.37). (See also Figs.8.1-8.9,17.1 in Ono et al. 1990). Suñaces. The centrallobe has three main surfaces (medial, dorsal, and opercular-insular), and multiple sulcal surfaces (such as the precentral, central, postcentral, and cingular). Sulci of the centrallobe. There are five sulci associated with the centrallobe: the precentral, central, postcentral, cingular, and
paracentral. Like other lobes, the centrallobe has incomplete separations from the frontallobe (marked by the interrupted precentral sulcus), from the parietallobe (marked by the postcentral sulcus-interrupted in 44%), from the temporallobe (marked by the Sylvian fissure), from the insula (marked by the superoposterior part of the cingular insular sulcus), and from the cingulate gyrus (marked by the ascending part of the cingular sulcus) (Fig. 1.38). The precentral sulcus is regularly interrupted by two or three superficially localized extension arms and by five or six transverse gyri. The central sulcus usually has no interruptions on the dorsal surface (only 8%), but the superior and inferior ends are always closed by gyral connections. In the depth of a sulcus, there are prominent transverse gyri. The postcentral sulcus is interrupted in 44% of cases by two superficial connection arms (superior and inferior) on the surface, and several transverse gyri in the depth of the sulcus (Figs. 1.39, 1.40). Coronal sections show that five pedunculi can be identified making up the white matter that constitutes the centrallobe: the cingular (medial portion), paracentral, superior, middle, and infe-
rior (Fig. 1.69A. d, e).
Topographic Anatomy for Microsurgical Approaches to Intrinsic Brain Tumors
45
Flg. 1.37 The centrallobe a Superiorview 01 the central lobes 01the two hemispheres. Note the asymmetrical architecture 01the precentral gyri, with two connections to the superior and middle Irontal gyri (arrows) on t.le left side, and one to the superior lrontal gyrus on the right (arrow) Pc Postcentral gyrus Pr Precentral gyrus b Superior view 01 the central lobe. Again, note the asymmetry Pc Postcentral gyrus Pr Precentral gyrus e, d Superior view 01 the right centrallobe. Note the variations in the communications between the precentral and postcentral gyri and the lrontal and parietallobes, respectively. The superior frontal gyrus is separated by the interme-
a
b
diatesulcus(d) F1 SuperiorIrontalgyrus Pc Postcentral gyrus Pr Precentral gyrus e MRI01thecentrallobe (horizontal section). Note the sym-
metricalpaUernof the right and leftprecentraland postcentral gyriinthispatient
e
e
d
46
1 Anatomy
Gyri of the centrallobe. The paracentral gyrus (the para-medial part of Brodmann areas 1, 2, 3, 4, and 5) is a wide quadrangular structure that encompasses both the precentral and the postcentral gyri on the medial hemisphere. It is divided anteriorly from the medial part of the frontal gyrus by the paracentral sulcus, and fram the precuneus by the upward limb of the cingulate sulcus (referred to as the marginallimb or ramus). A shallow sulcus may
Fig. 1.38 The left cerebral hemisphere (superomedial view) 8 The paracentrallobule is well demarcated, with gyral connections to the precentral and postcentral gyri, and to the medial part of the superior frontal gyrus (arrows)
The preeentral gyrus (Brodmann area 4). The precentral gyrus begins approximately at the junction of the proximal and middle third of the posterior ramus of the Sylvian fissure, and follows an oblique vertical course toward the interhemispheric fissure. The precentral gyrus was called aseendent pli frontal by Gratiolet, "anteroparietal gyrus" by Huxley, "ascending frontal gyrus" by Turner, and "anterior central gyrus" by Huschke. It is continuous with the paracentral lobule on the mesial surface, which bridges the mesial end of the central sulcus. It is also ~ontinuous with the postcentral gyrus. The precentral gyrus is partially separated from the three horizontally directed frontal gyri by the precentral sulcus, which regularly contains the precentral
incompletelydividethis into upper and lower halves of equal propartions. Its relations to the precentral and postcentral gyri are discussed above. The paracentral gyrus occasionally connects with an arm to the cingulate gyrus (Fig. 1.38). The paracentral gyrus has three or four t.ransverse gyri hidden within the cingulate gyrus. (For variations, see Figs. 13.1, 13.2, 13.4-13.6, 13.11, 13.12 in Ono et al. 1990.)
b b The same specimen after removal of the paracentrallobule, showing a partially arbitrary borderline
arteries of the MCA. This sulcus is usually divided at its midportion by a ridge of tissue that connects the precentral gyrus to the middle frontal gyrus. The precentral gyrus is also regularly connected inferoanteriorly with inferior gyri, inferoposteriorly with the postcentral, and superiorly with the paracentral gyri. The posterior limit of the precentral gyrus is the central sulcus, containing the central arteries. Within the central sulcus, three or four interdigitations (transverse gyri) are directed towards the postcentral gyrus, and vice versa. This gyrus is one of the broadest convolutions of the hemisphere, measuring approximately 9-15 mm in width and 10-12 cm in length. (Figs. 1.39, 1.40, 1.48, 1.51b.)
b Fig. 1.398, b The left centrallobe, between the white arrows (superolateralview). Note the variations in the precentral and postcentral gyri
Pc Postcentral gyrus Pr Precentral gyrus
Topographic Anatomy tor Microsurgical Approaches to Intrinsic Brain Tumors Thepostcentralgyrus (Brodmann areas 1, 2, 3, and inferiorly Brodmann43) is located parallel to the precentral gyrus, separatedfromit by the central su1cus(92% uninterrupted). Posteriorly,it isseparated from the parietallobe by the postcentral su1cus (46% uninterrupted). Its mesial extension into the paracentral lobuleis through a small isthmus which is not as pronounced as that ofthe precentralgyrus.The suprasylvianlimb of the supra-
marginalgyrusmerges with the base of the postcentral gyrus, preventingtheinferiormarginof the postcentral su1cusfrom reach-
47
ing the Sylvian fissure. A shallow su1cussometimes marks this point (posterior subcentralsu1cus)(Figs. 1.39,1.40).One to three connective arms posterior to the superior middle and inferior parietallobules divide the postcentral gyrus (in 44%) into superior and inferior parts. There are also numerous interdigitations (transverse gyri) in the depths of the central and postcentral su1ci and the Sylvian fissure (Fig. 1.48). Figure 1.15, p.21 shows a unique variation between the precentral and postcentral gyri.
b Fig.1.40a The right centrallobe, between the white arrows (superolateralview).Notethe significantly different patterns in both the size and theconnections01the precentral and postcentral gyri
Parietal
Lobe
Surfaces. The parietallobe has three main surfaces (medial, dorsal.andopercular-insular) and numerous su1calsurfaces.
Parietallobesulci. The parietallobe is associated with seven su1ci: the parielo-occipital, cingular, postcentral, interparietal superior, intermediate,and subparietal. On the lateral aspect of the parietal lobe.there are several semicircular, serpentine gyri, which have connectingarms to the postcentral gyri and to the superior temporalandoccipital gyri (Figs. 1.42-1.44). (For variations, see Ono el al. 1990,Figures 9.1-9.12,15.7-15.10,17.4.) Five parietal pedunculi can be identified: the cingular, precuneal,superior lobule, middle, and inferior lobule (Fig. 1.69A. f). The superior, middle and inferior parietal pedunculi arewellvisualizedon coronal MR images (Fig. 1.69B. f'). Parietallobules. The parietal gyri have a more pronounced serpenUneshape.and are therefore termed lobules. On coronal section, there are four lobules. The square-shaped precuneus (Brodmann areas7 and 31) constitutes the medial surface of the parietallobe, and lies between the ascending two or three rami of the cingulate
Fig. 1.40b The right centrallobe, between the white arrows (superolateral view).
gyrus and the superior parietallobule. It is formed by three substantial extension arms from the cingulate gyrus: an anterior, middle, and a posterior arm (though the middle arm may sometimes originate from either the anterior or the posterior arm) (Fig.1.41). (For variations, see Figures 13.1, 13.2, 13.5-13.12 in Ono et al. 1990.) The posterior arm runs in continuity with the isthmus of the cingulate gyrus to the origin of the medial temporooccipital (lingual) gyrus,separated from it by the anterior part of the ca1carinesu1cus(Fig.1.41). In a superior direction, the anterior gyrus of the precuneus is connected with the postcentral gyrus, and the middle and posterior arms of the precuneus are connected with the superior parietal lobule. The cingulate su1cus is not well-defined posteriorly, and a transverse suprasplenial su1cus (subparietal su1cus) divides this from the posterior portion of the cingulate gyrus. The parieto-occipital fissure (100% uninterrupted) divides the precuneus from the cuneus. Although no surface arms connect the precuneus and the cuneus, there are three or four crossing extension arms (transverse gyri) which do provide connections in the depths of the sulcus.(For variations of gyri and sulci, see Figures 9.1-9.12, 15.9-15.10; 17.1-4 in Ono et al. 1990.)
48
1 Anatomy
~
A
a
~ I
r\'« ""
.
""
.
Fig. 1.41 The right precuneus lobule (medial view) a Notethe degree of continuity with the cingulate gyrus
a
b
e
~
~' b The same specimen. after removal of the precuneus lobule. A totally arbitrary borderline with the cingulate gyrus has been created
The superior parietallobule (Brodmann areas 5, 7) is a complex lobule lying between the intraparietal sulcus and the interhemispheric fissure. It is continuous with the precuneus on the mesial surface (Fig. 1.41 a). Its complex appearance is caused by a folding of the gyrus forming two U-turns, that can be called the anterior and the posterior parts. It connects anteriorly with the postcentral gyrus and superiorly with the connective arms of precuneus. The complex morphology of the superior parietallobule is seen in Figures 1.42a and 1.43a). The interdigitations (transverse gyri) between the postcentral gyrus and the superior and inferior parietal lobules and the postcentral sulcus are shown in Figure 1.44. The middle and inferior parietallobule. The inferior parietal lobule (Brodmann areas 39 and 40) is divided into two parts, an inferior lobule (supramarginal or circumflex) and a middle lobule (angular gyrus). Their irregular shape is formed by the terminations of three large lateral sulci; the intraparietal sulcus, the Sylvian sulcus (fissure) and the superior temporal sulcus. The inferior (supramarginal) lobule encircles the end of the posterior ramus of the Sylvian fissure. Its suprasylvian limb is continuous with the postcentral gyrus, and its infrasylvian limb is continuous with the superior temporal gyrus. The middle (angular) lobule lies posterior to the inferior lobule, and covers the incursion of the superior temporal sulcus into this region. Posteriorly, the inferior parietallobule merges with the occipitallobe (Figs. 1.42, 1.43). There are numerous intersulcal hidden interdigitations (transverse gyri) between adjacent gyri and in the posterior part of the Sylvian fissure (Fig. 1.51b). The interdigitations between the superior and inferior lobules and the postcentral gyrus are well indicated in Figure 1.44b.
<J Fig. 1.42 The left cerebral hemisphere (superolateral view) a The divisions of the parietallobule (encircled by dots) P1 Superior parietallobule (Brodmann area 7) P2 Middle parietallobule (Brodmann area 39) P3 Inferior parietallobule (Brodmann area 40) b The same specimen, after removal of the superior parietallobule P2 Middle parietallobule (Brodmann area 39) P3 Inferior parietallobule (Brodmann area 40) e The same specimen (superior view). Note the multiple intrasulcal extensions to neighboring gyri. The transverse gyri are well visualized
.....
Topographic Anatomy ter Microsurgical Approaches to Intrinsic Brain Tumors
Fig.1.43a Theleft parietallobe (lateral view). The white arrows indicatetheinterparietalor intraparietal sulcus P1 Superiorparietallobule (Brodmann area 7) P2 Middleparietallobule (Brodmann area 39) P3 Inferiorparietallobule (Brodmann area 40)
49
b After removal of the middle parietallobule (P2), the transverse gyri are will visualized
....
'
/'
i\/'"i
/
~
\.
'\
1.
-~ )
~ Á..
\ ~-~
Fig.1.44a A posterolateralview of the parietallobe. The postcentral sulcusandintraparietalor interparietal sulcus are seen (arrow) P1 Superiorparietallobule (Brodmann area 7) P2 Middleparietallobule (Brodmann area 39) P3 Inferiorparietallobule (Brodmann area 40)
b The postcentral sulcus, with the transverse gyri (arrows) in the depths of the sulcus
OccipitalLobe
Within each su1cus, there are numerous short or long transverse gyri, especially within the ca1carine sulcus. This infolding allows a great increase in the surface area of the occipitallobe, resulting from the functional development of macular vision (and the visually-related association cortex). The occipital gyri are morphologically the most complex of the cortical gyri. Six gyrals pedunculi can be identified (Figs. 1.66 on p. 73 and 1.69A. g on p. 78). (See also Figs. 9.13, 9.14 in Ono et al. 1990):
Surfaces. Theoccipitallobe hasfour surfaces,the medial, dorsal, andfissural(ca1carine),and numeroussu1calsurfaces. basaL Occipitalsulci. The occipital lobe is associated with one fissure andfivesu1ci:the ca1carinefissure and the parieto-occipital, transverseoccipital, lateral occipital, lunate, and anterior occipital sulci.(Forvariations,seeKubik in Vol. III A, p. 350-68, and Ono etal.1990Figs. 14.1-14.16). An anterior occipital su1cusseparates theoccipitallobefrom the temporallobe (Fig. 1.45a-c). It usually liesslightlyposterior and parallel to a line drawn between the top of the parieto-occipital fissure and the occipital notch. Two or threelateraloccipital su1cidivide the lateral occipitallobe into secondarysu1ci.The lunate su1cusis often present just anterior to the occipitalpoleo Small subgyri may form extensive infolding pattemsdeep in the occipital su1ci. There are many axial su1ci and oneoperculated su1cus (lunate su1cus) in this area of the lobe.
1. Medial: cuneallobule 2. Mediobasal: gyrus tempero-occipitalis medialis (posterior part). MTO 3. Laterobasal: gyrus temporo-occipitalis lateralis (posterior part). LTO 4. Dorsal: superior occipital gyrus or lobule 5. Dorsal: medial occipital gyrus or lobule 6. Dorsal: inferior occipital gyrus or lobule
I
50
1 Anatomy Fig. 1.45a-c The left occipitallobe (lateral view). The lateral surface has three parts, the superior (1), middle (2), and inferior occipital (3) gyri. Note the numerous transverse gyri 01 Superior occipitallobe 02 Middle occipitallobe 03 Inferior occipita~ lobe
I
e
The cuneus (Brodmann areas 17, 18 and 19) is a medially welldemarcated lobule, lying between the parieto-occipital fissure and the calcarine sulci, and merging at the medial occipital pole with other occipital gyri. These sulci converge anteriorly near the isthmus of the cingulate gyrus. A tongue of cuneus projects deep beneath the precuneus in the parieto-occipital fissure. These gyri are sometimes called the cuneolingual gyri. The parieto-occipital branches of the PCA mark the fissure, while the calcarine arteries course within the calcarine fissure (Fig. 1.46). In this fissure, and also in the parieto-occipital sulcus, there are 6-10 transverse gyri. (See Fig. 14.1-14.6 in Ono et al. 1990.) The lingual gyrus (inferior part of Brodmann areas 17, 18, and 19) or medial temporo-occipital gyrus, is the other structure comprising the medial occipital poleo This is a long, tongue-like structure that sits above the tentorium cerebelli and forms the medial basal portion of the occipitallobe. It lies between the calcarine fissure and the collateral sulcus on the inferior surface of the temporo-occipitallobe. The collateral sulcus frequently contains the inferolateral trunk of the PCA, which gives rise to the anteroinferior, inferomedial, inferoposterior, and occipitotem-
poral branches. Anteriorly, this gyrus becomes continuous with the cingular gyrus (parahippocampal gyrus) near the isthmus (Fig. 1.47). (For variations, see Fig. 14.6-14.14 in Ono et al. 1990.) The transverse gyri within the collateral sulcus are short and small. The small superior middle and inferior occipital gyri (the lateral parts of Brodmann areas 17, 18, and 19) are divided by the lateral (or inferior) occipital sulcus. The middle occipital gyrus merges superiorly with the cuneus, while the inferior one merges laterally with the lateral occipitotemporal gyrus, and all of these merge at the occipital pole with the cuneus, MTO, and LTO gyri (Fig. 1.47). Again, numerous transverse gyri are found within each gyrus. The fusiform gyrus (lateral temporo-occipital gyrus, see page 52).
Topographic Anatomy tor Microsurgical Approaches to Intrinsic Brain Tumors
51
b
Fig.1.46 Right occipitallobe (posteromedial view) a Thecuneus is clearly defined by the parieto-occipital sulcus anteriorlyand the calcarine fissure inferiorly, but not laterally b The same specimen after removal of the cuneus. Note the transversegyri of the precuneus within the parieto-occipital sulcus (arrows) e The transverse gyri of the left lingual gyrus within the calcarine fissure (arrows)
Fig.1.47 The left temporo-occipital gyrus (inferomedial view). The lateraltemporo-occipital gyrus has been subtotally removed, and the occipital horn has been opened (arrow) LTO Lateral temporo-occipital (fusiform) gyrus MTO Medial temporo-occipital (lingual) gyrus P Parahippocampus U Uncus
e
52
1 Anatomy
Temporal Lobe Suñaces. The temporallobe has four main surfaces: the opercularinsular, superalateral, basal, and medial, as well as numeraus sulcal surfaces. Temporallobe suld. The temporallobe is associated with seven or eight sulci: the superior, middle, inferior temporal, angular, hippocampal, uncal, rhinal, and collateral sulci. Temporal gyri. Five temporal gyri (five temporal gyral pedunculi) can be identified on coranal MRI (Fig. 1.69): the superior, middle (between the middle and inferior temporal gyrus lies an interrupted sulcus), inferior, fusiform or lateral tempora-occipital (anterior two-thirds), and parahippocampal gyri. The superior temporal gyrus (TI with anterior, middle, and posterior segments, Brodmann areas 22, 38, 41 and 42) lies immediately inferior to the Sylvian fissure and forms the temporal operculum overlying the insula. This is a highly convoluted ribbon of cortex that merges with the supramarginal gyrus at the posterior end of the fissure. Approximately five to eight small gyri extend perpendicular to the long axis of the gyrus, into the depths of the adjacent superior temporal sulcus. Three or four short transverse gyri extend fram the anterior part of the superior temporal gyrus into the anterior Sylvian fissure, towards the inferior insular sulcus. One voluminous gyrus, or two parallel gyri five to six centimeters long, sometimes divided by a sulcus (the "transverse temporal gyri or Heschl gyri") run fram the posterior part of the superior temporal gyrus diagonally into the deep end of the Sylvian fissure. At this point, there is a ring portion surraunding the end of the transverse gyrus. The transverse gyrus faces very close to the posterior gyri of the insula (Fig. 1.48a-b). The superior temporal sulcus, consistentIy present in about a third of individuals, begins at the lateral aspect of the temporal pole and parallels the Sylvian fissure. It ends at the angular gyrus. This sulcus divides the superior fram the middle temporal gyrus (Figs. 1.17e, 1.45a). In 70% of cases, there are connections on the surface between the superior and middle temporal gyri (Fig. 1.15, p.21) and numerous interdigitations within the superior temporal sulcus (Fig.1.49b-c). The middle temporal gyrus (T2with anterior, middle, aÍ1dposterior segments, (Brodmann areas 28, anteriorly 38) is a horizontal gyrus with more serpentine parts, separated fram the superior temporal gyrus by the superior temporal sulcus and fram the interior temporal gyrus by the irregular mjddle temporal sulcus. This sulcus is incorrectIy called the inferior temporal sulcus. The upturned posterior end of the middle temporal sulcus parallels the superior temporal sulcus and ends in the posterior portion of the inferior parietal lobule. Unlike the superior temporal gyrus, which has a more unfolded form, the middle temporal gyrus infolds upon itself (similar to an accordion), thus enabling this large, narraw gyrus to occupy a small volume. Its anterior pole is encircIed by the anterior ends of the superior and inferior temporal gyri. There are connections on the surface to the superior and inferior temporal gyri, and numerous interdigitations (transverse gyri) in the depths of the sulci (Fig. 1.49b). The inferior temporal gyrus (T3 with anterior, middle, and posterior segments, (Brodmann areas 20, 37, anteriorly 38, and inferiorly 37) lies below the middle temporal sulcus and forms the narrow longitud inal convolution on the lateral inferobasal cerebral convexity.This gyrus is markedly smaller in width and length than the other temporal gyri. The inferior temporal sulcus sepa-
,
I
I
5
,
2 ......-)
.... ,.
t 1
r
Fig. 1.48a The left Sylvian fissure has been slightlyopened to show the course of the transverse gyrus, originating fromthe middle portion of the superior temporal gyrus (arrow)and traversing obliquely intothe depth of the posterior Sylvianfissure b Note the other gyral digitations posterior to the transverse gyri (Heschl) 1 Polar planum 2 Anteriortransverse gyrus 3 Intermediate transverse gyrus 4 Posterior transverse gyrus 5 Temporalplanum
rates the inferior temporal gyrus fram the lateral tempora-occipital gyrus, in which small and short transverse gyri are found (Fig. 1.49e). There are connective arms in the anterior part to the parahippocampal gyrus, and in the middle and posterior parts to the lateral tempora-occipital gyrus, as well as numeraus interdigitations in the sulci. The fusiform gyrus (lateral tempora-occipital gyrus, Brodmann are as 20 and 37) lies lateral to the parahippocampal and lingual gyri, which are separated by the collateral sulcus, and has a lazy, curving formo Its basal surface assumes a slight concavity as it conforms to the shape of the floor of the middle cranial fossa and tentorium cerebelli. Anteriorly, it merges into the inferior temporal gyrus, and posteriorly into the medial temporo-occipital gyrus. The interdigitations (transverse gyri) within the collateral sulcus are small, short, and flat (Fig. 1.50). The parahippocampal gyrus (see limbic lobe page 61).
Topographic Anatomy for Microsurgical Approaches to Intrinsic Brain Tumors
53
b
Fig.1.49 Therighttemporallobe a Alter removal01 the superior temporal gyrus (lateral view). The arrowsindicatethe transversegyri 01M2 b Alterremoval01the middle temporal gyrus {lateral view). Note the transverse gyri01the superior and inferior temporal gyrus (T1.T3) e Alterremovalof the inferiortemporal gyrus (inferior view). Note the transversegyri of T2 and the lateral temporo-occipital gyrus Iying between thesegyri and the surrounding gyri (arrows)
e
Fig.1.50 The right temporo-occipital lobe (inferior view). The lateral temporo-occipital (fusiform)gyrus has been removed. Note the gyral connectionsbetweenthe parahippocampal gyrus (arrow 1) and the Inferiortemporalgyrus, and between the isthmus cinguli (arrow 2) and themedialtemporo-occipitalgyrus MTOMedialtemporo-occipitalgyrus P Parahippocampus T Inleriortemporalgyrus
InsularLobe Theínsulais an invaginated set of gyri forming the floor of the Syl\ian fissure. Adequate visualization requires retraction of the frontal,frontoparietal, and temporal operculi overlying this structure. Its triangular shape resembles and unfolded fan or in verted pyramid,with the limen insulae forming the apex. There are five radially directed arms. The anterior three gyri are smaller and shorter (gyri brevi) than the two posterior (gyrus longus). These areseparatedfram each other by the precentral, central, and post-
central insular sulci, on which the adjacent transverse gyri of the inferior frontal, precentral, and postcentral gyri and inferior parietal lobule rest. The insula is the cortical covering of the claustrum and putamen of the lentiform nucleus. It is surrounded by the inferior and superior insular sulci (circular sulci). It is intimately related to the middle cerebral artery (Fig. 1.52a-e). The triangular insula is surrounded by a circular sulcus which has a superior horizontal part under the frontal and parietal oper
54
1 Anatomy
culum, a diagonal inferior part under the temporal operculum, and an anterior part at the level of the limen insulae (where the bifurcation of the MCA is usually located). The insular gyri extend anteriorly and posteriorly between the interdigitations (transverse gyri) of the F3,pre- and postcentral, and P3gyri, which are hidden underneath the operculum. The accessory insular gyri are connected with the orbital part of F3, and the transverse insular gyri with the lateral olfactory are a (Fig. 1.51b). The insula
receives its arterial blood supply from the M2 and M3 branches, which are numerous (30-40) tiny arteries. They supply the insula, the capsula extrema, and the claustrum, but not the putamen. The venous blood may drain into the cingular, deep Sylvian, or superficial vein. The anteroinferior part of the insular gyri has a peripaleocortical structure, the middle part a proisocortical structure, and the posterior part an isocortical structure (Fig. 1.51c).
Fig. 1.51 a The left insula (lateral view) after removal of the frontoopercular, centro-opercular, parieto-opercular, and temporal areas (anatomical specimen prepared by Dr. U. Türe) 1 Anterior short gyrus 2 Middle short gyrus 3 Posterior short gyrus 4 Anterior long gyrus 5 Posterior long gyrus
Fig. 1.51b The leftinsula (superolateral view)ina rubberized leftcerebral hemisphere 1 Transverse insular gyrus 2 Accessory insular gyrus 3 Short anterior gyrus 4 Short middle gyrus 5 Short posterior gyrus 6 Long anterior gyrus 7 Long posterior gyrus 8 Transverse gyrus (suborbital) 9 Transverse gyrus (subtriangular) 10 Transverse anterior and posterior gyri (subopercular) 11 Transverse gyri (subprecentral) 12 Transverse gyri (subpostcentral) 13 Anteriorparietal transverse gyrus 14 Middleparietal transverse gyrus 15 Posterior parietal transverse gyrus 16 Temporalplanum 17 Anteriortemporal transverse gyrus (Heschl) 18 Posterior temporal transverse gyrus (Heschl) C Central sulcus of insula
1 2 3 4 5
<J Fig. 1.51e The cytoarchitecture of the insula (fromMesulam and Mufson in Jones and Peters, eds.; Cerebral Cortex, vol. 4, New York:Plenum Press, 1985, pp. 179-226) 1 Archicortex 2 Peripaleocortex 3 Proisocortex 4, 51socortex
Topographic Anatomy tor Microsurgical Approaches to Intrinsic Brain Tumors
55
Fig 1.52a The right insula and the branches 01the middle cerebral artery (MCA) after removalof the opercular areas
a
b The left ínsula and the temporal operculum, with the arterial branches of the inferior trunk of Ihe MCA. Note the depth of the sulci in the frontal, central, and parietal opercular areas (ana¡omlcal specimens in Fig, 1.52 b-e prepared by Dr. U. Türe)
b
56
1 Anatomy
Fig. 1.52c The left cerebral hemisphere (lateral view) after removal of the opercular areas and insula 1 Internal capsule 2 Superior longitudinal fasciculus 3 Arcuate fasciculus, between the areas of Broca and Wernicke 4 Inferior fronto-occipital fasciculus 5 Uncinate fasciculus 6 External capsule 7 Putamen
d After removal of the left insula and the lentiform nucleus, the lenticulostriate branches of M1 are well seen.
e Magnified views of the areas with lenticulostriate arteries. Note the course the arteries take into the internal capsule
Topographic Anatomy for Microsurgical Approaches to Intrinsic Brain Tumors
57
Table 1.7a Connective-fiber system of the limbic lobe e uniquelimbiclobe bridges the phylogenetie gap between the newerneocerebralareas and the primitive central are as. The anatmist subdividesthe limbic lobe into cortical and nuclear areas. e corticalpart of the limbic lobe comprises the olfactory and hippocampalareas and the parahippocampal and cingulate gyrus. e cortexof the limbic lobe displays a variability in its complex'tyandlayeringfram three to five. Because of the complex folding f the cerebral hemispheres, the cortical areas of the limbic lobe ccupythe medialand basal surface of the brain in two circles, boveand beneath the callosum. The anatomy of the limbic lobe hasbeen comprehensively studied in numerous scientific papers nd monographs (McLean 1949-1973, Nauta 1956-1972, tephan1975,Akert 1985, Zilles 1987, Duvernoy 1988, Niewenuyset al. 1988,Holstege 1990). Figures 1.53a-b and 1.54a-d and 'able1.7a, b review the basic format, and should stimulate a teviewof limbiclobe anatomy in detail (see also Fig. 2.12 on page 37).
Cingulum Fornix Longitudinal striae Anterior eommissure Hippoeampal eommissure Genu and rostrum 01the corpus eallosum Uneinate laseieulus Posterior eommissure Stria terminalis Stria medullaris Ansa peduneularis Mamillothalamie traet (Vieq d'Azyr) Mamillary pedunele Diagonal band 01 Broca Habenulointerpeduneular traet (Meynert) Dorsal longitud inal traet (Sehütz) Medial lorebrain bundle
Ig 1.53a The right cerebral hemisphere. ediobasaltemporal areas, with the temporal le. Largearrow: the connections 01cingulate yrusto the paracentrallobule and precuneus Paraterminalgyrus Subcallosal gyrus Parollactory area i.pol Cingulate pole, with connection to the gyrus rectus and medial frontal gyrus (small
arrows)
TO Medial temporo-occipital (Iingual) gyrus Parahippocampus
Uncus
s,
. i = superior
and inferior rostral sulci
'g. 1.53 b "The medial aspect of the right rebral hemisphere (from Duvernoy, The 'uman Brain, Vienna: Springer, 1991, p. 39, ~ig.20) 1 Rhinal sulcus 2 Gyrus ambiens 2' Uncal notch 3 Semilunar gyrus 4 Endorhinal sulcus
a
5 Gyrus uncinatus 6 Uncal sulcus
" ""' ---1",. '"
7 Parahippocampal gyrus 8 Band 01 Giacomini 9 Intralimbic gyrus Fimbria (extraventricular
part of hippocam-
pal body) Margo denticulatus (extraventricular 01hippocampal body) Isthmus Lingual gyrus Anterior calcarine
-
--
part
sulcus
Splenium of the corpus Cingulate gyrus Body of fornix Anterior commissure
callosum
Anterior column of fornix (postcommissural
lornix) Paraterminal gyrus (septal Subcallosal gyrus Temporal pole
nuclei)
b
58
1 Anatomy Paracenlral Iobule
Anl. lubercle or thalamus Body or romix
Mammillolhal. lracl
Poslcrior column or rornix , \
, I
Slriae longiludinales ( lancisi)
Subeallasal J)'rus I
, I ,
Precuneus
I
Thalamo-cinlUlar projeclion
I I
I I ¡
I I I I
Gyrus cingu Ii I I I
I I I
.&.
I
, I
Cuneus
-, ,, , Isthmus
,,,'
Denlate IYrus
I
I
I
, I Hippocampus,
,
I I
area
,
Olr actory slalk (or Iracl)
, Medial
\
I I
I
, Un~us
IYruS,' I I AmYldala
Anlerior
Fig. 1.54a The hippocampo-mamillo-thalamocingular circuit (01 Papez) in a dissected human brain (Irom Haymaker, Anderson, and Nauta, The Hypothalamus, Springlield, IL: Thomas, 1969, p..140, Fig. 4.27). Fibers originating in the hippocampus (blue) extend via the fornix to the mamillary body. A relay (red) pass es upward in the mamillothalamictract to reach the anterior group 01 thalamic nuclei, Irom
Par'. olracl.
olraclory slria
"
Hippocampal
,
,
column
Temporal pole anneclanl fibers
Fusirorm IYrus or rornix
which libers radiate through the internal capsule to reach the cingular gyrus. Association libers bring the cingulum into communication with a large portíon 01 the cerebral cortex. Fibers reaching the hippocampus (green) are varied in their originoBelore entering the hippocampus, Ihey synapse in the dentate gyrus
Topographic Anatomy for Microsurgical Approaches to Intrinsic Brain Tumors
59
A. Allocortex bulbi olfactorii . ..~ Bulbus olfactorius
i: ~ Bulbus olfactorius
accessorius
B. Allocortex primitivus 1. Paleocortex 1 . Paleocortex
I or semicortex
~f a Regioretrobulbaris Regioperiamygdalaris
;:;;;;; Tuberculumolfactorium Septum and regio periseptalis ~ Regiodiagonalis
~
2. Paleocortex
11or eupaleocortex
00i Regioprepiriformis 11.
Archicortex
I
Subiculum . Cornu ammonls Fascia dentata
}
Of the hippocampus retrocommissuralis
Hippocampus
supracommissuralis
Hippocampus
praecommissuralis
C. Periallocortex 111. Peripaleocortex Regio peripaleocorticalis
claustralis
IV.Periarchicortex J
o', Regioentorhinalisand
I ~
area perirhinalis
Regio presubicularis
,
" ...'
,''1i<'(
and
Fig. 1.54b The allocortical and mesocortical limbic lobe, with different cortical areas (after Stephan, Allocortex, Berlin: Springer, 1975)
area parasubicularis Regioretrosplenialis
111 . Regio cingularis periarchicorticalis
;;;;;;;:i
andareasubgenualis
1
Fig. 1.54e The seven main semicircular connections of the limbic lobe 1 Cingulate 2 Longitudinal striae 3 Fornix 4 Stria terminalis 5 Stria medullaris thalami 6 Diagonal band and connection to the olfactory striae 7 Uncinate fasciculus Green Cingulate fibers Red Fibers of the hippocampus Blue Amygdaloid complex Yellow Connections between the mesodiencephalon
60
1 Anatomy
Fig. 1.54 d The major eonneetive fibers of the limbie lobe 1 Cingulategyrus 2 Longitudinal striae (indusium griseum) 3 Fornix 4 Terminalstriae 5 Medullarystriae 6 Habenulointerpeduneulartraet 7 Inferiorthalamie pedunele 8 Ventralamygdalofugal fibers 9 Mamillothalamietraet 10 Mamillotegmentaltraet 11 Mamillarypeduneles 12 Medialteleneephalie faseieles 13 Dorsallongitudinal faseieles (Sehütz) 14 Olfactory traet 15 Diagonal traet 16 Septal area 17 Anterior commissure 18 Anteriorthalamie nueleus 19 Dorsalmedial thalamie nueleus 20 Habenula 21 Interpeduneularnueleus 22 Mamillarybody 23 Hypothalamus 24 Amygdala body 25 Hippoeampus 26 Thalamoeingulatefibers LTO Lateraltemporo-oeeipital gyrus MTO Medial temporo-oeeipital gyrus Tpole Temporalpole
Table 1.7b
Nuclei ofthe limbic lobe
Telencephalic nuclei Amygdala Basolateral Corticomedial (olfaetory) Septal nuclei Basal telencephalic nuelei Ventral striatum Nucl~us aceumbens septi Ventral pallidum Substantia innominata Basal nucleus of Meynert Diencephalic nuclei Hypothalamic Periventricular Intermediate Lateral Mamillary body Thalamic Nuclei anterior and laterodorsalis Nueleus mediodorsalis Habenula Medial pulvinar Mesencephalic nuclei Interpeduncular nucleus Dorsal and deep tegmental nuclei (Gudden) Peripeduncular nucleus Ventral tegmental area (Tsai) Periaqueduetal gray Central superior nucleus (Bekhterev)
a
Topographic Anatomy for Microsurgical Approaches to Intrinsic Brain Tumors
61
The parahippocampalgyrus is an elongated, C-shaped gyrus betweenthe uncus and isthmus of the cingulate gyrus, where the lingualgyrus merges. Laterally, the collateral sulcus separates it fromthefusiformgyrus(lateral occipitotemporalgyrus). Anteriorly.the rhinal sulcus performs the same function. This gyrus encirclesthe lateral incisural space and forms a hook-like bulbous expansionanteriorly, the uncus. The parahippocampus and uncus
lie laterally and envelop the hippocampus, the dentate gyrus, and the cornu ammonis, and are separated fram them by the hippocampal sulcus. There are individual connecting arms between the parahippocampus gyrus, the anterior part of T3, and the anterior part of the LTO gyrus (Figs. 1.55, 1.56). (For variations of the rhinal and collateral sulci, see Fig. 12.1-7 in Ono et al., 1990.)
Fig.1.55 The left parahippocampus gyrus (inferomedialview) a Notethe connections (arrows) with both the medial and lateral temporo-occipitalgyri (the latter via a bridge across the collateral sulcus) LTO Lateraltemporo-occipital gyrus Ph Parahippocampus MTOMedialtemporo-occipital gyrus u Uncus
b u
The same specimen after removal of the parahippocampus Uncus
gyrus
b
Fig.1.56 The left parahippocampus gyrus (inferomedial view) 1 Parahippocampus 2 Lateraltemporo-occipital (fusiform) gyrus 3 Medialtemporo-occipital (Iingual) gyrus Uncus Notethe connection (arrow) to the inferior temporal gyrus (T3) Righthippocampal gyrus, with the collateral sulcus intact Le!thippocampalgyrus,withthe uninterruptedcollateral sulcus
e
62
1 Anatomy
The cingulate gyrus is a C-shaped convolution around the callosum, with numerous connecting arms with the medial frontal, paracentral, precuneal, cuneal, and parahippocampal gyri (Fig. 1.57a-b, 1.60a-d). In Figure 1.58, after removal of the cingu-
late gyrus, these intersected connective arms be come visible. The connecting arms to the mentioned areas are constant, but show individual variations. In the case of corpus callosum agenesis, the cingulate gyrus is also markedly changed. (see Fig. 1.6f, on p. 10.)
Fig.1.57a The left cerebral hemisphere (medial view).The cingulate pole (Ci.pol), with connections to the gyrus rectus (arrow 1), the rostral part of the medial frontal gyrus (arrow 2), the middle part of the medial frontal gyrus (arrow 3), and to the cingulate gyrus (arrow 4), and with connections to the precuneus (arrows 5 and 6)
Fig. 1.57b The right cerebral hemisphere, with an unusually-shaped callosal body. Note the connection of the cingulate gyrus to the precentral gyrus to the precuneus and to the medial frontalgyrus (arrows) Ci.pol Cingulate poie Cu Cuneus Fu Lateraltemporo-occipital (fusiform)gyrus Li(MTO) Medialtemporo-occipital (Iingual)gyrus Parac Paracentral gyrus Ph Parahippocampus Pr.cu Precuneus U Uncus
Fig.1.58 Removalof the cingulate gyrus by cutting the connective arms in the depths of the cingulate sulcus. The numerous transverse gyriof the superior frontalgyrus and the precuneus lobules are shown (arrows)
Fig. 1.59a The right cingulate gyrus (medial view). The parolfactory area deserves the term "cingulate pole," with its regular connectionlo the gyrus rectus and medial frontalgyrus
Topographic Anatomy for Microsurgical Approaches to Intrinsic Brain Tumors
63
Fig.1.59b The connective arms of the right cingulate gyrus with the lrontal,central, parietal, occipital, and temporallobes
Figure1.59a-e shows the axial position of the cingulate gyrus inthecenter of the cerebral hemisphere, like a wheel with spokes, withall the hemispheric gyri directly or indirectly merging. The subcallosal area like the para terminal or parolfactorial areasshouldbetter be termed the "cingulate pole"; the subcallosal gyrus,gyrusrectus,one or two inferior arms of the medial frontal gyrus(between the inferior, superior, rostral, and parolfactory su1ci)and the cingulate gyrus originate in this polar area. Posteroinferiorto the splenium, the cingular gyrus shows a regular reductionof size (isthmus), where the base of the calcarine and parieto-occipitalsulci are situated. Immediately inferior to the isthmusarea, there is a regular connection to the medial temporooccipitalgyrus and anteriorly to the parahippocampal gyrus. Withinthe cingulate and parieto-occipital sulci, there are numerousshortandsmalltransverse gyri, whereas within the callosal sulcusthere are no such gyri. The cytoarchitecture of the cingulate regionhas been comprehensively studied (Stephan 1975, Zilles 1987,1990). Figure 1.61shows the topography of the cytoarchitecture of the cingular gyrus. (For more detailed Stephan 1975.)
information,
see
Fig. 1.59c The two cingulate gyri (superior view). Note the development of the cingulum around the corpus callosum, and the relations to the frontal, temporal, and occipitallobes
r
64
1 Anatomy
a
d Fig. 1.60 The cingulate gyrus, with connective arms (arrows) a The right cingulum, with two arms to the precuneus (arrows 1 and 2), to the medial frontal gyrus (arrows 3 and 4), and to the gyrus rectus (arrow 5) b The right cingulum, with three arms (arrows) to the precuneus and two arms (arrows) to the medial frontal gyrus
e The left cingulum, with tion anterior to the callosal tral gyrus, and an unusual d The left cingulum, with two arms to the precuneus
a unique configuration involving an connecknee (arrows), a connection to the paracenconnection with the precuneus arms (arrows) to the medial frontal gyri and
Fig. 1.61 a, b The cytoarchitecture of the insular region (from Stephan, in: Leonhardt, Tondury and Zilles, Rauber/Kopsch: Anatomie des Menschen, Stuttgart: Thieme, 1987, vol. 3, p. 415, Fig. 13.76) 1 Cingular sulcus 2 Medioradiate area (isocortical) 3 Dorsal infraradiate area (pro-isocortex) 4 Inferior infraradiate area (periarchicortex) 5 Callosal sulcus 6 Subgenual area (periarchicortex) 7 Precommissural hippocampus 8 Septum 9 Anterior commissure 10 Fornix
Topographic Anatomy for Microsurgical Approaches to Intrinsic Brain Tumors
TheCentralZoneof White Matter (External andInternalCapsules)
internal capsules. For the most part, they are internal to the arcuate fibers, and they intersect the commissural fibers of the corpus callosum and the anterior commissure. At the periphery of the corpus striatum, they form the corona radiata. The medial aspect of the corona radiata is separated from the lateral ventric1e by the fronto-occipital fasciculus, and its lateral aspect is covered by the superior longitudinal fasciculus. Below, the corona radiata is directly continuous with the internal capsule, a thick, curved band of white matter that consists of all the projection fibers, and which cuts into the corpus striatum, dividing it almost completely into two parts, the lentiform and the caudate nuc1ei (Fig. 1.62a-b).
