Malignant Lymphoma
This page intentionally left blank
Malignant Lymphoma Edited by
BW Hancock MD FRCP FRCR Profess...
71 downloads
1327 Views
47MB 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
Malignant Lymphoma
This page intentionally left blank
Malignant Lymphoma Edited by
BW Hancock MD FRCP FRCR Professor of Clinical Oncology and YCR Director of Cancer Research, The University of Sheffield, UK
PJ Selby MD FRCP Director, ICRF Cancer Medicine Research Unit, St James's University Hospital, Leeds, UK
K MacLennan DM FRCP Professor of Tumour Pathology, ICRF Cancer Medicine Research Unit, St James's University Hospital, Leeds, UK and
JO Armitage MD Professor and Chairman, Department of Internal Medicine, Section of Oncology and Hematology, University of Nebraska Medical Center, Omaha, Nebraska
A member of the Hodder Headline Group LONDON Co-published in the USA by Oxford University Press Inc., New York
First published in Great Britain in 2000 by Arnold, a member of the Hodder Headline Group, 338 Huston Road, London NW1 3BH http://www.arnoldpublishers.com Co-published in the United States of America by Oxford University Press Inc., 198 Madison Avenue, New York, NY10016 Oxford is a registered trademark of Oxford University Press © 2000 Arnold All rights reserved. No part of this publication may be reproduced or transmitted in any form or by any means, electronically or mechanically, including photocopying, recording or any information storage or retrieval system, without either prior permission in writing from the publisher or a licence permitting restricted copying. In the United Kingdom such licences are issued by the Copyright Licensing Agency: 90 Tottenham Court Road, London W1P 9HE. Whilst the advice and information in this book are believed to be true and accurate at the date of going to press, neither the authorfs] nor the publisher can accept any legal responsibility or liability for any errors or omissions that may be made. In particular (but without limiting the generality of the preceding disclaimer) every effort has been made to check drug dosages; however it is still possible that errors have been missed. Furthermore, dosage schedules are constantly being revised and new side-effects recognized. For these reasons the reader is strongly urged to consult the drug companies' printed instructions before administering any of the drugs recommended in this book. British Library Cataloguing in Publication Data A catalogue record for this book is available from the British Library Library of Congress Cataloging-in-Publication Data A catalog record for this book is available from the Library of Congress ISBN 0 340 74207 0 12345678910 Commissioning Editor: Joanna Koster Project Editor: Sarah de Souza Production Editor: James Rabson Production Controller: Fiona Byrne Project Manager: Marian Haimes Typeset in 10/12 pt Minion by Phoenix Photosetting, Chatham, Kent Printed and bound in Great Britain by The Bath Press, Bath
To the memory of Tim McElwain and Mike Bennett for their major contributions to an understanding of the clinical management and pathology of malignant lymphoma
This page intentionally left blank
Contents
Color plates appear between pages 50 and 51 Contributors Preface Foreword
PART 1
1
HISTOPATHOLOGY
4
5
6
7
8
9
3
Hodgkin's disease
KA MacLennan, B Vaughan Hudson, G Vaughan Hudson 3
1
Lymphoma classification KA MacLennan
2
ix xi xiii
9
Follicular lymphoma KA MacLennan
21
Mantle cell lymphoma DD Weisenburger, JO Armitage
27
Diffuse indolent B cell neoplasms KA MacLennan
43
Diffuse aggressive B cell lymphoma KA MacLennan
49
T cell lymphoproliferative disorders AS Jack, SJ Richards, KA MacLennan
55
Extranodal lymphomas PG Isaacson
71
Cytogenetics WG Sanger, BJ Dave, MR Bishop
91
PART 2
PATHOGENESIS
105
10
Hodgkin's disease V Diehl, J Wolf
107
Viruses and malignant lymphoma LM Weiss, KL Chang
115
Molecular biology VI Pappa, BD Young
133
11
12
viii Contents
PART 3
EPIDEMIOLOGY
13
Hodgkin's disease NE Mueller
14
159
161
Non-Hodgkin's lymphoma RA Cartwright
169
PART 4
CLINICAL MANAGEMENT
179
15
Hodgkin's disease: clinical features PWM Johnson, PJ Selby, BW Hancock
16
Imaging of lymphoma K Sandrasegaran, PJ Robinson, A Sprigg
17
421
The way forward BW Hancock, PJ Selby, JO Armitage, KA MacLennan
Index
399
Long-term problems M Henry-Amar
31
385
Infections B Crosse, PJ Selby
30
371
Lymphoma in the elderly PWM Johnson
29
359
Pediatric lymphomas JS Malpas
28
351
Cutaneous lymphomas RT Hoppe, YH Kim
27
331
AIDS-related lymphoma AM Levine
26
325
High-dose therapy PJ Bierman, JM Vose, JO Armitage
25
309
Other low-grade non-Hodgkin's lymphomas JA Radford
24
299
Follicular lymphoma TA Lister, AZS Rohatiner
23
287
Lymphoblastic lymphoma in adults JW Sweeten ham
22
269
Aggressive non-Hodgkin's lymphoma ER Gaynor, Rl Fisher
21
247
Advanced Hodgkin's disease BW Hancock, PJ Selby
20
221
Localized non-Hodgkin's lymphoma SB Sutcliffe, MK Gospodarowicz, MH Robinson
19
205
Localized Hodgkin's disease SB Sutcliffe, AR Timothy, MH Robinson
18
181
437
439
Contributors
James 0 Armitage
Michel Henry-Amar
Department of Internal Medicine, University of Nebraska
Centre Francois-Baclesse, Service de Recherche Clinique,
Medical Center, Omaha, Nebraska, USA
Caen, France
Philip J Bierman
Richard T Hoppe
Department of Internal Medicine, University of Nebraska
Professor of Cancer Biology, Chairman, Department of
Medical Center, Omaha, Nebraska, USA Michael R Bishop National Institutes of Health, National Cancer Institute, Bethesda, Maryland, USA
Radiation Oncology, Stanford University Medical Center, Stanford, California, USA PG Isaacson Department of Histopathology, Royal Free and University College London Medical School, University Street, London, UK
Ray A Cartwright Director, Leukaemia Research Fund, Centre for Clinical Epidemiology, University of Leeds, 30-32 Hyde Terrace, Leeds, UK Karen L Chang Department of Pathology, City of Hope National Medical Center, Duarte, California, USA
Andrew S Jack Haematological Malignancy Diagnostic Service, Leeds Teaching Hospitals, Leeds, UK PWM Johnson CRC Department of Medical Oncology, Southampton General Hospital, Southampton, UK Youn H Kim
B Crosse ICRF Cancer Medicine Research Unit, St James's University Hospital, Leeds, UK Bhavana J Dave
Associate Professor of Dermatology, Stanford University Medical Center, Stanford, California, USA Alexandra M Levine Professor of Medicine, University of Southern California
Assistant Professor, Pathology/Microbiology and Pediatrics,
School of Medicine, Norris Cancer Hospital, Los Angeles,
University of Nebraska Medical Center, Omaha, Nebraska, USA
California, USA
V Diehl Department of Internal Medicine I, University of Cologne, Germany
T Andrew Lister Consultant Medical Oncologist, Department of Medical Oncology, St Bartholomew's Hospital, West Smithfield, London, UK
Richard I Fisher
KA MacLennan
Professor of Medicine, Director Division of
Consultant Histopathologist, ICRF Cancer Medicine Research Unit, St James's University Hospital, Beckett Street, Leeds, UK
Hematology/Oncology, Loyola University Medical Center, Maywood, Illinois, USA Ellen R Gaynor Professor of Medicine, Division of Hematology/Oncology, Loyola University Medical Center, Maywood, Illinois, USA
JS Malpas Masters Lodge, Charterhouse Square, London, UK Nancy E Mueller
Mary K Gospodarowicz
Professor of Epidemiology, Harvard School of Public Health, Boston, Massachusetts, USA
Professor of Radiation Oncology, Department of Radiation
VI Pappa
Oncology, University of Toronto, Princess Margaret Hospital,
Second Department of Internal Medicine, Athens University,
Toronto, Ontario, Canada
Evangelismos Hospital, Athens, Greece
Barry W Hancock
JA Radford
YCR Department of Clinical Oncology, Weston Park Hospital, Sheffield, UK
Christie Hospital NHS Trust, Wilmslow Road, Withington, Manchester, UK
x Contributors SJ Richards
John W Sweetenham
Haematological Malignancy Diagnostic Service, Leeds Teaching Hospitals, Leeds, UK
University of Colorado Health Sciences Center, Division of
Martin H Robinson
Adrian R Timothy
YCRC Department of Clinical Oncology, Weston Park Hospital
St Thomas' Hospital, Lambeth Palace Road, London
Medical Oncology, Denver, Colorado, USA
NHS Trust, Sheffield, UK
B Vaughan Hudson
PJ Robinson
The British National Lymphoma Investigation,
Department of Radiology, St James's University Hospital, Beckett Street, Leeds, UK
UCH/Middlesex Hospital, London, UK G Vaughan Hudson
Ama ZS Rohatiner
The British National Lymphoma Investigation,
Medical Oncology Unit, St Bartholomew's Hospital, West
UCH/Middlesex Hospital, London, UK
Smithfield, London, UK
Dennis D Weisenburger
K Sandrasegaran
Department of Pathology and Microbiology, University of
Department of Radiology, Birmingham Heartlands Hospital,
Nebraska Medical Center, Omaha, Nebraska, USA
Birmingham, UK
Lawrence M Weiss
Warren G Sanger
Chairman, Department of Pathology, City of Hope National Medical Center, Duarte, California, USA
Director, Cytogenetics Laboratories, Professor, Pathology/Microbiology and Pediatrics, University of Nebraska Medical Center, Omaha, Nebraska, USA PJ Selby Director, ICRF Cancer Medicine Research Unit, St James's
Jurgen Wolf Department of Internal Medicine I, University of Cologne, Germany Julie M Vose
University Hospital, Beckett Street, Leeds, UK
University of Nebraska Medical Center, Omaha, Nebraska,
A Sprigg
USA
Department of Radiology, Sheffield Children's Hospital, Western Bank, Sheffield, UK
BD Young Imperial Cancer Research Fund, Medical Oncology
Simon B Sutcliffe Vancouver Cancer Center, Vancouver, Canada
Department, St Bartholomew's Hospital Medical College, London, UK
Preface
This book gives fiilly referenced reviews of the many aspects of malignant lymphoma. It is a field where it is of particular importance that the clinician be aware of the variations in histological structure in the lesions with which he or she deals; we hope that the histopathology chapters are intelligible to the clinician, particularly as the most recent classification (Revised EuropeanAmerican Lymphoma/World Health Organisation) is likely to be universally accepted. An understanding of the pathogenesis (particularly the role of viruses and relevance of molecular biological discoveries) and epidemiology is likely to improve treatment strategies, so these subjects are also covered in depth. However,
clinical management, in all its aspects, comprises the major section of the book since it is towards improving the patient's lot that we all strive. We have invited an international panel of distinguished experts in all fields to contribute chapters in the hope that the book will give you a truly global interpretation of current and possible future strategies in understanding, diagnosing and treating this wide spectrum of diseases known collectively as the malignant lymphomas. Barry Hancock August 2000
This page intentionally left blank
Foreword
When the history of oncology research is finally written, the lymphomas will be the tumor group that dominates. As a group these tumors have opened up trails where there were no paths in understanding the biology of cancer, in general, and shown how biology can be effectively linked to cancer management. They have also served as a model for how multidisciplinary teams can coalesce to properly diagnose, stage and treat a group of complex cancers. A number of prominent firsts are associated with lymphomas. Hodgkin's disease was the first hematological malignancy described, followed in a few decades by the lymphocytic lymphomas. They were also the first tumors of a major organ system in adults cured by chemotherapy. Treatments developed for them have led to principles that have had applications in many other cancers as well. For many years, however, the field was plagued by a cacophony of pathology classification systems developed by prominent pathologists the world over. Most relied on empiricism and were not easy to learn and use. Always, it was difficult to match data across systems. A major advance occurred when the Working Formulation was developed because it brought some commonality to the language we used and allowed a more accurate crosscomparison of study results. It was, however, not a very scientific classification. Now we have the Revised European-American
Lymphoma (REAL) classification, which has brought order out of chaos by merging science with clinical practice, and it has brought us this book. The REAL classification was first met with considerable skepticism by lymphoma clinicians who feared it would be too unwieldy for practical use. On the contrary its strong scientific foundation and logical display of the numerous lymphoma subtypes has made it easy to remember and the most efficient lymphoma classification yet devised. So it is appropriate to have a new text on the subject that re-arrays the rich lode of information available to us on lymphomas under the new system. The editors and authors of this text have done just that. The content covers the science well, and distills all the information in a practical way and covers the new entities offered in the REAL classification. It is the judgement of this author that this is the definitive text in the field and, that it will be a necessary and welcome addition to the bookshelf of any physician who sees patients with lymphomas. Now if we can only call them what they are instead of what they are not! Vincent T. DeVita Jr, MD Professor of Medicine and Epidemiology and Public Health Yale University School of Medicine 30 June 2000
This page intentionally left blank
PART
Histopathology
Lymphoma classification Hodgkin's disease Follicular lymphoma
3 9 21
Mantle cell lymphoma
27
Diffuse indolent B cell neoplasms
43
Diffuse aggressive B cell lymphoma
49
T cell lymphoproliferative disorders
55
Extranodal lymphomas
71
Cytogenetics
91
1
This page intentionally left blank
1 Lymphoma classification KAMACLENNAN
Introduction Rappaport classification Lukes and Collins classification Kiel classification National Cancer Institute sponsored lymphoma classification project
3 3 4
Revised Kiel classification Revised European-American lymphoma classification Lymphoma classification project
4
WHO classification
5 6 6 7
References
8
4
INTRODUCTION The classification of the non-Hodgkin's lymphomas has been the subject of considerable controversy over many decades. Our understanding of the immune system has increased dramatically over recent years and thus our appreciation of the complexity of its neoplasms has been heightened. These factors, together with the introduction of new methodologies for the study of malignant lymphoma has greatly expanded the number of distinct entities that we now can recognize. This has inevitably led to terminological confusion, and many clinicians regard lymphoma classification with apprehension and distaste. The clinical management strategies for the treatment of malignant lymphoma still revolve around the concepts of low-grade lymphoma, which are regarded as indolent but incurable, and high-grade lymphomas, which are aggressive but potentially curable. However, embedded within these clinical groupings are distinctive biological entities that have widely varying clinical behavior. In order to understand how far we have progressed in the difficult field of the classification of the non-Hodgkin's lymphomas, it is of some value to review the history of lymphoma classifications briefly. Although earlier reports of lymphoproliferative disease are available, the first detailed study and documentation that malignant lymphoma was a distinctive entity was provided by Thomas Hodgkin's classical paper describing the disease,1 which Samuel Wilks generously gave the eponymous name Hodgkin's disease.2 This was
based upon macroscopic findings alone. In later years microscopic study of tissue from cases of malignant lymphoma was undertaken and the bewildering complexity of neoplasms of the immune system became apparent.3"5 The first classification of the modern era to gain widespread acceptance was proposed by Henry Rappaport.
RAPPAPORT CLASSIFICATION The Rappaport classification (Table 1.1) was submitted for publication as a fascicle of the Armed Forces Institute of Pathology Atlas of Tumor Pathology in 1959 but was not published until 1966.6 The classification paid particular attention to the architecture of non-Hodgkin's lymphoma, believing that there were nodular and diffuse subtypes of each cytological form. Although lacking any precision in terminology or lineage, the Rappaport classification proved extremely useful clinically and persisted for more than 20 years.
Table 1.1 Original Rappaport classification of 1956
1. 2. 3. 4. 5.
Lymphocytic type, well differentiated Lymphocytic type, poorly differentiated Mixed type (lymphocyticand reticulum cell) Reticulum-cell type Hodgkin's type
4 Lymphoma classification
Table 1.2 Original Lukes and Collins classification of 1974
Table 13 Original Kiel classification of 1974
I U cell (undefined cell) type II T eel I types Mycosis fungoides and Sezary syndrome Convoluted lymphocyte ?lmmunoblastic sarcoma (of T cells) ?Hodgkin's disease III B eel I types Small lymphocyte (chronic lymphocytic leukemia; CLL) Plasmacytoid lymphocyte Follicular center cell (FCC) types (follicular, diffuse, follicular, and diffuse and sclerotic) Small cleaved Large cleaved Small non-cleaved Large non-cleaved Immunoblastic sarcoma (of B cells) IV Histiocytictype V Unclassifiable
Low-grade malignancy Malignant lymphoma- lymphocytic (CLL and others) Malignant lymphoma - lymphoplasmacytoid (immunocytic) Malignant lymphoma - centrocytic Malignant lymphoma - centroblastic-centrocyticfollicular; follicular and diffuse; diffuse; with and without sclerosis High-grade malignancy Malignant lymphoma - centroblastic Malignant lymphoma - lymphoblastic Burkitt type Convoluted-cell type Others Malignant lymphoma - immunoblastic
In the 1970s the development of modern immunological concepts of T and B cell lineage impacted on lymphoma classification.7
LUKES AND COLLINS CLASSIFICATION The Lukes and Collins classification (Table 1.2) was developed based on cell lineage and morphological alterations associated with lymphocyte transformation in response to antigenic stimulus. It introduced new terminology to describe the varied cell morphology of follicle center cells, and terms such as cleaved and non-cleaved became widely accepted. Central to the classification was the belief that experienced hematopathologists could discriminate between B and T cell lineage lymphoma with reliability.7"13 In Europe, a similar ideology was used by Karl Lennert and members of the European Lymphoma Club to formulate the Kiel classification.
KIEL CLASSIFICATION The Kiel classification (Table 1.3) achieved popularity in Europe and as its central tenet was the concept that the cytology of lymphoma cells could be used as a grading system to predict clinical behavior. The presence of numerous transformed or blast cells was indicative of aggressive disease. Important in the structure of the Kiel classification was the belief that the nodular lymphomas proposed by Rappaport were the neoplastic equivalent of germinal center B cells. The Kiel classification introduced new terminology for follicle center B cells. Although originally termed germinoblasts and germinocytes as equivalent to the large non-cleaved and
small cleaved cells of the Lukes and Collins classification, this was subsequently changed as the terms centroblast and centrocyte were introduced.14"17 The Lukes and Collins and the Kiel classifications were the most scientifically appropriate classifications that existed in their day.18 However, four other classifications were also in use in the early 1970s. These were the Dorfman classification,19 World Health Organisation (WHO) classification,20 the British National Lymphoma Investigation classification21'22 and the updated Rappaport classification.23 These employed different terminology and had different criteria for diagnosis. It is therefore no surprise that there was a degree of confusion that pertained during this period and caused extreme difficulties in the comparison of therapeutic trials in malignant lymphoma. In an attempt to resolve these difficulties, the American National Cancer Institute sponsored a comparative study of lymphoma classification.
NATIONAL CANCER INSTITUTE SPONSORED LYMPHOMA CLASSIFICATION PROJECT The National Cancer Institute (NCI) lymphoma classification project studied the reproducibility and clinical value of the six major lymphoma classifications that were used in the 1970s. Six pathologists each representing a particular lymphoma classification and six expert hematopathologists who were to use each classification studied 1175 cases of malignant lymphoma accrued at four major oncology institutions (three Northern American and one European: Stanford University, Tufts-New England Medical Center, University of Minnesota Hospitals and the Milan National Tumour Institute). The prognostic significance and reproducibility of each classification was studied. As it transpired there was no significant difference between any of the six classifications in terms of clinical value and reproducibility. An attempt was made to select one classification for
Revised Kiel classification 5
Table 1.4 Working formulation of non-Hodgkin's lymphomas for clinical usage Low grade A. Small lymphocytic Consistent with CLL; plasmacytoid B. Follicular predominantly small cleaved cell Diffuse areas, sclerosis C. Follicular mixed small cleaved and large cell Diffuse areas, sclerosis Intermediate grade D. Follicular predominantly large cell Diffuse areas, sclerosis E. Diffuse small cleaved cell Sclerosis F. Diffuse mixed, small and large cell Sclerosis; epithelioid cell component G. Diffuse large cell Cleaved cell, non-cleaved cell, sclerosis High grade H. Large cell, immunoblastic Plasmacytoid, clear cell, polymorphous epithelioid cell component I. Lymphoblastic Convoluted, non-convoluted J. Small non-cleaved cell Burkitt's, follicular areas Miscellaneous Composite, mycosis, fungoides, histiocytic, extra medullary, plasmacytoma, unclassifiable, other
use world-wide. This was unsuccessful and so, after analysis of survival data, a working formulation of nonHodgkin's lymphomas for clinical usage was proposed (Table 1.4). This recognized three grades of lymphoma, low grade, intermediate grade and high grade, and was proposed, not as a classification, but as a common terminology for reporting lymphoma clinical trials.24 Serious criticisms were levelled at the working formulation, principally by Lukes and Lennert, who regarded it
as biologically imprecise with a division of distinctive entities between different clinical grades, and a lumping together of diverse lymphomas under terms such as diffuse mixed small and large cell and diffuse large cell lymphoma. They expressed the hope that the working formulation would not stifle research into the basic biology of malignant lymphoproliferative disease.24 Although proposed solely as a translational device, the working formulation was quickly adopted in North America as a classification, and during the 1980s and early 1990s became the standard classification. In Europe the working formulation was adopted by some centers, but the Kiel classification became predominant and was the most important in this continent, particularly after it was updated in 1988.25
REVISED KIEL CLASSIFICATION The Kiel classification was extensively updated in 1988 by Stansfeld and colleagues and now introduced clear delineation of B and T cell lymphoma entities.25 The classification was intended for nodal lymphomas and paid little attention to extra nodal disease. The updated Kiel classification (Table 1.5) was criticized for its level of complexity, particularly within the peripheral T cell lymphoma compartment where ten major subtypes were recognized. This was felt to lack clinical relevance and to be poorly reproducible by pathologists.26 The emergence of these two dominant classifications led to a major divergence of terminology between Europe and the United States, which led to difficulties in interpretation of the clinical and pathological literature. In order to resolve these difficulties, a group of 19 expert hematopathologists, who called themselves the International Lymphoma Study Group (ILSG), conducted a series of meetings in an attempt to identify distinct lymphoma entities within the field of lymphoproliferative
Table 1.5 Updated Kiel classification on non-Hodgkin's lymphoma
Low grade Lymphocytic-chronic lymphocytic and prolymphocytic leukemia; hairy cell leukemia Lymphoplasmacytic/cytoid Plasmacytic Centroblastic/centrocytic Centrocytic High grade Centroblastic Immunoblastic Large cell anaplastic Burkitt lymphoma Lymphoblastic Rare types
Lymphocytic-chronic lymphocytic and prolymphocytic leukemia Lymphoepithelioid Angioimmunoblastic Tzone Pleomorphic, small cell Pleomorphic, medium and large cell Immunoblastic Large cell anaplastic Lymphoblastic Rare types
6 Lymphoma classification
disease, which could be reproducibly diagnosed and on which they could establish an acceptable terminology.27
REVISED EUROPEAN-AMERICAN LYMPHOMA CLASSIFICATION In 1994, Harris and co-workers published the Revised European-American Lymphoma (REAL) classification. This classification was a listing of lymphoid neoplasms which were distinctive biological entities and which could be reproducibly diagnosed by hematopathologists (Table 1.6). This listing delineates precursor and peripheral lymphoid neoplasms of both B and T cell lineage as well as including plasmacytoma/multiple myeloma and Hodgkin's disease. The classification is applicable to nodal and extranodal lymphoma alike, and is thus a significant advance on the Kiel classification.27'28 Although fiercely criticized by some when first published, the classification has been broadly accepted and welcomed by hematopathologists world-wide. Criticisms that were levelled at the REAL classification were that it had not been tested for its clinical and prognostic value, and there were no data as to how well it could be applied by hematopathologists.29 In order to study these questions, a group was established under the chairmanship of Professor J.O. Armitage from the University of Nebraska Medical Center at Omaha, which was termed the Lymphoma Classification Project.
LYMPHOMA CLASSIFICATION PROJECT In order to evaluate the recently proposed ILSG classification of non-Hodgkin's lymphoma, a group of five expert hematopathologists visited eight major oncology institutions. The objectives of this study were to determine how well the REAL classification could be applied by expert hematopathologists, and to compare its applicability with the working formulation and Kiel classification to study the clinical value of the classification and to gain some idea of the geographic variability in the incidence of non-Hodgkin's lymphoma. The results of this study showed that the REAL classification could be applied with a high degree of accuracy with an interobserver concordance rate of 85 per cent for the major lymphoma subtypes and an intraobserver concordance rate of 94 per cent when clinically insignificant discrepancies were discounted.30 The REAL classification was a good predictor of survival and failure-free survival. This study has validated the ILSG proposal, and shown the REAL classification to be superior to the working formulation and updated Kiel classification in terms of reproducibility and prognostic significance. The REAL classification and the data generated by the Lymphoma Classification Project have been of great
Table 1.6 Lymphoid neoplasms recognized by the International Lymphoma Study Group B cell neoplasms I Precursor B cell neoplasms: B precursor lymphoblastic leukemia/lymphoma II Peripheral B cell neoplasms 1. B cell chronic lymphocytic leukemia/prolymphocytic leukemia/small lymphocytic lymphoma 2. Lymphoplasmacytoid lymphoma/immunocytoma 3. Mantle cell lymphoma 4. Follicle center lymphoma, follicular Provisional cytologic grades: I (small cell), II (mixed small and large cell), III (large cell) Provisional subtype: diffuse, predominantly small cell type 5. Marginal zone B cell lymphoma Extranodal (MALT type *monocytoid B cells) Provisional subtype: nodal (*monocytoid B cells) 6. Provisional entity: splenic marginal zone lymphoma (*villous lymphocytes) 7. Hairy cell leukemia 8. Plasmacytoma/plasma cell myeloma 9. Diffuse large B cell lymphoma Subtype: primary mediastinal (thymic) B cell lymphoma 10. Burkitt lymphoma 11. Provisional entity: high-grade B cell lymphoma, Burkitt-like T cell and putative natural killer (NK) cell neoplasms I. Precursor T cell neoplasm: T precursor lymphoblastic lymphoma/leukemia II. Peripheral T cell and NK cell neoplasms 1. T cell chronic lymphocytic leukemia/prolymphocytic leukemia 2. Large granular lymphocytic leukemia (LGL), T and NK cell types 3. Mycosis fungoides/Sezary syndrome 4. Peripheral T cell lymphoma, unspecified Provisional cytologic categories, medium-sized cell, mixed medium and large cell, large cell, lymphoepithelioid cell Provisional subtype: hepatosplenic y6 T cell lymphoma Provisional subtype: subcutaneous panniculiticTcell lymphoma 5. Angioimmunoblastic T cell lymphoma (AIL) 6. Angiocentric lymphoma 7. Intestinal T cell lymphoma (*enteropathy associated) 8. Adult T cell lymphoma/leukemia (ATLL) 9. Anaplastic large cell lymphoma (ALCL), CD 30+, T and null-cell types 10. Provisional entity: anaplastic large cell lymphoma, Hodgkin's like Hodgkin's disease I. Lymphocyte predominance II. Nodular sclerosis III. Mixed cellularity IV. Lymphocyte depletion V. Provisional entity: lymphocyte-rich classical Hodgkin's disease
WHO classification 7
value in formulating the forthcoming proposal from the WHO on the classification of hemopoietic neoplasms. WHO CLASSIFICATION The WHO classification, although not yet published, is being developed under the joint auspices of the Society for Hematopathology and the European Association for Haematopathology. A steering committee composed of
Drs C Berard, J Diebold, N Harris, E Jaffe and K Lennert has established ten committees, which are involved in the classification of hematolymphoid malignancy. Although a final version has not been published, the broad outlines of the classification are illustrated in Table 1.7, and the close similarity between the proposed WHO classification and the REAL classification are immediately apparent. It is to be hoped that, with the publication of the WHO classification, a period of stability in lymphoma
Table 1.7 World Health Organisation classification of neoplastic diseases of the hematopoietic and lymphoid tissues B cell neoplasias Precursor B cell neoplasms B cell lymphoblastic leukemia/lymphoma Peripheral B cell neoplasms B cell chronic lymphocytic leukemia/small lymphocytic lymphoma Variant: with monoclonal gammopathy/plasmacytoid differentiation; mu heavy chain disease B cell prolymphocytic leukemia Variant: hairy cell variant Lymphoplasmacytic lymphoma Variant: Waldenstrom's macroglobulinemia; gamma heavy chain disease Mantle cell lymphoma Variant: blastic Follicular lymphoma Grades: Grade 1 (centroblasts comprise <50 per cent of the follicle surface area); Grade 2 (centroblasts comprise >50 per cent of the follicle surface area) Variant: cutaneous follicular lymphoma Marginal zone B cell lymphoma of mucosa-associated lymphoid tissue Variant: alpha heavy chain disease Nodal marginal zone lymphoma ± monocytoid B cells Splenic marginal zone B cell lymphoma (± villous lymphocytes) Hairy cell leukemia Diffuse large B-cell lymphoma Variants: Centroblastic Immunoblastic Tcell or histiocyte rich Anaplastic large B cell Burkitt-like Lymphomatoid granulomatosistype Diffuse large B cell lymphoma, subtypes: Mediastinal (thymic) large B cell lymphoma Intravascular large B cell lymphoma Primary effusion lymphoma in HIV patients/pyotorax related Burkitt lymphoma Variant: with plasmacytoid differentiation (AIDSassociated) Plasmacytoma Variants: Solitary plasmacytoma of bone Extramedullary plasmacytoma
Plasma cell myeloma Variants: Indolent myeloma Smoldering myeloma Osteosclerotic myeloma (POEMS syndrome) Plasma cell leukemia Non-secretory myeloma Systemic light chain disease Primary amyloidosis T cell neoplasias Precursor Tcell leukemia/lymphoma Tcell lymphoblastic leukemia/lymphoma Peripheral T/NKcell neoplasms, predominantly leukemic/disseminated T cell prolymphocytic (T-PLL) T cell large granular lymphocyte leukemia NKcell leukemia Adult Tcell lymphoma/leukemia Peripheral T cell and NK cell neoplasms, predominantly nodal AIL Tcell lymphoma Peripheral T cell lymphoma (unspecified) T-zone Lymphoepithelioid (Lennert) lymphoma ALC lymphoma (T and null cell types) Peripheral T cell and NK cell neoplasms, predominantly extra nodal Mycosis fungoides Sezary syndrome Primary cutaneous CD 30-positive T cell lymphoproliferative disorders Primary cutaneous ALC lymphomas + borderline with lymphomatoid papulosis Subcutaneous panniculitic-like T cell lymphoma NK/Tcell lymphomas, nasal/nasal type Enteropathy-type intestinal T cell Hepatosplenic76Tcell lymphoma Hodgkin lymphoma (Hodgkin disease) Nodular lymphocyte-predominance Hodgkin lymphoma Classical Hodgkin lymphoma Hodgkin lymphoma, nodular sclerosis (Grades I and II) Hodgkin lymphoma, mixed cellularity Classical Hodgkin lymphoma, lymphocyte-rich Hodgkin lymphoma, lymphocyte depletion
8 Lymphoma classification
terminology will pertain, so that prospective data on the clinical behavior of defined lymphoma entities may be accrued.
15. Lennert K, Stein H, Kaiserling E. Cytological and functional criteria for the classification of malignant lymphomata. BrJ Cancer 1975; 31 (suppl 2): 29-43. 16. Lennert K. Malignant lymphomas other than Hodgkin's disease. New York: Springer-Verlag, 1978.
REFERENCES
17. Lennert K. Immunology: morphology and function. Adv Exp Med Biol 1979; 114:1-9. 18. Lennert K, Collins RD, Lukes RJ. Concordance of the Kiel
1. Hodgkin T. On some morbid appearances of the absorbent glands and spleen. Med Chir Trans 1832; 17:68-114. 2. Wilkes Sir S. Cases of enlargement of the lymphatic glands and spleen (or, Hodgkin's disease), with remarks. Guys Hosp Rep 1865; 11: 56-67. 3. Greenfield WS. Specimens illustrative of the pathology of lymphadenoma and leucocythaemia. Trans Path Soc Land. 1878; 29: 272-304. 4. EwingJ. Neoplastic diseases. Philadelphia, London: WB Saunders, 1919. 5. Fox H. Remarks on microscopic preparations made from some of the original tissue described by Thomas Hodgkin, 1832. Ann Med Hist 1926; 8: 370-4. 6. Rappaport H. Tumors of the hematopoietic system, series 1, section III. Washington, DC: Armed Forces Institute of Pathology. 1966. 7. Lukes RJ, Collins RD. Immunologic characterization of human malignant lymphomas. Cancer 1974; 34 (suppl): 1488-503. 8. Lukes RJ, Collins RD. New approaches to the classification of the lymphomata. BrJ Cancer 1975; 31 (suppl 2): 1-28. 9. Lukes RJ, Collins RD. Lukes-Collins classification and its significance. Cancer Treat Rep 1977; 61: 971-9. 10. Lukes RJ, Lincoln TL, Parker JW, Alavaikko MJ. An
and Lukes-Collins classifications of non-Hodgkin's lymphomas. Histopathology 1983; 7: 549-59. 19. Dorfman RF. Classification of non-Hodgkin's lymphomas (letter). Lancet 1974; 2: 961-2. 20. Mathe G and Rappaport H. Histological and cytological typing of neoplastic diseases of hematopoietic and lymphoid tissues. Geneva: World Health Organisation, 1976. 21. Bennett MH, Farrer-Brown G, Henry K, Jell iff e AM. Classification of non-Hodgkin's lymphoma. Lancet 1974; ii: 405. 22. Henry K, Bennett MH, Farrer-Brown G. Morphological classification of non-Hodgkin's lymphomas. Rec Results Cancer Res 1978; 64: 38-56.
23. Nathwani BN, Kim H, Rappaport H, Solomon J, Fox M. Non-Hodgkin's lymphomas: a clinicopathologic study comparing two classifications. Cancer 1978; 41: 303-25. 24. Anonymous. National Cancer Institute sponsored study of classifications of non-Hodgkin's lymphomas: summary and description of a working formulation for clinical usage. The Non-Hodgkin's Lymphoma Pathologic Classification Project. Cancer 1982; 49: 2112-35.
immunologic approach to classification of malignant
25. Stansfeld AG, Diebold J, Noe H, et al. Updated Kiel classification for lymphomas. Lancet 1988; 1: 292-3.
lymphomas: a cytokinetic model of lymphoid neoplasia. In: Clarkson B, et al., eds. Differentiation of normal and
26. Hastrup N, Hamilton-Dutoit S, Ralfkiaer E, Pallesen G. Peripheral T-cell lymphomas: an evaluation of
neoplastic hematopoietic cells. Cold Spring Harbor, NY:
reproducibility of the updated Kiel classification.
Cold Spring Harbor Laboratory, 1978; 935-52.
Histopathology 1991; 18: 99-105.
11. Lukes RJ, Taylor CR, Parker JW. Multi para meter studies in malignant lymphoma based on studies in 1186 cases. Prog Clin Biol Res 1983; 132E: 203-13. 12. Lukes RJ, Taylor CR, Parker JW, Lincoln TL, Pattengale PK, Tindle BH. A morphologic and immunologic surface marker study of 299 cases of non-Hodgkin lymphomas and related leukemias. Am} Pathol 1978; 90: 461-85. 13. Lukes RJ, Parker JW, Taylor CR, Tindle BH, Cramer AD, Lincoln TL. Immunologic approach to non-Hodgkin lymphomas and related leukemias. Analysis of the results of multiparameter studies of 425 cases. Semin Hematol 1978; 15: 322-51. 14. Gerard-Marchant R, Hamlin I, Lennert K, et al. Classification of non-Hodgkin's lymphoma. Lancet 1974; H: 406-8.
27. Harris NL, Jaffe ES, Stein H, et al. A revised European-American classification of lymphoid neoplasms: a proposal from the International Lymphoma Study Group. Blood 1994; 84:1361-92. 28. Chan JK, Banks PM, Cleary ML, et al. A proposal for classification of lymphoid neoplasms (by the International Lymphoma Study Group). Histopathology 1994; 25: 517-36. 29. Rosenberg SA. Classification of lymphoid neoplasms. Blood 1994; 84:1359-60. 30. Anonymous. A clinical evaluation of the International Lymphoma Study Group classification of non-Hodgkin's lymphoma. The Non-Hodgkin's Lymphoma Classification Project. Blood 1997; 89: 3909-18.
2 Hodgkin's disease KA MACLENNAN, B VAUGHAN HUDSON AND G VAUGHAN HUDSON
9
Introduction Cell of origin of Hodgkin's disease
10
Mixed cellularity Hodgkin's disease 13 Clinical significance of morphological pattern in Hodgkin's
Lymphocyte-predominant Hodgkin's disease Nodular sclerosis
10 12
disease
14
References
14
Lymphocyte-depleted Hodgkin's disease
13
INTRODUCTION The first descriptions of the disease currently termed Hodgkin's disease are attributed to Thomas Hodgkin1 and the term Hodgkin's disease was generously applied by Sir Samuel Wilkes;2 however, there are several texts describing a similar disease process that antedate both these workers' manuscripts.3'4 These early descriptions of Hodgkin's disease were all concerned with the macroscopic appearances and distribution of affected lymph nodes, spleen and visceral organs as observed at post mortem examination. There thus exists some doubt as to the exact nature of the disease processes being described. Herbert Fox,5 after histological examination of pathological material stored at Guy's Hospital from three of Thomas Hodgkin's original cases, believed two were typical of Hodgkin's disease, and one to be an example of lymphosarcoma or leukaemia; other skilled observers have concurred with this view.6,7 During the latter half of the nineteenth century, many workers undertook histological examination of cases that were felt to be similar to Hodgkin's disease.8'11 There emerged from these descriptions an awareness that the normal structure of the lymph node was replaced by abnormal tissue, often described as fibrous tissue, and the disease was associated with unusual giant cells. In 1898, Sternberg gave a masterful description of these giant cells,12 which 4 years later was followed by the publication of Dorothy Reed's classic paper.13 Following the meticulous histological descriptions of Hodgkin's disease by these early microscopists, terminological confusion was soon to emerge and, by 1933, Walhauser was able to find 52 synonyms for this condition
(an unusually large number even for the field of lymphoreticular pathology).14 This ridiculous state of affairs was only resolved by the widespread adoption of the Jackson and Parker classification" (see below). In the early part of the twentieth century, some workers attempted to correlate the histological appearances of Hodgkin's disease with the clinical course. In 1919, Ewing recognized a rapidly fatal form of the disease characterized by depletion of lymphocytes and a sheetlike growth of pleomorphic mononuclear and multinuclear cells, which he termed Hodgkin's sarcoma.16 Rosenthal, in 1936, observed the inverse relationship between the number of lymphocytes and abnormal reticulum cells;17 he was also able to correlate survival and response to orthovoltage radiotherapy to the histological appearances. Following the work of Rosenthal, Jackson and Parker15'18"20 proposed their classification of Hodgkin's disease. Three histological subtypes were recognized: paragranuloma, granuloma and sarcoma, which showed a good correlation with clinical behavior and prognosis. Unfortunately, the classification proved to be of limited value as the majority of cases were classified as Hodgkin's granuloma15'21 and this subtype showed an extremely variable clinical course.22 These problems were overcome by the widespread adoption of the Lukes and Butler classification and its modification proposed at the Rye conference,23-26 which has remained essentially unchanged for over 20 years, and most pathologists believe they are familiar with the terminology and criteria employed. It is therefore surprising to find high levels of disagreement between pathologists in establishing the diagnosis of Hodgkin's disease and its classification, which may range
10 Hodgkin's disease
from 13 per cent27'28 to a staggering figure of 47 per cent, reported by Symmers.29 The reasons for the difficulties pathologists experience in the diagnosis of HD are not too difficult to understand. Hodgkin's disease is rare and most pathologists will see only a few cases a year. This, combined with the complexity of the histological picture, which may be closely mimicked by a variety of neoplastic and nonneoplastic lymphoproliferative conditions, will lead to errors in diagnosis.
CELL OF ORIGIN OF HODGKIN'S DISEASE There has been considerable controversy over the years about the cellular origin of the putative malignant cell in Hodgkin's disease (HD), the Hodgkin's and ReedSternberg cell (H-RS cell). Many candidates have been proposed that include histiocytes, interdigitating cells, follicular dendritic cells and lymphoid cells of both B and T cell lineage. One significant problem in establishing the lineage of the H-RS cell has been their relative paucity in tissue affected by HD, where they often make up less than 5 per cent of the total cell population. This has led investigators to study Hodgkin's cell lines and cases of HD containing numerous H-RS cells, which may show major differences to typical HD.3031 An early study of a case of H-RS cell-rich nodular sclerosis revealed a clonal immunoglobulin gene rearrangement.32 Although initially controversial, subsequent reports also found immunoglobulin (Ig) gene rearrangements in a percentage of cases. Other workers demonstrated T cell receptor rearrangements of both beta and gamma chains in some cases of HD.33 A novel approach of single cell microdissection of HRS cells was established, which allowed analysis of the Ig gene configuration by polymerase chain reaction (PCR). Initial reports showed clonal Ig gene rearrangement with somatic hypermutation within the H-RS cell population in the vast majority of the cases studied.34 Controversy soon followed with other workers providing dramatically different results despite using very similar methodology. In summary, one group were unable to detect any evidence of Ig gene rearrangement within H-RS cells,35 one group showed polyclonal Ig rearrangement36 and a third demonstrated a mixture of results, with some cases showing no Ig gene rearrangement,37 some polyclonal rearrangement,36 some clonal,38-10 and some mixed polyconal and clonal Ig gene rearrangement.41 These results are clearly incompatible. Evidence is now accumulating that the majority of cases of classical HD have clonal Ig gene rearrangement, with somatic hypermutation clearly identifying the H-RS cells as a neoplastic, germinalcenter-derived B cell. In addition, many cases display 'crippling mutations' such as stop codons within the rearranged Ig genes.42
Similarly controversial results have been found in lymphocyte-predominant nodular HD, with some workers demonstrating somatically hypermutated, clonally rearranged Ig genes, while others have found polyclonal patterns of Ig gene rearrangement within populations of microdissected lymphocytic and/or histiocytic (L & H) cells. Recently, three separate groups have demonstrated clonal Ig gene rearrangements within populations of microdissected L & H cells. These rearrangements are somatically hypermutated but lack the 'crippling mutations' seen in classical HD, and show evidence of continuing antigen selection in an analogous manner to follicle-center-derived non-Hodgkin's lymphomas.43'44
LYMPHOCYTE-PREDOMINANT HODGKIN'S DISEASE For many years, Hodgkin's disease with a predominance of lymphocytes has been recognized as having a more indolent natural history15,17,22,23,25,45-47 than the usual type of HD. Lukes and Butler described a form of HD that contained a spectrum of cytological appearances, which ranged from a predominance of mature lymphocytes to a histiocyte-rich cellular background which they termed lymphocytic and/or histiocytic (L & H) Hodgkin's disease;24 they recognized nodular and diffuse architectural patterns. These cytological and architectural patterns of L & H HD were amalgamated at the Rye conference26 and termed lymphocytic predominance. Lymphocyte-predominant (LP) Hodgkin's disease makes up a variable percentage of cases of HD in large series, depending on the stringency of the diagnostic criteria applied.48 In the British National Lymphoma Investigation (BNLI) series of 4249 cases, LP comprises 5.7 per cent and in the European Organisation for Research and Treatment of Cancer (EORTC)-GELA studies of localized HD (H 8), which include 722 centrally reviewed cases, LP makes up 4 per cent.49 It usually presents with localized, asymptomatic disease and often involves unusual sites, such as the suprahyoid neck, the periparotid lymph nodes and the inguinal region. There is a marked male predominance and patients are usually a decade older than the peak age incidence of the usual type of HD.50 For many years there was a lack of precision in the diagnosis of LP, with many cases of the usual type of HD that displayed a lymphocyte-rich cellular background being included in this category. In a seminal paper published in 1979, Poppema and co-workers51 recognized the cytological similarities between LP nodular HD and a reactive condition affecting germinal centers, termed progressive transformation.52'53 They postulated that LP nodular HD was a distinct form of HD, which arose in
Lymphocyte-predominant Hodgkin's disease 11
the B cell regions of the lymph node and was related to progressive transformation of germinal centers. In order to emphasize the differences between LP nodular and other histological subtypes of HD, they proposed the term nodular paragranuloma, which has been adopted by some workers.
Morphologic features Lymph nodes affected by LP nodular HD are enlarged and can reach significant sizes (up to 5 cm); their cut surface has a uniform fleshy appearance and occasionally residual remnants of lymph node may be observed, which are compressed at the periphery of an expansile tumor mass.54 Microscopically LP nodular HD is characterized by the presence of a macronodular growth pattern, which is expansile rather than infiltrative (Plate 1); nodules do not penetrate the lymph-node capsule or extend into perinodal tissue. Diffuse areas are sometimes seen. Exclusively diffuse LP HD is exceptionally rare in our experience, with the majority of cases being nonHodgkin's lymphomas of peripheral T cell or T cell-rich B cell type.55 The nodules of LP nodular HD often have a 'motheaten' appearance at low power (Plate 2), and are composed of small round or slightly irregular lymphoid cells with admixed large lymphoid cells, epithelioid histiocytes, dendritic reticulum cells and a Reed-Sternberg (RS) cell variant called the L & H or popcorn cell. The lymphocytes within the nodules show a close cytological similarity with mantle zone lymphocytes, which is confirmed by their phenotype. Epithelioid histiocytes may be scattered within the nodules or form loose aggregates; well-formed, sarcoid-like granulomata, if present, are usually seen at the periphery of the nodules and may form encircling rings (Plate 3). The histiocytes bear a close resemblance to those seen in mantle cell lymphoma, and possess an open nuclear chromatin with a single prominent nucleolus and well-defined eosinophilic cytoplasm. The nuclei of follicular dendritic cells are easily identified and multinucleated forms, resembling Warthin-Finkeldy giant cells are common (Plate 4). L & H cells have a characteristic morphology with a large, irregular and often lobulated nucleus with a prominent nucleolus which is often amphophillic and irregular (Plate 5). Classical RS cells are rare and are not essential for the diagnosis of LP nodular HD; in fact, if classical RS cells can be found with ease, the diagnosis of LP nodular HD should be changed to one of mixed cellularity as the clinical behavior of these cases is different from typical LP HD.56 The number of L & H cells is very variable and ranges from scanty to very numerous, making up more than 10 per cent of the cellular composition of the nodules; the latter is often seen in relapses of LP nodular. The number of L & H cells present does not
seem to influence the clinical behavior.50 L & H cells may be confined to the nodules or may spill out into the internodular region of the node. Immunocytochemistry The nodules of LP nodular HD are composed of polytypic small B cells expressing CD 20 and CD 79a57^ (Plate 6) and showing co-expression of IgM and IgD in a similar fashion to mantle zone B cells.61 Within the nodules is a meshwork of follicular dendritic cells (FDCs) revealed by staining for CD 21 and CD 35, and their processes often wrap around the L & H cells (Plate 7). The L & H cells uniformly express a B cell phenotype with strong expression of CD 20 and CD79a62'63 (Plate 8). There is evidence of immunoglobulin synthetic capacity as shown by the presence of J chain within the L & H cells,64 and some workers have shown the presence of kappa-light-chain restriction either by immunocytochemistry65 or by in situ hybridization for light-chain messenger RNA.66,67 The markers of classical H-RS cells, CD 30 and CD 15 are not usually detected on L & H cells68 (though there is some evidence for expression of a heavily sialylated form of CD 15, which is undetectable without prior neuraminidase digestion69). The presence of these markers should prompt consideration of a diagnosis of follicular colonization by classical HD (see below). There is frequent expression of epithelial membrane antigen by L & H cells70 and the presence of EpsteinBarr virus is not usually detectable.48,63,71 Within the nodules are numerous T cells, which express CD 3. Numerous CD 57-positive T cells are also seen and these may form rosettes around the L & H cells (Plate 9). The number of CD 57 cells has proved useful in the differential diagnosis of LP HD and lymphocyterich classical HD (LRCHD). Cases of LP have been shown to have >200 CD 57-positive cells per high-power field compared to an average of 45 CD 57-positive cells per high-power field in LRCHD.48 Non-Hodgkin's lymphoma arising in patients with LP nodular It is now clear from several large studies of patients with LP nodular HD that there is a markedly increased risk of non-Hodgkin's lymphoma (NHL), which ranges from an incidence of 3.8 per cent72 to nearly 10 per cent.73 The lymphomas associated with LP nodular HD may occur simultaneously74-77 or after a period of many years.72,75,78,79 They are usually of B cell lineage72,74,75 and there is some evidence that there may be a clonal relationship between the original LP and the subsequent B cell NHL;79,80 other workers have been unable to confirm this.78 The morphology of these secondary, high-grade B cell lymphomas is variable: some show features typical of diffuse
12 Hodgkin's disease
large B cell lymphoma exhibiting centroblastic or immunoblastic cytology, while others resemble sheets of L & H cells. Since the first recognition of T cell lineage NHL following LP nodular HD,72 subsequent reports have confirmed this association,81'82 and Weisenburger and co-workers have reported the concurrent presentation of T NHL and LP nodular HD.83 These may have a variety of histological patterns but the majority appear to fall within the peripheral T cell lymphoma, unspecified group of the Revised European-American Lymphoma (REAL) classification.84 A single case of composite T and B lineage lymphoma has been described in the setting of LP HD.85
NODULAR SCLEROSIS The presence of fibrosis and the proliferation of fibroblastic cells in HD has been recognized for over a century.10-12,17 The recognition by Smetana and Cohen21 of a sclerosing variant of Hodgkin's granuloma15 and its associated superior survival were among the first steps in the delineation of nodular sclerosis (NS). Lukes and co-workers23-25,86 described the histological features of NS, and stressed the importance of nodularity, lacunar cells and birefringent collagen band formation. Rappaport and colleagues emphasized the unique nature of NS by demonstrating the consistency of this histological pattern in sequential biopsies and from different anatomical locations.87,88
Morphological features Lymph node involvement by nodular sclerosis may be partial or complete. There is usually capsular and intranodal fibrosis, which may impart a firm rubbery texture. The cut surface may have a coarsely nodular appearance and areas of necrosis may be macroscopically apparent. Histologically, capsular thickening is present in the majority of cases (Plate 10) and there is a variable degree of intranodal sclerosis, which may range from occasional thin collagen bands to large areas of collagenous sclerosis that obliterate most of the nodal structure. Nodularity is a constant feature of NS, and may be present partially or throughout the lymph node. NS is associated with a particular H-RS cell variant termed the lacunar cell. The lacunar cell is most obvious in specimens fixed in formalin in whom paraffin processing dissolves the lipid-rich cytoplasm to leave a clear space; specimens fixed in mercuric-based fixatives do not show this helpful artefact. The nucleus of lacunar cells is typically twisted or lobulated with a prominent eosinophillic nucleolus (Plate 11). In recent years there has been considerable confusion over the precise criteria required to diagnose NS and this
has centred around the entity termed cellular phase NS. Lukes86 required the presence of intranodal collagen band formation in association with lacunar cells to establish a diagnosis of NS and recognized a cellular phase in which only a single band of collagen was found in association with the typical • cellular background of NS. Cases lacking collagen band formation were classified as mixed cellularity by Lukes. Other workers have classified cases as cellular-phase NS when lacunar cells are seen in the absence of collagen band formation.87,89 The advantage of adhering to the strict criteria proposed by Lukes and Butler24 is that they do enable pathologists to achieve very high levels of interobserver and intraobserver concordance (97 per cent)90 in the diagnosis of the NS subtype. The cellular nodules of NS show a wide range of cytological appearances ranging from a lymphocyte-rich cellular background with scanty lacunar cells to one of lymphocyte depletion and sheets of lacunar and H-RS cells. This latter pattern may be associated with areas of necrosis. In many cases there are also admixed histiocytes, eosinophils and plasma cells with the lymphocytes and lacunar cells. The cytological diversity of the cellular nodules of NS has prompted workers to develop grading systems for NS, which might correlate with prognosis (reviewed in MacLennan et a/.91). In a series of publications the BNLI proposed a grading system which recognized low-grade (Grade I) and high-grade (Grade II) subtypes of NS. 50,56,90-93 The histologic criteria for this grading system have been published in detail elsewhere90 and are only outlined here. Cases were classified as Grade II NS if more than 25 per cent of the cellular nodules showed lymphocytedepleted cytology. These lymphocyte-depleted nodules are often composed of sheets of mononuclear Hodgkin's and lacunar cells; an appearance that has been termed 'syncytial Hodgkin's disease' by some workers.94'95 Central necrosis and eosinophilic abcess formation within these lymphocyte-depleted nodules is sometimes observed (Plate 12). Also classified as Grade II NS were cases in which more than 25 per cent of the cellular nodules contained numerous pleomorphic H-RS cells in the absence of lymphocyte depletion. The rarest form of lymphocyte-depleted cytology was the bland-appearing fibrohistiocytic variety; if more than 80 per cent of the cellular nodules showed this feature, the case was classified as Grade II NS.93 The adverse prognostic significance of fibroblastic proliferation was also reported by Colby et a/.96 All other cases were graded as Grade I including borderline cases. Using this system significant differences in survival and disease-free survival are seen between the grades of NS (see below). Other workers have confirmed the clinical value of this grading system;97'101 some have not been able to demonstrate a difference in prognosis between the two grades of NS.102'103
Mixed cellularity Hodgkin's disease 13
Immunocytochemistry The phenotype of NS differs from LP nodular HD in that the nodules are composed predominantly of T cells104'105 (CD 3-positive, CD 45 Ro-positive) with a prevalence of CD 4-positive cells centrally and a rim of CD 8-positive lymphocytes at the periphery. The lacunar cells exhibit strong staining for CD 15 in over 80 per cent of cases and this staining is usually membrane and golgi associated; CD 30 is also expressed in the majority of cases106-110 (Plate 13). The expression of lymphoid lineage-restricted antigens on H-RS cells remains controversial with some groups claiming expression of CD 3 in a percentage of cases,111,112 whilst others find expression of B lineage antigens on a small percentage of H-RS cells,113 which may be seen in up to 60 per cent of cases of HD.114 There is variable expression of BCL 6 protein and CD 138.115,116 The significance of these phenotypic differences is unclear but in human immunodeficiency virus (HIV)-associated HD there is a marked predominance of CD 138-positive/BCL 6-negative H-RS cells.117 Some cases of anaplastic large cell lymphoma (ALCL) may display morphological features, which are reminiscent of NSHD particularly the Grade II subtype.118 So close may these similarities be that some workers have introduced the term 'ALCL Hodgkin's like';84 other workers feel that the vast majority of these cases are in fact related to classical HD and have used the term 'malignant lymphoma with features of Hodgkin's lymphoma and ALCL'.119 Immunocytochemistry can be helpful in distinguishing between HD and ALCL; whilst CD 30 is usually expressed by both, CD 15 staining is uncommon in ALCL and, when present, does not exhibit the membrane and Golgi staining characteristic of HD. Leukocyte common antigen (CD 45) is expressed in a percentage of ALCL118 but in our hands has proved of limited value. Recently antibodies to the p80 NPM-ALK fusion protein, generated by the 2;5 translocation,120 have become available,121,122 which stain just over half the cases of ALCL studied; no case of HD was labelled.122 Of interest is the detection of t(2;5) in a minor population of microdissected CD 30-positive cells from cases of classical HD123 and in the peripheral blood of normal individuals.124
LYMPHOCYTE-DEPLETED HODGKIN'S DISEASE Lymphocyte-depleted (LD) Hodgkin's disease is the rarest form of Hodgkin's disease and its frequency appears to be diminishing. It includes two distinctive morphological entities from the Lukes and Butler classification: diffuse fibrosis and reticular HD. It is now clear from various studies that many of the cases formally classified as LD were in fact examples of non-Hodgkin's
lymphomas125 often of anaplastic large cell type or of other HD subtypes, such as the Grade II form of NS.56 In a review of cases from the BNLI, many of the cases that were originally diagnosed as LD HD were reclassified as NHL and the incidence of true LD HD was below 2 per cent in this series. Patients with LD HD tended to be elderly and often presented with advanced symptomatic disease. Bone marrow disease occurred in over 60 per cent of cases.50 There was a low attainment of complete remission with combination chemotherapy and survival was poor.
Morphological features Lymphocyte-depleted Hodgkin's disease has a high frequency of extranodal involvement. In particular, the bone marrow is affected in many cases and may be the site of initial diagnostic biopsy (Plate 14). When lymph nodes are affected, the architecture is completely effaced. The diffuse fibrosis variant is characterized by a hypocellular lymph node often showing areas of geographic necrosis (Plate 15). In the background there is a pink fibrillary appearances of non birefringement fibrosis. Lymphocytes are relatively scanty and bizarre mononuclear and multinuclear Hodgkin's cells are seen. Classical Reed-Sternberg cells are often difficult to find. The reticular subtype of LD HD is characterized by a numerical predominance of H-RS cells. It has been our experience that the majority of cases initially diagnosed as reticular HD represents examples of non-Hodgkin's lymphomas.
MIXED CELLULARITY HODGKIN'S DISEASE In the Lukes and Butler classification, mixed cellularity Hodgkin's disease was used to classify cases of Hodgkin's disease that did not conform to the pathological criteria of LP, NS and LD HD. It thus contained a spectrum of cytological appearances ranging from lymphocyte-rich forms, which contained classical RS cells, to subtypes which showed foci of lymphocyte-depleted cytology not involving the whole lymph node. Many cases of mixed cellularity Hodgkin's disease have similarities to nodular sclerosis, such as focal nodularity and the presence of lacunar cells, but lack sufficient criteria to be diagnosed as NS. Other cases showed distinctive morphological patterns, which often involved alterations in the structure of the germinal center and the marginal zone B cell region.
Morphological features Classical mixed cellularity Hodgkin's disease is characterized by a diffuse architecture, which effaces the nodal
14 Hodgkin's disease
architecture completely. Typically the cytological background contains lymphocytes, macrophages, plasma cells and eosinophils, as well as mononuclear and classical Reed-Sternberg cells that are easy to find (Plate 16). There may be small foci of necrosis but this is much less common than in either NS or LD HD. Some cases may show the presence of lacunar cells or even areas of indistinct nodularity - features that suggest a close association with NS. In the absence of the three essential criteria for the diagnosis of NS (nodularity, intranodal collagen band formation and lacunar cells), these are best classified as mixed cellularity. Some workers prefer to classify these cases with NS features as HD unclassified between mixed cellularity and nodular sclerosis54 and some even put them into the cellular phase of NS. Several striking morphological patterns have been observed in mixed cellularity HD. One such case is inter follicular HD, highlighted by the Stanford Group, which is characterized by florid reactive follicular hyperplasia and an easily overlooked interfollicular infiltrate containing typical mononuclear Hodgkin's cells and RS cells126 (Plate 17). A variant of this form of HD is characterized by a marginal zone hyperplasia where the H-RS cells are seen to sit within a sea of marginal zone B cells. This has been termed HD occurring in monocytoid B cell clusters127 (Plate 18). In some cases of mixed cellularity HD the germinal centers are replaced by large expansile masses of mantle zone lymphocytes within which H-RS cells are readily found. The mantle cell nodules contain an expanded meshwork of follicular dendritic cells and, in many cases, the Hodgkin's cells express B cell antigens in addition to CD 15 and CD 30. Some workers have termed this follicular Hodgkin's disease.128 The term lymphocyte-rich classical Hodgkin lymphoma has been applied to morphological variants of HD characterized by an abundance of small lymphocytes with relatively scanty classical H-RS cells and very few eosinophils and plasma cells. In the REAL classification, LRCHD is a provisional entity84 and it has been formally adopted in the forthcoming World Health Organisation (WHO) classification.119 This histological subtype may be nodular or diffuse. The nodular subtype corresponds closely to the entity of follicular HD described by Isaacson and co-workers128 (Plate 19). The diffuse subtype is characterized by a predominance of small T lymphocytes and shows no evidence of involvement of germinal centers. In some cases of HD there are marked regressive changes within germinal centers that come to resemble the dendritic cell-only germinal centers that are seen in the hyaline vascular variant of Castleman's disease (Plate 20). These are surrounded by H-RS cells, which seem to localize preferentially at the junction of the marginal and mantle zones. The reasons for these different patterns of germinal-center reaction in mixed cellularity HD are
unknown but one might postulate they are related to the pattern of cytokine expression by the H-RS cells. Rare cases with the morphological features of the plasma cell variant of Castleman's disease in association with HD have been described.129
CLINICAL SIGNIFICANCE OF MORPHOLOGICAL PATTERN IN HODGKIN'S DISEASE Many believe that histopathology has little part to play in the prognostic assessment of patients with HD, and that the role of the pathologist is limited to accurate establishment of the diagnosis of HD and documentation of involvement of extranodal sites.130-132 In a series of publications over the past 10 years, the BNLI has documented the value of accurate histopathological classification in HD and shown that there are differences in the clinical presentation, response to therapy, freedom from relapse and overall survival between the different histological subtypes of HD. In addition it has demonstrated that there is clinical value in the subdivision of NS into two prognostic grades.50,56,90-93,133 It can be seen from the cause-specific survival curves from over 4000 patients that there are distinct differences in the rate of death from Hodgkin's disease in the different histological subtypes (Fig. 2.1).
Figure 2.1 Cause-specific actuarial survival curve for 4578 patients with Hodgkin's disease subdivided according to histological type, 1970-97. (Data from the British National Lymphoma Investigation.)
REFERENCES 1. Hodgkin T. On some morbid appearances of the absorbent glands and spleen. Med Chir Trans 1832; 17: 68-114. 2. Wilkes Sir S. Cases of enlargement of the lymphatic glands and spleen, (or, Hodgkin's disease), with remarks. Guys Hosp Rep 1865; 11: 56-67.
References 15 3. Malpighi. De viscerum structura. Bonn: Omnia Opera, 1666.
26. Lukes RJ, Craver LF, Hall TC, Rappaport H, Rubin P.
4. Craigie D. Elements of general and pathological anatomy. Edinburgh: Adam Black, 1828.
1966; 26:1311. 27. Chelloul N, Burke J, Motteram R, LeCapon J, Rappaport
5. Fox H. Remarks on microscopic preparations made from some of the original tissue described by Thomas Hodgkin, 1832. Ann Med History 1926; 8: 37(M. 6. Symmers WStC. The lymphoreticular system. In: Symmers WStC, ed. Systemic pathology. Edinburgh: Churchill Livingstone, 1978: 784-5. 7. Lennert K. Borderlands of pathological entities. In: Magrath IT, ed. The non-Hodgkin's lymphomas, 2nd edn. London: Arnold, 1997:133-67.
Report of the nomenclature committee. Cancer Res
H. HL-A antigens and Hodgkin's disease. Report on the histological analysis. In: Dausset J, Colombani J, eds. Histocompotability testing, Copenhagen: Munksgaard, 1972:769-71. 28. Miller TP, Byrne GE, Jones SE. Mistaken clinical and pathologic diagnoses of Hodgkin's disease. A Southwest Oncology Group study. Cancer Treat Rep 1982; 66: 645-51. 29. Symmers WStC. Survey of the eventual diagnosis in 600
8. Virchow R. Die Krankhaften Geschwuelste, Vol 2. Berlin: Hircwald, 1864.
cases referred for a second histologic opinion after an
9. Murchison C. Case of lymphadenoma of the lymphatic
Pathol 1968; 21: 650-3.
initial biopsy diagnosis of Hodgkin's disease. J Clin
system, spleen, liver, lungs, heart, diaphragm, dura mater etc. Trans Path Soc London 1870; 21: 372-89. 10. Langhans T. Das maligne lymphosarkom (pseudoleukaemia). Virchow Arch 1872; 54: 509-37.
30. Drexler HG and Minowada J. Hodgkin's disease derived cell lines: a review. Hum cell 1992; 5: 42-53. 31. Drexler HG. Recent results on the biology of Hodgkin and Reed-Stern berg cells. I. Biopsy material. Leuk
11. Greenfield WS. Specimens illustrative of the pathology
lymphoma 1992; 8: 283-313. 32. Linch DC, Jones HM, Berliner N, et al. Hodgkin-cell
of lymphadenoma and leucocythaemia. Trans Path Soc London 1878; 29: 272-304. 12. Sternberg C. Uber eine eigenartige unter dem Bilde der Pseudoleukamie verlaufende Tuberculose des lymphatischen Apparates.Z. Heilk. 1898; 18: 21-90. 13. Reed DM. On the pathological changes in Hodgkin's disease, with especial reference to its relation in tuberculosis. Johns Hopkins Hosp Rep 1902; 10:133-96. 14. Walhauser A. Hodgkin's disease. Arch Pathol 1933; 16: 522-62, 672-712. 15. Jackson H Jr, Parker F Jr. Hodgkin's disease and allied disorders. Oxford: Oxford University Press, 1947.
leukaemia of B-cell origin. Lancet 1985; 1: 78-80. 33. Griesser H, Feller AC, Mak TW, Lennert K. Clonal rearrangements of T-cell receptor and immunoglobulin genes and immunophenotypic antigen expression in different subclasses of Hodgkin's disease. IntJ Cancer 1987; 40:157-60. 34. Kuppers R, Hansmann ML, Diehl V, Rajewsky K. Molecular single-cell analysis of Hodgkin and Reed-Stern berg cells. Mol Med Today 1995; 1: 26-30. 35. Roth J, Daus H, Trumper L, et al. Detection of immunoglobulin heavy-chain gene rearrangement at
16. EwingJ. Neoplastic diseases. Philadelphia, London: WB Saunders, 1919.
the single-cell level in malignant lymphomas: no
17. Rosenthal SR. Significance of tissue lymphocytes in the
cells. IntJ Cancer 1994; 57: 799-804.
rearrangement is found in Hodgkin and Reed-Sternberg
prognosis of lymphogranulomatosis. Arch Pathol 1936; 21:628-46. 18. Jackson H Jr, Parker F Jr. Hodgkin's disease. I. General considerations. N EnglJ Med 1944; 230:1-8.
36. Ohshima K, Suzumiya J, Mukai Y, et al. Classical Hodgkin and Reed-Sternberg cells demonstrate a nonclonal immature B lymphoid lineage: evidence from a single cell assay and in situ hybridization. Hematol
19. Jackson H Jr, Parker F Jr. Hodgkin's disease. II.
Oncol 1996 14:123-36. 37. Trumper LH, Brady G, Bagg A, et al. Single-cell analysis
Pathology. N EnglJ Med 1944; 231: 35-44. 20. Jackson H Jr, Parker F Jr. Hodgkin's disease. III. Symptoms and course. N EnglJ Med 1994; 231: 636-46. 21. Smetana HF, Cohen BM. Mortality in relation to histologic type in Hodgkin's disease. Blood 1956; 11: 211^4. 22. Jelliffe AM, Thompson AD. The prognosis in Hodgkin's disease. BrJ Cancer 1955; 9: 21-36. 23. Lukes RJ. Relationship of histological features to clinical stages in Hodgkin's disease. AmJRoengenol 1963; 90: 944-55. 24. Lukes RJ, Butler JJ. The pathology and nomenclature of Hodgkin's disease. Cancer Res 1966; 26:1063-81. 25. Lukes RJ, Butler JJ, Hicks EB. Natural history of Hodgkin's disease as related to its pathologic picture. Cancer 1966; 34: 317-44.
of Hodgkin and Reed-Sternberg cells: molecular heterogeneity of gene expression and p53 mutations. Blood 1993; 81: 3097-115. 38. Hansmann ML, Kuppers R. Pathology and 'molecular histology' of Hodgkin's disease and the border to nonHodgkin's lymphomas. Baillieres Clin Haematol 1996; 9: 459-77. 39. Kuppers R, Rajewsky K. The origin of Hodgkin and Reed/Stern berg cells in Hodgkin's disease. Annu Rev /tfW7HA70/1998;16:471-93. 40. Vockerodt M, Soares M, Kanzler H, etal. Detection of clonal Hodgkin and Reed-Sternberg cells with identical somatically mutated and rearranged VH genes in different biopsies in relapsed Hodgkin's disease. Blood 1998;92:2899-907.
16 Hodgkin's disease 41. Hummel M, Marafioti T, Ziemann K, Stein H. Ig rearrangements in isolated Reed-Stern berg eel Is: conclusions from four different studies. Ann Oncol 1996;7(suppl4):31-3. 42. Kanzler H, Kuppers R, Hansmann ML, Rajewski K. Hodgkin and Reed-Sternberg cells represent the outgrowth of a dominant tumour clone derived from (crippled) germinal centre B cells. J Exp Med 1996; 184: 1495-505. 43. Ohno T, Stribley JA, Wu G, Hinrichs SH, Weisenburger DD, Chan WC. Clonality in nodular lymphocyte predominant Hodgkin's disease. N EnglJ Med 1997; 337: 459-65. 44. Marafiota T, Hummel M, Anagnostopoulos I, el al. Origin of lymphocyte predominant nodular Hodgkin's disease from a clonal expansion of highly mutated germinal-center B cells. N EnglJ Med 1997; 337: 453-8. 45. Harrison CV. Benign Hodgkin's disease (Hodgkin's paragranuloma).) Path Bact 1952; 64: 513-18. 46. Lumb G, Newton KA. Prognosis in tumours of lymphoid tissue. Cancer 1957; 10: 976-93. 47. Lennert K, Mohri N. Histologische Klassifizierung und Vorkommen des M. Hodgkin. Internist 1974; 15: 57-65. 48. von Wasielewski R, Werner M, Fischer R, et al. Lymphocyte-predominant Hodgkin's disease: an immunohistochemical analysis of 208 reviewed Hodgkin's disease cases from the German Hodgkin Study Group. Am} Pathol 1997; 150: 793-803. 49. Henry-Amar M, MarnayJ. Personal communication, 1997. 50. MacLennan KA, Bennett MH, Bosq J, et al. The histology and immunohistology of Hodgkin's disease: the relationship to prognosis and clinical behavior. In: Sommers R, Henry-Amar M, Carde P, eds Treatment strategy in Hodgkin's disease. London, Paris: John Libbey, 1990:17-25. 51. Poppema S, Kaiserling E, Lennert K. Nodular paragranuloma and progressively transformed germinal centres: ultrastructural and immunohistologic findings. Virchows Arch B Cell Path 1979; 31: 211-25. 52. Lennert K, Muller-Hermelink HK. Lymphocyten und ihre Funkionsformen - Morphologic, Organisation und immunologische Bedeutung (lecture). Verhandl Anat Gesellschaft 1975; 69:19-62. 53. Muller-Hermelink HK, Lennert K. The cytologic, histologic and functional basis for a modern classification of lymphomas. In: Lennert K, in collaboration with Stein H, Mohri N, Kaiserling E, Muller-Hermelink HK, eds Malignant lymphomas other than Hodgkin's disease. New York: Springer, 1978: 38^1. 54. Neiman RS. Current problems in the histopathologic diagnosis and classification of Hodgkin's disease. Pathol Annu 1978; 13: 289-328. 55. Ramsey AD, Smith WJ, Isaacson PG. T-cell rich-B-cell lymphoma. AmJSurg Pathol 1988; 12: 433-43.
56. Bennett MH, MacLennan KA, Vaughan Hudson B, Vaughan Hudson G. The clinical and prognostic relevance of histopathological classification in Hodgkin's disease. ProgSurg Pathol 1989; 10:127-51. 57. Tiemens W, Visser L, Poppema S. Nodular lymphocyte predominance type of Hodgkin's disease is a germinal centre lymphoma. Lab Invest 1986; 54: 457-61. 58. Hansmann ML, Wacker HH, Radzun HJ. Paragranuloma is a variant of Hodgkin's disease with a predominance of B-cells. VirchowArch (Pathol Anat) 1986; 409: 171-81. 59. Coles FB, Cartun RW, Pastuszak WT. Hodgkin's disease, lymphocyte predominant type: immunoreactivity with B-cell antibodies. Mod Pathol 1988; 1: 274-8. 60. Pinkus GS. Said JW. Hodgkin's disease, lymphocytes predominance type, nodular - further evidence for a Bcell derivation. Am J Pathol 1988; 133: 211-17. 61. Poppema S. Lymphocyte-predominance Hodgkin's disease. Int Rev Exp Pathol 1991; 33: 53-79. 62. Kuzu I, Delsol G, Jones M, Gatter KC, Mason DY. Expression of the Ig-associated heterodimer (mb-1 and B 29) in Hodgkin's disease. Histopathology 1993; 22: 141-4. 63. Mason DY, Banks PM, Chan JKC, et al. Nodular lymphocyte predominance Hodgkin's disease: a distinct clinicopathological entity. AmJSurg Pathol 1994; 18: 526-30. 64. Stein H, Hansmann M-L, Lennert K, Brandtzaeg P, Gatter KC, Mason DY. Reed-Sternberg and Hodgkin's cells in lymphocyte predominance Hodgkin's disease of nodular subtype contain J cha\n.AmJ Clin Pathol 1986; 86: 292-7. 65. Schmidt C, Sargent C, Isaacson PG. L and H cells of nodular lymphocyte predominant Hodgkin's disease showimmunoglobulin light chain restriction. Am J Pathol 1991; 139:1281-9. 66. Hell K, PringleJH, Hansmann M-L, et al. Demonstration of light chain mRNA in Hodgkin's disease .J Pathol 1993; 17:137-43. 67. Stoler MH, Nichols GE, Symbula M, Weiss LM. Lymphocyte predominance Hodgkin's disease: Evidence for k light chain restricted monotypic B cell neoplasm. Am J Pathol 1995; 146: 812-18. 68. Nicholas DS, Harris S, Wright DH. Lymphocyte predominance Hodgkin's disease: an immunohistochemical study. Histopathology 1990; 16: 157-65. 69. Hsu SM, Ho YS, Li PJ, et al.t&H variants of Reed-Sternberg cells express sialyated Leu M1 antigen. AmJ Pathol 1986; 122:199-203. 70. Jack AS, Cunningham D, Soukop M, Liddle CN, Lee FD. Use of Leu M1 and antiepithelial membrane antigen monoclonal antibodies for diagnosing Hodgkin's diseasej Clin Pathol 1986; 39: 267-70. 71. Bosq J, Audouin J, Henry-Amar M, et al. Relationship between EBV infection, clinical, biological and histologic characteristics and response to therapy in
References 17
patients with Hodgkin's disease. In: Proceedings of the Third International Symposium on Hodgkin's Lymphoma. 1995: Abstracts. 72. Bennett MH, MacLennan KA, Vaughan Hudson B, Vaughan Hudson G. Non Hodgkins lymphoma arising in patients treated for Hodgkin's disease in BNLI: a 20 year experience. Ann Oncol 1991; 2 (suppl 2): 83-92. 73. Miettinen M, Franssila KO, Saxen E. Hodgkin's disease, lymphocytic predominance nodular increased risk for subsequent non-Hodgkin's lymphoma. Cancer 1983; 51: 2293-300. 74. Sundeen JT, Cossman J, Jaffe ES. Lymphocyte predominant Hodgkin's disease with coexistent 'large cell lymphoma': histological progression or composite malignancy? Am JSurgPathol 1988; 12: 599-606. 75. Hansmann ML, Stein H, Fellbaum C. etal. Nodular paragranuloma can transform into high-grade malignant lymphoma of B type. Hum Pathol 1989; 20: 1169-75. 76. Whittaker M, Foucar K, Keith T, McAneny B. Letter. Am J Surg Pathol 1989; 13: 715-16. 77. Grossman DM, Hanson CA, Schnitzer B. Simultaneous lymphocyte predominant Hodgkin's disease and large cell lymphoma. Am J Surg Pathol 1991; 15: 668-76. 78. Pan LX, Diss TC, Peng HJ, Norton AJ, Isaacson PG.
lymphocyte predominant Hodgkin's disease. Ann Diagn Pathol 1999; 3: 23-34. 86. Lukes RJ. Criteria for involvement of lymph node, bone marrow, spleen and liver in Hodgkin's disease. Cancer Res 1971; 31:1755-67. 87. Strum SB, Rappaport H. Interrelations of the histological types of Hodgkin's disease. Arch Pathol 1971; 91:127-34. 88. Strum SB, Rappaport H. Consistency of histological subtypes in Hodgkin's disease in simultaneous and sequential biopsy specimens. Natl Cancer Inst Monogr 1973;36:253-60. 89. Dorfman RF. The enigma of Hodgkin's disease: current concepts based on morphologic, clinical and immunologic observations. In: Hanaoka M, Kadin ME, Mikata A, Watanabe S, eds Lymphoid malignancies, immunocytology and cytogenetics. New York: Field and Wood, 1990:167-76. 90. MacLennan KA, Bennett MH, Vaughan Hudson B, Vaughan Hudson G. Diagnosis and grading of nodular sclerosing Hodgkin's disease: a study of 2190 patients. Int Rev Exp Pathol 1992; 33: 27-51. 91. MacLennan KA, Bennett MH, Tu A, etal. Prognostic significance of cytologic subdivision in nodular sclerosing Hodgkin's disease: an analysis of 1156
Nodular lymphocyte predominance Hodgkin's disease:
patients. In: Vavallia F, Bonadonna G, Rozencweig M,
a monoclonal or polyclonal B-cell disorder? Blood 1996; 87: 2428-34.
eds Malignant lymphomas and Hodgkin's disease:
79. Greiner TC, Gascoyne RD, Anderson ME, et al. Nodular lymphocyte-predominant Hodgkin's disease associated with large cell lymphoma: analysis of Iggene rearrangements by V-J polymerase chain reaction. Blood 1996; 88: 657-66. 80. Wickert RS, Weisenburger DD, Tierens A, Greiner TC, Chan WC. Clonal relationship between lymphocytic predominance Hodgkin's disease and concurrent or subsequent large cell lymphoma of B lineage. Blood 1995:86:2312-20. 81. Tefferi A, Wiltsie JC, Kurtin PJ. Secondary T cell
experimental and therapeutic advances. Dordrecht: Martinus Nijhoff Publishing, 1985:187-200. 92. Bennett MH, MacLennan KA, Easterling MJ, Vaughan Hudson B, Vaughan Hudson G, Jelliffe AM. Analysis of histological subtypes of Hodgkin's disease in relation to prognosis and survival. In: Quaglino D, Hayhoe FGJ, eds. The cytobiology of leukaemia and lymphomas, Serono Publications, Vol. 20, New York: Raven Press, 1985: 15-32. 93. MacLennan KA, Bennett MH, Tu A, Vaughan Hudson B, Vaughan Hudson G. The relationship of histopathology to survival and relapse. A study of 1659 patients. Cancer
82. Rysenga E, Linden MD, Carey JL, Ross CW, Schnitzer B,
1989;64: 1686-93. 94. Strickler JG, Michie SA, Warnke RA, Dorfman RF. The 'syncytial variant' of nodular sclerosing Hodgkin's disease. Am J Surg Pathol 1986; 10: 470-7.
Sawdyk M, Maeda K. Peripheral T-cell non-Hodgkin's
95. Ben-Yehuda-Salz D, Ben-Yehuda A, Polliack A, etal.
lymphoma in the setting of nodular lymphocyte predominance Hodgkin's disease. AmJHaematol 1992; 40: 232-3.
lymphoma following treatment of nodular lymphocyte
Syncytial variant of nodular sclerosing Hodgkin's
predominance Hodgkin's disease. Arch Pathol Lab Med 1995;119:88-91.
disease. A new clinicopathologic entity. Cancer 1990,
83. Delabie J, Greiner TC, Chan WC, Weisenburger DD. Concurrent lymphocyte predominance Hodgkin's disease and T-cell lymphoma. Am J Surg Pathol 1996; 20: 355-62. 84. Harris NL, Jaffe ES, Stein H, et al. A revised European-American classification of lymphoid neoplasms: a proposal from the International Lymphoma Study Group. Blood 1994; 84:1361-92.
65: 1167-72. 96. Colby TV, Hoppe RT, Warnke RA. Hodgkin's disease: a clinicopathologic study of 659 cases. Cancer 1981; 49: 1848-58. 97. Gartner HV, Wherman M, Inniger R, Steinke B. Nodular sclerosing Hodgkin's disease: prognostic relevance of morphological parameters. In: First International Symposium on Hodgkin's Lymphoma, Cologne, 27A, 1987. 98. Jairam R, Vrints LW, Breed WPM, Wijlhuizen TJ, Wijnen
85. Hancock JC, Wells A, Hailing KC, et al. Composite B-cell
TJM. Histological subclassification of the nodular sclerotic
and T-cell lymphoma arising 24 years after nodular
subtype of Hodgkin's disease. NethJ Med 1988; 33:160-7.
18 Hodgkin's disease 99. Wijlhuizen TJ, Vrints LW, Jairam R, et al. Grades of nodular sclerosis (NSI-NSII) in Hodgkin's disease: are they independent prognostic values? Cancer 1989; 63: 1150-3. 100. Ferry JA, Linggood RM, Convery KM, Efird JT, Eliseo R, Harris ML Hodgkin's disease, nodular sclerosis type implications of histologic subdassification. Cancer 1993; 71:457-63. 101. Georgii A, Hasenclever D, Fischer R, etal. Histopathological grading of nodular sclerosing Hodgkin's reveals significant differences in survival and relapse rates under protocol-therapy. Proceedings of the Third International Symposium on Hodgkin's Lymphoma, Kbln, 1995, Abstract 83. 102. Masih AS, Weisenburger DD, Vose JM, Bast MA, Armitage JO. Histologic grade does not predict prognosis in optimally treated advanced stage nodular sclerosing Hodgkin's disease. Cancer 1992; 69: 228-32. 103. Hess JL, Bodis S, Pinkus G, Silver B, Mauch P. Histopathologic grading of nodular sclerosis Hodgkin's disease: lack of prognostic significance in 254 surgically staged patients. Cancer 1994; 74: 708-1714. 104. Borowitz MJ, Croker BP, Metzger RS. Immunohistochemical analysis of the distribution of lymphocyte subpopulations in Hodgkin's disease. Cancer Treat Rep 1982; 66: 667-74. 105. Abdulaziz S, Mason DY, Stein H, Gatter KC, Nash JRG. An immunohistological study of the cellular constituents of Hodgkin's disease using a monoclonal antibody panel. Histopathology 1984; 8:1-25. 106. Pinkus GS, Thomas P, Said JW. Leu M1 - a marker for Reed-Stern berg cells in Hodgkin's disease. AmJPathol 1985;119:244-52. 107. Hall PA, D'Ardenne AJ. Value of CD15 immunostaining in diagnosing Hodgkin's disease: a review of published literature.7 Clin Pathol 1987; 40:1298-304. 108. Hall PA, D'Ardenne AJ, Stansfield AJ. Paraffin section immunohistochemistry. II. Hodgkin's disease and large cell ana plastic (Ki1) lymphoma. Histopathology 1988; 13:161-9. 109. Chittal SM, Caveriviere R, Schwarting R, et al. Monoclonal antibodies in the diagnosis of Hodgkin's disease: the search for a rational panel. AmJSurg Pathol 1988; 12: 9-21. 110. Werner M, Georgii A, BernhardsJ, Hubner K, Schwarze E-W, Fischer R. Characterization of giant cells in Hodgkin's lymphomas by immunohistochemistry applied to randomly collected diagnostic biopsies from the German Hodgkin trial. Haematol Oncol 1990; 8: 241-50. 111. Cibull ML, Stein H, Gatter KC, Mason DY. The expression of the CD3 antigen in Hodgkin's disease. Histopathology 1989;15:597-605. 112. Casey TT, Olson SJ, Cousar JB, Collins RD. Immunophenotypes of Reed-Stern berg cells: a study of 19 cases of Hodgkin's disease in plastic-embedded sections. Blood 1989; 74: 2624-8.
113. Korkolopoulou P, Cordell J, Jones M,et al. The expression of the B-cell marker mb-1 (CD 79a) in Hodgkin's disease. Histopathology 1994; 24: 511-15. 114. Isaacson PG, Ashton-Key M. Phenotype of Hodgkin and Reed-Stern berg cells. Lancet 1996; 347: 481. 115. Carbone A, Gloghini A, Gaidano G, et al. Expression status of BCL-6 and syndecan-1 identifies distinct histogenetic subtypes of Hodgkin's disease. Blood 1998; 92: 2220-8. 116. Carbone A, Gloghini A, Gattei V, et al. Reed-Sternberg cells of classical Hodgkin's disease react with the plasma cell-specific monoclonal antibody B-B4 and express human syndecan-1. Blood 1997; 89: 3787-94. 117. Carbone A, Gloghini A, Larocca LM,etal. Human immunodeficiency virus-associated Hodgkin's disease derives from post-germinal center B cells. Blood 1999; 93: 2319-26. 118. Agnarrson BA, Kadin ME. Ki1 positive large cell lymphoma: a morphological study of 19 cases. AmJ Surg Pathol 1988; 12: 264-74. 119. Stein H. Hodgkin's disease. AmJ Surg Pathol 1997; 21: 119-20. 120. Morris SW, Kirstein MN, Valentine MB, Dittmer KG, Shapiro DN, Saltman DL, Look AT. Fusion of a kinase gene, ALK, to a nucleolar protein gene, NPM, in nonHodgkin's lymphoma. Science 1994; 262:1281-4. 121. Shiota M, Nakamura S, Ichinohasama R, et al. Anaplastic large cell lymphomas expressing the novel chimeric protein p80NPM/ALK: a distinctive clinicopathologic entity. Blood 1995; 86:1954-60. 122. Pulford K, Lamant L, Morris SW, et al. Detection of anaplastic lymphoma kinase (ALK) and nucleolar protein nucleophosphormin (NPM) - ALK proteins in normal and neoplastic cells with the monoclonal antibody ALK1. Blood 1997; 89:1394-404. 123. Trumper L, Daus H, Merz H, etal. NPM/ALK fusion mRNA expression in Hodgkin and Reed-Stern berg eel Is is rare but does occur: results from single-cell cDNA analysis. Ann Oncol 1997; 8 (suppl 2): 83-7. 124. Trumper L, Pfreundschuh M, Bonin FV, Daus H. Detection of the t(2;5)-associated NPM/ALK fusion cDNA in peripheral blood cells of healthy individuals. BrJ Haematol 1998; 103:1138^4. 125. Kant JA, Hubbard SM, Longo DL, Simon RM, DeVita VT, Jaffe ES. A critical appraisal of the pathologic and clinical heterogeneity of lymphocyte depleted Hodgkin's disease, y Clin Oncol 1986; 4: 284-94. 126. Doggett RS, Colby TV, Dorfman RF. Interfollicular Hodgkin's disease. AmJSurg Pathol 1983; 7:145-9. 127. Mohrmann RL, Nathwani BN, Brynes RK, Sheibani K. Hodgkin's disease occurring in monocytoid B-cell clusters. AmJ Clin Pathol 1991; 95: 802-8. 128. Ashton-Key M, Thorpe PA, Allen JP, Isaacson PG. Follicular Hodgkin's disease. Am J Surg Pathol 1995; 19: 1294-9. 129. Maheswaran PR, Ramsay AD, Norton AJ, Roche WR. Hodgkin's disease presenting with the histological
References 19
features of Castleman's disease. Histopathology 1991; 18: 249-53. 130. Torti FM, Dorfman RF, Rosenberg SA, Kaplan HS. The changing significance of histology in Hodgkin's disease. ProcAm Assoc Cancer Res 1979; 20: 401 (C-454). 131. Dorfman RF, Colby TV. The pathologists role in the management of patients with Hodgkin's disease. Cancer Treat Rep 1982; 66: 675-80.
132. Culline S, Henry-Amar H, Diebold J, et al. Relationship of histological subtypes to prognosis in early stage Hodgkin's disease: a review of 312 cases enrolled in a controlled clinical trial. EurJ Cancer 1989; 25: 551-6. 133. Vaughan Hudson B, Vaughan Hudson G, MacLennan KA, Bennet, MM, Jelliffe AM. A retrospective evaluation of radiotherapy as a curative agent in localised Hodgkin's disease. BrJ Cancer 1987; 56: 872.
This page intentionally left blank
3 Follicular lymphoma KAMACLENNAN
Introduction
21
Transformation
Morphology
21
Extranodal disease
Immunophenotype
22
References
INTRODUCTION The first description of the entity now termed follicular lymphoma was provided by Ghon and Roman in 1916,' but it was the publications of Brill et al.2 and Symmers3'4 that brought the entity of follicular lymphoma (FL) to the attention of clinicians and pathologists. Although initially there was confusion as to whether FL was a neoplastic condition, it gradually became clear that it was a distinctive form of low-grade non-Hodgkin's lymphoma. The favorable prognostic significance of a follicular pattern and a superior response rate to radiotherapy of FL were recognized.5 Studies undertaken by Rappaport and co-workers6 delineated histologic criteria for the recognition of FL; however, they felt that an origin from the germinal center was unproven and applied the term nodular lymphoma. Detailed morphologic and ultrastructural studies by Lennert and co-workers7-9 and Lukes and Collins10 clearly identified FL as a germinal center neoplasm. FL is characterized by the neoplastic proliferation of germinal center B cells (both centroblasts and centrocytes) arranged in rounded aggregates which recapitulate the non-neoplastic germinal center. Within the neoplastic follicle are reactive cellular elements, which inhabit the normal germinal center. These are follicular dendritic cells, macrophages and T cells. A description of the cytogenetic and molecular biologic events that occur in follicular lymphoma is provided in Chapters 9 and 12.
MORPHOLOGY Lymph nodes affected by FL are enlarged and usually
22 23 23
show complete effacement of their normal architecture by neoplastic follicles. These are relatively uniform in size when compared to reactive germinal centers and are closely packed with compression of the surrounding lymphoid architecture (Plate 21); this process involves the entire lymph node in the majority of cases.11 The neoplastic follicles show a uniformity in their cellular composition and lack the distinct 'zoning' seen in reactive germinal centers (Plate 22). The mantle zones of the follicles in FL are usually ill-defined and difficult to appreciate,12 a factor associated with an increased mortality rate by some workers.13 Follicular lymphoma may be completely follicular in architecture or may have diffuse areas, which may range from small foci to large expanses; the latter is often seen in association with centroblastic cytology. The prognostic significance of diffuse areas in FL remains unclear and some workers believe they are associated with a decreased median survival.14'15 This area was addressed as part of a larger investigation of lymphoma classification.16 It was found that, if there was unequivocal follicularity within a follicle center lymphoma, the extent of the diffuse area was unimportant prognostically.16 Follicular lymphomas may be associated with intranodal sclerosis either as collagen bands or as fine compartmentalizing fibrosis.17-19 Sclerosis associated with FL often occurs in retroperitoneal and inguinal lymph nodes, and is associated with a more favorable prognosis.17-19 In some cases of FL the neoplastic B cells may develop cytoplasmic vacuolation and resemble the signet ring cells of poorly differentiated adenocarcinoma. This change may affect a minor population of cells or be extensive (Plate 23). The cytoplasmic vacuoles in the signet ring variant of FL may be either clear or associated with pink globules; the former has been associated with
22 Follicular lymphoma
immunoglobulin G (IgG) production and the latter with IgM.20~23 Occasionally large quantities of extracellular eosinophilic material may be found within the follicles.24'25 Other rare variants of FL include cases showing marginal zone differentiation,26-30 rosette formation,31 a 'floral pattern' to the neoplastic follicles,32 epithelioid granuloma formation33'34 and pronounced plasma cell differentiation.35 With the exception of marginal zone differentiation, which is associated with a decreased survival,36 these variants do not affect prognosis. The neoplastic follicles of FL contain a variable composition of centroblasts and centrocytes, which may range from a predominance of centrocytes to rare cases in which the follicles contain sheets of centroblasts. This cytological variability has formed the basis for a series of grading systems for FL. These grading systems employ either a subjective assessment of the percentage of large cells present within the follicles37-40 or by counting the number of large non-cleaved cells per high-power field41 or both.42 Some have used a proliferative index as assessed by automated image analysis of Ki 67 staining in FL to predict prognosis; a high proliferative index also showed a close correlation with grading.43 Although differences in survival can be demonstrated between different grades of FL (Fig. 3.1), they are small and all the methods so far employed suffer from very poor interobserver concordance rates.44,45 Objective evidence of'cure' as evidenced by a plateau in the actuarial survival curve, is not seen for the majority of patients.