Thisarea is best described by Williams and Warwick (1989: 973-5)and,withthe permission of the authors, their description is adaptedhere with some additions. The projection fibers of the anterior,superior, posterior, and inferior pedunc1es, the thalamic radiations,the supralenticular, retrolenticular, and sublenticular pedunculiof the corticopontine fibers, and the supralenticular pedunclesof the corticospinal fibers, converge from all directions intothecentral zone of the white matter to form the external and
Cortieospinal traet Thalamocortieal fibers
Thalamus
",
, , .-
, Foramen of Magendie
,,
,,
,
'."
,,
---Pyrarnid
Fastigium "
Fig.1.62a Projectionlibers in relaliontothelateralventriele (preparation by A Vierling,AnatomicInstitute01the University01 Heldelberg; IromElze,Anatomie desMenschen, 2nded., Berlin: Spnnger, 1960,p. 578)
65
/
"
,,
,,
,,
, , Amygdala
, Temporal horn
Lateral recessus of fourth ventricle
Glossopharyngeal nerve Vagal nerve
66
1 Anatomy
j
/
j .:
I
"-
.
.}.:: ::
Fig. 1.62b Projection fibers in relationto the basal ganglia, central nuclei, and ventricle (coronal view)
The Internal Capsule In horizontal section through the cerebral hemisphere, the internal capsule is a broad band of white fibers, bent with a lateral concavity,which accommodates itself onto the convex medial surface of the lentiform nuc1eus and medially onto the head of the caudate nuc1eus and into the concavity of the lateral thalamus (Fig. 1.63a). It can be divided into an anterior limb, a genu, a posterior limb, a retrolentiform part and a sublentiform part. The anterior limb is interposed between the lentiform nuc1eus on the lateral side and the head of the caudate nuc1eus on the medial side. The posterior limb has the thalamus on its medial side and the lentiform nuc1euson its lateral side. The fibers of the internal capsule continue to converge as they pass downwards, and at the same time the frontal fibers tend to pass backwards and medially, while the temporal and occipital fibers pass forwards and laterally. At the lower limit of the lentiform nuc1eus, they are crossed by the optic tract and enter the midbrain. The corticofugal fibers enter
the crus cerebri, where the frontal fibers are placed on the medial side and the temporal, parietal and occipital fibers on the lateral side. The topography of the projection fibers within the mesencephalon are presented in Fig. 1.63b-d. The anterior limb of the internal capsule contains frontopontine fibers, which arise in the cortex of the frontallobe and synapse about the cells of the nuc1eus pontis, the axons of which pass to the cerebellar hemisphere of the opposite side. In addition, there are the fibers of the anterior thalamic radiation, which interconnect with the medial and anterior nuc1ei of the thalamus and the corte x of the frontallobe. The genu contains the corticonuc1ear fibers, which arise mainly from area 4 oí the cerehral cortex and terminate in the motor nuc1ei 01' the cranial nen'es to the head, mostly of the opposite side. The most anterior fibers oí the superior thalamic radiation (interconnecting the thalamus and cerebral cortex) also extend into the gcnu.
Topographic Anatomy for Microsurgical Approaches to Intrinsic Brain Tumors Flg.1.63a The topography 01projeetion liberswithinthe internal eapsule (alter Zilles in Leonhardt,Tóndury and Zilles, Rauber / Kopsch: Anatomiedes Menschen,Stuttgart: Thieme, 1987, vol. 3, p. 465)
67
Head 01 eaudate nucleus
1.2 Anterior crus 1 Frontopontine traet 2 Anteriorthalamietraet 3 Genu 3 Cortieonuelear traet (dark red: head, neek) 4-12 Posteriorcrus 4 Cortieospinal traet (dark red: anteriorupperlimb;middle-trunk; posterior-
Putamen Lateral pallidum
Medial pallidum
lower limb)
5
Cortieorubral and eortieotegmental traet (beige) 6 Superiorthalamie traet 7 Inleriorthalamie traet 8 Auditory radiation 9 Optie radiation 10 Posteriorthalamie traet 11 Temporoparieto-oeeipital libers 12 Cortieoteetaland eortieotegmental libers Retrolenticularpart of the capsula itema: light green7; dark green 7, 8; tract 10- 12 Yellow Cortieopontine libers Orange Thalamoeortieal libers
Thalamus
Medial genieulate body Lateral genieulate body
Griseum eentrale meseneephali Aqueduet Nueleus nervi oeulomotorii Collieulus superior Nucleus ruber
Nueleus euneilormis Nucleus paraterminalis Teetum
Substantia nigra
Tegmentum
Peduncle Oecipitopontine traet Crus Temporopontine traet Parietopontine traet Cortieospinal traet Cortieobulbar traet Frontopontine traet Nueleus interpeduneularis Oeulomotor nerve
Fig 1.63b Thetopography01projeetion libers within the meseneephalon
68
1 Anatomy
Fig. 1.63c The topography of projection fibers within the mesencephalon, in greater detail (from Brodal, Neurological Anatomy,Oxford: Oxford University Press, 1981, p. 186)
Stratummedium Slralum zonale
I
I Slralum
\
Slralum cinereum \ (= Cappa cinera1'-,
Stralumprofundum,"-
,
I
\\ I
--::,$
I
I
oplicum
Nucleus
The vascularization of the internal capsule is fram perforating branches of the ICA, A¡, and M¡ (see Table 1.12b, p. 103). The posterior limb includes the corticospinal tract disposed in scattered bundles, with the fibers concerned with the innervation of the upper limb anterior, followed by those to the trunk and lower limbs. Other descending fibers here, include frantopontine fibers from the frantallobe (in particular areas 4 and 6), corticorubral fibers from the frontal lo be to the red nucleus, and fibers from the globus pallidus contained in the subthalamic fasciculus. Most of this portion of the internal capsule contains fibers of the superior thalamic radiation which carry general sensory impulses from the ventral thalamic nuclei to the postcentral gyrus.
tractus
mesencephalici n. trigemini I
Tractus spinotectalis
I
I
"
Nucleus reticularis laterali;:-"'I:: -" I I Corpus teniculatum mediale
In the retralentiform sector there are parietopontine and occipitopontine fibers, and fibers from the occipital cortex to the superior colliculus and pretectal region. In addition, there is the posterior thalamic radiation, which includes the optic radiation, and interconnections between the cortices of the occipital and parietallobes and the caudal portions of the thalamus (especially the pulvinar). The fibers of the optic radiation arise in the lateral geniculate body and sweep backwards in the angle between the central section and the inferior horn of the lateral ventricle. In their course, they are intimately related to the superior and lateral surfaces of the inferior horn and the lateral surface of the posterior horn of the lateral ventricle, and are separated fram the latter by the tapetum. The sublentiform part contains the temporopontine and some parietopontine fibers and the acoustic radiation, running fram the medial geniculate body to the superior temporal and the transverse temporal gyri (areas 41 and 42). There are also a few fibers interconnecting the thalamus with the cortex of the temporal lobe and the insula. The fibers of the acoustic radiation sweep forwards and late rally below and behind the lentiform nucleus to reach the cortex.
"
,.-\..._
I I I I I I I Tractus temporopontinus
Fig. 1.63d Section through the mesencephalon at the rever of the superior colliculus. N. r. = red nucleus, surrounded by fibers of the centrallongitudinal fasciculus. 111= Oculomotar Nucleus. N. 111= Oculomo-
tar Nerve. From: H. Ferner and J. Staubesand. In: Benninghoff and Goetler,eds., Anatomy, Vol. 3. Urban & Schwarzenberg, 1975, p. 201, Fig. 213.
The External Capsules, Capsula Extrema, and Claustrum (Fig. 1.68A. e, p. 76).
This layer of the white matter is interposed between the lateral aspectof the lentiform nucleus and the claustrum. The fibers of the frontoparietal (and temporal operculum?) of the insula pass acrass the lateral surface of the lentiform nucleus, and turn medially below this nucleus and the ansa lenticularis. The capsula extrema is located lateral to the claustrum and medial to the insula. For nearly a century, the identity and function of the claustrum has presented a puzzle. Its location, adjacent to the putamen, suggests that it belongs to the basal ganglia, but its afferent and efferent connections are mainly with the cortex, implying that, despite its subcorticallocation, the claustrum deals primarily with cortical information. (For more information, see Sherk 1986.)The arterial vascularization of the external capsule is from perforating arteries, whereas the capsula extrema receives perforating branches from the M2-M3 segments of the MCA. In the retrolenticular area, however, there is a branch from the MCA to the extrema, external, and internal capsule (see Vol.IV B), which needs to be known for surgery of insular tumors. The venous drainage is outwards in the capsula extrema, and inwards in the external capsule and claustrum (the external capsule may have both).
Topographic Anatomy for Microsurgical Approaches to Intrinsic Brain Tumors
69
Summary Thecyto-,myelo-, angio- and chemo-architecture of the CNS has beenwell described by neuroanatomists, neurophysiologists, neuropathologists, neurosurgeons, and neurochemists. However, beyondthesedescriptions, there are particular structural patterns that should concern all of us, especially neurosurgeons, neuroradiologists,and neuropathologists (Fig. 2.1, p. 127). l. Divisions o[ the cerebrum.
'ieocortical Isocortical are a Unimodal area Heteromodal area Transitional (allocortical and mesocortical) Archicortical area Paleocortical area Periallocortical area Proisocortical area Central nudei Ventricular 2. Compartmentalization o[ the CSF system. Cisterns, su1ci, fissures(presented in Vols. 1, III A), ventricles with their subcompartments such as the lateral ventricle (again in five different parts),third and fourth ventricles, and aqueduct. 3. Compartmentalization o[ the arterial and venous systems (seeVol.III A). 4. Lobar systemo[ the telencephalon. There are seven cerebrallobesin each hemisphere: the frontal, central, parietal, occipital. temporal, insular, and limbic. The gyral convolutions and lobespresent a continuous pattern. For pragmatic and teaching purposesthe separation of gyri into distinct lobes is recommended. 5. Gyral convolutions: The frontal, parietal and temporal lobeseach consist of four gyri; the centrallobe of three; the occipitallobe of five; the insular lobe of five; and the limbic lobe has threedistinguishableareas (the olfactory, septal, and hippocampal areas) and two gyri (the parahippocampal and cingulate gyrus).Eachgyrushas several transverse gyri, which are hidden in thedepthsof the sulci and fissures. Each gyrus consists of numerousconical (pyramidal) segments extending from the surface downto the periventricular matrix. 6. Segmentalwhite-matter sectorial patterns. The routes of arteriesandveins,as well as the pathways of fibers of different subsystemseachhave a distinct pattern within these sectors of gyral segments.In anatomical and MRI sections of the brain, the cascadeof sectorswithin the white matter is well recognizable: subcortical,subgyral,gyral, lobar, and capsular sectors. The projectionfibersenter and leave the gyral segmental system in sector 4 (Iobarpedunde), also where the coronate radiation and thalamic radiationconvergeinto the external and internal capsule and the callosalradiations transverse them. There is a distinct relation betweenthe sectors and the pattern of arterial blood supply and venousdrainage. The white matter can be subdivided into two zones.a large peripheral zone intercepted by numerous gyral segmentsbetweenthe cortex and the ventricular wall, and a central zoneat the base of the telencephalon, with the internal capsule andhemisphericpeduncles. Each gyral segment consists of four sectors,through which the pathways of several subsystems course (see pp.28-32 and Table
1.6 on page 32).
7. Brodmann's maps (1909) o[ the human cerebral cortex, whichappear in almost every book of neuroanatomy and neurophysiology,have proved to be very usefuI ideograms for neu-
a
e Fig. 1.64a-c Area map 01the human cortex (after Brodmann, Vergleichende Lokalisationslehre der Grosshirnrinde, Leipzig: Barth, 1909) a Lateral view, b medial view Frontallobe (yellow) 6, 8, 9, 10, 11, 12, 32, 46, 47 Precentral gyrus (dark blue) 4, Postcentral gyrus (Iight blue) 1, 2, 3, 5 Parietallobe (beige) 7,39,40, Occipitallobe (dark green) 17,18, 19 Limbic lobe (orange) 23,24, 25, 26, 27, 28, 29, 30, 31,33,34,35,36 Temporallobe (Iight green) 20, 21,22, 37, 38 Broca area (white) 44, 45 Primary auditory cortex (white) 41, 42 Gustatory cortex (white) 43 Fig. 1.64c Area map 01 the human cortex, Irontobasal view (after Duvernoy, The Human Brain, Vienna: Springer, 1991, p. 45)
70
1 Anatomy
roscientic communications. Based on a cytoarchitectural classification with functional interferences, they may yet stand the test of time for clinical use, though they still represent only an approximation that is functionally unreliable in individual cases (Fig. 1.64). The text and atlas on the cytoarchitecture of the human adult cortex by von Economo and Koskinas, published in 1925, is also well known, but less frequently discussed. Szikla et al. (1977), in their
monograph on the monumental work of von Economo and Koskinas, presented schematic diagrams and nomenclature that are unsurpassed to date, and of immense pragmatic valve at the present stage of neuroimaging. Three figures from Economo and Koskinas are therefore presented here (Fig. 1.65a-e). For general clinical use a reduced version is required; this is provided in Fig. 1.66 on page 73.
Fig.1.65a
[ Fig.1.65b
Fig. 1.65c t> I
l
Topographic Anatomy for Microsurgical Approaches to Intrinsic Brain Tumors
Fig.1.65a-c Areamap 01the human cortex (alter von Economo and Koskinas, Die Cytoarchitektonikder Hirnrinde des erwachsenen Menschen,Vienna:Springer,1925).The key may at lirst appear daunting,
F1 F1Q, F20, F30
butwitha little patience it will prove to contain inlormation extremely use-
F3 (pF3) F3pt F3t Fi. Fo Fus 11,12
lultothesurgeon.a Lateralview, b medial view, e inlerior view AB Area parollactoria Broca (carrelour ollactil) Ang Lobulus angularis Aq Aquaeduct AR Gyri Andreae Retzii BB Broca'sband BG Bandelettede Giacomini B.011 Bulbusollactorius C FissuraCalcarina Ca Gyrus centralis anterior Cc. Corpus callosum Ch Chiasmanervi optici Coa Commissuraanterior Cp Gyrus centralis posterior Cu Cuneus e Cap cmg Sulcuscallosomarginalis d Sulcusdiagonalis (operculi) E Gyrus (occipitalis) descendens (perpendicularis) Ecker Gyrus lrontalis primus, Pars mediana 01
secundus,
tertius
F1 F30p
l. dt l. m f. pa Is. c f. Sy Gsm g. a. a., g. a. m., g. a. p. g. g. g. g. g.
amb ant. o ant. d. ant. proc br. ac. (a.)
g.
br. 1, 11, 111
g. br. imd
71
Pars orbitalis gyri etc. Pars opercularis 01 = Broca's point Pars praetriangularis 01 F3 Pars triangularis 01 F3 (Cap) Fimbria Fornix Gyrus lusilormis Sulcus Irontalis superior, inlerior Fascia dentata Sulcus lrontalis medius Fossa paracentralis Fasciola cinerea Fissura Sylvii Lobulus supramarginalis Gyrus arcuatus anterior, medius, posterior lobuli parietalis superioris Gyrus ambiens Gyrus anticentralis operculi Gyrus antidiagonalis operculi Gyrus antipraecentralis operculi Gyrus brevis accessorius (anterior) insulae Gyrus brevis primus, secundus, tertius insulae Gyrus brevis intermedius insulae Legend continued on next page
1>
72
1 Anatomy
g. cl. P g. dt g. d. u g. F, F3 g. fl. a. (p.) g. fs. g. g. g. il g. imd g.lg.s. (i.) g.ol.lt g. 01.mi g. pip g. pl. a. (p.) g. po. i. (s.) g. po. is. 1,11
g. prois g. r g. rl g. sc g. sg. i, (m), (s) g. sml g.~ g. U.a. (p) g.~a.S. g.~~ g. tr. op.
1, 11, 11I
Hi h hi IP Is ic ig ip ipo it J Lg L. s. a. (p), Lr 1 1.a Ig It m mg. a., (s.), (p.) O" O2,03 Op. P Op. R Opt ot Pa Pb Pi Pr Ps PT pF3 p. f. po p.Sy R Rst rC rh ri
Gyrus cuneolingualis posterior Gyrus dentatus Gyri digitati unci orbital Gyrus frontolimbicus anterior (posterior) Gyrus fasciolaris Gyrus geniculatus Gyrus intralimbicus Gyrus brevis intermedius (insulae) Gyrus lingualis superior (inferior) Gyrus olfactorius lateralis Gyrus olfactorius medialis Gyrus parietalis inferior posterior Gyrus parietolimbicus anterior (posterior) Gyrus parietooccipitalis inferior (superior) Gyrus postcentralis insuale primus et secundus Gyrus praecentralis insulae Gyrus rectus Gyrus retrolimbicus (rhinencephalolingualis) Gyrus sub callo sus Gyrus sagittalis cunei inferior, (medius), (superior) Gyrus semilunaris Gyrus subtriangularis operculi Gyrus temporolimbicus anterior, (posterior) Gyri temporales transversi anteriores Schwalbe Gyrus transversus insulae Gyrus transversus operculi parietalis primus, secundus, tertius Gyrus Heschl primus, secundus (gyri temporales profundi, gyri temporales transversi posteriores) Gyrus hippocampi Ramus horizontalis fissurae Sylvii Fissura hippocampi Insulare pole Isthmus Incisura capi Induseum griseum Sulcus interparietalis Incisura preoccipitalis Incisura temporalis Incisura Jensen (Sulcus intermedius primus) Lingula Gyrus limbicus superior pars anterior, (pars posterior) Gyri limbici pars retrosplenialis Sulcus intralimbicus Lamina affixa. Sulcus lingualis Lamina terminalis Corpus mamillare Margo anterior, (superior), (posterior) sulci circularis insulae Gyrus occipitalis primus, secundus, tertius Operculum parietale Operculum rolando nervus opticus Fissura occipitotemporalis (F. collateralis) Lobulus paracentralis Regio parietalis basalis Lobus parietalis inferior Praecuneus Lobus parietalis superior Gyrus (temporo-) polaris 1nferior frontal gyrus Pli falciforme Fissura parietooccipitalis Ramus posterior fissurae Sylvii Sulcus Rolando Rostrum corporis callosi Fissura retrocalcarina Fissura rhinalis Sulcus rostralis inferior
rl rs S.p.a Spl s.a s. a. rh s. B s. br. 1,1I s. cc s. c. is s. d s. fd S. frmg, mI., (md.), (11.) s. g. F, s. imd.l, s.1 S01
11
S02 S. oa s.ol S. oro It, (ml.), (imd.), (tr) s. pa S. po. i, (s) S. po. is S. poI. a. (m-), (p.), (ps.) s. prc S. prd S. pro i, (s) S. pr. is S. p. s S. p. tr S. rh. i s.san S. sc. a, (p.) S. sg. s, (i) S. so S. sor S. sp S. tp. 1, 11 S. tr. a. S s. tr. op. 1, 11
Tr.o Tu. o t" t2, t3 t", t1", t2" t2" U V v. cmg
Sulcus retrolingualis Sulcus rostralis superior Substantia perforata anterior Splenium corporis callosi Sulcus acusticus Sulcus Sulcus Sulcus Sulcus Sulcus Sulcus Sulcus
arcuatus rhinencephali (fissura rhinica) Brissaud brevis primus, secundus insulae corporis callosi centralis insulae (parolfactorius) diagonalis fimbriodentatus
Sulcus frontomarginalis medialis, (medius), (Iateralis) Sulcus gyri frontalis primi Sulcus intermedius primus (Jensen), secundus Sulcus lunatus Sulcus occipitalis primus (praeoccipitalis, interoccipitalis) Sulcus occipitalis (secundus) lateralis Sulcus occipitalis anterior Sulcus olfactorius Sulcus orbitalis lateralis, (medialis), (intermedius), (transversus) Sulcus paracentralis Sulcus postcentralis inferior (superior) Sulcus postcentralis insulae Sulcus parolfactorius anterior, (medius), (posterior), (postremus) Sulcus precunei Sulcus prediagonalis Sulcus precentralis inferior, (superior) Sulcus precentralis insulae Sulcus parietalis superior Sulcus parietalis transversus Sulcus rhinencephali internus Sulcus semiannularis Sulcus subcentralis anterior, (posterior) Sulcus sagittalis cunei superior, (inferior) Sulcus suboccipitalis Sulcus supraorbitalis Sulcus subparietalis Sulcus temporalis profundus primus, secundus Sulci temporales transversi anteriores Schwalbe Sulcus transversus operculi parietalis primus, secundus Gyrus temporalis primus, secundus, tertius Thalamus Truncus fissurae parietooccipitalis et Calcarinae Trigonum (or tuber) olfactorium Tuberculum olfactorium or colliculus nuclei caudati Sulcus temporalis superior, medius, internus caudal end of temporal sulci Uncus Ramus verticalis fissurae Sylvii Ramus verticalis sulci callosomarginalis
Topographic Anatomy for Microsurgical Approaches Fig 1.66 The labeling a Lateralview a F1 F2 F3
m 01 02 03 Op Orb p Pc1 Pc1 Pc1 Pi Pm Pr1 Pr2 Ps S s T1 T2 T3 Tr b Medialview a Ci Cu Fa Is LTO
1
of gyral areas for analyzing MR images Anterior area or part Superior frontal gyrus Middle frontal gyrus Inferior frontal gyrus Inferior area or part Motor precentral gyrus Middle are a or part Superior occipital gyrus Middle occipital gyrus Inferior occipital gyrus Opercular part of the inferior frontal gyrus Orbital part of the inferior frontal gyrus Posterior area or part Superior postcentral gyrus Middle postcentral gyrus Inferior postcentral gyrus Inferior parietallobule Middle parietallobule Superior precentral gyrus Middle precentral gyrus Superior parietallobule Sensory precentral gyrus Superior are a or part Superior temporal gyrus Middle temporal gyrus Inferior temporal gyrus Triangular part of the inferior frontal gyrus Anterior part Cingulum Cuneus Fornix Isthmus Lateral temporo-occipital Middle part Mamillary body Medial frontal gyrus Medial temporo-occipital Posterior part Paracentral gyrus precuneus
gyrus
I
2
I
\-.-/ I /t
3
73
to Intrinsic Brain Tumors 4
I
i I'c
V
I
U I ./Pr
I
5
6
1/
7
I
l I VI \
'Cuneus
I 1
I
2
a I
3
I
4
I
5
I
6
I
7
Rostral gyrus Subcallosal gyrus
gyrus
Septum Inferiortemporal gyrus Temporalpole uncus
AReviewof Cerebral Anatomy in Correlation with MRI Thethree-planesectional MRI pictures of the living brain impress at the veryfirst moment of inspection. Neuroscientists, everyone however,who analyze the gray scale thoroughly to depict reliable detailsof structural changes, are obliged to check and correlate lopographicand functional anatomical knowledge with the qualilativelyimproving MRIpictures,in a reciprocal processthat helps 10 develop accurate criteria for better interpretation of cIinical observationsand pathological anatomic findings. At the end of the section on gyral and white-matter anatomy, il Ihereforeseemed advisable to add correlative pictures in three planesof sectional anatomy and corresponding MR images side by side to offer evidence for the concept of anatomy proposed here.Insagittaland horizontalplanes, stained brain sections are
I
I
b
used in Figures 1.67A. a-g and Figures 1.68A. a-f. In coronal section (Fig. 1.69A and C), drawings are presented that have been derived from anatomical sections to elucidate the landmarks. The continuum of gyral convolutions is well seen in Figures 1.67A. a, b and 1.68 A. a. The holistic architecture of the brain is obvious, as the different lobes can only be identified approximately and for teaching purposes. The transverse gyri within the depth of the fissures and suIci cause the jigsaw-puzzle effect on MRI pictures. They are better visualized in three-dimensional pictures. The horizontal anatomical sections are not the same as those previously published in Vol. III B, because the sections are performed in different planes. With MRI (Fig. 1.68 B. c-d), the optic radiation is well visualized. In the MRI Figure 1.68B. e, the medial temporo-occipital gyrus is demonstrated in its en tire length.
74
1 Anatomy
a
b
d
e
9
Fig. 1.67A. a-g Cerebral anatomy approximately correlated to MR images (sagittal view): whole sections of cerebrum (stained) (courtesy of Prof. P.Croscurth, Institute of Anatomy, University of Zurich) Fig. 1.67B. a-f MRI sections (noncontrast). The continuum is wellDo visualized on the lateral pictures. Labeling has been omitted to maintain clarity
-
Topographic Anatomy tor Microsurgical Approaches to Intrinsic Brain Tumors
75
F!g 11.67B.a-f
a
b
d
e
f
76
1 Anatomy
+~~ , 4;
,
1 ~ I i~~ .¡, )~
Fig. 1.68A.a-f Cerebral anatomy approximately correlated to MR images (horizontal views) The sections of cerebrum (prepared by M. P. Lusk and 1. Orbay at the Dept. of Neurosurgery, University of Zurich)
d
Topographic Anatomy for Microsurgical Approaches to Intrinsic Brain Tumors
77 1.68A
f
1.688
e
e
f
78
1 Anatomy
Medial frontal gyrus Insular sulcus
Inter mediate sulcus
Inferior frontal sulcus
Callosal knee
Superior rostral sulcus
Olfactory tract and sulcus
a
e
Temporo-occipital
d
e
Temporooccipital sulcus
Parietooccipital sulcus Calcarine fissure Anterior occipital sulcus
9
Temporooccipital sulcus
Collateral sulcus
f
sulcus
Fig. 1.69A.a-g Cerebral anatomy (coronai MRI sections) a A section through the anterior part of the right frontallobe, approximately correlated, just anterior to the callosal knee b A section through the middle frontallobe, just behind the temporal pole level e A section through the posterior part of the frontallobe at the level of the foramen of Momo d A section through the precentral gyri e A section through the postcentral gyri f A section through the parietallobe 9 A section through the occipitallobe
Temporo-occipital
sulcus
Collateral sulcus
Topographic Anatomy for Microsurgical Approaches to Intrinsic Brain Tumors b'
e'
d'
f'
g'
h'
Flg 1.69B a' Anteriorsection Middlesection Posteriorfrontalsections Precentralsection.Note the pyramidal tracts Postcentralsection.Note the pyramidal tracts and decussatio f' Parietalsection.Notethe pyramidal tracts and decussatio g' Parietalsection h' Occipitalsection
<: CI Cu F1 F2 F3 HI
Cingulategyrus Cuneus Superiorfrontalgyrus Middlefrontalgyrus Inferiorfrontalgyrus Hippocampus 11"1 Insula L Lateralfronto-orbitalgyrus LTO Lateraltemporo-occipitalgyrus M Medialfronto-orbitalgyrus MTO Medialtemporo-occipitalgyrus 01 Superioroccipital gyrus 02 Middleoccipital gyrus 03 Inferioroccipitalgyrus P1 Superiorparietallobule P2 Middleparietallobule
P3 Parac Ph PostC1 PostC2 PostC3 Prc.1 Prc.2 Prc.3 Prec R T1 T2 T3 U
Inferior parietallobule Paracentral gyrus Parahippocampus Superior peduncle of the postcentral gyrus Middle peduncie of the postcentral gyrus Inferior peduncle of the postcentral gyrus Superior peduncle of the precentral gyrus Middle peduncle of the precentral gyrus Inferior peduncle of the precentral gyrus Precuneus gyrus rectus Superior temporal gyrus Middle temporal gyrus Inferior temporal gyrus Uncus
79
80
1 Anatomy
I
Culmen
~ ILingula
a'
SSL
Anterior cerebellar
section
ISL Horizontal fissure
Tonsillouvular
a
suleus AVe Primary fissure
./
~
Superoposterior
b' Middle cerebellar
section
/"""" Fig. 1 2 3 4 5
1.69C. a-c and a'-c' White matter traet Subcortical Sublobular lobular lobar Center of white matter-middle cerebellar peduncle AOl Anterior quadrangular lobule AVe Anterior velum BI Biventrallobule GR Gracilis lobule ISl Inferior semilunar lobule Me Mesencephalon No Nodulus POl Posterior quadrangular lobule PVe Posterior velum S Superior cerebellar peduncle SSl Superior semilunar lobule TO Tonsil
Posterolateral
fissure Tonsillouvular
b
suleus
.
Superoposterior fissure
Horizontal fissure
c' Posterior cerebellar section
c T onsillouvular
suleus
Topographic Anatomy for Microsurgical Approaches to Intrinsic Brain Tumors
Infratentorial
Topographic Anatomy
The anatomy and physiology of the cerebellum is well known. Detailed information is readily available in numerous publications(seealsoVol. III A, pp. 312-19). Some specialized aspects of microsurgical anatomy will however be presented here. A methodological approach to understanding the microsurgical anatomyof the white and gray matter of the cerebellum is important in tumor surgery. Our previously proposed scheme, used in discussingthe architecture of the cerebral white matter, is also applicableto the architecture of the white matter of the cerebellum (Table
81
Divisions of the Cerebellum The cerebellum has a median part, the vermis, and two large lateral parts, the cerebellar hemispheres. The paramedian sulci separate the vermis from the hemispheres on the superior surface. Caudally, the cerebellar hemispheres are separated from the vermis by the distinctly narrow and deep posterior cerebellar incisura (cerebellar notch), which contains a fold of dura (the falx cerebelli).
1.8).
Cerebellarlobes, lobules, connective fibers, and peduncles and their functions Lobes
Fissures
Lobusanterior (Pafeocerebef- Precentral fum, spiocerebellar)
Lobules Hemispheric
Vermian
Vinculum
Lingula (archicerebel/um) Centralis Culmen
LOP
Declive
LSS
Folium
LSI + lobulus gracilis
Tuber
Biventer
Pyramis (paleocerebellum)
Ala lobuli centralis LOA
Peduncles
Connective fibers Afferent
Efferent
Brachium conjunctivum (superior cerebellar peduncle)
Tractus spinocerebellaris ventralis Tractus tectocerebellaris
Tractus cerebellorubralis Tractus cerebellotectalis Tractus dentatothalamicus (fasciculus uncinatus ascendens)
Brachium pontis (middle cerebellar peduncle)
Tractus pontocerebellaris
Brachium posterior (inl. cerebellar peduncle) (corpus restiforme)
Tractus spinocerebellaris dorsalis Tractus reticulocerebellaris Tractus olivocerebellaris Tractus vestibulocerebellaris
Primary
Lobusmedius (neocerebelfum,pontocerebellar,tectocerebellar)
Lobuspostenor(neocerebe/lum,pontocerebellar,tectocerebellar)
Postclival (posterosuperior)
(Pyramidal unilateral)
Function
Major output to pons and midbrain Extrapyramidal, motoric ton e
Massive input from neocortex Voluntary motor
Horizontal
Prepyramidal (prebiventer) Secondary (postpyramidal)
Tonsilla
Uvula (paleocerebellum)
Flocculus
Nodulus
Posterolateral
Lobusflocculonodularis (archicerebeffum,vestibulocerebellar) Anteriorquadrangulare lobule Posteriorquadrangular lobule Inferiorsemilunar lobule Superiorsemilunar lobule
Vestibulofastigial
Tractus corticonuclearis
Tractus fastigiobulbaris
Input from vestibular system. Posture and balance
82
1 Anatomy
Cerebellar Lobes and Lobules Conventional anatomical descriptions distinguish two (or sometimes three) lobes in the cerebellum (Williams and Warwick 1989). The cerebellum can be divided into two fundamental parts, termed the tlocculonodular lobe and the corpus cerebelli, the latter comprising an anterior and middle lobe. These subdivisions possess functional as well as morphological and embryological significance. The corpus cerebelli comprises the remainder of the cerebellum, and is separated from the tlucculonodular lobe by the posterolateral fissure, which is the first to appear on the cerebellum both in phylogeny and on,togeny. From a surgical anatomical point of view, we favor distinguishing four lobes (ten hemispheric lobules and nine vermian lobules) (Table 1.8 on p. 81). The corpus cerebelli is subdivided by the fissura prima into anterior lobes, and by the great horizontal fissure into posterior lobes (Fig. 1.70). The anterior lobe lies in front of the fissura prima, and comprises the anterior quadrangular lobule, the ala of the centrallobules, and the lingula, central lobule and culmen. The middle lobe comprises the posterior quadrangular and superior semilunar lobules and the declive and folium. The posterior lobe (inferior to the horizontal sulcus) comprises the inferior semilunar, gracilis, biventer, tonsillar lobules, tuber, pyramis, and uvula. The fourth lobe is separated from the
corpus cerebelli by the posterolateral sulcus, and comprises the tlocculus, the peduncle of the tlocculus, and the nodule. Each of the ten hemispheric and nine vermian lobules contains at least 10-15 primary folia (a total of 100-200 primary folia), and a similar or greater number of secondary and tertiary folia. Between the folia there are many unusual spiral cascades or semicircular connecting extensions (Figs. 1.77 a-b, 1.78a-c, p. 89-90). Certain sectors of the cerebellum are phylogenetically older than the rest. The tlocculonodular lobe, which is predominantly vestibular in its connections, together with the lingula, which receives spinocerebellar connections in addition to vestibular, constitutes the oldest part of the cerebellum, or archicerebellum. The anterior lobe (excluding the lingula, but including the pyramid and uvula of the posterior lobe) is phylogeneticallythe next part to appear, and is predominantly spinocerebellar in its connections, constituting the paleocerebellum. At this stage in phylogeny, this newly developed lobe separated the archicerebellum into two parts, the lingula in front and the tlocculonodular behind. With the evolution of the neopallium in the mammal, there is a further expansion of the cerebellum, with an addition to the middle and posterior lobes (the pyramid and uvula). This addition constitutes the neocerebellum, and is predominantly corticopontocerebellar in its connections. Like the paleocerebellum, the neocerebellum separates the anterior and tlocculonodular lobe.
Preculminar
fissure
Postclival fissure
\ \
Posterolateral fissure
Fig. 1.70 The cerebeliar fissures, peduncles, hemispheric lobules, and vermis (medial view) Red Alterent Blue Elterent Yeliow Fastigiobulbar and cerebeliovestibular fibers AQL Anterior quadrangular lobule BIV Biventral lobule Ce Centrallobule Cu Cuneus De Declive Fo Folia GR Gracilis lobule ISL Inferior semilunar lobule Li Lingula No Nodulus PQL Posterior quadrangular lobule Py Pyramis SSL Superior semilunar lobule Tonsil TO Tuber Tu Uv Uvula
Topographic Anatomy tor Microsurgical Approaches to Intrinsic Brain Tumors Unique structures at the superior and inferior ends of the cerebellumare the medullary velae (best described by Williams and Warwick 1989). Thesuperior medullary ve/um is a thin lamina of white substance that stretches between the superior cerebellar peduncles (brachiaconjunctiva), and with them forms the roof of the cranial part of the fourth ventricle; its deep surface is covered with the ventricular ependyma. The velum is narrow anterosuperiorly, II'hereit extends into the interval between the inferior colliculi, and broader posteroinferiorly, where it is continuous with the whitesubstance of the superior part of the vermis. The folia of the lingulaare prolonged onto the dorsal surface of its lower half, and a medianridge, termed the frenulum veli, descends upon its superior part from between the inferior colliculi (Fig. 1.71a-b). The trochlearnervesemerge at the side of the frenulum.
83
The inferior medullary vela are two thin, somewhat crescentishaped, sheets placed one on each side of the nodule. Each consists of a thin layer of white matter and neuroglia, surfaced over internally by the ventricular ependyma, and externally by the pia mater. Its internal surface forms the lower wall of the lateral dorsal recess of the fourth ventricle; its external surface is related to the superior aspect of the ton sil. Its convex peripheral margin is continuous with the white core of the cerebellum and with the sides of the pyramid, uvula and nodule; its anterior (sometimes inferior) border is free, and from it the ventricular ependyma is prolonged downwards in close apposition with the pia mater to form the thin part of the roof of the ventricle and to reach the taeniae. At its anterolateral comer, the velum is continuous with the dorsal part of the peduncle of the flocculus, from which most, if not all, of its nerve fibers are derived.
"í
lit
.
Rg. 1.71a-b The surfaces 01the cerebellum a A superoanterior view into the cerebellar mesencephalic Notethe hidden superoanterior surlace
b
lissure.
A superoanterior
view 01 the superior medullary velum and lingula
84
1 Anatomy
Cerebellar Hemispheric Borders and Surfaces The cerebellar hemispheres have a unique wedge shape in the .
craniocaudal direction, and present numerous surfaces (Fig. 1.72a-e). Anatomists describe two surfaces (superior and inferior) or three (superior, inferior, and ventral). In order to describe surgical approaches fully, additional anatomical details are required.
Borderlines. There are five cerebellar borderlines (Fig. 1.73a-c). Surfaees. There are six main surfaces and numerous fissural surfaces to identify. 1. The superoanterior (rostral) surface (not mentioned by neuroanatomists). This lies between the anterior lobule and the dorsal inferior mesencephalon, where the cerebellomesencephalic fissure separates the two entities (Fig. 1.7la).