IMMUNOPHENOTYPE Phenotypically FL is seen to be composed of rounded aggregates of B cells with similarities to normal germinal center B cells. There is expression of pan-B cell antigens CD 19, 20 and 22, surface immunoglobulin and
Figure 3.1 Actuarial survival of patients subdivided using the Berard criteria into follicular lymphoma small, mixed and large cell types.
expression of CD 10. Antibodies to CD 10 are now available that work in routinely fixed and processed tissue46 (Plate 24). CD 5 and CD 43 are usually negative and there is no nuclear cyclin Dl expression. These are important discriminating features from mantle cell lymphomas.47-52 The neoplastic B cells in FL differ from normal germinal center B cells in the presence of BCL 2 protein within their cytoplasm in up to 85 per cent of cases53"55 (Plate 25). Cases of follicular lymphoma with predominantly large cell cytology express BCL 2 protein less frequently.55 The expression of BCL 2 protein is of practical value in the discrimination of FL from florid follicular hyperplasia as it is not expressed in the B cells of the latter condition. Cell adhesion molecules, such as VLA 4-VCAM and LFA1-ICAM, mediate the interactions between the neoplastic B cells and the follicular dendritic cells (FDCs), and appear to be significant in the retention of a follicular architecture.56-65 Follicular dendritic cells within reactive germinal centers show a degree of heterogeneity in their antigenic profile, in particular expression of CD 21 is restricted to FDCs in the apical light zone.66 As CD 21 is expressed in many examples of FL (Plate 26) this suggests there may be similarities between FL and the light zone of the germinal center. Numerous reactive T cells are also seen scattered within the follicles, which express pan-T cell antigens CD 3 and CD 2. Many are of the helper-subtype expressing CD 4 and show expression of CD 40 ligand.67 Some also express the germinal center T cell-associated marker CD 57.
TRANSFORMATION Transformation of FL into a diffuse, high-grade nonHodgkin's lymphoma is relatively common and is associated with a poor prognosis.68 The frequency of histologic transformation is variable and ranges from around 30 per cent69'70 to an actuarial prediction that 60 per cent of patients will transform.71 Transformation tends to occur early in the course of the disease, and is associated with adverse prognostic factors or failure to achieve complete remission; the rate of transformation shows a tendency to plateau at 6 years.69 Most transformed FLs show histologic features of diffuse large B cell lymphoma with centroblastic cytology.72 The genetic factors involved in transformation are at present unknown, somatic mutations in the translocated bcl 2 gene,73"75 cytogenetic abnormalities at 6q23-26 and 17p,76 and, mutations in p5377'78 or overexpression of p53 protein79 have been identified as possible factors. Rare case of FL may transform into high-grade nonHodgkin's lymphoma with variant histology, sometimes with 'Burkitt-like' features and an aggressive leukaemic course; these are often associated with t(8;14) with deregulation of c-myc, in addition to t(14;18).80-86 Rare
References 23
cases of transformation of FL to CD 30-large cell lymphoma with anaplastic features have been described.87
EXTRANODAL DISEASE
5. Gall EA and Mallory TB. Malignant lymphoma. A clinicopathologic survey of 618 cases. AmJ Pathol 1941; 18: 381-429. 6. Rappaport H, Winter WJ, Hicks. Follicular lymphoma: a re-evaluation of its position in the scheme of malignant lymphoma, based on a survey of 253 cases. Cancer
Follicular lymphoma is typically disseminated at presentation with the involvement of a wide range of organs whose function may or may not be compromised. One of the commonest sites of extranodal disease is the bone marrow, which shows histological evidence of disease in approximately three-quarters of cases.88 The earliest morphological features of bone-marrow infiltration by FL are paratrabecular aggregates of small B cells (Plate 27),89'91 which are admixed with follicular dendritic cells and T cells.54,92-94 More extensive marrow disease shows the presence of neoplastic follicles within the hemopoietic marrow.95 When extensive marrow involvement is present, the normal marrow elements are displaced by sheets of coalescing follicles and normal hemopoietic function may be compromised. Cytologically, small irregular B cells predominate. In the presence of heavy bone-marrow infiltration, FLs may develop a leukemic phase in which cytologically atypical lymphoid cells are found in the peripheral blood; the leukemic cell count may be very high96,97 (Plate 28). The spleen is commonly involved by FL. The lymphoma cells preferentially home in to the white pulp regions98"101 and form expansile white nodules measuring several millimetres in diameter, imparting a miliary appearance to the cut surface (Plate 29). Large tumor masses are not usually seen in the absence of high-grade transformation. Many other organs may be infiltrated by FL, including the liver,95 soft tissues102 and skin.103 In extranodal locations the follicular pattern may be difficult to discern and may require the immunohistochemical demonstration of a follicular dendritic cell meshwork.
1956; 9: 792-821. 7. Lennert K. Germinal centers and germinal center neoplasms. Nippon Ketsueki Gakkai Zasshi 1969; 32: 495-500. 8. Lennert K. Giant follicular lymphoma. Dtsch Med Wochenschr 1973; 98: 335-6. 9. Lennert K, Stein H, Kaiserling E. Cytological and functional criteria for the classification of malignant lymphomata. BrJ Cancer 1975; 31 (suppl 2): 29-43. 10. Lukes RJ, Collins RD. New approaches to the classification of the lymphomata. BrJ Cancer 1975; 31 (suppl 2): 1-28. 11. Nathwani BN, Winberg CD, Diamond LW, Bearman RM, Kim H. Morphologic criteria for the differentiation of follicular lymphoma from florid reactive follicular hyperplasia: a study of 80 cases. Cancer 1981; 48: 1794-806. 12. Crocker J, Jones EL, Curran RC. A quantitative study of the size of benign and malignant lymphoid follicles. 7 Clin Pathol 1983; 36:1055-61. 13. West KP, Potter LJ, Henderson SD, Lauder I. A retrospective study of follicular lymphomas. Histopathology 1989; 14: 629-36. 14. Ostrow SS, Diggs CH, Sutherland JC, Gustafson J, Wiernik PH. Nodular poorly differentiated lymphocytic lymphoma: changes in histology and survival. Cancer Treat Rep 1981; 65: 929-33. 15. Vose JM, Bierman PJ, Lynch JC, et al. Effect of follicularity on autologous transplantation for large-cell non-Hodgkin's lymphoma. J Clin Oncol 1998; 16: 844-9. 16. Anonymous. A clinical evaluation of the International Lymphoma Study Group classification of non-Hodgkin's lymphoma. The Non-Hodgkin's Lymphoma Classification Project. Blood 1997; 89: 3909-18.
REFERENCES 1. Ghon A, Roman B. Ueber das Lymphosarkom. Frankfurt ZPathol 1916; 19:1-138. 2. Brill NE, Baehr G, Rosenthal N. Generalised giant lymph follicle hyperpalasia of the lymph follicles and spleen: a hitherto undescribed type. JAMA 1925; 84: 668-71. 3. Symmers D. Follicular lymphadenopathy with
17. Bennett MH. Sclerosis in non-Hodgkin's lymphomata. Br J Cancer 1975; 31 (suppl 2): 44-52. 18. Bennett MH, Millett YL Nodular sclerotic lymphosarcoma: a possible new clinico-pathological entity. Clin Radiol 1969; 20: 339^3. 19. Millett YL, Bennett MH, Jelliffe AM, Farrer-Brown G. Nodular sclerotic lymphosarcoma. A further review. BrJ Cancer 1969; 23: 683-92. 20. Kim H, Dorfman RF, Rappaport H. Signet ring cell
splenomegaly: a newly recognised disease of the
lymphoma. A rare morphologic and functional
lymphatic system. Arch. Pathol 1927; 3: 816-20.
expression of nodular (follicular) lymphoma. Am J Surg
4. Symmers D. Giant follicular lymphadenopathy with or without splenomegaly: its transformation into polymorphous cell sarcoma of the lymph follicles and its association with Hodgkin's disease, lymphatic leukaemia, and an apparently unique disease of the lymph nodes and spleen - a disease entity believed heretofore undescribed. Arch Pathol 1938; 26: 603-47.
PatholWJS; 2:119-32. 21. Harris M, Eyden B, Read G. Signet ring cell lymphoma: a rare variant of follicular lymphoma. J Clin Pathol 1981; 34:884-91. 22. Silberman S, Fresco R, Steinecker PH. Signet ring cell lymphoma. A report of a case and review of the literature. AmJ Clin Pathol 1984; 81: 358-63.
24 Follicular lymphoma
23. Spagnolo DV, Papadimitriou JM, Matz LR, Walters MN. Nodular lymphomas with intracellular immunoglobulin inclusions: report of three cases and a review. Pathology 1982; 14: 415-27. 24. Talerman A, Platenburg HP. Follicular lymphoma with deposits of amorphous hyaline material, y Pathol 1974; 112:27-31. 25. Chittal SM, Caveriviere P, Voigt JJ, et al. Follicular
38. Lukes RJ, Collins RD. Immunologic characterization of human malignant lymphomas. Cancer 1974; 34 (suppl): 1488-503. 39. Henry K, Bennett MH, Farrer-Brown G. Classification of the non-Hodgkin's lymphomas. In: Anthony P and Woolf N, eds. Recent advances in histopathology 10. Edinburgh: Churchill Livingstone. 1978: 275-302. 40. Anonymous. National Cancer Institute sponsored study
lymphoma with abundant PAS-positive extracellular
of classifications of non-Hodgkin's lymphomas:
material. Immunohistochemical and ultrastructural observations. AmJSurg Pathol 1987; 11: 618-24.
summary and description of a working formulation for
26. Chan JK, Ng CS, Hui PK. An unusual morphological variant of follicular lymphoma. Report of two cases. Histopathology 1988; 12: 649-58. 27. Hernandez AM, Nathwani BN, Nguyen D, et al. Nodal benign and malignant monocytoid B cells with and without follicular lymphomas: a comparative study of follicular colonization, light chain restriction, bcl-2, and t(14;18) in 39 cases. Hum Pathol 1995; 26: 625-32. 28. Nathwani BN, Hernandez AM, Deol I, Taylor CR. Marginal zone B-cell lymphomas: an appraisal. Hum Pathol 1997; 28: 42-6. 29. Ree HJ, Leone LA. Prognostic significance of parafollicular small lymphocytes in follicular lymphoma: clinicopathological studies of 82 cases of primary nodal origin. Cancer 1978; 41:1500-10. 30. Schmid U, Cogliatti SB, DissTC, Isaacson PG. Monocytoid/marginal zone B-cell differentiation in follicle centre cell lymphoma. Histopathology 1996; 29: 201-8. 31. Frizzera G, Gajl-Peczalska K, Sibley RK, Rosai J, Cherwitz D, Hurd DD. Rosette formation in malignant lymphoma. Am J Pathol 1985; 119: 351-6. 32. GoatesJJ, Kamel OW, LeBrun DP, Benharroch D, Dorfman RF. Floral variant of follicular lymphoma. Immunological and molecular studies support a neoplastic process. AmJSurg Pathol 1994; 18: 37-47. 33. Kojima M, Nakamura S, Motoori T, et al. Centroblastic and centroblastic-centrocytic lymphomas associated with prominent epithelioid granulomatous response without plasma cell differentiation: a clinicopathologic study of 12 cases. Hum Pathol 1996; 27: 660-7. 34. Naresh KN. Morphological spectrum of follicle center cell lymphomas associated with infiltration of epithelioid histiocytes. Hum Pathol 1997; 28:114-15. 35. Frizzera G, Anaya JS, Banks PM. Neoplastic plasma cells in follicular lymphomas. Clinical and pathologic findings in six cases. Virchows Arch A Pathol Anat Histopathol 1986; 409:149-62. 36. Nathwani BN, Anderson JR, Armitage JO, et al. Clinical significance of follicular lymphoma with monocytoid B-cells. Non-Hodgkin's lymphoma classification project. Hum Pathol 1999; 30: 263-8. 37. Rappaport H. Tumors of the hematopoietic system.
clinical usage. The Non-Hodgkin's Lymphoma Pathologic Classification Project. Cancer 1982; 49: 2112-35. 41. Mann RB, Berard CW. Criteria for the cytologic subclassification of follicular lymphomas: a proposed alternative method. Hematol Oncol 1983; 1:187-92. 42. Jaffe ES, Raffeld M, Medeiros LJ. Histopathologic subtypes of indolent lymphomas: caricatures of the mature B-cell system. Semin Oncol 1993; 20: 3-30. 43. Martin AR, Weisenburger DD, Chan WC, et al. Prognostic value of cellular proliferation and histologic grade in follicular lymphoma. Blood 1995; 85: 3671-8. 44. Metter GE, Nathwani BN, Burke JS, et al. Morphological subclassification of follicular lymphoma: variability of diagnoses among hematopathologists, a collaborative study between the Repository Center and Pathology Panel for Lymphoma Clinical Studies.7 Clin Oncol 1985; 3: 25-38. 45. Nathwani BN, Metter GE, Miller TP, et al. What should be the morphologic criteria for the subdivision of follicular lymphomas? Blood 1986; 68: 837-45. 46. Mclntosh GG, Lodge AJ, Watson P, et al. NCL-CD10-270: a new monoclonal antibody recognizing CD10 in paraffin-embedded tissue. Am J Pathol 1999; 154: 77-82. 47. Banks PM, Chan J, Cleary ML, et al. Mantle cell lymphoma. A proposal for unification of morphologic, immunologic, and molecular data. Am J Surg Pathol 1992; 16, 637^10. 48. Contos MJ, Kornstein MJ, Innes DJ, Ben-Ezra J. The utility of CD20 and CD43 in subclassification of low-grade Bcell lymphoma on paraffin sections. Mod Pathol 1992; 5: 631-3. 49. Isaacson PG. Malignant lymphomas with a follicular growth pattern. Histopathology 1996; 28: 487-95. 50. Swerdlow SH, Zukerberg LR, Yang Wl, Harris NL, Williams ME. The morphologic spectrum of nonHodgkin's lymphomas with BCL1/cyclin D1 gene rearrangements. Am J Surg Pathol 1996; 20: 627-40. 51. Treasure J, Lane A, Jones DB, Wright DH. CD43 expression in B cell lymphoma. J Clin Pathol 1992; 45: 1018-22. 52. Vasef MA, Medeiros LJ, Koo C, McCourty A, Byrnes RK. Cyclin D1 immunohistochemical staining is useful in distinguishing mantle cell lymphoma from other low-
Washington, DC: Armed Forces Institute of Pathology,
grade B-cell neoplasms in bone marrow. AmJ Clin
1966.
Pathol 1997; 108: 302-7.
References 25 53. Ashton-Key M, DissTC, Isaacson PG, Smith ME. A comparative study of the value of immunohistochemistry and the polymerase chain reaction in the diagnosis of follicular lymphoma. Histopathology 1995; 27: 501-8. 54. Ben-Ezra JM, King BE, Harris AC, Todd WM, Kornstein MJ. Staining for Bcl-2 protein helps to distinguish benign from malignant lymphoid aggregates in bone marrow biopsies. Mod Pathol 1994; 7: 560-4. 55. Gaulard P, d'Agay MF, Peuchmaur M, etal. Expression of the bcl-2 gene product in follicular lymphoma. AmJ Pathol 1992; 140:1089-95. 56. Lampert IA, Van Noorden S. Acetyl cholinesterase is expressed in the follicular dendritic cells of germinal centres: differences between normal and neoplastic follicles.yPof/70/1996; 180:169-74. 57. Liu YJ, Grouard G, de Bouteiller 0, Banchereau J. Follicular dendritic cells and germinal centers. Int Rev Cytol 1996; 166:139-79. 58. Vyth-Dreese FA, Dellemijn TA, van Oostveen JW, Feltkamp CA, Hekman A. Functional expression of adhesion receptors and costimulatory molecules by fresh and immortalized B-cell non-Hodgkin's lymphoma cells. Blood 1995; 85: 2802-12. 59. Ishii G, Harigaya K, Soeta S, Mikata A. VLA-4-dependent adhesion in follicular non-Hodgkin's lymphomas. Hematol Pathol 1995; 9:155-69. 60. Kuriyama Y, Nakano M, Kawanishi Y, Iwase 0, Kuge S, Toyama K. Significance of VLA-4 and LFA-1 expressions in neoplastic follicle formation and its deterioration in B-cell non-Hodgkin's lymphomas. Leuk Lymph 1994; 13: 123-9. 61. Ree HJ, Khan AA, Elsakr M, Liau S, Teplitz C. Intercellular adhesion molecule-1 (ICAM-1) staining of reactive and neoplastic follicles. ICAM-1 expression of neoplastic follicle differs from that of reactive germinal center and is independent of follicular dendritic cells. Cancer 1993; 71: 2817-22. 62. Gloghini A, Carbone A. The nonlymphoid microenvironment of reactive follicles and lymphomas of follicular origin as defined by immunohistology on paraffin-embedded tissues. Hum Pathol 1993; 24: 67-76. 63. Petrasch S, Kosco M, Schmitz J, Wacker HH, Brittinger G. Follicular dendritic cells in non-Hodgkin lymphoma express adhesion molecules complementary to ligands on neoplastic B-cells. BrJ Haematol 1992; 82: 695-700. 64. Freedman AS, Munro JM, Morimoto C, et al. Follicular non-Hodgkin's lymphoma cell adhesion to normal germinal centers and neoplastic follicles involves very late antigen-4 and vascular cell adhesion molecule-1. S/oo1992;79:206-12. 65. Freedman AS, Saporito L, Rhynhart K, Morimoto C, Nadler LM. Adhesion of follicular lymphoma cells to lymphoid germinal centers-a potential mechanism of tumor cell homing following autologous transplantation. Leuk Lymph 1994; 13: 47-52.
66. Imal Y, Yamakawa M. Morphology, function and pathology of follicular dendritic cells. Pathol Int 1996; 46: 807-33. 67. Carbone A, Gloghini A, Gruss HJ, Pinto A. CD40 ligand is constitutively expressed in a subset of T cell lymphomas and on the microenvironmental reactive T cells of follicular lymphomas and Hodgkin's disease. AmJ Pathol 1995; 147: 912-22. 68. ArmitageJO, Dick FR, Corder MD. Diffuse histiocytic lymphoma after histologic conversion: a poor prognostic variant. Cancer Treat Rep 1981; 65: 413-18. 69. Bastion Y, Sebban C, Berger F, et al. Incidence, predictive factors, and outcome of lymphoma transformation in follicular lymphoma patients. J Clin Oncol 1997; 15:1587-94. 70. Cullen MH, Lister TA, Brearley Rl, Shand WS, Stansfeld AG. Histological transformation of non-Hodgkin's lymphoma: a prospective study. Cancer 1979; 44: 645-51. 71. Acker B, Hoppe RT, Colby TV, Cox RS, Kaplan HS, Rosenberg SA. Histologic conversion in the nonHodgkin's lymphomas. J Clin Oncol 1983; 1:11-16. 72. Lennert K, Feller AC. Centroblastic-centrocytic lymphoma. Histopathology of non-Hodgkin's lymphomas (based on the updated Kiel Classification). Berlin: Springer-Verlag, 1990. 73. Matolcsy A, Warnke RA, Knowles DM. Somatic mutations of the translocated bcl-2 gene are associated with morphologic transformation of follicular lymphoma to diffuse large-cell lymphoma. Ann Oncol 1997; 8 (suppl 2): 119-22. 74. Matolcsy A, Casali P, Warnke RA, Knowles DM. Morphologic transformation of follicular lymphoma is associated with somatic mutation of the translocated Bcl-2 gene. Blood 1996; 88: 3937-44. 75. Raghoebier S, Broos L, Kramer MH, et al. Histological conversion of follicular lymphoma with structural alterations of t(14;18) and immunoglobin genes. Leukemia 1995; 9:1748-55. 76. Tilly H, Rossi A, Stamatoullas A, et al. Prognostic value of chromosomal abnormalities in follicular lymphoma. Blood 1994; 84:1043-9. 77. Sander CA, Yano T, Clark HM, et al. p53 mutation is associated with progression in follicular lymphomas. Blood 1993; 82:1994-2004. 78. Dalla-Favera R, Ye BH, Lo Coco F, et al. Identification of genetic lesions associated with diffuse large-cell lymphoma. Ann Oncol 1994; 5 (suppl 1): 55-60. 79. Symmans WF, Katz RL, Ordonez NG, Dalton H, Romaguera JE, Cabanillas F. Transformation of follicular lymphoma. Expression of p53 and bcl-2 oncoprotein, apoptosis and cell proliferation. Acta Cytol 1995; 39: 673-82. 80. Thangavelu M, Olopade 0, Beckman E, et al. Clinical, morphologic, and cytogenetic characteristics of patients with lymphoid malignancies characterized by both t(14; 18)(q32;q21) and t(8;14)(q24;q32) or t(8;22)(q24;q11). Genes Chromosomes Cancer 1990; 2:147-58.
26 Follicular lymphoma 81. Gauwerky CE, Huebner K, Isobe M, Nowell PC, Croce CM. Activation of MYC in a masked t(8;17) translocation results in an aggressive B-cell leukemia. Proc Natl Acad SciLISA 1989; 86: 8867-71. 82. Gauwerky CE, Hoxie J, Nowell PC, Croce CM. Pre-B-cell leukemia with a t(8; 14) and a t(14; 18) translocation is preceded by follicular lymphoma. Oncogene 1988; 2: 431-5. 83. Sham RL, Phatak P, Carignan J, Janas J, Olson JP. Progression of follicular large cell lymphoma to Burkitt's lymphoma. Cancer 1989; 63: 700-2. 84. Yano T, Jaffe ES, Longo DL, Raffeld M. MYC rearrangements in histologically progressed follicular lymphomas. Blood 1992; 80: 758-67. 85. Bisiau H, Daudignon A, Le Baron F, et al. Transformation of follicular lymphoma with both t(14;18) and t(8; 22). Nouv Rev Fr Hematol 1995; 37: 241--4. 86. Brito-Babapulle V, Crawford A, Khokhar T, et al. Translations t(14;18) and t(8;14) with rearranged bcl-2 and c-myc in a case presenting as B-ALL (L3). Leukemia 1991; 5: 83-7. 87. Alsabeh R, Medeiros LJ, Glackin C, Weiss LM. Transformation of follicular lymphoma into CD30-large cell lymphoma with anaplastic cytologic features. AmJ Surg Pathol 1997; 21: 528-36. 88. Criel A, Pittaluga S, Verhoef G, et al. Small B cell NHL and their leukemic counterpart: differences in subtyping and assessment of leukemic spread. Leukemia 1996; 10: 848-53. 89. Salisbury JR, Deverell MH, Seaton JM, Cookson MJ. Three-dimensional reconstruction of non-Hodgkin's lymphoma in bone marrow trephines J Pathol 1997; 181:451-4. 90. Juneja SK, Wolf MM, Cooper IA. Value of bilateral bone marrow biopsy specimens in non-Hodgkin's lymphoma. J Clin Pathol 1990; 43: 630-2. 91. Baroni CD, Manente L, Occhionero M, Marzullo A, Mandelli F, Biagini, C. Involvement of the bone marrow by non-Hodgkin's lymphomas: incidence, histology and pathologic correlations. Tumori 1981; 67:191-6. 92. Meuge-Moraw C, Delacretaz F, Baur AS. Follicular dendritic cells in bone marrow lymphoproliferative diseases: an immunohistochemical study including a new paraffin-resistant monoclonal antibody, DR53. Histopathology 1996; 28: 341-7.
93. Chetty R, Echezarreta G, Comley M, Gatter K. Immunohistochemistry in apparently normal bone marrow trephine specimens from patients with nodal follicular lymphoma. J Clin Pathol 1995; 48:1035-8. 94. Chilosi M, Pizzolo G, Fiore-Donati L, Bofill M, Janossy G. Routine immunofluorescentand histochemical analysis of bone marrow involvement of lymphoma/leukaemia: the use of cryostat sections. BrJ Cancer 1983; 48: 763-75. 95. Kim H, Dorfman RF. Morphological studies of 84 untreated patients subjected to laparotomy for the staging of non-Hodgkin's lymphomas. Cancer 1974; 33: 657-74. 96. Spiro S, Gallon DAG, Wiltshaw E, Lohmann R. Follicular lymphoma: a survey of 75 cases with special reference to the syndrome resembling chronic lymphocytic leukaemia. BrJ Cancer 1975; 31: 60-72. 97. Come SE, Jaffe ES, Andersen JC, et al. Non-Hodgkin's lymphomas in leukemic phase: clinicopathologic correlations. AmJ Med 1980; 69: 667-74. 98. Alkan S, Ross CW, Hanson CA, Schnitzer B. Follicular lymphoma with involvement of the splenic marginal zone: a pitfall in the differential diagnosis of splenic marginal zone cell lymphoma. Hum Pathol 1996; 27: 503-6. 99. Burke JS. Splenic lymphoid hyperplasias versus lymphomas/leukemias. A diagnostic guide. Am J Clin Pathol 1993; 99: 486-93. 100. Piris MA, Mollejo M, Campo E, MenarguezJ, FloresT, Isaacson PG. A marginal zone pattern may be found in different varieties of non-Hodgkin's lymphoma: the morphology and immunohistology of splenic involvement by B-cell lymphomas simulating splenic marginal zone lymphoma. Histopathology 1998; 33: 230-9. 101. Warnke R, Levy R. Immunopathology of follicular lymphomas. A model of B-lymphocyte homing. N EnglJ Med 1978; 298: 481-6. 102. Salamao DR, Nascimento AG, Lloyd RV, Chen MG, Habermann TM, Strickler JG. Lymphoma in soft tissue: a clinicopathologic study of 19 cases. Hum Pathol 1996; 27: 253-7. 103. Garcia CF, Weiss LM, Warnke RA, Wood GS. Cutaneous follicular lymphoma. AmJ Surg Pfltfjo/,1986; 10: 454-63.
4 Mantle cell lymphoma DD WEISENBURGER AND JO ARMITAGE
Introduction Pathologic features
27 27
Immunologic features Cytogenetic and molecular genetic features
31
Normal cellular counterpart Differential diagnosis
33 33
31
INTRODUCTION In the mid-1970s, Berard and colleagues1"4 coined the term lymphocytic lymphoma of intermediate differentiation to describe a group of non-Hodgkin's lymphomas that were not readily classifiable as either well-differentiated (small lymphocytic) or poorly differentiated (small cleaved cell) lymphoma. In lymph node sections, the tumors usually had a diffuse pattern of growth and were composed of a mixture of small lymphoid cells, some with round nuclei like those of small lymphocytic lymphoma, and others with indented and cleaved nuclei like those of small cleaved cell lymphoma. Thus, the term intermediate was used to describe the intermediate morphologic appearance of the tumors. About the same time, Lennert and co-workers5~7 described a similarappearing lymphoma, termed centrocytic, which was characterized by a predominance of irregular and cleaved lymphoid cells. Early immunologic studies of these tumors revealed a B-cell phenotype with the neoplastic cells showing moderate to intense staining for monoclonal surface immunoglobulin (Ig).3'4'7 Cytochemical stains for surface alkaline phosphatase suggested to Berard and colleagues3'4 that intermediate lymphocytic lymphoma corresponded to the cells of primary lymphoid follicles and the mantle zones of secondary follicles, whereas Lennert and co-workers" believed that centrocytic lymphoma was a germinal center cell lymphoma. In the early 1980s, Weisenburger and associates8 and Palutke and colleagues9 described a distinctive type of
Clinical features Treatment and survival Tumor grade Conclusion References
34 35 36 37 37
follicular lymphoma that was characterized by the proliferation of atypical small lymphoid cells in wide mantles around benign germinal centers. Weisenburger and associates8'10 coined the term mantle zone lymphoma for this entity and suggested that it represented the follicular counterpart of diffuse intermediate lymphocytic lymphoma. More recent studies, which are detailed herein, have characterized all of these various lymphomas clinically and at the molecular level, and have led to the conclusion that they represent a closely related spectrum of tumors that correspond to lymphocytes in the primary lymphoid follicles and mantle zones of secondary follicles. Thus, the term mantle cell lymphoma (MCL) is now the accepted name for this group of lymphomas.11,12
PATHOLOGIC FEATURES The pathologic features of MCL have been refined and the histologic spectrum of the disease has been expanded in recent years through the use of immunologic, cytogenetic and molecular techniques. A number of large, wellstudied series with detailed descriptions of the pathology of the MCL have been published.13-19 Lymph nodes The non-Hodgkin's lymphomas of mantle cell type usually consist of atypical small lymphoid cells, and have
28 Mantle cell lymphoma
Image Not Available
Figure 4.1 Mantle cell lymphoma composed of nodules with reactive germinal centers surrounded by wide mantles of neoplastic small lymphoid cells. Reproduced with permission from Weisenburger DD. Mantle cell lymphoma. In: Knowles D, ed. Neoplastic hematopathology. Baltimore: Williams & Wilkins, 1992: 617-28.
either a nodular or diffuse pattern of growth, or a combination of the two patterns. Modularity is present, at least focally, in approximately 30 per cent of cases of MCL at the time of initial diagnosis. Early in the course of disease, nodular MCL may have a distinctly nodular or a vaguely nodular growth pattern at low magnification. In nodular MCL, some or many of the nodules may consist of follicles with reactive germinal centers surrounded by broad and expansive mantles of small lymphoid cells (Fig. 4.1), the so-called mantle zone pattern.8,20 In such cases, however, some neoplastic nodules without germinal centers, which mimic primary follicles, are also present. In other cases, these latter nodules may predominate or be present exclusively (Fig. 4.2), and the process may be confused with a follicular center cell lymphoma of the small cleaved cell type. Later in the course of disease, invasion
Image Not Available
Figure 4.2 Mantle cell lymphoma composed of nodules of neoplastic small lymphoid cells without reactive germinal centers. Reproduced with permission from Weisenburger DD. Mantle cell lymphoma. In: Knowles D, ed. Neoplastic hematopathology. Baltimore: Williams & Wilkins, 1992: 617-28.
and obliteration of the reactive germinal centers and interfollicular areas by neoplastic cells results in a diffuse pattern of growth. Residual vague nodularity may be seen in such cases, and naked germinal centers lacking a normal lymphocyte cuff are found within the diffuse areas in approximately one-quarter of the cases (Fig. 4.3). Cytologically, a MCL usually consists of a monotonous population of atypical small to medium-sized lymphoid cells with irregular and indented nuclei, moderately coarse chromatin, inconspicuous nucleoli, and scant cytoplasm (typical 'intermediate' cytology). Small round lymphocytes, some of which are T cells, are admixed in variable numbers, and neoplastic cells with cleaved nuclei are often present as well. However, cases of MCL with predominantly round nuclei or only slight nuclear irregularity, and cases with markedly angulated and cleaved nuclei ('centrocytic' cytology) or even cerebriform nuclei do occur. Although the neoplastic lymphoid cells show a spectrum of nuclear irregularity from case to case, ranging from slight to marked, the cells usually show little variation in an individual neoplasm. In about 20 per cent of cases of MCL, the neoplastic cells are larger than usual and have more finely dispersed nuclear chromatin and small nucleoli. Such cases have been referred to as large cell ('anaplastic centrocytic') or blastic variants of MCL.6,15-17 Sometimes, a mixture of atypical small cells and larger blastic cells is present, imparting a more pleomorphic cytologic picture. In other cases, the blastic cells are quite monotonous, ranging from medium to large in size, with very fine chromatin and multiple small nucleoli ('centrocytoid centroblastic' cytology). Thus, the diversity of morphologies seen in different cases of MCL is broad, ranging from small cells with round or slightly irregular nuclei at one end of the spectrum to large transformed cells with distinct nucleoli at the other end. The cytologic spectrum of MCL is shown in Fig. 4.4.
Image Not Available
Figure 4.3 Diffuse mantle cell lymphoma with a naked germinal center lacking a normal lymphocyte cuff. Reproduced with permission from Weisenburger DD. Mantle cell lymphoma. In: Knowles D, ed. Neoplastic hematopathology. Baltimore: Williams & Wilkins, 1992: 617-28.
Pathologic features 29
Figure 4.4 Cytologic spectrum of mantle cell lymphoma including cases with round to slightly irregular nuclei (a), markedly irregular nuclei (b), blastic nuclei (c), and large vesicular nuclei with prominent nucleoli (d).
In general, large transformed lymphoid cells with vesicular nuclei and prominent nucleoli (large noncleaved cells or centroblasts) are not seen in the lymphocytic forms of MCL, and plasma cells are usually absent or only present in small numbers and polyclonal in nature. The mitotic rate is generally low in the lymphocytic forms of MCL, but an increased mitotic rate is usually seen in the pleomorphic and blastic variants, and is often accompanied by admixed benign histiocytes. These rather distinctive histiocytes have abundant pink cytoplasm and may contain phagocytized cellular debris (Fig. 4.5). Histologic progression from a nodular pattern to a diffuse pattern may be evident in a subsequent biopsy, and progression from lymphocytic to blastic cytology with a high mitotic rate is not uncommon. Norton and colleagues17 noted histological transformation to blastic
cytology upon rebiopsy in 17 per cent of their cases of MCL, and found blastic cytology in 70 per cent of their cases at autopsy. However, transformation of MCL to the more common forms of diffuse large cell lymphoma is a rare event. Cytology, peripheral blood and bone marrow In lymph node touch preparations and other cytologic specimens, the neoplastic cells are small to medium sized, with irregular, indented and cleaved nuclear contours, moderately clumped (smudged) to more finely dispersed chromatin, one or more conspicuous nucleoli, and small to moderate amounts of cytoplasm (Fig. 4.6).
Figure 4.6 Touch preparation of mantle cell lymphoma Figure 4.5 Blastic mantle cell lymphoma with admixed benign
showing medium-sized lymphoid cells with irregular nuclear contours, moderately clumped (smudged) chromatin, small
histiocytes, one of which contains phagocytized cellular debris.
nucleoli and scant cytoplasm.
30 Mantle cell lymphoma
Larger cells with round nuclei, fine chromatin, prominent nucleoli and moderate amounts of basophilic cytoplasm are seen in the blastic variants of MCL. Smears of involved bone marrow and peripheral blood generally reflect the lymphoid population present in the lymph nodes.10'21'22 The neoplastic cells in the blood and bone marrow of a given patient maybe quite heterogeneous in appearance (Fig. 4.7). In bone marrow sections, the neoplastic cells may infiltrate in either a focal, often paratrabecular, pattern or a diffuse pattern. However, one should not make a diagnosis of MCL based upon the examination of peripheral blood or bone marrow alone because of the lack of precise criteria for such a diagnosis. However, immunological studies by flow cytometry (see 'Immunologic features') may be very useful in the diagnosis of such specimens.21'22
Spleen The spleen is often enlarged in patients with MCL and particularly in those with the nodular (mantle zone) type.8'20 Spleen weights have ranged from 600 to 3700 g.8'13 Macroscopically, the cut surface often reveals numerous lymphoid nodules measuring 1-3 mm in diameter. Microscopically, the white pulp areas are markedly expanded by a proliferation of atypical lymphoid cells similar to those found in the lymph nodes. Benign and reactive-appearing germinal centers may be present in the white pulp areas, and the marginal zone is sometimes preserved (Fig. 4.8). Nodular infiltration of the red pulp is usually also present, albeit in varying degrees.
Image Not Available
Figure 4.7 Peripheral blood smear of mantle cell lymphoma showing a spectrum of lymphoma cells. Reproduced with permission from Weisenburger DD. Mantle cell lymphoma. In: Knowles D, ed. Neoplastic hematopathology. Baltimore: Williams & Wilkins, 1992: 617-28.
Image Not Available
Figure 4.8 Spleen in mantle cell lymphoma showing an expanded white pulp area with a small germinal center surrounded by a wide mantle of small lymphoid cells. Note that the marginal zone is preserved. Reproduced with permission from Weisenburger DD. Mantle cell lymphoma. In: Knowles D, ed. Neoplastic hematopathology. Baltimore: Williams & Wilkins, 1992: 617-28.
Other organs Involvement of other organs by MCL is not uncommon because the patients usually have advanced stage disease at the time of diagnosis. Liver involvement is common and is characterized by atypical portal lymphoid infiltrates. Involvement of other extranodal sites is also common. Extranodal sites that are most likely to be involved primarily or as part of a disseminated process include the gastrointestinal tract and Waldeyer's ring (20-30 per cent of cases). Gastrointestinal involvement as multiple lymphomatous polyposis (Fig. 4.9), often
Image Not Available
Figure 4.9 Multiple lymphomatous polyposis of the intestine with myriads of confluent polyps within the ileum (top) and scattered, more sessile polyps within the colon (bottom). Reproduced with permission from Moynihan MJ, Bast MA, Chan WC, et al. Lymphomatous polyposis: a neoplasm of either follicular mantle or germinal center cell origin. Am J Surg Pathol 7996; 20: 442-52.
Cytogenetic and molecular genetic features 31
accompanied by a large localized mass, has been reported23'24 but is not entirely specific for MCL.25
rates in the lymphocytic forms and high rates in the biastic variants, but with considerable overlap.15'19'28'29
IMMUNOLOGIC FEATURES
CYTOGENETIC AND MOLECULAR GENETIC FEATURES
The immunohistologic features of MCL reveal a characteristic phenotype.13-15,19,26,27 In frozen sections, the cells have a monoclonal B cell phenotype, almost always bearing surface IgM and often slgD. Surface IgG is expressed along with slgM in about 20 per cent of cases. The kappa to lambda light-chain ratio is reversed in MCL with about 60 per cent of cases expressing monoclonal lambda light chains, and the residual germinal centers are polyclonal. The neoplastic cells also stain for a variety of pan-B cell antigens (CD 19, 20, 22 and 24) and HLA-DR antigen. Interestingly, the cells usually have the pan-T cell antigen CD 5 on the surface (Fig. 4.10) and are negative for CD 10 (CALLA) antigen. The neoplastic cell$ may also bear the T cell-associated antigens CD 43 and Leu 8, but fail to stain for other pan-T cell antigens. The cells are usually also negative for CD 23 antigen. Antibodies to dendritic reticulum cells reveal large aggregates of these cells in cases with a nodular or mantle-zone pattern, whereas a more sparse and irregular meshwork of dendritic cells is usually found in diffuse areas. Cases of blastic MCL are less likely to express slgD, CD 5 and CD 43, and may express CD 10 antigen. The phenotype of MCL is remarkably similar to that of small lymphocytic lymphoma and chronic lymphocytic leukemia, except for more intense slg and CD 20 staining and lack of CD 23 expression in MCL. Studies of cellular proliferation rates in MCL have generally found low
The characteristic cytogenetic abnormality in MCL is the t(ll;14)(ql3;q32), which is seen in the majority of cases.3M3 Variant translocations involving the Ilql3 breakpoint have also been reported,33,34 whereas the presence of trisomy 12 appears to be a secondary abnormality.35 The presence of a complex karyotype with hyperdiploidy has been associated with large atypical cells36 and may suggest a more aggressive clinical course in MCL. However, the t(ll;14)(ql3;q32) also occurs, albeit infrequently, in other types of non-Hodgkin's lymphoma, lymphocytic leukemia and multiple myeloma. Therefore, cytogenetic findings need to be carefully correlated with the pathologic and immunologic features to confirm a diagnosis of MCL. The molecular counterpart of the t(ll;14) involves an error in V-D-J joining during Ig heavy-chain gene rearrangement, resulting in the movement of a putative cellular oncogene adjacent to the bd-1 (Ilql3) breakpoint into proximity of the enhancer region of the Ig heavy chain gene (14q32). Breaks in the latter region are thought to occur during early B cell development and are mediated by the recombinase system, whereas 1 Iql3 appears to be a common fragile site.37 The breakpoints in the bd-1 locus are not tightly clustered, although 30-40 per cent of cases of MCL have breaks in the major translocation cluster (MTC) region (Fig. 4.II).38"42 However, using multiple probes including those for a number of minor breakpoint regions, a variety of investigators have detected clonal rearrangements in 50-70
Figure 4.11 Diagram of the t(11; 14) in mantle cell lymphoma. An 1lql3 breakpoint at the bcl-1 major translocation cluster is shown; other described breakpoints (arrows in upper panel) exist as close as 1 kbfrom the first exon of the PRAD1 gene. Figure 4.10 Mantle cell lymphoma stained with CD 5 antibody.
Reproduced with permission from Arnold A. The cyclin
Note that the tumor cells do not stain as intensely as the few admixed T cells.
43: 543-9.
D1/PRAD1 oncogene in human neoplasia. ] Invest Med 7995;
32 Mantle cell lymphoma
per cent of patients with MCL.38"42 A polymerase chain reaction (PCR) assay has been developed and used to detect most of the breaks in the MTC region.9'43"45 This assay may also be used on DNA extracted from paraffinembedded tissues. The use of chromosome 11 paints and fluorescence in situ hybridization (FISH) to detect the t(ll;14) in interphase cells of MCL has also been reported.46'47 Both the PCR and FISH techniques will be useful to detect minimal residual disease in MCL. The putative oncogene deregulated by the t(ll;14) was identified by two groups and is located approximately 120 kb telomeric from the MTC breakpoint (Fig. 4.II). 48,49 The gene was named PRAD1 because of its original recognition in parathyroid adenoma50 but has been officially named CCND1. The gene encodes for cyclin Dl and is overexpressed in nearly all cases of MCL, whereas it is expressed only rarely in other forms of hematopoietic cancer.39,41,42,48,51,52 Since overexpression of this gene at the RNA level has also been noted in most cases of MCL without detectable bd-1 rearrangements, additional minor breakpoint sites outside of those detected by the available probes are likely to be involved in the translocation of chromosome Ilql3. Alternatively, deregulation could occur by mutation or deletion of negative regulator elements adjacent to the gene, or extra copies of the gene may result in its overexpression. All of the known breakpoints leave the CCND1 coding region structurally intact and result in increased protein expression. In some cases, loss of 3' end regulatory sequences may also increase the half-life of cyclin Dl.53 Antibodies to cyclin Dl, which work on paraffin-embedded material (Fig. 4.12), have been shown to be highly sensitive and specific for MCL54-59 and are very useful diagnostically. Alternatively, in situ hybridization detection of cyclin Dl mRNA in cytopreps or paraffin tissues can be used to elucidate the nature of diagnostically difficult cases.60,61 The mechanism by which cyclin Dl overexpression facilitates lymphomagenesis is not yet well understood,
but its key role in cell cycle regulation and the progression of cells through the main commitment checkpoint in Gl to S phase is certainly important.62"64 Overexpression of cyclin Dl results in a shortened Gl phase, probably through its physical interaction with the tumor-suppressor retinoblastoma protein (RB).65 Cyclin Dl binds to and activates important enzymes called cyclin-dependent kinases (CDK; mainly CDK4 and CDK6), whose activity is needed to propel cells through the Gl checkpoint. Cyclin D1-CDK4 complexes then bind to and hyperphosphorylate RB, which in turn prevents RB from binding important transcription factors such as E2F. Thus, the growth-restraint effect of RB via its binding of transcription factors is removed and the cells are propelled into S phase (Fig. 4.13). Mutations of the tumor-suppressor p53 gene have been reported in aggressive variants of MCL.66'67 This gene regulates the expression of p21 protein, which is an important universal inhibitor of cyclin-CDK complexes, including cyclin D-CDK4.68 Thus, mutations of the p53 gene with loss of this inhibition could further enhance the effects of overexpressed cyclin Dl. The normal p53 gene acts as a molecular monitor of the genome.69 If DNA is damaged, p21 protein accumulates and switches off replication to allow extra time for DNA repair. If the repair fails, p53 may trigger cell suicide by apoptosis. However, tumor cells in which p53 is inactivated by mutation cannot carry out this arrest and are genetically unstable. Such cells will accumulate mutations and chromosomal rearrangements at an increased rate, thus leading to the rapid selection of highly malignant clones. Thus, p53 also plays a critical role at the Gl checkpoint62'68'69 and mutations of p53 are important in the progression of MCL.66'67 Evidence has accumulated that CCND1 can function like an oncogene by causing abnormalities of cellular growth control, cell cycle progression and gene expression, as well as malignant transformation.70"72 Studies
Image Not Available
Figure 4.12 Mantle cell lymphoma stained with a polyclonal antibody to cyclin Dl. Note the nuclear positivity of variable intensity in a majority of the cells.
Figure 4.13 Schematic diagram showing the functional interrelationships of cyclin Dl. Reproduced with permission from Arnold A. The cyclin D1/PRAD1 oncogene in human neoplasia. J Invest Med 7995; 43: 543-9.
Differential diagnosis 33
using transgenic mice have shown that CCND1 cooperates with myc genes in the generation of B cell lymphomas, although CCND1 was not oncogenic by itself.73,74 However, these studies demonstrated subtle alterations in cell cycle progression and the number of bone marrow B cells due to CCND1 overexpression.73'74 Activation of cyclin D genes by proviral insertions in murine lymphomas has also been reported.75,76 Also, cyclin Dl was shown to induce mammary hyperplasia and carcinoma in a different transgenic model.77 Thus, it appears that CCND1 is a bona fide oncogene whose activity appears to depend on the specific cell type as well as specific co-operating partner genes72-74,78 in order to induce tumors. Further elucidation and study of such gene interactions is needed in MCL.
NORMAL CELLULAR COUNTERPART Currently, the various types of B cell neoplasia are thought to represent cells arrested at various stages in the normal differentiation scheme.79 The histologic and immunologic features of MCL suggest that the neoplastic cells correspond to normal, naive B lymphocytes that home to and reside in primary lymphoid follicles and the mantle zones of secondary follicles. These cells seem to correspond phenotypically to a major population of fetal B cells that leave the bone marrow and form the primary lymphoid follicles in the lymph nodes and spleen.80"82 At birth, 68 per cent of cord blood B cells and approximately half of peripheral blood B cells are CD 5positive83'84 and these cells are morphologically similar to the cells of MCL.85 In the adult, CD 5-positive B cells circulate in small numbers8"18 and are found in the inner area of mantle zones of lymphoid follicles.86'89 Normal mantle cells have been shown to express a diverse repertoire of unmutated Ig heavy-chain variable region genes, as one would expect of naive pregerminal center B cells.90 Identical findings have also been reported in MCL.91 Furthermore, CD 5-positive B cells can be induced to differentiate to CD 5-negative cells with the immunologic features of germinal center cells.92 Thus, the cells of MCL appear to correspond to precursor cells of the
normal germinal center reaction.79,93 The possibility that some cases of MCL arise from a subset of mantle zone lymphocytes with the immunologic features of marginal zone cells94,95 has been suggested by some authors.96,97
DIFFERENTIAL DIAGNOSIS A diagnosis of MCL with a mantle zone pattern may be difficult to make when incomplete obliteration of the normal lymph node architecture and numerous benignappearing germinal centers are present. However, the presence of wide follicular mantles in mantle zone lymphoma is distinctly different from the thin mantles found in most cases of reactive follicular hyperplasia. In the spleen, involvement of the red pulp by lymphoma is also a helpful diagnostic feature. However, in lymph nodes, mantle zone hyperplasia and angiofollicular lymphoid hyperplasia of the hyaline-vascular type (Castleman's disease) are two uncommon reactive processes that may be difficult to distinguish from mantle zone lymphoma. Mantle zone hyperplasia usually occurs as an isolated small node in the neck of a young individual. In mantle zone hyperplasia, the follicles are usually localized to the cortex of the node, and the architectural effacement and diffuse areas of involvement characteristic of lymphoma are lacking. Angiofollicular lymphoid hyperplasia also usually presents in young individuals, but as a large and localized mass, and is characterized by typical hyaline-vascular germinal centers without diffuse areas of involvement. Immunohistochemical stains may be very helpful in distinguishing mantle zone lymphoma from these reactive processes. The monoclonality, and CD 5 and CD 43 positivity, of the neoplastic cells clearly separate mantle zone lymphoma from the follicular, mantle zone and angiofollicular hyperplasias. This rule is also valid for separating diffuse MCL from diffuse reactive lymphoid proliferations of B cell type. A variety of non-Hodgkin's lymphomas may also be confused with the lymphomas of mantle cell origin. However, immunohistochemical and cytogenetic or molecular studies may be very useful in differentiating the various entities (Table 4.1). The nodular (primary
Table 4.1 Phenotypes of various B lymphocytic lymphomas
+
Mantle cell lymphoma Follicular center cell lymphoma Small lymphocytic lymphoma/CLL Monocytoid B cell lymphoma Mucosa-associated lymphoma
M±D G±M M±D M M
-/+ -/+ -/+
Lymphoplasmacytic lymphoma
M
+
+
+ -
+/+ -/+
+ + -/+
+ +
t(11;14)(q13;q32) t(14;18)(q32;q21) Trisomy 12, del 13q14 Trisomy 3 and 18 t(11;18)(q21;q21) Trisomy 3 and 18 t(9;14)(p13;q32), de!6q23
slg = surface immunoglobulin, clg = cytoplasmic immunoglobulin, DRC = dendritic reticulum cell network, CLL = chronic lymphocytic leukemia, + = >80 per cent positive, +/- = >50 per cent positive, -/+ = <50 per cent positive, - = <20 per cent positive.
34 Mantle cell lymphoma
follicular) form of MCL may be difficult to differentiate from follicular small cleaved cell lymphoma. However, the cells of MCL are usually not as markedly angulated and cleaved as those of follicular center cell lymphoma. Also, large transformed cells are usually absent in MCL, although residual large cells from invaded benign germinal centers may occasionally confuse the issue. Harris and Bhan98 have described a rare form of follicular center cell lymphoma in which small cleaved cells exit the neoplastic germinal centers and accumulate in the adjacent mantle zones. However, such cases should not be considered as mantle zone lymphoma since they arise from germinal center cells. In difficult cases, immunohistochemical stains can be used to separate follicular center cell lymphoma from MCL by the fact that the former has monoclonal, CD 10-positive germinal centers and CD 5-negative mantle zones, whereas polyclonal germinal centers and monoclonal, CD 5-positive mantle zones are seen in nodular MCL. Immunologic studies may also be helpful in separating diffuse MCL from diffuse small cleaved follicular center cell lymphoma, which is CD 5-negative and CD 10-positive, and often exhibits immunoglobulin heavy-chain switching to a more mature phenotype." Also, transformed large cells, and small cells with the markedly elongated and twisted nuclei of follicular center cell lymphoma, are generally absent in diffuse MCL. Interfollicular small lymphocytic lymphomas may encroach upon and invade reactive follicles and produce a pseudo-mantle zone pattern.100 This pattern is characterized by reactive germinal centers with thin, residual mantle zones that are surrounded by the neoplastic infiltrate. However, the predominance of small lymphocytes with uniformly round nuclei and the presence of pseudofollicular proliferation centers and paraimmunoblasts in small lymphocytic lymphoma are useful differential features, since they do not occur in MCL. Perry and associates101 have also shown that lymphocytic lymphomas composed of cells with irregular nuclei, but having pseudofollicular proliferation centers, should be classified as small lymphocytic lymphoma for clinical purposes. The immunophenotypes of MCL and small lymphocytic lymphoma are similar, but the presence of numerous dendritic reticulum cells and the absence of CD 23 antigen in MCL are useful diagnostic features. A pseudo-mantle zone pattern may also be seen in monocytoid B cell lymphoma, centrocyte-like B cell lymphoma occurring in mucosa-associated lymphoid tissues, and peripheral T cell lymphoma composed of atypical small lymphoid cells. These lymphomas arise in the parafollicular or interfollicular regions, and may secondarily invade reactive lymphoid follicles. Each of these lymphomas has distinctive histologic and immunologic features that are useful in the differential diagnosis (Table 4.1). However, the presence of lymphoepithelial lesions is not useful in differentiating MCL from centrocyte-like lymphoma of mucosa-associated lymphoid
tissue, since they have also been described in MCL.24,25,102 Pileri and associates103 have described cases of apparent 'mantle zone' lymphoma of the lymphocytic type with plasma cell differentiation. However, it is not clear whether these authors are describing interfollicular lymphoplasmacytoid/cytic lymphomas with a pseudo-mantle zone pattern or a rare form of MCL with plasma cell differentiation. In the spleen, MCL must also be distinguished from splenic marginal zone lymphoma. The presence of predominantly white pulp involvement with a prominent follicular dendritic network, cellular monotony with the absence of large transformed cells (centroblasts) or clonal plasma cells, and CD 5 positivity favor a diagnosis of MCL.104 The blastic variants of MCL may sometimes be confused with B lymphoblastic lymphoma105 or granulocytic sarcoma, although the chromatin pattern is usually somewhat more coarse in blastic MCL. Immunologic studies are usually helpful in this regard, since blastic MCL is surface Ig- and CD 5-positive, and terminal deoxynucleotidyl transferase (tdt)-negative, whereas B lymphoblastic lymphoma is CD 5-negative, usually surface Ig-negative, and tdt-positive. The presence of CD 10 is not helpful since it may be expressed in either entity. In addition to the above features, a number of myeloid markers including myeloperoxidase, lysozyme and specific esterase, will clearly delineate granulocytic sarcoma from MCL. Although MCL may be diagnosed in extranodal sites, such as the gastrointestinal tract, one should hesitate to make a primary diagnosis of MCL in extranodal sites, such as the bone marrow, liver or soft tissue, because of the nuclear irregularities that may occur as a result of the surrounding fibrous tissue reaction. Such cases are better diagnosed as lymphocytic lymphoma, not further classified, if corroborating evidence for MCL cannot be obtained. Similarly, one should not make a diagnosis of MCL based on bone marrow or peripheral blood smears alone, since criteria for such a diagnosis have not been well defined. A lymph node biopsy with immunologic studies is often necessary to categorize such cases precisely, although flow cytometric and molecular studies of blood and bone marrow may also be useful.21,22,27
CLINICAL FEATURES Mantle cell lymphoma comprises 2.5-4.0 per cent of all non-Hodgkin's lymphomas in the USA, whereas higher rates of 7-9 per cent are found in Europe. A number of detailed studies have defined the clinical features of MCL (Table 4.2).13-15,17-19,106-12 Patients with MCL have a median age of approximately 60 years and males predominate. Patients generally present with advanced (stage III/IV) disease, usually with generalized lymphadenopathy and bone marrow and liver involvement, but fewer than a
Treatment and survival 35
Table 4.2 Clinical features of mantle cell lymphoma at initial presentation Median age Male to female ratio
60 years 4:1
Generalized lymphadenopathy Splenomegaly Hepatomegaly Peripheral blood lymphocytosis Bone marrow infiltration Gastrointestinal involvement Waldeyer's ring involvement
90% 60% 30% 30% 80% 20% 10%
Ann Arbor Stage IM/IV B symptoms Bulky disease Poor performance status Elevated lactate dehydrogenase Elevated (32 microglobulin
90% 40% 30% 20% 40% 55%
Image Not Available
Figure 4.14 Overall survival of patients with mantle cell lymphoma (MCL) compared with those having Working Formulation (WF) types A through E. Reproduced with permission from Fisher Rl, DahlbergS, Nathwani BN, Banks PM, Miller TP, Grogan TM. A clinical analysis of two indolent lymphoma entities: mantle cell lymphoma and marginal zone lymphomas, including mucosa-associated lymphoid tissue and monocytoid B-
half of the patients have systemic (B) symptoms. Splenomegaly is present at initial diagnosis in approximately 60 per cent of the patients. In the nodular (mantle zone) type, 80 per cent of the patients have splenomegaly, which may be massive. Other extranodal sites are also frequently involved, particularly the gastrointestinal tract and Waldeyer's ring (20-30 per cent of cases). A particularly striking extranodal presentation is multiple lymphomatous polyposis of the intestine, which should suggest a diagnosis of MCL. Mild anemia is not uncommon at presentation, whereas thrombocytopenia occurs in fewer than 15 per cent of the patients. A peripheral blood lymphocytosis of greater than 4000/mm3 occurs in 20-40 per cent of the cases, but absolute counts above 20000/mm3 are uncommon. Hypogammaglobulinemia, a monoclonal gammopathy and a positive Coombs' test are also decidedly uncommon.
cell subcategories: a Southwest Oncology Group study. Blood 7995; 85:1075-82.
combination chemotherapy, which usually did not contain doxorubicin, various investigators obtained complete remission (CR) rates of 20-40 percent. 106-108,114In a prospective comparative analysis of MCL and folliclecenter lymphoma, the German Low Grade Lymphoma Study Group found that patients with MCL had more disease at presentation, a lower and slower response to such chemotherapy and a worse prognosis.115 Meusers and colleagues107 reported a CR rate of 58 per cent when using a combination of cyclophosphamide, doxorubicin, vincristine and prednisone (CHOP), but few long-term remissions or cures have been reported in any series
TREATMENT AND SURVIVAL Mantle cell lymphoma is a vexing and increasingly frequent problem for oncologists. The disease brings together the worst characteristics of high-grade and lowgrade lymphomas, i.e. the course is not indolent and the disease is rarely curable. The median survival of patients with MCL has ranged between 3 and 4 years in large series.13,15,17-19,106,107,109-111 This is significantly shorter than the survival of the patients with similar forms of lymphoma (Fig. 4.14). In two detailed studies,14,15 patients with a nodular (mantle zone) pattern had a significantly longer median survival (77-88 months) when compared to those with diffuse MCL (30-33 months) (Fig. 4.15). The prognostic significance of a predominantly nodular pattern has recently been confirmed by others.19,110,112,113 With the use of non-curative, pre-first-generation
Image Not Available
Figure 4.15 Overall survival of patients with mantle cell lymphoma having a mantle zone pattern compared with those with a diffuse pattern. Reproduced with permission from Weisenburger DD, Duggan MJ, Perry DA, Sanger WG, ArmitageJO. Non-Hodgkin's lymphomas of mantle zone origin. In: Rosen PP, Fechner RE, eds Pathology annual. East Norwalk, CT: Appleton & Lange, 7990:739-58.
36 Mantle cell lymphoma
because of disease recurrence and progression. Zucca and associates111 have reported a benefit from aggressive chemotherapy in a subset of patients with good prognostic indicators. Teodorovic and colleagues114 recently reported a CR rate of 52 per cent for patients treated with aggressive chemotherapy and suggested that improved survival may be achieved if a CR after CHOPlike aggressive chemotherapy, also containing bleomycin, is obtained. In elderly patients, less aggressive therapy, such as cyclophosphamide, vincristine and prednisone (CVP), may be justified. Therapeutic experience with the purine analogs, fludarabine and 2-CDA, and interferon has been disappointing.12'114 Although a small proportion of patients may benefit from observation only, Bookman and colleagues108 found that CRs could not be obtained when such patients were later treated for progressive symptomatic disease. Most studies14,15,20,106-108,111,114 have shown, however, that patients who achieve a CR have a longer survival than those who do not achieve a CR, but few patients are cured. Since the long-term prognosis of patients receiving conventional therapy for MCL is rather poor, the use of aggressive combination chemotherapy with stem-cell transplantation for younger patients has been suggested.17'107'108'116 Stewart and colleagues117 treated nine such patients with high-dose therapy and stem-cell transplantation, and three were progression-free at 7, 12 and 25 months post-transplantation. However, longer follow-up of greater patient numbers will be required to determine whether high-dose therapy can overcome the chemoresistance and increase the cure rate of MCL. The optimal timing for high-dose therapy may be early as part of front-line treatment.117 Against the background of these early promising results, the European Bone Marrow Transplantation guidelines118 suggest that appropriate patients should be offered high-dose chemotherapy with autotransplant; however, Coiffier in a review of reported series found that, whilst response rates were improved, no patient was cured.119 Purging of harvested stem cells with anti-B cell antibody does not improve the situation.120 Coiffier concludes that, outside of randomized prospective studies, high-dose therapy is not appropriate for MCL patients and suggests that this (and the promising role of new agents, such as rituximab, the anti-CD 20 antibody)121 should be prospectively compared with standard CHOP therapy. A number of clinical and pathologic features are predictive of survival in MCL. The clinical features predicting a poor prognosis are generally the same as those found in other lymphoma subtypes, and include advanced age and stage, B symptoms, poor performance status, peripheral blood lymphocytosis, elevated lactate dehydrogenase or fte-microglobulin levels, high risk with the International Prognostic Index (Fig. 4.16), and failure to achieve a good clinical response to therapy.13,14,17,106,107,109,111 The pathologic features that predict a
Image Not Available
Figure 4.16 Overall survival of patients with mantle cell lymphoma grouped according to the International Prognostic Index. Reproduced with permission from Kluwer Academic Publishers; Zucca E, Roggero E, Pinotti G, et al. Patterns of survival in mantle cell lymphoma. Ann Oncol 7995; 6:259-62.
poor prognosis are a diffuse pattern, a high mitotic rate or proliferative fraction, blastic cytology and p53 overexpression 13-15,17,19,42,64,65,106,110,112,113
TUMOR GRADE The Working Formulation divides the different subtypes of non-Hodgkin's lymphoma into low-grade, intermediate-grade and high-grade categories according to their median survivals, whereas the Kiel classification utilizes only low-grade and high-grade categories based on cytologic features rather than survival. Since MCL was not well understood when the Working Formulation was prepared, it was not included as a category in the Working Formulation. However, most cases of MCL were probably included in the intermediate-grade category of diffuse small cleaved cell lymphoma in that study. In contrast, MCL is considered to be a low-grade lymphoma in the Kiel classification. Based on the studies cited herein, we believe that the predominantly nodular (mantle zone) form of MCL composed of small lymphoid cells should be considered a low-grade lymphoma (median survival, >5 years), whereas diffuse MCL composed of small cells should be considered an intermediate-grade lymphoma (median survival, 3 years). However, the lymphocytic types of MCL are similar to other lymphomas of lowgrade malignancy in that they are generally incurable, except for the uncommon case with low-stage disease. The blastic variants of MCL have cytologic features and proliferative rates similar to those of the high-grade lymphomas (small non-cleaved and lymphoblastic) in the Working Formulation and, regardless of pattern, should probably be considered as such (median survival, <2 years), and this is recognized in the proposed WHO classification (see Chapter 1). De novo cases of blastic MCL may be the most responsive to high-dose therapy and should probably be treated with curative intent.
References 37
Future classifications of non-Hodgkin's lymphoma should include MCL, with all of its various patterns and cytologies, so that future clinical studies can clearly delineate and further characterize this important entity.
CONCLUSION
4. Namba K, Jaffe ES, Braylan RC, Soban EJ, Berard CW. Alkaline phosphatase-positive malignant lymphoma. A subtype of B-cell lymphomas. AmJ Clin Pathol 1977; 68: 535-42. 5. Lennert K, Stein H, Kaiserling E. Cytological and functional criteria for the classification of malignant lymphomata. BrJ Cancer 1975; 31 (suppl II): 29-43. 6. Lennert K. Malignant lymphomas other than Hodgkin's
Numerous recent studies have confirmed that MCL is a distinct clinicopathologic entity. The neoplastic cells of MCL appear to correspond to naive B cells that normally home to and reside in primary lymphoid follicles and the mantle zones of secondary follicles. As such, they correspond to a subset of normal follicular B cells, which are thought to transform into germinal center cells in response to antigen. The relationship between the nodular (mantle zone) and diffuse lymphocytic forms of MCL is biologically analogous to the germinal center cell lymphomas of follicular and diffuse types, respectively, whereas the blastic forms of MCL are analogous to the transformed lymphomas arising in other low-grade lymphomas. New and better therapies are badly needed for this group of lymphomas,122 including further exploration of high-dose therapy with stem-cell rescue,117 immunotherapy with anti-shared idiotype monoclonal antibodies,123 radioimmunotherapy with 13T-labelled B cell-specific antibodies,124 and innovative approaches, which take advantage of cell-cycle checkpoints and proliferation controls.125'126 Until such progress is made, MCL will continue to be one of the worst forms of nonHodgkin's lymphoma, a clinically aggressive disease with little hope of a cure.
ACKNOWLEDGEMENT
disease. Berlin: Springer-Verlag, 1978: 284-302. 7. Tolksdorf G, Stein H, Lennert K. Morphological and immunological definition of a malignant lymphoma derived from germinal-center cells with cleaved nuclei (centrocytes). BrJ Cancer 1980; 41:168-82. 8. Weisenburger DD, Kim H, Rappaport H. Mantle-zone lymphoma. A follicular variant of intermediate lymphocytic lymphoma. Cancer 1982; 49:1429-38. 9. Palutke M, Eisenberg L, Mirchandani I, Tabaczka P, Husain M. Malignant lymphoma of small cleaved lymphocytes of the follicular mantle zone. Blood 1982; 59: 317-22. 10. Weisenburger DD, Nathwani BN, Diamond LW, Winberg CD, Rappaport H. Malignant lymphoma, intermediate lymphocytic type. A clinicopathologic study of 42 cases. Cancer 1981; 48:1415-25. 11. Banks PM, Chan J, Cleary ML, etal. Mantle cell lymphoma: a proposal for unification of morphologic, immunologic, and molecular data. Am J Surg Pathol 1992; 16: 637-40. 12. Zucca E, Stein H, Coffier B. European Lymphoma Task Force (ELTF). Report of the Workshop on mantle cell lymphoma (MCI). Ann Oncol 1994; 5: 507-11. 13. SwerdlowSH, HabeshawJA, Murray LJ, Dhaliwal HS, Lister TA, Stansfeld AG. Centrocytic lymphoma. A distinct clinicopathologic and immunologic entity. A multiparameter study of 18 cases at diagnosis and relapse. AmJ Pathol 1983; 113:181-97. 14. Weisenburger DD, Duggan MJ, Perry DA, Sanger WG, ArmitageJO. Non-Hodgkin's lymphomas of mantle-zone
This work was supported in part by USPHS CA36727 awarded by the National Cancer Institute, Department of Health and Human Services, and is reproduced in part from a review by the authors (Blood; 87: 4483-94).
REFERENCES
origin. In: Rosen PP, Fechner RE, eds Pathology annual. East Norwalk, CT: Appleton and Lange, 1990:139-58. 15. Lardelli P, Bookman MA, Sundeen J, Longo DL, Jaffe ES. Lymphocytic lymphoma of intermediate differentiation. Morphologic and immunophenotypic spectrum and clinical correlations. Am J Surg Pathol 1990; 14: 752-63. 16. Lennert K, Feller AC. Histopathology of non-Hodgkin's lymphomas. Berlin: Springer-Verlag, 1992: 93-102. 17. Norton AJ, Mathews J, Pappa V, etal. Mantle cell
1. Berard CW, Dorfman RF. Histopathology of malignant lymphomas. Clinics Haematol 1974; 3: 39-76. 2. Berard CW. Reticuloendothelial system: an overview of neoplasia. In: RebuckJW, Berard CW, Abell MR, eds The reticuloendothelial system. International Academy of Pathology Monograph no. 16. Baltimore: Williams & Wilkins, 1975: 301-17. 3. Jaffe ES, Braylan RC, Namba K, Frank MM, Berard CW.
lymphoma: natural history defined in a serially biopsied population over a 20-year period. Ann Oncol 1995; 6: 249-56. 18. Pittaluga S, Wlodarska I, Stul MS, et al. Mantle cell lymphoma: a clinicopathologic study of 55 cases. Histopathology 1995; 26:17-24. 19. Segal GH, Masih AS, Fox AC, Jorgensen T, Scott M, Braylan RC. CDS-expressing B-cell non-Hodgkin's
Functional markers: a new perspective on malignant
lymphomas with bcl-1 gene rearrangement have a
lymphomas. Cancer Treat Rep 1977; 61:1953-62.
relatively homogeneous immunophenotype and are
38 Mantle cell lymphoma
20.
21.
22.
23.
24.
25.
26.
27.
28.
29.
30.
31.
32.
associated with an overall poor prognosis. Blood 1995; 85: 1570-9. Duggan MJ, Weisenburger DD, Ye YL, et al. Mantle zone lymphoma. A dinicopathologic study of 22 cases. Cancer 1990; 66: 522-9. Pombo de Oliveira MS, Jaffe ES, Catovsky D. Leukemic phase of mantle zone (intermediate) lymphoma: its characterization in 11 cases. J Clin Pathol 1989; 42: 962-72. Criel A, Billiet J, Vandenberghe E, et al. Leukemia intermediate lymphocytic lymphomas: analysis of twelve cases diagnosed by morphology. Leuk Lymph 1992; 8: 381-7. O'Brian DS, Kennedy MJ, Daley PA, etal. Multiple lymphomatous polyposis of the gastrointestinal tract. A clinicopathologically distinctive form of non-Hodgkin's lymphoma of B-cell centrocytic type. AmJ Surg Pathol 1989; 13: 691-9. Lavergne A, Brouland J, Launay E, Nemeth J, RuskoneFourmestraux A, Galian A. Multiple lymphomatous polyposis of the gastrointestinal tract. An extensive histopathologic and immunohistochemical study of 12 cases. Cancer 1994; 74: 3042-50. Moynihan MJ, Bast MA, Chan WC, et al. Lymphomatous polyposis: a neoplasm of either follicular mantle or germinal center cell origin. Am J Surg Pathol 1996; 20: 442-52. Zuckerberg LR, MedeirosJL, Ferry JA, Harris NL. Diffuse low-grade B-cell lymphomas. Four clinically distinct subtypes defined by a combination of morphologic and immunophenotypic features. AmJ Clin Pathol 1993; 100: 373-85. Molot RJ, Meeker TC, Wittwar CT, et al. Antigen expression and polymerase chain reaction amplification in mantle cell lymphomas. Blood 1994; 83:1626-31. Ott MM, Ott G, Kuse P, et al. The anaplastic variant of centrocytic lymphoma is marked by frequent rearrangements of the bcl-1 gene and high proliferation indices. Histopathology 1994; 24: 329-34. Czader M, Porwit A, Tahi E, Ost A, Mazur J, Auer G. DMA image cytometry and the expression of proliferative markers (proliferating cell nuclear antigen and Ki-67) in non-Hodgkin's lymphomas. Modern Pathol 1995; 8: 51-8. Weisenburger DD, Sanger WG, Armitage JO, Purtilo DT. Intermediate lymphocytic lymphoma: immunophenotypic and cytogenetic findings. Blood 1987;69:1617-21. Rimokh R, Berger F, Cornillet P, etal. Break in the bcl-1 locus is closely associated with intermediate lymphocytic lymphoma subtype. Genes Chromosomes Cancer 1990; 2: 223-6. Leroux D, Le Marc'hadour F, Gressin R, etal. NonHodgkin's lymphomas with t(11;14)(q13;q32): a subset of mantle zone/intermediate lymphocytic lymphomas. BrJ Haematol 1991; 77: 346-53.
33. Vandeberghe E, de Wolfe-Peeters C, Wlodarska I, et al. Chromosome 11q rearrangements in B non-Hodgkin's lymphoma. BrJ Haematol 1992; 81: 212-17. 34. Komatsu H, lida S, Yamamoto K, et al. A variant chromosome translocation at11q13 identifying PRADI/Cyclin D1 as the bcl-1 gene. Blood 1994; 84: 1226-31. 35. Neilson JR, Cai M, Bienz N, Leyland MJ. Leukemic mantle cell lymphoma with t(11;14) and trisomy 12 showing clinical features of stage AO B cell chronic lymphocytic leukemia. J Clin Pathol Molec Pathol 1995; 48: M165-6. 36. Daniel MT, Tagaud I, Flexor MA, Nogueira ME, Berger R, Jonveaux P. Leukemic non-Hodgkin's lymphomas with hyperdiploid cells and t(11;14)(q13;q32): a subtype of mantle cell lymphoma? BrJ Haematol 1995; 90: 77-84. 37. Chary-Reddy S, Prasad VS, Ahuja YR. Expression of common fragile sites in untreated non-Hodgkin's lymphoma with aphidicolin and folate deficiency. Cancer Lett 1994; 86:111-17. 38. Williams ME, Swerdlow SH, Rosenberg CL, Arnold A. Chromosome 11 translocation breakpoint at the PRADI/Cyclin D1 gene locus in centrocytic lymphoma. Leukemia 1993; 7: 241-5. 39. Rimokh R, Berger F, Delsol G, et al. Rearrangement and overexpression of the bcl-1/PRAD-1 gene in intermediate lymphocytic lymphomas and in t(11 q13)bearing leukemias. Blood 1993; 81: 3063-7. 40. de Boer CJ, Loyson S, Kluin PM, Kluin-Nelemans HC, Schuuring E, van Krieken JHJM. Multiple breakpoints within the Bcl-1 locus in B-cell lymphoma: rearrangements of the cyclin D1 gene. Cancer Res 1993; 53: 4148-52. 41. Hayashi T, Ohno H, Yamabe H, etal. Clinical aspects of B-cell malignancy involving the BCL-1/PRAD1 locus. IntJ Hematol 1995; 59: 281-96. 42. Bosch F, Jares P, Campo E, etal. PRADI/Cyclin D1 gene overexpression in chronic lymphoproliferative disorders: a highly specific marker of mantle cell lymphoma. Blood 1994; 84: 2726-32. 43. Williams ME, Swerdlow SH, Meeker TC. Chromosome t(11;14)(q13;q32) breakpoints in centrocytic lymphoma are highly localized at the bcl-1 major translocation cluster. Leukemia 1993; 7:1437-40. 44. Rimokh R, Berger F, Delsol G, etal. Detection of the chromosomal translocation in t(11;14) by polymerase chain reaction in mantle cell lymphomas. Blood 1994; 83: 1871-5. 45. Luthra R, Hai S, Pugh WC. Polymerase chain reaction detection of the t(11;14) translocation involving the bcl1 major translocation cluster in mantle cell lymphoma. Diagnos Molec Pathol 1995; 4: 4-7. 46. Zucca E, Soldati G, Schlegelberger B, et al. Detection of chromosome 11 alterations in blood and bone marrow by interphase cytogenetics in mantle cell lymphoma. Br yWomwto/1995; 89: 665-8.
References 39 47. Lionel JA, Schuuring E, Kibbelaar RE, et al. Detection of 11q13 rearrangements in hematologic neoplasias by double-color fluorescence in situ hybridization. Blood 1996;87:1512-19. 48. Rosenberg CL, Wong E, Petty Em, et al. PRAD1, a candidate BCL1 oncogene: mapping and expression in centrocytic lymphoma. Proc NatAcadSci USA 1991; 88: 9636-42. 49. Withers DA, Harvey RC, Faust JB, Melnyk 0, Carey K, Meeker TC. Characterization of a candidate bcl-1 gene. Molec Cell Biol 1991; 11: 4846-53. 50. Motokura T, Bloom T, Kim HG, et al. A novel cyclin encoded by a bcl 1-linked candidate oncogene. Nature 1991;350:512-15. 51. Oka K, Ohno T, Kita K, et al. PRAD1 gene overexpression in mantle-cell lymphoma but not in other low-grade Bcell lymphomas, including extranodal lymphoma. BrJ Haematol 1994; 86: 786-91. 52. de Boer CJ, van Krieken JHJM, Kluin-Nelemans HC, Kluin PM, Schuuring E. (1995). Cyclin D1 messenger RNA overexpression as a marker for mantle cell lymphoma. Oncogene 1995; 10:1833-40. 53. Rimokh R, Berger F, Bastard C, et al. Rearrangement of CCND1 (BCL1/PRAD1) 3' untranslated region in mantlecell lymphomas and thet(11q13)-associated leukemias. Blood 1994; 10: 3689-96. 54. Yang Wl, Zukerberg LR, Motokura T, Arnold A, Harris NL. Cyclin D1 (bcl-1, PRAD1) protein expression in low-grade B-cell lymphomas and reactive hyperplasia. AmJ Pathol 1994:145:86-96. 55. Zuckerberg LR, Yang Wl, Arnold A, Harris NL Cyclin D1 expression in non-Hodgkin's lymphomas. Detection by immunohistochemistry. AmJ Clin Pathol 1995; 103: 756-60. 56. Swerdlow SH, Yang Wl, Zuckerberg LR, Harris NL, Arnold A, Williams ME. Expression of cyclin D1 protein in centrocytic/mantle cell lymphomas with and without rearrangement of the BCL1/cyclin D1 gene. Human Pathol 1995; 26: 999-1004. 57. Nakamura S, Seto M, Banno S, et al. Immunohistochemical analysis of cyclin D1 protein in hematopoietic neoplasms with special reference to mantle cell lymphomas. Japj Cancer Res 1994; 85: 1270-9. 58. Kuroda H, Komatsu H, Nakamura S, etal. The positive nuclear staining observed with monoclonal antibody against PRADI/cyclin D1 correlates with mRNA expression in mantle cell lymphoma. Jap J Cancer Res 1995:86:890-8. 59. de Boer CJ, Schuuring E, Dreef E, etal. Cyclin D1 protein analysis in the diagnosis of mantle cell lymphoma. Blood 1995; 86: 2715-23. 60. Delmer A, Ajchenbaum-Cymbalista F, Tang R, et al. Over-expression of cyclin D1 in chronic B-cell malignancies with abnormality of chromosome 11q13. BrJ Haematol 1995; 89: 798-804. 61. Williams ME, Nichols GE, Swerdlow SH, Stoler MH. In
62. 63.
64. 65.
66.
67.
68.
69. 70.
71.
72.
73.
74.
75.
76.
77.
situ hybridization detection of cyclin D1 mRNA in centrocytic/mantle cell lymphoma. Ann Oncol 1995; 6: 297-9. Hunter T, Pines J. Cyclins and cancer II: cyclin D and CDK inhibitors come of age. Cell 1994; 79: 573-82. Lukas J, Jadayel D, Bartikova J, et al. BCL-1/cyclin D1 oncoprotein oscillates and subverts the G1 phase control in B-cell neoplasms carrying the t(11;14) translocation. Oncogene 1994; 9: 2159-67. Arnold A. The cyclin D1/PRAD1 oncogene in human neoplasia.y Invest Med 1995; 43: 543-9. Daudy SF, Hinds PW, Louie K, Reed SI, Arnold A, Weinberg RA. Physical interactions of the retinoblastoma protein with human D cyclins. Cell 1993;73:499-511. Louie DC, Offit K, Jaslow R, et al. p53 overexpression as a marker of poor prognosis in mantle cell lymphomas with the t(11;14)(q13;q32). Blood 1995; 86: 2892-9. GreinerTC, Moynihan MJ, Chan WC, etal. p53 mutations in mantle cell lymphoma are associated with variant cytology and predict a poor prognosis. Blood 1966; 87: 4302-10. Hirama T, Koeffler HP. Role of cyclin-dependent kinase inhibitors in the development of cancer. Blood 1995; 86: 841-54. Lane DP. p53, guardian of the genome. Nature 1992; 358:15-16. Jiang W, Kahn SM, Zhou P, et al. Overexpression of cyclin-D1 in rat fibroblasts cause abnormalities in growth control, cell cycle progression and gene expression. Oncogene 1993; 8: 3447-57. Hinds PW, Daudy SF, Eaton EN, Arnold A, Weinberg RA. Function of a human cyclin gene as an oncogene. Proc NatAcadSci USA 1994; 91: 709-13. Lovec H, Sewing A, Lucibello FC, Miiller R, Mbrby T. Oncogenic activity of cyclin D1 revealed through cooperation with Ha-ras: link between cell cycle control and malignant transformation. Oncogene 1994; 9: 323-6. Bodrug SE, Warner BJ, Bath ML, Lindeman GJ, Harris AW, Adams JM. Cyclin D1 transgene impedes lymphocyte maturation and collaborates in lymphomagenesis with the myc gene. EMBOJ 1994; 13: 2124-30. Lovec H, Grzeschiezek A, Kowalski MB, Mbrb'y T. Cyclin D1/bcl-1 cooperates with myc genes in the generation of B-cell lymphoma in transgenic mice. EMBOJ 1994; 13: 3487-95. Lammie GA, Smith R, Silver J, Brookes S, Dickson C, Peters G. Proviral insertions near cyclin D1 in mouse lymphomas: a parallel for BCL1 translocations in human B-cell neoplasms. Oncogene 1992; 7: 2381-7. Hanna Z, Jankowski M, Tremblay P, et al. The vin-1 gene, identified by proviral insertional mutagenesis, is the cyclin D2. Oncogene 1993; 8:1661-6. Wang TC, Cardiff RD, Zukerberg L, Lees E, Arnold A, Schmidt EV. Mammary hyperplasia and carcinoma in MMTV-cyclin D1 transgenic mice. Nature 1994; 369: 669-71.
40 Mantle cell lymphoma 78. Afar DEH, Mclaughlin J, Sherr CJ, Witte ON, Roussel MF. Signaling by ABL oncogenes through cyclin D1. Proc Nat Acad Sci USA 1995; 92: 9540-4. 79. Weisenburger DD, Harrington DS, Armitage JO. B-cell neoplasia. A conceptual understanding based on the normal humoral immune response. In: Rosen PP, Fechner RE, eds Pathology annual. East Norwalk, CT: Appleton and Lange, 1990: 99-115. 80. Bofill M, Janossy G, Janossa M, et al. Human B cell development: II. Subpopulations in the human fetus. J Immunol 1985; 134:1531-8. 81. Antin JH, Emerson SG, Martin P, Gadol N, Ault KA. Leu-1+ (CD5+) B cells. A major lymphoid subpopulation in human fetal spleen: phenotypic and functional studies.) Immunol 1986; 136: 505-10. 82. Asano S, Akaike Y, Muramatsu T, Mochizuki M, Tsuda T, Wakesa H. Immunohistologic detection of the primary follicle (PF) in human fetal and newborn lymph node anlages. Pathol Res Practice 1993; 189: 921-7. 83. Rabian-Herzog C, Lesage S, Gluckman E. Characterization of lymphocyte subpopulations in cord blood. Bone Marrow Transplant 1992; 9 (suppl 1): 64-7. 84. Durandy A, Thuillier L, Forvielle M, Fischer A. Phenotypic and functional characteristics of human newborns' B lymphocytes. J Immunol 1990; 144: 60-5. 85. Hamburg A, Brynes RK, Reese C, Golomb HM. Human cord blood lymphocytes. Ultrastructural and immunologic surface marker characteristics; a comparison with B- and T-cell lymphomas. Lab Invest 1976; 34: 207-15. 86. Gobbi M, Caligaris-Cappio F, Janossy G. Normal equivalent cells of B cell malignancies: analysis with monoclonal antibodies. BrJHaematolWSl; 54: 393-403. 87. Gadol N, Ault KA. Phenotypic and functional characterization of Leu-1 (CDS) B cells. Immunol Rev 1986; 93: 23-34. 88. Freedman AS, Boyd AW, Bieber FR, et al. Normal cellular counterparts of B cell chronic lymphocytic leukemia. fi/oorf1987;70:418-27. 89. Abe M, Tominga K, Wakesa H. Phenotypic characterization of human B-lymphocyte subpopulations, particularly human CD5+ Blymphocyte subpopulation within the mantle zones of secondary follicles. Leukemia 1994; 8:1039-44. 90. Kiippers R, Zhoa M, Hansmann ML, Rajewsky K. Tracing B cell development in human germinal centers by molecular analysis of single cells picked from histological sections. EMBOJ 1993; 12: 4955-67. 91. Hummel M, Tamaru J, Kalvelage B, Stein H. Mantle cell (previously centrocytic) lymphomas express VH gendes with no or very little somatic mutations like the physiologic cells of the follicle mantle. Blood 1994; 84: 403-7. 92. Caligaris-Cappio F, Riva M, Tesio L, Schena M, Gaidano GL, Bergui L. Human normal CD5+ B lymphocytes can be induced to differentiate to CDS- B lymphocytes with germinal center features. Blood 1989; 73:1259-63.
93. Holder MJ, Abbot SD, Milner AE, et al. II-2 expands and maintains IgM plasmablasts from a CD5+ subset contained within the germinal centre cell-enriched (surface lgD-/CD39- buoyant) fraction of human tonsil. //7t/mmwA70/1993; 5:1059-66. 94. van den Oord JJ, de Wolf-Peeters C, Desmet VJ. The marginal zone in the human reactive lymph node. AmJ Clin Pathol 1986; 86: 475-9. 95. van Krieken JHJM, von Schilling C, Kluin PM, Lennert K. Splenic marginal zone lymphocytes and related cells in the lymph node. A morphologic and immunohistochemical study. Human Pathol 1989; 20: 320-5. 96. van den Oord JJ, de Wolf-Peeters C, Pulford KAP, Mason DY, Desmet VJ. Mantle-zone lymphoma. Immuno-and enzyme-histochemical studies on the cell of origin. AmJ Surg Pathol 1986; 10: 780-8. 97. Takeshita M, Masuda Y, Sumiyoshi Y, et al. Clinicopathologic, enzyme and histochemical studies of centrocytic (mantle cell) lymphoma: comparison with other types of low grade lymphoma based on the updated Kiel Classification. Acta Pathol Jap 1993; 43: 244-52. 98. Harris NL, Bahn AK. Mantle-zone lymphoma. A pattern produced by lymphomas of more than one cell type. AmJ Surg Pathol 1985; 9: 872-82. 99. Weisenburger DD, Linder J, Daley DT, Armitage JO. Intermediate lymphocytic lymphoma. An immunohistologic study with comparison to other lymphocytic lymphomas. Human Pathol 1987; 18: 781-90. 100. Ellison DJ, Nathwani BN, Cho SY, Martin SE. Interfollicular small lymphocytic lymphomas. The diagnostic significance of pseudofollicles. Human P0tf?o/1989;20:1108-18. 101. Perry DA, Bast MA, Armitage JO, Weisenburger DD. Diffuse intermediate lymphocytic lymphoma. A Clinicopathologic study with comparison to small lymphocytic lymphoma and diffuse small cleaved cell lymphoma. Cancer 1990; 66:1995-2000. 102. Fraga M, Leoret E, Sanchez-Verde L, et al. Mucosal mantle cell (centrocytic) lymphomas. Histopathology 1995;26:413-22. 103. Pileri S, Rivano MT, Gobbi M, Taruseio D, Lennert K. Neoplasticand reactive follicles within B-cell malignant lymphomas. A morphological and immunological study of 30 cases. Hematol Oncol 1985; 3: 243-60. 104. Pittaluga S, Verhoef G, Criel A, et al. 'Small' B-cell nonHodgkin's lymphomas with splenomegaly at presentation are either mantle cell lymphoma or marginal zone cell lymphoma. AmJ Surg Pathol 1996; 20: 211-23. 105. Cheng AL, Su IJ, Tien HF, Wang CC, Chen YC, Wang CH. Characteristic Clinicopathologic features of adult B-cell lymphoblastic lymphoma with special emphasis on differential diagnosis with an atypical form probably of blastic lymphocytic lymphoma of intermediate differentiation origin. Cancer 1994; 73: 706-10.
References 41 106. BrittingerG, Bartels H, Common H, etal. (Kiel Lymphoma Study Group). Clinical and prognostic relevance of the Kiel classification of non-Hodgkin's lymphomas. Results of a prospective multicenter study by the Kiel Lymphoma Study Group. Hematol Oncol 1984; 2: 269-306. 107. Meusers P, Engelhard M, Bartels H, etal. Multicentre randomized therapeutic trial for advanced centrocytic lymphoma. Anthracycline does not improve prognosis. Hematol Oncol 1989; 7: 365-80. 108. Bookman MA, Lardelli P, Jaffe ES, Duffey PL, Longo DL. Lymphocytic lymphoma of intermediate differentiation: morphologic, immunophenotypic, and prognostic factors. J Nat Cancer Inst 1990; 82: 742-8. 109. BergerF, Felman P, Sonet A, et al. Nonfollicular small B-cell lymphomas: a heterogeneous group of patients with distinct clinical features and outcome. Blood 1994; 83: 2829-35. 110. Fisher Rl, DahlbergS, Nathwani BN, Banks PM, Miller TP, Grogan TM. A clinical analysis of two indolent lymphoma entities: mantle cell lymphoma and marginal zone lymphoma (including mucosa-associated lymphoid tissue and monocytoid B cell categories): a Southwest Oncology Group study. Blood 1995; 85:1075-82. 111. Zucca E, Roggero E, Pinotti G, etal. Patterns of survival in mantle cell lymphoma. Ann Oncol 1995; 6: 259-62. 112. Plank L, Lennert K. Centrocytic lymphoma. Am J Surg Pathol 1993; 17: 638-9. 113. Garcia-CondeJ, Cabanillas F. Mantle cell lymphoma: a new lymphoproliferative entity with definite histopathological patterns, clinical characteristics and prognostic factors, and an investigational therapeutic approach. Ann Oncol 1995; 6: 305-6. 114. Teodorovic I, Pittaluga S, Kluin-NelemansJC, etal. Efficacy of four different regimens in 64 mantle-cell lymphoma cases: clinicopathologic comparison with 498 other non-Hodgkin's lymphoma subtypes.) C/m Oncol 1995; 13: 2819-26. 115. Hiddemann W, Unterhalt M, Herrmann R, et al. Mantlecell lymphomas have more widespread disease and a slower response to chemotherapy compared with follicle-center lymphomas: results of a prospective comparative analysis of the German Low-Grade Lymphoma Study Group. 7 Clin Oncol 1998; 16: 1922-30.
116. Haas R, Brittinger G, Meusers P, etal. Myeloablative therapy with blood stem cell transplantation is effective in mantle cell lymphoma. Leukemia 1996; 10: 1975-9. 117. Stewart DA, Vose JM, Weisenburger DD, etal. The role of high-dose therapy and autologous hematopoietic stem cell transplantation for mantle cell lymphoma. Ann Oncol 1995; 6: 263-6. 118. Goldman JM, Schmitz N, Niethammer D, Gratwohl A for the Accreditation Sub-Committee of the European Group for Blood and Marrow Transplantation. Allogeneic and autologous transplantation for haematological diseases, solid tumours and immune disorders: current practice in Europe in 1998. Bone Marrow Transplant 1998; 21:1-7. 119. Coiffier B. Which treatment for mantle-cell lymphoma patients in 1998?y Clin Oncol 1998; 16: 3-5. 120. Freedman AS, Neuberg D, Gribben JG, et al. High-dose chemoradiotherapy and anti-B-cell monoclonal antibody-purged autologous bone marrow transplantation in mantle-cell lymphoma: no evidence for long-term remission.7 Clin Oncol 1998; 16:13-18. 121. Coiffier B, Haioun C, Ketterer N, etal. Rituximab (antiCD20 monoclonal antibody) for the treatment of patients with relapsing or refractory aggressive lymphoma: a multicenter phase II study. Blood 1998; 92:1927-32. 122. Coffier B, Hiddemann W, Stein H. Mantle cell lymphoma: a therapeutic dilemma. Ann Oncol 1995; 6: 208-10. 123. Ohno T, Oka K, Yamaguchi M, Kita K, Shirakawa S. Frequent expression of shared idiotypes in mantle cell lymphoma and extranodal small lymphocytic/nonmantle cell diffuse small cleaved lymphoma. Leukemia 1995:9:1935-9. 124. Kaminski MS, Zasadny KR, Francis IR, et al. Radioimmunotherapy of B-cell lymphoma with [131I] anti-B1 (anti-CD20) antibody. N EnglJ Med 1993; 329: 459-65. 125. Hartwell LH, Kastan MB. Cell cycle control and cancer. Science 1994; 266:1821-8. 126. Peng B, Mehta NH, Fernandez H, Chan CC, Ravechi E. Growth inhibition of malignant CD5+ B(B1) cells by antisense IL-10 oligonucleotide. Leukemia Res 1995; 19: 159-67.