2. The superior surface. This is the well-known largest surface, above the two hemispheres and the vermis, and underneatb the tentorium. 3. The posterior surface. The second largest surface, lying beneath the superior parts of the cerebellar hemispheres (Fig.1.73b). 4. The interhemispheric surface. This is found between the inferior parts of the cerebellar hemispheres (inferior semilunar lobule, gracile lobule, biventer lobule, and tonsil) and the inferior parts of the vermis (such as the tuber, pyramis, and uvula) (Fig.1.73b). 5. The inferior surface. This lies between the tonsillar and biventer lobules and the medulla oblongata, separated by the cerebellomedullary fissure (Fig. 1.73b). 6. The anterior surface. A triangular surface located behind the posterior surface of the petrosal bone, and marked by the convergence of all the cerebellar lobules along the horizontal fissure (Fig. l.72a-c, 1.73c). 7. Fissural and sulcal surfaces.
a
b
Fig. 1.72a-c A lateral view of the right anterior surface. The topography of the horizontal fissure and the cerebellar peduncle is shown b A lateral view of the left anterior surface e Anterior view
Topographic Anatomy for Microsurgical Approaches to Intrinsic Brain Tumors
85
4
d
6
e
6 6
15 Flg 173a-f The lobes and borders of the cerebellar hemispheres (from Villiger and Ludwig, Atlas of Cross-Section Anatomy of the Brain.Philadelphia:Blakiston, 1951, pp. 112-114). a, d superior view; b,e posterior view; e, f anterior view
d-f 1 2 3 4 5 6 7 8 9 10 11 12
Centrallobule Culmen Centralalarlobule Anteriorquadrangularlobule Posteriorquadrangularlobule Superiorsemilunarlobule Inferiorsemilunarlobule Tuber Declive Uvula Flocculus Siventrallobule
Ce Cu Ceal AQL PQL SSL ISL Tu De Uv Floc Si
13 14 15 16 17 18 19 20 21 22 23 24
Gracilislobule Pyramid Tonsil Superiorcerebellarpeduncle Lingularalarlobule Middlecerebellarpeduncle Anteriorhorizontalsulcus Flocculonodular lobule Inferiorcerebellarpeduncle Nodule Lingula Folium
Gr Py To Scp Li al Mcp FlocI Icp No Li Fo
86
1 Anatomy
Cerebellar Flssures and Sulci (Table1.8,Fig.1.70) Fissural surfaces (within the major intracerebellar fissures): Precentral Preculminal Primary Postclival (posterosuperior) Great horizontal Prepyramidal Secondary (pretonsillar) Parafloccular Posterolateral Vallecular (uvulotonsillar) Choroid These are illustrated in Fig. 1.70 on page 82. Cerebellar sulci and folía: There are approximately 300 primary folia; most are oriented transversely and in parallel. Each folium is further subdivided into secondary and tertiary folia. At least 300 more sulci lie between all of these folia. Due to this remarkable arrangement, the cortical area of the cerebellum approaches nearly three-quarters that of the cerebrum, though the mass of the cerebellum is about one-tenth that of the cerebrum (HumeAdams 1988). Estimates suggest that six-sevenths of the cerebellar surface is buried within the fissures and sulci: it would measure over one meter if traced out and laid flat. Fissures (Table 1.8, Fig. 1.70). The fissures of the cerebellum are classified into two categories; those that divide the cerebellum itseU and those that divide the cerebellum from the brainstem. The fissures that divide the cerebellum from the brainstem are of special importance, as they allow access to the superior cerebellar artery (SCA), the anterior cerebellar artery (AICA), the posterior inferior cerebellar artery (PICA), and the entry and exit zones of the cranial nerves. The horizontal fissure is also of surgical significance, as it divides the hemispheres and vermis into superior and inferior halves (the cerebellum could be likened externally to a clam), and extends anteriorly over the brachiu~ pontis to either side of the foramen of Luschka. The horizontal fissure also represents the plane of division separating the vascular territories of the SCA and PICA). The great horizontal fissure runs, on each side, between the superior semilunar lobulus (and also midline, the folium vermis) and the inferior semilunar lobulus (and also midline, the tuber vermis). It extends all the way around laterally and then ventrally, as far as the foramen of Luschka, where the flocculus meets the cerebellopontine cistern. The fissure also extends ventrally to the bulbopontine sulcus and the foramen cecum. All cerebellar hemisphere lobuli and their sulci converge along the horizontal fissure on each side (Fig. 1.72a, b). Other fissures are presented in Figures 1.70 and Table 1.8 on page 81). The fissural separation dividing the cerebellum from the brainstem is arbitrarily divided into three (in order to facilitate anatomical description): the cerebellomesencephalic, the cerebellopontine, and the cerebello-pontine fissures. The lateral and posterior infratentorial cisterns also aid in demarcating these boundaries. They form part of a continuous sulcal channel in the space between the brainstem and the cerebellum. The W-shaped cerebellomesencephalic fissure divides the anterior rostral cerebellum (specifically, the lobulus quadrangu-
laris and ala lobuli centra lis of the cerebellar hemisphere and the culmen and lingula of the vermis) from the dorsal surface of the mesencephalon. It is continuous over the dorsal surface with the same fissure of the opposite side. It extends below to the superior medullary velum, the superior cerebellar peduncle, and the dorsal surface of the middle cerebellar peduncle (Fig. 1.71a-b). The lateral pontomesencephalic segment (second portion) of the superior cerebellar artery (SCA) encircles the lateral midbrain in close relation to the cerebral peduncle and brachium of the colliculi to enter into the fissure anterior to the lobulus quadrangularis. The third portion of the SCA (the cerebellomesencephalic segment) loops within the fissure and gives off tightlylooped perforating branches to the mesencephalon and cortex. The fourth segment exits the fissure, by recurving posteriorly over the medial side of the lobulus quadrangularis, to gain the dorsal cerebellar surface. The medial and lateral mesencephalic and precentral cerebellar veins and the trochlear nerve course through the fissure. Exposure of this area is difficult, beca use the ability to retract the superomedial cerebellum posteriorly is limited, due to the proximity of the semirigid tentorial edge. The cerebellomedullary fissure divides the anterior inferior cerebellum from the medulla. The cerebellomedullary fissure extends upwards to the tela choroidea and to the arachnoid, cover. ing the dorsolateral fourth ventricle and the inferior cerebellar peduncle. The medial and anterior aspect of the biventer overo hangs the approach to this fissure. The PICA originates and loops within the fissure, giving off branches to the inferior medullary velum, the inferior cerebellar peduncle, and the suboccipital cerebellum in its intrafissural course. Cranial nerves IX, X, XI, and XII course late rally to restrict complete exposure. Access to the proximal portion of the posterior inferior cerebellar artery, as it arises from the vertebral artery is, however, possible (Fig. 1.72a-e). The cerebellopontine flSsure relates to the lateral recesses of the fourth ventricle, and is commonly referred to as the cerebelo lopontine angle, although its shape is really triangular. It is borde red medially by the pons and the ventral aspect of the brachium pontis superiorly and laterally by the posterior quadranguiar lobule and superior semilunar lobulus, and inferiorly and laterally by the flocculus, inferior semi lunar lobule, lobulus gracilis and biventer lobulus. The apex of the triangle leads into the horizontal fissure of the cerebellum. The fissure contains the foramen of Luschka, and frequently has choroid plexus projecting into it. Cranial nerves V, VII, VIII, and the anterior inferior cerebellar artery and its branches traverse this fissure (Fig. 1.72a-c). The vallecula cerebelli separates the two cerebellar hemispheres inferiorly, which form a deep median fossa. The uvula and nodulus of the vermis project into the vallecula in the midline. flanked by posterolateral fissures (Fig. 1.74a-e). The foramen of Magendie opens into the cisterna magna at the inferior portion of the vallecula. By opening the fourth ventricle, it is possible to gain access to the dorsal (posterior) median brainstem and lateral recesses of the fourth ventricle and restiform body. Opening more superiorly (between the uvula and the nodulus and the tonsillar lobules) permits the choroidal point of the PICA to be reached. An overview of the whole of the fourth ventricle can be obtained by dissecting and displacing the choroid plexus (Fig. 1.74a-e).
Topographic Anatomy tor Microsurgical Approaches to Intrinsic Brain Tumors
87 ...,
b Fig.1.74a-c An inferiorview through the vallecula into the fourth venIricle.Theposterolateralfissure between the pyramis-uvula and tonsil (arrows) a Thecerebellarvalleculum, with the median foramen of the fourth ventricle(theforamenof Magendie). Note the sulci between the tonsils anduvula(arrows)
b Alter removal of the right tonsil, the extension of the posterolateralfissure between the tonsil and the uvula is well demonstrated (arrow)
"
\
¡ .
, I
e e Another specimen, with a lelt-sided injected PICA. On the right side, the tonsil has been removed. Entrance into the fourth ventricle (arrow)
88
1 Anatomy
The White Matter 01 the Cerebellum The numerous overlapping, wafer-like segments of the parallelrunning folia that constitute the cerebellar cortex make it difficult to grasp the architectural arrangement of the cerebellum from the surface. The nomenclature of the different parts of the vermian and hemispheric lobes is not easy to memorize. As with the cerebrum, the architecture of the cerebellum can be better delineated by an inside-out approach. The white matter of the cerebellum can be divided into the following identifiable regions (Fig. 1.75). 1. Cortex, sector O 2. Subcortical (folial, sector 1; peduncle in relation to the projecting fibers) 3. Sublobular, sector 2 (peduncle) 4. Lobular, sector 3 (peduncle) 5. Lobar, sector 4 (peduncle) 6. Hemispheric peduncles (superior, middle, and inferior).
The white matter forms a central core, which is much thicker in the medial parts than it is in the lateral area, where it forms a fIattened strip connecting the enlarged lateral portions with each other. From its surfaces, a series of nearly parallel plates or laminae (pedunculi) project towards the surface, and these give off secondary pedunculi. The next laminae are small and short, and finally the subcortical (folial) pedunculi are very short and small (Figs. 1.76, 1.77a-b, 1.78a-c). On MR images, it is possible to see the white matter cIearly and to recognize all three cerebellar pedunculi (superior, middle, and inferior), and the dentate nucleus (Figs. 1.68Bf on p. 77 and 1.69 Ca-c on p. 80). Other nuclei and tracts cannot be distinguished in vivo at present. However, Tz sequence MR images on anatomical specimens after fixation have been able to demonstrate brainstem tracts and nuclei (Armington 1988, Hirsch et al. 1989, Martin et al. 1992). Future advances in MR imaging may provide tract and nuclear identification that will be useful in cases of displacement of these structures by cerebellar tumors. The white matter of the cerebellum (similar to the cerebral white matter) contains several networks of subsystems within the described peduncular are as, and these are shown in Table 1.9a, b.
Fissurae cerebelli I
Stratum granulosum I
Stratum moleculare
Velum medullare rostrale [superius]
I
Posterolateral
fissure Fig. 1.75 Thewhite-matterarchitectureofthe cerebeilum (coronalsections) 1a Subcortical (folial)peduncle AOL Anteriorquadrangular lobule Bi Biventrallobule 1b Sublobularpeduncle ISl Inferiorsemilunar lobule 1c lobular peduncle POl Posterior quadrangular lobule 1d lobar peduncie SSl Superior semilunar lobule 1e Hemisphericpeduncle To Tonsil
Laminae albae Fig. 1.76 The laminae albae (subcortical peduncle) inside the lobulus precentralis cerebeili (from Zenker, ed., Benninghoff, Anatomie, Vol.3. 1985, Munich: Urban and Schwarzenberg, 1985, p. 296, Fig. 15.3)
Topographic Anatomy for Microsurgical Approaches to Intrinsic Brain Tumors
a
I
r
....
b
Fig.1.77a, b The cerebellarfoliae and white matter of the cerebellum
89
90
1 Anatomy
-a
--
b
-
~ 1...
e
Fig. 1.78 a-e The cerebellar folia a Cascading folia within a fissure b, e Folia transgressing sulci
Table 1.9a Contents of the white matter (subsystems)
of the eerebel-
Table 1.9b
Projeetion fiber system of the eerebellum
lum 1. Conneetive fibers a. Assoeiation fibers
b. Commissural fibers Anterosuperior Posteroinferior e. Projeetion fibers Conneet the eerebellum with other parts of the brain and spinal eord. They are grouped into three large bundles or peduncles on eaeh side
Superior peduncle Cerebellomeseneephalic and Cerebellothalamie Spino-eerebellar Middle pedunele Pontoeerebellar Inferior pedunele, reeiproeal Cerebellobulbar Cerebellospinal (see Table 1.9b) Spinoeerebellar
Superior eerebellar pedunele
lar traet Trigeminoeerebellar traet Ceruloeerebellartraet Teetocerebellartraet Middle cerebellar peduncle
Arterial, eapillary, 6. CSF transeerebral
venous pathways?
{ Corpus retisforme
migration
pathways 5. Vaseular network
{
Inferior eerebell,ar pedunele Juxta-
The cerebellum is attached to the mesencephalon, pons, and medulla by three pairs of thick, connecting fiber bundles, the cerebellar pedunculi (found on the ventral aspect of the cerebellum). These are the superior (brachium conjunctivum, with mainly effer-
)
Alterent pathways
}
Pontoeerebellarfibers
Cerebello-olivarytraet Cerebellonueleartraet Cerebelloretieulartraet Cerebellovestibular C",ebello,p;nalt,.c! traet
Elterent pathways
}
Vestibuloeerebellar traet restiform body
Cerebellar peduncles
Elterent pathways
Anterior spinoeerebel-
2. Neurotransmitter 3. Neuroimmune network 4. Neuroglia
Dentatothalamie traet Dentatorubral traet Interpositorubral traet Interpositothalamie traet Cerebellotectal traet Fastigiothalamie traet Fastigiovestibular traet
Conneet adjaeent or more distand foliae. including the vermis. They do not eross the midline Interconneet the two hemispheres
j
Posterior spinoeerebellar traet Olivocerebellar traet Cuneocerebellar traet Nucleoeerebellar tract Retieuloeerebellar traet
Alterent pathways
ent projecting fibers, towards the midbrain and thalamus), middle (brachium pontis, entirely afferent pontocerebellar fibers to the neocerebellum) and inferior (corpus restiforme and juxtarestiforme, with a mixture of afferent and efferent fibers) cerebellar pedunculi (Figs. 1.79a-c, 1.80a-c, see also Table 1.8 on p.81).
Topographic Anatomy for Microsurgical Approaches to Intrinsic Brain Tumors Lingula cerebelli I
II I
91
Colliculus caudalis I I I
I
Pedunculus
cerebri
Sulcus lateralis mesencephali
Pedunculus
Pedunculus
cerebellaris
cerebellaris
cranialis
medius
_n - -- - - - - -
..
--~-- ---_
J
\
\\
~
.~.
Trigonum nervi vagi
----
Trigonum nervi hypoglossi Area retro-olivaris
_n_
/) ..
Taeniaventriculi quarti
---
---Obex
-- - -
Tuberculumtrigeminale - ---Tuberculumcuneatum - - --~-~
-- -Flg.1.79a Thecerebellar peduncles (dorsalview)(alter Leonhardt,Tónduryand Zilles, Rauber/ Kopsch: Anatomiedes Menschen, Stut1gart: Thieme,1987,vol. 3, p.134,Fig.7.16)
Sulcus medianus dorsalis
- - - Sulcus intermedius Fasciculus cuneatus Funiculus lateralis
--
h__
dorsalis
Sulcus dorsolateralis
Lateral geniculate
body
Superior colliculus Medial geniculate body Inferior colliculus
Superior cerebellar (brachium
Middle cerebellar (brachium
pontis)
Inferior cerebellar peduncle (corpus restiformum)
Fig.1.79b The cerebellar peduncles Lateralview (from Brodal, Neurological Anatomy, Oxtord: Oxtord UniversityPress, 1981, p. 298, Fig. 5.3)
peduncle
conjunctivum) peduncle
92
1 Anatomy Mamillary body
Infundubulum Olfaetory nerve Optie nerve
Spinoteetal traet //~
;.,/
. . Medlallemnlseus
-~;; ,/
Laterallemniseus
,/
Ventral spinoeerebellar traet ,/ Trigeminal
traet //
Mediallongitudinal traet
/
Dentate --'nucleus
/
..-:::------
Olive Pyramidal deeussation
, , "
traet
,
,,/
", External areuate fibres
_Olivoeerebellar
""
',_
Restiforrn body Coehlear nerve
Vestibular
nerve
Facial nerve
Fig. 1.79c Lateral view (from Elze, Anatomie des Menschen, Berlin: Springer, 1960, p. 460, Fig. 460)
Meseneephalon
I I 1
Medulla oblongata
--
Fig. 1.80a The cerebellar connections (sagittal view). The cortex and central nuclei are dotted, and the corticonuclear fibers and hatehed (from Zenker, ed., Benninghoff, Anatomie Vol. 3. Munich: Urban and Schwarzenberg, 1985, p. 307, Fig. 15.14) Red Afferent pathways Blue Efferent pathways 1 Superior cerebellar peduncie 2 Middle cerebellar peduncle 3 Inferior cerebellar peduncie
Topographic Anatomy for Microsurgical Approaches to Intrinsic Brain Tumors
Fig.1.80b Cerebellarefferent pathways (from Friek, Leonhardt, and Starck,Human Anatomy, Stuttgart: Thieme, 1991, vol. 2, p. 307, Fig.109) 1 Axonsof Purkinjeeells (eerebellareortieonuelear fibers) 2 Paleoeerebellar nueleus 3 Neocerebellar nueleus(dentate nueleus) 4 Cerebello-olivary traet 5 Inferiorolivarynueleus 6 Oentatorubral fibers 7 Rednucleus 8 Oentatothalamie traet
93
Fig. 1.80c Cerebellar afferent pathways (from Friek, Leonhardt, and Starek, Human Anatomy, Stuttgart: Thieme, 1991, vol. 2, p. 305, Fig.108) 1 Posterior spinoeerebellar traet 2 Anterior spinoeerebellar traet 3 Inferior olivary nueleus 4 Olivoeerebellar traet 5 Vestibular nueleus 6 Vestibuloeerebellar traet 7 Pontine nueleus 8 Pontoeerebellar fibers 9 Cortieopontine fibers 10 Teetoeerebellar traet
ThePredilectiveLocationof Intrinsic CerebellarTumorsand Surgical Planning Inplanningoperativestrategyin relation to tumors of the cerebellum,it is helpfulto consider the cerebeIlum as a three-component organ(Fig.1.81).The cerebeIlum is divided into median, paramedian,and lateral zones. This is based on similar efferent projectionsby neurons located in each zone. Intrinsic lesions arise withinthe white matter of the cerebeIlar hemispheres, predominantlywithineither of the hemispheric pedunculi, or within the I'ermianpedunculus (and grow expansively). These lesions start inthe pedunculusof the foIlowing specific hemispheric and vermian locations.
Hemispheric(a lateral, b paramedian) lb Anterior quadrangular lobule 2a,bPosteriorquadrangular lobule 3a,bSuperioror inferior semilunar lobule 4 Biventrallobule 5 Tonsillarlobule
Vermian (e median) le Centrallobule, culmen (superior) 2e Declive, folium (middle) 3e Tuber, pyramids (inferior) 4e Uvula (inferior) Se Nodules (inferior) In OUTseries, not a single tumor arase in the tIocculonodular area (i. e., arising from archicerebeIlar territory). From an anatomical perspective, hemispheric lesions can thus be classified as arising either superiorly, mediaIly, or inferiorly, and arising either fram the lateral, paramedian (paravermal), or median (vermal) zones of the cerebeIlum, as illustrated in Fig. 1.81. The main sites involved closely foIlow the main arterial supply to the cerebeIlum, as is shown in Fig. 1.82.
94
1 Anatomy Fig. 1b 1e 2a 2b 2c 3a 3b 3c 4a 4b 4e 5
SSL
1.81 The localization of cerebellar tumors (posterosuperior view) Paramedian part of anterior quadrangular lobule Culmen Lateral part of posterior quadrangular lobule Paramedian part of posterior quadrangular lobule Declive Lateral part of superior and inferior semilunar lobule Paramedian part of superior and inferior semilunar lobule Tuber and pyramis Lateral part of biventral lobule Paramedian part of biventral lobule Uvula Tonsil
ISL Si
a Lateral
b Paramedian
e Median
-1 -_/ Fig. 1.82 The territory of the PICA, AICA, and SCA (from Ya:?argil. Microneurosurgery,Stuttgart: Thieme, 1987, vol. 3A, p. 275, Fig. 6.15d) 1 Anteriorand posterior quadrangular lobuli: supplied by the lateral and paramedian branches of the superior cerebellar artery 1a Superior vermian: supplied by the median branches of the superior cerebellar artery 2 Paravermian:supplied by SCA and PICA branches
2a Inferior vermian: supplied by SCA and PICA branches 3, 4 Semilunar lobuli, paramedian and lateral: supplied by SCA,PICA, and AICA branches 5 Biventer-tonsilar: supplied by PICA branches and also by A/CA branches 6 Cerebellopontine: supplied by AICA, PICA, and SCA branches
Topographic Anatomy for Microsurgical Approaches to Intrinsic Brain Tumors
95
VascularAnatomy Arteries Thearterialsupply of the brain is exquisite in its form and degree of complexity (Fig. 1.83). The microsurgical removal of CNS tumorsrequires a knowledge of the brain vasculature in order to
preserve the blood supply to normal structures while extirpating tumors. The blood supply to a tumor must be specifically eliminated, while bypassing vessels are spared. In addition, this knowledge provides the surgeon with an idea of the regional blood supply that a tumor parasitizes in its initial and intermediate
Flg.1.83 A unique latex cast of the human Intracranialand
extracranialarteries (prepared
by Mr A. Lang, Dept. of Anatomy, University of Zurich) a Coronalview b Sagittalview
b
96
1 Anatomy
PICA BA AICA
c=J ~
SCA PCA
c=J
AChA
a
CJ CJ
-
ACA MCA LSA
Fig. 1.84 A uselul review 01 the blood supply through the major cerebral vessels a Vascular territories (cited by Barnett, Stroke, Edinburgh: Churchill Livingstone, 1992, Irom Savoiardo, Ital J Neurol 1986;7:405, with permission). The territory 01the perlorating lenticulostriate arteries (LSA) 01 the middle cerebral artery (MCA) is indicated separately. The posterior cerebral artery (PCA) territory includes that 01the perforating branches 01the posterior communicating artery. Variationsand overlapping' 01
territories are Irequent, particularly in the posterior lossa, between the posterior inlerior cerebellar artery (PICA) and the anterior inlerior cerebe llar artery (AICA), and at the genu 01 the internal capsule ACA Anterior cerebral artery AChA Anterior choroidal artery BA Basilar artery SCA Superior cerebellar artery Fig. 1.84b ~
growth phases. There is a tendency for intrinsic tumors to stay for a long time in their are a of origin, and this may be related to the segmental gyral arterial and venous patterns of the neocerebrum or paleocerebrum, or to the segmental arterial and venous patterns of the basal ganglia and brainstem. It is possible that the vascular system plays a rol e in this confinement, although the means by which it does so are unknown. In addition, the growth pattern of glial tumors in their early and intermedia te phases (in a conical configuration with the apex directed toward the ventricu-
lar system) may be related to this segmental gyral vascular supply. A precise analysis of the brain's arterial and venous systems is beyond the scope of the present chapter. Only general features of the cerebral vasculature, as it relates to the pathophysiology and surgical removal of intrinsic cerebral tumors are presented here. Three illustrations (Savoiardo 1986) show the vascularization are as of the cerebral hemispheres, providing instant orientation as to the difierent arterial areas of supply (Fig. 1.84a-b).
Topographic Anatomy ter Microsurgical Approaches to Intrinsic Brain Tumors
-D
97
If PICA
I d
AICA SCA
I
e
WSCA
I
b
_
.?~~ .'1 P PA LPA
IM':i:"¡¡j DPA
a
a
b
d
e
e
Flg. 1.84b A detailed illustration of the vascular territories in the cerebellumand brainstem derived from CT and MRI studies. a-d axial sectlons,e sagittal paramedian sections, f coronal sections. The course of thepenetrating branches of the brainstem is superimposed on the territaries.The arientation 01the vermian and hemispheric branches of the SCAis indicated in d; the distribution of infarcts in these territories is indicated by shading (cited by Barnett, Stroke, Edinburgh: Churchill LivIngstone,1990, from Savoiardoet al., Am J Neuroradiol 1987; 8:199, with permission)
f AICA DPA LPA PICA PPA WSCA
Anterior inferior cerebellar artery Dorsal penetrating arteries Lateral penetrating arteries Posterior inferior cerebellar artery Paramedian penetrating arteries Watershed area in the deep white matter, mostiy supplied by the SCA
98
1 Anatomy
Arteries within Sulci In general, cortical vessels traverse the free surface of the gyrus with which they are associated. They enter more or less obliquely in the depth of a sulcus, and climb to the surface to be come visible again. However, an artery that runs in the bottom of one sulcus may enter the depth of another communicating sulcus without appearing on the free surface of the initial gyrus. The vasculature within the subarachnoid space of one sulcus may contribute to the perfusion and drainage of several gyri. This is an important surgical point in preventing morbidity (Figs. 1.85, 1.86). Szikla et al. (1977) pro vide detailed information on the relationship between the sulci, fissures, and arteriés, and is highly recommended.
Fig.1.85 The course of a leptomeningeal artery withina sulcus. Note how the vessel supplies the transverse gyri withina sulcus and then exits fromthe sulcus to supply other neighboring gyri
Fig.1.86 An injected dissection demonstrating the course of the left [> central artery branches (cent) withinthe central sulcus (fromSziklaet al., Angiography of the Human Brain Cortex, Berlin: Springer, 1977, p. 9, Fig.13)
Two basic patterns of vasculature exist. One characteristic pattern is for the vessel to dip into a sulcus, give off several branches, and then regain the surface en route to its next destination. The other common pattern is for a cortical vessel to traverse the external surface of a gyrus, enter into a sulcus to supply the gyri on either side of that sulcus, and then continue within the depth of that sulcus en route to another without reappearing 00 the free surface. This particular phenomenon is not unusual at the level of the precentral sulcus, which freely communicates with the frontal sulcus. Thus, wide exposure of a sulcus is important to differentiate between end vessels, transit vessels, and tumor-supplying vessels, and to eliminate only those vessels which directly supply blood to the tumor.
Topographic Anatomy for Microsurgical Approaches to Intrinsic Brain Tumors
BloodSupplyof Cerebraland Cerebellar Tumors
Table 1.10 Intrinsic cerebral tumors and their vascularization in relation to their location
Thevascularityof the neocerebrum and neocerebellum has been describedin many publications, and will not be repeated here. Thereaderisalsoreferredto Vol.1,p. 54-143-intracranial arteries;pp. 144-64-perforating arteries; pp. 165-7-cerebral veins; and Vol. III A, pp. 327-32-venous system, pp. 338-44-blood vesselsof the cerebral cortex; pp. 345-9-blood vessels of the cerebellar cortex. Thearterialandvenoussupplyof each individualsegment of thecerebrum(and cerebellum) is distinctive, in terms of the proportionsderived from leptomeningeal versus perforating arteries (Table 1.10). The vascularization of tumors arising in an individualsegment assumes this same arrangement. Figures 1.87-1.91summarizethe common patterns of leptomeningeal versus perforator supply for cerebral and cerebellar tumors. Table 1.11on p. 102outlines the vascular supply to individual cerebral
99
Arterial supply
Neocerebral and cerebellar Paleo- and archicerebral, cerebellar Paralimbic Limbic Central Nuclei Intraventricular
Venous drainage
Leptomeningeal
Perlorator
+++
(+)
+++
(+)
+++ (+)
+ ++ +++ +++
++ ++
+ + +++ +++
Superlicial
Deep
lobes and gyri.
b
Fig.1.87a The vascularization 01 a neocerebral tumor The majar blood supply is via leptomeningeal arteries, supply Irom the deep perlorating vessels. This pattern throughoutal! neocerebral areas (see also Fig. 1.84 b, p.
(Irontallobe). with a small is the same 97)
Fig. 1.87b Diagrammatic representation 01the vascular supply to a neocerebral tumor. In the case 01a hypervascularized tumor,the deep perlorating vessels can be enormously dilated
100
1 Anatomy
a
b
Fig. 1.88 The vascularization of a mediobasal limbic lobe tumor. The vascularization of tumors in this location is via branches from the anterior choroidal, uncal, and posterior cerebral branches
a b
a
Fig. 1.89 The vascularization of a central nucleus tumor a A right thalamic tumor (medial view). Note the major arterial supply, via branches from the deep perforating vessels (PCOA, P1, P2, P3, A1).
Coronal view Superior view
b
The leptomeningeal supply, if present, is scanty, but the vessels may become dilated in vascularized gliomas b A schematic representation of a
Topographic Anatomy tor Microsurgical Approaches to Intrinsic Brain Tumors
101
b Fig. 1.90. The vascularization of a tumor in the pineal area a A right pineal tumor (medial view). The vascularization pattern of tumorsinthis area is via direct branches from the medial or lateral poste-
rior choroidal, collicular, posterior callosal, posterior cerebral, and pericallosal arteries b A schematic representation of a
CerebellarBlood Supply
vasculature (see Vols. I and III A, B). Figures 1.81 and 1.82 on p. 94 and Fig. 1.91 are given in this volume to illustrate the common pattern of the arterial supply of a cerebellar tumor (Fig. 1.91).
Themicrosurgicalremoval of cerebellar tumors (as in the cerebrum) requires a knowledge of the cerebellar and brainstem
~Flg.1.91 The vascularization of a paramedian cerebellar tumor a A left posterior quadrangular lobule tumor (superior view). Note the extensiveleptomeningeal artery supply (via superior, anterior inferior,
and posterior inferior arteries), and the small supply from the deep perforating vessels (paramedian pontine vessels) b A schematic representation of a (see Fig. 1.84 b, p. 97)
--
102
1 Anatomy
Table 1.11 Vascular supply of cerebrallobes and gyri Lobe
Frontal Superior frontal gyrus Middle frontal gyrus Inferior frontal gyrus Central lobe Precentral gyrus
Postcentral gyrus Paracentral gyrus
Temporal Superior temporal gyrus
Feeders
Cortical drainage
Terminal drainage
A3
Frontal ascending veins Frontal ascending and Sylvian veins Frontal ascending and Sylvian veins
SSS
A 3 and M 3
M3
M 3>A 3
M 3>A 3 A3
ICA, M 1, M 2 and M 3
Central Ascending and descending veins Central descending veins Frontal ascending and descending veins Sylvian, ascending, descending, and laterobasal veins
Middle temporal gyrus
M 3>P 3
Sylvian, ascending, descending and lateralobasal veins
Inferior temporal gyrus
P 3>M 3
Sylvian, ascending, descending and lateralobasal veins
Temporal occipital lateral gyrus Temporal occipital medial gyrus
P 3-anterior choroidal artery P3
Laterobasal vein Mediobasal vein
SSS and sphenoparietal sinus SSS and sphenoparietal sinus SSS, Sylvian veins to sphenoparietal sinus
Lobe
Feeders
Cortical drainage
Terminal drainage
Parietal Superior parietallobule Inferior parietal lobule
A 3, M 3 and P3 M 3 and A 3
SSS
Precuneus
A 3 and P 3
Ascending parietal vein Ascending and descending parietal veins Ascending parietal vein
Occipital Cuneus
P3
Basal gyri
P3
Dorsal gyri
P3
Ascending and descending veins
Insular lobe Insular gyri
M 2 and M 3
Ascending and descending veins
Ascending and descending veins Medial and lateral veins
SSS
Sphenoparietal Sinus, Sylvian vein and vein of Labbe, superior petrosal, sinus Sphenoparietal sinus, Sylvian vein and vein of Labbe, superior petrosal, sinus Sphenoparietal sinus, Sylvian vein and vein of Labbe, superior petrosal, sinus Superior petrosal and tentorial veins Basal vein
Arterial Supply of Central Nuclei and the Internal Capsule Basal ganglia. The basal ganglia can be divided into the caudate nucleus, putamen, globus pa11idus,subthalamic nucleus, and substantia nigra. The majority of the blood supply to a11basal ganglia structures comes from only two sources (Table 1.12a). The caudate nucleus is divided into head, body, and tail. The ventromedial head is supplied by Heubner's artery, which originates from the Al segment of the anterior cerebral artery (ACA) (Perlmutter and Rhoton 1976). The dorsolateral head is fed by the lateral striate perforators of the middle cerebral artery (MCA), which also supply the body. In some cases, one of these two vessels may dominate and supply the whole head of the cau-
SSS, Sylvian vein, or vein of Labbe SSS
SSS & vein of Galen Vein of Galen, sigmoid and transverse Sinus and SSS Temporooccipital vein and Sigmoid Sinus Sylvian Vein Frontal ascend. Temporal vein
SSS: Superior sagittal sinus
date (Marinkovic et al. 1986). The dual perfusion of the tail arises from branches of the anterior choroidal artery (AChoA) and lateral posterior choroidal artery. The major blood flow to the putamen comes from lateral striate branches of the MCA. The ventromedial part and the caudal part are nourished from the recurrent artery of Heubner and branches of the AChoA respectively (GilIilan 1968, Dunker and Harris 1976). The globus pa11iduscan be divided into a lateral and medial region. The medial area receives blood flow via many perforators from the internal carotid (lCA), ACA and AChoA. While the majority of the lateral area is perfused by lateral striate branches of the MCA, contribution also arises from medial striate branches of the ACA and occasiona11yAChoA perforators.
Topographic Anatomy tor Microsurgical Approaches to Intrinsic Brain Tumors The subthalamic nucIeus has three sources: peduncular branchesof the AChoA, the posterior cerebral artery (PCA), and the premamillarybranches of the posterior communicating artery (PCoA) (Haymaker 1976). Similarly,the rostral substantia nigra receives blood from the peduncularbranches of the AChoA and premamillary branch of the PCoA (Duvernoy 1978, Rhoton et al. 1979). The caudal area obtains supply from peduncular branches of PCA, collicular, medial posterior choroidal, and superior cerebellar arteries (SCA). Internalcapsule.Perfusion of the internal capsule is anatomically subdivided.The anterior limb receives a dual supply: the ventromedialportion is supplied by branches of Heubner's artery, whileperforators of the middle cerebral artery irrigate the dorsolateral portion (Dunker and Harris 1976, Rosner et al. 1984, Marinkovicet al. 1987) (Table l.Ub). The dorsal portion of the genu receives blood from the lateralstriate branches of the MCA, and the ventromedial aspect fromperforators of the ACA, ICA, and AChoA. The lateral striate branches of the MCA perfuse the dorsal aspectof the posterior limb (Marinkovic et al. 1987). Perforators of the ICA and ACA supply the part closest to the genu (the rostral part of the ventral portion). The caudal aspect is supplied by branchesof the AChoA. The thalamogeniculate branches of the PCAmayalso feed the posterior limb (Schlesinger 1976). The retrolenticular part is irrigated by capsular branches of the AChoA(Herman et al. 1966, Rhoton et al. 1979).
Table 1.12 a
103
The arterial blood supply of Basal Ganglia (central nuclei)
Caudate Ventral medial head by Heubner artery Dorsal lateral head and body by MCA Entire head can be irrigated by MCA perforators or Heubner artery (in some cases) Tail by AChA, lateral posterior choroidal artery
Putamen supplied
by several groups of perforators Rostral ventromedial part from Heubner artery Most of putamen by lateral striate of MCA Occasionally, caudal putamen-AChA Globus pallidus lateral segment - Rostroventrallateral-medial striate of ACA - Large remainder. later striate of MCA, perforators AChA Medial-internal carotid, ACA, AChA
Subthalamic nucleus -
Peduncular branches of ACha PCA Premamillary of PCoA
Substantia nigra Rostral
-
Peduncular
-
Premamillary of PCoA
branches
of AChA
caudal peduncular branchesof PCA COllicular,medial postchoroidal Superiorcerebellararteries
Table 1.12 b The arterial blood supply of the internal capsule Anterior limb
ArterialSupply of the Thalamus
-
Dorsolateral:perforatorsof MCA
-
Dorsal portion: laterial striates of MCA Ventromedial: perforators of ACA, ICA, AChA
Genu Thethalamusreceives its blood supply from anterolateral, lateral posterolateral,medial, and dorsal arteries. The premamillary branchof the PCoA givesrise to the anterolateral arteries. In additionto irrigatingthe anterior ventral (AV), reticular, ventral anterior(VA),and ventral medial dorsal (VMD) thalamic nuclei, they supplytelencephalic and mesencephalicstructures.These include thecrus,substantianigra, optic tract, posterior caudal and lateral hypothalamus, and the subthalamus (Haymaker 1976, Schlesinger1976).The AChoA provides the lateral arteries to the thalamus.They supply the ventral anterior (VA) thalamic reticular nucleus,the ventral lateral (VL), and occasionally the ventroposterolateral(VPL) and the pulvinar. Outside the thalamus, they supply the crus, substantia nigra, red nucleus, and subthalamicnuclei (Schlesinger 1976). The thalamogeniculate arter¡esserveastheposterolateralarteries and perfuse the VPL, ventroposteromedial(VPM), VL, LP and part of the centromedian nuclei(Graff-Radfordet al. 1985, Caplan et al. 1988). Also, they irrigatethe crus cerebri, the brachium of the superior colliculus, and the medial geniculate body (MGB) prior to supplying the thalamicnuclei(Table l.12c on p. 104). The diencephalic interpeduncular (thalamoperforating) branchescomprise the medial thalamic arterial supply. Prior to reachingthe thalamus,they provide blood fIow to the red and subthalamicnuclei. The thalamic supply goes to the VPM, VPL, medialdorsal,centromedian,parafascicular, AV, VL, LD, and LPM(Schlesinger,1976,Graff-Radford et al. 1985). The dorsal branches arise from P2 as branches from the medialandlateralposteriorchoroidalarteries. Mainlythe geniculate bodies,pulvinar,lateral dorsal, medial dorsal, posterior
Ventromedial: Heubner artery
-
Posterior limb Dorsal aspect: lateral striates of MCA Rostral part of ventral portion. close to gen u - Perf. of ICA - ACA Caudal dorsal and caudal ventral branches of AChA
Retrolenticular part: capsular
branches of AChA
nuclei, anterior and midline thalamic nuclei are supplied (Schlesinger 1976, Zeal and Rhoton 1979). The habenula, pineal body, and midbrain receive branches of the medial posterior choroidal artery.
104
1 Anatomy
Table 1.12e The arterial blood supply of the thalamus Anterolateral artery 1. From premamillary branches 01PCoA a. Premamillary branch belore reaching thalamus (1) Supplies crus, optic tract, substantia nigra (2) Part 01posterior hypothalamus, caudal lateral part (3) Subthalamic area b. Once in thalamus may irrigate (1) Anterior ventral (2) Part 01ventral lateral and medial dorsal nuclei 2. Some authors say rostral thalamus receives Irom a. Perlorators 01internal carotid b. Perlorators 01ACA Lateral arteries 01the thalamus-originating Irom AChA 1. At rostral crus, also supplies: nigra, subthalamic and red nuclei 2. Ventral anterior, reticular, and ventrolateral thalamic nuclei 3. Occasionally irrigate VPL and Pulvinar Posterolateral (thalamogeniculate) arteries supply: 1. Crus, brachium 01superior colliculus, MGB 2. Terminal branches supply: VPL, VPM, part 01centromedian nuclei, ventral lateral, posterior lateral nuclei, some cases paralascicular and medial dorsal nuclei Medial (diencephalic) group 01interpeduncular (thalamoperlorators) 1. Irrigate ventral to thalamus: subthalamic, red nuclei 2. Thalamic: medial dorsal, centromedian, paralascic, VPM, VPL, anterior ventral, ventral lateral, lateral dorsal, lateral posterior 3. These perlorators can perluse part 01the mamillary body, posterior hypothalamic area, third nerve, crus 4. Occasional diencephalic perforators 01thalamoperlorator give off mesencephalic perlorators Dorsal: Irom medial and lateral posterior choroidal arteries 1. Supply geniculate bodies and pulvinar 2. Medial dorsal nuclei, lateral dorsal posterior nuclei, anterior and midline nuclei 01thalamus 3. Medial posterior choroidal: branches to habenula, pineal, midbrain
Arterial Supply of the Hypothalamus The specific blood supply of the hypothalamus and pituitary gland has been studied in detail (Haymaker 1976). The hypothalamic vasculature is dense and has many anastomoses and the richest blood supply of any brain area. The hypothalamus is supplied by arteries derived from the proximal branches of the circle of Willis. The hypothalamic blood supply can be divided into rostral, middle, and caudal divisions. The rostral hypothalamic vessels derive from the ACA, ACoA, and ICA, and form a dense network. The ACA gives rise to the preoptic, commissural, suprachiasmatic and supraoptic arteries (Duvernoy et al. 1969, Haymaker 1976, Marinkovic et al. 1989, 1990). The ACoA may give off branches to the median preoptic and anterior median commissural areas. The supraoptic, paraventricular, and superior hypophyseal arteries may stem from the ICA. The middle hypothalamic arterial division comes from the ICA or PCoA. These give off tuberoinfundibular branches. The ICA also gives off the superior hypophyseal artery (Haymaker 1976). The PCoA and PCA provide branches to the caudal hypothalamic area. Both supply the mamillary region, with the PCA also providing thalamic peforators (Table l.12d).