This page intentionally left blank
5 Diffuse indolent B cell neoplasms KAMACLENNAN
Introduction Small lymphocytic lymphoma/B cell chronic lymphocytic leukemia
43 43
INTRODUCTION Diffuse low-grade B cell neoplasms form a heterogeneous group having in common relatively small cell size and a B cell phenotype. These lymphomas may have a variety of morphological appearances and in many cases may prove difficult to separate on the basis of morphology alone requiring detailed phenotyping for accurate characterization.1 Included within the diffuse indolent B cell lymphomas are a number of well-defined clinicopathological entities as well as a group of tumors that are less clearly defined.
SMALL LYMPHOCYTIC LYMPHOMA/B CELL CHRONIC LYMPHOCYTIC LEUKEMIA The vast majority of patients with lymphadenopathy showing the typical features of small lymphocytic lymphoma will have evidence of peripheral blood involvement by a clonal B cell population. Likewise, many patients who have a lymphocytosis due to B cell chronic lymphocytic leukemia (B-CLL) will show evidence of lymph node involvement demonstrating a morphological pattern identical to small lymphocytic lymphoma. The morphology and phenotype of small lymphocytic lymphoma and B-CLL are identical, and it is now clear that they are one and the same disease process.2 The question whether very rare cases of typical small lymphocytic lymphoma that never develop a leukemic phase exist remains controversial.3 Lymphadenopathy in B-CLL is typically of modest size and does not usually exceed 2 cm. Macroscopically,
Lymphoplasmacytic lymphoma Marginal zone lymphoma References
44 45 46
the lymph nodes have a delicate capsule and their cut surface has a pale white grey coloration and is soft in texture. Histologically, the lymph node architecture is diffusely replaced by small round lymphocytes, which typically have a clumped chromatin pattern and scanty cytoplasm4 (Plate 30). Scattered among these are slightly larger cells with more abundant pale cytoplasm, an open nuclear chromatin and a visible nucleolus; these are prolymphocytes. Occasional large cells with immunoblastic features termed paraimmunoblasts are also present.5 Prolymphocytes and paraimmunoblasts may be aggregated together to form rather ill-defined nodules termed pseudofollicles (Plate 31) and these are important diagnostic features, as they are not seen in other low-grade B cell lymphomas.6 Pseudofollicles occur in approximately 85 per cent of cases of B-CLL.5 Uncommonly, large aggregates of prolymphocytes may extensively infiltrate and replace the nodal architecture, which has been termed the tumor-forming subtype of B-CLL by Lennert;5 a very similar morphologic pattern has been termed the paraimmunoblastic variant of small lymphocytic lymphoma/leukemia by Pugh.7 The prognostic significance of this histologic pattern remains unclear but there is some evidence that these patients may have a worse prognosis.8 In addition, the relationship of this histologic pattern to CLL/prolymphocytic leukemia (PL) has yet to be clarified. A small percentage of cases of B-CLL may show evidence of plasmacytoid differentiation, and contain a number of plasmacytoid lymphocytes and mature plasma cells, which are clonally related to the B-CLL clone. These have previously been separated from B-CLL and termed lymphoplasmacytoid immunocytoma.2'5 However, many now believe when pseudofollicles are present these cases represent examples of B-CLL, which
44 Diffuse indolent B cell neoplasms
have acquired immunoglobulin secretory capacity and behave in an identical fashion to more typical B-CLL.9'10 Bone marrow infiltration is extremely common in BCLL and a variety of histological patterns may be observed in the bone marrow trephine, some of which have prognostic significance. The bone marrow trephine may show interstitial infiltration where B-CLL cells permeate between normal hemopoietic elements and fat cells. There may be nodular infiltration where rounded aggregates of small B cells are seen in the intertrabecular regions. On occasions these may be associated with proliferation centers. There is generally preservation of normal hemopoietic marrow around these nodules. There may be also a mixture of interstitial and nodular infiltration. In cases of heavy bone marrow involvement by BCLL, there may be diffuse effacement of the normal marrow architecture by sheets of small lymphocytes, often with admixed pseudofollicles and replacement of fat cells.11 The pattern of bone marrow infiltration correlates with prognosis with a median survival of 90 months for the nodular pattern, 46 months for the interstitial and 28 months for diffuse marrow replacement.12 B-CLL has a rather unique immunophenotype, which distinguishes it from normal B cells and other indolent B cell lymphomas. There is weak expression of surface immunoglobulin usually of IgM, or IgM and IgG type. There is expression of pan-B cell antigen CD 19 and CD 20. CD 5 and CD 23 are expressed in the majority of cases. FMC 7 is usually not seen.13'14 This phenotype has formed the basis of a scoring system for the diagnosis of B-CLL.13 Some cases of B-CLL will demonstrate atypical morphological features or possess a divergent phenotype with loss of CD 23 or 'bright' expression of Sig and CD 20; expression of atypical markers, such as FMC 7, CD 38 and CD lla may be seen. These cases have been termed 'atypical CLL' and are associated with a more advanced stage and have a more aggressive clinical course. They are also associated with complex cytogenetic abnormalities, such as trisomy 12.15-18 A minority of cases of B-CLL undergo transformation. This may take the form of increased numbers of prolymphocytes in the peripheral blood, termed prolymphocytoid transformation, or CLL/PL, when there are between 10 and 55 per cent of prolymphocytes in the peripheral blood.19 This is associated with a tendency to be refractory to chemotherapy and a poor prognosis. In some patients there is a striking increase in the size of lymphadenopathy, which may achieve diameters of 5 cm or more. This is often associated with development of a diffuse large B cell lymphoma, often showing pleomorphic immunoblastic cytology20 (Plate 32). This has been termed Richter's syndrome. In many cases the large B cell lymphoma is clonally related to the original CLL but in some it appears to be a clonally independent second tumor.21'22 Rare cases of high-grade transformation with lymphoblastic cytology and nuclear terminal
deoxynucleotidyl transferase (tdt) expression have been described. The occurrence of cells with the morphology and phenotype of classical Hodgkin's disease have been described and termed the Hodgkin's variant of Richter's transformation (Plate 33).23-25 These show evidence of the presence of Epstein-Barr virus in some cases.26'27 Microdissection studies of the H-RS cells have shown them to be clonally related to the original B-CLL clone.28 The clinical course of the Hodgkin's disease variant of Richter's syndrome is still not entirely clear as few cases have been reported; some at least seem to follow the clinical pattern of Hodgkin's disease with persistent B-CLL. Rare cases of nodular lymphocyte-predominant Hodgkin's disease have been described in association with B-CLL.29
LYMPHOPLASMACYTIC LYMPHOMA A variety of terms have been used to describe immunosecretory small B cell neoplasms without features of other lymphoma subtypes including small lymphocytic lymphoma plasmacytoid, lymphoplasmacytoid and lymphoplasmacytic immunocytoma, and Waldenstrom's macroglobulinemia. Included under these terms are a heterogeneous group of B cell lymphomas showing variable degrees of plasma cell differentiation. In the Kiel classification three subtypes of immunocytoma have been recognized: lymphoplasmacytoid, lymphoplasmacytic and polymorphic.5 The polymorphic subtype was abandoned to avoid confusion.30 The Kiel definition of immunocytoma assumes that there is no evidence of follicle center cell, mantle cell or marginal zone differentiation, and that the tumor is composed predominantly of small lymphocytes showing variable degrees of plasmacytoid differentiation and, in the lymphoplasmacytic subtype, mature or Marshalko plasma cells. More recently, the International Lymphoma Study Group has redefined immunocytoma and excluded cases that show features of B-CLL, and have restricted the diagnosis of immunocytoma to one of a small B lymphocytic lymphoma showing lymphoplasmacytoid or plasma cell differentiation without features of any other lymphoma subtype.9,10 Lymph nodes affected by lymphoplasmacytic immunocytoma are usually only modestly enlarged. The normal architecture is partially or completely effaced by diffuse proliferation of small lymphoid cells with scattered lymphoplasmacytoid forms and in some cases mature plasma cells (Plate 34). Periodic acid-Schiff (PAS) staining may reveal the presence of positive intranuclear inclusions called Dutcher bodies in up to half the cases. There may be scattered large lymphoid cells with immunoblastic morphology, some showing plasmacytoid features. In some cases these may be
Marginal zone lymphoma 45
numerous and the borderline between diffuse large B cell lymphoma of immunoblastic type and immunocytoma can be difficult to define. In lymph nodes where there is not complete effacement of the architecture, the growth pattern tends to be interfollicular and the peripheral sinuses are often preserved. The mere presence of lymphoplasmacytoid and plasma cells in a B cell neoplasm does not define the tumor as immunocytoma. The plasmacytoid component should show the same light-chain restriction as the small B cell component and there should be no evidence of BCLL, such as proliferation centers and prolymphocytes. The spleen is commonly infiltrated in cases of advanced-stage immunocytoma, where there may be infiltration of both the red and the white pulp, which may cause diagnostic difficulties in the separation of immunocytoma from splenic marginal zone lymphoma. Precise criteria for this distinction are not currently available and the precise inter-relationship of these two entities is still unclear. Bone marrow infiltration is common and may be nodular or interstitial. Immunophenotypically, lymphoplasmacytic immunocytoma differs from B-CLL in lacking CD 5 and CD 23. There is usually strong surface immunoglobulin, usually of the IgM type, which shows the same light-chain restriction as lymphoplasmacytoid cells containing cytoplasmic immunoglobulin (Plate 35). There is expression of the pan-B cell markers CD 19, CD 20, CD 22 and CD 79a, and in cases showing pronounced plasma cell differentiation, CD 79a is the most reliable marker for paraffin-section immunohistochemistry. It is clear from the above descriptions of the morphology and phenotype of lymphoplasmacytic immunocytoma that the diagnosis is to a certain extent one of exclusion of other lymphoma subtypes. In a recent study it has been shown that the interobserver and intraobserver concordance rates for the diagnosis of lymphoplasmacytic immunocytoma are relatively poor compared to other better-defined lymphoma entities. A chromosomal translocation that has been associated with lymphoplasmacytic immunocytoma is the t(9;14)(p!3;q32).This has been shown to deregulate the pax 5 gene, which encodes the B cell specific activator protein.31 There is considerable controversy in the literature as to whether the prognosis of lymphoplasmacytic immunocytoma differs significantly from small lymphocytic lymphoma/B-CLL. Preliminary data from St Bartholomew's Hospital in London would suggest that any degree of plasmacytoid differentiation is associated with an inferior overall and disease-free survival.32
MARGINAL ZONE LYMPHOMA Proliferations of marginal zone B cells have been recognized for many years in reactive lymphadenopathies, in
particular, toxoplasmosis.33'34 Although initially considered to be of monocyte/macrophage origin, their B cell lineage was later clarified.35'36 Malignant lymphomas showing cytological features of monocytoid B cells were later recognized and termed monocytoid B cell lymphoma37-46 or parafollicular B cell lymphoma by some.47 Lymph nodes involved by monocytoid B cell lymphoma show infiltration around and within the nodal sinuses, which extends in to the parenchyma and often surrounds reactive germinal centers (Plate 36). Cytologically there is variability in cell size and a significant degree of nuclear irregularity; there is often relatively abundant clear or grey cytoplasm45 (Plate 37). It became apparent there was a morphologic overlap with extranodal MALT lymphomas (see Chapter 8) and this commonality became more obvious when cases presenting as nodal disease later developed typical extranodal MALT-type disease.45,46 There developed a considerable controversy as to whether nodal monocytoid B cell lymphoma actually existed as an entity or whether it was a manifestation of metastatic MALT lymphoma involving lymph nodes.48 In 1994, the International Lymphoma Study Group published the revised European-American lymphoma classification (REAL),9 which included the term marginal zone B cell lymphoma (MZL). Under this heading were three entities, extranodal MZL of MALT type (see Chapter 8) and two provisional entities of nodal and splenic MZL. Nodal MZLs were regarded as having identical morphology to their extranodal equivalent and it was believed that the majority were disseminated MALT lymphomas with only a small minority being true primary, nodal MZLs.9 Recent work suggests primary nodal MZL does exist and shows significant differences in the distribution of disease, clinical parameters and survival.49 Nodal MZLs present with peripheral and para-aortic lymphadenopathy more commonly than extranodal MZL; they also have a significantly worse overall and failure-free survival.49 Primary splenic lymphomas with involvement of the red and white pulp have been described.50,51 These may show the presence of villous lymphocytes in the peripheral blood.50,52 In the REAL classification these are termed splenic marginal zone lymphoma. Splenic infiltration is characterized by white pulp involvement, which may surround or replace germinal centers. In the red pulp there may be small nodules or diffuse sinusoidal infiltration.52 Cytologically the lymphoma cells may contain small lymphocytes, typically located in the mantle zone or larger 'monocytoid' cells present in the marginal zone. Bone marrow involvement is common and usually nodular. Circulating lymphoma cells are frequently detected in the peripheral blood and may possess thin cytoplasmic projections, termed villous lymphocytes50 (Plate 38). Splenic MZL corresponds closely to the entity termed splenic lymphoma with circulating villous lymphocytes.53,54
46 Diffuse indolent B cell neoplasms
REFERENCES 1. Harris NL Low grade B-cell neoplasms. In: Weiss LM, ed. Pathology of lymph nodes. New York: Churchill Livingstone,1996: 236-74.
18. Matutes E, Oscier D, Garcia-Marco J, et al. Trisomy 12 defines a group of CLL with atypical morphology: correlation between cytogenetic, clinical and laboratory features in 544 patients. BrJ Haematol 1996; 92: 382-8.
2. Jaffe ES, Raffeld M, Medeiros LJ. Histopathologic subtypes of indolent lymphomas: caricatures of the mature B-cell system. Semin Oncol 1993; 20: 3-30.
19. Bennett JM, Catovsky D, Daniel MT, et al. Proposals for the classification of chronic (mature) B and T lymphoid leukaemias. French-American-British (FAB) Cooperative
3. Ben-Ezra JM. Small lymphocytic lymphoma. In: Knowles
Group.7 Clin Pathol 1989; 42: 567-84. 20. Armitage JO, Dick FR, Corder MP. Diffuse histiocytic
DM, ed. Neoplastic hematopathology. Baltimore: Williams &Wilkins, 1992:603-16. 4. Pangalis GA, Nathwani BN, Rappaport H. Malignant lymphoma, well differentiated lymphocytic: its relationship with chronic lymphocytic leukemia and
lymphoma complicating chronic lymphocytic leukemia. Cancer 1978; 41: 422-7. 21. Matolcsy A, Casali P, Knowles DM. Different clonal origin of B-cell populations of chronic lymphocytic leukemia
macroglobulinemia of Waldenstrom. Cancer 1977; 39:
and large-cell lymphoma in Richter's syndrome. Ann NY
999-1010. 5. Lennert K. Malignant lymphomas other than Hodgkin's disease. New York: Springer-Verlag, 1978. 6. Dick FR, Maca RD. The lymph node in chronic
AcadSci 1995; 764: 496-503. 22. Matolcsy A, Inghirami G, Knowles DM. Molecular genetic demonstration of the diverse evolution of Richter's
lymphocytic leukemia. Cancer 1978; 41: 283-92. 7. Pugh WC, Manning JT, Butler JJ. Paraimmunoblastic variant of small lymphocytic lymphoma/leukemia. Am J Surg Pathol 1988; 12: 907-17. 8. Bonato M, Pittaluga S, Tierens A, et al. Lymph node histology in typical and atypical chronic lymphocytic leukemia. Am J Surg Pathol 1998; 22: 49-56. 9. Harris NL, Jaffe ES, Stein H, et al. A revised EuropeanAmerican classification of lymphoid neoplasms: a proposal from the International Lymphoma Study Group. Blood 1994; 84:1361-92. 10. Harris NL. Low grade B-cell neoplasms.ln: Weiss LM, ed. Pathology of lymph nodes. New York: Churchill Livingstone, 1996: 247. 11. Rozman C, Hernandez-Nieto L, Montserrat E, Brugues R. Prognostic significance of bone-marrow patterns in chronic lymphocytic leukaemia. BrJ Haematol 1981; 47: 529-37. 12. Frisch B, Bartl R. Biopsy pathology of bone and bone marrow, 2nd edn. London: Arnold, 1999. 13. Moreau EJ, Matutes E, A'Hern RP, et al. Improvement of the chronic lymphocytic leukemia scoring system with the monoclonal antibody SN8 (CD79b). AmJ Clin Pathol 1997; 108: 378-82. 14. Molica S, Levato D, Dattilo A, Mannella A. Clinicoprognostic relevance of quantitative immunophenotyping in B-cell chronic lymphocytic leukemia with emphasis on the expression of CD20 antigen and surface immunoglobulins. EuroJ Haematol 1998;60:47-52. 15. Mould S, Gardiner A, Corcoran M, Oscier DG. Trisomy 12 and structural abnormalities of 13q14 occurring in the same clone in chronic lymphocytic leukaemia. BrJ Haematol 1996; 92: 389-92. 16. Juliusson G, Merup M. Cytogenetics in chronic lymphocytic leukemia. Semin Oncol 1998; 25:19-26. 17. Oscier DG. Cytogenetic and molecular abnormalities in chronic lymphocytic leukaemia. Blood Rev 1994; 8: 88-97.
syndrome (chronic lymphocytic leukemia and subsequent large cell lymphoma). Blood 1994; 83: 1363-72. 23. Brecher M, Banks PM. Hodgkin's disease variant of Richter's syndrome. Report of eight cases. AmJ Clin Pathol 1990; 93: 333-9. 24. Fayad L, Robertson LE, O'Brien S, et al. Hodgkin's disease variant of Richter's syndrome: experience at a single institution. Leuk Lymphoma 1996; 23: 333-7. 25. Williams J, Schned A, Cotelingam JD, Jaffe ES. Chronic lymphocytic leukemia with coexistent Hodgkin's disease. Implications for the origin of the Reed-Stern berg cell. Am J Surg Pathol 1991; 15: 33-42. 26. Momose H, Jaffe ES, Shin SS, Chen YY, Weiss LM. Chronic lymphocytic leukemia/small lymphocytic lymphoma with Reed-Sternberg-like cells and possible transformation to Hodgkin's disease. Mediation by Epstein-Barr virus. AmJ Surg Pathol 1992; 16: 859-67. 27. Rubin D, Hudnall SD, Aisenberg A, Jacobson JO, Harris NL. Richter's transformation of chronic lymphocytic leukemia with Hodgkin's-like cells is associated with Epstein-Barr virus infection. Mod Pathol 1994; 7: 91-8. 28. Ohno T, Smir BN, Weisenburger DD, Gascoyne RD, Hinrichs SD, Chan WC. Origin of the Hodgkin/Reed-Sternberg cells in chronic lymphocytic leukemia with 'Hodgkin's transformation'. Blood 1998; 91:1757-61. 29. Weisenberg E, Anastasi J, Adeyanju M, Variakojis D, Vardiman JW. Hodgkin's disease associated with chronic lymphocytic leukemia. Eight additional cases, including two of the nodular lymphocyte predominant type. AmJ Clin Pathol 1995; 103: 479-84. 30. Lennert K, Feller AC. Histopathology of non-Hodgkin's lymphomas (based on the updated Kiel classification). Berlin: Springer-Verlag, 1990. 31. Amakawa R, Ohno H, Fukuhara, S. t(9;14)(p13;q32) involving the PAX-5 gene: a unique subtype of 14q32 translocation in B cell non-Hodgkin's lymphoma. IntJ Hematol 1999; 69: 65-9.
References 47 32. Rohatiner A, personal communication, 1998. 33. Stansfeld AG. The histologic diagnosis of toxoplasmic lymphadenitis.) Clin Pathol 1961; 14: 565-73. 34. Dorfman RF, Remmington JS. Value of lymph node biopsy in the diagnosis of acute acquired toxoplasmosis. N EnglJ Med 1973; 289: 878-81. 35. Cardoso DA, Harris NL, Bhan AK. Characterization of immature sinus histiocytes (monocytoid cells) in reactive lymph nodes by use of monoclonal antibodies. Human Pathol 1984; 15: 330-5. 36. Kojima M, Nakamura S, Itoh H, et al. Occurrence of monocytoid B-cells in reactive lymph node lesions. Pathol Res Pract 1998; 194: 559-65. 37. Nathwani BN, Mohrmann RL, Brynes RK, Taylor CR, Hansmann ML, Sheibani K. Monocytoid B-cell lymphomas: an assessment of diagnostic criteria and a perspective on histogenesis. Human Pathol 1992; 23: 1061-71. 38. Sheibani K, Burke JS, Swartz WG, Nademanee A, Winberg CD. Monocytoid B-cell lymphoma. Clinicopathologic study of 21 cases of a unique type of low-grade lymphoma. Cancer 1988; 62:1531-8. 39. Sheibani K, Sohn CC, Burke JS, Winberg CD, Wu AM, Rappaport H. Monocytoid B-cell lymphoma. A novel Bcell neoplasm. Amj Pathol 1986; 124: 310-18. 40. Arber DA, Sheibani K, Weiss LM. UCL3D3 and UCL4D12 reactivity in small B-cell neoplasms with special emphasis on monocytoid B-cell lymphoma. Human Pathol 1994; 25:1084-90. 41. Shin SS, Sheibani K. Monocytoid B-cell lymphoma. Am J C//>7P0tfjo/1993;99:421-5. 42. TraweekST, Sheibani K. Monocytoid B-cell lymphoma. The biologic and clinical implications of peripheral blood involvement. Am J Clin Pathol 1992; 97: 591-8. 43. Traweek ST, Sheibani K, Winberg CD, Mena RR, Wu AM, Rappaport H. Monocytoid B-cell lymphoma: its evolution and relationship to other low-grade B-cell neoplasms. Blood 1989; 73: 573-8.
44. Plank L, Hansmann ML, Fischer R. The cytological spectrum of the monocytoid B-cell reaction: recognition of its large cell type. Histopathology 1993; 23: 425-31. 45. Cogliatti SB, Lennert K, Hansmann ML, Zwingers TL. Monocytoid B cell lymphoma: clinical and prognostic features of 21 patients. J Clin Pathol 1990; 43: 619-25. 46. Nizze H, Cogliatti SB, von Schilling C, Feller AC, Lennert K. Monocytoid B-cell lymphoma: morphological variants and relationship to low-grade B-cell lymphoma of the mucosa-associated lymphoid tissue. Histopathology 1991; 18: 403-14. 47. Davis GG, York JC, Glick AD, McCurley TL, Collins RD, CousarJB. Plasmacytic differentiation in parafollicular (monocytoid) B-cell lymphoma. A study of 12 cases. Amj Surg Pathol 1992; 16:1066-74. 48. Campo E, Miquel R, Krenacs L, Sorbara L, Raffeld M, Jaffe ES. Primary nodal marginal zone lymphomas of splenic and MALT type. Amj Surg Pathol 1999; 23: 59-68. 49. Nathwani BN, Anderson JR, Armitage JO, et al. Marginal zone B-cell lymphoma: a clinical comparison of nodal and mucosa-associated lymphoid tissue types. 7 Clin Oncol 1999; 17: 2486-92. 50. Neiman RS, Sullivan AL, Jaffe R. Malignant lymphoma simulating leukemic reticuloendotheliosis: a clinicopathologic study of ten cases. Cancer 1979; 43: 329-42. 51. Schmid C, Kirkham N, DissT, Isaacson PG. Splenic marginal zone cell lymphoma./4/r?7 Surg Pathol 1992; 16: 455-66. 52. Isaacson PG, Matutes E, Burke M, Catovsky D. The histopathology of splenic lymphoma with villous lymphocytes. Blood 1994; 84: 3828-34. 53. Mulligan SP, Matutes E, Dearden C, Catovsky D. Splenic lymphoma with villous lymphocytes: natural history and response to therapy in 50 cases. BrJ Haematol 1991; 78: 206-9. 54. Mulligan SP, Catovsky D. The chronic B-cell leukaemias. Austral N ZealJ Med 1993; 23: 42-50.
This page intentionally left blank
6 Diffuse aggressive B cell lymphoma KAMACLENNAN
Burkitt's lymphoma High-grade B cell lymphoma Burkitt-like
49 49
BURKITTS LYMPHOMA Burkitt's l y m p h o a was first recognized as an endemic malignancy of childhood in equatorial Africa where it comprised 50 per cent of childhood cancer.1'5 The association between Epstein-Barr virus (EBV) and Burkitt's lymphoma was established, and it has been shown that all the cells of the malignant clone in endemic Burkitt's lymphoma show evidence of EBV infection.6-10 Later the characteristic chromosomal translocation t(8;14)(q24;q32) and the variant translocations involving chromosome 2pl 1 and 22ql 1, at the loci of the K and l immunoglobulin light-chain genes were described. These translocations deregulate the c-myc gene and are present in all cases of Burkitt's lymphoma whether endemic or non-endemic.11-23 There are marked differences in the clinical presentation between endemic and non-endemic Burkitt's lymphoma. Endemic Burkitt's lymphoma shows a high frequency of involvement of the jaw, which is more common in younger patients (70 per cent in children under 5 reducing to 25 per cent in patients over 14) and is centred around the developing dental lamina. Abdominal disease is frequently encountered in endemic Burkitt's lymphoma, being present in over half the patients at presentation. The pattern of disease within the abdomen differs from non-endemic Burkitt's (see below) with a relatively high frequency of mesenteric involvement and ascites. While localized ileocecal disease is uncommon,24 sporadic Burkitt's frequently presents with disease centred around the terminal ileum and cecum. It may present with bulky disease at this site or relatively small volume disease, which comes to clinical attention because it forms the apex of an intussusception. There is also a high frequency of involvement of peripheral lymph nodes, tonsils and oropharynx and pleura.
Diffuse large B cell lymphoma References
50 51
Burkitt's lymphoma, whether endemic or sporadic shows an identical histological picture. The tumor adopts a diffuse growth pattern with a starry sky appearance caused by actively phagocytic macrophages engulfing apoptotic debris. The tumor cells are monomorphous medium-sized lymphoid cells arranged in cohesive sheets (Plate 39). They have an open nuclear chromatin with multiple small nucleoli and a thin rim of basophilic cytoplasm that is best appreciated in Giemsa-stained preparations.25-28 The proliferation rate is high29 and numerous cells are seen to undergo apoptosis. The histological appearances of Burkitt's lymphoma are critically dependent on good fixation and the diagnosis maybe problematic in poorly fixed material. Imprint cytology is often extremely helpful in establishing the diagnosis where the typical L3 cytology with basophilic, vacuolated cytoplasm is apparent. Burkitt's lymphoma displays a mature B cell phenotype expressing CD 19, CD 20 and CD 79a. CD 5 and CD 23 are absent and CD 10 is almost invariably present (Plate 40). There is expression of strong monoclonal surface immunoglobulin, usually of IgM type. Almost 100 per cent of the cells are labeled with the proliferationassociated antigen Ki 67 (Plate 41).
HIGH-GRADE B CELL LYMPHOMA BURKITT-LIKE High-grade B cell lymphoma Burkitt-like is included as a provisional entity in the Revised European-American Lymphoma (REAL) classification to denote cases that are borderline in their morphology between true Burkitt's and large B cell lymphoma.30 Some of these cases correspond to the entity of Burkitt's lymphoma
50 Diffuse aggressive B cell lymphoma
with cytoplasmic immunoglobulin or the plasmablastic subtype of Burkitt's lymphoma described by the Kiel group.28 Morphologically, there is considerably more pleomorphism in high-grade B cell lymphoma Burkittlike than in true Burkitt's, with the presence of some large pleomorphic cells, medium-sized cells with more abundant cytoplasm than true Burkitt's and scattered small B cells. Phenotypically, there is expression of panB cell antigen CD 19 and CD 20, but CD 10, which is almost invariably present in true Burkitt's lymphoma, is absent from a significant percentage of cases of highgrade B cell lymphoma Burkitt-like. Recent data from the lymphoma classification project show that the diagnostic reproducibility of a diagnosis of high-grade B cell lymphoma Burkitt-like is poor. The clinical presentation and prognosis are identical to diffuse large B cell lymphoma. It would thus appear that there is little value in the separate recognition of this intermediate group between true Burkitt's lymphoma and diffuse large B cell lymphoma, and these cases should be classified as diffuse large B cell lymphoma.31
DIFFUSE LARGE B CELL LYMPHOMA
Introduction Large cell lymphomas were initially regarded as arising from a variety of cell lineages, including reticulum cells and histiocytes, before immunologic techniques clearly defined them as being of lymphoid origin.32"41 The majority of diffuse large cell lymphomas (90 per cent) are of B cell lineage and make up between 30 and 40 per cent of all non-Hodgkin's lymphomas (NHLs).31 Up to 40 per cent of cases present with extranodal disease, and there is an approximately equal division between localized and advanced stage disease. Diffuse large B cell lymphoma may be primary or secondary to a pre-existing or synchronous low-grade B cell NHL, usually follicular lymphoma. A small percentage of diffuse large B cell lymphomas arise in the setting of lymphocyte-predominant nodular Hodgkin's disease, which usually has undergone multiple relapses over a long period of time.
Morphology Diffuse large B cell lymphomas are composed of large lymphoid cells whose nuclei are approximately twice the size of a small lymphocyte. There is an open nuclear chromatin with often multiple punctate nucleoli (Plate 42). The cytoplasm is basophilic and may be scanty or abundant with evidence of plasmacytoid differentiation. This diversity of morphology has been used as the basis for subdivision of large cell NHL of B cell lineage into numerous subtypes.28,42-48 Centroblastic lymphoma,
immunoblastic lymphoma, large cell, sclerosing B cell lymphoma of the mediastinum and large cell anaplastic lymphoma of B cell lineage are all separately designated in the updated Kiel classification and subsequent publications.49'50 Centroblastic lymphoma has been subdivided into four cytologic groups in the Kiel classification. These are: monomorphic, where the cellular composition consists of over 60 per cent of typical centroblasts, which are medium sized to large cells with round or oval pale nuclei, multiple peripheral nucleoli and a thin rim of basophilic cytoplasm. The polymorphic subtype is characterized by an admixture of centroblasts and immunoblasts, the latter comprising more than 10 per cent of the cellular population. The multilobated variant is diagnosed when more than 10 per cent of the tumor cells show three or four nuclear lobulations. Centrocytoid Centroblastic lymphoma is characterized by intermediate cytology between large centrocytes and centroblasts.28 The latter subtype is felt by many to include many cases of blastically transformed mantle cell lymphoma. Using the Kiel criteria, B immunoblastic lymphoma is defined as a monomorphic tumor composed of sheets of immunoblasts that comprise more than 90 per cent of the cellular population. Immunoblasts are characterized by large oval nuclei with clear nucleoplasm and a single large central nucleolus or multiple central nucleoli (Plate 43). They may show pronounced plasmacytic differentiation.28 A small study by the International Lymphoma Study Group, included in the publication of the REAL classification showed that subdivision of large B cell lymphoma was poorly reproducible, even by experienced hematopathologists. The REAL classification amalgamated these entities and termed them diffuse large B cell lymphoma, and included primary mediastinal (thymic) large B cell lymphoma as a subtype.
Phenotype Diffuse large B cell lymphomas express pan-B cell antigens CD 19, CD 20, CD 22 and CD 79a. In conventionally fixed and processed sections, over 80 per cent of diffuse large B cell lymphomas express CD 20 (Plate 44). Cases that are CD 20 negative often show immunoblastic cytology; the majority of these are CD 79a positive. The germinal centre-associated marker, CD 10, is seen in a significant percentage of diffuse large B cell lymphomas. CD 5 is present in around 10 per cent of cases and may be associated with immunoblastic morphology. Expression of BCL 2 protein has been shown to be an independent prognostic marker in diffuse large B cell lymphoma and is seen in approximately 50 per cent (Plate 45). Overexpression is associated with a significantly reduced relapse-free and cause-specific survival.28,51-54
References 51
Rare subtypes of diffuse large B cell lymphoma A distinctive type of large B cell lymphoma has been described, which arises in the mediastinum, probably from a population of thymic B cells.55'56 This tumor affects a younger age group than typical large B cell lymphomas, with a median age ranging from 25 to 32 years. There is a female predominance with a male to female ratio of 1:2. Most patients present with localized disease (Ann Arbor stage I and II). However, there is often extension into adjacent structures such as lung, pericardium and anterior chest wall.57 Histologically, mediastinal large B cell lymphoma shows distinctive features: it is characterized by compartmentalizing fine fibrosis, which separates lobules of large lymphoid cells with clear cytoplasm (Plate 46). There are a wide variety of different nuclear cytological features that range from irregular nuclei resembling large centrocytes to a typical centroblastic morphology and some cases show immunoblastic nuclear features. Immunophenotypically, mediastinal large B cell lymphomas express pan-B cell antigen CD 19 and CD 20 and, in common with normal thymic B cells, are usually CD 21 negative. It is rarely possible to demonstrate the presence of monoclonal immunoglobulin in these tumors. Genotypically, mediastinal large B cell lymphomas show rearrangement of both heavy- and light-chain immunoglobulin genes. There are point mutations or translocations involving the c-myc oncogene.58 The bd 2 gene is in germline configuration.59 Large cell anaplastic lymphoma (LCAL) of B cell lineage shows similar morphologic features to anaplastic large cell lymphoma (ALCL) of T/null type. Phenotypically LCAL shows weaker expression of CD 30, which, rather than having a characteristic distinct membrane and Golgi distribution, shows a diffuse cytoplasmic staining. Recently, an unusual subtype of large B cell lymphoma has been described which shows immunoblastic cytology, lacks B lineage markers and contains monoclonal cytoplasmic IgA. Epithelial membrane antigen is expressed in all described cases but there is no reactivity with CD 30. This lymphoma is unusual as it expresses the full-length form of the anaplastic lymphoma kinase, which is only detected as a truncated fusion protein in malignant lymphomas in association with t(2;5) in ALCL of T/null type; this translocation has not been identified in this type of large B cell lymphoma. Rare cases of diffuse large B cell lymphoma demonstrate a small number of neoplastic large B cells, comprising 5-10 per cent of the total cellular population, in a numerically predominant background of non-neoplastic small T cells or histiocytes. These cases have been termed pseudo-T cell lymphoma and T cell-rich B cell lymphoma (Plate 47). Morphologically T cell-rich B cell lymphoma shows a diffuse growth pattern composed of
sheets of small lymphocytes sometimes admixed with epithelioid histiocytes. The large neoplastic B cells may be inconspicuous and only become apparent on immunostaining with B cell markers. Small B cells are notable by their absence. The morphology of the large B cell component is variable and they may show immunoblastic, centroblastic or pleomorphic cytology. A common cytologic subtype shows marked similarities with the popcorn or lymphocytic and/or histiocytic cell of lymphocyte-predominant Hodgkin's disease, which causes difficulties in the differential diagnosis of diffuse lymphocytic predominance. The clinical behavior of T cell-rich B cell lymphoma does not appear to differ from other diffuse large B cell lymphomas in terms of the complete response rate, overall and relapse-free survival. Some workers have demonstrated an unusually high rate of bone marrow disease, which is in excess of 50 per cent.60 Intravascular large cell lymphoma is a rare but distinctive form of high-grade lymphoma characterized by the accumulation of large lymphoid cells within vascular lumina. These may lie free within the vascular space or may be enmeshed within fibrin (Plate 48). Rarely, lymphoma cells may spread into adjacent lymph node or extranodal structures. Most cases exhibit a B cell phenotype,61"67 but rare T lineage cases have been described.68,69 The clinical manifestations of intravascular large cell lymphoma are variable, but central nervous system and cutaneous disease are extremely common.63,70 Other extranodal sites, such as lung, kidney and gastrointestinal tract, maybe involved.61,71-73 Primary effusion lymphoma (PEL), originally termed body cavity-based lymphoma, is a rare form of large cell lymphoma that is characterized by the development of lymphomatous effusions in the pleural and abdominal cavity in the absence of solid tumor masses. Initially reported cases were all associated with human immunodeficiency virus (HIV) infection, EBV and a newly described human herpes virus type 8 (HHV-8).74"80 Recently two distinctive forms of PEL have been described in those that are HIV and EBV positive; the majority of these are associated with HHV-8 and show a germline configuration of c-myc. These cases show immunoblastic or anaplastic cytology. The remaining cases are not associated with HHV-8 and show rearrangements of c-myc. These show Burkitt's or Burkitt-like morphology.81 Most cases lack B cell marker expression and show evidence of clonal immunoglobulin gene rearrangement indicating their B cell lineage. Structural analysis of the IgH genes suggest that PELs may arise from different stages of B cell development.82
REFERENCES 1. Burkitt D. Malignant lymphomata involving the jaws in Africa. 7 Laryngol Otol 1965; 79: 929-39.
52 Diffuse aggressive B cell lymphoma 2. Burkitt D, Wright D. Geographical and tribal distribution of the African lymphoma in Uganda. Br MedJ 1966; 1: 569-73. 3. Wright DH. The epidemiology of Burkitt's tumor. Cancer Res 1967; 27: 2424-38. 4. Burkitt D. Burkitt's lymphoma. JAMA 1972; 222:1164. 5. Burkitt DP. Classics in oncology. A sarcoma involving the jaws in African children. CA Cancer] Clin 1972; 22: 345-55. 6. Niedobitek G, Agathanggelou A, Rowe M, et al. Heterogeneous expression of Epstein-Barr virus latent proteins in endemic Burkitt's lymphoma. Blood 1995; 86: 659-65. 7. Geser A, Lenoir GM, Anvret M, et al. Epstein-Barr virus markers in a series of Burkitt's lymphomas from the West Nile District, Uganda. EurJ Cancer Clin Oncol 1983; 19: 1393^04. 8. Purtilo DT, Manolov G, Manolova Y, Harada S, Lipscomb H, Tatsumi E. Role of Epstein-Barr virus in the etiology of Burkitt's lymphoma. lARCSci Publ 1985; 231-47. 9. Rowe M, Rowe DT, Gregory CD, et al. Differences in B cell growth phenotype reflect novel patterns of Epstein-Barr virus latent gene expression in Burkitt's lymphoma cells. EM BO J 1987; 6: 2743-51. 10. Tao Q, Robertson KD, Manns A, Hildesheim A, Ambinder RF. Epstein-Barr virus (EBV) in endemic Burkitt's lymphoma: molecular analysis of primary tumor tissue. Blood 1998; 91:1373-81. 11. Siebert R, Matthiesen P, Harder S, et al. Application of interphase fluorescence in situ hybridization for the detection of the Burkitt translocation t(8;14)(q24;q32) in B-cell lymphomas. Blood 1998; 91: 984-90. 12. Niwano H, Aoki S, Tsukada N, et al. An aggressive case of Burkitt's lymphoma with t(8;14) and c-myc rearrangement transformed from CD5+ B-cell lymphoma. Ann Hematol 1997; 75: 221-5. 13. Berger R, Bernheim A. Cytogenetics of Burkitt's lymphoma-leukaemia: a review. lARCSci Publ 1985; 65-80. 14. Berger R, Bernheim A. Cytogenetic studies on Burkitt's lymphoma-leukemia. Cancer Genet Cytogenet 1982; 7: 231^4. 15. Whang-Peng J, Lee EC, Sieverts H, Magrath IT. Burkitt's lymphoma in AIDS: cytogenetic study. Blood 1984; 63: 818-22. 16. Lenoir GM, Land H, Parada LF, Cunningham JM, Weinberg RA. Activated oncogenes in Burkitt's lymphoma. Curr Top Microbiol Immunol 1984; 113: 6-14. 17. Douglass EC, Magrath IT, Lee EC, Whang-Peng J. Cytogenetic studies in non-African Burkitt lymphoma. Blood 1980; 55:148-55. 18. Taub R, Kirsch I, Morton C, et al. Translocation of the cmyc gene into the immunoglobulin heavy chain locus in human Burkitt lymphoma and murine plasmacytoma cells. Proc Natl Acad Sci USA 1982; 79: 7837-41. 19. Lai JL, Fenaux P, Zandecki M, Nelken B, Huart JJ, Deminatti M. Cytogenetic studies in 30 patients with
20.
21.
22.
23.
24.
25.
26. 27.
28.
29. 30.
31.
32.
33.
34.
35.
Burkitt's lymphoma or L3 acute lymphoblastic leukemia with special reference to additional chromosome abnormalities. Ann Genet 1989; 32: 26-32. Emanuel BS, Selden JR, Chaganti RS, Jhanwar S, Nowell PC, Croce CM. The 2p breakpoint of a 2;8 translocation in Burkitt lymphoma interrupts the V kappa locus. Proc Natl Acad Sci USA 1984; 81: 2444-6. Gelmann EP, Psallidopoulos MC, Papas TS, Dalla-Favera R. Identification of reciprocal translocation sites within the c-myc oncogene and immunoglobulin mu locus in a Burkitt lymphoma. Nature 1983; 306: 799-803. Dalla-Favera R, Bregni M, Erikson J, Patterson D, Gallo RC, Croce CM. Human c-myc one gene is located on the region of chromosome 8 that is translocated in Burkitt lymphoma cells. Proc Natl Acad Sci USA 1982; 79: 7824-7. Aventin A, Mecucci C, Guanyabens C, et al. Variant t(2;18) translocation in a Burkitt conversion of follicular lymphoma. BrJ Haematol 1990; 74: 367-9. Magrath I and Bhatia K. Pathogenesis of small noncleaved cell lymphomas (Burkitt's lymphoma). In: Magrath I, ed. The non-Hodgkin's lymphomas. London: Arnold, 1997: 385-410. Wright DH. Burkitt's lymphoma: a review of the pathology, immunology, and possible etiologic factors. PatholAnnu 1971; 6: 337-63. Wright DH. Definition of Burkitt's tumor. Int J Cancer 1968; 3: 410. Rappaport H, Wright DH, Dorfman RF. Suggested criteria for the diagnosis of Burkitt's tumor. Cancer Res 1967; 27: 2632. Hui PK, Feller AC, Lennert K. High-grade non-Hodgkin's lymphoma of B-cell type. I. Histopathology. Histopathology 1988; 12:127-43. Cooper EH, Frank GL, Wright DH. Cell proliferation in Burkitt tumours. EurJ Cancer 1966; 2: 377-84. Harris NL, Jaffe ES, Stein H, et al. A revised European-American classification of lymphoid neoplasms: a proposal from the International Lymphoma Study Group. Blood 1994; 84:1361-92. Anonymous. A clinical evaluation of the International Lymphoma Study Group classification of non-Hodgkin's lymphoma. The Non-Hodgkin's Lymphoma Classification Project. Blood 1997; 89: 3909-18. Berard CW, Jaffe ES, Braylan RC, Mann RB, Nanba K. Immunologic aspects and pathology of the malignant lymphomas. Cancer 1978; 42: 911-21. Stein H, Papadimitriou CS, Bouman H, Lennert K, FuchsJ. Demonstration of immunoglobulin production by tumor cells in non-Hodgkin's and Hodgkin's malignant lymphomas and its significance for their classification. Recent Results Cancer Res 1978; 64:158-75. Taylor CR. Immunocytochemical methods in the study of lymphoma and related conditions.) Histochem Cytochem 1978; 26: 496-512. Habeshaw JA, Catley PF, Stansfeld AG, Brearley RL Surface phenotyping, histology and the nature of non-
References 53
36.
37.
38.
39.
40.
41.
42.
43.
44.
45.
46.
47.
48. 49.
Hodgkin lymphoma in 157 patients. BrJ Cancer 1979; 40:11-34. Berard CW, Cossman J, Jaffe ES. Malignant lymphomas as tumours of the immune system. BrJ Cancer 1980; 42:, 1-20. Warnke R, Miller R, Grogan T, Pederson M, DilleyJ, Levy R. Immunologic phenotype in 30 patients with diffuse large-cell lymphoma. N EnglJ Med 1980; 303: 293-300. Jaffe ES, Strauchen JA, Berard CW. Predictability of immunologic phenotype by morphologic criteria in diffuse aggressive non-Hodgkin's lymphomas. AmJClin Pathol 1982; 77: 46-9. Warnke RA, Strauchen JA, Burke JS, Hoppe RT, Campbell BA, Dorfman RF. Morphologic types of diffuse large-cell lymphoma. Cancer 1982; 50: 690-5. Lindemalm C, Christensson B, Johansson B, Mellstedt H, Ost A, Biberfeld P. A clinico-pathological and immunological study of unfavourable non-Hodgkin lymphomas. Comparison of the Rappaport and Kiel classifications. Ada Pathol Microbiol Immunol Scand A 1983; 91:435-43. Horning SJ, Doggett RS, Warnke RA, Dorfman RF, Cox RS, Levy R. Clinical relevance of immunologic phenotype in diffuse large cell lymphoma. Blood 1984; 63:1209-15. Strauchen JA, Young RC, DeVita VT Jr, Anderson T, Fantone JC, Berard CW. Clinical relevance of the histopathological subdassification of diffuse 'histiocytic' lymphoma. N EnglJ Med 1978; 299:1382-7. Armitage JO, Dick FR, Platz CE, Corder MP, Leimert JT. Clinical usefulness and reproducibility of histologic subdassification of advanced diffuse histiocytic lymphoma. Am J Med 1979; 67: 929-34. Barcos M, Herrmann R, Pickren JW, etal. The influence of histologic type on survival in non-Hodgkin's lymphoma. Cancer 1981; 47:2894-900. Whitcomb CC, Cousar JB, Flint A, et al. Subcategories of histiocytic lymphoma: associations with survival and reproducibility of classification. The Southeastern Cancer Study Group experience. Cancer 1981; 48: 2464-74. Herrmann R, Barcos M, Stutzman L, et al. The influence of histologic type on the incidence and duration of response in non-Hodgkin's lymphoma. Cancer 1982; 49: 314-22. Stryker JA, Abt AB, Eyster ME, Lowman JG. Histopathologic subdassification of diffuse histiocytic lymphoma and response to therapy. Radiology 1982; 142: 501-6. Nathwani BN. Classifying non-Hodgkin's lymphomas. Monogr Pathol 1987; 18-80. Stansfeld AG, Diebold J, Noel H, et al. Updated Kiel classification for lymphomas [letter]. Lancet 1988; 1: 292-3.