Table 1.12 d The arterial blood supply of the hypothalamus and the pituitary gland Hypothalamic vasculature (all Irom circle 01Willis) 1. Rostral a. ACA (1) Medial and lateral preoptic artery (2) Right and left commissural artery (3) Suprachiasmatic (4) Supraoptic artery (5) Supply medial and lateral preoptic nuclei, suprachiasmatic nuclei, organum vasculosum, lamina terminalis, supraoptic nuclei b. AChoA (1) Median preoptic (2) Anterior median commissural artery (a) Vascular network in organum vasculosum (b)and lamina terminalis (3) Supply part 01preoptic area, median part 01anterior commissure, portion 01lornix column, septal region c.ICA (1) Supraoptic artery to supraoptic nuclei (2) Paraventricular artery, tuberoinlundibular branches-paraventricular nuclei (3) Superior hypophyseal artery, posterior median commissural artery (a) Runs ventorostral to optic chiasm (b)Anastomoses with other commissural arteries 2. Middle hypothalamic arteries (Irom ICA, PCoA) a. ICA gives tuberoinlundibular artery, superior hypophyseal artery (1) Supply intermediate portion 01 hypothalamus b. PCoA gives off several tuberoinlundibular vessels (1) Supply lateral hypothalamus, later tubo nuclei (2) Arcuate nuclei, ventromedial dorsomedial nuclei (3) Posterior part 01hypothalamic area 3. Caudal hypothalamic arteries (PCoA, PCA) a. PCoA (1) Mamillary branches (2) Premamillary artery (3) Supplies mamillary body and caudal lateral and posterior hypothalamus b. PCA (1) Mamillary branches (2) Thalamic perlorators (3) Interpeduncular perlorators 4. Hypothalamus gets supply lrom many vessels and with rich anastomoses, is well protected Pituitary gland vasculature 1. Inlerior hypophyseal artery a. From C 3-intracavernous portion b. At sella, divides into ascending and descending C. Anastomose around posterior lobe d. Supply posterior lobe 01neurohypophysis e. Connect to superior hypophyseal artery by right and lel! middle artery 2. Superior hypophyseal artery lrom C 4 opthalmic segment a. Collateral branches (1) Optic nerve (2) Chiasmatic branches: ventral, rostral, dorsal chiasm (3) Tuberoinlundibular vessels
Topographic Anatomy for Microsurgical Approaches to Intrinsic Brain Tumors
ArterialSupplyof the Midbrain
lar, medial posterior choroidal, PCA, SCA, PCoA, and AChoA (Duvernoy 1978, Rhoton et al. 1979 Zeal and Rhoton 1979). These branches supply the crus, substantia nigra, and mediallemmscus. The SCA, medial posterior choroidal, collicular, and accessory collicular comprise the lateral arteries. After entering the midbrain at the lateral mesencephalic su1cus, these vessels supply the lateral tegmentum. The laterallemniscus, the central tegmental tract, and the reticular formation are served by these vessels. The posterior arteries are formed by branches of the SCA and collicular artery, and form a plexus overlying the collicular plate (Duvernoy 1978). They enter the dorsal midbrain via the tectum and supply the superior and inferior colliculi as well as the periaqueductal grey.
(Fig. 1.92 a).
The circumference of the midbrain is penetrated by mesencephalicperforators. They can be separated into anteromedial (paramedian),anterolateral (short circumflex), lateral, and posteriorbranches(Duvernoy 1978) (Table 1.13a). Theanteromedial arteries divide into lateral and median subgroups.The medial branches supply the red nudeus and the periaqueductal gray, trochlear, and oculomotor nudei (Duvernoy 1978).The mediallemniscus, substantia nigra, and decussation of ¡hesuperiorcerebellar pedunde are perfused by the lateral group. Theanterolateral arteries to the midbrain are called peduncular branches.They arise from many vessels, induding the collicu-
654
3 2 1 32
31
30 29
í".
, F)
7 8
i;W . l. .,
,Jiu¡.
J
;.
"'\\
..' .......
1
, 28
9 10 11
(
r
-
\
\ Jo..L A
J(lrt-Wff/')\27 ", ",.
"
. _ .. f _..._ ..."'....-.p ___. , ""
11
/
UJIb'W,""":.:n././//1'/21
."'..u"'.....,_ 1lI;/Jlmll
26 25
24
12 13
14
Fig.1.92 The vaseularizationof the brainstem (G. Tondury et al. In: Leonhardt,et al, eds. Rauber/Kopsch: Anatomie des Menschen, Stuttgart:Thieme, 1987,vol.3, pp. 213-15, Figs. 8.35-8.37) a Thearterialbloodsupply of the meseneephalon 1
2 3 4 5 6 7 8 9 10 11 12 13 14
Aqueduet Centralgray matter Superioreollieulus Perlianueleus Dorsaltrigeminallemniseus Centralsegmental faseicle Laterallemniseus Medialgenieulate body Spinothalamietraet Substantia nigra temporopontine fibers Trigeminallemniseus mediallemniseus Cortieospinalfibers
105
15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31
Nucleus ruber Frontopontinefibers Mediallongitudinal fascicle Nucleus of Westphal-Edinger Nucleus of the oculomotor nerve Basilar artery Interpeduncular perforating arteries Superior cerebellar artery Posterior cerebellar artery (P2 segment) Short circumferentialarteries Thalamogeniculate arteries Quadrigeminal arteries Posteromedial choroidal artery Geniculate body Posterolateral choroidal artery Pulvinar Quadrigeminal plate
32 Pineal body
--
106
1 Anatomy
Arterial Supply of the Pons
Arterial Supply of the Medulla
(Fig. 1.92 b).
(Fig. 1.92 e, d).
The pontine blood supply is separated into three groups: anteromedial (paramedian), lateral, and dorsal (Table 1.13b). The anteromedial arteries are derived fram the terminal vertebral artery and basilar artery, and can be further subdivided into middle, rastral and caudal vessels. These arteries supply the paramedian tegmentum, including the fascicles of the pyramidal tract, mediallemniscus, mediallongitudinal fasciculus, reticular formation, and the raphe, para median pontine, and abducens nuclei. The lateral perforators aris~ from SCA, AICA, and long pontine arteries (Duvernoy 1978). They supply the lateral pons including the superior cerebellar peduncle, central tegmental tract, lateral lemniscus, locus coeruleus, motor and principal sensory nuclei of V, abducens nucleus, facial nucleus, superior olivary nucleus, oral pontine reticular nucleus, laterallemniscus, and pyramidal tract. Terminal branches of the SCA constitute the posterior arterial supply to the pons (Duvernoy 1978). They perfuse the superior cerebellar peduncle, the mesencephalic nucleus of the trigeminal nerve, and the locus coeruleus.
Like other are as of the brain stem, the medulla receives circulation from anteromedial (paramedian), anterolateral, lateral, and dorsal arteries (Duvernoy 1978) (Table 1.13c). The vertebral and anterior spinal arteries give rise to the paramedian medullary branches, which supply the pyramidal tract, mediallemniscus, central reticular formation, medial accessory olivary nucleus, reticular formation, and inferior olivary nucleus. The antera lateral arteries perfuse the pyramidal tract and inferior olivary nucleus. The lateral arteries arise fram branches of the PICA, AICA, vertebral, and basilar arteries (Duvernoy 1978) and supply the lateral dorsal medulla, including the inferior cerebellar peduncle, inferior olivary nucleus, spinothalamic and spinocerebellar tracts, spinal trigeminal nucleus, central reticular formation, dorsal motor nucleus of the vagus, nucleus and tractus solitarius, and the hypoglossal, vestibular, cochlear, cuneate, and ambiguous nuclei. The PICA gives rise to the posterior arteries supplying the dorsal medulla (Marinkovic and Milisavljevic 1990). The gracile and cuneate nuclei, area postrema, and vagal, solitary, and medial vestibular nucleus are supplied by these branches.
5 4 3 2 1
28 27 26
,1 6 25
7
8 9 10 11 12
13 14 15 16 17 18 19 20 21 22 Fig. 1 2 3 4 5 6 7 8 9 10 11 12 13 14
1.92 b The arterial blood supply of the pon s Teetospinal traet Mediallongitudinal faseiele Retieular formation Central segmental faseiele Loeus eoeruleus Superior eerebellar faseiele Ventral spinoeerebellar traet Rubrospinal traet Middle eerebellar pedunele Trigeminal nueleus Motor trigeminal nueleus Spinal traet of the trigeminal nerve Laterallemniseus Spinothalamie traet
15 16 17 18 19 20 21 22 23 24 25 26 27 28
23 Trigeminal lemniseus Medial lemniseus Cortieospinal fibers Pontine nueleus Pontine raphe Basilar artery Median pontine branehes Short pontine branehes Anteroinferior eerebellar artery Long pontine branehes Superior eerebellar artery Branehes to the fourth ventriele Fourth ventriele Anterior medullary velum
Topographic Anatomy for Microsurgical Approaches to Intrinsic Brain Tumors
2a
1a
13 37
36
107
30
31 3 4 8 5 6 7 32 12 11 14
17 33 18 19 20 21
22 23 24
25 e
30 29
1a 2 2a 3 4 5 6 7 8 9 10 11 12 13 14 15 16
28
27 26
17 18 19 20 21
22
23 24
25 d
Fig. 1.92c, d The arterial blood supply of the e Rostral,d caudal 13 1 Funieulusgraeilis 14 1a Nueleusgraeilis 2 Funieuluseuneatus 15 2a Nueleuseuneatus 16 17 3 Spinaltraet 01the trigeminal nerve 18 4 Spinal nueleus of the trigeminal nerve 19 5 Solitarynueleus 6 Dorsalnueleus of the vagal nerve 20 7 Nueleusambiguus 21 8 Dorsalspinoeerebellar traet 22 23 9 Rubrospinaltraet 24 10 Trigeminallemniseus 11 Spinothalamietraet 25 12 Ventralspinoeerebellar traet 26
medulla
oblongata.
Mediallongitudinal faseiele Caudal olivary nueleus Medial aeeessory olivary nueleus Olivospinal traet Nueleus of the hypoglossal nerve Mediallemniseus Cortieopontine traet Median medullary branehes Ventromedian fissural artery (from the basilar artery) Anterior spinal artery Inferior pyramidal arteries Anterolateral suleal artery Vertebral artery Olivary artery
27 Posterolateral suleal artery 28 Posteroinferior eerebellar artery 29 Posterior arteries of the medulla oblongata 30 Posterior spinal artery 31 Inferior vestibular nueleus 32 Olivoeerebellar traet 33 Retieular formation 34 Superior olivary artery 35 Posterior olivary artery 36 Ventrieular branehes 37 Teetospinal traet
--
108 Table 1.13a
1 Anatomy The arterial blood supply of the midbrain
Mesencephalic perforators 1. Anteromedial (paramedian) a. Median subgroup (1) Median mesencephalic branches 01thalamoperlorating arteries (2) Enter PPS, go through tegmentum, close to raphe (a)short-medial red nuclei supply (b)Long-toward periaqueductal gray supply: i) Trochlear nuclei ii) III nuclei iii) Ventral part 01periaqueductal gray b. Lateral subgroup-small mesencephalic perlorators (1)Enter midbrain at entry zone 01 III (2) Supply: medial substantia nigra, lateral red nuclei part 01 mediallemniscus, and decussation 01the superior cerebellar peduncle in caudal midbrain 2. Anterolateral (short circumflex) peduncular branches originating Irom collicular, acessory collicular, and medial posterior choroidal arteries, and PCA, SCA, PCoA, AChA a. Supply and penetrate crus, substantia nigra, and mediallemniscus 3. Lateral branches-small numerous perforating vessels Irom: a. Collicular b. Accessory collicular c. Medial posterior choroidal d. Superior cerebellar e. Enter midbrain at level 01mesencephalic sulcus and lemniscal trigone l. Supply lateral tegmentum, laterallemniscus, central tegmental tract, and adjacent reticular lormation 4. Posterior branches-small perlorators lrom plexus over colliculi a. From colliculi, and superior cerebellar artery b. Enter tectum, supply: superior and inlerior colliculi, dorsal periaqeductal gray Major structures supplied 1. Red nucleus a. Mesencephalic paramedian perlorators lrom P1 b. Diencephalic paramedian perforators 2. Substantia nigra a. Rostral (1)peduncular branches 01AChA (2) Premamillary 01PCoA b. Caudal (1) Peduncular branches 01PCA (2) Collicular, medial postchoroidal (3) Superior cerebellar arteries
Table 1.13 e
Table 1.13 b
Arterial blood supply of the pons
Anteromedial (paramedian) 1. Middle: short and long pontine arteries a. Short, to the medial lascicles 01the pyramidal tract b. Long, to the medial part 01the mediallemniscus, MLF, paramedian pontine nuclei 2. Rostral, irrigate a. Paramedian tegmentum b. MLF c. Reticular lormation 3. Caudal: irrigate paramedian tegmentum a. MLF b. Caudal reticular lormation c. Abducens nuclei
Anterolateral arteries lrom paramedian arteries supply 1. Ventral lateral pyramidal tract 2. Pontine nuclei 3. Mediallemniscus
Lateral (many perlorators lrom long pontine artery) 1. SCA perlorators: supply rostrolateral pon s a. Peduncle b. Central tegmental tract c. Oral pontine reticular nuclei d. Laterallemniscus e. Part 01locus coeruleus 2. AICA perlorators irrigate the caudalateral pons, supply a. VII Nucleus b. Superior olivary nucleus c. Laterallemniscus d. Middle cerebellar peduncle e. Portion 01VI nucleus l. Principal sensory nucleus 01V g. Central tegmental tract 3. Lateral pons also supplied by long pontine arteries a. Supply lateral lascicle 01pyramidal tract b. Pontine nuclei c. Laterallemniscus d. Central tegmental tract (motor and principal trigeminal nuclei)
Posterior (dorsal) branches 01the terminal stem 01the SCA 1. Superior cerebellar peduncle 2. Mesencephalic trigeminal nucleus 3. Locus coeruleus
Arterial blood supply of the medulla oblongata
Anteromedial (paramedian) 1. From vertebral or anterior spinal artery 2. Enter medulla at anteromedian sulcus a. Short-supply (1)Mediallemniscus and pyramidal tract (2) Central reticular lormation and medial accessory olivary nuclei, medial inlerior olivary nuclei b. Long-between raphe and mediallemniscus to Iloor 01lourth ventricle (1)Mediallemniscus, dorsal accessory olivary nuclei (2) Central reticular lormation, XII nuclei Anterolateral-supply
pyramidal tract and inlerior olivary nuclei
Lateral lrom vertebral, AICA, PICA, BA 1. Penetrate olive and lateral medullary lossa, ICP a. Supply lateral and lateral dorsal inlerior olivary nuclei b. Spinothalamic and spinocerebellar tracts c. Spinal trigeminal nuclei central reticular lormation d. Dorsal nuclei 01X e. Solitary nuclei and lasciculus l. Part 01XII nuclei vestibular and cochlear nuclei g. Cuneate nuclei, ambiguus, ICP in caudal medulla Posterior-lrom PICA and posterior spinal artery 1. Supply gracile and cuneate 2. Area postrema 3. Caudal parts 01vagal nucleus and solitary nucleus 4. Occasionally medial vestibular nucleus
Topographic Anatomy for Microsurgical Approaches to Intrinsic Brain Tumors
Veins
109
Table1.14 Venous drainage o, the brain
Comprehensivedescriptions of the dural sinuses and superficial anddeepvenoussystems of the brain can be found in Vol 1 and 11 and in the publications of Huang 1964, 1965, 1974, Duvernoy andMatsushimaet al. 1983.For present purposes, we 1975,1978, wJl limit our considerations to the parenchymal white-matter venoussystem,as this relates directly to the location of intrinsic braintumors.
The DeepWhite-Matter Veins of the Brain
1. Corticalareas ~ corticalveins~ superficialmedullary veins~ leptomeningeal veins ~
'
2 . Whlema t tter
Subcorticalareas (1-2 mmdeep) ~ superficialpialveins ....---Deep white matter and central nuclel Ventricular wall Choroid plexus
}
Deep medullary veins~lnternal
}
. Subdependymal velns
I
cerebral
vein . Basilar veln Veinof Galen
3. Transcerebral vein (Kaplan) Transanastomotic vein (Schlesinger 1939, Hassler 1966)
Thewhite-mattervenous drainage can be divided into three main typesof small veins: The short superficial medullary veins, the longerdeepmedullaryveins, and the larger transcereberal anastomoticveins.The superficialmedullaryveins begin 1-2 cm below thecortex and run a straight course through the cortex to the pia, to join the superficial cortical draining veins (Tables 1.14, 1.15) (seealsoVol.III A, 7.1, 7.2, pp. 327-8). Table1.15 Synopsis o, the cerebral veins (from Krayenbühl and Ya9argil 1972; see also Vol. 111 A of the present work Tables 7.1, 7.2, pp. 327, 328)
" Internalcerebral vein
I External(cortical) cerebral veins (drainage) (1)Ascending(superior) cerebral and cerebellar veins
Septalvein
Frontal veins
Precentralveins Parietalveins Occipitalveins Superiorcerebral veins Fluccular
veins
Petrasalsinus
}
Superior longitudinal sinus
{ }
Great cerebral vein Straight sinus Inferior petrosal sinus
Temporo-occipital veins (veinof Labbé) Inferiorcerebellarveins
} { }
)
Insular vein Hippocampal Olfactorial vein Anterior cerebral vein Peduncular vein
'"temal oembat
,el"
Basilar vein
j
:> Great cerebral vein
Ponlinevein
(2)Descendingcerebral and cerebellar veins Superficialmiddle cerebralveins (Sylvian)
Thalamostriate vein Caudate vein Choroidal vein
Internaloccipitalveins Cavernous sinus Sphenoparietal sinus Transverse
sinus
Transverse
sinus
Sigmoid
sinus
Dorsal callosal vein
Superiorcerebellarvein Precentralcerebellarvein Medialfrontoparietalveins Callosal
dorsal veins
j ]
}
Straight sinus Inferior longitudinal
sinus
i
(3)Inferiorcerebralveins Medialand lateral frontobasal veins
Superior longitudinal sinus
Temporobasalveins
Superficial petrosal sinus
Occipitobasalveins
Transverse sinus Great cerebral vein
{
The deep medullary veins similarly begin 1-2 cm below the cortex,but run a long, deep course along the white-matter fiber tracts toreachthe ventricular surface
and join the deep venous sys-
tem (through the subependymal veins). These veins are of very smallcaliber(although larger than the superficial medullary), and unlikesystemic veins, they do not in crease in caliber as tributaries join (at rightangles) along their course (Fig. 1.93). These vessels are not randomly positioned, but are grouped in fan-shaped bundles that converge on the ventricular surface (Fig. 1.94). Theseare the vesselsthat can be identified angiographically as a groupof cIoselypacked, dilated veins adjacent to glioblastomas.
It is noteworthy that in this instance these vessels do not deviate (as a result of tumor mass effect) but have a spray-like appearance, and can be traced to dilated, tortuous subependymal veins. Similar changes in these veins are no! seen in less malignant tumors, metastatic lesions, or meningiomas. With such lesions, the deep medullary veins surrounding the tumor are instead displaced and dilated in an arcuate fashion around the periphery of the tumor. This certainly suggests that highly malignant gliomas produce dilatation without compression, perhaps on the basis of a vascular reactive substance.
110
1 Anatomy
a
Fig. 1.93 The white-matter venous system of the brain. The venous circulation in the white matter,and the connections between the superficial cortical and ventricular veins through the intracerebral and extracerebral anastomotic veins (from Ya$argil, Microneurosurgery, Vol. 111 B, Stuttgart: Thieme, 1988, p. 209, Fig. 4.90) 1 Pialartery 2 Pialvein 3 Basal artery 4 Basal vein 5 Cortical artery 6 Cortical vein 7 Cortical and long subcortical vein 8 Subventricularartery 9 Subventricularvein 10 Intracerebralanatomic vein 11 Extracerebralanatomic vein 12 Great vein of Galen 13 Intravenousanastomosis 14 Junction of the cortex and the white matter
b
Fig. 1.94 Angioarchitecture of the deep cerebral venous system(from Hassler, Neurology 1966;16:507, Figs. 2a, b) a A coronal section through the anterior part of the caudate nucleus. The arrow indicates an anastomotic vein T Terminal vein b A similar venogram from a coronal section at the level of the thalamus, which is drained by direct branches from the internal cerebral vein B Basal vein I Internal cerebral vein T Terminal vein
Topographic Anatomy tor Microsurgical Approaches to Intrinsic Brain Tumors The transcerebral anastomotic veins connect the superficial (cortical)and the deep (ventricular) venous systems. These veins are larger than the medullary veins, but fewer in number (Fig.1.95).Both the deep medullary veins and the anastomotic veinsfollowthe path of developmental migration of the cortical neurons andsupportingastroglialcells (from the ventricular SUffact to the cortex), perhaps deriving their position from this embryologicaldevelopment. It is interesting to note that glial tumors.whichcornrnonlyoriginate in the subcortical region, eventuallyexpandin the direction of the deep white matter veins,
2
111
toward the ventricular surface. Hassler has noted that internal medullary veins from a particular regio n of white matter empty into subependymal veins in a defined anatomical pattern (Fig. 1.96). As a result, one can devise a white-matter topographical schema related to these subependymal venous drainage patterns (Fig. 1.97). It is interesting to note the similarity between the morphology of these white-matter venous drainage zones and the morphology of peritumoral edema on neuroimaging changes that frequently occur with tumors and other diseases of the white-matter. Perhaps the pathogenesis of these changes will be found to
3
4
2 5
6
10
9
8
7
Flg 1.95 The anastomotic veins between the internal and external cerebralvenous systems (coronal view) (from Leonhardt, Tbndury and Zllles,Rauber/ Kopsch: Anatomie des Menschen, Stuttgart: Thieme, 1987 vol.3, p. 222,Fig.8.43) 1 Superiormarginalsinus 2 Externalcerebral vein 3 Medullary anastomotic vein 4 Central semioval vein 5 Medullary vein 6 Superficial middle cerebral vein 7.
8
9 10 11 12 13 14 15
Centralnucleusvein Perforatingvein Deep middle cerebral vein Thalamostriate vein Internal cerebral vein Choroidal vein Central nuclei veins
Fig. 1.96 The drainage area of the deep central venous system (coronal sections). The brain siices were were at 2, 4, 6, 8, 11.5, and 13 cm posterior to the frontal pole. The choroid vein drained a small portion of the white matter adjacent to the choroid plexus, but it was too narrow to be indicated on the diagrams. Note the similarity between the whiter-matter patterns of venous drainage and the shape and location of peritumoral white-matter changes (from Hassler, Neurology 1966;16:507, Fig. 2a, b) B Basal vein I Internal cerebral vein P Vein of the posterior horn S Septal vein T Terminal vein
112
1 Anatomy
relate to abnormalities of these venous networks (Chapter 3, Cases 3.52-3.59 on pp. 243-5). The prolific and exquisite nature of the cerebral venous system was recognized by William Harvey centuries ago, but only now are we beginning to appreciate its real significance (Fig. 1.98a-b).
Fig. 1.97 The centripetal drainage 01 the white matter (horizontal view). Four-lifths 01 the white matter and the central nuclei drain via deep medullary veins to the internal cerebral vein, basilar vein, and vein 01 Galen
Topographic Anatomy for Microsurgical Approaches to IntrinsicBrain Tumors
113
Fig.1.98a The architecture of the anastomoticwhite-mattervenous system. A latexinjectedspecimen (lateral view) (preparation by Mr.A. Lang, Dept. of Anatomy, University of Zurich) b Closedview. Note the fine structure, size, andcourse of the anastomotic white-matter veln2joining the superficial and deep venous systems
b
114
1 Anatomy
Conclusions The combination of improved surgical anatomical knowledge with modern neuroimaging would seem to have solved many, if not all, of the localization problems that traditionally have been uppermost in neurosurgeons' minds when dealing with CNS tumors. This is certainly true to a degree for many extrinsic lesions. In addition, the general localization of intrinsic lesions does not normally pose any difficulty. For this we are most gratefuI to the impressive neuroradiological achievements of the last decade. . At the same time, however, our traditional neuroanatomical views of the cerebrum must be redefined so that we can better understand the three-dimensional teality of micro-surgical anatomy and pathology and make full use of the information displayed on MR and PET images.The direct triplanar capabilities and high resolution quality of modern MR imaging has truly made possible an in vivo evaluation of neuroanatomy with a precision scarcely thought possible eight years ago. Attention to the additional neuroanatomical information that MR imaging can routinely provide is of paramount importance in the optimum microsurgical remo val of tumors. The problem so far has be en how to make useful sense of all the available in vivo anatomical data. At present, much of this available topographical information is, at best, simply not seen, and is at worst grossly misinterpreted, leading to erroneous surgical approaches for many intrinsic tumors. The traditional lobar concept of the cerebrum continues to dominate the diagnostic reporting of neuroradiologists and neuropathologists, when describing gross specimens, as well as neurosurgical planning. Yet triplanar MR images display the anatomy of the hemispheres in more detail. Not only are the grossly discernible features of the brain, such as lobes, gyri, fissures, sulci, and cisterns easily identified, but it is also now possible to identify their substructures, such as the architectural substructure within the white matter, the main pathways of the associative, commissural, and projection systems (Schnitzlein and Murtagh, 1990) of cortical are as, the individual central nudei, and the distinctly different ventricular spaces (and associated choroidal plexuses). In addition, the pathomorphological effects of pathophysiological changes can be recognized in similar detail on MR images. These effects are attributed to anisotropically restricted diffusion with the white matter. The technique may have application in a wide range of neurological diseases and result in better localization of lesions and improved detection of disease (Doran et al. 1990). As a result, our traditionallobar descriptive terms (with reference to frontal, temporal, parietal, occipital), though carrect, are no longer adequate. The resolution capabilities of modern neuroimaging have facilitated the diagnosis and location of lesions within the CNS to the extent that an updated scheme is necessary to undestand the topographic anatomy and its relationship with the underlying white matter. Such a scheme must correspond directly with the anatomical detail that neuroimaging currently displays. In addition, this scheme must overcome the difficulties associated with the tremendous variability in the gyral and sulcal surface anatomy of the cerebrum. A precise correlation between the gyral antomy and the white-matter substructures requires a detailed knowledge of the surface gyral topography. Widely employed functional maps
(Brodmann) in ~urrent use are excellent for electrophysiological studies. For neurosurgical purposes, however, the topographical gyral pattern presented by von Economo and Koskinas almost 70 years ago is recommended (Fig. 1.65a-c, pp. 70-71). We must reassert our knowledge of the gyri and sulci in precise anatomical terms, as we correlate them with ever-improving neuroimaging studies. The anatomical scheme presented in this Chapter is based on the sectoral and peduncular white-matter architecture of the cerebrum and cerebellum. This methodological approach has enhanced our understanding of the surgical anatomy of the white and gray matter, and is the foundation of our approach to intrinsic CNS tumors. It permits exact localization of these tumors based on MR images. It allows more exact determination of operability, and subsequently more precise microsurgical removal. As a result, this anatomical strategy has significantly improved our ability to care for brain tumor patients. This anatomical design may in the future lead to a better understanding of the pathophysiology of CNS tumors. It is interesting to note that each distinct part of the brain (neocerebral, archi- and paleocerebral, central nudei, and ventricular) is not only anatomically and physiologically distinct, but also each has special pathophysiological characteristics. These characteristics probably determine to a large degree the frequency of tumor occurrence in specific locations, the biological activity of tumor growth, especially in its early and intermediate phases (perhaps related to vascular supply or embryological astroglial migration, or both), and location of peritumoral edema. The answers to these questions await future neurophysiological and neuropathological studies based on a solid neuroanatomical foundation.
2
RichardPau!Lohse Thirtyverticalsystematic colourseries""ithred diagonals,1943/1970. x 165cm,Oil on canvas. @RichardPaulLohse Foundation,Zürich
Neuropathology
I
116
2 Neuropathology
Introduction
Historical Perspective Attempts to classify cerebral tumors centered on gross pathological characteristics until Virchow (1847), the founder of cell pathology, identified the glial constituents of nervous tissue and coined the term Nervenkitt or neuroglia.Previously, the concept of epigenesis as a central theory of embryology had been established by von Baer (1828 to 1837), and the cell theory had been introduced by Schwann (1837). Improvements in the microscope at that time enabled Remak (1854), von Koelliker (1859), and Virchow (1858) to confirm that nerve cells and neuroglial cells are the basic units of organization in the nervous system. Subsequently, Virchow gave the name "glioma" (1864) to tumors arising from intrinsic brain tissue, and he developed a histological tumor classification system that was accepted for decades. Better staining techniques and improved microscope resolution around the turo of the century provided the means for revolutionary discoveries in histology. Golgi (1894), von Koelliker (1896), Ramón y Cajal (1908), and Rio-Hortega (1933), were at the forefront in developing an understanding of the cellular components of the nervous system. With this new knowledge, several classification systems aros e, notable among them being those of Tooth (1912) and Ribbert (1918). In 1926, Bailey and Cushing published their landmark monograph A Classification ofTumors ofthe Glioma Group. Their categorization consisted of 14 main groups, and correlated clinical features, histology, and outcome. Based on the identification of 20 cell types, it reflected similarities between the cellular morphology of the tumors studied and the cellular appearance of developing (primitive) cells seen in normal embryogenesis. However, the notion that tumors arise from aberrations in the normal development of the various components of the central nervous system remained controversial. Roussy and Oberling (1932) arrived at a similar classification scheme in their atlas, but were careful to point out that cell resemblances should not be regarded as proof of origin in the etiology of tumors. Other notable contributors to neuropathology during this period were Ostertag (1932), and, again, Rio-Hortega (1934), who focused on cell morphology, and gave comprehensive combination names to tumors. The complexity of Rio-Hortega's classification and its lack of clinical correlation made his scheme impractical. Scherer (1940) advocated complete examinations of brain tumors in serial section after fixation. Willis (1960) summarized contemporary knowledge in an excellent review. The realization that the degree of anaplasia seemed to correlate with malignant behavior was emphasized by Kernohan et al. (1949). Following Broders (1926), who conceived a grading system for epithelial tumors, a classification scheme subsequently proposed by Keroohan and Sayre (1952) did away with all but five groups in the categorization of gliomas (an unwarranted reduction), but, more usefully, also suggested a grading system (with grades 1-4, in ascending order of malignany). The simplicity of this scheme and the universal practicality of the grading system assured its ready adoption. However, with widespread usage, drawbacks have become evident. As with any grading system, the classification is limited by sampling errors to such an extent that
there is little assurance that a single sample is agrade 2 (and not a grade 3). This scheme, with usage, has demonstrated better prognostic correlation in purely astrocytic tumors than in oligodendrocytic lesions. The elimination of other major groups that account for rare (but real) glioma types (mainly of an embryonal character) appears, in retrospect, to have been a mistake. The classification debate has continued over the last forty years, with significant differences aired by Ringuertz (1950), Henschen (1955), Zülch (1949, 1951, 1956, 1971, 1980, 1986), Rubinstein (1970, 1972, 1982, 1985), Russell and Rubinstein (1977. 1989), Burger and Vogel (1982), Voth et al. (1982), Jellinger (1978. 1987), Burger, Scheithauer and Vogel (1991) and conceroing the intratumoral heterogeneity of gliomas (Daumas-Duport et al. (1979). Nevertheless, considerable consensus has been achieved through the efforts of a panel established under the auspices of the World Health Organization. This resulted in the monograph Histological Typing of Tumors of the Central Nervous System (Zülch 1979), which was recently updated following the WHOpanel meeting in Zurich, in Apri11990, and published by P. Kleihues, P. C. Burger, and B. W. Scheithauer, 1993.
The Scope of Modern Neuropathology The pioneering contributions of the German, French, and Spanish histopathologists of the nineteenth century were essential in defining the morphological aspects (both macroscopic and microscopic)of neuropathology. The twentieth century has seen phenomenal growth in the expertise provided by other scientific disciplines, particularly the basic sciences. A greater fundamental understanding of the nature of CNS tumors requires multidisciplinary cooperation on a large scale. Funding for these projectsis enormous. Important clues as to the most beneficial allocation of research efforts continue to come from studies of epidemiology (incidence and prevalence studies, genetic and familial studies, endogenous and exogenous environmental studies, etc.) and biology (molecular neurobiology, neurohistochemistry, tumorcell kinetics, immunocytochemistry, tumor immunology,etc.) (Table 2.1).
Introduction
Table 2.1
The range of neurapathological
features
encountered
in CNS tumors
Epidemiology
Age, Sex Hereditary factors (germ cell mutations) Familial incidence (tumour clustering within lamilies) Genetic lactars (somatic mutations) Prenatal environmental factars, postnatal factars Trauma Radiation Progressive multifocalleukoencephalopathy Multiple sclerosis
Morphology
Location (supratentorial, infratentarial, spinal, or combinations)
Macroscopicmorphology
Supratentorial and inlratentarial
Spinal
Number Size
Microscopicmorphology
Shape Extension Tissue pattern Tissue consistency Translarmation (instant or gradual)
Changed cell marphology Changed ultrastructure Mitosis Proliferation Tumor cell culture
Biology Etiologyand pathogenesis (notknown)
Initiallocus (unicentric, multicentric,or global
Molecular biology signals the imp.artance 01mutations in proto-oncogenes that may result in oncogenes with subsequent neoplastic change. Loss 01narmal tumar supressar gene lunction may be equally importan!. Extrinsic
Intrinsic
Number
Unicentric Multicentric Global
Growthdynamics
Growth pattern Growth kinetics
117
Single multicentric, or global Small<2 cm Moderate 2-4 cm Large>5 cm Circumscribed ar diffuse Expansive ar inliltrative Homogeneous, heterogeneous, or mixed Soft, hard, or mixed From normal cell to tumor cell (WHO Grade 1) From normal cell to anaplastic cell (WHO Grade IV), rapid changes From tumor cell Grade 1to 11,to 111, ar to IV
Parenchymal Vascular Cell kinetics Proliferation Dormant pool
Sone, cartilage, connective tissue, sinus mucosa, vascular tissue, Iymphatic tissue, paraganglial tissue, dural, ar arachnoid Cortical, subcartical, central nuclei, subependymal plate, ependyma, choroid plexus, ar intraventricular Unilateral, ar bilateral (midline lesions) Unilateral, bilateral, cerebrocerebellar, ar cerebrospinal Gliomatosis, meningiomatosis, neurofibromatosis, or Iymphomatosis Diffuse (local, hemispheric, ar global) Demarcated (circumscribed) Preoperative Inactive (dormant) Active Progressive (slow or last rate of growth) Alternating (active-inactive) Regressive
Extrinsic (epidural, intradural ar subdural; epiarachnoidal ar subarachnoidal) Intrinsic (neocerebral, transitional, central nuclei, or intraventricular regions) Mixed: extrinsic ar intrinsic Extrinsic (epidural, intradural ar subdural; epiarachnoidal ar subarachnoidal) Intrinsic (juxtamedullary, intramedullary, or central canal)
118 Table 2.1
2 Neuropathology The range of neurapathological features encountered in CNS tumors (Continuation)
Growth tactics
Spread (migration)
Postoperative Regrowth (recurrent) or growth of residual tumor Local Perifocal Global Expansive Displacement or destruction Infiltrative Insinuation between fiber tracts wTiiiout destruction or transgression of fiber tracts and penetration with destruction Extrinsic Intrinsic
1
"1
Interactions between tumor and surrounding tissue
Interface between tumor and normal tissue Types of membrane or pseudomembrane formation
Cleavage No cleavage (different degrees of adherence) Dural
Arachnoidal
Arachnoidal-pial Glial Ependymal
Alteration
of tumor tissue
Alteration
of vasculature
(vasogenic
factors)
Alterations of normal tissue
Metabolic changes (increased glucose utilization, etc.) Necrosis Cyst formation Lipomatosis Embryonal tissue differentiation Changed arterioles, capillaries and veins (hypervascularization or hypovascu larization) Thrombosis of arterioles Thrombosis of intratumoral veins (typical of glioblastomas) Changes in flow rate (slow flow, or high flow) Changes in direction of flow Angioneogenesis Erosion, destruction, or hyperostosis of bone Changes in dura and arachnoidea Compression, stenosis or occlusion of veins, dural sinuses, or arteries Alterations in CSF dynamics Compression of neural parenchyma
Calcification, ossification
thickening or thinning
obstructive hydrocephalus Edema, displacement, or herniation
Epidural, intradural, or subarachnoid (cisternal) Along parenchymal vascular system Along CSF pathways (intraventricular or subarachnoid) Along: 1) White matter association fibers (short distance fibers e. g., uncinate tract, or long-distance- fi bers-unknown) 2) Projection systems (cerebralunknown, or brainstem-bulbopontomesencephalic system) 3) Commissural systems (callosal system or anterior and posterior commisure-unknown)
Pituitary adenoma, epi-, intradural meningioma, chordoma, chondroma, paraganglioma, dermoid and epidermoid Craniopharyngioma, pinealoma, teratoma, germinoma, neurinoma, dermoid, epidermoid and arachnoid cyst Meningioma, craniopharyngioma, and optic glioma Cavernoma, hemangioblastoma, and glioma Ependymoma, subependymoma, neurocytoma, and choroid plexus papilloma
Introduction
119
Table2.1 The range 01 neurapathological leatures encountered in CNS tumors (Continuation) Biochemicaland pharmacologicaleffects (endocrine or exocrinesecretion, exudation) Changesin immunological activity(humoralor cellular immunity) Combination01different tumortypes (see Table 2.4 onp. 125)
Combination with other diseases
On neural parenchyma On CSF constituents
Focal Global Reabsorption impairment (increased protein)
Within the CNS Systemic Meningioma and neurinoma Meningioma and glioma Meningioma and cavernous Meningioma and choroid plexus papilloma Pituitary adenoma and glioma Optic glioma Aneurism Angioma Occlusive vessel disease Inlectious disease Degenerative disease
Intraventricular subependymal
Epidemiology and Pathogenesis
Biological Activity
Theepidemiologyand pathogenesis of brain tumors is reviewed ina reviewarticle by P. M. Black (1991), from which the following passagesare quoted by permission:
The propensity of the central nervous system to grow neoplasms is unrivaled by any other organ system. The variable expression and wide spectrum of biological behavior observed in these tumors provide constant amazement and unending challenges to those involved in the clinical and research neurosciences.
..New brain tumors develop in approximately 35000 adult Americanseach year ~alker et al. 1985, Mahaley et al. 1989). In children,astrocytomas and medulloblastomas are the most commontumors; in adults, the most common are metastatic tumors, astroglial neoplasms (including glioblastoma multiforme),meningiomas, and pituitary adenomas. There is recent evidence that the incidence of primary tumors among the elderlyis increasing (Greig et al. 1990). More adults die each year of primary brain tumors than of Hodgkin's disease or multiple sclerosis. Malignant gliomas alone account for 2.5% of deaths due to cancer and are the third leading cause of death from cancer in persons 15 to 34 yearsof age (Levin et al. 1989, Salcman, 1990). Thereis little information about the relation of environmental factorsto primary brain tumors. Tobacco, alcohol, or dietary patterns have not been associated with them. Cranial irradiationand exposure to some chemicals may lead to an increased incidenceof both astrocytomas and meningiomas (Schoenberg 1991,Nicholson et al. 1982). Head injury may possibly potentiate meningiomas,but does not appear to cause astrocytomas (Schoenberg1991). Sixteenpercent of patients with primary brain tumors have a familyhistory of cancer (Mahaley et al. 1989). Specific predisposinggenetic disorders include neurofibromatosis, with associated acoustic neuromas, meningiomas, and childhood gliomas;tuberous sclerosis with astrocytomas; von Hippel- Lindau disease with hemangioblastomas; Turcot's syndrome with glioblastomas, medulloblastomas, and colon carcinoma (Schoenberg1991); and Li-Fraumeni syndrome with multiple malignantfamilial tumors, including glioblastoma (Malkin et al. 1990)."