50. Lennert K and Feller AC. In: Anonymous, Histopathology of non-Hodgkin's lymphomas (based on the updated Kiel classification). Berlin: Springer-Verlag, 1990. 51. Gascoyne RD. Pathologic prognostic factors in diffuse aggressive non-Hodgkin's lymphoma. Hematol Oncol Clin North Am 1997; 11: 847-62.
52. Gascoyne RD, Adomat SA, Krajewski S, et al. Prognostic significance of Bcl-2 protein expression and Bcl-2 gene rearrangement in diffuse aggressive non-Hodgkin's lymphoma. Blood 1997; 90: 244-51. 53. Hermine 0, Haioun C, Lepage E, et al. Prognostic significance of bcl-2 protein expression in aggressive non-Hodgkin's lymphoma. Groupe d'Etude des Lymphomes de I'Adulte (GELA). Blood 1996; 87: 265-72. 54. Hill ME, MacLennan KA, Cunningham DC, et al. Prognostic significance of BCL-2 expression and bcl-2 major breakpoint region rearrangement in diffuse large cell non-Hodgkin's lymphoma: a British National Lymphoma Investigation Study. Blood 1996; 88: 1046-51. 55. Davis RE, Dorfman RF, Warnke RA. Primary large-cell lymphoma of the thymus: a diffuse B-cell neoplasm presenting as primary mediastinal lymphoma. Hum Pathol 1990; 21:1262-8. 56. Perrone T, Frizzera G, Rosai J. Mediastinal diffuse largecell lymphoma with sclerosis. A clinicopathologic study of 60 cases. Am J Surg Pathol 1986; 10:176-91. 57. Haioun C, Gaulard P, Roudot-Thoraval F, et al. Mediastinal diffuse large-cell lymphoma with sclerosis: a condition with a poor prognosis. Am J Clin Oncol 1989; 12: 425-9. 58. Scarpa A, Borgato L, Chilosi M, et al. Evidence of c-myc gene abnormalities in mediastinal large B-cell lymphoma of young adult age [see comments]. Blood 1991,78:780-8. 59. Tsang P, Cesarman E, Chadburn A, Liu YF, Knowles DM. Molecular characterization of primary mediastinal B cell lymphoma. Am J Pathol 1996; 148: 2017-25. 60. Skinnider BF, Connors JM, Gascoyne RD. Bone marrow involvement in T-cell-rich B-cell lymphoma. Am J Clin Pathol 1997; 108: 570-8. 61. KoYH, Han JH, GoJH,et al. Intravascular lymphomatosis: a clinicopathological study of two cases presenting as an interstitial lung disease. Histopathology 1997; 31: 555-62. 62. Murase T, Nakamura S, Tashiro K, et al. Malignant histiocytosis-like B-cell lymphoma, a distinct pathologic variant of intravascular lymphomatosis: a report of five cases and review of the literature. BrJ Haematol 1997; 99: 656-64. 63. DiGiuseppe JA, Nelson WG, Seifter EJ, Boitnott JK, Mann RB. Intravascular lymphomatosis: a clinicopathologic study of 10 cases and assessment of response to chemotherapy. J Clin Oncol 1994; 12: 2573-9. 64. Wick MR, Mills SE. Intravascular lymphomatosis: clinicopathologic features and differential diagnosis. Semin Diagn Pathol 1991; 8: 91-101. 65. Ferry JA, Harris NL, Picker LJ, et al. Intravascular lymphomatosis (malignant angioendotheliomatosis). A B-cell neoplasm expressing surface homing receptors. Mod Pathol 1988; 1: 444-52. 66. Bhawan J. Angioendotheliomatosis proliferans systemisata: an angiotropic neoplasm of lymphoid origin. Semin Diagn Pathol 1987; 4:18-27.
54 Diffuse aggressive B cell lymphoma
67. Sheibani K, Battifora H, Winberg CD, etal. Further
demonstration of human herpesvirus-8: a case report.
evidence that 'malignant angioendotheliomatosis' is an angiotropic large-cell lymphoma. N EnglJ Med 1986; 314: 943-8. 68. Au WY, Shek WH, Nicholls J, Tse KM, Todd D, Kwong YL
Am J Surg Pathol 1998; 22: 493-9.
T-cell intravascular lymphomatosis (angiotropic large cell lymphoma): association with Epstein-Barr viral infection. Histopathology 1997; 31: 563-7. 69. Sepp N, Schuler G, Romani N, et al. 'Intravascular lymphomatosis' (angioendotheliomatosis): evidence for a T-cell origin in two cases. Hum Pathol 1990; 21:1051-8. 70. Glass J, Hochberg FH, Miller DC. Intravascular lymphomatosis. A systemic disease with neurologic manifestations. Cancer 1993; 71: 3156-64. 71. Bogomolski-Yahalom V, Lossos IS, Okun E, Sherman Y,
76. Horenstein MG, Nador RG, Chadburn A, et al. Epstein-Barr virus latent gene expression in primary effusion lymphomas containing Kaposi's sarcomaassociated herpesvirus/human herpesvirus-8. Blood 1997; 90:1186-91. 77. Karcher DS, Alkan S. Human herpesvirus-8-associated body cavity-based lymphoma in human immunodeficiency virus-infected patients: a unique Bcell neoplasm. Hum Pathol 1997; 28: 801-8. 78. Carbone A, Gaidano G. HHV-8-positive body-cavity-based lymphoma: a novel lymphoma entity. BrJ Haematol 1997; 97: 515-22. 79. Said W, Chien K, Takeuchi S, et al. Kaposi's sarcoma-
Lossos A, Polliack A. Intravascular lymphomatosis - an
associated herpesvirus (KSHV or HHV8) in primary
indolent or aggressive entity? Leuk Lymphoma 1998; 29: 585-93.
effusion lymphoma: ultrastructural demonstration of
72. Levin KH, Lutz G. Angiotropic large-cell lymphoma with
herpesvirus in lymphoma cells. Blood 1996; 87: 4937-43. 80. Cesarman E, Chang Y, Moore PS, Said JW, Knowles DM.
peripheral nerve and skeletal muscle involvement: early
Kaposi's sarcoma-associated herpesvirus-like DMA
diagnosis and treatment. Neurology 1996; 47:1009-11.
sequences in AIDS-related body-cavity-based lymphomas.
73. Demirer T, Dail DH, Aboulafia DM. Four varied cases of intravascular lymphomatosis and a literature review. Cancer 1994; 73:1738^5. 74. Uphoff CC, Carbone A, Gaidano G, Drexler HG. HHV-8 infection is specific for cell lines derived from primary effusion (body cavity-based) lymphomas. Leukemia 1998; 12:1806-9. 75. Hsi ED, Foreman KE, Duggan J, et al. Molecular and
N EnglJ Med 1995; 332:1186-91. 81. Nador RG, Cesarman E, Chadburn A, et al. Primary effusion lymphoma: a distinct clinicopathologic entity associated with the Kaposi's sarcoma-associated herpes virus. Blood 1996; 88: 645-56. 82. Matolcsy A, Nador RG, Cesarman E, Knowles DM. Immunoglobulin VH gene mutational analysis suggests that primary effusion lymphomas derive from different
pathologic characterization of an AIDS-related body
stages of B cell maturation. AmJ Pathol 1998; 153:
cavity-based lymphoma, including ultrastructural
1609-14.
7 T cell lymphoproliferative disorders AS JACK, SJ RICHARDS, KA MACLENNAN
Introduction
55
Nodal-based peripheral T cell lymphomas
60
Precursor T cell lymphoblastic leukemia/lymphoma
56
References
64
Predominately leukemic peripheral T cell lymphoproliferative disorders
57
INTRODUCTION T cells greatly outnumber B cells in the peripheral blood and also form a majority of cells in peripheral lymphoid tissue and in bone marrow. Despite this, T cell lymphoproliferative disorders are much less common than their B cell counterparts. The incidence of all types of T cell lymphoma and leukemia in the UK is around 2.0/ 100 000 per annum and, as such, constitute around oneseventh of the total number of lymphoproliferative disorders. The reason for the much higher rate of malignant transformation in B cells has not been satisfactorily explained but several factors may be important. B cells are produced throughout life by differentiation of lineage-committed precursors in the bone marrow. In contrast, de novo production of T cells in the thymus from bone mar row-derived precursors declines rapidly after the age of 40 years and T cell numbers in later life are maintained by division of existing peripheral lymphocytes. Peripheral T cells respond to antigenic stimulation by division and differentiation, but do not undergo further episodes of genetic recombination or somatic hypermutation as B cells do within germinal centers. This may result in a lower risk of a T cell sustaining an aberrant recombination event during antigen receptor gene rearrangement. T cell lymphoproliferative disorders were first recognized in the late 1970s using sheep red blood cell resetting technique.1 A few years later the introduction of the first monoclonal antibodies to CD 3 and other T cell surface molecules greatly increased the ease with which these tumors could be recognized.2 It is only in the past few years with the introduction of new antibodies, tyramide-mediated amplification and heat-mediated antigen retrieval that relatively detailed immunophenotypic
studies can be performed on fixed paraffin- or resinembedded tissues.3'5 The use of Southern blotting and later polymerase chain reaction (PCR)-based techniques made it possible to study T cell clonality in clinical samples.6"8 The widespread use of these techniques has raised interesting questions that have not been fully resolved about the boundaries between T cell malignancy and reactive inflammatory conditions. It is only recently that some consensus has begun to emerge as to the classification of T cell lymphoproliferative disorders. The National Cancer Institute Working Formulation did not subdivide lymphomas by immunophenotype. As a consequence, variable numbers of T cell lymphomas were included in several different categories. Until recently, the Working Formulation was used in most clinical trials throughout the world, and this has led to a paucity of good-quality data on the clinical behavior and response to treatment. T cell lymphomas were separately recognized in the Kiel classification.9 Subdivision of tumor types was based partly on cells size and morphology, and partly on the nature of reactive non-neoplastic elements in the tumor.10,11 This was much less robust and clinically relevant than the Kiel classification of B cell lymphomas and was criticized for lack of reproducibility.12 Like their B cell counterparts, T cell lymphomas were subdivided into high and low grade according to cell size, but these gradings did not correlate well with clinical behavior.13 Very few T cell lymphomas can be regarded as indolent, and the distinction between high and low grade, even if this could be done reproducibly, was of limited value in the selection of treatment.14 The Revised European-American Lymhoma (REAL) classification15 differs from other lymphoma classifications in that it is based on the recognition of clinical-
56 T cell lymphoproliferative disorders
pathological entities that are defined in terms of cell lineage morphology, genetic and clinical features. Using this approach, there is now an emerging consensus as to the classification of T cell and natural killer (NK) cell neoplasms, although the definition of entities is less clear than for B cell lymphomas. The starting point of the classification is the identification of tumors of precursor T cells from those peripheral immunocompetent T cells. Precursor T cell leukemia/lymphoma is a relatively homogenous entity that mainly occurs in children and young adults. Peripheral T cell malignancies derived from immunocompetent cells are very heterogeneous and occur mainly in the older adults. Peripheral T cell lymphoproliferative disorders can be subdivided according to their clinical presentation into those that are predominately leukemic, nodal-based disease and primary extranodal lymphomas (see Chapter 8). Although this is a convenient framework for discussing this group of tumors, it must be recognized that there is considerable overlap between these groups and these distinctions have limited clinical significance. Peripheral T cell leukemias, such as T cell prolymphocytic leukemia (T-PLL) often have significant levels of nodal or splenic disease. It is almost certainly the case that systemic nodal-based T cell lymphomas more frequently present with extranodal disease than their B cell counterparts and some tumors, such as the rare panniculitic T cell lymphoma, may be widely disseminated while remaining mainly extranodal. It is now recognized that a very small proportion of tumors previously considered to be T cell lymphomas are in fact true NK cell lymphomas. Most of these tumors appear to occur in the nose and surrounding tissues, although they may be found at other extranodal sites and occasional cases of NK cell leukemia are found.15'16 Improved understanding of the nature of NK cells is likely to facilitate the recognition of these tumors.
PRECURSOR T CELL LYMPHOELASTIC LEUKEMIA/LYMPHOMA All lymphoid cells arise from a committed lymphoid progenitor cell. This is absent in mice deficient in the IKAROS transcription factor. There is evidence that such cells may initially differentiate into B cell or T/NK progenitors. Some of the latter cells remain in the bone marrow to become NK cells, while others enter the thymus and become committed to the T cell lineage.17-24 The first identifiable lineage-committed T cells express CD 1, CD 2, CD 5 and cCD 3. As gene rearrangement occurs, the cells acquire CD 4 and CD Salpha. Non-reactive and selfreactive cells are deleted by positive and negative selection. As the remaining cells enter the thymic medulla, they lose CD 1 expression, switch from CD 45RO to RA and express either CD 4 or CD 8alpha/beta to become immunocompetent peripheral T cells. A small proportion
of T cells are produced by thymic-independent pathways mostly in the gastrointestinal tract. These cells characteristically have express CD 8alpha/alpha.25-30 Precursor T cell lymphoblastic leukemia/lymphoma consists of cells with phenotypic and genotypic features resembling normal thymocytes. There is variation between patients in the extent of marrow involvement and solid tumor formation, but there appears to be little merit in attempting to make an arbitrary distinction between T cell acute lymphoblastic leukemia (T-ALL) and T cell lymphoblastic lymphoma. For simplicity T-ALL will be used as the preferred term. As with B lineage ALL, T-ALL is a disease of children and becomes very rare in later life. T-ALL accounts for around 10 per cent of all childhood acute lymphoblastic leukemias and 25 per cent of cases in adults. It has been suggested, although far from proven, that the incidence may be relatively constant throughout the world in contrast to B-lineage ALL, where the peak in early childhood may be a feature of developed countries. This would suggest a lesser role for environmental factors in the pathogenesis of T-ALL. The only strong risk factor for the development of T-ALL that has been identified is ataxiatelangiectasia. The ATM gene that is mutated in this condition is involved in cell-cycle arrest of cells with sublethal DNA damage.31-33 The reason why precursor T cells are particularly susceptible to loss of this function is not clear. T cell neoplasms also develop in mice deficient in ATM. The blast cells in T-ALL are relatively large cells, often with highly convoluted irregular nuclei, distinct nucleoli and a moderate amount of cytoplasm (Plate 49). The cytoplasm stains positively with PAS and displays dotlike reactivity for acid phosphatase, although these are now obsolete diagnostic tests. The cellular morphology corresponds in most cases to the FAB L2 morphological type but it is not possible to predict the immunophenotype of ALL from the morphology with any degree of certainty. There is considerable variation in the immunophenotype of individual cases of ALL reflecting the complexity of normal thymic differentiation. A small number of cases have features of prothymocytes with expression of CD 7, CD 2 and terminal deoxynucleotidyl transferase (tdt). These cases are the most likely to show CD 34 expression, which is much less common in T lineage than B lineage ALL. It has been suggested that these patients have a poorer prognosis.34-36 The most common immunophenotype corresponds to the common thymocyte stage of normal differentiation. These cells coexpress CD 4 and CD 8 together with cCD 3, CD 2, CD 5 and CD 7. The cells lack surface expression of the T cell receptor and the presence of tdt is evidence of continuing gene rearrangement. In most of these cases CD la will be present. A third group demonstrate a late thymocyte phenotype with loss of tdt expression and some of the cells showing evidence of surface antigen receptor expression. The level of CD 2 expression may also be an
Predominately leukemic peripheral T cell lymphoproliferative disorders 57
independent prognostic factor within this group.37'38 Around 15 per cent of patients with T-ALL will show expression of the myeloid-associated antigens CD 13 and CD 3339 and occasional cases have weak expression of CD 20.40 This is not taken as evidence of biphenotypic leukemia. The influence of CD 13 and CD 33 expression on prognosis appears to be small. Cases of T-ALL expressing CD 10 appear to have a poor prognosis. In ALL as a whole the key clinical presenting features used to assign risk are age (1-9 years; favorable), blast cell count (<50 x 109/1; favorable) and the presence of solid tumors (unfavorable). In most clinical trials, patients with T-ALL are found to have a higher incidence of unfavorable risk factors. These patients tend to be older, with higher blast cell counts. Mediastinal/thymic involvement is a characteristic feature of T-ALL, and lymphadenopathy and hepatosplenomegaly are often present.41 Children with T-ALL are about twice as likely to be placed in a poor risk category than those with Blineage ALL and disease-free survival at 5 years is around 50 per cent. In adults, T-ALL has a better outcome than B-lineage disease. This is most likely due to the higher frequency of bcr-abl translocations in the B cell group. ALL can be classified according to the degree of aneuploidy present. Most cases of T ALL are pseudodiploid, implying a normal chromosomal number with structural abnormalities. A subset of cases with a poor outcome has a near tetraploid chromosome content. The most common non-random chromosomal abnormality in T-ALL involves deregulation of expression of TAL-1, which is present in 20-30 per cent of cases.42-44 This gene is a critical regulator of hematopoiesis in mice but is not normally expressed in lymphoid cells. The gene is capable of forming alternative transcripts that form heterodimers with other DNA binding proteins. The most common rearrangement leading to deregulation of TAL-1 in cases of childhood, but not adult, T-ALL, is a small interstitial deletion on chromosome 1 that leads to the juxtaposition of TAL-1 to the regulatory elements of the SIL gene that is normally expressed in T cells.43,46 The sequences flanking the deletion are homologous to recombination sequences of the TCR genes, suggesting that generation of the abnormality is closely related to TCR rearrangement and is mediated by recombinase activity. In around 3 per cent of cases, TAL-1 is activated by a t( 1:14) involving the TCR alpha/delta locus or by a t(l;7) involving the TCR beta gene.44,47-49 Antibodies to the TAL-1 protein have been produced and abnormal expression of the gene can be shown by immunocytochemistry to be present in cases without evidence of a TAL-1 rearrangement.50 Dysregulation of TAL-1 by whatever mechanism, has not been found to have strong prognostic significance." In addition to TAL-1, a large number of transcription factor genes have been found to be activated by translocations involving the TCR gene loci on chromosome 14 and 7. It is reported that a translocation of this type is present
in 40-50 per cent of cases of T-ALL. The best characterized is the t( 11; 14), which appears exclusive to T-ALL and leads to activation of the tcl-2 gene.52-55 This translocation also appears to be due to recombinase activity.56 The wide range of transcription factors involved in T-ALL-associated translocations belong to three broad groups classified on the basis of structural motifs that mediate binding to DNA or other proteins. The helix HLH group includes TAL-1, TAL-2 and c-wyc.57^0 RBTN1 and RBTN2 are LIM-domain-containing proteins, which interact with TAL-161-63 and HOX11 is a homeoboxcontaining gene.64 Deregulation of HOX11 may be the commonest abnormality in adult T-ALL and is associated with a more favorable outcome. In some cell lines, combinations of abnormalities, such as TALI and RBTN1, may act synergistically in the dysregulation of cell growth.65 In addition to structural chromosomal abnormalities, a variety of specific tumor suppressor gene abnormalities have been described. Inactivation of the cell cycle regulator Rb appear to be the most common, but loss of activity of a range of cyclin-dependent kinase inhibitors have also been described in some cases and may have prognostic significance.66-68 P53 mutations are rare at presentation but are more common in relapse cases.69,70 Individual cases and studies in transgenic mice have also suggested that abnormalities of Ets-1 and IKAROS, both critical genes in lymphoid cell development, may also have a role in the pathogenesis of T-ALL.
PREDOMINATELY LEUKEMIC PERIPHERAL T CELL LYMPHOPROLIFERATIVE DISORDERS Large granular lymphocytosis Large granular lymphocytes (LGLs) are identified by their size and the presence of cytotoxic granules in their cytoplasm (Plate 50). In normal individuals following viral or other infections the LGL fraction will be a mixture of cytotoxic T cell and NK cells. Although morphologically similar, these cells belong to separate lineages. Most cytotoxic T cells recognize the presentation of viral or other endogenous proteins presented on class I major histocompatibility complex (MHC) molecules. In contrast, NK cells recognize cells that have lost expression of MHC class I as a result of infection by some types of virus or neoplastic transformation. Both types of cell contain cytotoxic granules capable of perforating the cell membrane or inducing apoptosis in the target cell. Persistent large granular lymphocytosis is a relatively common finding in older patients. In the majority of patients, this will be detected on a routine blood sample; the lymphocyte count is rarely above 10 x 109/1 and is usually considerably less.71-73 Associations with many diseases of the elderly, such as ischemic heart disease, are likely to be coincidental. However, large granular
58 T cell lymphoproliferative disorders
lymphocytosis is much more common in patients with rheumatoid arthritis and Felty's syndrome. Asymptomatic patients with large granular lymphocytosis may be more likely to have rheumatoid factor and a few patients develop rheumatoid disease after diagnosis of large granular lymphocytosis.74'75 The second important association is with chronic neutropenia.76 In these cases the neutropenia appears to be due to reduced marrow production, possibly as a result of cytokine production by the T cells, although the exact mechanism is poorly understood. Thirdly, LGL proliferations may be associated with underlying B cell malignancy77 and occasionally other non-hematological malignancies.78 In contrast to acute reactive states where a mixture of T cells and NK cells are found, persistent large granular lymphocytosis is almost always due to an expansion of cytotoxic T cells. There is considerable variation between patients in the immunophenotype of the LGL population. Almost all cases express CD 3/TCR alpha/beta and, of these, 60 per cent will be CD 8 positive. The remainder will express CD 4 sometimes in association with weak CD 8.79,80 LGL expansions consisting of CD 3/TCR gamma/delta are rare. A characteristic of LGLs is the expression of the NKa markers CD 16, CD lib, CD 56 and CD 57 in various combinations. CD 16 in particular is strongly associated with rheumatoid arthritis.81 Large granular lymphocytosis due to pure NK cell proliferation is very uncommon. The question of monoclonality has been extensively investigated in large granular lymphocytosis. The immunophenotype CD 8+, CD 16+, CD 56- is strongly predictive of the presence of monoclonality using Vgamma PCR.82 Cases with CD 4 and 8 co-expression are also likely to show monoclonality. It must be emphasized that the presence of monoclonality does not form the basis of a distinction between benign and malignant LGL proliferations. Monoclonality can occur frequently in reactive T cell proliferations, such as infectious mononucleosis, and there is no difference in the prognosis between cases of large granular lymphocytosis with monoclonality and those without. Genetic abnormalities appear to be uncommon, although recently a case showing a t(3;5) has been described.83 It is doubtful whether it is justified to use the term leukemia to describe LGL proliferations. Except in a very few cases, patients do not have progressive disease and tissue or bone marrow infiltration is also very rare. A high level of CD 56 expression maybe a feature of the more aggressive variant.84 In most cases no treatment is required with the exception of those with severe neutropenia who may require granulocyte colony stimulating factor (G-CSF).
T cell prolymphocytic leukemia T cell prolymphocytic leukemia is a very rare entity with an incidence of 1 in 1000 000 per year. The term
prolymphocyte is used to describe the morphological features seen in most cases, and has no implication for the lineage or state of differentiation of the tumor cells. The cells are of medium size with a round nucleus and a large prominent nucleolus (Plate 51). The cell cytoplasm is agranular. Almost every case has the immunophenotype of a CD 4+ peripheral T cell with expression of CD 45RO or CD 45RA and RO. Some cases may have CD 4 and CD 8 co-expression, and there is variable expression of T cell activation markers, such as CD 38 and HLA-DR. CD 4-, CD 8+ cases are rare.85 Expression of NKa markers are not seen in T-PLL. The presence of inv 14(qllq32) is a characteristic feature of this type of leukemia.86'87 A possible role for inactivation of the ATM tumor supressor gene in the pathogenesis of T-PLL has been suggested.88-90 Occasional cases are seen in which a small proportion of the cells express tdt. This raises the possibility that T-PLL may consist of cells at the immediate post-thymic stage of differentiation. The clinical features of T-PLL are distinctive with a very high lymphocyte count that may rise rapidly even within a few days. The peripheral lymphocyte count often exceeds 100 x 109/1 and may rise to 1000 x 109/1. Bone marrow replacement, bulky nodal disease and splenomegaly are usually present, and extranodal involvement may be seen. The overall tumor bulk at presentation is very high. As may be expected from these clinical features, the prognosis is very poor with a median survival around 6-7 months. Until recently, there was little evidence that any treatment was effective but a recent study has suggested that anti-CD 52 can rapidly reduce the lymphocyte count.91 It has been suggested that there is a subgroup of patients with a more stable clinical course. One feature of these patients appears to be expression of CD 45RA. There are a small number of patients who have similar clinical features and outcome, but in whom the tumor consists of small lymphocytes with condensed nuclear chromatin and small indistinct nucleoli. These cases invariably have a CD 4+ peripheral T cell immunophenotype and typically are associated with lymphadenopathy and hepatosplenomegaly. There seems little merit at present in classifying these cases separately. In particular, use of the term T cell chronic lymphocytic leukemia (T-CLL) would seem to be misleading since the clinical features are certainly not those of a chronic leukemia analogous to B cell chronic lymphocytic leukemia (B-CLL).
Adult T cell leukemia/lymphoma In spite of its rather uninformative name, adult T cell leukemia/lymphoma (ATLL) is a specific clinical and pathological entity. The key defining feature of this condition is the association with human T cell leukemia/lymphoma virus type 1 (HTVL-1). This virus
Predominately leukemic peripheral T cell lymphoproliferative disorders 59
is endemic in the islands of the Caribbean and in Southwest Japan. Areas of high incidence have also been described in South America and sporadic cases are found world-wide.92'95 As well as the association with T cell leukemia and lymphoma, this virus appears to play a role in the pathogenesis of tropical spastic paraparesis and possibly a variety of systemic inflammatory disorders.96 HTLV-1 is transmitted through infected cells in transfused blood, breast milk and by sexual contact.97'98 HTLV-1 is a retrovirus that does not carry a conventional viral oncogene and the incidence of ATLL in those infected is low. Viral infection of CD 4+ T cells appears to be an essential first step but other genetic events are required before the onset of overt malignant disease. Carriers of the virus have a polyclonal T cell expansion99 and the emergence of a monoclonal population as judged by the viral integration pattern increases the risk of subsequent malignant progression. The mechanism by which the virus facilitates neoplastic transformation of T cells is not fully understood. It appears likely that one of the main effects of the virus is through the viral tax gene, which is capable of transactivation of a wide variety of cellular genes coding for cytokine and adhesion molecules. Transfection experiments show that this gene is capable of immortalizing T cells in vitro. One of the most important genes activated by tax is IL-2R (interleuk-2 receptor), which may prolong the life of the cell increasing the probability that critical genetic abnormalities will develop. Among the abnormalities described in ATLL are constitutive activation of the JAK-STAT signaling pathway100 and alterations in the tumor suppresor gene pi6.101 Overexpression of IL-4Rhas been associated with aggressive disease.102 A wide range of other cellular genes are activated by HTLV-1. These include the cytokines IL-4, IL-10, IL-1, G-CSF and MIP-1.103,104 Virally infected cells may express high levels of the adhesion molecules L-selectin (CD 62 L) and HML-1 (CD 103), and it has been suggested that this is the basis of the widespread dissemination of tumor cells seen in many patients.105,107 The cellular features of ATLL are relatively distinctive. The characteristic cells are medium-sized lymphocytes that show a high degree of nuclear pleomorphism including a proportion of cells with polylobated nuclei. Large lymphoid cells with more abundant cytoplasm and prominent nucleoli may also be seen in tissue biopsies. The tumor cells are CD 4+, CD 45RO+ peripheral T cells. The presence of IL-2R identified by the anti-CD 25 is always present. Other more variable features include weak or absent surface TCR alpha/beta/CD 3 expression, weak expression of CD 8 and CD 38 expression. The rate of cell proliferation defined by the cell-cycle marker Ki67 has been suggested as an important prognostic factor. ATLL causes diffuse infiltration of lymph nodes with effacement of the normal architecture. The cells present in the node may be more variable in morphology than those seen in the peripheral blood.
There are no reliable morphological features that distinguish ATLL from other peripheral T cell malignancies. In endemic areas the diagnosis will not be a problem. In non-endemic areas, cases may be classified as T-PLL, Sezary syndrome or peripheral T cell lymphoma unless HTLV-1 serology is carried out. Attempting to subclassify or grade ATLL on the basis of cellular morphology is of little value. ATLL occurs in older patients with a peak in the fifth decade. A number of distinctive clinical syndromes are recognized.108 The majority of patients have an acute rapidly progressive clinical course. The lymphocyte count is usually raised and can be very high. In some patients this may be associated with anemia or thrombocytopenia. Involvement of lymph nodes is clinically detectable in most of these patients, and infiltration of skin, liver and spleen is also present in around one-third of cases. A subset of patients with acute presentation may have a predominately lymphomatous pattern of disease with only low levels of peripheral blood and marrow involvement. A subset of these patients may show an angiocentric and angiodestructive pattern.109 One of the most typical features of ATLL is the presence of hypercalcemia, which is seen in 75 per cent of patients.110 Only a few of these patients have lytic bone lesions and the mechanism is thought to involve aberrant production of IL-1 and parathyroid hormone (PTH)-related peptide by the tumor cells.111 The conventional treatment of ATLL has been CHOP (cyclophosphamide, doxorubicin, vincristine and prednisone) or similar types of combination chemotherapy, but with an almost uniformly poor outcome. The median survival from diagnosis is around 12 months. Two other more indolent clinical syndromes are recognized. Patients with chronic ATLL have peripheral blood and marrow involvement, but no clinically detectable solid tumor. The overall tumor bulk estimated by serological tests is low with lactate dehydrogenase (LDH) within twice the normal upper limit. These patients progress to the acute form of the disease, although this can take a few years. The Ki67 fraction is reported as being predictive of the time to progression. The presence of p53 abnormalities may also be associated with progressive disease. In smouldering ATLL, the peripheral T cell count is low or normal with less than 3 per cent of the T cells showing abnormal morphology. The main clinical feature is the presence of cutaneous infiltration. When biopsied, these lesions show a dense dermal infiltrate of cells similar to those described above. In some patients the histological features resemble mycosis fungoides. In endemic areas, serological studies for HTLV-1 should always be carried out before a definitive diagnosis of mycosis fungoides is made.112 Skin lesions in the smouldering phase are responsive to steroids. Small numbers of circulating morphologically atypical cells may be found in carriers with no other clinical features.
60 T cell lymphoproliferative disorders
Sezary syndrome The term Sezary syndrome is used to described a CD 4+ peripheral T cell leukemia in association with erythroderma and lymphadenopathy. This probably includes two distinct entities. In patients with typical cutaneous T cell lymphoma of mycosis fungoides type, small numbers of circulating tumor cells can be detected in the peripheral blood, particularly where sensitive flow cytometric or PCR tests are used. In the terminal stages of disease, some patients may progress to overt leukemia, sometimes defined as tumor cells greater than 1 x 109/1 or more than 10 per cent atypical lymphocytes. The more common form of Sezary syndrome is where patients present with a T cell leukemia and generalized skin infiltration, but with no preceding history of mycosis fungoides. These patients often have a very high lymphocyte count and evidence of generalized lymphadenopathy. These cases of de novo Sezary syndrome have many features in common with T-PLL, which may also involve the skin.113,114 In both forms of Sezary syndrome, the leukemic cells are of small to intermediate size with highly convoluted nuclei. All cases have the immunophenotype of an activated CD 4+,TCR alpha/beta, CD 45RO+ peripheral T cell, sometimes with loss of CD 7 expression. Phenotypic differences may be reflected in the functional properties of the neoplastic T cells.115"117 The prognosis of Sezary syndrome, whether de novo or secondary to mycosis fungoides, is generally poor.118
NODAL-BASED PERIPHERAL T CELL LYMPHOMAS Peripheral T cell lymphomas found in lymph nodes are morphologically and immunophenotypically complex leading to the problems in classification described above. At present there are three main entities with distinctive cellular and clinical features. Anaplastic large cell lymphoma of T cell type has in the past been defined in terms of morphology and immunophenotype. More recently, the presence of the t(2;5) with abnormal expression of ALK protein has become a key feature. As the definition has become clearer it has become apparent that this group has a significantly better prognosis than other nodal peripheral T cell lymphomas. The second distinctive entity is angioimmunoblastic T cell lymphoma, which is characterized by proliferation of stromal elements and B cells, and is associated with various autoimmune phenomena. The remaining T cell lymphomas, although they are morphologically heterogeneous, are currently considered as a single group known variously as common or unspecified type. It is likely that, in the future, other distinctive entities may be separated from this group on the basis of cytogenetic or other
features. Most cases of nodal peripheral T cell lymphoma are of common type, up to one-third are anaplastic lymphomas and angioimmunoblastic lymphoma accounts for a further 15-20 per cent.
Anaplastic lymphoma of T cell lineage The history of the term anaplastic large cell lymphoma illustrates one of the key problems encountered by those new to hematopathology. A diagnostic term may be in use for many years during which time the underlying concept may change almost beyond recognition. This can make valid retrospective comparisons very difficult. Subsequent developments have also shown the problem inherent in defining a diagnostic entity on the basis of reactivity with a single antibody. Anaplastic large cell lymphoma was first defined in terms of reactivity with the anti-CD 30 antibody Ki-1 that had been raised against the Hodgkin's cell line L428. Ki-1 lymphoma was a synonymous term for many years. CD 30 is a member of the tumor necrosis factor (TNF) receptor superfamily and is expressed on normal activated T and B cell and, of course, Reed-Sternberg cells in classical Hodgkin's disease. In vitro studies suggest that CD 30, like other members of the TNF superfamily may have a role in the induction of apoptosis and the release of cytokines.93 Circulating soluble CD 30 may block induction of tumor cell death by CD 30 ligandbearing lymphocytes.119 The types of tumor included in the Ki-1 lymphoma category had been previously diagnosed as histiocytic lymphoma, malignant histiocytosis, Hodgkin's disease, various types of large cell lymphoma and even metastatic carcinoma. CD 30 expression appeared to correlate strongly with highly pleomorphic cellular morphology and a sinusoidal pattern of nodal infiltration. Since these original descriptions were given, the definition of this condition has undergone almost continuous evolution in terms of morphological features, immunophenotype, cytogenetics and the separate recognition of the primary cutaneous CD 30-positive lymphoma. The common type of ALCL consists of large lymphoid cells with an eccentric bean-shaped nucleus and a welldemarcated Golgi apparatus. This cell is sometimes referred to as the 'hallmark' cell.120 Multinucleated and other bizarre forms may also be present (Plate 52). This cellular morphology is often associated with invasion of the marginal sinus of the lymph node (Plate 53). However, it is now recognized that the morphological spectrum of this tumor is much wider and only about 60-70 per cent of cases have a common type of morphology. Although hallmark cells are a key defining feature of the condition and correlate closely with the t(2;5), in some cases they may be present in relatively small numbers. In the small-cell variant, the predominant population consists of small to intermediate-sized
Nodal-based peripheral T cell lymphomas 61
lymphoid cells with cerebriform nuclei (Plate 54). These cells are considered to be part of the neoplastic clone. In other variants, large numbers of reactive epithelioid histiocytes or neutrophils maybe present.121'124 There is also a rare sarcomatoid variant where the cells may be spindle shaped or have other bizarre morphological features not normally seen in lymphoid cells. The term Hodgkin's-like ALCL has also been proposed, although it is now clear that most of these cases are part of the spectrum of classical Hodgkin's disease. With the increasing development of cell markers, it became apparent that many cases included in the original description of Ki-l+ ALCL were of B cell origin, with strong expression of CD 20, CD 19 and other B cell markers. These cases of CD 30+ B cell lymphoma are part of the spectrum of diffuse large B cell lymphoma (DLBCL) and do not appear to be clinically different from C30-negative DLBCL. They should no longer be classified as ALCL. The remaining cases have a T cell or null phenotype. Around 50 per cent of cases have expression of CD 3 or CD 2, although the proportion with definitive evidence of T cell differentiation is increasing with improvements in immunocytochemistry techniques. Most null cases can be shown to have a clonal T cell gene rearrangement. The term ALCL is, therefore, now used exclusively to describe a subset of T cell lymphoma. The presence of a T cell phenotype is a key feature in the distinction between ALCL and Hodgkin's disease. Some cases of Hodgkin's disease have in the past been reported as showing evidence of T cell differentiation; however, recent studies in the pathogenesis of Hodgkin's disease suggest that the term should be probably reserved for tumors of B lineage. A total of 60 per cent of cases of ALCL express CD 45 and, when present, this is also a useful feature distinguishing this tumor from classical Hodgkin's disease in which this marker is typically absent. The expression of epithelial membrane antigen is a relatively constant feature of ALCL and this is not seen in classical Hodgkin's disease. Conversely, CD 15 expression is rare in ALCL. A wide range of T cell activation markers including CD 71, class II MHC and CD 25 are commonly seen in ALCL and cells frequently contain cytotoxic granules as demonstrated by the presence of perforin, granzyme B or perform.125 A very important observation in clarifying the nature of ALCL has been the description of the t(2;5)(p23;q35). This produces a fusion protein containing elements of anaplastic lymphoma kinase (ALK) and nuclephosmin. The resulting product has tyrosine receptor kinase activity. ALK may also be deregulated by variant translocations including t(l;2)(q25;p23) and inv(2)(p23;q35). These translocations can be identified by cytogenetics, fluorescence in situ hybridization (FISH) or reverse transcription PCR (RT-PCR). The ALK protein is not expressed in normal hematolymphoid cells and its presence, demonstrated by immunocytochemistry, is in most cases indicative of the translocation.126 The exception is a
recently described rare B cell lymphoma variant in which the p80 fusion protein is present without the translocation.127 In most cases, the protein will be found in the cytoplasm and nucleus (Plate 55). A recent study has suggested that the absence of nuclear staining correlates with variant translocation not involving the nucleophosmin gene.120 The proportion of cases classified as ALCL reported to contain the t(2;5) varies widely from 15 to 60 per cent.128'130 The reason for these differences lies in the individual composition of the series. These and other studies show that the translocation is much more frequent in younger patients with disseminated predominately nodal disease.131'132 Several studies have failed to show a prognostic difference between CD 30-positive and other types of large cell lymphoma.117'118 However, tumors showing expression of ALK appear to have a more favorable outcome.132"136 These differences in clinical features and outcome have led to the proposal that the presence of the translocation should be the defining feature of the disease entity.120 While there is clearly a strong case for adopting this proposal there remains the problem of small numbers of cases that have identical morphology and immunophenotype but lack evidence of ALK dysregulation. A further problem in the definition of ALCL has been the relationship between nodal systemic disease and primary cutaneous CD 30-positive lymphoma. Lymphomatoid papulosis is characterized by relapsing selfhealing cutaneous nodules. Two forms of this condition are described. The rare type A form has morphological features similar to mycosis fungoides but with distinctive clinical features. In the type B form the nodules consist of an inflammatory background population admixed with highly pleomorphic large CD 30-positive cells, some of which may resemble Reed-Sternberg cells.137 Typically these cells have a more normal T cell phenotype than is seen in ALCL and do not show expression of ALK protein.130,131,138 These cells are monoclonal and in a few cases it has been possible to show that recurrences are derived from the same clone.139 In a small and as yet undefined proportion of patients, there is progression of lymphomatoid papulosis to a non-regressing CD 30positive lymphoma.140 Similar tumors may also arise in patients with no previous history of lymphomatoid papulosis. Most studies have shown that the prognosis of this group of patients is excellent with few recorded cases of death from disease.141 One of the problems in accurately defining this group of patients is that systemic ALCL, like other types of T cell lymphoma, frequently involves the skin either at presentation or subsequently, and cases of apparently localized cutaneous disease may be encountered that subsequently have a very aggressive clinical course. The presence of ALK or a t(2;5) may be helpful in the differential diagnosis of a lesion, although the absence of ALK in a CD 30-positive T cell lymphoma presenting in the skin should not lead to the assumption that this is a primary cutaneous lymphoma.
62 T cell lymphoproliferative disorders
Angioimmunoblastic T cell lymphoma Angioimmunoblastic lymphadenopathy was first described in the early 1970s and at that time was considered to be a reactive condition. Many series from that time contained cases that would now be considered to be angioimmunoblastic T cell lymphoma mixed with patients with lymphadenopathy caused by hypersensitivity to drugs such as phenytoin or viral infections. This led to considerable confusion about the prognosis and appropriate management of such patients. There are a number of key features that must be present in a lymph node biopsy to make a firm diagnosis of angioimmunoblastic T cell lymphoma (AIL). As with all types of lymphoproliferative disorders, the nodal architecture should be replaced by lymphoma. A point of particular importance in the case of AIL is the disruption of normal B cell development in the node. A greatly expanded population of plasma cells, preplasma cells and sometimes B-immunoblasts is present but with the absence of organized B cell follicles and reactive germinal centers. In a few cases a significant monoclonal population of B cells may be detected by PCR. Only in very rare cases are intact B cells follicles with reactive germinal centers present.142 A further key feature is loss of definition of the peripheral sinus of the node with spread into the adjacent connective tissues. Again this would not be expected in benign conditions. The T cell population in AIL is usually highly pleomorphic and includes small lymphocytes, intermediate sized cells with irregular nuclei, which are usually the predominant cell type, and large blast cells. In some cases the blast cells have abundant clear cytoplasm (Plate 56). Occasionally cells resembling Reed-Sternberg cells may be present but Hodgkin's disease is not usually a feasible differential diagnosis. The immunophenotype of the neoplastic cells is usually that of a normal CD 4+, CD 45RO+ peripheral T cell; data on aberrant phenotypes are sparse. Variable numbers of CD 8+ T cells are present and these are assumed to be non-neoplastic. The rate of cell proliferation, as defined by Ki-67, is usually relatively low. In most cases that satisfy the strict morphological criteria of this condition, it would be expected that a clonal T cell gene rearrangement would be detected by PCR. In many cases Epstein-Barr virus can be detected in the T cell population. The virus is usually present in a small minority of the cells and its significance to the pathogenesis of the condition remains uncertain.143-145 It has been suggested that, in AIL and other T cell lymphomas, abnormalities of the viral LMP-1 gene may be associated with a poorer prognosis. Chromosomal abnormalites are frequently found in the neoplastic T cells. The most common are trisomies of chromosome 3,5 and x.146'147 A feature of AIL appears to the presence of multiple subclones showing variable cytogenetic features within the same tumor.148
One of the defining features of AIL is the proliferation of stromal elements within the node. The most striking feature on routinely stained sections is the extensive branching vascular networks that consist mainly of large endothelial cells, similar to those seen in high endothelial venules, surrounded by a greatly thickened basement membrane (Plate 57). This can be demonstrated using a periodic acid-Schiff (PAS) stain. Proliferation of normal high endothelial venules may occur in the paracortex of reactive lymph nodes and abnormal vessel morphology is an important feature in distinguishing AIL from benign conditions. The second main stromal abnormality in AIL is the proliferation of follicular dendritic cells. These cells, which are of non-hemopoietic origin, normally display antigens to B cells in the centrocyte-rich zone of the germinal center. In AIL these cells are present as extensive diffuse networks (Plate 58). These may be associated with ill-defined B cell aggregates but organized germinal centers are not seen. The presence of this feature is not apparent on hematoxylin and eosin section but can be readily demonstrated using a variety of antibodies, including CD 21, CD 35 or CD 23. The final stromal element is the presence of a general increase in collagen and other extracellular matrix elements. As a result of this, parts of the tumor may appear hypocellular in comparison with normal lymphoid tissue. The final key morphological feature is the presence of increased numbers of other non-neoplastic cell types. There is almost always significant eosinophilia and epithelioid macrophages may be present, sometimes forming microgranulomas. The clinical features of AIL are distinctive and form an integral part of the diagnosis. Almost all patients have generalized nodal disease sometimes with hepatosplenomegaly. A wide range of systemic effects, including arthropathy, skin rashes and neurological effects are described.149'150 However, in many cases, the systemic effects may appear disproportionate to the bulk of the tumor determined by imaging studies. Fever and raised plasma viscosity are very common in these patients. In most cases the skin lesions do not appear to be due to direct tumor infiltration, although a pre-AIL state presenting as a skin rash has been described.151 Gastrointestinal ulceration may also be found and, in a few patients, gastrointestinal symptoms maybe severe. Again, endoscopic biopsies do not usually show infiltration by tumor. One of the most common systemic manifestations of disease is the presence of hypergammaglobulinemia reflecting the plasmacytosis seen in lymph node biopsies. This is likely to reflect excess IL-6 production by the tumor cells.152 In a few cases circulating plasma cells may be found and occasional patients may present with extramedullary plasmacytomas in the skin or other sites, or associated B cell lymphomas may be found.153 A paraprotein may be present at presentation or develop during the course of the illness. Around 25 per cent of patients have a Coomb's positive hemolytic anemia and a wide
Nodal-based peripheral T cell lymphomas 63
variety of other autoantibodies may be seen. Lymphopenia is often present but in some patients there may be a large granular lymphocytosis. Examination of the bone marrow may show a variety of abnormalities in patients with AIL. Focal infiltration by tumor may be present, although marrow replacement is not common. Erythroid hyperplasia may be present in patients with haemolytic anemia. In other patients there may be suppression of normal hematopoiesis. Many of the cellular and clinical features described above may be found in other types of peripheral T cell lymphoma. The accurate definition of AIL relies on demonstrating the pattern as a whole. If this is done, it is unlikely that confusion with reactive states will be a problem in most cases. When this approach to diagnosis is used, it is clear that AIL has a relatively poor prognosis. There appears to be little merit in attempting to treat these patients with steroids or alkylating agents. Using combination chemotherapy, only a minority of patients will have a sustained complete remission and the median survival appears to be less than 2 years. In addition to the usual prognostic factors it has been suggested that the number of systemic symptoms may correlate with prognosis.154
Peripheral T cell lymphoma - common type The common type of peripheral T cell lymphoma is a morphologically heterogeneous group of tumors that include a number of well-described histological subtypes. The rationale for regarding this as a single group is that many cases are difficult to classify by the Kiel criteria and distinct clinical entities have not been described. The value in continuing to describe these subtypes is that it illustrates the range of morphological features that may be encountered in diagnosing this group of tumors. The most morphologically distinct subtype in this group, and the one with the greatest claim to be a distinct entity, is Lennert's lymphoma (lymphoepithelioid lymphoma). The defining feature of this tumor is the presence of very large numbers of epithelioid macrophages (Plate 59). These are activated macrophages that have undergone further differentiation leading to increased secretory activity and mutual cohesion. The reactive macrophage population is often the major cellular component in the affected lymph node. The neoplastic T cells are found in small clusters lying between the macrophages. Usually these are mainly medium-sized cells with irregular nuclei together with variable numbers of blast cells. In most cases the tumor cells have a normal CD 4+, CD 45RO+, peripheral T cell phenotype. In most cases the rate of cell proliferation in the tumor cells is low. Lennert's lymphoma in most cases is a disseminated disease, although some patients have more
localized disease. Fever and other systemic symptoms are common. Like AIL this is a tumor that has a low proliferative potential and was regarded in the Kiel classification as low grade. However, the prognosis with any type of therapy appears to be poor with a median survival of around 16 months.155 The diagnosis of Lennert's lymphoma has in the past given rise to considerable problems. It is important to realize that other types of lymphoma may be rich in epithelioid macrophages. These include follicle center and diffuse large B cell lymphoma as well as both classical and lymphocyte-predominant nodular Hodgkin's disease. The use of appropriate marker studies will readily resolve most of these cases. Some cases of AIL may also be rich in epithelioid macrophages. In these cases the clinical features may be helpful in making the distinction. The majority of cases within this group consist of a variable mixture of medium- and large-sized neoplastic cells. The most typical morphological feature is the presence of highly irregular nuclei that may show a wide variety of bizarre shapes and contain multiple nucleoli. In some cases, T immunoblasts predominate. Like their B cell counterparts, these cells have a coarse chromatin pattern and one or more large nucleoli. A feature of some T immunoblasts is the presence of abundant clear cytoplasm. Eosinophils and reactive macrophages may be present in variable numbers. Although these features may be suggestive of T cell lineage, a reliable distinction from diffuse large B cell lymphoma is only possible with marker studies. In most cases, these tumors are CD 4+, CD 45RO+ but abnormal T cell immunophenotypes are common in this group. Peripheral T cell lymphomas of small cell type are rare. They consist of a relatively monomorphic population of small lymphocytes with irregularly shaped nuclei and sometimes a rim of clear cytoplasm. In these cases reactive non-neoplastic cells are generally sparse or absent. Like most other types of peripheral T cell lymphoma, the tumor cells are CD 4+, CD 45RO+. The main morphological differential diagnosis in these cases is mantle cell lymphoma but this distinction is not a problem when adequate marker studies are performed. Most peripheral T cell lymphomas present with disseminated nodal disease and extranodal deposits may also be present. A proportion of patients may have extensive bone marrow disease. Fever and other systemic symptoms are often present. It is suspected that the prognosis of peripheral T cell lymphomas is worse than that for diffuse large B cell lymphoma, although definitive data are difficult to obtain. The results obtained will depend on the exact population of cases included. Large numbers of anaplastic lymphomas will tend to improve the overall survival, while the opposite is the case with T lymphoblastic leukemia/lymphoma and possibly AIL.