Growth Kinetics Every benign or malignant tumor type retains the ability to change. This may be intrinsic to the cell type of origin (genetically stable or unstable and mutational) or extrinsic (hormonal, local environmental such as growth factor, distant environmental such as radiation, chemical carcinogenic, or viral, or unknown stimuli). The unpredictability of individual tumor behavior makes any estimation of the prognosis in CNS tumors particularly difficult. Frequently, there is significant growth variation over time, with periods of tumor hyperactivity, inactivity, or even regression. The classical morphological changes used to distinguish benign from malignant tumors do not hold quite the same significance in the current appreciation of CNS neoplasms. These histological hallmarks are the presence or absence of nuclear atypia, mitoses, endothelial vessel proliferation, and necrosis. Mitoses and proliferation of vessels are seen in glioblastomas and anaplastic astrocytomas, but also in many "benign" meningiomas. The tendency of meningiomas, subtotally removed, to regrow is far from a "benign" characteristic. Hyperactive behavior without mitoses is observed in some meningiomas, as well as in some pituitary adenomas, dermoids, epidermoids, and low-grade astrocytomas, e. g., pilocytic astrocytoma. There are examples of meningiomas, pituitary adenomas, epidermoids, and craniopharyngiomas penetrating the arachnoid and dura into the epidural space, and even penetrating the pia into the brain parenchyma. Yet, histologically, there is no difference between these more locally invasive lesions and similar noninvasive tumors of identical pathological type found in the same locations. Conversely, some intrinsic tumors, including the classic infiltrative glioblastoma, on occasion remain confined to their area of
--
120
2 Neuropathology
origin for a prolonged period of time, becoming behaviorally malignant and functionally aggressive only in their final phase of exponential growth. Thus, even when the pathology is known, the unpredictability of tumor behavior still creates notorious difficulties and uncertainties in patient management. Future attempts to improve the prognostic yield through an examination of cell kinetics and genetic expression may prove promising in this regard. Cell Kinetics The reader is referred to recent work regarding the reliability of the various methods used to evaluate the proliferative rate of tumors (Schiffer 1991, Black 1991). The monoclonal antibody Ki-67 labeling index has been shown to corre late with the mitotic index of cerebral tumors. It appears useful in differential diagnosis between astrocytomas, anaplastic astrocytomas, and glioblastomas, especially in small biopsy specimens. It corre late s inversely with the duration of preoperative symptoms, but not with survival (Raghavan et al. 1990). Another reliable method of assessing the proliferation rate of a tumor is the in-vitro bromodeoxyuridine (BUDR) labeling index, which seems to correlate both with the clinical course of disease as well as with patient survival (Nishizaki et al. 1990). Other approaches to evaluating cell turnover are techniques that measure increased RNA and DNA activity. Using a silver staining method, the mean number of nucleolar organizer regions (NORs) per nucleus is estimated. NORs are the sites of genes transcribing to ribosomal RNA. An increased number and reduced size are considered signs of malignancy. The mean number of Ag-NORs per cell has been found to parallel the degree of histological malignancy in gliomas, and corre late s well with Ki-67 immunoreactivity (Plate et al. 1990, 1991, 1992) and with survival (Kajiwara et al. 1992). Finally, an analysis of DNA content by flow cytometry has also been shown to be a reliable method of assessing malignancy. A higher proliferative index correlates with poorer survival and increasing malignancy in gliomas, as do higher nuclear antigen p 105 measurements (Appley et al. 1990). Nuclear antigen p 105 increases throughout the cell cycle and distinguishes cycling from noncycling cells. Genetic Expression An intensive study of the genetics of certain individual tumors has provided important insights into possible mechanisms leading to the development of CNS neoplasms. Neuroblastoma is one of the most common malignant tumors in children, exceeded only by leukemia and lymphoma and brain neoplasms. It is derived from neural crest celllines, and undergoes spontaneous differentiation probably more frequently and to a greater extent than any other human cancer (Bolande 1985, Bove 1981). Schwab (1990) provides an exciting review of the clinical debut of oncogene research in an article on the amplification of the myc-n oncogene and deletion of the putative tumor suppressor gene in human neuroblastoma. Human neuroblastoma cells often carry nonrandom chromosomal abnormalities signalling regular alterations. Quite frequent are "double minutes" (DMS), homogeneously staining regions (HSRS) both cytogenetic manifestations of amplified DNA and chromosome 1 p deletions (indicating a loss of genetic informa-
tion). DMS have been shown to be a chromosomal manifestation of amplified DNA, and in neuroblastoma cells, this amplified DNA has always been found to encompass the cellular oncogene myc-n (Schwab 1985). The deletions seen in chromosome Ip appear to overlap at 1p36.1-2 (Weith 1989), indicating a loss of specific genetic material from the tumor cell that may well tum out to be identified as a tumor-suppressor gene. It seems likely that both the amplification of cellular oncogenes and the loss of tumor-suppressor genes play important roles in neuroblastoma. The amplification of myc-n is an indicator of a poor prognosis, even when classical morphological criteria would suggest a better outcome. More aggressive therapy can be offered early to patients with this fatal mark of biological destiny. Of further interest is the existence of a special subgroup (stage IV) characterized by frequent spontaneous regressions. In this group, only 7% showed myc-n amplification, and all underwent tumor regressions. Thus, in neuroblastoma, two abnormalities are consistently found that suggest rapid tumor growth: an excess of amplified DNA (encompassing the cellular oncogene myc-n), and deletions in chromosome 1p (corresponding to a likely tumorsuppressor gene). Along the same lines, it is known that an alteration or deletion of the likely tumor-suppressor gene in the RB1 gene (resident in chromosome 13, band q 14) predisposes not only to retinoblastoma but also to osteosarcoma, small-cell carcinoma of the lung, and breast cancer (Gennett and Cavenee 1990). Chromosome abnormalities are common in neural tumors. Cytogenetic analyses of glioma cultures have demonstrated structural abnormalities on cromosomes 1, 6, 4, 8, 9, 10, 17 and 22 (Bigner et al. 1984, 1986, 1988). Loss of part or all of chromosome 10 has been reported consistently in glioblastoma, with nonrandom alterations in other chromosomes in other glioma grades. Questions continue to be raised as to whether these genotypic changes are secondary to tumor initiation, or whether they are related to tumor progression. Allelic deletions of chromosome 17 p may contribute to the oncogenesis of all malignant grades of astrocytic tumors (lames et al. 1989). Alterations of 17pare also found in many other cancers. Other abnormalities in malignant glial tumors are amplified erb-b oncogene expression with trisomy 7 (chromosome 7 in excess) (80%), monosomy of chromosome 10 (60%), monosomy involving chromosomes 6, 14, or 22, loss of chromosome Y, or gain of chromosomes 19, 20, or X. The commonest deletion (40% gliomas) is on chromosome 9p. In oligodendrogliomas, a loss of chromosome Y, deletions of 1 p, and abnormalities of chromosome 6 (in the 6q band) and of chromosome 11 (in the 11q band) have been found. Many meningiomas show a loss of chromosome 22 (monosomy 22); others have, in addition, coexistent monosomy 8 and 14 and loss of chromosome Y, with deletions of 1p and 11 p. Based on cytogenetic analysis, James (1988) has concluded that, irrespective of the pathologic heterogeneity noted in glioblastoma, this tumor is donal in nature. These findings increasingly support the concept that the loss of a tumor-suppressor gene is important in the pathogenesis of brain tumors, and raise the possibility of ultimately treating the disease by replacing the defective gene (Weissmann et al. 1987). The full significance of oncogene amplification as a predictor for poor prognosis has become clearer with the identification of amplified erb-b 2 in aggressively growing breast cancers (Slamon 1989).
Introduction Other elements are also potentially important in the pathophysiologic development of brain tumors (Black 1991). Many malignant tumors induce the formation of new blood vesseis,increasingtheir own nutrient supply and possibly enhancing their growth (Folkman et al. 1987). Factors that stimulate angiogenesis include acidic fibroblast growth factor, also called endothelial-cell growth factor; transforming growth factors exand ~; angiogenin; tumor necrosis factor ex; and basic fibroblast growth factor (Folkman et al, 1987). Malignant astrocytomas wereimportant stimulators of angiogenesis in a physiologic assay (Brem 1976),and angiogenesis may be an important component in the progression of astroglial tumors. Benign tumors are also dependenton their blood supply; Takamiya et al. (193) could prove the inhibition of angiogenesis and the growth of human nervesheathtumors with AGM-1470 in 348 nude mice (nu/nu). Both benign and malignant tumors may cause edema in the surrounding brain; some appear to secrete factors that in crease vascularpermeability (Bruce et al. 1987). The extracellular matrix producedby malignant astrocytomas may promote the growth andinfiltration of tumor cells; the precise differences between normal cells and neoplastic cells require elaboration (Rutka et al. 1987).Finally, the host immune response to an astrocytoma is indequate(Bullard et al. 1986), partly as a result of the secretion of immunosuppressivesubstances, including prostaglandins and a glioblastoma-derived T-cell suppressive factor now identified as
transforming growthfactor 132 (Fontana et al. 1984).
.
Clearly the biology of tumor growth is multifactorial and highlyredundant (Table 2.2). One can only hope that future treatmentproceduresfor brain-tumor patients will involve the identification of amplified cellular oncogenes that better identify those patients who have a poor prognosis (and who may subsequently require more specific or aggressive therapeutic management.) More recent evidence suggests that mutations of the p53 tumor suppressor gene, also located on chromosome 17, are common in glioblastoma multiforme and represent an escape from the normal cell cycle control (Newcomb et al. 1993).
Table2.2 The regulatory factors involved in tumor growth, with positive or negative effects (C. MolI, Institute 01 Neuropathology, University 01Zurich)
+
Oncogenes
+/-
Chromosomal alterations Tumorsuppressorgenes
Al! levels
Growth lactors
Vascular endotheliar growth
Extracellular matrix
Chemical mutagens
Immunosuppression
121
The Neurosurgeon's Viewpoint Neurosurgeons and neuropathologists view CNS tumors from different perspectives. At a daily, practical level, neurosurgeons deal with tumors as active lesions in symptomatic patients who need solutions. Neuropathologists, on the other hand, are called upon to make decisions on the nature of tumors by processing fixed or cultured tumor material in isolation (albeit with an increasing array of cell-specific diagnostic techniques). It is important to realize that tumors have not only macroscopic and microscopic aspects (which are well known to neuropathologists), but in addition mesoscopic aspects, which come from long-term clinical follow-up, from three-dimensional neuroimaging, from position emission tomography (PET) and from the stereoscopic views most vividly witnessed during surgery. Mesoscopic means the intermediate (or interreacting) three-dimensional characteristics as seen "in situ" through the operating microscope. Neurosurgeons, through their close clinical contacts with patients, have the best opportunity to observe the natural history of tumors in their initial, intermediate, and final stages. During surgery, neurosurgeons are struck by the spectrum of dramatic consequences brought about by the biological nature of tumors. These include: Perilesional changes causing local distortion of the brain. Displacement and herniation of brain tissue. Elongation and distortion of cranial nerves and vessels, which are not usually visualized on even the most modern MR images. Changes in the vascularity or neovascularization of the adjacent brain. Variation in the type and quality of tumor vascularity, with capillary and venous changes. Mention is also not made of the local or general changes seen within the arachnoidal and ependymallayers. Wide variability in the adherence or adhesiveness at the interface between tumor and brain layers (dura, arachnoid, pia, ependyma, choroid plexus). Displacement, invasion into, or adherence or adhesiveness to major arteries, veins, venous sinuses, or cranial nerves. Changes of biological nature from grades I-III to III-IV within a short period of time (2-3 months) or over longer periods (3-4 years). In some cases a recurrence of malignant tumors occurs at increasingly shorter periods but in other cases is seen over longer periods (9-15 months). There is a surprising survival rate in 1% of glioblastoma multiforme grade IV cases. The reason for the regular venous thrombosis seen in glioblastomas and other malignant tumors. The histological diagnosis is not always clear in some cases. Extracranial metastases from malignant gliomas are extremely rare, in the absence of previous surgical intervention. Leifer et al. (1989) collected only two previously reported cases from the literature, in which the glial nature of the metastasis had been confirmed histologically. One of these patients had a multicentric glioblastoma and liver metastases. In Zürich extracranial metastases have not been observed prior to surgery but have been seen in a single case following surgery. Neurosurgeons continue to seek answers to the fundamental questions regarding the altered pathophysiology found around tumors. With the presence of high-resolution MRI scanners, neu-
122
2 Neuropathology
ropathologists have become more interested in the pathological behavior of tumor cells, as they attempt to understand and better interpret what can now be imaged. As a result, the thrust of current neuropathological research is directed at the biological origins of tumors, rather than the physiological events that occur both around them and remote from them. Nevertheless, neurosurgery has cause to be grateful for the pivotal role played by modern neuropathologists in the development of our present understanding of the origin, nature, and growth characteristics of brain tumors. Indeed, in some respects, neuropathology has outpaced neurosurgery in the adaptation and incorporation of basic scientific knowledge into practice (e. g., molecular biology techniques). Neuropathologists made the conceptual leap to electron-microscopic studies twenty-five years ago, and to molecular, biological, and genetic investigations fifteen years ago. The results of the latter have been the establishment of neuropathology as a vital meeting-point between molecular biology, biochemistry, and physiology on the one hand, and the clinical neurosciences on the other. Our experiences in the management and treatment of over 3400 tumor patients has provided us with what we believe are important new insights into the ways tumors progress. Some of these views may appear, at first sight, to conflict with conventional neuropathological teaching, or to contradict the conclusions drawn from well-thought-out projects by experienced neurooncology researchers, neurosurgeons, and neuropathologists. We make no apology for presenting our hypotheses. It is our belief that the adaptation of microsurgical techniques for the removal of extrinsic as well as intrinsic, and malignant as well as benign, CNS tumors will results in a significant improvement in the quality of life for tumor patients. Of course, no claim can be advanced that the inexorable biological destiny of anaplastic meningiomas, neurinomas, astrocytomas, or glioblastoma multiforme tumors can be altered, but some patients survive for a longer time (see Case 2.36 and Vol. IV B). It is our belief, though, that the quality of survival for patients with high-grade malignant tumors, and the quality and duration of survival for a significant number of patients with low and intermediate-grade primary brain tumors, is significantly enhanced by the application of microsurgicql principies and techniques to these tumors.
Our microneurosurgical perspective stimulated our interest in carefully oberserving tumors with respect to their site of origin, initial growth, location, local behavior, extension, and vascularization. This made it evident that many of the traditional tumor concepts still currently in use in brain-tumor surgery and management have outlived their usefulness. It is fascinating to review the classical atlases of brain-tumor neuropathology with our anatomical architectural schema in mind. The white-matter sectorial (peduncular) cascade (as identified on neuroimaging) permits precise gyrallocalization of cerebral tumors. In reviewing the photographs from these works, one is thrilled at the accuracy of topographical identification using this scheme, and amazed at the inaccuracy resulting when the older methods of lobar anatomy where applied (Zülch 1971, Burger et al. 1982, 1991, Russel-Rubinstein 1983, Treip 1978, Janish et al. 1976). At first sight, the topography, micromorphology, and biological diversity of tumors may seem overwhelming. Fortunately, a closer study of CNS tumors reveals distinct structural and behavioral consistencies. Within defined limits, many tumors, regardless of their location, have predictable characteristics (e. g.. topography, lesion consistency, brain interaction, vascularity, pattern of spread, mode of presentation, etc.). Exceptional or unusual tumor behavior is not common. In fact, a recurrent surprise is the consistent absence of specific neurological deficits with large-sized, critically located tumors. In reviewing our series, we have drawn special attention to the surgical-pathological characteristics of tumors in an attempt to correlate preoperative radiological imaging. Our general observations are described in the conceptual synopsis of CNS tumor pathology that follows. More space is givento a speculative discussion on the nature of intrinsic tumors and how they grow, as these tumors continue to present formidable neurosurgical management problems. Indeed, there is a wide spectrum of operative management strategy in neurosurgical practice that stems from a disparity in the interpretation of the fundamental growth characteristics of many intrinsic tumors. Facts supporting our views on how tumors progress are discussed in detail in the following sections. The application of these concepts to the management of individual tumors is reinforced by many examplesgiven in the cases presented at the end of this chapter.
General Considerations: Categorization 01 CNS Tumors The following conceptualization is based on in vivo anatomicalpathological observations of tumors operated on during the past three decades. These ideas have served as a useful guide for the precise localization and removal of these tumors. CNS tumors are generally classified as follows. Extrinsic tumors. These tumors are derived from tissues that encase and support the brain (bone, dura, arachnoid), or from tissues that are strictly outside the brain (pineal and pituitary glands and their associated tissues, vascular structures, cell-residuetissue, craniofacial structures, and the mucosa of sinuses). Extrinsic tumors are extrapial (i. e., outside the pia) in origin and loca-
tion (or, at least, they start out that way). These tumors have special relationships with the bone, dura and arachnoid layers. with the subarachnoid cisterns and intracisternal structures (such as the cranial nerves and basal vessels), and with the gyral surfaces and sulci of the brain. Definition: extrinsic tumors are extrapial or epipial in origino Intrinsic tumors. These tumors are derived from nerve cells and their processes, from nonneuronal specialized support cells (which are collectively called the glia), or from cells of mesenchymal origino Neuronal tumors include gangliocytomas, gangliogliomas, central neurocytomas, and dysembroplastic neu-
Specific Considerations roepithelial tumors.The glialtumors are identifiedby their cellof origin,such as astrocytes, oligodendrocytes, ependymal cells, or microglialcells.The microglial cells are specialized macrophages, andarethe main type of derived CNS immune cell. These cells are associatedwith the uncommon (but increasingly frequent) CNS lymphomas.
Choroid plexus papillomas and intraventricular ependymomasand meningiomas are topographically regarded as intrinsic tumors. Primitive neuroectodermal tumors (PNETS), the groupof tumors that appear to arise from cells resembling the primitive embryonic cells of the developing brain, are also included.Secondary tumors that metastasize into cerebral parenchymavia the blood stream are also intrinsic, by definition. lntrinsictumors originate beneath the piallayer and beneath the "glia limitans"-the basement membrane (formed by a specialgroupof astrocytes) that completely invests the CNS. Consequently,during surgery both the pia and the glia limitans must be crossedin order to remove intrinsic tumors. Definition:intrinsic tumors originate beneath both the pia andthe glia limitans.
123
Some may question the usefulness of another classification scheme that amplifies parts of traditionally useful schemes. Inspection of the tables and outlines that follow will reveal few major differences from the accepted World Health Organization schema (WHO 1990) or with the consideration of tumors as "extra-axial" or "intra-axial." Still, the terms extrinsic and intrinsic, as defined above, are preferred here. The topographical distribution of neuraxis tumors into supratentorial, infratentorial, spinal, and mixed remains valido Thus, while our classification of extrinsic tumors is similar to that in standard texts, we find it useful to further subdivide the conventional classification systems for intrinsic tumors into more precise and surgically relevant neuroanatomical regions. This has enhanced our surgical conceptualization of the presice demarcation of tumor origin, of the tumor's three-dimensionallocation, of its likely pathology, and of the best routes for exploring and removing it.
SpecificConsiderations ExtrinsicCranialTumors ExtrinsicCNS tumors are defined as lesions that originate from outsidethe pia. They are thus extrapial (or epipial), and arise fromtissuesthat are neither neuronal nor glial. They include all tissuesthat encase or support the brain substance, and the pituitary,pineal, cranial nerve sheaths, dysembryogenic tissues, and extrinsicmetastases.We prefer the precise term "extrinsic" to the traditional term "extra-axial." The designation "extraraxial" denotesthat the tumor s located outside the neuraxis, but does not preciselydefine the relationship of the tumor to the brain linings or to the brain parenchyma. The one exception to these guidelines is optic glioma, which arisesfrom subpial glial cells in the optic nerve or anterior tract (i.e.,anterior to the lateral geniculate ganglion). Because it arises fromsubpialglialcells,it should strictly be classified as "intrinsic." However,because its expansion is mainly extrapial, into the arachnoidmembranes and cisternal spaces, it is categorized as "extrinsic."The affinity of certain epidural, extrinsic tumors for particularareas of the skull is well known. Examples from this seriesof casesare presented in Vol. IV B of the present work in detail.
GrowthPattern Extrinsictumors display two basic growth patterns, well circumscribedor diffuse.The biological and genetic factors that govern thegrowthpatterns of particular extrinsic tumors (e. g., meningiomas),haveattracted special interest with the advent of techniquesforproteinsequencing,gene analysis,and the cloning of
cells in culture, and with the development of assays for hormonereceptor status (see the section on biological activity above). Most extrinsic tumors (90%) are well demarcated, noninfiltrative, and surgically completely resectable using microtechniques. They almost always display readily dissectable leptomeningeal planes, which act as a barrier between the tumor and the brain surface. As a result, there is little or no adherence to neurovascular structures in the epidural, subdural, or subarachnoid spaces. Tumors with these characteristics include most meningiomas, neurinomas, chordomas, chondromas, glomus tumors, cranipharyngiomas, epidermoids, adenomas, and optic gliomas. Approximately 10% of "benign" extrinsic tumors, however, exhibit a tendency to grow diffusely and aggressively. This unfavorable tendency is most notable in meningiomas, adenomas, craniopharyngiomas, chordomas, glomus jugulare tumors, epidermoids, dermoids, and optic gliomas (Table 2.11, p. 149). The majority of tumors that exhibit this diffuse, expansive, and destructive growth pattern are located on the ventral side of the brain or spinal cord and their encasements. These tumors may initially grow expansively outwards, but under some unknown biological influence, they often proceed to insinuate themselves into every space available (be it crevice, fissure, cistern, sinus, canal, foramena, etc.). Instead of possessing easily dissectable arachnoid planes, these tumors become densely adherent to nerves, vessels, leptomeninges, and other surrounding structures, almost as if glue had been poured into the area and was in the process of setting. Many extrinsic tumors also have peculiar growth kinetics. Periodic neuroimaging follow-up of nonoperated tumors allows accurate documentation of growth patterns, and provides insight into their highly variable natural history. Some tumors remain inactive for years, causing no symptoms, or producing only minor
--
124
2 Neuropathology
symptoms that rema in stable. Some grow slowly but steadily. Stil! others, however, grow suddenly and aggressively after years of inactivity, penetrating barriers such as the skul!, scalp, sinuses, dura, and arachnoid. In this situation, it is common to find no histological evidence of mitoses and no other changes suggestive of abnormal proliferation. The trigger for this sudden alteration in behavior is unknown. Conversely, spontaneous regression of other tumors is also documented. Thrombosis of veins within a tumor or in the parenchyma immediately adjacent to a tumor is a sign of malignancy. This phenomenon, though common with highly malignant intrinsic tumors, is also seen with maligI).ant meningioma and neurinoma. The thrombosis cannot be explained by the external compression of the veins, as it is found both within and adjacent to the tumor.
Extrinsic tumors are assigned to categories based on their primary area of growth (Table 2.3). These categories subdivide al! tumors orginating outside the CNS parenchyma into four types based on their relationship to the meninges. The relationships of expanding extrinsic masses to the leptomeningeal membranes are important from a surgical viewpoint. It is crucial for the surgeon to visualize the tumor compartment as a separa te space containing the growing mass, so that the dissection planes likely to be encountered at surgery can be recognized. For example, osteomas, chordomas, glomus tumors, and most paranasal sinus tumor extensions into the anterior and middle cranial fossae grow in the epidural space. The dura is invaginated into the brain and surrounding neurovascular structures as they grow.Pituitary adenomas, on the other hand, grow in an
Table 2.3 Extrinsic tumors Epidural Tumors(after Dr. J. Vignaud, Department of Radiology, Rothschild Hospital, Paris) Tumors originating in the skull, spine, or their linings
Cartilage
Chondroma, chondrosarcoma
Embryonal remnants
Chordoma, craniopharyngioma, dermoid, epidermoid Osteoma, osteosarcoma, osteoblastoma, aneurysmal bone cyst, giant-cell tumor, Paget's disease Fibrous dysplasia, fibrosarcoma, ossifying fibroma Multiple myeloma, leukemia, Iymphoma, eosinophilic granuloma Hemangiomas, cavernomas Glomus jugulare tumor Mucoceles, granuloma, cholesteatoma, malignancy Teratoma, angiofibroma, squamous-cell carcinoma, adenocarcinoma, cylindroma, rhabdomyosarcoma, Iymphoma Mucocele, osteoma, nasal glioma, inverting papilloma, esthesioneurobastoma, malignancy of epithelial, neurogenic, or mesenchymal origin Metastatic
Bone
Fibrous tissue Hematologic and reticuloendothelial Blood vessels Paraganglionic tissue Pneumatic cavities Extracranial tumors invading the skull base
Oronasopharynx
Nose and paranasal sinuses
Mixed-invading tumors arising in the intradural, subdural, or subarachnoid areas
Intradural (sella turcica,
sinus cavernosa)
Subdural (intrathecal)
Subarachnoid (cisternal)
Neoplasms invading all tour areas and also extending into the brain or vice versa
Meningioma . Meningiomsarcoma Neurinoma Chemodectoma Encephaloceles Pituitary adenoma Pituitary adenomas, meningioma, epidermoid, dermoid, cavernomas Metastasis Meningioma, angioblastic meningioma, hemangiopericytoma, chondroma, osteochondroma, osteoma of falx, fibrosarcoma, meningeal sarcomatosis, mesenchymal chondrosarcoma Metastasis, Leukemia (chloroma), Iymphoma Craniopharyngioma, neurinoma, epidermoid, dermoid, optic glioma, pineal tumors, lipoma, meningeal carcinomatosis, subarachnoid cysts, Sarcoidosis, melanoma Metastasis Meningioma Neurinoma Chemodectoma Arachnoid cysts Pituitary adenoma Craniopharyngioma Glioma Metastasis
Specific Considerations intraduralcompartment formed by the splitting of the dura into a sellarleaf and diaphragm. Thediaphragma sellae often acts as a restraint on the growth of giant pituitary adenomas, as they expand into the supratentorial compartment. The diaphragm must be transgressed (along withthe arachnoid of the parasellar cisterns) when removal is attemptedsupratentorially. Conversely, a patulous diaphragm can beseen herniating down into the sella after transsphenoidal removal of suchtumors. Similarly, tentorial and cavernous sinus meningiomasarise and grow between the dural sheets. Still other tumors arise in the subdural compartment (such as meningiomas, hemangiopericytomas, osteochondromas, etc.). These tumors pushawayand are invested by the arachnoid layers as they expand,and are thus "epiarachnoidal." Meningiomasof the planum or jugum sphenoidale may presenttheoppositesituation to that discussed above for pituitary adenomas.These tumors can invaginate the chiasmatic cistern and diaphragma sella into the sella turcica, causing the radiagraphic appearance of a giant pituitary adenoma. They can be difficult to differentiate radiographically, but one clue is that supratentorial tumorsherniating into the sella usually do not "fill up" or expand the sella on plain skull films, whereas most giant pituitary adenomas do. Craniopharyngiomas, meningiomas, optic and hypothalamic gliomas, pineal tumors, and epidermoids grow within the subarachnoidspace. Depending on their size, they are confined to one or more cisterns, meaning that one or more arachnoid layers mustbe transgressed prior to removal. Sometumor types appear on the list in multiple categories, indicatingthe diversity of their presentation and the unpredictability of their clinical behavior.
Table2.4 Sites of Predilection
Intrinsic Tumors Predilection The predilection of intracranial and spinal extrinsic tumors to originate from specific anatomical sites is well known (Table 2.4). Intrinsic tumors also demonstrate a specific anatomic and functional territories within the CNS. We have noted the tendency of gliomasto arise in white matter beneath homotypicalisocortex, in the limbic and paralimbic systems, and in the caudate nucleus, thalamus, hypothalamus, and brainstem. The theory that glial tumors have a predilection for specific areas of the brain has been the object of intensive studies. Schwartz (1932, 1936) reported, in 400 cases of gliomas, a predilection for these tumors to arise in the frontal, temporal, and parietal lobes. Ostertag (1932, 1936, 1941, 1949, 1950, 1952) described an ontogenetic tendency of tumors at certain locations. Zülch (1949, 1951, 1956, 1975) studied the relationships between his pathological classification and the frequency of site of origin in over 9000 brain-tumor cases,and was able to confirmthe concept of "preferential sites of origin of brain tumors, in the vast majority of cases," especially for glioblastomas. Interestingly, he concluded that this intrinsic tumor characteristic was of diminished importance, because the technical capability to image these regions had not yet been developed. These observations have been recognized and confirmed in many different parts of the world, but, surprisingly, little additional research has been carried out along the way. Contemporary neurosurgical practice has continued to focus primarily on the histology of the tumor with regard to prognosis, and on whether a
for CNS Tumor Intrinsic
Extrinsic Supratentaríal
Infratentaríal
Spín.al
Chordoma, chondroma Craníopharyng ioma
Adenoma Fíbroma Epídermoid Opticglíoma Pínealoma Glomus tumor Plexus
papilloma
Glíoblastoma Astrocytoma
Olígodendroglíoma Ependymoma
Ganglioglíoma Medulloblastoma Hemangíoblastoma
Cavernoma Lymphoma +++ Verycommon ++ Common + Occurs
+++ (+) ++ +++ +++ ++ ++ +++ +++
-
++ +++ ++ (+)
++ ++ + -
+ ++
(+) (+)
Infratentorial
Supratentoríal Subpallíal
Meningioma Neurínoma
125
-
í. v.
Subpallial
Spinal. íntramedullary í. v.
+
(+) (+) -
-
-
+
+++
-
-
-
-
+++
-
+++
-
-
-
+++ ++ +++ ++ ++
++ ++ ++ -
+
(+) ++
+
++ ++ + +
++ 7
(+)7 ++
(+) ++ (+) ++ ++ ++ +++ ++
++
-
+
+
+
-
-
-
(+)
-
í. v.
Rare Not seen intraventrícular
+ +
126
2 Neuropathology
generous lobar removal (on occasion in functionally important eloquent areas or in so-called "noneloquent" areas) can be accomplished. Unfortunately, this "lobar" concept of tumor location is far too broad to allow any consideration of the cytoarchitecture, myeloarchitecture, and angioarchitecture of the single cerebral or cerebellar gyrus. As pointed out by Russell and Rubinstein (1989), the initial genetic event or "hit" that transforms a cell into a malignancy may be far remo te in time from the eventual phenotypic expression of the tumor. In order for a cell or population of cells to be vulnerable, there must be active replication or the cell must be capable of returning to that state. Therefore, it is reasonable to suppose that a genetic event, occurring during the fetal or early postnatal period, might not be phenotypically expressed until adult life, after subsequent gene tic or epigenetic events have occurred. Knudson (1971), formulating a two-hit concept of cancer in his study of retinoblastoma, provides strong evidence for an initial genetic event resulting in a predisposed cell or cell line, which later, following a further genetic event, progresses to form a tumor. Other types of tumor may also require several genetic mutations before initiation. Russel and Rubinstein (1989) favored the idea of predilection with regard to intrinsic tumors such as the astroblastoma (predilection for paraventricular location), ganglioglioma (predilection for hippocampal regions), desmoplastic infantile ganglioglioma (frontal and parietal predilection), and desmoplastic (cerebellar hemisphere predilection) and vermal (midline predilection) medulloblastomas. The concept of predilection (based on Rubinstein's idea of a neoplastic "window of vulnerability") would stand on firmer scientific ground if the sites and periods of CNS myelinogenesis and prolonged gliocytogenesis (both known to occur well into postnatallife, possibly up to seven years) were precisely known, especially in relation to the individual gyri and subgyri. Precise studies using high-resolution MRI to map the timing of myelination in the fetal and infantile brain, and comparing these results to similarly studied early childhood tumors, would lend support to the concept of a predilection of tumors to originate at the sites of most active cellular turnover during development. Another possible explanation for the predilection of tumors to develop at specific sites is that oncogene initiation promotion, or both, are more likely to occur in phylogenetically more recently developed regions of the brain (e. g., the association cortex), or in older regions of the brain that are constantly active, creating new information (e. g., the memory and limbic system). In his concept of the "triune brain", McLean (1952) traced the evolutionary development of the human forebrain. Each evolutionary brain are a has unique connectivity, chemistry, and function. The "reptilian" brain is composed of a striatal complex of basal ganglia, and is rich in dopaminergic and cholinergic systems. The "paleomammalian" brain is represented by the development of the limbic system, which is rich in opiate receptors. Finally, the "neomammalian" brain emerges as the neocortex increases in volume and complexity (with an associated ascent of the mammalian species up the evolutionary ladder). Although the interconnectivity of regions has increased with evolution from lower animals to the human, each area maintains its identity, and is capable of functioning more or less independently of the other. Within the neocortex, there is also evidence of phylogenetic differences between regions. While they are familiar with the cerebral maps of Brodmann (1908) and others who focused on the unique architectonic fea-
tures of individual brain regions, some neurosurgeons, neuropathologists, and neuroradiologists may be unfamiliar with the work of brain theoreticians interested in the functional organization of the brain as related to regions sharing common architectonic characteristics (Fig. 2.1). Mesulam (1985) used these architectonic differences in the regional cortical layering pattern and divided the entire cortical mantle into five subtypes, which display a gradual increase in structural complexity and differentiation from phylogenetically primitive areas (limbic or allocorticaI), through transitional cortical regions (paralimbic or mesocortical), to neocortical regions (homotypical and idiotypical). Studies of this type that examine regional differences in the neuronallayering patterns do not suggest any associated regional difference in the architecture and biology of glial cells. But there is increasing evidence to show that glial cells have regional differences in their biology, and are not a homogeneous family of cells spread throughout the nervous system. Applying these concepts to intrinsic tumors, we note that certain patterns emerge. Neocerebral gliomas arise most frequently in the white matter subjacent to the heteromodal association areas (homotypical isocortex), less commonly in the unimodal areas, and only rarely beneath idiotypical cortex (primary sensorimotor cortex). Gliomas arising in the transitional zones are also common. more often in paleocerebral than archicerebral zones. Basal ganglia and central nuclei gliomas also have a predilection to arise within the white matter of the nuclei. Thus, pulvinar thalamic, caudate (head), and hypothalamic tumors are seen in decreasing order of frequency. We ha ve observed no instances in which a central glioma arose within the putamen, pallidum, caudate tail, substantia nigra, red nucleus, or dentate nucleus. Cerebellar tumors arise within the anterior, middle, and posterior lobes and the vermis, whereas gliomas of the flocculonodular lobes (archipallium) are not observed. Real intraventricular gliomas most commonly arise fram the septum pellucidum of the third ventricle and from the floor of the fourth. Less commonly, they originate in the lateral ventricular trigone, but no cases have been noted of these tumors beginning in the frontal, temporal, or occipital horns, which are often the sites of extension of periventricular tumors. Those tumors arising from the septum pellucidum may extend into the body of lateral ventricle (see Fig. 2.18 b). Upon initiation, these tumors also demonstrate a predilection to remain within the white matter. We have not witnessed the spread of these tumors by white-matter connecting fibers, but we have seen spread along the CSF in benign gliomas (Case 2.40). Spread along the uncinate and callosal systems can be observed, but not along the projection fibers (optic or auditive radiations), pyramidal radiations, thalamocortical, or corticothalamic, pathways and even not along the optic tracts. This restriction on migration may be related to the vascularization pattern, or to as yet unidentified immunological borderlines. In some situations, surgical intervention or radiation may playa role in the loss of these defense borderlines, leading to diffuse local and perifocal growth. This occurs uncommonly (5%), and it is certainly possible that this change in growth pattern would have occurred naturally in the late phase anyway (and may be not related to surgery). These observations have stimulated many questions about the origin of glial-cell tumors. Why do oligodendrogliomas not occur in the cerebellum or spinal cord and only rarely in the cenI ! jJ
.
Specific Considerations Extrapersonal space
Perhaps our ability to diagnose and remove tumors at an earlier course in their evolution, the future emergence of tumor growth factor suppressor, and the eventual creation of tumorspecific chemotherapeutic agents, will revitalize a wider interest in the topic of predilection.
Primary sensory and motor areas Idiotypic cortex Modality-specific
(unimodal) association
127
areas
Homotypical isocortex High-order (heteromodal) association areas
Initial Growth Pattern
Temporal poi e-caudal orbitofrontal anterior insula Cingulate-parahippocampal Paralimbic areas Septum-s.innominata amygdala-piriform
c. hippocampus
Limbic areas (corticoid and allocortex)
Hypothalamus Internal milieu Flg.2.1 Thecortical zones ot the human brain (from Mesulam, PrincipIesof Behavioral Neurology, Philadelphia: Paris, 1985)
tral nuclei?Is this difference in predilection for glial-cell tumors to arisein certain cortical zones related to the glial-cell density of that
zone?
Is the
embryological
glial-neuronal
matrix
responsible for the predilection of these tumors to remain associated with the neurons they "grew up" with? And where, along this primitive glial network, do the bias tic astroglial cells firstbecome neoplastic?
There is general agreement that astrocytic tumors arise in white matter. In the initial and intermediate growth phases, the tumor nidus grows by expansion, the most actively dividing cells being predominately those at the periphery. Clusters of these tumor cells extend into and separate local surrounding axons, neurons, and glia, splitting the fiber tracts (Fig. 2.2). Low-grade gliomas (generally), anaplastic gliomas (frequently), and even some glioblastomas (occasionally) remain confined within their site of origin for a long time. Littie is definitely known about the factors that limit tumor growth in these early stages. Convincing evidence exists for the presence of both a humoral and a cell-mediated immune response to astrocytic tumors. A specific humoral immune response to their tumors is produced by a majority of patients with astrocytic tumors (Gately et al. 1982). In addition, patients with clinically significant tumor masses, and those with more malignant gliomas, show evidence of impairment of cellmediated immune response (Roszman and Brooks 1982). Prominent perivascular cuffing of lymphocytic T-cells is found in the blood vessels within and around up to 60% of poorly differentiated gliomas.The majority of these appear to be T suppressorcytotoxic cells (von Hanewehr et al. 1984). The significance of these immune-response changes continues to be a matter of intense research activity, as well as a controversial debate. Nonetheless, it appears that early in the developmental growth of astrocytic tumors, there are biological factors that limit growth (tumor-suppressor proteins and an active local immune reaction). At some point, however, two other phases occur, either
b Fig. 2.2 The initial growth pattern of a subcortical glioma a A subcortical glioma nidus expanding between the surrounding white matter fiber bundles. Projection fibers: pink, association fibers: blue, commissural fibers: yellow. Note the separation of fibers by the growing tumor. They are not interrupted b Expansile subgyral tumor growth, with separation of surrounding fiber pathways
128
a
2 Neuropathology
\ ~
\ '-- J
sequentially or concomitantly. First, the tumor changes its behavior (by further mutation) into a more malignant, less differentiated formo Second, the effectiveness of both the local and systemic immune responses becomes progressively impaired. It may be that the more malignant tumors secrete suppressor factors that prevent or limit any further effective host immune response. It is at this very advanced stage that we envision the unrestricted spread of tumor cells away from the main original nidus. Comparisons of serial stereotactic biopsies have suggested a wide area of tumor-field change at this late stage (Kelly 1987). However, just how many of the abnormal cells away from the main tumor mass are migrated tumor cells, as opposed to newly mutated tumor cells or re active astrocytic cells, remains open to speculation. It may also be that interval areas devoid of tumor cells are indicative of some continued efficacy of the local immune response. Another possible explanation for the restricted localization of early astrocytic tumors may be a dependence on the local subgyral or segmental blood supply. We have strong suspicions that tumors not yet capable of secreting angiogenic factors are forced to confine themselves to the areas of their segmental blood supply. With expansive growth in the subgyral region, the' astrocytic tumor may reach a size large enough for severe compression and displacement of adjacent gyri (with effacement of normal sulci) and impairment of venous drainage to occur. This results in distortion, but not distruction, of these gyri (see Fig. 2.8). With contrastenhanced MRI imaging, it can be very difficult, or even impossible, to distinguish a gyrus containing a low-grade tumor from adjacent distorted and perhaps edematous gyri. As a result, a false impression of extensive "lo bar" involvement may be produced when, in fact, the tumor is confined to only one subgyral sector.
I
Tumor growth tends to be directed centrally, down through the sectors of the involved gyrus towards the periventricular matrix (Fig. 2.3). Commonly, with astrocytic tumors, the anatomical integrity of the basal ganglia and thalamus is well respected. Tumor mass effect often compresses and distorts these structures, which then indent and deviate the ventricular walls. Similarly,the central nuclei may be compres sed and significantly displaced, without invasion or functional effect (until very late) (Cases 2.2, 2.30). Our observations favor the concept that astrocytic tumors grow initially from a focus of abnormal cells in white matter, in b
I r
I ..L ~
specific
architectonic
are as that
are
phylogenetically
more
recently evolved. In the early and intermediate phase of their biologicallife, these tumors grow expansively, but remain primarily restricted to their sectors of origino Only later, at an advanced stage, do they extend into other white-matter sectors, usually following either changes in the bioloy of the tumor (more malignant) or impairment of the immune response (against the tumor).
e
\
\./
/ //
n
I
<JFig.2.3 The common growth pattern of a subcortical glioma a A subcortical glioma expanding into the gyral sector (along the segmental vascular and embryological glial pathway) towards the ventricular matrix b Overview of a, illustrating the most frequent pattern of growth towards the ventricular wall (coronal view, right side). Note that the pyramidal shape of this tumor growth is similarto that seen withsegmental infarction (perhaps suggesting that the direction of tumor expansion is somehow related to the local segmental vascularity). (See Vol.1I1B, pp. 43-4) e The pattern of growth into neighboring gyri (and across sulci) This occurs only later in tumor expansion, and is not seen initally
.
Specific Considerations
Localization 'These tumors are considered here under four main headings: 1 neocerebral(subpallial); 2 limbic and paralimbic (paleocerebral andarchicerebral);3 central nuclei; and 4 intraventricular. Neocerebral and Neocerebellar Tumors (Fig. 2.4a, b).
Thevastmajority of glial tumors arise from the subpallial regions ofthebrain in either the supratentorial or infratentorial compartments.Thesetumors arise predominantly in areas of white matter madeup of numerousbundles
of myelinated nerve fibers that
leadeithertowardsor awayfrom, the corticalneurons of the gray matter (Fig.2.2). Only the rarely occurring protoplasmatic astrocytomaarises from the cerebral corticallayer itself (namely ¡hetruepalliumor mantle layer). The thickness of the cerebral cortexvariesfrom nearly 6 mm in the precentral gyrus to 1.5 mm inthedepthsof the calcarine sulcus. It is estimated that the corte x containsover 14 billion neurons, made possible by the gross expansion of the total area of the cortex to 1800 cm2 through the infoldedfissures and gyri.
Neuroglial cells include &strocyte, oligodendrocyte, ependyma, and microglia cells. The neuroglial cells constitute support cells for the neurons, and outnumber neurons by a ratio of ten to one. It is from one or (some times more) of these types of support cells that most subpallial tumors arise. Thus, it is from the fibrillary astrocyte, oligodendroglia, or microgtia of the white matter of the cerebral and cerebellar hemispheres that the more common intrinsic brain tumors originate (astrocytic tumors, oligodendrogliomas, and lymphomas, respectively). Tumors arise much less commonly from neurons (neurocytoma, ganglioglioma), or from the protoplasmic type of astrocyte (which is more prevalent in the gray matter of the brain). Regardless of the initiating factors in the genesis of a tumorcell line, a particular subpallial cell type undergoes transformation and gives rise to a group of cells that grow abnormally. This group of cells most commonly originates within a sector of a gyrus, predominantly in the regio n of the subcortical white matter. As this group of cells replica tes into a recognizable tumor, is maintains a close-knit relationship within the white matter of its gyrus of origin (Fig. 2.4). This special, confined relationship is then preserved throughout its initial and intermediate growth phases. In
Flg2.4 Theinitialsites of glioma formation in the cerebrum. a coronal view,b horizontalview Lightblue Subcortical (most frequent) Oarkblue Subgyral (common) Red/blue Lobar (rare), (or MRI made in the late phase)
129
. t)
Red
Callosal (very rare)
Green
Limbic / paralimbic (frequent)
--
130
2 Neuropathology
our estimation, the gyrallocation of each individual tumor has a considerable bearing on treatment planning, and must be defined as accurately as possible, especia11yif one is to be able to sensibly compare different treatment modalities. (This is clearly important now that so many different treatment options are being tried). Consequently, subpallial tumors should be defined not just in terms of lobes (e. g., frontal lobe, parietal lobe, temporal lobe, Table 2.5 Intrinsic tumors 1. Neocerebral tumors (subcortical, subgyral, gyral, lobar peduncles) a. Frontallobe Superior frontal gyrus Middlefrontal gyrus Anterior,- middle, and posterior parts } Inferiorfrontal gyrus b. Centrallobe Precentral gyrus Postcentral gyrus Superior, - middle, and inferiorparts } Paracentrallobule c. Parietallobe Superior parietallobule Middle parietallobule (angular gyrus) Inferiorparietallobule (supramarginal gyrus) Precuneus d. Occipitallobe Cuneus Superior occipital gyrus Inferioroccipital gyrus Medialtemporo-occipital gyrus Lateral temporo-occipital gyrus (posterior part) e. Temporallobe .. . Superior temporal gyrus Middletemporal gyrus } Antenor, - mlddle, and postenor parts Inferiortemporal gyrus Lateral occipital-temporal gyrus (anterior two-thirds) 2. Paleocerebral and archicerebral tumors (transitional area) a. Limbic(allocortex) Amygdala, hippocampus, subcallosal gyrus, substantia innominata, septal areas b. Paralimbic (temporal pole, fronto-orbital, anterior insular, cingulate, parahippocampus [see also table 1.3c, p. 18]) 3. Central Gray Matter a. Basal ganglia (caudate, putamen, pallidum) b. Central nuclei (thalamus, meta-, epi-, subo, and hypothalamus) c. Brainstem (mesen-, meten-, myel-encephalon) 4. Intraventricular Infratentorial tumors 1. Neocerebellar (subfolial, lobular, lobar, medial peduncle) Posterior quadrangular lobe Superior semilunar lobe Inferiorsemilunar lobe Declive Folium Vermis } Tuber 2. Paleocerebellar (subcortical, sublobular, lobular, lobar peduncles) Lingula,centrallobe Anteriorquadrangular lobe Biventerlobe . Pyramis } Vermls Tonsil 3. Archicerebellar Flocculonodular lobe
etc.), but also in terms of the specific sector of the particular gyrus involved (Table 2.5). The anatomical origin of every tumor (wherever possible) should be precisely defined, on a named gyral basis. Although this presents an apparent cha11enge at first sight, our surgical experience from the pre-MRI era, and a careful study of MRI scans over the last ten years (especia11ycoronal sections) demonstrate that it is routinely possible to define the most likely gyrus oí origin in the vast majority of intrinsic tumors. The range of presentations and patterns of behavior of subpa11ialtumors are characteristic. The site of origin of these tumors within the white matter demonstrates a site "preference" that has be come more evident with the advent of high-resolution CT and MRI. Recalling the division of white matter into subcortical, gyral, subgyral, lobar, and capsular sectors as presented in Chapter 1, we may once again note that most gliomas arise in the subcortical or subgyral sectors of the cerebrum and cerebe11um. The second most frequent site is within the center of gyrus itself. Tumors that originate at deeper levels (e. g., within the lobar sector) are rare (see Fig. 2.9 on p. 134), and tumors that begin at the level oí the hemispheric peduncle (capsular sector) are very rare (see Cases 2.20, 2.21 on pp. 171-2). Initial growth phase. After beginning as a focus of abnormal cellsa tumor nidus-the subpa11ialtumor grows by spherical extension in a11directions. The multiplying ce11sform groups or columns that grow between the fiber tracts in a manner analogous to tha! of a cauda equina tumor spreading between the nerve roots, separating them as it grows. In fact, the initial expansive growth of subpa11ial tumors is analogous to that of a cerebral abscess, so that, as the tumor enlarges, it conforms to the shape of the gyrus within which it originates. Up to about 2 cm in diameter, the tumor mass is commonly somewhat spherical (Fig. 2.2 on p. 127). During this early phase oí growth, invasion and destruction of white-matter fibers (association, commissural, and projection myelinated nerve fibers) do no! usua11yoccur. Presumably at this stage, host tumor-suppressor factors and tumor factors promoting growth are in a steady state. This state of equilibrium, coupled with the plasticity and overlap of functional capabilities inherent within the central nervous system, sheds light on the frequent paucity of signs and symptoms in many patients with subpa11ial tumors. Only when such tumors reach a larger size do they exert direct destructive effects on the fiber bundles, due to invasion and transection. In the earlyphases of growth within some gyri, more often than not it is the geometric shape of the gyrus (rather than any biological constraint) that causes the physical contour of many gliomas to appear bizarre. Until this phenomenon is recognized, the true location of these tumors is easily misinterpreted. The radiological interpretation of these complex three-dimensional structures can be difficult, frequently leading to a misleading impression of more extensive involvement (i. e., lobar rather than gyral see the example cases at the end of this Chapter and the classic neuropathologic atlases). Tumors originating within the gyral sector expand in a somewhat e11iptical shape either centrifugally toward the subcortical sector or centripeta11y toward the lobar sector, or in both directions (Fig. 2.5). However, the prodominant direction of growth is towards the ventricle. With growth, these tumors assume a more pyramidal shape, related perhaps to the white-matter segmental vascular distribution pattern (see Chapter 1, Fig. 1.25a-b, p. 31),
Specific Considerations
131
or to the embryological migration of glial cells (from the periventricular region), or to lobar neurochemical properties (Fig. 2.5). Tumors arising in the lobar sector also grow in direction of the ventricle to resulting in an impressive indentation of the ventricular walls. The ependyma and modified internal glia limitans layers of the ventricle constitute a powerful biological and physical boundary to tumor penetration. Only very rarely do intrinsic tumors penetrate through the ependymal lining of the ventricular wall. The mode of expansion demonstrated by lobar tumors once the subgyral regio n is reached is the same as that for tumors originating there (Figs. 2.5, 2.6).
Fig.2.5 Advancingpatterns01subgyral tumor growth (coronal views) a Thetumorexpandsalong its own axis to the neighboring gyral segment b Expansionto neighboring subgyral white matter libers (advanced stage) e Theprelerentialdirection 01growth is towards the ventricular cavity (earlyandintermediatestages) d Expansion alongthe callosal radiation (advanced stage)
Intermediate growth phase. From its favored origin in the subcortical-subgyral sector, the typical glioma expands centripetally, growing mainly downwards and inwards through the white matter in the direction of the sectors of its own gyrus. Once through and deep to the originating gyrus, growth then proceeds in a predominantly inward and deep direction towards the ventricle (Fig. 2.3). The direction of this growth (though initially in a similar direction to the gyral white-matter sectorial cascade) is most intimately related to the direction and pattern of white-matter segmental vascular supply, neuroglial embryological migration, or chemical and immunological factors within a special functional area, such as the limbic lobe (see Cases 2.9-2.19, 4.1, 4.3, 4.24, 5.16-5.19, and Fig. 1.25). The architecture of the white matter was discussedon pages 27-39 above. The course of projection fibers passing proximally through sections 1,2,3, and 4 (though initially towards the ventricular cavity) is altered medially as they enter the hemispheric peduncle (sector 5) and are distributed to the internal capsule. The course of short associationfibers is through sectors0,1, and 2, and that of long association fibers through 3 and 4 and then along the fasciculi. Commissural fibers pass through all four sectors, but leave sector 4 to enter the corpus callosum. It is apparent that the cascading white matter fiber system is not aligned with the ventricle wall.
Fig.2.6a, b A glioma has expanded towards the ventricle and indented into the ventricularcavity
132
2 Neuropathology
The vascularization of the white-matter sectors (both arterial and venous), the embryological astroglial network (directing neuronal migration) and hitherto unknown biochemical and immunological entities are, however, oriented in this cortical-periventricular axis. (Case 2.1 on p. 154). The embryological astroglial network (that directs neuronal migration during development) may retain the capacity to direct traffic back towards the ventricle, or perhaps the ventricular-matrix astroglial cells retain their mitotic capacity and are somehow responsible for tumor genesis. The answers to these questions await further neuropathological work, but the fact remains that one or both of these entities seem to best explain the patterns of growth most commonly observed in subpal. lial gliomas (Fig. 2.7). Growth in an anterior-posterior or medial-lateral direction can also occur, but these are not the predominant vectors of growth. When significant growth occurs in these directions, a progressive stage of tumor involvement is seen es it sweeps around
the subgyral region of the affected gyrus and into the subgyral regions of adjacent gyri (Fig. 2.5 b). Once two gyri are involved, the intervening sulcus is compressed, and its associated subarachnoid space is obliterated. Further increase in tumor size preferentially expands the subgyral region, altering the pattern of venous drainage and causing obliteration of adjacent sulci. In the final stages of growth, the tumor tends to evert and turn out one or more of these compressed sulci, at which point the tumor can be considered truly "lobar" (Fig. 2.8). There is a predilection area in the frontallobar sector (prelenticular) at the confluence of lobar areas from the temporal and parietal lobes (retrolenticular) (Fig. 2.9 on p. 134). The term "lobar" implies that a tumor involves sector 4 with involvement of more than one gyrus (out of five that normally make up a lobe). Tumors can, in their advanced phases of growth, grow still deeper into the capsular sector (Case 2.25, p. 175), but this is a very late phenomenon.
~,' l.
¡~,:. I
e
_'c:
Fig.2.7 The lollowing cases are Irom the collection 01Prolessor P. Kleihues, Institue 01 Neuropathology, University 01 Zurich, and show the tumor in an intermediate growth phase a A well-circumscribed left medial F 1 (posterior part) glioblastoma. Note the precise delineation 01 the tumor margins
b An oligodendroglioma involvingthe left superior parietal lobules, with compression 01the procuneus medially and displacement 01the middle parietallobules laterally.Note the absence 01inliltrationintothe precuneus
e A well-circumscribed left Irontal lobar type oligodendroglioma, Grade 11.Notethat the surrounding cingulate gyrus and superior Irontal gyriare not involved.The tumor has inliltratedintothe corpus callosum, and there isassociated compression 01the caudate and lateralventricularsystem
d A left lrontal glioblastoma originating in the lobar sector and expanding into the F 1, F 2, and cingulate gyri without inliltration
Specific Considerations
133
Fig.2.7e A well-delineated astrocytoma, Grade 11,within the right Insula.Thetumor appears to be infiltrating within the insula, but there is a well-defined medial border with displacement of the putamen and caudatemedially. Extension of the tumor into the sylvian fissure
f A large, subcortical, right superior temporal gyrus glioblastoma, extending along the arcuate fiber system to the insula. There is associated displacement of the external capsule, claustrum, putamen and pallidum. T 2, T 3, LTOand the parahippocampal gyri are not infiltrated
9 A large glioblastoma of the left fronto-orbital, insular, and T 1 gyri in thelatephase. It extends and infiltrates medially into the putamen and caudate. The sylvian fissure is entirely blocked by the tlimor. The operculararea of F 3 is displaced, but not infiltrated
h A huge, left middle temporal gyral glioblastoma, herniating in a mushroom-fashionthrough the pia membrane. There is tremendousdisplacement of the sylvian fissure and insula upwards, and of the striatum medially
i A circumscribed oligodendroglioma of the left anterior F 1 gyrus, compressing the F 2 gyrus laterally, and the cingulate gyrus medially, without infiltration. However, the lobar sector of the frontal lobe is infiltrated, as well as the callosal body
134
2 Neuropathology
Fig. 2.8 A glioblastoma in the late phase, with its origin in the lobar sector and expansion into the frontal gyri and callosal radiation
Late growth phase. At the late stage, the tumor may become more aggressive (dedifferentiating into a more malignant, hybrid form) and escape the inhibiting host immune response. When this happens, destruction of the white matter surrounding the tumor may occur as a consequence of several processes (invasion, toxic metabolites, infarction, etc.). It is common for the original shape of the tumor to change from a spherical one to a more complex form at this point. The original center of the tumor may become ischemic, leading to cystic changes. The more malignant peripheral areas of the tumor may grow faster in one direction, as a result of a dependency on adequate blood supply, of a need for high concentrations of growth factors,or as a consequence of the density of white-matter connecting fibers. At this advanced stage of tumor growth, invasion along whitematter fiber pathways do es occur. Growth along short and long association fibers may promote true lobar involvement (Fig. 2.5a). Extension along commissural fibers may demonstrate spread into the corpus callosum (frequently at the forceps major and minor) (Fig.2.5d). Extension along the projecting fibers is extremely rare (Case 2.25 on p. 175).
Unusual patternsof spread.Only about one percent of tumors arising in a subcortical sector in the cerebrum or cerebellum present as an exophytic growth on the surface of the hemisphere. This pattern of tumor growth, through the corticallayers, and expanding in a mushroom-like fashion over the neocortical surface, is most commonly encountered in tumors originating from astrocytes and in the late phase of glioblastomas (see Fig. 2.7h). The external glia limitans membrane and pia mater remain intact in this situation, preventing tumor passage into the subarachnoid space. As a result, these tumors very rarely cross the sulci (Fig. 2.10). An exception to this role are insular tumors (anaplastic gliomas) that have a tendency to extend through the piallayer and extrude into the fissural space. In addition, certain benign tumors (piloid astrocytoma) (see Case 2.40) and malignant (glio-
Fig.2.9 Frequent sites of tumors in lobar sector 4: 1 prelenticular, 2 retrolenticular
Fig.2.10 An unusual pattern of subgyral glioma growth (coronal views) Mushroom-like gyral expansion, with compression of the surrounding gyri and overlying gyrus, creating an exophytic mass (but still retained in the pial membrane) that compresses and conceals the surrounding gyri. Note the scanty growth towards the ventricular cavity in this instance
Specific Considerations blastoma)tumors may emerge at presentation in multicentric, disseminated locations; this phenomenon is not well understood. lntrinsic cerebellar tumors have also predilection sites: median (vermian) and hemispheric (paramedian and lateral). Depending on the organization of the cerebellar white matter, theyexpand within one lobule and along the segment of the anterior. middle, 01' posterior lobule. They usually are directed towardsthe periventricular matrix, but occasionally they may also growoutwardslike a mushroom (Fig. 2.10; Cases 3.2, 3.33). Unique characteristics of cerebellar glial tumors. Hemispheric lesionsexpand locally and may extend medially towards the vermis. but rarely infi!trate across to the opposite side. Thus, hemispheric tumors are usually not bihemispheric. Lesions which arisein the vermis (superior, middle or inferior part) may expand bilaterally and give the impression of a bilateral hemispheric lesion. Tumors of the anterior quadrangular lobule do not extend alongthe superior cerebellar peduncle to the mesencephalon. Gliomas arising in the tonsillouvular region do not extend down alongthe inferior cerebellar peduncle. In contrast, lesions of the middlepart of the hemisphere may extend into and along the middlecerebral peduncle to infiltrate the pons, and vice versa (Tomita1986). Summaryof subpallial tumor growth. The physical development of neocerebral and neocerebellar subpallial tumors may be summarizedas follows: 1. Initiation occurs in white matter (as a nidus of abnormal
cells),in either: a) the subcortical subgyral sectors, 01'b) the gyral sectors;less commonly, in c) the lobar sector, and d) the capsular sector. 2. Early growth is expansive, in three dimensions, conformingto the shape of the gyrus. 3. Thereafter, the main direction of growth is centripetal, towardsthe ventricle. 4. Upon reaching the lobar sector, deformity and swelling of adjacentgyri may occur, but still without invasion of these gyri. Localinterstitial edema due to venous outflow obstruction may be marked. 5. Additional growth occurs in an anterior-posterior or verticaldirection,in the subgyral sector, in the gyral sector, 01'later in thelobar sector. This results in compression and, later, obliterationof suIci. 6. There is no invasion by neocerebral gliomas into the al/ocorticaland mesocortical areas (or vice versa), the central Ill/clei,or the ventrieles, until the end stages of growth (see the examplesat the end of this Chapter).
Thephylogeneticrestriction of intrinsic tumors. When tumors are followedup over the long term in patients who refuse surgery, 01' aredeniedit (for whatever reason), another consistent feature is noted.Thereis a tendency of these tumors to stay within their phylogenetic01'architectonic zone of origin and not to invade adjacentterritories with a different phylogeny and with different architectonicand functional features (such as the central, frontal, parietal,occipital,01'temporallobes, or the transitional areas, centralnuclei,01'intraventricular areas). This phenomenon has been observedin many patients who were later proved (often years later)to be suffering from astrocytomas, anaplastic astrocytomas, oligodendrogliomas, 01'other "infiltrative" glial types of tumor. Tumorsin the transitional areas show the same phenomenon.
135
These tumors originate in the white matter beneath allocortex and mesocortex, and their growth and spread remains almost entirely within the limbic 01'paralimbic system (as if there were a "barrier" prohibiting their spread into the central nuclei 01'into the neocortical zones). These interesting tumors are discussed in further detaillater in this Chapter. Peculiarities of the site of origin of intrinsic tumors. It is rare for gliomas of the cerebral pallium to originate in the white matter beneath highly specialized cortical areas such as the primary motor area, the primary sensory strip, the primary auditory area, or Brodmann areas 1,2,3,4,22, and 44. Gliomas in these areas present diffuse growth tendency (Cases 2.25-2.26 on p. 175). We have seen only one case in which the tumor apparently arose from the primary visual area (area 17), but autopsy revealed the origin to be area 19. There is a striking prevalence of glial tumors arising in the unimodal or heteromodal association areas of the cerebral pallium. In the cerebellum, they arise most commonly in the anterior, middle and posterior lobes and vermis. In our series, there are none that arise from the flocculonodular lobes (archicerebellum). It is tempting to explain this phenomenon on the basis of volume (i. e., association areas >primary sensorimotor areas 01' frontal >temporal >parietal >occipital). This volume theory is commonly put forward, yet it seems far too simplistic. Perhaps the explanation is based on the fact that those are as phylogenetically more recent are more susceptible to soma tic mutations than the more primitively envolved areas. We await further neuropathological information for definitive answers to this intriguing observation. Final observations. Our observations concern the initial site of origin of tumors, and the pattern of their expansion, development, and progression. As stated above, low-grade gliomas, anaplastic gliomas, and sometimes even glioblastomas demonstrate a propensity to remain localized within their initial sector of origin (subcortical, gyral, subgyral), confined there by little-known factors for a long time. Clear infiltration from subpallial and pallial areas into transitional 01' central nuclear areas 01' vice versa does not occur. In addition, early, intermedia te, and even late growth phase gliomas never extend along the long association and projection pathways, such as the superior or inferior longitudinal, the frontooccipital, occipitotemporal, or cingulate fasciculi, the optic 01' auditory radiations, 01' the sensorimotor pathway systems. Similarly, the short and long commissural pathways, such as the anterior and the hippocompal commissures, only very exceptionally transmit tumor to the opposite side. Only about one per cent of malignant tumors extend through the corpus callosum (forceps major and minor), these having arisen either in the lobar peduncles of the frontal or parietal lobes 01' within the genu 01' splenium of the corpus callosum, and having expanded bilaterally. Infiltration across the middle portion of the corpus callosum is extremely rare (Case 2.38 on p. 184). These observations are even true for most recurrent tumors, the vast majority of which are primarily located at their site of origin, regardless of time interval 01' prior adjuvant treatment therapies. We continue to ask ourselves why these limitations occur, what the factors are that, regulate the growth (at least for a while) of most glial tumors so well, why the general belief that these tumors are infiltrative at any early stage is not borne out by our observations, and why indeed the vast majority of gliomas do not metastasize outside the CNS.
136
2 Neuropathology
Tumors in Limbic and Paralimbic Areas lntrinsic tumors arising in the transitional paBial regions are unique, and deserve special recognition for their unusual range of presentation and behavior. These tumors arise in white matter adjacent to the allocortex or mesocortex, areas that are more commonly known as the limbic and paralimbic cortex. The phylogenetic, embryological, and architectonic features of these regions differ from the remainder of the cerebral hemispheres. The allocortical structures are phylogenetically primitive, and appear early in the evolution of mammals. The number of cortical layers ranges from none to two. I.nman, the allocortical structures include the septum, subcallosal gyrus, substantia innominata, amygdala, and hippocampus. The mesocortical structures have three to five layers, and include the temporal pole, caudal orbitofrontal cortex, insula, and cingulate and parahippocampal gyri. In addition, the operculae of the frontal, frontoparietal, and temporallobes that abut the Sylvian fissure also demonstrate a transitional five-Iayered cortex (Figs. 2.11, 1.54, Table 1.3c on p. 18). The vascular anatomy of the limbic areas differs from the other transitional areas. lnstead of receiving blood exclusively either from the leptomeningeal vessels or from the perforating vessels, allocortical limbic areas receive a dual arterial supply from both (though more from the perforators). The arterial supply to the paralimbic are as is principally from the leptomeningeal arteries. The limbic areas drain into the deep cerebral venous system, whereas the paralimbic territories drain into both the deep and superficial system (Table 1.10 on p. 99).
Intrinsic transitional tumors originate and grow in a fashion similar to glial tumors in neocortical are as. Initial growth begins as a tumor nidus, gradually expanding in a spherical fashion. Later, this shape often expands to appear oval. As previously described, the predominant tendency is for these tumors progressively to take the form of a cone, with the base representing the initial spherical site of origin and the apex representing growth towards the ventricular system. However, with large tumors, the C-shape of the limbic system makes the identification of the nidus of origin difficult. Still, it is clear that these tumors remain confined within this system, and the vast majority respect the more recently developed regions of the cortex (neocortex), central nuclei, and ventricles. The reasons for this striking location remain open to speculation. It is worth repeating that these transitional glial tumors have a remarkable tendency during initial and intermediate growth, and expansion, to be confined to the white-matter beneath allocortical and mesocortical zones. It seems as if this group of tumors has an "affinity" for these phylogenetically primitive areas, an idea which was introduced by Filiminoff (1947) and later developed by Yakovlev (1959). Spread to adjacent neocortical (six-Iayeredisocortex) white matter is unusual, and tends to occur late (it at all) in the course of the disease. The pattern of growth within the allocortical/ mesocortical zones shows some very interesting sequences (Table 2.6, Fig. 2.12).
a
b
Fig.2.11 The divisionof the telencephalon (after Leonhardt, Tóndurg and Zilles, Rauber/Kopsch: Anatomíe des Menschen, Stuttgart: Thieme,1987,vol.3, p. 44, Fig. 13.28) Red Neocorticalareas Yellow Paleocortical areas Oarkblue Archicorticalareas Lightblue Periarchicorticalareas
a
Coronal view through the striatum
b Coronalview through the mid-thirdventricle
Specific Considerations Table2.6 Patterns of growth in limbic and paralimbic tumors 1a b e
d e f 9 2a b e 3 4 5
Temporalpole (mediobasal) Amygdala Hippocampus Uncus Ammon'shorn Dentate Parahippocampus Indusiumgriseum Fornix Mamillarybody Septalarea Cingulum,- anterior part Cingulum,- middle part Cingulum,- posterior part Insula Subcallosaland Ironto-orbital (in combination with types 1-3 above) Global(includes all 01the above types)
Patternsof growth. Type 1 tumors: region of the mediobasal temporalarea. A tumor arising in of one these areas (1a, 1b, or 1e) remainsrestrietedto this one area for a long time (initial and intermediategrowthphases). It remains reeognizably distinet and discrete,whilegradually expanding. Rarely, a fusiform tumor expan-
sion involving all three regions (1 a, 1 b, and 1e) may rapidly oceur on sequential seanning. In exeeptional cases, a tumor may spread up along the isthmus into the eingular gyrus (Cases 2.12, 4.1, 4.3, 4.24,5.17,5.18). Tumor growth does not proeeed along the fimbria to the fornix (1 e) or along the indusium griseum (1 d) to the eallosum. True fornieal tumors (as distinet from third ventrieular tumors obstrueting the foramina of Momo), seem to arise bilaterally (two observations of our own. Tumors in the 1f region, arising in the mamillary body, remain loeally eonfined and show no propensity of extend along the mammillary-thalamie fibers. Similarly, tumors that arise outside the limbie and paralimbie regions-for example, from the lateral temporo-oeeipital or medial temporo-oeeipital gyri, or from the superior, middle, or inferior temporal gyri-do not invade the limbie or paralimbie systems, or this only oeeurs in the late phase of tumor growth (Fig. 2.7h on p. 133). Type 2 tumors: cingular tumors. These arise in the posterior, middle, or anterior seetions of the eingulum. 1. Posterior eingular tumors (2e) arise at the base of the preeuneal regio n (often ealled the parasplenial region), and may extend up to the preeuneus. They do not extend bilaterally, as true splenial tumors frequently do. If there appears to be bilateral extension, it is due to herniation of exophytie parts of the tumor aeross the midline (Case 2.19).
/
Flg.2.12 Corticaland nuclear connections01the limbic-paralimbic system(sagittalview) 1a Temporalpole 1b Amygdala with the stria terminalis 1e Uncus/hippocampus parahippocampus 1d Indusiumgriseum 1e Fornix 1f Mamillary body 19 Septalregion 2a Anteriorcingulum 2b Middle cingulum 2c Posteriorcingulum 3 Insula 4 Fronto-orbital area Is Isthmuscinguli MTO Medial temporo-occipital gyrus
137
138
2 Neuropathology
2. Middle cingular tumors arising along the midsection of the cingulum (2 b) may expand into the paracentral areas (Case 2.18 on p. 169). 3. Anterior cingular tumors (2a) either stay well isolated within the cingulum, or expand downwards around the genu of the corpus callosum into the subcallosal and paraterminal areas, and later into the fronto-orbital areas (Case 2.16 and 2.17 on pp. 167-8). Type 3 tumors: insular tumors. These are also seen to arise from three sections, the anterior (a), middle (b), and posterior (c). 1) A tumor may arise in, and remain well confined to, one section a (anterior), b (middle), or c (posterior).
2) It may expand to occupy combinations of sections: a/b, b/c, or a/b/c. 3) Other tumors may expand into the opercular are as, so that: a) 3 a or 3 b types expand more in the frontal opercular region; b) 3 c types expand more in the temporal opercular areas (the gyrus transversus and gyrus circumflexus); and c) 3 a types may also present as a combined frontoinsulotemporal type (see Cases 2.13, 2.14, and other cases in Vol. IV B, chapter on insular tumors). Type 4 tumors: fronto-orbital, parolfactory, and septal types.The tumor expands within the fronto-orbital, parolfactorial gyrus, innominate substance, and basal parts of the pallidum (Fig. 2.13, Case 2.17, p. 168).
Fig.2.13 In our clinical experience, some limbic tumors were found exactly within the fronto-orbital areas shown here (adapted, with permission, from Alheid et al. in Paxinos, Human Nervous $ystem, Basal ganglia, 1990, pp. 491, Fig. 19.7) a A direct print of an acetylcholinesterase-stained coronal section from the Macaque brain. Note the continuation of the striatum to the base of the brain in the region of the accumbens (Acb) and olfactory tubercle (Tu), which together form the ventral striatum. Numerous cell bridges across the internal capsule indicate the continuity between the caudate and putamen. Internal variations in the intensity of the acetylcholinesterase staining within the striatum apparently represent a compartmentalization of this latter structure that is thought to result from a mosaic of afferents that end in terminal clusters
Fig. 2.13 b The relationships between the basal ganglia, extended amygdala, and magnocellular corticopetal forebrain cell groups (including the diagonal band complex and the basal nucleus of Meynert). Observe that caudomedial portions of the nucleus accumbens may represent a mixed territory of striatal and extended amygdaloid elements. Acb, nucleus accumbens; B, basal nucleus of Meynert; BL, baso lateral amygdala; B8T, bed nucleus of the stria terminalis; CeMe, centromedial amygdala; OB, nucleus of the diagonal band; GP, globus pallidus; VP, ventral pallidum; \ \ \ \ \, caudate, putamen, and ventral striatum; / / / I 1, extended amydala, stippled areas represent corridors occupied by neurons of the magnocellular corticopetal cell complex.
a
----
Specific Considerations Type5 tumors,involvingcombinations of the above three regions andthefronto-orbitalarea, are relativelycommon.These are: 1) 4a Fronto-orbital, anterior insula, and temporomediobasal(temporal pole + amygdala + uncus). 2) 4b Frontoorbital, anterior insula, and parahippocampus. 3) 4c Frontoorbital, mediobasal temporal, amygdala, uncus, hippocampus,and parahippocampus (Case 2.15 on p. 166). 4) 4d Bilateral extension: temporal pole, insula, fronto-orbital.septalareas, anterior commissure to opposite fronto-orbital, insula,and temporal poleo(Case 5.19). Tumorsinvolving the totality of the transitional system are seen.Al! of the above types (1-4) are involved. These patients mayhavea certain degree of intellectual impairment and memory butwedidnot observethe Klüver-Bucy syndrome in any deficits, case.The tumors arising in transitional areas in their initial and intermediatephases do not spread to adjacent neocerebral areas, suchasthe superior,middle,or inferior temporal gyri, the superior.middle or inferior frontal gyri, or the central or occipital regions.In addition, these lesions do not grow into the basal ganglia.central nuclei, or into the ventricles (Cases 2.13, 2.14, 2.15; morecaseswillbe presentedin Vol.IV B). Furtherobservations. 1. Tumors arising in the amygdala (corticoidorzero-Iayeredcortex) usually remain within the amygdala anddo not expand or infiltrate into architectonically or phylogenetical!ymore advanced are as (such as the neighboring hippocampus).Amygdala tumors do not infiltrate through the anterior commissureto the other side. 2. Tumors arising in the hippocampus usually do not infiltrateintothefornix,cingulum,or parahippocampalgyri.Thispat-
ing feature of transitional system tumors, even though some of these tumors are among the largest encountered. Transitional tumors, as a result of this propensity to expand within the confines of the allocortical or mesocortical zones, respect the adjacent neocerebral are as, basal ganglia, and central nuclei (such as the claustrum, putamen, pallidum, nucleus caudatus, thalamus, hypothalamus), and internal capsule. The spread of these tumors within the transitional zones tends to be toward phylogenetically or architectonically more primitive areas, and not in the opposite direction, toward more advanced or complex zones. The biological factors responsible for these phenomena are open to speculation, but may involve containment by the segmental vascular system, or inhibition by neurotansmitter concentrations distinctive to these areas, or both.
Tumors of the Basal Ganglia and Central Nuclei The basal ganglia and central nuclei represent the gray matter of the central telencephalon, diencephalon, mesencephalon, metencephalon, and myelencephalon. Though the term "basal ganglia and central nuclei" is used, we perceive these tumors as arising from the glial cells supporting white-matter fibers that run between and through these nuclei. The tumors of the central nuclei are divided into supratentorial and infratentorial groups, as outlined in Table 2.7. Primary tumors of the basal ganglia and central nuclei are rare. In our series, there are no cases of tumors originated in the putamen, pallidum, or claustrum. Those that occur in this region arise from the head of the caudate, the thalamus, and the hypothalamus.
ternholdstrue for a majorityof limbicand paralimbic tumors.
3. Unlike tumors in the neocerebral subpallium, tumors of the caudal orbitofrontal lobe or anterior insula-temporal pole readilytraverse one of the association bundles (the uncinate fasciculus)to infiltrate rapidly from one are a to another. 4. The limbic and paralimbic areas can be divided into four primaryregions. A variety of tumor growth patterns are observed,but early tumors tend to remain confined to a single sub(suchas1a, 1b,or 1c, temporal pole, amygdala,uncus hipregion pocampus,parahippocampus). 5. Insular tumors may remain isolated (region 3), spread withintheinsularregion,or spread to neighboringregions (such asthefrontaland temporal opercular regions, along the uncinate fasciculus ).
6. Tumorsof the cingulum occur in subregions 2 a, 2 b, and 2c.Solitarytumors of the fornix (1 e) and mamillary body (lf) alsooccur. 7. Many intraventricular tumors (paraforaminal) may originatefromtheseptal region (1g). 8. Extensionalong the connective fibers of the limbic lobe to themesencephalonhas been seen only in one case, which will be presentedin Vol. IV B. 9. Involvement of the en tire limbic and paralimbic system canbeseen unilaterally or bilaterally. Conclusions. Transitional tumors remain well localized at their siteoforiginfor a considerable period of time, and even up to the finalstagesofgrowththey remain confined within their architectonic(andthus,phylogenetic)system of originoIn the initial and intermedia te phases of tumor growth, this confinement is a strik-
139
Table 2.7
Tumors of the basal ganglia and central nuclei
Supratentorial Telencephalic Caudate Claustrum Putamen Globus pallidus Diencephalon Epithalamus Metathalamus Thalamus Subthalamus Hypothalamus
basal ganglia
}
Lentiformnucleus }
Infratentorial Mesencephalic Dorsal (tectum: colliculi, brachii) Tegmentum Base of Peduncles (crus cerebrii) Metencephalic Pons Pedunculi cerebelli Superior (cranial, rastral, brachium conjunctivum) Middle (brachium pontis) Inferior (caudal), corpus restiforme Cerebellar nuclei Dentate, fastigial, globiform, emboliform Myelencephalic
--
140
2 Neuropathology Much like a glioma in the subpallial region, a glioma arising in one of the supratentorial central nuclei begins as a small, discrete nidus that expands centrifugally in all directions (Figs. 2.14-2.16). With enlargement, it distorts and displaces the surrounding structures, such as the internal and external capsules
and adjacent nuclei. Nonetheless, with growth it continues to
a
b
e Fig.2.14 Growth patterns for central nuclei tumors (coronal views) a Glioma originating in the left thalamus and expanding locally b Expansion beyond the thalamus into surrounding white matter. Note that the affected fiber bundles are separated, as are the neighboring vascular structures e Right thalamic glioma expanding (without ependymal penetration)
into the lateral ventricular
cavity
"respect" their anatomical fiber tract borders. In most cases, supratentorial central nuclear gliomas are unilateral and mesoscopically well-demarcated from the surrounding structures. Medially, these gliomas may indent and distort the ventricular wall, but violation of the ependyma and internal glial membrane (with intraventricular penetration and CSF dissemination) is rare. Laterally and superiorly, these tumors may expand into the subpallial white-matter of an adjacent lobe or gyrus, but not into the transitional are as. It is also characteristic for a tumor arising in a central nucleus to remain in the segment of the nucIeus from which it originated (i. e., the dorsal or caudal thalamus, pulvinar, head of the caudate nucleus, etc.). Infiltration into long projection, association, or commissural fiber systems does not occur. The most common areas of whitematter involvement are the connections between the caudate and lentiform nuclei anteriorly in the retrolenticular regio n posteriorly. Most thalamic and hypothalamic lesions remained unilateral. However, occasionally bilaterallesions occur (see IVB). Extension through the internal capsule into the lentiform nuclei occurs only in the final phase (Fig. 2.14 b). Similarly, tumors of the lentiform nuclei infiltrate through the internal capsule to reach the claustrum and thalamus only at a late stage. Gliomas of the caudate nucleus occur only in the head of the caudate, and never extend into the tail; in one case, bilateral occurrence was observed by the neuropathologist. This contrast with tumors of the fornix, which can display extension along the body of the fornix bilaterally to reach the trigonum, without extensions or penetration into the thalamus, hippocampus, or parahippocampus. Unlike their counterparts in the supratentorial compartment, infratentorial central nuclei gliomas (brainstem) demonstrate a dichotomy in their behavior that depends upon their location (Fig. 2.17). Dorsal, lateral, and ventral mesencephalic tumors are frequently well circumscribed. Basal and tegmental mesencephalic tumors, on the other hand, tend to be diffuse. Only a minority of pontine tumors (10%) are well demarcated and unilateral. The remaining 90% are primarily diffuse and bilateral, remaining within the pons or extending cranially to the mesencephalon and diencephalon, or caudally to the myelencephalon. Pontine gliomas commonly infiltrate along the middle cerebellar peduncle into the cerebellar hemispheres. Pontine lesions expand caudally along the corpus restiforme, but do not infiltrate the inferior cerebellar peduncle and vice versa. No explanation exists for the propensity of central brainstem gliomas to grow in such a diffuse fashion within the pons. However, they frequently present an exophytic growth tendency dorsolaterally or ventrolaterally. Perhaps an explanation for this growth of pontine gliomas may be found in the complex weaving of the horizontal and vertical white-matter tracts that characterize the basis pontis. Alternatively, the vascularizationpattern of this area may be relevant. In any case,even in those patients with a pontine glioma who survive for two or more years, infiltration into the spinal cord is not seen, and only rarely is there spread rostrally into the peduncles of the telencephalon. This lack of infiltration into these areas does not appear to be related to the patients' relatively short survival periodo Perhaps tumors in this location
Specific Considerations arelessaggressivebiologically, or the tumor-suppressor environmentrnaybe more effective. Might it be that a "borderline" exists between the brainstem (embryological metencephalon and mesencephalon)and the other anatomical subdivisions of the nervous system (myelencephalon, diencephalon, and teleneephalon)?Such a "borderline" (based on as yet unrecognized biologicalor vascular factors) might constitute a functional "line resistance"to tumor growth, perhaps related to higher concentrations of tumor-growth inhibitors, or better blood supply and immuneresponse in some are as. Answers to these speculative questionsremain elusive. A final observation noted in brainstem tumors is the relatively common occurrence of astrocytomas, gangliogliomas, ependyrnomas,and cavernomas, while glioblastomas are rare, and oligodendrogliomas are extremely rare (see Cases 3.32, 4.32-4.37,5.28-5.30,5.33-5.38). lt is surprising that patients with brainstem tumors and presentingwith minor symptoms, do not have involvement of long traetsignsin the initial phase. (One frequently sees patients with diplopia,who are able to walk and can even be active in some sports.)Thisis because the fiber tracts are not destroyed but just displaeedby the intervening tumor growth.