64 T cell lymphoproliferative disorders
REFERENCES 1. Waldron JA, Leech JH, Glick AD, Flexner JM, Collins RD. Malignant lymphoma of peripheral T-lymphocyte origin: immunologic, pathologic, and clinical features in six patients. Cancer 1977; 40:1604-17. 2. Warnke RA, Link MP. Identification and significance of cell markers in leukemia and lymphoma. Annu Rev Med 1983; 34:117-31. 3. Jacobs W, Dhaene K, Van ME. Tyramine-amplified immunohistochemical testing using'homemade' biotinylated tyramine is highly sensitive and costeffective. Arch Pathol Lab Med 1998; 122: 642-3. 4. King G, Payne S, Walker F, Murray Gl. A highly sensitive detection method for immunohistochemistry using biotinylated tyramine.7 Pathol 1997; 183: 237-41. 5. Surowiak P, label M, Kasprzak A, Seidel J, SurdykZasada J, Lesisz I. Application of biotinylated tyramine to amplify reactions with horseradish peroxidase labelling. Folia Histochem Cytobiol 1999; 37: 57-8. 6. Langerak AW, Wolvers-Tettero IL, van DJ. Detection of T cell receptor beta (TCRB) gene rearrangement patterns in T cell malignancies by Southern blot analysis. Leukemia 1999; 13: 965-74. 7. Hayashi Y, Fukayama M, Funata N, Hishima T, Oba T, Koike M. Polymerase chain reaction screening of immunoglobulin heavy chain and T cell receptor gamma gene rearrangements: a practical approach to molecular DMA analysis of non-Hodgkin's lymphoma in a surgical pathology laboratory. Pathol Int 1999; 49: 1109-13. 8. Kaul K, Petrick M, Herz B, Cheng TC. Detection of clonal rearrangement of the T-cell receptor gamma gene by polymerase chain reaction and single-strand conformation polymorphism (PCR-SSCP). Mol Diagn 1996; 1:131-7. 9. Stansfeld AG, Diebold J, Noel H, etal. Updated Kiel classification for lymphomas [letter]. Lancet 1988; 1: 292-3. 10. Lennert K, Feller AC, Godde-Salz E. Morphology, immunohistochemistry and genetics of peripheral T cell lymphomas. Onkologie 1986; 9: 97-187. 11. Suchi T, Lennert K, Tu LY, el al. Histopathology and immunohistochemistry of peripheral T cell lymphomas: a proposal for their classification.) Clin Pathol 1987; 40:995-1015. 12. Hastrup N, Hamilton-Dutoit S, Ralfkiaer E, Pallesen G. Peripheral T-cell lymphomas: an evaluation of reproducibility of the updated Kiel classification. Histopathology 1991; 18: 99-105. 13. ArmitageJO, Greer JP, Levine AM, etal. Peripheral T-cell lymphoma. Cancer 1989; 63:158-63. 14. Anonymous. A clinical evaluation of the International Lymphoma Study Group classification of non-Hodgkin's lymphoma. The Non-Hodgkin's Lymphoma Classification Project. Blood 1997; 89: 3909-18. 15. Harris NL, Jaffe ES, Stein H, et al. A revised
16.
17.
18.
19.
20.
21.
22.
23.
24.
25.
26.
27. 28.
29.
European-American classification of lymphoid neoplasms: a proposal from the International Lymphoma Study Group. Blood 1994; 84:1361-92. Jaffe ES, Chan JK, Su IJ, et al. Report of the Workshop on Nasal and Related Extranodal AngiocentricT/Natural Killer Cell Lymphomas. Definitions, differential diagnosis, and epidemiology. Am JSurg Pathol 1996; 20:103-11. Cortes M, Wong E, Koipally J, Georgopoulos K. Control of lymphocyte development by the Ikaros gene family. CurrOpin Immunol 1999; 11:167-71. Georgopoulos K. Transcription factors required for lymphoid lineage commitment. CurrOpin Immunol 1997; 9: 222-7. Georgopoulos K, Bigby M, WangJH, etal. The Ikaros gene is required for the development of all lymphoid lineages. Cell 1994; 79:143-56. Georgopoulos K, Moore DD, Derfler B. Ikaros, an early lymphoid-specific transcription factor and a putative mediator for T cell commitment. Science 1992; 258: 808-9. Georgopoulos K, Winandy S, Avitahl N. The role of the Ikaros gene in lymphocyte development and homeostasis. Annu Rev Immunol 1997; 15:155-76. Sun L, Heerema N, Crotty L, et al. Expression of dominant-negative and mutant isoforms of the antileukemic transcription factor Ikaros in infant acute lymphoblastic leukemia. Proc Natl Acad Sci USA 1999; 96: 680-5. Kuo CT, Leiden JM. Transcriptional regulation of T lymphocyte development and function. Annu Rev Immunol 1999; 17:149-87. Klug CA, Morrison SJ, Masek M, Hahm K, Smale ST, Weissman IL. Hematopoietic stem cells and lymphoid progenitors express different Ikaros isoforms, and Ikaros is localized to heterochromatin in immature lymphocytes. Proc Natl Acad Sci USA 1998; 95: 657-62. Guy-Grand D, Rocha B, Mintz P, et al. Different use of T cell receptor transducing modules in two populations of gut intraepithelial lymphocytes are related to distinct pathways of T cell differentiation.) Exp Med 1994; 180: 673-9. Regnault A, Cumano A, Vassalli P, Guy-Grand D, Kourilsky P. Oligoclonal repertoire of the CD8 alpha alpha and the CD8 alpha beta TCR-alpha/beta murine intestinal intraepithelial T lymphocytes: evidence for the random emergence of T cells. 7 Exp Med 1994; 180: 1345-58. Rocha B, Guy-Grand D, Vassalli P. Extrathymic T cell differentiation. CurrOpin Immunol 1995; 7: 235-42. Rocha B, Vassalli P, Guy-Grand D. Thymic and extrathymic origins of gut intraepithelial lymphocyte populations in mice. 7 Exp Med 1994; 180: 681-6. Spencer J, Cerf-Bensussan N, Jarry A, et al. Enteropathyassociated T cell lymphoma (malignant histiocytosis of the intestine) is recognized by a monoclonal antibody (HML-1) that defines a membrane molecule on human mucosal lymphocytes. Am7 Pathol 1988; 132:1-5.
References 65
30. Fung-Leung WP, Kishihara K, Gray D, Teh HS, Lau CY, Mak TW. Intestinal T cells in CDS alpha knockout mice and T cell receptor transgenic mice. Adv Exp Med Biol 1995; 371A:
121^.
31. Jeggo PA, Carr AM, Lehmann AR. Splitting the ATM: distinct repair and checkpoint defects in ataxiatelangiectasia. Trends Genet 1998; 14: 312-16. 32. Stankovic T, Kidd AM, Sutcliffe A, et al. ATM mutations and phenotypes in ataxia-telangiectasia families in the British Isles: expression of mutant ATM and the risk of leukemia, lymphoma, and breast cancer. AmJ Human Genet *\ 998; 62: 334-45. 33. Brown KD, Tagle DA. Molecular perspectives on cancer, the cell cycle and the inherited disorder ataxiatelangiectasia. Prog Clin Biol Res 1997; 396:101-13. 34. Digel W, SchultzeJ, Kunzmann R, Mertelsmann R, Lindemann A. Poor prognosis of prethymic phenotype acute lymphoblastic leukemia (pre-T-ALL). Leukemia 1994; 8: 1406-8. 35. Garand R, Voisin S, Papin S, et al. Characteristics of pro-T ALL subgroups: comparison with late T-ALL. The
T-cell acute lymphoblastic leukemia. Ann Hematol 1992; 64: 305-8. 44. Brown L, Cheng JT, Chen Q, et al. Site-specific recombination of the tal-1 gene is a common occurrence in human T cell leukemia. EMBOJ1990; 9: 3343-51. 45. Breit TM, Beishuizen A, Ludwig WD, et al. tal-1 deletions in T-cell acute lymphoblastic leukemia as PCR target for detection of minimal residual disease. Leukemia 1993; 7:2004-11. 46. Janssen JW, Ludwig WD, Sterry W, Bartram CR. SIL-TAL1 deletion in T-cell acute lymphoblastic leukemia. Leukemia 1993; 7:1204-10. 47. Bernard 0, Lecointe N, Jonveaux P, et al. Two sitespecific deletions and t(1;14) translocation restricted to human T-cell acute leukemias disrupt the 5' part of the tal-1 gene. Oncogene 1991; 6:1477-88. 48. Jonsson OG, Kitchens RL, Baer RJ, Buchanan GR, Smith RG. Rearrangements of the tal-1 locus as clonal markers for T cell acute lymphoblastic leukemia. J Clin Invest 1991; 87: 2029-35. 49. Bernard 0, Guglielmi P, Jonveaux P, et al. Two distinct
Groupe d'Etude Immunologique des Leucemies.
mechanisms for the SCL gene activation in the t(1 ;14)
Leukemia 1993; 7:161-7.
translocation of T-cell leukemias. Genes Chromosomes
36. Cascavilla N, Musto P, D'Arena G, et al. Are 'early' and 'late' T-acute lymphoblastic leukemias different diseases? A single center study of 34 patients. Leuk Lymphoma 1996; 21: 437-42. 37. Uckun FM, Steinherz PG, Sather H, et al. CD2 antigen expression on leukemic cells as a predictor of event-free survival after chemotherapy for T-lineage acute
Cancer 1990; 1:194-208. 50. Delabesse E, Bernard M, Meyer V, et al. TAL1 expression does not occur in the majority of T-ALL blasts. BrJ Haematol 1998; 102: 449-57. 51. Bash RO, Crist WM, Shuster JJ, et al. Clinical features and outcome of T-cell acute lymphoblastic leukemia in childhood with respect to alterations at the TAL1 locus:
lymphoblastic leukemia: a Children's Cancer Group
a Pediatric Oncology Group study. Blood 1993; 81:
study. Blood 1996; 88: 4288-95.
2110-17.
38. Uckun FM, Gaynon P, Sather H, et al. Clinical features and treatment outcome of children with biphenotypic CD2+ CD19+ acute lymphoblastic leukemia: a Children's Cancer Group study. Blood 1997; 89: 2488-93. 39. Uckun FM, Sather HN, Gaynon PS, et al. Clinical features and treatment outcome of children with myeloid antigen positive acute lymphoblastic leukemia: a report from the Children's Cancer Group. Blood 1997; 90: 28-35.
52. Cheng JT, Yang CY, Hernandez J, Embrey J, Baer R. The chromosome translocation (11;14)(p13;q11) associated with T cell acute leukemia. Asymmetric diversification of the translocational junctions. J Exp Med 1990; 171: 489-501. 53. Finver SN, Martiniere C, Kagan J, Cavenee W, Croce CM. The chromosome 11 region flanking the t(11;14) breakpoint in human T-ALL is deleted in Wilms' tumor hybrids. Oncogene Res 1989; 5:143-8. 54. Royer-Pokora B, Fleischer B, Ragg S, Loos U, Williams D.
40. Warzynski MJ, Graham DM, Axtell RA, Zakem MH, Rotman RK. Low level CD20 expression on T cell
Molecular cloning of the translocation breakpoint in TALL 11;14 (p13;q11): genomic map of TCR alpha and
malignancies. Cytometry 1994; 18: 88-92. 41. Boucheix C, David B, Sebban C, et al.
delta region on chromosome 14q11 and long-range map of region 11p13. Human Genet 1989; 82: 264-70.
Immunophenotype of adult acute lymphoblastic leukemia, clinical parameters, and outcome: an analysis of a prospective trial including 562 tested patients (LALA87). French Group on Therapy for Adult
55. Ribeiro RC, Raimondi SC, Behm FG, Pui CH. T-cell acute lymphoblastic leukemia with t(11;14) in children. Leuk Lymphoma 1992; 7: 351-8. 56. McGuire EA, Hockett RD, Pollock KM, Bartholdi MF,
Acute Lymphoblastic Leukemia. Blood 1994; 84:
O'Brien SJ, KorsmeyerSJ. Thet(11;14)(p15;q11) in a T-cell
1603-12.
acute lymphoblastic leukemia cell line activates multiple transcripts, including Ttg-1, a gene encoding a potential
42. Goldfarb AN, Greenberg JM. T-cell acute lymphoblastic leukemia and the associated basic helix-loop-helix gene SCL/tal. Leuk Lymphoma 1994; 12:157-66. 43. Borkhardt A, Repp R, Harbott J, et al. Frequency and DNA sequence of tal-1 rearrangement in children with
zinc finger protein. Molec Cell Biol 1989; 9: 2124-32. 57. Xia Y, Hwang LY, Cobb MH, Baer R. Products of the TAL2 oncogene in leukemic T cells: bHLH phosphoproteins with DMA-binding activity. Oncogene 1994; 9:1437-46.
66 T cell lymphoproliferative disorders
58. Baer R. TAL1, TAL2 and LYL1: a family of basic helixloop-helix proteins implicated in T cell acute leukaemia. Semin Cancer Biol 1993; 4: 341-7. 59. Goldman J, McGuire WA. H-ras and c-myc RNA expression in human T-cell ALL and in normal human lymphocytes. Pediat Hematol Oncol 1992; 9: 309-16. 60. Xia Y, Brown L, Yang CY, et al. TAL2, a helix-loop-helix gene activated by the (7;9)(q34;q32) translocation in human T-cell leukemia. Proc Natl Acad Sci USA 1991; 88:11416-20. 61. Wad man I, Li J, Bash RO, et al. Specific in vivo association between the bHLH and LIM proteins implicated in human T cell leukemia. EMBOJ 1994; 13: 4831-9. 62. Ono Y, Fukuhara N, Yoshie 0. Transcriptional activity of TAL1 in T cell acute lymphoblastic leukemia (T-ALL) requires RBTN1 or-2 and induces TALLA1, a highly specific tumor marker of T-ALL. J Biol Chem 1997; 272: 4576-81. 63. Sanchez-Garcia I, RabbittsTH. LIM domain proteins in leukaemia and development. Semin Cancer Biol 1993; 4: 349-58. 64. Hatano M, Roberts CW, Minden M, Crist WM, Korsmeyer SJ. Deregulation of a homeobox gene, HOX11, by the t(10;14) in T cell leukemia. Science 1991; 253: 79-82. 65. Lichty BD, Ackland-Snow J, Noble L, Kamel-Reid S, Dube
73. Sivakumaran M, Richards S. Immunological abnormalities of chronic large granular lymphocytosis. Clin Lab Haematol 1997; 19: 57-60. 74. Bowman SJ, Sivakumaran M, Snowden N, et al. The large granular lymphocyte syndrome with rheumatoid arthritis. Immunogenetic evidence for a broader definition of Felty's syndrome. Arthritis Rheum 1994; 37: 1326-30. 75. Yoe J, Gause BL, Curti BD, Longo DL, Bagg A, Kopp WC, Janik JE. Development of rheumatoid arthritis after treatment of large granular lymphocyte leukemia with deoxycoformycin. Am J Hematol 1998; 57: 253-7. 76. Scott CS, Richards SJ, Sivakumaran M, etal. Persistent clonal expansions of CD3+TCR gamma delta+ and CD3+TCR alpha beta+CD4-CD8- lymphocytes associated with neutropenia. LeukLymphoma 1994; 14: 429-40. 77. Frolova EA, Richards SJ, Jones RA, et al. Immunophenotypic and DMA genotypic analysis of T-cell and NK-cell subpopulations in patients with B-cell chronic lymphocytic leukaemia (B-CLL). Leuk Lymphoma 1995; 16: 307-18. 78. Claudepierre P, Bergamasco P, Delfau MH, et al. Unusual CD3+, CD4+ large granular lymphocyte expansion associated with a solid tumor.7 Rheumatol 1998; 25:1434-6. 79. Richards SJ, Sivakumaran M, Parapia LA, et al. A distinct
ID. Dysregulation of HOX11 by chromosome
large granular lymphocyte (LGL)/NK-associated (NKa)
translocations in T-cell acute lymphoblastic leukemia: a
abnormality characterized by membrane CD4 and CD8
paradigm for homeobox gene involvement in human
coexpression. The Yorkshire Leukaemia Group. BrJ
cancer. LeukLymphoma 1995; 16: 209-15.
Haematol 1992; 82: 494-501.
66. Cayuela JM, Gardie B, Sigaux F. Disruption of the multiple tumor suppressor gene MTS1/p16(INK4a)/CDKN2 by illegitimate V(D)J recombinase activity in T-cell acute lymphoblastic leukemias. Blood 1997; 90: 3720-6. 67. Diccianni MB, Batova A, Yu J, et al. Shortened survival after relapse in T-cell acute lymphoblastic leukemia patients with p16/p15 deletions. Leukemia Research 1997; 21: 549-58. 68. Yamada Y, Hatta Y, Murata K, et al. Deletions of p15 and/or p16 genes as a poor-prognosis factor in adult T-cell leukemia. J Clin Oncol 1997; 15:1778-85. 69. Hsiao MH, Yu AL, Yeargin J, Ku D, Haas M. Nonhereditary p53 mutations in T-cell acute lymphoblastic leukemia are associated with the relapse phase. Blood 1994; 83: 2922-30. 70. Jonveaux P, Berger R. Infrequent mutations in the P53 gene in primary human T-cell acute lymphoblastic leukemia. Leukemia 1991; 5: 839-40. 71. Scott CS, Richards SJ, Sivakumaran M, et al. Transient and persistent expansions of large granular lymphocytes (LGL) and NK-associated (NKa) cells: the Yorkshire Leukaemia Group Study. BrJ Haematol 1993; 83: 505-15. 72. Chen CL, Sadler RH, Walling DM, Su IJ, Hsieh HC, Raab-
80. Sala P, Tonutti E, Feruglio C, Florian F, Colombatti A. Persistent expansions of CD4+ CD8+ peripheral blood T cells. Blood 1993; 82:1546-52. 81. Scott CS, Richards SJ. Classification of large granular lymphocyte (LGL) and NK-associated (NKa) disorders. Blood Rev 1992; 6: 220-33. 82. Sivakumaran M, Richards SJ, Hunt KM, et al. Patterns of CD16 and CD56 expression in persistent expansions of CD3+NKa+ lymphocytes are predictive for clonal T-cell receptor gene rearrangements. The Yorkshire Leukaemia Group. BrJ Haematol 1991; 78: 368-77. 83. Schmidt HH, Pirc-Danoewinata H, Panzer-Grumayer ER, et al. Translocation (3;5)(p26;q13) in a patient with chronic T cell lymphoproliferative disorder. Cancer Genet Cytogenet 1998; 104: 82-5. 84. Prieto J, Rios E, Parrado A, Martin A, de Bias JM, Rodriguez JM. Leukaemia of natural killer cell large granular lymphocyte type with HLA-DR-CD16-CD56bright+ phenotypej Clin Pathol 1996; 49:1011-13. 85. Matutes E, Brito-Babapull V, Swansbury J, et al. Clinical and laboratory features of 78 cases of T-prolymphocytic leukemia. Blood 1991; 78: 3269-74. 86. Maljaei SH, Brito-Babapulle V, Hiorns LR, Catovsky D.
Traub N. Epstein-Barr virus (EBV) gene expression in
Abnormalities of chromosomes 8,11,14, and X in
EBV-positive peripheral T-cell lymphomasj Virol 1993; 67: 6303-8.
T-prolymphocytic leukemia studied by fluorescence in situ hybridization. Cancer Genet Cytogenet 1998; 103:110-16.
References 67 87. Brito-Babapulle V, Catovsky D. Inversions and tandem translocations involving chromosome 14q11 and 14q32 in T-prolymphocytic leukemia and T-cell leukemias in patients with ataxia telangiectasia. Cancer Genet Cytogenet 1991; 55:1-9. 88. Yuille MA, Coignet LJ, Abraham SM, et al. ATM is usually rearranged in T-cell prolymphocytic leukaemia. Oncogene 1998; 16: 789-96. 89. Uhrhammer N, Bay JO, Bignon YJ. Seventh International Workshop on Ataxia-Telangiectasia. Cancer Res 1998; 58: 3480-5. 90. Stilgenbauer S, Schaffner C, Litterst A, et al. Biallelic mutations in the ATM gene in T-prolymphocytic leukemia. Nature Medicine 1997; 3:1155-9. 91. Pawson R, Dyer MJ, Barge R, et al. Treatment of T-cell prolymphocytic leukemia with human CD52 antibody. J Clin Oncol 1997; 15: 2667-72. 92. Pombo de Oliveira MS, Matutes E, Schulz T, et al. T-cell malignancies in Brazil. Clinico-pathological and molecular studies of HTLV-l-positive and -negative cases. Int J Cancer 1995; 60: 823-7. 93. Masuda M, Ishida C, Arai Y, et al. Dual action of CD30 antigen: anti-CD30 antibody induced apoptosis and interleukin-8 secretion in Ki-1 lymphoma cells. Int J Hematol 1998; 67: 257-65. 94. Harrington WJ Jr, Ucar A, Gill P, et al. Clinical spectrum of HTLV-I in south Florida. J Acquired Immune Defic Syndr Human Retrovirol 1995; 8: 466-73. 95. de Oliveira M, do S, Hamerschlak N, Chiattone C, Loureiro P. HTLV-I infection and adult T-cell leukemia in Brazil: an overview. Rev Paulista de Medicina 1996; 114: 1177-85. 96. Sharata HH, Colvin JH, Fujiwara K, Goldman B, Hashimoto K. Cutaneous and neurologic disease associated with HTLV-I infection. J Am Acad Dermatol 1997; 36 (Pt 2): 869-71. 97. Hu CY, Lin MT, Yang YC, et al. Familial transmission of human T-lymphotropk virus type 1 (HTLV-1) in patients with adult T-cell leukemia/lymphoma or HTLV-1associated myelopathy.y Formosan MedAssoc 1998; 97: 101-5. 98. Kanzaki T, Setoyama M, Katahira Y. Human T lymphotropic virus-1 infection. Austral] Dermatol 1996; 37(suppl1):S20-2. 99. Cavrois M, Wain-Hobson S, Gessain A, Plumelle Y, Wattel E. Adult T-cell leukemia/lymphoma on a background of clonally expanding human T-cell leukemia virus type-1positive cells. Blood 1996; 88:, 4646-50. 100. Takemoto S, Mulloy JC, Cereseto A, etal. Proliferation of adult T cell leukemia/lymphoma cells is associated with the constitutive activation of JAK/STAT proteins. Proc Natl Acad Sci USA 1997; 94:13897-902. 101. Uchid T, Kinoshita T, Murate T, Saito H, Hotta T. CDKN2 (MTS1/p16INK4A) gene alterations in adult T-cell leukemia/lymphoma. Leuk Lymphoma 1998; 29: 27-35. 102. Mori N, Shirakawa F, Murakami S, Oda S, Eto S. Characterization and regulation of interleukin-4
103.
104.
105.
106.
107.
108.
109.
110.
111.
112.
113.
114.
115.
116.
receptor in adult T-cell leukemia cells. EurJ Haematol 1996; 56: 241-7. Mori N, Gill PS, Mougdil T, Murakami S, Eto S, Prager D. lnterleukin-10 gene expression in adult T-cell leukemia. Blood 1996; 88:1035^5. Yamada Y, Ohmoto Y, Hata T, et al. Features of the cytokines secreted by adult T cell leukemia (ATL) cells. Leuk Lymphoma 1996; 21: 443-7. Tatewaki M, Yamaguchi K, Matsuoka M, et al. Constitutive overexpression of the L-selectin gene in fresh leukemic cells of adult T-cell leukemia that can be transactivated by human T-cell lymphotropic virus type 1 Tax. Blood 1995; 86: 3109-17. Uchiyama T, Ishikawa T, Imura A. Adhesion properties of adult T cell leukemia cells. Leuk Lymphoma 1995; 16: 407-12. Otsuka M, Hanada S, Utsunomiya A, Ishitsuka K, Uozumi K, Arima T. Expression of the intestinal T-lymphocyte-associated-molecule recognized by the HML-1 antibody on mononuclear cells from HTLV-Iinfected subjects. Am J Hematol 1995; 50:1-8. Cohen PL, ButmarcJ, Kadin ME. Expression of Hodgkin's disease associated antigen BLA.36 in anaplastic large cell lymphomas and lymphomatoid papulosis primarily of T-cell origin. Am J Clin Pathol 1995; 104: 50-3. Ohtake N, Setoyama M, Fukumaru S, Kanzaki T. A case of adult T-cell leukemia/lymphoma (ATLL) with angiocentric and angiodestructive features.) Dermatol 1997; 24:165-9. Macera MJ, Hyde P, Peddanna N, Szabo P, Gogineni SK, Verma RS. T-cell receptor J beta 1/J beta 2 locus rearrangements in an HTLV-1-positive T-cell lymphoma with complex chromosomal aberrations. Am] Hematol 1996; 52: 53-7. Roodman GD. Mechanisms of bone lesions in multiple myeloma and lymphoma. Cancer1997; 80 (suppl): 1557-63. D'incan M, Antoniotti 0, Gasmi M, et al. HTLV-Iassociated lymphoma presenting as mycosis fungoides in an HTLV-I non-endemic area: a viro-molecular study. BrJ Dermatol 1995; 132: 983-8. Pawson R, Matutes E, Brito-Babapulle V, et al. Sezary cell leukaemia: a distinct T cell disorder or a variant form of T prolymphocytic leukaemia? Leukemia 1997; 11:1009-13. Mallett RB, Matutes E, Catovsky D, MacLennan K, Mortimer PS, Holden CA. Cutaneous infiltration in T-cell prolymphocytic leukaemia. BrJ Dermatol 1995; 132: 263-6. Yagi H, Tokura Y, Furukawa F, Takigawa M. CD7-positive Sezary syndrome with a Th1 cytokine profile. J Am Acad Dermatol 1996; 34: 368-74. TendlerCL, Burton JD, JaffeJ, etal. Abnormal cytokine expression in Sezary and adult T-cell leukemia cells correlates with the functional diversity between these T-cell malignancies. Cancer Res 1994; 54: 4430-5.
68 T cell lymphoproliferative disorders 117. Saed G, Fivenson DP, Naidu Y, Nickoloff BJ. Mycosis fungoides exhibits a Thl-type cell-mediated cytokine profile whereas Sezary syndrome expresses a Th2-type profile.7 Invest Dermatol 1994; 103: 29-33. 118. Hoppe RT, Wood GS, Abel EA. Mycosis fungoides and the Sezary syndrome: pathology, staging, and treatment. Curr Prob Cancer 1990; 14: 293-371. 119. Younes A, Consoli U, Snell V, etal. CD30 ligand in lymphoma patients with CD30+ tumors. J Clin Oncol 1997; 15: 3355-62. 120. Benharroch D, Meguerian-Bedoyan Z, Lamant L, et al. ALK-positive lymphoma: a single disease with a broad spectrum of morphology. Blood 1998; 91: 2076-84. 121. Kadin ME. Anaplastic large cell lymphoma and its morphological variants. Cancer Surveys 1997; 30: 77-86. 122. McCluggage WG, Walsh MY, Bharucha H. Anaplastic large cell malignant lymphoma with extensive eosinophilic or neutrophilic infiltration. Histopathology 1998;32:110-15. 123. Ott G, Bastian BC, Katzenberger T, et al. A lymphohistiocytic variant of anaplastic large cell lymphoma with demonstration of the t(2;5)(p23;q35) chromosome translocation. BrJ Haematol 1998; 100: 187-90. 124. Pileri SA, Pulford K, Mori S, et al. Frequent expression of the NPM-ALKchimeric fusion protein in anaplastic large-cell lymphoma, lymphohistiocytic type. AmJ Pathol 1997; 150:1207-11. 125. Foss HD, Demel G, Anagnostopoulos I, Araujo I, Hummel M, Stein H. Uniform expression of cytotoxic molecules in anaplastic large cell lymphoma of null/T cell phenotype and in cell lines derived from anaplastic large cell lymphoma. Pathobiology 1997; 65: 83-90. 126. Pulford K, Lamant L, Morris SW, et al. Detection of anaplastic lymphoma kinase (ALK) and nucleolar protein nudeophosmin (NPM)-ALK proteins in normal and neoplastic cells with the monoclonal antibody ALK1. Blood 1997; 89:1394-1404. 127. Delsol G, Lamant L, Mariame B, et al. A new subtype of large B-cell lymphoma expressing the ALK kinase and lacking the 2;5 translocation. Blood 1997; 89:1483-90. 128. Sarris AH, Luthra R, Papadimitracopoulou V, et al. Longrange amplification of genomic DNA detects the t(2;5)(p23;q35) in anaplastic large-cell lymphoma, but not in other non-Hodgkin's lymphomas, Hodgkin's disease, or lymphomatoid papulosis. Ann Oncol 1997; 8 (suppl 2): 59-63. 129. Nakamura S, Shiota M, Nakagawa A, et al. Anaplastic large cell lymphoma: a distinct molecular pathologic entity: a reappraisal with special reference to p80(NPM/ALK) expression. Am J Surg PatholW97; 21: 1420-32. 130. Sarris AH, Luthra R, Cabanillas F, Morris SW, Pugh WC. Genomic DNA amplification and the detection of t(2;5)(p23;q35) in lymphoid neoplasms. Leuk Lymphoma 1998; 29: 507-14.
131. Pittaluga S, Wiodarska I, Pulford K, etal. The monoclonal antibody ALK1 identifies a distinct morphological subtype of anaplastic large cell lymphoma associated with 2p23/ALK rearrangements. Am J Pathol 1997; 151: 343-51. 132. Shiota M, Mori S. The dinicopathological features of anaplastic large cell lymphomas expressing pSONPM/ALK. Leuk Lymphoma 1996; 23: 25-32. 133. Zinzani PL, Bendandi M, Martelli M, etal. Anaplastic large-cell lymphoma: clinical and prognostic evaluation of 90 adult patients. J Clin Oncol 1996; 14: 955-62. 134. Longo G, Federico M, Pieresca C, et al. Anaplastic large cell lymphoma (CD30+/Ki-1+). Analysis of 35 cases followed at GISL centres. EurJ Cancer 1995; 31 A: 1763-7. 135. Weisenburger D., Gordon BG, VoseJM, etal. Occurrence of the t(2;5)(p23;q35) in non-Hodgkin's lymphoma. Blood 1996; 87: 3860-8. 136. Kadin ME, Morris SW. The t(2;5) in human lymphomas. Leuk Lymphoma 1998; 29: 249-56. 137. LeBoit PE. Lymphomatoid papulosis and cutaneous CD30+ lymphoma. AmJ Dermatopathol 1996; 18: 221-35. 138. Herbst H, Sander C, Tronnier M, Kutzner H, Hugel H, Kaudewitz P. Absence of anaplastic lymphoma kinase (ALK) and Epstein-Barr virus gene products in primary cutaneous anaplastic large cell lymphoma and lymphomatoid papulosis. BrJ Dermatol 1997; 137: 680-6. 139. Chott A, Vonderheid EC, Olbricht S, Miao NN, Balk SP, Kadin ME. The dominant T cell clone is present in multiple regressing skin lesions and associated T cell lymphomas of patients with lymphomatoid papulosis. J Invest Dermatol 1996; 106: 696-700. 140. Amagai M, Kawakubo Y, Tsuyuki A, Harada R. Lymphomatoid papulosis followed by Ki-1 positive anaplastic large cell lymphoma: proliferation of a common T-cell clone. J Dermatol 1995; 22: 743-6. 141. Paulli M, Berti E, Rosso R, etal. CD30/KM-positive lymphoproliferative disorders of the skin clinicopathologic correlation and statistical analysis of 86 cases: a multicentric study from the European Organization for Research and Treatment of Cancer Cutaneous Lymphoma Project Group.7 Clin Oncol 1995; 13:1343-54. 142. Ree HJ, Kadin ME, Kikuchi M, Ko YH, Go JH, Suzumiya J, Kim DS. Angioimmunoblastic lymphoma (AILD-type T-cell lymphoma) with hyperplastic germinal centers. AmJSurg Pathol 1998; 22: 643-55. 143. Anagnostopoulos I, Hummel M, Stein H. Frequent presence of latent Epstein-Barr virus infection in peripheral T cell lymphomas. Leuk Lymphoma 1995; 19: 1-12. 144. Khan G, Norton AJ, Slavin G. Epstein-Barr virus in angioimmunoblastic T-cell lymphomas. Histopathology 1993;22:145-9.
References 69 145. Ohshima K, Takeo H, Kikuchi M, et al. Heterogeneity of Epstein-Barr virus infection in angioimmunoblastic lymphadenopathy type T-cell lymphoma. Histopathology 1994; 25: 569-79. 146. Kumaravel TS, Tanaka K, Arif M, et al. Clonal identification of trisomies 3, 5 and X in angioimmunoblastic lymphadenopathy with dysproteinemia by fluorescence in situ hybridization. Leuk Lymphoma 1997; 24: 523-32. 147. Schlegelberger B, Zwingers T, Hohenadel K, et al. Significance of cytogenetic findings for the clinical outcome in patients with T-cell lymphoma of angioimmunoblastic lymphadenopathy type. J Clin Onaj/1996;14: 593-9. 148. Schlegelberger B, Zhang Y, Weber-Matthiesen K, Grote W. Detection of aberrant clones in nearly all cases of angioimmunoblastic lymphadenopathy with dysproteinemia-type T-cell lymphoma by combined interphase and metaphase cytogenetics. Blood 1994; 84: 2640-8. 149. Sonobe M, Yasuda H, Okabe H, et al. Neuropathy associated with angioimmunoblastic lymphadenopathylike T-cell lymphoma. Int Med 1998; 37: 631-4. 150. Layton MA, Musgrove C, Dawes PT. Polyarthritis, rash
151.
152.
153.
154.
155.
and lymphadenopathy: case reports of two patients with angioimmunoblastic lymphadenopathy presenting to a rheumatology clinic. Clin Rheumatol'\998; 17: 148-51. Schmuth M, Ramaker J, Trautmann C, et al. Cutaneous involvement in prelymphomatous angioimmunoblastic lymphadenopathy. J Am Acad Dermatol 1997; 36: 290-5. Hsu SM, Waldron JA Jr, Fink L, et al. Pathogenic significance of interleukin-6 in angioimmunoblastic lymphadenopathy-type T-cell lymphoma. Human Pathol 1993; 24:126-31. Abruzzo LV, Schmidt K, Weiss LM, et al. B-cell lymphoma after angioimmunoblastic lymphadenopathy: a case with oligoclonal gene rearrangements associated with Epstein-Barr virus. Blood 1993; 82: 241-6. Simonitsch I, Panzer-Gruemayer ER, Ghali DW, etal. NPM/ALK gene fusion transcripts identify a distinct subgroup of null type Ki-1 positive anaplastic large cell lymphomas. BrJ Haematol 1996; 92: 866-71. Patsouris E, Engelhard M, Zwingers T, Lennert K. Lymphoepithelioid cell lymphoma (Lennert's lymphoma): clinical features derived from analysis of 108 cases. BrJ Haematol 1993; 84: 346-8.
This page intentionally left blank
8 Extranodal lymphomas PG ISAACSON
Introduction Mucosa-associated lymphoid tissue: the MALT lymphoma concept Malignant lymphoma of the gastrointestinal tract Malignant lymphoma of the salivary glands Malignant lymphoma of the lung Malignant lymphoma of the thyroid Lymphomas of the ocular adnexa and eye Malignant lymphoma of the skin Malignant lymphoma of the upper aerodigestive tract
71 72 74 77 77 79 79 80 81
INTRODUCTION Reliable data on the proportion of lymphomas that arise in tissues or organs other than lymph nodes are not readily available, but the figure probably lies somewhere between 20 and 40 per cent.1'2 One reason for this imprecision derives from the problem of definition. Nodal lymphomas frequently spread to extranodal sites, which can lead to a mistaken diagnosis of primary extranodal disease, and, equally, extranodal tumors often disseminate to lymph nodes and may erroneously be thought to arise there. No watertight definition of extranodal lymphomas exists but a reasonable operational definition is that a lymphoma may be considered to be extranodal if it presents with the main bulk of disease in an extranodal site, provided that primary lymphomas of that site, without nodal involvement, have previously been described.3 Almost all extranodal lymphomas are of non-Hodgkin's type, with the single exception of the thymus; extranodal Hodgkin's disease is vanishingly rare. Sites of origin Extranodal lymphomas may arise from primary lymphoid organs such as the spleen, thymus or Waldeyer's ring, from organs with a significant lymphoid tissue
Malignant lymphoma of the central nervous system Malignant lymphoma of the thymus Malignant lymphomas of the spleen Malignant lymphoma of the urogenital tract Extranodal lymphomas of miscellaneous sites Multifocal extranodal lymphoma Extranodal lymphoma associated with immunodeficiency References
83 83 84 85 85 86 86 87
component, such as the gastrointestinal tract, or from organs or tissues normally devoid of lymphoid tissue, such as the brain or skeletal muscle. In effect, lymphomas may arise anywhere in the body but certain sites are more prone than others. The gastrointestinal tract is by far the commonest site of extranodal lymphoma followed by the skin, Waldeyer's ring and the salivary glands.
Classification The first rational classifications of non-Hodgkin's lymphomas, which emerged in the late 1970s and early 1980s, emphasized the architectural and cytological resemblances between lymphomas and normal lymph nodes.4,5 With the single exception of cutaneous T cell lymphomas, which were categorized separately in the Kiel classification, site of origin was not considered a significant factor. Its importance, however, became apparent following the description of lymphomas of mucosa-associated lymphoid tissue (MALT) in 1983,6 which was soon followed by a description of other clinicopathologically distinct extranodal lymphomas. It was partly this omission of extranodal entities from the currently used lymphoma classifications that provoked the proposal of a new lymphoma classification in 1994. The Revised European-American Lymphoma (REAL) classification7 (see Table 1.6, page 6) essentially lists those
72 Extranodal lymphomas
lymphomas that are recognized as disease entities and is the first classification specifically to include extranodal lymphomas. In this chapter extranodal lymphomas will be grouped together according to their site of origin and described following the principles of the REAL classification.
MUCOSA-ASSOCIATED LYMPHOID TISSUE: THE MALT LYMPHOMA CONCEPT In the 1970s, Lukes and Collins4 and Lennert5 observed that lymphomas tended to recapitulate the histological features of normal lymph nodes and emphasized the importance of this in lymphoma classification. In attempting to apply these principles to the classification of certain extranodal B cell lymphomas, Isaacson and Wright6'8 noted that their histology was more closely related to that of mucosa-associated lymphoid tissue than of lymph nodes and coined the term 'MALT lymphoma'. Histology of MALT The features of MALT can best be observed in the gastrointestinal tract where it comprises the Peyer's patches, lymphocytes in the lamina propria, intraepithelial lymphocytes and mesenteric lymph nodes. Peyer's patches are distinguished by a central B cell follicle, which is surrounded by a prominent marginal zone; small T cell areas are present between these prominent B cell foci (Fig. 8.1). These lymphoid aggregates form dome-like protuberances covered by intestinal epithelium, which characteristically contains clusters of intraepithelial marginal zone B cells; this so-called lymphoepithelium is a defining structure of MALT. The intraepithelial B cells must be distinguished from the more prominent intraepithelial T cells, which are present throughout the intestinal epithelium with the exception of the dome areas above Peyer's patches. Functional properties There are no afferent lymphatics in MALT and antigen passes directly across the covering mucosa into the lymphoid tissue. The antigen-stimulated B cells leave MALT via efferent lymphatics and the B cells then 'home' back to the mucosal lamina propria where they comprise the antibody forming plasma cells.9 The behavior of Peyer's patch T cells is less well understood but is probably similar. Most lamina propria plasma cells synthesize immunoglobulin A (IgA) antibodies, which, together with secretory component, are transported to the surface of the intestinal epithelium where they provide local immunity.
Figure 8.1 Peyer's patch showing a central B cell follicle (F) with surrounding marginal zone (M) and intraepithelial B cells (arrows). Tcell areas are also present (T).
Acquired MALT Paradoxically, MALT lymphomas arise most commonly in organs such as the stomach, salivary gland and thyroid, which normally contain no lymphoid tissue. A common feature, however, is that the lymphomas arise in a setting of a variety of chronic inflammatory, often autoimmune disorders (chronic gastritis, Sjogren's syndrome, Hashimoto's thyroiditis, etc.), which result in the accumulation of lymphoid tissue characterized by the presence of abundant B cell follicles with an adjacent lymphoepithelium (i.e. MALT).10-12 More details of acquired MALT as it relates to the different sites where MALT lymphomas occur are given later in the following sections. MALT lymphoma concept The term MALT lymphoma refers to a group of B cell extranodal lymphomas that recapitulate the features of the B cell component of MALT. They are thought to be derived from the marginal zone B cells that are found
Mucosa-associated lymphoid tissue 73
surrounding B cell follicles and within the adjacent lymphoepithelium. First characterized in the gastrointestinal tract,6 it is now evident that a significant proportion of extranodal B cell lymphomas arising in a wide variety of organs are of MALT type and share common clinical, histological and molecular genetic properties.3 The features described below are those common to all MALT lymphomas and will be repeatedly referred to in the discussions of lymphomas arising in various extranodal sites that are described later in this chapter. Clinical presentation MALT lymphomas occur predominantly, but not exclusively, in individuals over 50. The presentation is subtle since the symptoms and signs tend to merge with those of the chronic inflammatory disorder that almost invariably precedes the emergence of lymphoma. In keeping with their indolent course (see later) MALT lymphomas are usually at stage IE or IIE (Musshoff)13 when diagnosed.