141
Internal rnedullary lamina
Fig.2.15 Predilection sites 01thalamic tumors 1 Anterior nucleus 2 Ventral nucleus 3 Not observed 4 Pulvinar 5 Centrallocalized gliomas, originating Irom the internal medullary lamina, with diffuse growth A Anterior nucleus LO Lateral dorsal nucleus LP Lateral posterior nucleus MO Medial dorsal nucleus VA Ventral anterior nucleus VL Ventral lateral nucleus VPL Ventral posterior nucleus VPM Ventral posterior nucleus
b
ir I
LJ Fig.2.16 Growth paUern 01 gliomas arising in the caudate nucleus head a Sphericalinitial intermediate growth within the capsule 01 the head 01the left caudate nucleus with compression and deviation (but no penetration)01the surrounding structures
(coronal view) b Late-stage growth across the striatal liber cOllnections (into the lentilorm nucleus), with destruction 01 this pathway but preservation 01 other surrounding projection libers and nuclear structures
--
142
2 Neuropathology Fig.2.17 Artistic diagram 01 predilection sites 01 gliomas within the brainstem C Central, mostly dilluse DI Oorsolateral
L V a b e, d
Lateral mostlycircumscribed } Ventral Mesencephalon Pons Medullaoblongata
The circumscribed type of brainstem tumors have a tendency to grow in a paraaxial fashion to adjacent cisterns, or with centrally periventricular placed lesions, to the fourth ventricle. The diffusely growing lesions generally remain within an anatomic area (e. g. within the pons), but occasionally transgress these anatomical borderlines to extend rostral into the encephalon (see Case 5.36, page 357 and also chapter Brainstem Tumors in Vol. IVB).
d
Intraventricular Tumors The intraventricular designation includes tumors of the lateral, third, and fourth ventricles, tumors of the aqueduct, tumors arising from the walls of the ventricles and from the septal areas, and tumors of the choroid plexus (Table 2.8 a, b). Intraventricular tumors are divided into: 1) those that arise primarily within the ventricle, and 2) those that secondarily invade the ventricle from beyond its walls. As stated in the section on subpallial tumors, the vast majority of tumors that appear on imaging studies to invade the ventricle secondarily do not actually do so, as they are in fact contained by the internal glia limitans and the ependyma. Compared to the subpallial and central lesions, a wider variety of tumors arise in and around the ventricles (Table 2.8a, b). This is not surprising, considering the developmental migration of many neurons and glia from the subependymal plate layer.
Frequently encountered lesions in this group are meningioma, choroid plexus papilloma, colloid cyst, neurocytoma, astrocytoma, subependymal giant-cell astrocytomas, teratoma, and ependymoma. A large majority of tumors in this region are well demarcated and removable. At surgery, it is clear that most do not transgress the ventricular ependymallining (an exception is the ependymoblastoma, which extends and invades in both intraventricular and periventricular directions) (Fig. 2.18 a-e). Curiously, meningiomas of the third ventricle are distinctly rare (see Case 5.19, p. 345), while those arising in the trigone are not uncommon. Choroid plexus papillomas of the third ventricle and meningiomas of the fourth ventricles are rarely seen. Some of the lateral ventricular tumors that appear to originate within the ventricular system actually arise from the septum pellucidum and be long to the transitional tumors types.
Specific Considerations Table 2.8a
Intraventricular
Lateralventricle
Thirdventricle
Aqueduct Fourthventricle
tumors: anatomical
sites of origin
Frontal horn Body (celia media) Septum peliucidum Trigone Posterior horn Temporal horn Anteroinferior (infundibular part) Anterosuperior (foramen of Momo) Posterior
Table 2.8b
Intraventricular tumors: origins of common tumors
Site of origin
Tumor type
From the ventricular wali
Ependymoma, subependymoma Astrocytic tumors Neurocytoma Oligodendroglioma Tuberous sclerosis (giant celi) Craniopharyngioma Optic glioma Ectopic pituitary adenoma Pinealoma Germinoma Teratoma Thalamic glioma Medulioblastoma Acoustic neurinoma Choroid plexus papilioma
From the periventricular region, expanding into the ventricle
Superior Lateral (foramina of Luschka) Inferior (foramen of Magendie)
From within the ventricle, especially the trigone and fourth ventricle From within the ventricle, especially the trigone (and occasionally from the third ventricle)
Fig.2.18 Predilection sites of intraventricular and periventricular tumor s a Lateralventricle 1 Frontobasal medial-origin mostly from septal area 2 Frontobasallateral-originmostly from caudate head 3 Gliomasfromthe septum peliucidum 4 Dorsaliyextendingthalamic gliomas 5 Precunealgliomas in the precuneus, isthmus area, and lingual gyrus 6 Cunealgliomasfrom the cuneus 7 Gliomasfromthe fusiform gyrus, 02, and 03 8 Gliomasfromthe parahippocampus and lateral temporo-occipital gyrus GliomasIromthe parahippocampus and inlerior temporal gyrus Temporalmediobasalgliomas lrom the amygdala, uncus, and hippocampus Retrolenticular gliomas Intraventriculartumors, such as oligodendrogliomas ependymomas,plexuspapiliomas,and meningiomas
143
Meningioma Craniopharyngioma (in the third ventricle)
b Intraventricular and periventricular tumors 01 the third ventricle, and periventricular tumors extending into the ventricles mimicking an intraventricular les ion Hy Proper third ventricle tumors: glioma, neurocytoma, meningioma, cavernoma, and metastasis 1 Gliomas 01 the amygdala and parahippocampal areas 2 Adenomas, Rathke's cyst 3 Craniopharyngeomas, epidermoids 4 Optic gliomas 5 Subcallosal-septal gliomas 6 Gliomas 01 the caudate head 7 Gliomas 01the septum pellucidum and lateral ventricle 8 Gliomas 01 the lornix 9 Colloid cyst in the loramen 01 Momo 10 Plexus papilloma 11 Gliomas 01 the thalamus 12 Hamartoma 01the mamillary body 13 Gliomas 01 the ventral mesencephalon 14 Gliomas 01 the dorsal mesencephalon 15 Pineal and parapineal tumors 16 Gliomas 01 the isthmus area
144
2 Neuropathology Fig.2.18c Tumorin the fourth ventricle 1a Superior median and paramedian 1b Inferiormedian and paramedian 2 Posterior (fastigial) 3 Lateral
Tumors of the fourth ventricle may arise primarily whithin the confines of the ventricle (subependymoma, ependymoma, plexus papilloma) or extend secondarily into the cavity (astrocytoma, ganglioma, cavernoma, medulloblastoma, dermoid and epidermoids). Although the fourth ventricle is a very small space, there are four main predilection sites (Fig. 2.18c), which can be differentiated.
Tumor Infiltration Burger (1990) considered the great affinity of neoplastic cells for fiber pathways (Maxwell1946, Matsukado et al. 1961, Burger et al. 1988, Burger 1990). He describes the corpus callosum as the best known of these routes, especially for glioblastomas that arise from the lobar peduncle of the frontal and parietallobes. A proposed second conduit is via the fornix (Burger 1990), reached either in the superior-medial temporal lobe orsubjacent to the
corpus callosum. Another alleged route of spread is via the anterior commissure, through which frontal and temporal lesions cross the midline to reach the contralateral side. Burger states that the optic radiations, cerebral peduncles, and multiple association pathways that link functional regions also serve as cornmon routes of spread (Fig. 2.19).
Fig.2.19 Proposed common pathways of glioma spread (fromBurger et al. 1982, 1990, 1991).Our observations suggest that routes 1,4, and 5 may occur, but only in the late phases of glioma growth. Routes2 and 3 have not been observed 1 Corpus callosum 2 Fornix 3 Optic radiations 4 Association fibers 5 Anterior commissure
Tumor Infiltration The view that glial tumors frequently demonstrate widespreadearly dissemination, and the belief that peritumoral density(intensity) changes on imaging tests represent infiltration, are refutedby the followingobservations. (1) Recurrent tumors, even thosethat recur months or years after removal, tend to be found at thesamesite (Fig.2,20a, b). This is a well known phenomenon Ihat is widelydocumented
in the literature. (2) The incidence of
multicentric tumors is low, averaging 5-6% in most large series (4% macroscopic and 6% microscopic in Russell and Rubinstein's series). In most of these cases, the tumors prove to be glioMultiple gliomas of other cell types are rare. (3) Glioblastomas. matosiscerebri is an extremely rare condition (0.05%). (4) A higherincidenceof gliomatosis cerebri has not been reported in fue MRI era. Our observations concerning the initial site of gliomasand the pattern of expansion do not confirm the descriptionsof Burger. As asserted in the section on subpallial tumors, low-gradeglioma,anaplastic glioma, and even glioblastoma show apropensityto remain confined within their initiallocation (subcortical,segmental,gyral, and lobar sectors). The routes of spread are restricted for a long time by the local vascular system and otherunknown factors. We have discussed and confirmed this obvious discrepancy withBurger,who concedes that most neuropathological observationsaremadeon the brains of patients who survived until the terminalphasesof the disease and who have undergone surgery and radiationtherapy. We also have observed dissemination along the CSFpathwaysin the terminal phase of glioma patients who have undergonesurgery and radiotherapy. This mode of spread, however,isnot seen in these tumors during the initial and intermediatephases. Certain gliomasoriginating within the subcortical, gyral, and lobarsectors,however, may show extension (expansion with and without infiltration) along the short associative fibers to reach neighboringsubgyri, through the uncinate fasciculus (transitional tumors)or acrossthe corpus callosum (rare tumors of the frontal or parietallobar pedunde). This mode of infiltrative spread may
Fig.2.20 Recurrentgliomas: commonsites(horizontalviews) a Recurrent rightIrontallobe glioma.Recurrentgliomasmost frequently ariseat the site 01initial presentation. Thechronology may be:glioblastoma alter one year, anaplasticgliomaalter 3-5 years, benign(Grade11)astrocytomaalter 8-10years b Thepatterns01spread lor recurren!gliomas.Alter initialrecurrence attheprimarysite(in the walis 01the prevlous tumorcavity),recurrent tumorsIrequentlyspread via surgical pathways (throughthe pia into the subarachnoid space,or through the ependyma intothe CSF),contralateralspreadacrossthe lorceps major(orminor)mayalso occur (arrow)
145
relate to the special arterial and venous patterns in these territories (Case 2.38). Primary su1cisuch as the sylvian, parieto-occipital, ca1carine, transverse, and interhemispheric fissures are not traversed by intrinsic tumors. Gliomas respect the double layer of pia, which must be traversed in crossing a su1cus to invade adjacent gyri. AIso, these tumors usually respect the layers of the membrana limitans gliae (external and internaI) and pia. Occasionally, they are seen to distend these layers and expand, mushroom-like, over the neighboring gyri. It is extremely rare for gliomas to penetrate the pia or arachnoid and spread along the cranial-spinal axis via the subarachnoid space with the exception of insular anaplastic gliomas) (see Case 2.40 on p. 186). This is only observed in the postoperative terminal phase. The belief that gliomas diffusely invade brain parenchyma, and that they have spread beyond the confines of the gross and radiological boundaries of the tumor by the time of their discovery, is one that pervades not only the literature, but daily neurosurgical practice. We do not dispute that malignant gliomas spread in this way, but we add the proviso that this does not occur until the terminal phases of tumor growth. The infiltration of gliomas into neighboring gyri, into the ependymallayer, or into the pia, is a late occurence, the tumor having been present at its initial locus for some time. Several examples of these observations are presented at the end fo this chapter. Intrinsic tumors frequently lie in one place for a long time with no infiltration into surrounding tissues. On the other hand, there are examples (see Case 2.31 on p. 179) of these same tumors (5-10%) that behave aggressively, respecting no margins, and infiltrating deeply into surrounding structures. In these cases, there is severe adhesion to surrounding tissues, with destruction of all bordering layers. The reason for this sudden change in tumor characteristics is not well understood, but may reflect the same aggressive change of behavior observed in some intrinsic lesions.
146
2 Neuropathology
Peritumoral Changes The peritumoral intensity changes dramatically displayed on CT and MRI studies have received a great deal of attention as to their significance. The issue is, which imaging changes are due to tumor per se, and which are related to nontumoral changes (i. e., are reactive). Much has been written correlating peritumoral tissue changes evident on neuroimaging studies with postmortem examinations. It has been proposed that the changes are secondary to compression of normal brain, to edema, to vasoactive substances acting on the blood-brain barrie¡;, to infiltrating cells, or to metabolic tissue changes. Perhaps the answer is a combination of some or all of the above. Although the pattern of peritumoral change is often identical in a wide variety of pathologic situations, many believe that the peritumoral changes have different causation in different processes. Still, the question always comes down to what the changes mean in an individual case. In the face of considerable peritumoral edema, some tumors (especially meningiomas, anaplastic gliomas, hypernephromas, and melanomas) can be expected to exhibit significant adherence or infiltration into surrounding brain, or both, while others (metastatic cancers, lymphomas, and. plexus papillomas) demonstrate no such behavior. Conversely, some tumors (acoustic neuroma, epidermoid, dermoid, craniopharyngioma, adenoma, oligodendroglioma, chordoma, etc.) may manifest striking adherence, infiltration, or both, without any evidence of peritumoral edema (see Table 3.7, p. 204, and pp. 203-207). Overall, marked hyperintensity from peritumoral focal or holohemispheric changes is observed in glioblastoma (90%), anaplastic astrocytoma (75%), metastatic tumor (75%), meningioma (70%), abscess (95%), parasite infestation (25%), radionecrosis (90%), and neurinoma (5%). There are suggestions that these findings depend on lesion location, as both benign and malignant tumors situated in regions of rich venous drainage often have no peritumoral changes. We have observed that the pattern of these changes closely resembles the venous drainage patterns of the cerebral white matter, as outlined by Hassler (1964) (Figs. 1.94, 1.96, pp. 110,111, Cases 3.52-3.56, Table 3.7, p. 204). The presence of these changes around a meningioma signals an increased adherence of the tumor to the arachnoid and pia (as discussed on p. 147). It has also been shown that these peri-
tumoral findings may relate to (1) estrogen or progesterone activo ity by the tumor; (2) malignant change; (3) compression ofnormal brain, with alterations in cerebral blood flow, the cerebral metabolic rate of oxygen, glucose utilization, and oxygen excretion (compressiveischemia with compromise of the blood-brain barrier); (4) venous mechanical obstructions; and (5) elevated hydrostatic pressure within the tumor (secondary to hemorrhageor rich vascularization). In glioblastoma, it is postulated that the peritumoral hyperintensity changes may represent (1) edema secondary to the release of vaso active peptides by the tumor with damaged parenchyma and consequent blood-brain barrier breakdown; (2) tumor cells infiltrating along the white-matter tracts; (3) metabolic changes in the white-matter, such as alterations in cerebral blood flow, the cerebral metabolic rate of oxygen, glucose utilization, and oxygen excretion; or (4) compression or thrombosis of the deep medullary venous system. Although (1) and (2) maybe present in some cases, they are not constant factors. Burger and Vogel (1988) studied eleven cases to determine the microscopic extensions of gliomas in relation to the contrast-enhanced rimon CT, and concluded that "the distribution of tumor cells cannot be inferred from CT images, since peritumoral changes may over-or underestimate the tumor's extent." Metastatic tumors often ha ve the most peritumoral changes, but no studies have shown that these changes reflect tumor invasion. Usually the contrary is found at operation. It is erroneous to always equate the presence of a tense and swollen brain with edema formation. After the removal of a lesion with marked peritumoral changes, the brain is usually very relaxed, even though the white-matter changes may persist for months or years after removal. This is particularly true in meningiomas (Case 3.59, p. 245). The white-matter changes in malignant gliomas usually disappear after complete removal, clearly refuting the argument that these changes reflect tumor invasion. It is quite clear that the presence of peritumoral changes,per se, should not deter the neurosurgeon from planning a complete tumor removal. At the present time, there are no definitive answers as to the significance of these changes in any individual case (Cases 3.52-3.58, pp. 243-245).
Tumor Demarcation The concept of demarcation (i. e., clear separability, or a distinguishable border) between different biological tissues is easily appreciated with regard to most extrinsic tumors. In addition though, many intrinsic tumors are also well-demarcated when observed under operating conditions. This point stands in contradiction to the statement of Greig (1989): "An important aspect of primary brain tumors is the diffuse border that exists between the tumor tissue and the surrounding brain, which cannot be demarcated macroscopically... it is difficult for a neuropathologist to determine where exactly brain begins that is completely free from neoplastic invasion." This is correct for the gliomas in their final stage.
Under the mesoscopic view of the surgical microscope, a discernible plane of demarcation can be found in most intrinsic tumors in the early phase (Fig.2.21). Of course, microscopic tumor extension beyond this apparent demarcation is outside the resolution of the operating scope (5-10 x), so the smallest extensions of infiltrating astrocytic tumor cells (detectable histologically) cannot be appreciated during surgery. No reliable method of detecting tumor extension at the time of surgery exists. It is clear that recurrent glial tumors nearly always return at the site of previous occurence (Fig. 2.20). Those factors (vascular or fiber topography, inhibitory proteins) responsible for the circumscription of the initial tumor continue to restrict the spread of
Brain-Tumor Interface: Adherence and Adhesiveness
147
b Flg.2.21 Glialtumor demarcation a Coronalsection.A left lrontallobe glioma which is well demarcated by glioticchanges in the surrounding white matter. This peritumoral 'capsule"maybe produced by either hardening or softening 01the surrounding whitematter
b The "encapsulation" 01 a similar glioma, resembling that 01 an
therecurrence.In addition, any tumor cells remaining after initial
sis can be totally deceiving.A note of warning must be given. It has occasionally been our experience to have easily removed a discrete and nicely circumscribed meningioma-like lesion, only to have the histopathological diagnosis return as glioblastoma and, conversely, to have removed a diffuse, poorly marginated gliomalike mass, only to have the final pathology return as meningioma.
surgicalremoval must reside in the immediate vicinity of the originaltumor.The only other possible explanation is that those factors(genetic,molecular, viral) responsible for inducing the initial tumorarestillpresent, resulting in another new tumor in the same location.
abscess
Thetendency of surgeons to equate the presence or absence ofacleardemarcation plane with a presumed histological diagno-
Brain-TumorInterface:Adherence and Adhesiveness Benignand malignant tumors in both the intrinsic and extrinsic categories haveunpredictable reactions with local surrounding tissues. As a consequence, a pseudocapsular formation arises aroundsomeextrinsic and intrinsic tumors (Table 3.8, p. 205). The interfacebetween the tumor mass and the surrounding tissues is a \'itallyimportant zone for the neurosurgeon, for it is at this site thatthefinal determination of resectability is made. The existence anddegreeof adherence of a particular tumor to the pia and arachnoid,blood vessels, dura, sinuses, and glial surfaces often cannotbe reliably predicted preoperatively by angiogram, CT, or MRI. Usually, it is familiarity with tumor adherence characteristicsframpast experience that provides the best guide as to what to expectwith future tumors. However, unwelcome surprises involvingunbelievableadhesive difficulties are more common than we wouldlike (Fig.3.47, p. 239). Our experience has demonstrated that the degree of adherenceisnot particularly related to the tumor type or classification. Meningioma,craniopharyngioma, glioma, papilloma, teratoma, some neurinomas, and metastatic tumors may have partial, subtotal,or total adherence to the surrounding tissue. The explanation for this tremendous variability defies current analysis and meritsfuture study.The physiology and reactive properties oi the
arachnoid system are poorly understood. Perhaps the extreme degrees of adherence reflect exaggerated (and abnormal) biochemical, inflammatory or autoimmune reactions between the arachnoid and the tumor surface. A better understanding oi the mechanisms of this process may lead to preoperative or intraoperative targeted agents that soften or reduce tissue adhesiveness (see Table 2.9). MRI has shown some promise in characterizing tumor consistency (e. g., differentiating Antoni A from Antoni B types of acoustic schwannoma). More studies correlating tumor texture, composition, and adherence characteristics (as demonstrated at surgery) with MRI would be beneficial in operative recommendations, especially as regards the timing of surgery and the assessment of morbidity risks. Specific points relating to the adhesion characteristics of tumors are: (1) Small and large meningiomas with imaging evidence of increased peritumoral intensity of MRI or decreased density on CT usually exhibit marked arachnoid-pia adherence, with atrophic glial changes on the underlying cortex (see Cases 3.43, 3.45 on pp. 236-7). Meningiomas, with minimal or no peritumoral reaction on imaging, are usually dissected from the adjacent tissue
148
2 Neuropathology
with little difficulty (see Cases 3.42, 3.44, 3.46, and 3.51 on pp. 235-242). (2) Neurinomas rarely have peritumoral hypodensity, and rarely demonstrate adhesion with the pia layer of the cerebellar lobulus, pons, and medulla, but do adhere with nerves seven and eight. (3) Craniopharyngiomas demonstrate surprisingly little peritumoral edema and tissue reaction on imaging (even in large, cystic lesions). Yet despite the paucity of peritumoral imaging changes, these tumors are notoriously very adherent (see Case 3.47 on p. 239). (4) Glioblastomas can be very adherent to the arachnoid and dura; this feature appears tó be independent of the degree of perilesional changes. (5) Lymphomas do not seem to produce perilesional changes. (6) Metastases often produce a high degree of peritumoral change, but are rarely adherent (see Table 3.7, p. 204).
(7) In some cases, piloid astrocytomas, gangliogliomas, and especially oligodendrogliomas, there is a tendency to escape through the piallayer and to infiltrate into the arterial adventitia. Table 2.9 Surgically observed variables between peritumoral socalled edema and changed interface (adherence and adhesiveness) Edema
++ / Adherence
++
Edema Edema
++ /Adherence / Adherence
++
Edema
-
/ Adherence -
Meningioma. metastasis (hypernephroma). acoustic neurinoma. glioma Metastasis (Ca). Iymphoma. abscess Craniopharyngioma. adenoma. chordoma. dermoid. oligodendroglioma. plexus papilloma Cavernoma. hemangioblastoma. chondroma. optic glioma. adenom. craniopharyngioma. meningioma, neurinoma
++ Severe - None
Tumor Vascularization A quantitative and qualitative determination of the peritumoral vascular state with contrast CT scan and MRI, together with angiography and a presumptive diagnosis of the tumor type, typically provides the neurosurgeon with some assurance that the patient can be offered safe surgery. Nevertheless, our surgical experience indicates that many features of tumor vascularity are unmeasurable and unpredictable. The angiographic features of both extrinsic and intrinsic tumors were studied extensively between 1930 and 1970. The patterns of vessel shift tumor blush, venous stasis, and neovascularity, and the changes and qua lity of the cerebral blood flow in and around the tumor are described thoroughly in numerous publications and monographs. For the performance of tumor surgery, this information concerning vascularity and cerebral blood flQw (in and around the lesion) is of great value. The vascularity of a tumor is estimated on CT and MRI by noting the intensity of contrast enhancement. MRI "flow voids", which may penetrate or encircle a tumor, can also been seen. Magnetic resonance angiography (MRA) can be very helpful in the precise indentification of peritumoral and intratumoral vascular structures. Other vascular changes in and around tumors may be angiographically undetectable, including the development of increased local collaterals, fistulas, microaneurysms, and sinusoids. Huang et al. (1964) noted that this increased vascularity in the normal white-matter surrounding glioblastomas may correspond to "remarkable dilatation of the adjacent normal medullary vessels." The relationship of these changes to peritumoral edema seems clear. On histological examination, pathological vessels are often found to be deficient in the normal anatomicallayers. These vesseIs demonstrate marked fragility on manipulation and an inability to constrict reflexly when cut. Along these lines, about 5% of tumors in our experience show a marked and unpredictable fragility of their vascularity, which occurs independently of tumor type (benign and malignant). Due to these abnormalities, difficulties
with proper hemostasis can occur in the resection of some adenomas, gliomas, ependymomas, plexus papillomas, medulloblastomas, and angioblastomas. Normally vascularized or hypovascularized tumors do not present problems with blood loss or hemostasis, provided they are devascularized properly and in sequential steps. Even many hypervascularized lesions can be controlled easily with stepwise devascularization using standard microsurgical techniques. A highly vascular tumor with high flow characteristics should be approached with special attention directed towards reducing the input of the arterial feeders before any attempt is made to remove the tumor. Draining veins are preserved until later, when the tumor is well isolated from its arterial blood supply. Embolization of larger feeders in meningiomas, neurinomas, and glomus tumors does not guarantee a relatively bloodless field, and even the experienced surgeon may have almost insurmountable difficulties when encountering very vascular tumors of these types. Historically, neuropathology has provided valuable descriptions of the idiosyncracies of tumor vasculature, primarily from a histopathologic perspective. Tumor capillary endothelial abnormalities, such as failed tight junctions and endothelial fenestrations, explain, at least in part, tumor vascular "leakiness" and peritumor al edema formation. The identification of a series of tumorsecreted angiogenic factors has opened up a promising field oí research, with possible therapeutic applications. Continued investigation into the dependence of tumor growth upon the coincident development of its vasculature is needed. Venous thrombosis. Thrombosis of arteries around and within intrinsic tumors occasionally occurs (meningiomas, craniopharyngiomas). However, the thrombosis of small, modera te, and large veins within a intrinsic tumor is a pathognomonic sign of malignancy. While this is very frequent in glioblastoma, it is also seen other gliomas (including the anaplastic variety), with angioblastic meningiomas, with neurinomas PNET, and with lymphomas.
The Numbers and Types of Tumors Unfortunately,this firm indicator of malignancy cannot be visualizedon imaging studies, including MRI, MRA, or selective angiography.This tendency for venous thrombosis cannot be
The
149
explained by mechanical and hemodynamic factors alone, as it never occurs with benign extrinsic and intrinsic tumors. It certainly appears to be a biologically induced phenomenon.
Numbersand Types 01 Tumors
Although some pathological entities are most commonly seen as
processes (i. e., demyelinating or infectious), neoplastic diseaseof the CNS is most frequently solitary (95%), with multidiffuse
centricity less common (0.3%) (Cases 2.41, 2.42), (Case 2.24). This pattern orderssuchas aneurysma
(4.5%), multihistological types rare and diffuse varieties very rare (0.1 %) is similar to that seen in vascular dis-
and AVMs (Table 2.10).
Definitions.A. A unicentric tumor is a solitary, singular, separateoindividualprocess that
nonetheless may be bilateral if
encounteredin the midline. B. Mu/ticentric tumors grow simultaneously or consequently,but are topographically remote from each other, withoutevidenceof disseminationby seeding or direct extension andwithout demonstrablemicroscopiccontinuity (Reichenthal 1983).In simple terms, multicentric tumors develop totally independently. Accordingto Russel and Rubinstein (1989) most cases of assumed multicentricity take the form of one or more small inconspicuous foci, situated on the periphery of the main tumor
Table 2.10
CNS disease processes:
patterns
Disease of the CNS
Pattern of morphological changes
Neoplasms
Unicentric > Multicentric > Multihisto- > Globallogical type type "-matosis" Unicentric > Multicentric > Diffuse
Vascular disease Infectious disease
Unicentric < Multicentric < Global (Abscess) (Meningitis) (Parasites)
Demyelina-
(Encephali-
tis) Unicentric < Multicentric < Global
tion
mass. True multicentricity, in which the tumors are widely separated is exceptional. The pattern of growth for common CNS tumors are shown in Table 2.11. The occurence of multicentric
Table2.11 Patterns of occurrence in CNS tumors Tumor
Unilateral Unicentric
Extrinsic Meningioma Neurinoma Chordoma Chondroma
Craniopharyngioma Adenoma Fibroma Glomustumor Pineal tumor
Teratoma
Epidermoid Dermoid Intrinsic Glioblastoma Astrocytoma
Xanthoastrocytoma Oligodendroglioma Ependymoma Ganglioglioma Medulloblastoma Hemangioblastoma Lymphoma Plexuspapilloma Metastasis
Multicentric
Bilateral
Diffuse
Global
+++ +++ +++
++ + -
+ +
(+) (+)
+ +++ +++
+
-
-
+++ +++ +++ +++ +++
-
+ + +
-
+
-
+ -
-
-
-
-
-
-
-
-
Unicentric
+ -
+ + + + + + + +
Metastasis
Aggressive growth
Multicentric
+ +
-
-
-
+ -
-
-
+ (+)
+ + + +
-
+ +
+ + +
+
-
+
-
-
+
+ + ...:.
+ -
+ + -
+ + -
+ -
+ +
-
-
-
-
-
+++ +++ ++
(+) -
+++ +++ +++ +++ +++ +++ +++ +
-
-
-
++ + ++
+ -
-
-
-
+ -
+
-
-
-
-
+ + + -
+ + +
+
-
?
-
+
+ + + + + (+) ? + +
150
2 Neuropathology
gliomas has been reported by many authors (Solitare 1962, Batzdorf et al. 1963, Solomon et al. 1969, Borovich et al. 1976, Bussone et al. 1979, Chadduk et al. 1983, Reichenthal et al. 1983, and Barnard et al. 1987). Kato et al. (1990) recently reviewed these cases. There is a considerable variation in the reported incidence of multicentric tumors (0.9%-10.0%) (Batzdorf et al.), but it average s 2-3%. For meningiomas, however, the introduction of modern neuroimaging has increased this rate to 5-10% (Domenicucci et al. 1989), with more than 90% of these in women. The pathogenesis of multicentric tumors remains elusive. Storch (1899) suggested that gliomas have the ability to induce neoplastic change in adjacent ~lial tissue. Ostertag (1936, 1941) postulated that these tumors arise from primitive cells (with a blastomatous potential) displaced during development. Later the concept was developed that these lesions arise from multiple embryonal rests scattered at different sites. This dysoncogenetic theory, however, do es no! explain the chronological delay between brain maturation and tumor appearance. Perhaps this is what led Zülch (1957) to suggest that the only difference between multiple and multicentric lesions is that the latter disseminate by as yet unrecognized pathways. Other theories were presented by Willis (1960), who suggested that multifocallesions develop as a result of a staged neoplastic transformation occurring over diffuse areas and progressive proliferations at specific sites simultaneously, as a result of biochemical, hormonal, and mechanical factors. The possibility of such a premalignant state prior to the development of frank anaplasia was also suggested by Moertel et al. (1961), who proposed that it might be brought about by specific intrinsic or extrinsic carcinogenic influences acting on susceptible tissue over a sufficient period of time.
A familial occurence of glial tumors was noted by Heuch el al. (1986), who reported on glioblastomas in three family members, with one case being multicentric. Similar associations were made by Baughman et al. (1969) and linked to Turcot's syndrome. Kato (1989) suggested that underlying diseases such as neurofibromatosis, tuberous sclerosis, or multiple sclerosis are often responsible for promoting multicentric neoplastic changes. The etiology of multicentric meningiomas is probably the presence of multicentric dural foci, as multiple minute foci are frequently found around solitary meningiomas (Borovich and Doron 1986). Final answers on the genesis of multicentric glial tumors clearly await a better explanation of unicentric tumor formation. From a histological perspective, a solitary tumor characteristically has one cellular type (unitype), though on very are occasions (0.1%), more than one histological cell type (multitype) can be seen in the same neoplasm. Multitype tumors are observed more frequently in multicentric lesions, though even in this instance it is rare (0.3%) (Table 2.12). The majority of unicentric and unitype CNS tumors are unilateral (two-thirds), but bilateral tumors (both intrinsic and extrinsic) are common (one-third). These are all midline tumors, in the parase llar, callosal, brainstem, vermian, and paraspinal regions. Unicentric and multitype tumors are extremely rare (0.01%), with only two cases in the present series (one case of meningioma + glioblastoma, one case of optic glioma + dermoid). Multicentric and unitype tumors are frequenl (5-10%), and may be found in a variety of combinations of CNS divisions (Table 2.12, Cases 2.38-2.42). Unitype or multitype tumors can be uni- or multicentric (unilateral, bilateral, supratentorial and infratentorial, cranial and spinal).
Table 2.12 Unicentricity and multicentricity, histological unitypes and multitypes in tumors and vascular diseases in own series Glo Glomus tumor Meningioma Neurinoma Adenoma Craniopharyngioma Chordoma, Chondroma Dermoid Optic glioma Glioblastoma Astrocytoma Oligodendroglioma Ependymoma
Plexuspapilloma Ganglioglioma Medulloblastoma Germinoma Lymphoma Sarcoma Melanoma Hemangioblastoma Cavernoma AVM Venous angioma Aneurysm
+ + -
+ + +
M
+ + -
N + + + -
Ad
Cr
-
-
+ -
+ + -
+
+ -
-
-
-
-
-
-
+ + + +
-
O -
+ -
+
-
O Always unicentric and histological unitype (+) Multicentric, unitype + Multicentric and combined type
De
-
-
-
-
Op
Gli
As
-
+ + -
+ + -
+ + -
+ + +
+ +
-
-
-
-
-
-
-
-
-
+
01
Ep PIPa G.GI Med Ger
-
Ly
Sar Mel Hbl
-
-
-
-
-
-
-
-
-
-
-
-
-
+ -
-
-
-
+ + +
+ + +
-
-
-
-
-
-
-
O -
O -
-
-
-
-
-
-
O
Ca AVM V.A A
-
+
+ +
+ +
-
-
-
+
-
+ +
-
-
-
+
+
+
+ + -
+ + +
-
-
-
+
(+) ? +
? (+) + +
+ +
+
O -
+
-
Ch
+
-
-
-
+ + +
+ +
+ +
-
+ -
-
(+) +
-
-
-
-
-
-
-
-
-
+
O -
(+) -
O -
(+) -
(+) -
? + (+) + ? (+)
Conclusions Locationof Multicentric Tumors Within the CNS Cerebralor cerebellar (unilateral or bilateral)
Cerebellar and spinal: Cerebral, cerebellar and spinal:
meningioma, neurinoma, glioblastoma, astrocytoma, cavernoma, hemangioblastoma, medulloblastoma, metastasis. hemangioblastoma. gliomatosis, neuromatosis.
meningiomatosis
Multicentric-multitypelesions are quite rare (0.3%), with only onecasein the present series (see Case 2.41 on p. 187). Some com-
151
binations, such as oligodendroglioma and ependymoma, hemangioblastoma and glioblastoma, medulloblastoma and glioblastoma, cavernoma and glioblastoma, etc., ha ve not been reported. It is c1ear that the spectrum of CNS tumors encompasses a diverse group of lesions. This led Courville (1936) to conc1ude that practically every combination of tumors can be found within the cranial cavity from the standpoint of location, tissue of origin, degree of malignancy, etc. Modern neuroimaging for CNS neoplasms has enhanced our ability to study their patterns of growth. With careful scrutiny by many observers over long periods of time, we will eventually come to a better understanding of the biological pathogenetic features (as related to location, cell type, and multiplicity) of CNS tumors.
Conclusions
Predilectivesites. For extrinsic tumors, predilective sites at the skullbase and along the calvarium are well known. For intrinsic tumors, predilectivesitesare welldocumented (see atlasesof neuropathologyand MRI neuroradiology) but not well systematized (Table2.13).Tumorsoriginatingin a particular area remain in thatareaeitherdueto the segmentalvascularpattern the anatomi-
calborderline,the embryonic glial-neuronal migration network, the sectional CSF transcerebral pathway, or the host immune defensesystem.
Tumorbiology.Forall CNStumors (both benign and malignant), growthoccursin initial, intermediate, and final phases. The duration ofeachphaseis generally predictable for
a group of similar
tumors,but in an individual case it can be quite unpredictable. Therate of growth for a majority of tumors is more or less linear, butina minorityof neoplasms, unpredictable growth is seen (suddenwildgrowth, slow growth and rapid growth alternating, no growth,regression, etc.). Edema (perilesional hyperintensity), interfacialadherence, penetration and infiltration (with destructionofborderlines),and migration(dissemination)occur in both benignand malignant tumors. Fortunately, a majority of neoplasms(bothbenign and malignant) demonstrate respect for their borderlines,and usually they remain circumscribed and demarcated. Histologicalinvestigations have reached their limits. Future understandingof tumor activities depends on molecular and cellular biology(i. e., what substances are produced by the tumor parenchymato inactivate the macrophages, to paralyze the defensiveactivityof the neuroglia, and to increase the local supply for their own needs).
Extrinsictumors. (1) While respecting dural and arachnoid planes,these may grow to a very large size. (2) Some initially well demarcated tumors (e. g., meningiadenomas,chordomas,and craniopharyngiomas)suddenly amas, changetheir growth characteristics and behave aggressively, with infiltrationinto surrounding structures. (3) Someextrinsic tumors are associated with extraordinary degreesof adherence and adhesiveness, making total removal impossiblewithout morbidity (e. g., hypothalamic damage with somecraniopharyngiomas, cranial-nerve injury with cavernous sinusmeningiomas, chordomas,and chondromas).
/ntrinsic tumors. (1) These tumors show a precise predilection to arise from certain localized areas of the brain. The reasons for this remain unc1ear, but it may be related to cytoarchitectonic structure, phylogenetic development, or vascularization patterns. (2) Descriptions of tumor location based on "lobar" terminology are no longer tenable. Instead, the tumor location must be defined in terms of precise topographic anatomy. This approach aids the neurosurgeon in preoperative planning, and optimizes the precise removal of tumor without injury or removal of uninvolved brain. It also allows a more valid comparison between specialists and between institutions comparing treatment protocols. (3) Multicentric, diffuse, and global forms of astrocytic tumors are rare, comprising about 1-5% of the total tumor incidence. They do occur, but late in the biological course. During the early and intermediate growth phases, 90-95% of intrinsic tumors are unicentric and mesoscopically well demarcated from the surrounding tissues. (4) Demarcations formed by pial, arachnoidal, and durallayers are readily respected. The biological nature of the well-demarcated surgical plane at the interface between intrinsic tumor and nontumorous glial white-matter is poorly understood. Many tumor demarcation borders are well defined and easily identified, while some are indeterminate. Still other groups are characterized by adherence and adhesion, so that it is more difficult to dissect at the borderline. The demarcation characteristics of tumors are, at present, inadequately predicted with neuroimaging, and can only be accurately assessed at surgery. (5) Most intrinsic tumors have a predilection to originate at certain sites. This is especially true for subpallial (neocortical and allocortical) tumors. Tumors arising in the highly specialized are as of the visual, auditory, and sensorimotor areas (inc1uding language areas) are very rare and more of a diffuse type. Further research, incorporating modern data on phylogenetic and architectonic concepts may provide relevant information on tumor origin and spread. (Diffuse intrinsic tumors occur in the limbic lobe, central lobe, brainstem, and spinal cord.) (Cases 2.23-2.28.)
152
2 Neuropathology
Table 2.13
Location 01 CNS tumors by pre1erred site 01 origin Infratentorial
Supratentorial
Extrinsic Epidural Intradural Subdural Interarachnoidal
++ ++ +++ +++
Intrinsic 1. Neocortical areas a. Primary sensorimotor isotypical cortical areas (1,2,3,4,17,41,42, 43, 4~) b. Unimodal association, homotypical cortical areas (5,7,8,18,19,20,21,37) c. Heteromodal association, hOlÍlotypical cortical areas (9, 10, 11, 12,39,45,46,47) 2. Allocortical and mesocortical areas (Limbic lobe) a. Paleocortical b. Archicortical c. Periarchicortical d. Transitional septal, amygdala 3. Central nuclei
" /
a. Putamen-Pallidum
basal dorsal
b. Head 01caudate Tail 01caudate c. Thalamus Pulvinar Medial Lateral 4. Ventricular Frontal Temporal Occipital Trigonum Third ventricle +++ ++ + (+) ?
Frequent, Moderate, Rare, Very rare, Unknown
Extrinsic Epidural Intradural Subdural Interarachnoidal
(+)
+ +++
++ (+) +++ ++ (+) ? + ? + (circumscribed)
}
+ (Diffuse)
Horn
++ ++ +++ +++
Intrinsic 1. Cerebellar a. Neocerebellar (LOP, LSS, LSI) (Declive, lolium) b. Paleocerebellar LOA, centrallobule, c. Archicerebellar Flocculonodular 2. Central nuclei
(+)
Extrinsic Epidural Intradural Subdural Interarachnoidal
+ + ++ ++
Intrinsic Juxtamedullary Intramedullary Central canal
++ ++ +
++ ++ culmen
++ ? ?
3. Ventricular
++
Brainstem Mesencephalon Tectum Tegmentum
Ruber Substantia nigra
}
Spinal
Base Pons Base
++ (+) Diffuse ? ? +
Tegmentum
+ (circumscribed) + (Diffuse)
Medullaoblongata
+
+ + LOA LOP LSI LSS
Anterior quadrangular lobule Posterior quadrangular lobule Inlerior semilunar lobule Superior semilunar lobule
(6) Our experience indicates that the majority of intrinsic tumors grow by spherical expansion. The proliferating cells of intrinsic tumors enlarge by insinuating themselves between the myelinated neuronal tracts. Concentric expansions split the whitematter tracts apart as the tumor grows. Neuronal axons appear remarkably resistant to actual destruction. Functional abnormalities, even with large tumors, are compensated for by the capacity of the central nervous system for parallel processing. (7) Astrocytic tumors arise most commonly in the subcortical white-matter and maintain a localized relationship with the subgyral, gyral, or lobar sector of origino Similarly, tumors arising in the central nuclei (although growing to a large size) rema in confined to their central origina. Intraventricular tumors stay within the ependymallining, even when causing huge distortions within the brain parenchyma. The biologic factors underlying these relative or absolute containments are poorly understood. (8) Most recurrent tumors arise at the previous original site
of the tumor, lending further support to the concepts of unicentricity and containment. (9) The spread of intrinsic tumors to a separate site within the CNS is rare (1 %) within two years of initial operation. The spread of astrocytic tumors outside the CNS is extremely rare (Leifer et al. 1989). (10) Our observations on the spread of astrocytic tumors conflict with commonly reported pathological series. These reports are based on less common examples of end-stage tumors following multiple therapies, and should nOt be regarded as representing the "normal" natural history. During the early and intermediate growth phases, most astrocytic tumors do not infiltrate along the large association, commissural, or projection white-matter bundles, with two exceptions: (a) the uncinate fasciculus and (b) the corpus callosum. Only poor-grade tumors in their endstages of growth have been pathologically proved to have diffusely infiltrated the brain.
Conclusions (11)Thrombosis of veins within intrinsic, and also extrinsic, tumorsisusuallya hallmark of malignancy that can be recognized intraoperatively.It is a biologically induced feature of the tumor, ratherthan being due to mechanical venous compression.
TheFuture (1) Moleculargenetics and cytogenetic studies have proved very usefui in identifying abnormal chromosomes in various tumors. Manyhumangeneticdiseasesare caused by genetic deletions. If thedeletionsinvolve important tumor-suppressor functions, then thepatient may be predisposed to develop one or more tumors ",hena further genetic event ("hit") occurs. Alternatively, or in addition,if a patient inherits an oncogene with a predisposition towardstumor expression, or if a mutation develops in a normal oncogene.that patient has the possibility of tumor development. inneuroblastomaand retinoblastomahave provided modStudies elsofthe likelygenetic mechanisms of tumor initiation. Why they developin any one individual is still unknown. Similar studies in glialtumorssuggesta cascadeof gene abnormalities,particularly involvingchromosomes 17, 9 and 10, which upset the balance the actionsof oncogenesand tumor-suppressor genes. between Lossofa tumor-supressorgene on chromosome22 appears relevantinmeningiomas. Further geneticresearch maylead to an abil-
itytodetectthose patients at risk for CNS neoplasms, and perhaps tothediscoveryof methods of repairing abnormal oncogenes. In thenearfuture, the optimal approach to brain tumors will involve thedeterminationof the cytogenetic profile of every tumor and a precisedefinition of its prognostic features. Only then can completeand specific therapy be properly planned. (2) Immunohistochemistry has proved very useful in identifyingcell types within tumors. Research on functional tumor pathophysiologymay identify more of the factors elaborated by tumorsthat relate to why certain tumors are soft or hard, why somearehighlyvascularbut others not, and why some are very adherentto surrounding structures while others have clear and easilydissectedsurgicalplanes. More has to be known about these featuresif further reductions in microsurgical morbidity are to be achieved. (3) Our observations have not provided a breakthrough in the development of adjunctive therapy. However, in this regard, we know that 95% of these tumors retain a unicentric, unitype appearance during their initial and intermediate phase of growth (in many cases). This implies that if effective adjunctive therapy can be developed, there is a "window" during which it can be applied. Thus, the outlook in this regard is not hopeless. Hopefully,the future treatment of these tumors will resemble that witnessedwith CNS abscesses, so that adjuvant therapy will be curativewithoutsurgery.
153
The Present (1) Our experience favors more precisely defined approaches, and surgical removal of all CNS tumors. The indications for reoperative surgery for intrinsic lesions are greater than previously supposed. (2) The goal of surgery in extrinsic tumors is complete removal using microsurgical techniques, but not at the expense of neurological morbidity. The indications for early reoperation are less clear now that more precise follow-up is possible with MRI. (3) Adjuvant therapies must be specifically adapted to the biological characteristics of each tumor. Mesoscopic insights into tumor behavioral characteristics, in combination with advances in neuropathological knowledge have resulted in a better conceptualization of how operability must be defined. In addition, this permits a better assessment as to a genuine cure, as opposed to a macroscopic "surgical cure." Neurosurgeons must move forwards into the twenty-first century and apply microsurgical philosophy to the full spectrum of CNS tumors in routine practice. It seems very likely that effective biological and chemotherapeutic therapies will be come standard in the management of many intrinsic tumors within the next five years. Precise microsurgical tumor resection to less than log 1 remaining tumor cells will be vital for both the quality and length of survival. Stereotactic methods, though having tremendous patient appeal, are unlikely to achieve this. The goals to be achieved are ancient: to relieve symptoms, to achieve complete tumor removal so far as the nature of the pathology allows, and to accomplish this without injury to normal structures. Within the following section, 45 cases are presented to illustrate the various pathological problems that we have discussed in this chapter. The cases include Case 2.1-2.20 2.1-2.9 2.10-2.19 2.20-2.21 2.22 2.23-2.26 2.27 2.28 2.29 2.30 2.31 2.32 2.33-2.37 2.38-2.40 2.41-2.42 2.43-2.45
Well circumscribed tumors Neocerebral cases Limbic and para-limbic system Rare internal capsular tumors Retrolenticular area Diffuse tumors Slow growing tumor within a highly eloquent area Fast growing recurrence (within 4 weeks). Natural history of a glioma Fast growing unicentric recurring tumor Recurrence occurring outside anatomic borderlines Recurrence at an unexpected site Unpredictable behavior of multicentric tumors Multicentric, multitype tumors Problematical histological diagnosis
Additional relevant cases can be found at the end of chapters 3, 4, and 5 of this volume and within volume IVB.
r i
154
,
2 Neuropathology
I
Cases
Gyral Localization 01 Neocerebral Tumors
a
e
b
---.
Case 2.1 A 28-year-old male who had grandmal epilepsy for 1 year, without any neurological deficits. MRI views: a sagittal (T1), b coronal (T,). There is a well-delineated gyrallesion in the posterior part of the left superior frontal gyrus (F 1). Note the displacement of the middle frontal gyrus (F 2) laterally, and the conical extension in
d the direction of the ventricle. Postoperative views (T1) (two years postoperatively): e sagittal (paramedian), d corona!. The exact position of the tumor (anaplastic astrocytoma) within a segment 01F 1 is now well seen.
Cases
a
b
e
d
Case 2.2 A 41-year-old female who had had two Jacksonian seizures (with postictal aphasia) over 1 year. Neuroimaging views: a MRI(T1),coronal, b MRI (spin-echo), horizontal. e MRI (T1), sagittal. There is a well-delineated lesion in the opercular part of the left inferior frontal gyrus (F 3), with displacement of the inferior part of the precentral gyrus posteriorly. Herniation into the sylvian fissure. The neurological examination revealed a moderate receptive dysphasia. d Postoperative CT(contrast, one week postoperatively, horizontal view). The exact position of the tumor (oligodendrog/ioma, Grade 11)is now well seen. The tumor was approached through F 2 superior and anterior to Broca's area, and the patient's language difficulty resolved postoperatively.
e
155
156
2 Neuropathology
a
b
e
d
Case 2.3 A 25-year-old female suffering increasing focal seizures
lesion lies within the postcentral gyrus, where some hyperintensily
(dysesthesia in the right leg) over 1 year. MRI views: a Horizontal (T2),b sagittal (T1).There is a well-delineated lesion in the supf3rior parietal lobule, adjacent to the postcentral gyrus (superior part). Note that the sagittal view gives the mistaken impression that the
can be seen. Postoperative
MRI (T1, 3 weeks postoperatively):
e hor-
izontal, d sagittal. The exact position of the tumor (astrocytoma, Grade 11)is now clear. The patient remains neurologically normal.
Cases
157
Case2.4 A 66-year-old lemale withacute left leg weaknessand incontinence.MRIviews(T1): a sagittal,b coronal. Thereis a welldelineatedlesion in the rightparacentralgyrus. Notethe miId displacement01the surrounding gyri.Theright cingulate sulcusis occluded. Postoperativeviews(two weekspostoperatively): e sagittal,d coronal. Onlya small-intensity changeidentiliesthe site 01thiscavernomaafter removal.
b
e
d Case2.5 A 12-year-old male who had suffered parietal complete selzuresover 2 years. MRI view (T1): a sagittal. There is a welldelineatedlesion in the right precuneus, herniating into the lower partof the parietooccipital sulcus. Postoperative view: (31/2years
postoperatively): b sagittal. The exact position of the tumor (pilocytic astrocytoma, Grade 1)is now clear. The child remains seizurefree.
158
2 Neuropathology
a
b
e
d
Case 2.6 A 25-year-old woman with progressively severe leftsided frontal seizures over 10 years. MRI views: a horizoQtal(T1), b coronal (T2).There was a well-encapsulated lesion in the right middle parietal lobule. Its behavior was followed up with regular MRIexaminations over 6 years. Note that the lesion extended to the
surface. Postoperativeviews (1 year postoperatively):e horizontal (T2),d coronal (T1).A transsulcal approach was necessary, as this partially calcified oligodendroglioma (Grade 1/)was covered by a displaced thin layer of cortex. The patient remains seizure-free. Preoperatively and postoperatively, there were no neurological deficits.
Cases
159
b
d
Case2.7 A 40-year-old female suffering from right-sided epilepsy (aver10 years), progressing to mental slowing. MRI views (T1): a sagittal, b coronal,e horizontal. There is a well-delineated lesion in the left inferior lobule. Note the displacement of the deep structures herniation. The patient had no dysphasia, sensomotoric or visual field deficits. Postoperative views (one week postoperatively) d caronal, e sagittal. The exact position of the tumor (anaplastic astrocytoma) is now well seen. The patient's language function and parietallobe deficit improved, and she was able to return to work (untilthe tumor recurred 21/2 years later).
e
160
2 Neuropathology
a
e Case 2.8 A 35-year-old female suffering migraines (associated with scintillating scotomas in the right visual field). MRI views (T1): a horizontal, b sagittal. There is a well-delineated lesion in the right cuneus, both superior and inferior to the calcarine sulcus. This cystic lesion caused no visual-field deficit. Postoperative views
d (2 weeks postoperatively): e horizontal, d sagittal. This cystic ganglioglioma (Grade 1)was removed from a small incision in the occipital pole through a "yellowish"-Iooking transparent part of gyrus. The patient's visual fields remain normal.
Cases
161
Tumors of the Limbic Lobe Case2.9 A 16-yearoldfemalewith multipiepartialcomplex seizures(Iip smacking,generalized weakness,dizziness).MRIviews (T1)a coronal, b sagittal.A welldelineatedlesion in theleftamygdala. Postoperativeviews (14weekspostoperatively):e coronal,d sagittal. A parietalhippocampectomy(anterior two-thirds)revealed thisganglioglioma, Grade1.The patient remainsseizure-free.
a
b
e Case2.10 A29-year-oldman withgrandmal epilepsy.MRIview: acoronal.A welldelineatedlesion seemsto be spreadingtheentiretemporallobe.Postoperativeview(1 month postoperatively): b coronal.The positionof the tumor in theparahippocampusandamygdala (displacingthe temporalgyriIaterally, andnotinfiltrating)is wellseen.This anaplasticastrocytoma (GradeJ//)was removedvia a pterional-transsylvian approach.The patientremainssymptom-free 3yearsaftersurgery.
a
b
162
2 Neuropathology
a
e Case 2.11 A 39-year-old male with intermittent petitmal seizures, progressing to focal (right side) and then generalized seizures over 3 years. Despite radiation therapy, language deficits then arase. MRI views (T1):a horizontal, b sagittal, e corona!. There is a welldelineated medial-basal temporal lobe lesion, extending fram the temporal pole into the parahippocampus gyrus and fillingthe paramesencephalic cisterns. Note the displacement of the cerebral peduncle (a) by the herniating mass (but producing no clinical
signs). Postoperative view (1 week postoperatively):d corona!. A vascularized, very adherent pleomorphic xanthoastrocytoma (Grade 11I)was removed fram CN 111, and the anterior choroidal and posterior communicating arteries (and branches) in the paramesencephalic cisterns. The position of the tumor in the parahippocampus gyrus is well seen. Twoyears after surgery, the patient remains asymptomatic.
Cases
a
b
e
d
Case2.12 A 39-year-old female with partial complex seizures and progressivemental deterioration. MRI views (T1): a horizontal, b sagittal.A well-delineated right hippocampal and parahippocampallesion.Note its extension into the lateral temporo-occipital gyri andisthmuscinguli, with herniation into the trigonum and paramesencephaliccistern. There is also displacement of the cerebral
163
peduncle, without clinical signs. Postoperative views (1 year postoperatively): e horizontal d sagittal. The exact position of this tumor (anapIas tic astrocytoma) within the transitional zone (parahippocampus gyrus and hippocampus) is clear. This patient continues to do well 2 years postoperatively, with only an upper-Ieft quadrantanopsia, as a residual effect.
164
2 Neuropathology Case 2.13 A 54-year-old woman who had been suffering seizures for 2 years (with an aura), and a CT scan that suggested left temporal infarction. Over four years, she developed miId language deficits and right-sided apraxia. MRI views: a sagittal (T2), b horizontal (T1). e coronal (T1).There is a well-delineated lesion within the left insula, filling the sylvian fissure. Note the sharp medial border, with displacement and compression (but no infiltration) of the putamen and pallidum. Pastaperative views (3'/2years postoperatively): d sagittal (T1),e horizontal (T1),f coronal (T1).The precise location of the tumor (mixed gliama, Grade 1/)is evident. This patient remains asymptomatic 4 years following surgery.
e
Cases Case 2.131
1>
a
Case2.14 A 33-year-old female with focalseizures(Ieft hand) progressingto generalized seizures over1year.MRI views: a horizontal (T2),b coronal(T1),e sagittal (T1). A well-delineatedlesion in the right insula,withcompressionof the corpusstriatumand herniation into the sylvianfissure.Postoperative views (6 monthspostoperatively): d coronal(T1),e sagittal (T1).The exactpositionof the tumor (fibrillary astrocytoma,Grade 11)is well seen. Thepatientremainsseizure-free 4 yearspostoperatively,with no neurologicaldeficits. She is fully rehabilitated.
d
e
165
166
2 Neuropathology
b
a
e
Case 2.15 A 34-year-old male, initially with focal
e
seizures (right-sided dysesthesia, dysacusis, and loss of speech) and then a grandmal seizure. MRI views (T1): a horizontal, b coronal e sagittal. A large left limbic lobe lesion in the part of the insula with extension into the fronto-orbital, temporal pole, and amygdala. Note the sharp delineation against the striatum. Postoperative views (6 weeks postoperatively): d sagittal, e horizontal. The exact position of the tumor (anaplastic astrocytoma) is better seen. The patient remains seizure-free 5 years postoperatively, with no neurological deficits.
Cases Case2.16 A 50-yearoldmalewith a history 01headaches,mental slowingdiplopia, and rightlegweakness, over2 months.MRI views:a horizontal(T2), b coronal(T1),e sag¡ttal(T1).A welldelineatedlesion in the leltIronto-orbital region,involvingthe septaland paraollactoryareas.Note the displacementsurroundingthewhitematter, withoutinliltration.Postoperativeviews (7monthspostoperatively):d sagittal, e horizontal, f coronal.The position 01thetumor(glioblastomaGradeIV) in the paraolfactoryarea is clear.Thepatient remainsneurologically intact11/2years postoperatively.
167
a
b
e
d
e
168
2 Neuropathology
a
e
d
Case 2.17 A 63-year-old female with focal seizures
e
(speech arrest over 6 months, with pragressive headaches, mental slowing, loss of balance, and bilateral dysdiadochokinesia. MRI views: a horizontal (T2). b sagittal (T1, e coronal (T1). There is a lesíon in the paraolfactorial, septal areas and anterior cingulate gyrus. Note the severe herniation into the ventricular system. without penetration or infiltration. Occiusive hydrocephalus. Postoperative views (1 month postoperatively): d sagittal (T1), e horizontal (T1). The position of the tumor (anaplastic astrocytoma, Grade 111) is clear. The patient remains seizure-free more than 2 years postoperatively, with no neuralogical deficits. The psycho-organic syndrame has partially improved.
Cases
a
Case2.18 A 16-year-old female who had been suffering headaches,diplopia, and apathy for 3 months. MRI views: a horizontal (T2),b coronal (T1),e sagittal (T1). A well-delineated lesion in the middlepartof the cingulate gyrus, with severe ventricular compressionand subfalcial herniation,but no infiltration into the opposite side.Thereismarked perilesional hyperintensity within the right fron-
169
b
tal white matter, extending peripherally to sector 1. Postoperatively view (10 days postoperatively): d sagittal (T1). The location of the tumor a giant-cell glioblastoma, Grade IV), in the middle portion of the cingulate gyrus with small extension into the corpus callosum, is now well seen. Postoperative radiation therapy. The patient has been doing well for 2 years, with no seizures.
170
2 Neuropathology
a
e ~
b
d
J.
Case 2.19 A 39-year-old male with petitmal seizures over 2 years. MRI(TI) views:a sagittal,b corona!.A well-delineatedlesioninthe posterior portion of the cingulate gyrus, extending into the precuneus. Postoperative views (1 year postoperatively): e sagHtal,
d corona!. The histology indicated a mixed oligoastrocytoma, 11. Note that the posterior body of the corpus callosum has
Grade
also been partially removed. The patient remains symptom-free.
I I
I
~
Cases
171
RareLocalizations TumorsArising from the Internal Capsule
a
b
Case2.20 A 12-yearoldlemale,right handed,with a history 01 dizziness,latigue, diplopiaand right-sided weaknessover 3 months. MRIviews:a horizontal (T1),b coronal(T1), e sagittal(TJ A well-circumscribed,partially cysticlesion compressingthe lentilorm nucleuslaterally,the head01the caudate anterosuperiorly, and the thalamusmedially.Note thesevereventricular displacement.Postoperativeviews(8 monthspostoperatively): d sagittal(T,), e horizontal(T2),f coronal (T1).A vasculartumor (pilocyticastrocytoma) wasremovedusing an interhemispheric, transcallosalapproach. Entrancewas gained throughthe anterior thalamus,and the tumor originseemedto be in theregion01the internal capsulenearthe genu. Surprisingly, the patient remainsneurologically intact,withno speech difficulties. e
f
172
2 Neuropathology Case 2.21 A 4-year-old female, right handed, with progressive right-sided weakness (Ieg more than arm) and dysarthria over 2 years. MRI views (T1): a horizontal, b horizontal, e sagittal, d coronal. A large, well-circumscribed les ion in the area of the thalamus had been assumed preoperatively. The origin of this tumor is unclear. Note the tremendous displacement of the surrounding structures and the herniation of the tumor across the midline and into the left frontal horno Postoperative views: e sagittal, f coronal, 9 horizontal. A left-sided interhemispheric transcallosal approach was used to remove this tumor (a pilocytic astrocytoma, Grade 1), between the b thalamus and the head of the caudate. The tumor originated from the area of the posterior limb of the internal capsule. The medial internal capsule and lenticular nucleus are intact from its origino There was improvement of the dysphasia, and weakness of the right distalleg. Follow-up 4 years.
a
d
e
e
f
9
Cases
173
Rentrolenticular Tumors
a
b
e
d
e
f
Case2.22 A 9-year-old female with papilledema, left dysdiadochokinesia,dysmetria, loss of balance, and a left lower quandrantanopsia.MRIviews:a horizontal(T1),b coronal(T1),e sagittal(T1). Awell-circumscribedcystic lesion in the posterior part of the insula, beneaththe long insular gyrus. Postoperative views (9 months post-
operatively): d sagittal (T1), e horizontal (spin-echo), f coronal (T,). The exact origin of the tumor (pilocytic astrocytoma) was in the retrolenticular region. Postoperatively, the child had a moderate hemiparesis, which improved to permit fine motor skills after one year.
174
2 Neuropathology
Diffusely Growing Gliomas
a
b
Case 2.23 A 63-yearold female with occasional partial complex seizures and no neurological deficits. MRI views (T2): a coronal, b horizontal. A diffuse lesion in the left frontal, temporal, parietal, and occipitallobes, with displacement of the deep structures. Four years later, the patient remains without any neurological deficits and professionally active. Where is the tumor origin? Note that the anterior commissure can be well visualized, and that the putamen, pallidum, thalamus, and caudate all remain intact.
Case 2.24 A 62-yearold male with progressive headaches and seizures. MRI views: a coronal (T1), b horizontal (T,). A diffuse brainstem and cerebellar lesion, with multiple other supratentorial and bilateral mediobasal temporallesions. After a biopsy, this proved to be
a glioblastoma.
a
Cases
a
175
b
Case2.25 A 37-year-old female, right-handed, with focal sensory seizuresover 10 years. MRI views (T1): a horizontal, b sagittal, ecorona!.A diffuse lesion throughout the central and parietallobes, with extension into the pyramidal system. The patient refused surgery,and remained normal for 8 years (very vital). Her only troublefaryearswas occasional seizures, despite optimal and regularpharmacotherapy.She then developed a right hemiparesis and diedwithin 3 months (of paraneoplastic effects). An autopsy was refusedby the family. The assumed diagnosis was astrocytoma,
Grade11-11I.
e
a Case2.26 A 24-year-old male who had suffered several focal sensory seizures (Ieft side) and then two grandmal seizures. MRI v¡ewsa coronal (T2)' A diffuse les ion in the right superior and medial parietallobulus, with displacement of the ventricle. Postoperatively view(2 weeks postoperatively): b coronal (T1). An oligoastrocytoma
(Grade 11) was partially removed,
and the patient
remains
completely neurologically normal, but the seizures could not be controlled with pharmacotherapy. A thalamic stereotactic lesion was performed 1 year later to further treat the seizure disorder, and this was very effective in repressing the seizures. Thereafter, the patient regained full working capacity 2 years ago.
176
2 Neuropathology
Case 2.27 A 12-year-old male with epileptic seizures. MRIviews (T1). a horizontal, b coronal. There is a large, circumscribed les ion in the left thalamus, with elevation of the caudate. A stereotactic biopsy was performed, revealing an astrocytoma, and the child continues to do well (after 4 years), with no deficits, and attends school.
a Case 2.28 A 4-year-old girl with 10 days of headache and vomiting, but no neurological deficits. a MRI (T1), horizontal. A large, insular primitive neuroectodermal tumor (Grade IV) was totally removed, reducing the mass effect. The patient enjoyed an immediate recovery. b Follow-up view (4 weeks postoperatively). The arrows indicate residual or recurrent tumor. e MRI (T1, 8 weeks postoperatively). A large recurrent tumor at the same site. Clinically, there was headache and vomiting, but no neurological deficits. d Contrast CT, horizontal (1 week after the second operation). A subtotal removal, with reduction in mass effect, was carried out, and the patient enjoyed a recovery. However, 4 weeks later, the same symptoms arase again. No further therapy was applicable, and the child died 6 weeks later radiotherapy
was rejected.
a
b
Cases
b
a
d
e
Case 2.29 A 35-year-old male who had had a epileptic seizure 14 years previously and recent focal seizures, later progressing to another grandmal seizure. MRI views (T1):a horizontal, b sagittal. No clear lesion is identified. Possible intensity changes in the left mesiobasal temporal region were suggested. The patient remained asymptomatic tor 21/2 years, but then developed papilledema and a right homonymous hemianopsia. e sagittal. A definitive, well-delineated lesion is seen in the inferior temporal gyrus (T 3) of the temporal lobe. The patient wanted to discuss other options tor treatment. Subacute impairment occured 6 months later, and the MRI showed a large temporal lesion. Postoperative views (3 weeks postoperatively): d sagittal, e coronal. The lesion (glioblastoma, Grade IV) was removed. Postoperative radiotherapy and chemotherapy were provided. The patient died 3 months later.
177
r 178
2 Neuropathology
a
b
e
d
e Case 2.30 A 32-year-old male, right-handed, suffering generalized seizures, dysphasia, and weakness of the right side for only a few hours. MRI view: a A left insular tumor with hematoma in the center. A glioblastoma, Grade IV, was removed. b Postoperative CT. The patient refused radiotherapy, and consulted with a molecular biologist and immunologist. He continued to work as a surgeon. e First recurrence of the tumor, 5 months later. Postoperatively MRI views after the second operation: d Recurrent tumor at the same
site. The patient underwent further two operations within 6 months for recurrence at exactly the same site as the original tumor. He remaind neurologicaliy and mentaliy intact until the very end, and had no speech difficulties in three languages. In April 1992, there was acute deterioration. MRI view: e There was a local recurren! tumor, and he died. Particularly in the case of this young colieague, the absence of effective and adjuvant therapy was intolerable.
Cases
a b Case2.31 A 43-year-old male suffering epileptic seizures and speechdifficulties for a few hours. a MRI view: A small, circumscribedlesion in the left frontal operculum, part F 1. Extensive perilesional hyperintensitywithin the white maUer of the left'hemisphere. Befareand during surgery, a metastatic lesion was assumed. The
179
e tumor was sharply delineated. b Postoperatively CT:The histology, however, revealed a glioblastoma, Grade IV: Postoperative radiotherapy was applied. There was positive follow-up for only 4 months without neurological deficit. Therafter, there was acute deterioration with hemiparesis and aphasia. e MRI view: A huge recurrent tumor.
Case 2.32 A 31-yearold male, with headaches, papilledema, cerebellar signs, apathy, and a left lateral field cut MRI views: a horizontal (T2), b sagitlal (T). There is a diffuse leslon in the dorsal mesencephalon
and
adjacent thalamus, extending into the paraplneal region. This anaplastic astroeytama (Grade 111)was completely removed from its origin in the right pulvinar thalamic Postoperative
area.
a
b
radiother-
apy was provided,
and
the patient had a full recovery, working
as a
carpenter far 14 months, before developing
a pro-
gresslve psycho-organic syndrome. Note the second tumor in the frontal basal area. e sagitlal (T.) Surprisingly, was no evidence
there of
tumor recurrence
in the
mesencephalic region, but there was a second tumor In the right medial frontal area, which was removed by a second operation. d sagitlal
(T1,
2 weeks postoperatively).
d
180
2 Neuropathology
Unpredictable Behavior 01 Tumors
Case 2.33 A 40-yearold woman with occasional Jacksonian seizures manifested by right-sided weakness. Previously she had undergone removal of a left posterosuperior frontal (F 1) gyrus tumor (astrocytoma, Grade 1), with full recovery. Eighl years later, she began lo have headaches, increased seizures, and right leg weakness. MRI views (T1): a coronal, b sagittal. A recurrent tumor in the anterior parl of the previous tumor cavity was removed. Postoperative MRI views (4 months postoperatively): e coronal, d sagiltal. Histology: anaplastic astrocytoma, Grade 111. Radiotherapy was recommended. The patient is symptom-free 2 years after 2nd operation.
a
d
Cases
a
181
b
e Case2.34 A 26-year-old male suffering subacute headaches followedby a grandmal seizure. No neurological deficits. a Contrast Cl horizontal.There is a lesion in the left posterior temporal inferior gyrus(T3). The tumor (fibrillary astrocytoma Grade 11)has been removed.b Postoperativeview (1 month postoperatively). Preoperativelyand postoperatively, there were no visual field defects. Four yearslater,another seizure occured. e MRI (T1), horizontal view:
Tumor recurrence at the site of the original tumor, with perilesional diffuse edema. The location of the recurrent tumor is clear. Note the displaced position of the left optic radiations in this patient, whose visual fields remained intact. d Postoperative MRI view (2 weeks postoperatively). The patient continues to do well (2 years after 2nd operation), but has homonymous hemianopsia.
182
2 Neuropathology
d
Case 2.35 A 20-year-old woman with increasingly frequent migraines, diplopia, papilledema, hydrocephalus, and ataxia. a Horizontal contrast CT. A large, delineated septallesion with occlusive hydrocephalus. The tumor (neurocytoma) has been completely removed, and the hydrocephalus relieved. Recovery was good. Postoperative views (3 weeks postoperatively): b horizontal con-
trast CT.The same symptoms recurred six years later.e Horizontal MRI (T1): A large recurrence in the area of the original tumor,with occlusive hydrocephalus. Postoperative view (2 weeks postoperatively): d MRI (T1), horizontal view. Complete removal of the tumor. Note that the septum pellucidum has been removed. The patient continues to do well.
Cases
183
b
--
l.
i I
I
'"
.,."
J
Case2.36 A 58-year-old lemale with progressive conlusion, fatigue,loss 01 concentration, left hemiparesis, and lelt homonymaushemianopsia.a Horizontal (contrast) CT (1978). There is a dil¡uselesionin the right pulvinar thalami, with a large extension into thetrigonum.b Contrast CT, 10 years later. This tumor (glioblastoma,Grade IV) was removed radically. Postoperatively, conventianalradiotherapy was applied. The patient recovered lully and workedwithoutany disability lor 5 years until her retirement at the
age 0165. She died 01urinary carcinoma with metastases 14 years postoperatively. e, d The histology 01her tumor, which was lound to be glioblastoma also by Prol. Zülch and Prol. Rubinstein. Histology: A highly cellular, anaplastic tumor with increased polymorphism, multinuclear tumor cells, atypical mitotic ligures, and local necrosis with pseudopaíisading 01 tumor cells. In some areas vascular endothelial prolileration has also been observed. The diagnosis 01 glioblastoma multilorme (WHO Grade IV) was made.
184
2 Neuropathology Case 2.37 A 34-year-old man with epilepsy. a Contrast Cl A lesion in the left fronto-orbital area, with calcification. The tumor (anap/astic astrocytoma) was radically removed. Postoperative radiotherapy was given. The patient remains seizure-free, with no mental or neurologic defiQits. Postoperative views (9 years postoperatively): b MRI (T1),horizontal, e MRI (T1) coronal. No recurren! tumor.
a
b
Multiple Tumors
b
a
Case 2.38 A 58-year-old male, suffering headache, psychoorganic syndrome, and rapid deterioration. MR/ (T1): a horizontal, b sagittal. An unusual tumor within the entire callosum, septum pellucidum, anterior and posterior cingulate gyrus. There is a butterflytype bilateral extension along the forceps minor. No surgery was performed. The patient died before any treatment. Histology: G/iob/astoma.
Cases
185
Case2.39 A 68-year-old womansufferingrapidly progressivedeterioration. a-d MRIviews(T1).Multiple lesionsare identified bilaterally inthecerebraland cerebellar hemispheresand cisterns. The autopsyrevealed glioblastoma.
a
e
b
186
a
e
2 Neuropathology Case 2.40 A 10-year-old male with partial complex seizures and progressive forgetfulness, fatigue, and headaches. MRIviews(T1): a horizontal, b sagittal. Multiple lesions, particularly in the right mediobasal temporal region, with extension into the cisterns (right sylvian, interhemispheric, bilateral ambient, and left collateral). Postoperative view (6 weeks postoperatively):e horizontal. The large compact tumor (pilocytic astrocytoma) in the right amygdala and hippocampus was removed. The cisternal parts of the tumor (adhesive to arteries) could not be resected. No additional tumor growth was seen within 4 years, but tumor remains in the cisternal and su leal regions. The patient is symptom-free, with no neurological deficits, and attends school after 4 years. Follow-up views: d coronal, e sagittal.
Cases
Multicentricand Multitype Tumors Case2.41 A 59-year-old lemalewith subacute headaches,papilledema, homonymOJShemianopsia,and ataxia.CTcontrast views: a horizontal,b horizontal. Two lesionsare identified in the righttrigone(with surrounding perilesionalchanges) and alongthe falx. 80th tumors, a lalx meningiomaand oligoastrocytoma,Grade 11with adjacentcavernomawhich bled,werecompletely removedin one session. The patientremainedsymptomfreelor 7 years. Postoperativeviews (1 month postoperatively): e horizontal,d horizontal.
187
188
2 Neuropathology
b
Case 2.42 A 43-yearold female who had suffered a grandmal seizure and sudden coma. MRI views (T1): a coronal, b horizontal, e sagittal. Three lesions are identified. The right sphenoidal meningioma is associated with a significant shift in the underIying deep structure, while the left-sided thalamic lesion appears with a hematoma, and there is a third lesion in the third ventricle. Postoperative view (1 week postoperatively): d coronal. A left thalamic and third ventricular tumor (cavernoma) has been removed. At a second session, the sphenoid meningioma was removed 3 months later. The patient persisted with significant disability in a convalescent home.
d
e
Problematic Histological Diagnosis
Case 2.43a, b.
a
b
Cases
189
d
e
9 Case2.43 A 31-year-old female with progressive headaches and dizziness.No neurological deficits. MRI views (T1): a horizontal, b sagittal,a large lesion is identified in the right parietallobe. This strangetumorwas subcortical, with no relation to the dura and to the ventriclewall. The patient's visual fields remained intact. The histologyremainedunclear as to whether it was an atypical meningiamaoran anaplasticoligodendroglioma. MRI views (4 months postoperatively):e horizontal (T2).d coronal (Postero-anterior view). No residualtumor seen. Fourteen months later horizontal MRI view e withrecurrenttumor present, which was removed. Again the histoiogyremainedunclear. f horizontal MRI view. One year following the secondoperation. there is no evidence of tumor recurrence,
9 deeper horizontal view (T1). The optic radiations are preserved and the visual fields are normal and the patient remains asymptomatic. h Histology: A highly cellular tumor with mitotic activity consisting of small cell strands and nests, sometimes resembling whorls. A mucoid intercellular substance and PAS-positivity of tumor cells was noted. Immunohistochemically Pancytokeratin antibody, Epithelial Membrane Antigen (EMA) and Glial Fibrillary Acidic Protein (GFAP) were completely negative. Positive expression has been shown with anti-Vimentin and focally also with anti-S100 antibodies. In summary, histology and immunohistochemistry were compatible with anaplastic meningioma of a chordomatoid subtype (WHO Grade 111).It may be an atypical meningioma.
190
2 Neuropathology
.
e ~
'" "'''""",:,..", ;
.'~
'" ,.t. ,_". .# .
."
. ..
..
,¡. ..
.
.., ....
'"
d
..:
Case 2.44 A31-year-old male suffering grandmal seizures. Neurologicallyand mentally,there were no deficits. MRI views (T1):a sagittal, b sagittal, e corona!. Multiplelesions are identified in the left F 3, leftanterior cingulate gyrus, and right paraolfactorial gyrus. ThBhistology from the tumor biopsy was not definitive, suggesting either low grade tumor or infection. The patient has remained asymptomatic for2 years, withentirely unchanged MRIviews. d, e Histology;
- -.
lt,..
!'
,...
...
,
.
.. ~~.$
. '" , . .,~ i
'
. ..'....
i' :t '¡'V,. ¡ ~. \.N . ~~ ~. .,
.. ,..:(,
"
J(
"
, .. .
,
.1
J
,"" ..1:
,. ).. ,.
'N
.,
....
A lesion with reactive gliosis and inflammatory (Iymphoid) cell infiltration was observed histologically. There was no sign of a neoplastic tissue, monoclonality of infiltrating Iymphocytes, or necrosis. In special stains (Ziehl-Neelsen, toxoplasmosis-antibodies, fungal stainings) no germs or parasites could be identified. Thus, the definitive nature of the lesion could not be diagnosed histologically.
Cases
a
e Case2.45 A 26-year-old male suffering headaches, diminished balance, and right-sided dysmetria. MRI views (T1): a coronal (posteroanteriorview), b horizontal, e coronal, d sagittal. Multiple lesions are identilied. Open biopsy revealed a medulloblastoma, Grade IV. Radiotherapywas applied. Clinically the cerebellar lesion within the rightAOL concurs with the diagnosis 01 medulloblastoma, but what is unique is the preoperative dissemination 01 this lesion to both cavernous sinuses as well as into the sellar and parasellar areas. Histology:A malignant, highly cellular, and mitotic tumor wifh typical
191
b
d carot-shaped nuclei. Focal neuronal differentiation could be detected histologically and demonstrated with positive synaptophysin immunohistochemistry. Glial fibrillary acidic protein stained locally positive in trapped astrocytes. According to its morphology the tumor was graded WHO IV and termed Primitive Neuroectodermal Tumor (PNET). II its primary site would have been the cerebellum, medulloblastoma would be the name for the otherwise histologically identical neoplasia. Meningial seeding is a well-known leature 01these highly malignant tumors.