Low-grade MALT lymphoma HISTOPATHOLOGY
The cells of low-grade B cell MALT lymphomas (Fig. 8.2) characteristically resemble centrocytes and are often referred to as centrocyte-like (CCL) cells even when, as is often the case, their appearance is not entirely characteristic and they more closely resemble small round lymphocytes or monocytoid B cells. The CCL cells invade adjacent epithelial structures to form the characteristic lymphoepithelial lesions. A variable number of nucleolated transformed blasts is always present. Plasma cell differentiation is common and is usually maximal adjacent to epithelium. The presence of benign reactive B cell follicles, comprising a follicle center and mantle zone, within the neoplastic infiltrate is an invariable feature and, especially in small foci of disease, it is clear that the neoplastic CCL cells are distributed around the follicles in the marginal zone. The CCL cells may specifically colonize reactive follicle centers where they may undergo blast transformation or differentiate into plasma cells. This 'follicular colonization' can lead to a remarkable resemblance to follicular lymphoma (Fig. 8.3). MALT lymphomas are frequently multifocal and in paired organs, such as the salivary glands and conjunctivae, bilateral presentation often occurs, which poses problems in staging the disease. Both clinically and histologically, the emergence of low-grade MALT lymphoma in conditions such as chronic gastritis and Sjogren's syndrome in which MALT is acquired can be extremely difficult to diagnose. Under these circumstances demonstration of B cell monoclonality, which is indicative of neoplastic rather than
Figure 8.2
Low-grade B cell gastric MALT lymphoma showing
reactive follicle (F) surrounded by an infiltrate of centrocyte-like eel Is forming lymphoepithelial lesions (arrows). Note the similarity of the structure to a Peyer's patch.
Figure 8.3 Follicular colonization in a low-grade gastric B cell MALT lymphoma.
reactive B cell proliferation is very useful. This can be done either using immunohistochemistry to demonstrate Ig light-chain restriction or molecular genetic methods to show monoclonal Ig gene rearrangement.
74 Extranodal lymphomas
sharp contrast to comparable nodal lymphomas. Fiveand 10-year survivals of 91 and 75 per cent, respectively, have been reported for low-grade gastric MALT lymphoma at stage IE with slightly lower figures (82 per cent at 5 years) for stage IIE.16
High-grade B cell MALT lymphoma
Figure 8.4 Gastric lymph node infiltrated by a low-grade MALT lymphoma causing marked expansion of the marginal zones.
LYMPH NODE INVOLVEMENT AND DISTANT SPREAD Low-grade MALT lymphoma cells first colonize the marginal zones of regional lymph nodes (Fig. 8.4) and this infiltrate extends into the whole interfollicular area with eventual replacement of the entire node. Follicular colonization may occur in lymph nodes. Distant spread occurs late in the course of the disease; the bone marrow is characteristically, but not always, spared. Some cases manifest a tendency to spread to other mucosal sites. IMMUNOPHENOTYPE The cells of low-grade MALT lymphoma express all mature B cell antigens, and may be CD 43+, but are CD 5- and CD 10-. Like marginal zone cells they also express CD 21 and CD 35. In most cases MALT lymphomas synthesize surface and cytoplasmic IgM, and more rarely IgA or IgG.
High-grade transformation of low-grade MALT lymphoma is well documented.17 Coexistence of both lowand high-grade disease is not uncommon and can lead to difficulty in grading the disease as a whole. In many cases of apparently primary (i.e. de novo] extranodal B cell lymphoma arising in sites where low-grade MALT lymphomas are known to occur, careful searching will reveal microscopic foci of low-grade disease suggesting that the high-grade lymphoma is secondary. However, there remain those cases in which no low-grade component can be found and these must be categorized as primary high-grade lymphomas. Whether this group of lymphomas should be categorized as high-grade MALT lymphoma or more properly be designated large B cell lymphoma is problematical. There are several points in favor of classifying such cases as high-grade MALT lymphomas. Histologically, they are indistinguishable from secondary high-grade MALT lymphoma and they manifest similar molecular genetic properties, which distinguish them from nodal large B cell (high-grade) lymphomas. Some reports suggest that the clinical behavior of high-grade MALT lymphoma is similar to that of the low-grade tumors,18 while others have shown that high-grade MALT lymphomas have a less favorable outlook but nevertheless better than that of comparable nodal disease.17
MALIGNANT LYMPHOMA OF THE GASTROINTESTINAL TRACT
MOLECULAR GENETICS
The translocations that characterize other low-grade B cell lymphomas are not found in MALT lymphomas. By contrast, trisomy 3 is a common abnormality.14 Abnormalities of chromosome Ip22, in particular translocation t(l;14)(p22;q32), have been identified as uncommon but recurrent events. A novel gene bd-10 has been shown to be involved in this translocation.15 More recently t(ll;18)(q21;q21) has been identified as a characteristic property of MALT lymphoma and the breakpoint has been cloned.153 Clinical behavior The extremely indolent behavior of low-grade MALT lymphoma and its tendency to remain localized are in
The gastrointestinal tract is the commonest site of extranodal lymphoma and there is considerable geographic variation in its incidence. The intestine is a major secondary lymphoid organ so it is not surprising that it can be the primary site of a wide spectrum of lymphoma types (Table 8.1).
Low-grade B cell MALT lymphoma Low-grade MALT lymphomas may arise anywhere in the gastrointestinal tract but, in Western countries occur most frequently in the stomach. In the Middle East, where gastrointestinal lymphomas are much commoner, the distribution between stomach and intestine is more even.
Malignant lymphoma of the gastrointestinal tract 75
Table 8.1 Primary gastrointestinal non-Hodgkin's lymphoma
Bcell Mucosa-associated lymphoid tissue (MALT) type Low-grade High-grade with or without a low-grade component Immunoproliferative small intestinal disease Low-grade High-grade with or without a low-grade component Mantle cell (lymphomatous polyposis) Burkitt'sand Burkitt-like Other types of low- or high-grade lymphoma corresponding to lymph node equivalents
Tcell Enteropathy-associated T cell lymphoma (EATL) Other types unassociated with enteropathy Rare types (including conditions that may simulate lymphoma)
Low-grade gastric MALT lymphoma Because the stomach is by far the commonest site, gastric MALT lymphoma has served as the paradigm for the group as a whole. Under normal circumstances there is no lymphoid tissue in gastric mucosa. However, accumulation of lymphoid tissue with MALT characteristics follows infection with H. pylori1* (Fig. 8.5). Indeed, the presence of MALT in the stomach is virtually pathognomonic of H. pylori infection and there is strong evidence that the infection is a necessary first step in the pathogenesis of gastric MALT lymphoma. H. pylori is present in over 90 per cent of cases of gastric MALT lymphoma10'19 and epidemiological studies have confirmed the association between the infection and the development of lymphoma.20 Clinically, gastric MALT lymphoma presents as nonspecific dyspepsia; the endoscopic findings are usually similar to those of chronic gastritis with erosions and/or ulceration and not suggestive of malignancy. The histological features of low-grade gastric MALT lymphoma conform to the classical description of MALT lymphoma as outlined above3 (Fig. 8.2). Low-grade gastric MALT lymphomas remain localized to the stomach and gastric lymph nodes for long periods; unlike low-grade B cell lymphomas of lymph nodes, early bone marrow involvement is not characteristic. It was this favorable clinical behavior that resulted in the use of the term 'pseudolymphoma' for these tumors. However, in time, and without treatment, systemic spread occurs. The presence of the B cell clone which will become predominant in the transformation to MALT lymphoma has been demonstrated in H. pylori gastritis specimens taken several years before the development of gastric lymphoma.21'23 A possible explanation for the remarkably favorable clinical behavior of low-grade gastric MALT lymphoma
Figure 8.5
Gastric mucosa showing the effects of H. pylori
infection. There is a prominent B cell follicle with an associated lymphoepithelium (arrows) and surrounding chronic inflammation of the lamina propria.
is that it is immunoresponsive and, if so, H. pylori would be the most likely antigen. In vitro studies have confirmed that this is the case by showing specific responses of lymphoma cells to individual strains of H. pylori.24 Several clinical studies have now shown that eradication of H. pylori with appropriate antibiotics can lead to regression of the lymphoma.25-27a High-grade gastric MALT lymphoma There are few objective studies on the incidence of highversus low-grade gastric lymphoma but the present consensus is that high-grade disease is commoner. Many, if not most, cases evolve from the low-grade MALT lymphoma, therefore, not surprisingly, the same relationship to H. pylori infection can be shown.20 However, highgrade gastric lymphoma is not an immunoresponsive tumor, therefore, although eradication of H. pylori is probably a prudent step in terms of avoiding recurrences, it will not per se have any therapeutic effect on the lymphoma.
76 Extranodal lymphomas
MALT lymphoma of the intestine Distinction must be made between the usual type of intestinal lymphoma that has no apparent epidemiological specificity, sometimes called the 'western type', and immunoproliferative small intestinal disease (IPSID), which is a special type of MALT lymphoma that occurs in the Middle East and other defined geographic areas. 'WESTERN-TYPE* INTESTINAL MALT LYMPHOMA
These lymphomas can arise anywhere in the intestinal tract but, despite the concentration of MALT (Peyer's patches) in the terminal ileum, they are more common in the jejunum,28 where they may well arise from some form of 'acquired MALT'. High-grade lymphoma is considerably commoner and the relationship between low-grade intestinal MALT lymphoma is similar to that already described.28 The clinical behavior is not as favorable as that of gastric MALT lymphoma with reported 5-year survivals of 44-75 per cent for lowgrade and 25-37 per cent for high-grade disease.28 IMMUNOPROLIFERATIVE SMALL INTESTINAL DISEASE
This special variant of small intestinal MALT lymphoma occurs in the Middle East29 and has also been reported in the Indian subcontinent and the Cape region of South Africa.30 Most cases occur in patients under 30 and present with profound malabsorption. An alpha heavychain paraprotein is present in the serum in 70 per cent of cases, which has led to the term alpha chain disease. IPSID usually involves long segments of the jejunum. Histologically, the features are similar to other MALT lymphomas20 except that all cases show marked plasma cell differentiation. These plasma cells synthesize alpha heavy chain and, with the exception of a few cases, no light chain. IPSID runs a prolonged course and spreads out of the abdomen only in the terminal stages, usually following high-grade transformation. IPSID is similar to lowgrade gastric MALT lymphoma in that some cases respond to broad-spectrum antibiotics presumably following the eradication of an as yet unknown antigen in the small intestine.31 Mantle cell lymphoma Although mantle cell lymphoma typically presents as a nodal lymphoma, a minority of cases appears to arise primarily in the gastrointestinal tract.32 The clinical presentation is usually that of lymphomatous polyposis with multiple lymphomatous polyps peppering the intestinal and sometimes the gastric mucosa. The histological features are those of the nodal disease, and comprise a diffuse or nodular infiltrate of cells, which often
Figure 8.6
Small intestinal infiltration by mantle cell
lymphoma. Note the 'naked'follicle centers.
resemble centrocytes and typically surround 'naked' follicle centers (Fig 8.6). In small biopsy specimens the differential diagnosis between mantle cell lymphoma and MALT lymphoma can be extremely difficult. Unlike MALT lymphoma, lymphoepithelial lesions and transformed blasts are rare or absent in mantle cell lymphoma, and CD 5 expression is common. In contrast with low-grade MALT lymphoma, mantle cell lymphoma is an aggressive disease that soon disseminates beyond the gastrointestinal tract to involve peripheral lymph nodes and bone marrow. The median survival after diagnosis is 3-5 years.33 Burkitt's and Burkitt-like lymphoma In the Middle East, classical Burkitt's lymphoma commonly presents as an intestinal, usually ileocecal, tumor.34 The disease has a propensity to occur in boys under 10 years of age. In Western countries, childhood gastrointestinal lymphoma is rare but the commonest form, although not fulfilling all the criteria, bears a close resemblance to Burkitt's lymphoma as described above. Other types of B cell nodal lymphoma Although, theoretically, any type of nodal lymphoma could arise from gastrointestinal MALT, if strict criteria for primary disease are applied, primary gastrointestinal lymphomas other than those described earlier are extremely rare. Thus, for example, although follicular lymphoma is the commonest nodal non-Hodgkin's lymphoma, only a handful of cases of truly primary gastrointestinal follicular lymphoma have been described. Enteropathy-associated T cell lymphoma
(|^[|L_________ It has been known for many years that primary intestinal lymphoma occurs with greater frequency in patients
Malignant lymphoma of the lung 77
with celiac disease. The lymphoma has since been characterized as a distinct type of T cell lymphoma, which probably arises from intraepithelial T cells.35'36 Clinically, EATL usually presents with acute intestinal hemorrhage, perforation or obstruction in a patient with a history of gluten-sensitive enteropathy. In some cases there is no clinical history of malabsorption but villous atrophy is found in the resection specimen, while, more rarely, the intestinal mucosa appears normal. Currently, it is thought that patients representative of this last group are suffering from a latent from of enteropathy in which histological changes in the mucosa are minimal or absent. Multiple 'benign' chronic inflammatory small intestinal mucosa ulcers are often present in EATL and may be the dominant, or only, pathological lesion. Careful histological and, more recently, molecular genetic analysis has shown that these ulcers in fact are lymphomatous.37 The proximal small intestine is more commonly involved in EATL, which produces either large tumor masses, ulcers or an extensive diffuse cryptic lesion. In most cases the histological appearances of EATL are those of a highly pleomorphic large cell lymphoma, often with considerable accompanying inflammation, which can be of such a degree as to mask the neoplastic T cells. Intraepithelial tumor cells, sometimes smaller and less pleomorphic, may be prominent. The clinical course of EATL is very unfavorable since in most cases the lymphoma involves multiple foci of the small intestine and has already disseminated peripherally.
MALIGNANT LYMPHOMA OF THE SALIVARY GLANDS For a number of reasons, which include recent changes in diagnostic criteria, the presence of entrapped lymph nodes in the parotid glands and ectopic salivary tissue within these nodes, the true incidence of salivary gland lymphoma is difficult to estimate. It probably accounts for 10-15 per cent of all malignant salivary gland tumors.38 The normal salivary gland contains no organized lymphoid tissue but, as a consequence of Sjogren's syndrome and other, as yet undefined, chronic inflammatory conditions, MALT accumulates in salivary tissue as Peyer's patch-like structures around salivary ducts associated with a prominent lymphoepithelium11 (Fig. 8.7). The term myoepithelial sialadenitis (MESA) is often used to describe this lesion. It is from this acquired MALT that most salivary lymphomas arise. Patients with primary salivary gland lymphoma are usually over 50 and there is a marked female preponderance; this is a reflection of the much higher incidence of Sjogren's syndrome and accompanying MESA in females. There is often a long history of salivary gland (usually parotid) enlargement prior to the development
Figure 8.7 Parotid gland from a patient with Sjogren 's syndrome showing the accumulation of MALT around the dilated ducts.
of lymphoma. In some cases lymph node dissemination with cervical lymphadenopathy is the first evidence that the salivary gland is harboring a lymphoma. Although lymphoma can arise in any of the major or minor salivary glands, the parotid is by far the commonest site. In most cases the lymphoma is of the low-grade MALT type with prominent lymphoepithelial lesions (Fig. 8.8); high-grade transformation can occur as previously described. Dissemination is a late phenomenon and results in the typical marginal zone or interfollicular infiltrate in regional lymph nodes. This appearance has frequently led to the mistaken diagnosis of primary nodal 'monocytoid' B cell lymphoma in cervical lymph node biopsies from patients with Sjogren's syndrome.39 The clinical course of salivary gland MALT lymphoma is often prolonged but unpredictable with systemic dissemination to lymph nodes and bone marrow occurring at any time from a few months to many years following the diagnosis.40 Other types of lymphoma, including T cell lymphoma, occur with extreme rarity as primary salivary gland tumors. Most reports of such cases are, in fact, describing lymphomas of intraparotid lymph nodes rather than the salivary gland itself.
MALIGNANT LYMPHOMA OF THE LUNG Because almost all cases of pulmonary MALT lymphoma were, until fairly recently, called pseudolymphoma, the incidence of primary pulmonary lymphoma, most of which are of the MALT type, is unknown. The normal lung contains no organized lymphoid tissue but a number of conditions, including so-called follicular bronchiolitis, lead to the accumulation of MALT in relation to small bronchi or bronchioles.3'41 Although pulmonary MALT lymphoma must be assumed to arise from this acquired MALT, unlike the stomach and to a lesser extent the salivary glands, no defined prelymphomatous
78 Extranodal lymphomas
Figure 8.8
Low-grade salivary gland MALT lymphoma showing
a reactive B cell follicle surrounded by the neoplastic infiltrate within which is a prominent lymphoepithelial lesion (arrows).
chronic inflammatory condition has been defined in the lung. Low-grade pulmonary MALT lymphomas usually occur in elderly patients and most are discovered accidentally as a well-defined opacity following a routine chest X-ray. Histologically, the typical features of MALT lymphoma are present (Fig. 8.9) with lymphoepithelial lesions produced by lymphomatous invasion of bronchiolar epithelium. Dissemination to local lymph nodes or systemic spread is rare and high-grade transformation, although well documented, is rare. The clinical behavior of low-grade pulmonary MALT lymphoma is remarkably indolent,42 and cure is easily achieved following either surgical resection or radiotherapy. There are insufficient data for evaluation of the behavior of high-grade B cell lymphomas of the lung.
Lymphomatoid granulomatosis The term lymphomatoid granulomatosis was originally coined to describe a curious progressive chronic inflammatory condition.43 The lymphomatous nature of this disease only gradually became apparent and, until
Figure 8.9
MALT lymphoma of the lung showing the
neoplastic infiltrate surrounding a follicle center and forming lymphoepithelial lesions with bronchiolar epithelium (arrows).
recently, it was thought that the lymphoma was of T or NK cell origin and fell into the loosely characterized group of angiocentric lymphomas. The B cell nature of lymphomatoid granulomatosis and its association with the Epstein-Barr virus (EBV) has recently been clarified.44 Lymphomatoid granulomatosis presents with hemoptysis due to multifocal necrotizing lesions of the lungs. In a minority of cases there is a history of therapeutic immunosuppression. Dissemination outside the lung may be present at the time of diagnosis. Histologically, there is a mixed lymphoid infiltrate often with an angiocentric pattern and large areas of necrosis. Most of the cells are small T lymphocytes but also present are scattered large transformed B blasts, which can be shown to be monoclonal. These large blasts contain EBV. Lymphomatoid granulomatosis is an aggressive disease. Most patients die within 2 years of diagnosis despite intensive chemotherapy.
Other pulmonary lymphomas Primary T cell lymphomas of the lung with features similar to lymphomatoid granulomatosis are rare but well
Lymphomas of the ocular adnexa and eye 79
documented.45 High-grade B cell lymphomas associated with immunodeficiency, other than lymphomatoid granulomatosis also occur and will be described later in this chapter. Hodgkin's disease rarely arises primarily in the lung. Lymphoma of the pleura A distinctive type of high-grade B cell lymphoma arising in the pleura has been reported in Japan46 and, more recently, similar cases have been described in France.47 This type of lymphoma arises in the setting of longstanding pyothorax and is associated with the presence of EBV genome in all the neoplastic cells.
MALIGNANT LYMPHOMA OF THE THYROID The close association between Hashimoto's thyroiditis and thyroid lymphoma has been appreciated for many years but, until the advent of the MALT concept, it was thought that thyroid lymphoma was heterogeneous, comprising the full spectrum of nodal tumors. It is now clear that virtually all lymphomas arising in the thyroid are low- or high-grade MALT lymphomas.12 The lymphoid tissue that accumulates in the thyroid gland in Hashimoto's disease is characterized by B cell follicles, a minor perifollicular component and pronounced plasma cell differentiation. Focally, a lymphoepithelium, comprising B cells within the thyroid acinar epithelium, can be identified but this is not as florid as in other examples of acquired MALT. Like Hashimoto's disease, most cases of thyroid MALT lymphoma occur in females. A sudden increase in the size of the thyroid in a patient with Hashimoto's disease should raise the suspicion of lymphoma. The histological features of low-grade MALT lymphoma of the thyroid are notable for the prominence of lymphoepithelial lesions, which often comprise thyroid acini stuffed with lymphoma cells; reactive B cell follicles are often prominent and follicular colonization is common (Fig. 8.10). High-grade transformation maybe seen. There are few adequate follow-up studies of patients with either high- or low-grade thyroid MALT lymphoma. The presence of capsular invasion is thought to worsen the prognosis significantly.
LYMPHOMAS OF THE OCULAR ADNEXA AND EYE The ocular adnexa comprise the eyelids, conjunctiva, lacrimal gland and orbit. No significant clinicopathological differences exist between lymphomas arising in
Figure 8.10
Low-grade MALT lymphoma of the thyroid with a
lymphoid follicle surmounted by the neoplastic infiltrate, which forms a lymphoepithelial lesion with a thyroid acinus.
individual adnexa and, in any event, lymphoma arising in any one of these structures may soon extend into another. Ocular adnexal lymphomas are therefore best considered as a single entity. This homogeneity extends to histological type; virtually all lymphomas arising in the ocular adnexa are of the MALT type and most are low grade. Like other sites where MALT lymphomas occur, there is normally no organized lymphoid tissue in the ocular adnexa but a post-mortem study has shown that MALT is acquired in the ocular adnexa with increasing frequency according to age.48 Lymphomas of the ocular adnexa occur characteristically in patients over 50. The orbit is the commonest site, often together with lacrimal gland involvement, followed by the conjunctiva. Bilateral involvement is present in 15 per cent of cases. The histology of ocular adnexal lymphomas is typical of MALT lymphoma with a low frequency of high-grade transformation. Use of the term 'MALT' in describing localized orbital lymphomas is open to question, since there is no mucosa or epithelium in the orbit. However, with the exception that lymphoepithelial lesions are not present, the architecture and cytology of orbital
80 Extranodal lymphomas
lymphomas is identical to that of the other adnexa. Purists should, perhaps, use the term 'marginal zone' to designate orbital lymphomas. Progression, with systemic involvement, occurs in 15-20 per cent of patients diagnosed with stage IE lymphoma of the ocular adnexa. The interval between diagnosis and dissemination is very variable and unpredictable.
Table 8.2 Lymphomas of the skin
Tcell Epidermotropic small cerebriform cell Mycosis fungoides Sezary syndrome Non-epidermotropic small cell Non-epidermotropic pleomorphic large cell Anaplastic large cell
Bcell
Intraocular lymphoma This very rare tumor is usually a manifestation of highgrade B cell lymphoma of the brain. It carries a very poor prognosis.
MALIGNANT LYMPHOMA OF THE SKIN The skin is the third commonest site of primary extranodal lymphoma after the gastrointestinal tract and Waldeyer's ring.1 Unlike other extranodal lymphomas, those arising in the skin, or at least some of them, have been recognized as a special entity in various lymphoma classifications for many years, and also discussed at length in many dermatological and hematopathological publications. A full discussion of this complicated group of lymphomas is, clearly, beyond the scope of this chapter and only a concise summary will be presented here. The high concentration of immunological accessory cells present in the skin as the intraepidermal Langerhans cell population suggests that the skin is an important immunological organ. However, the normal resident lymphoid population is confined to a thin scattering of intraepidermal T cells, which are often in contact with the Langerhans cells, and an even more scanty intradermal T cell population. These resident cutaneous T cells express a distinctive 200 kDa surface antigen (HECA-452),49 which suggests that, despite their small numbers, they are especially committed to the skin. Many inflammatory dermatoses lead to the accumulation of organized lymphoid tissue in the skin and some cutaneous lymphomas may arise from this in a manner similar to MALT lymphoma. A wide variety of both T and B cell lymphomas can arise in the skin as summarized in Table 8.2.
Epidermotropic small cerebriform cell lymphomas This form of lymphoma, which accounts for more than half of all cutaneous T cell lymphomas, is commonly known as mycosis fungoides or, when circulating cells are present, Sezary syndrome. Dermatologists recognize multiple clinical variants.
Low-grade marginal zone type High-grade
Mycosis fungoides characteristically occurs in patients over 50 and presents as widespread erythematous patches, which, as the disease progresses, evolve into plaques and, finally, nodular tumors. The presence of circulating lymphoma cells (Sezary syndrome) is associated with pronounced erythroderma. In the early (patch) phase the histological changes resemble those of inflammatory dermatoses and the diagnosis can be very difficult. The classical features of a dense subepithelial infiltrate and intraepithelial collections of atypical cells with cerebriform nuclei are hallmarks of the later plaque phase, while large sheets of tumor cells, often showing high-grade transformation, are present in the final tumor phase. The clinical course is variable and related to the clinical stage at presentation. Over 50 per cent of patients die within 10 years of diagnosis following visceral or nodal involvement.
Non-epidermotropic small T cell lymphoma This is a poorly characterized and rare group of lymphomas whose main importance is to draw attention to the fact that epidermotropism is not a sine qua non for the diagnosis of cutaneous T cell lymphoma.
Non-epidermotropic pleomorphic large T cell lymphoma This type of lymphoma accounts for 5-10 per cent of cutaneous T cell lymphomas. It presents as distinct, sometimes multiple tumor nodules in the skin and, histologically, shows sheets of intradermal large and often pleomorphic lymphoma cell without epidermotropism. The lack of expression of the CD 30 antigen serves to define the poor prognosis of this group.
Anaplastic large cell lymphoma (ALCL) Until its lymphoid nature was clarified in 1985,50 this tumor was thought to be derived from histiocytes. In the skin it was known as 'regressing atypical histiocytosis'.
Malignant lymphoma of the upper aerodigestive tract 81
Primary cutaneous ALCL can occur at any age but is commoner in patients over 60. It presents as single or multiple ulcerating nodules. Histologically, ALCL is characterized by sheets of large cells with abundant eosinophilic cytoplasm and pleomorphic nuclei with prominent nucleoli (Fig. 8.11). An important defining feature is strong expression of the CD 30 antigen.51 Unlike its nodal equivalent, the t(2;5) translocation is not a feature. Some cases of cutaneous ALCL regress spontaneously. Overall, the disease has a remarkably good prognosis with tumor-related death occurring in only 10 per cent.52 This is in striking contrast to nodal ALCL, which is a much more aggressive tumor. Lymphomatoid papulosis This disease, which may be related to ALCL, is characterized by continuously developing, self-healing papular eruptions, usually on the trunk and limbs, in otherwise healthy individuals. It occurs in all age groups. Histologically, the appearances are those of a malignant lymphoid neoplasm with many features in common with ALCL except that the CD 30+ atypical cells tend to occur in small groups or singly, separated by a chronic inflammatory infiltrate.53,54 The T cell proliferation has been shown to be monoclonal.55 In the majority of patients, lymphomatoid papulosis 'burns itself out' after a few years but a minority progress to develop T cell lymphoma.
Low-grade B cell lymphoma of marginal zone type These uncommon tumors present as single or multiple violaceous plaques, or nodules on any part of the body. Histologically,3 their structure is similar to that of MALT lymphoma with reactive follicles surrounded by a monotypic population of small, often centrocyte-like or monocytoid B cells in the marginal zone. However, lymphoepithelial lesions are only rarely seen. Scattered blasts are characteristic and plasma cell differentiation is present in some cases. Primary low-grade marginal zone lymphoma of the skin runs an extremely indolent clinical course. Multiple recurrences and/or high-grade transformation may occur but the lymphoma remains localized in the skin for many years. High-grade B cell lymphoma These tumors characteristically occur in patients over 50 and present as red or violaceous nodules, which usually do not ulcerate. They tend to occur on the head, neck or back.56 In some cases there is histological evidence that the lesion has evolved from a low-grade B cell lymphoma of marginal zone type. The tumor is composed of sheets of large B cells, which may be separated from the epidermis by a so-called Grenz zone. Despite their histology, high-grade B cell lymphomas of the skin, like the low-grade B cell lymphomas, are remarkably indolent tumors. Only 2-5 per cent ever spread beyond the skin.
B cell lymphomas of the skin B cell lymphomas may arise primarily in the skin. Because their pathology has only recently been defined,3 their true incidence is unknown but they are uncommon. The majority are high-grade lymphomas but in some of these a residual low-grade component can be recognized.
Figure 8.11 (a) An anaplastic large cell lymphoma of the skin. 30 expression by tumor cells.
MALIGNANT LYMPHOMA OF THE UPPER AERODIGESTIVE TRACT This loosely defined anatomic area includes the nose, paranasal sinuses, mouth, tongue and the lymphoid
The same case stained with antibody to CD 30 showing strong CD
82 Extranodal lymphomas
tissue comprising Waldeyer's ring. By convention, the salivary glands are not included. The upper aerodigestive tract is the second commonest site of extranodal lymphoma1 but is also commonly involved in disseminated nodal disease; careful staging is therefore necessary before making a diagnosis of extranodal lymphoma at this site. The upper aerodigestive tract is an important site of lymphoid tissue, which occurs as well-recognized structures, such as the palatine tonsils and other elements of Waldeyer's ring, and also as less well-defined collections distributed throughout the mucosae. There are many anatomical similarities between the lymphoid tissue of Waldeyer's ring and Peyer's patches. Reactive B cell follicles are the most prominent component and between these there is the T cell zone. Beneath the overlying squamous epithelium is a mixture of plasma cells and marginal zone B cells, both of which appear to pass freely into the epithelium. There is, thus, intimate contact between the lymphoid tissue and covering squamous epithelium, which is enhanced in the palatine tonsil by the formation of deep complex crypts. Like Peyer's patches, the lymphoid tissue of Waldeyer's ring lacks afferent lymphatics and antigen gains access directly across the covering epithelium. The palatine tonsils account for most (over 40 per cent) of upper aerodigestive tract lymphomas with the next most common sites in order of frequency being the nasopharynx, the soft palate and the base of tongue. The lymphomas that arise in the upper aerodigestive tract comprise a mixture of nodal and extranodal types (Table 8.3). Low-grade B cell lymphomas of nodal type Immunocytoma, plasmacytoma, follicle center lymphoma and mantle cell lymphoma may all rise in the upper aerodigestive tract. Of these, only plasmacytoma occurs with any frequency elsewhere in the aerodigestive tract. Most plasmacytomas occur in the nose but they account for over 20 per cent of lymphomas of the oral cavity.57 A feature of plasmacytoma of the aerodigestive tract is the high frequency of amyloid deposition within the tumor.
Table 8.3 Lymphomas of the upper aerodigestive tract Bcell Low-grade nodal types, including plasmacytoma MALT type Large cell types (high-grade) Tcell Angiocentric (natural killer cell)
Low-grade MALT B cell lymphoma MALT lymphomas are infrequent in the organized native lymphoid tissue of Waldeyer's ring, a situation similar to that in Peyer's patches. Rare cases have been reported in the nose, nasopharynx, larynx and trachea.
Large (high-grade) B cell lymphoma These account for approximately 85 per cent of lymphomas of Waldeyer's ring and 30-50 per cent of lymphomas of the oral soft tissues.58'59 There may be an associated residual low-grade component of a nodal or extranodal type of B cell lymphoma. Localized, stage IE high-grade lymphomas of Waldeyer's ring have a good prognosis, with 80 per cent surviving 5 years or more. More advanced lymphomas and those of the oral cavity have a worse prognosis.58"61
Angiocentric T and/or NK cell lymphoma For many years a progressive necrotizing lesion of the upper aerodigestive tract has been recognized as an inflammatory disorder, most often called 'lethal mid-line granuloma'. The lymphomatous nature of this condition gradually became apparent and it's now established that this is a single clinicopathological entity probably derived from NK cells.62 The disease occurs most commonly in the Far East and certain parts of South America including Peru.63 The clinical presentation is that of a slowly enlarging ulcer involving, in order of decreasing frequency, the nose, paranasal sinuses, palate, mouth and tongue. The ulceration is progressive and may erode into the facial skeleton. The histology is extremely variable and the neoplastic nature of the infiltrate is often masked by intense chronic inflammation and extensive necrosis. Characteristically, the malignant cells have pleomorphic nuclei and delicate pale staining cytoplasm are arranged in an angiocentric pattern with invasion of the walls of small muscular arteries. In some cases the immunophenotype is typical of NK cells (CD 2+, CD 3-, CD 56+) but in others the cells express the T cell antigen CD 3. However, in favor of an origin from NK cells, rather than T cells, is the absence of T cell receptor gene rearrangement. Another consistent finding is the presence of monoclonal EBV genome in the malignant cells.64 Left untreated or undertreated these lymphomas are relentlessly locally progressive and can cause grotesque disfigurement. They preferentially spread to the skin, lungs and gastrointestinal tract, and seldom involve lymph nodes; meaningful data on the prognosis are not yet available.
Malignant lymphoma of the thymus 83
MALIGNANT LYMPHOMA OF THE CENTRAL NERVOUS SYSTEM Previously rare, there is now a rising incidence of primary lymphoma of the central nervous system (CNS).65 This increase is independent of those lymphomas occurring in immunocompromised patients, which will be discussed as a group elsewhere in this chapter. The peak incidence is in the sixth decade and most occur in the superficial parts of the brain; however, any site, including the spinal cord, may be involved and CNS lymphoma may be multifocal. The clinical presentation is that of any space occupying lesion. CNS lymphomas are almost always B cell tumors but occasional cases of T cell lymphoma have been described. Low-grade B cell lymphomas account for 10-20 per cent of cases. Most of these are lymphoplasmacytic lymphomas (immunocytomas). The pattern of infiltration is diffuse, without necrosis, and with conspicuous perivascular cuffing (Fig. 8.12). Most primary CNS lymphomas are high grade and result in large necrotizing masses; perivascular cuffing is also a characteristic finding in these high-grade tumors. Primary CNS lymphoma has a very poor prognosis with a median survival of less than 2 years.66 A pattern of recurrences within the CNS is the rule but distant spread is well recognized.
The thymus essentially consists of an epithelial skeleton extending from cortex to medulla where it forms the HassalPs corpuscles. Packed within this epithelial framework are proliferating T cells whose phenotype steadily matures with progression from cortex to medulla. Less well appreciated is the recently identified intramedullary B cell population, which is most concentrated around Hassall's corpuscles.67
Hodgkin's disease The thymus may be the site of presentation of Hodgkin's disease. Almost all cases are of the nodular sclerosis subtype. Hodgkin's disease presenting with mediastinal widening due to thymic involvement is especially common in young women.
T cell lymphoblastic lymphoma Thymic enlargement maybe the presenting sign of T cell acute lymphoblastic leukemia. The use of the term 'lymphoma' in this disease should be restricted to those cases in which the tissue phase is clearly separate from the leukemia; leukemia, however, invariably follows. This disease pattern is commonest in young males.
Low-grade B cell lymphoma of MALT type MALIGNANT LYMPHOMA OF THE THYMUS Most primary tumors of the thymus are epithelial rather than lymphoid in origin. The thymus is also the only extranodal site where Hodgkin's disease occurs with any frequency. Not surprisingly, given its role in T cell ontogenesis, T cell lymphoblastic lymphoma, usually associated with acute lymphoblastic leukemia, is the commonest non-Hodgkin's lymphoma in the thymus, but both low- and high-grade B cell lymphoma can arise in the thymus, the former being of the MALT type.
Three cases of this variant of MALT lymphoma have recently been described68,69 and the author has seen a further case. Interestingly, three of the four patients have been Chinese. One patient had a long history of Sjogren's syndrome. This form of lymphoma presents with mediastinal widening. The histology is typical of low-grade MALT lymphoma with prominent lymphoepithelial lesions formed by lymphomatous infiltration of Hassall's corpuscles. The indolent clinical behavior is typical of MALT lymphomas as a group.
High-grade B cell lymphoma of the thymus
Figure 8.12 A low-grade (lymphoplasmacytic) lymphoma of the brain showing marked perivascular cuffing.
This type of lymphoma was first recognized in 1980 and called 'sclerosing high-grade lymphoma of the mediastinum'.70 As more cases were reported, less emphasis was placed on the sclerosis and the thymic origin of the lymphoma was clarified.71,72 Although mentioned as a possible variant of large B cell lymphoma, high-grade B cell lymphoma of the thymus is not yet recognized as a disease entity in the REAL classification.7 There is speculation that thymic high-grade B cell lymphoma may arise from the distinctive thymic intramedullary B cell population. Seventy per cent of cases occur in patients under 35, 60 per cent of whom are females. The disease presents as a locally invasive mediastinal mass.
84 Extranodal lymphomas
Figure 8.13
High-grade B cell lymphoma of the mediastinum
(thymus). There is striking sclerosis with pocketing of the tumor cells.
Histologically, the large pleomorphic lymphoma cells are usually compartmentalized by sclerotic bands (Fig. 8.13). Thymic remnants, usually in the form of epithelial cysts, or less commonly as Hassall's corpuscles, are trapped within the tumor. High-grade B cell lymphoma of the thymus is locally highly invasive. Dissemination commonly involves extranodal sites, such as the kidney, liver and brain. The prognosis is dependent on the stage and is similar to that for nodal high-grade B cell lymphoma.
MALIGNANT LYMPHOMAS OF THE SPLEEN Although the spleen is a major lymphoid organ and very often involved by disseminated nodal lymphoma, primary lymphomas of the spleen are uncommon, accounting for less than 2 per cent of all lymphomas. The histology of splenic lymphoid tissue is complex and can only be summarized briefly here. The so-called white pulp comprises a sheath of T cells around the branching splenic arteries from which B-cell follicles are 'suspended' at regular intervals. Splenic B cell follicles are distinguished by the presence of a prominent marginal zone (Fig. 8.14) but otherwise are similar to those of the lymph nodes. The splenic red pulp consists of a complex network of anastomosing venous sinuses separated by an intervening sponge-like network of macrophages and 'reticular' cells. Small aggregates of B and T cells are present in the red pulp. The splenic circulation is remarkably complex, being designed in the interests of its functions as an organ concerned with mounting immune responses and acting as a red cell filter. Any type of B or T cell lymphoma can arise in the spleen, where its clinicopathological features are otherwise identical to those of the nodal disease. Although relatively rare, this group accounts for the majority of cases of 'primary' splenic lymphoma. To complicate matters further, splenic infiltration is often a feature of the B cell
Figure 8.14
A section of a normal spleen showing the white
pulp comprising a B cell follicle surrounded by a prominent marginal zone.
and other leukemias that not infrequently present with splenomegaly. Only two types of lymphoma characteristically arise in the spleen as opposed to peripheral lymph nodes; these are splenic marginal zone B cell lymphoma and gamma/delta T cell lymphoma.
Splenic marginal zone lymphoma The incidence of this recently described disorder is not yet known.73'74 It has a median age of onset of 60 years and there appears to be a female preponderance. Patients characteristically present with massive splenomegaly, which may be associated with the various cytopenias and anemias of hypersplenism. In a significant number of cases a peripheral lymphocytosis is present due to 'spillage' of the lymphoma into the peripheral blood. In many of these cases the circulating lymphoma cells are characterized by short, often polarized villi and this has given rise to the term 'splenic lymphoma with villous lymphocytes'.75 In splenic marginal zone lymphoma, the B cell follicles are greatly enlarged by an infiltrate of small and medium-sized lymphocytes that surround and replace the follicle center with obliteration of the mantle zone (Fig. 8.15). The medium-sized cells resemble marginal zone cells and tend to concentrate at the periphery of the nodules, which lends a 'marginal zone' appearance to the lymphoma. The red pulp is also heavily infiltrated with prominent intrasinusoidal invasion. Despite its name, splenic marginal zone lymphoma is probably not derived from splenic marginal zone cells and bears no resemblance to MALT lymphoma, which exhibits true marginal zone cell characteristics. In splenic marginal zone lymphoma neither the architecture nor the immunophenotype resemble those of the normal splenic marginal zone. Clinical experience with this newly described entity is still limited. Currently, it seems that splenectomy rather than chemotherapy is the treatment of choice.76 Although some cases progress and transform to a high-grade
Extranodal lymphomas of miscellaneous sites 85
Primary testicular lymphoma High-grade B cell lymphoma is the commonest testicular tumor in elderly males.78 These usually present as a painless unilateral testicular swelling, although both synchronous and metachronous bilateral examples are well recognized. Histologically, the large high-grade neoplastic B cells surround and replace the seminiferous tubules, often imparting a nodular appearance to the tumor. The prognosis of stage IE and IIE high-grade B cell testicular lymphoma is relatively favorable with a 60 per cent probability of surviving 5 years. Lymphoma of the bladder Bladder lymphoma is commoner in females, probably reflecting the higher incidence of chronic follicular cystitis in women, a condition which is characterized by the acquisition of MALT by the bladder mucosa. Most bladder lymphomas are of the low-grade MALT type79 and manifest the indolent clinical behavior that is characteristic of this group. High-grade B cell lymphomas of the bladder are rare and probably result from transformation of a low-grade MALT lymphoma. Lymphomas of the female genital tract
Figure 8.15
Splenic marginal zone lymphoma showing two
white pulp nodules, one of which contains a residual reactive follicle center (arrow). The red pulp is diffusely infiltrated.
Primary lymphomas of the ovary, fallopian tube, uterus, cervix and vagina have all been documented but are rare. Most reported cases have been high-grade B cell lymphomas but low-grade MALT lymphomas have also been described.
lymphoma, the majority run an indolent course following splenectomy. Splenic gamma/delta T cell lymphoma
EXTRANODAL LYMPHOMAS OF MISCELLANEOUS SITES
This rare lymphoma77 usually presents with splenomegaly and hepatomegaly. Bone marrow infiltration is also usually present or soon follows, and variable numbers of tumor cells are present in the peripheral blood. The splenic white pulp may be preserved, but there is heavy infiltration of the red pulp by small cells with pale cytoplasm and oval nuclei. The tumor cells express CD 3 and the gamma/delta T cell receptor hetrodimer, but are CD 4- and CD 8- and usually also CD 5- and CD 7-. This is an aggressive disease.
Primary lymphomas have been described arising in almost every organ or tissue in the human body. Some, including lymphomas of bone, breast, liver and soft tissues have been reported as substantial series, while others, such as lymphomas of kidney, heart, adrenal, gall bladder and pancreas, feature essentially as single case reports.
MALIGNANT LYMPHOMA OF THE UROGENITAL TRACT Apart from lymphomas of the testis and, to a lesser extent, the urinary bladder, urogenital lymphomas are very rare indeed.
Lymphomas of bone Excluding myeloma, primary lymphomas of bone are rare and most reports of large series have included substantial numbers of cases associated with nodal disease. Fifty per cent of bone lymphomas80 present as single osteolytic lesions in the long bones but any bone may be involved. Polyostotic cases have been described. Virtually all reported cases have been high-grade B cell
86 Extranodal lymphomas
lymphomas. The clinical behavior depends on the stage with a 55 per cent overall survival for stage IE and IIE cases. With soft tissue or nodal involvement this falls to 10-20 per cent. Lymphomas of the breast Lymphoma accounts for approximately 0.5 per cent of all malignant breast tumors.81 Twenty per cent of reported cases have been examples of Burkitt's lymphoma and these will be discussed separately together with other aspects of this disease. Lymphoma of the breast usually presents in older women as a unilateral lump indistinguishable from carcinoma. Most breast lymphomas are high-grade B cell tumors; there are occasional reports of low-grade MALT lymphoma arising in the breast. The clinical behavior of breast lymphoma is not well documented. Lymphoma of the liver More than 50 cases of primary hepatic lymphoma have been reported. Most have occurred in males over 50 and a history of chronic liver disease has been documented in a substantial minority of cases. Almost all reported cases of liver lymphoma have been highgrade B cell tumors,82 although recently a series of lowgrade MALT lymphomas of the liver has been described.83
Endemic Burkitt's lymphoma Burkitt's original descriptions were of an endemic tumor, later characterized as a lymphoma, that classically presented in the jaw, more often in the maxilla.84 Multiple quadrants of the jaw are frequently involved. The next most frequent site of presentation is the abdomen. Here the retroperitoneum, kidneys, adrenal glands and ovaries may be involved. Less frequent sites of presentation include the eyelids (lacrimal glands), salivary glands, testes and female breasts. In the breasts, the tumor is bilateral and typically associated with pregnancy; interruption of the pregnancy is rapidly followed by involution of the lymphoma. Even in advanced cases there is relative sparing of the lymph nodes and bone marrow.85 Sporadic Burkitt's lymphoma Although the histology of sporadic Burkitt's lymphoma is identical to that of the endemic disease, its biology, including sites of origin, is quite different.86 Jaw masses are only rarely encountered and most cases present with lymphoma in the abdomen due to involvement of the intestine, particularly the ileocecal region, ovaries and retroperitoneal lymph nodes. Other sites of presentation include Waldeyer's ring, peripheral lymph nodes and the bone marrow, where there is often an associated leukemic component. Angiotropic large cell lymphoma (Fig. 8.16)
Lymphomas of soft tissues Lymphomas have been reported arising in most soft tissues. The thigh, abdominal wall and limbs are the commonest sites. The histological type is highly variable and includes both low- and high-grade, mostly B cell, lymphomas.
MULTIFOCAL EXTRANODAL LYMPHOMA Two types of extranodal lymphoma, namely Burkitt's lymphoma and angiotropic (intravascular) large-cell lymphoma, typically present at multiple sites simultaneously.
This curious type of lymphoma87,88 was originally thought to be a tumor of endothelial origin and cases were labeled as variants of angioendotheliomatosis. Most cases present de novo with ill-denned skin lesions, CNS signs, including blindness, and occasionally adrenal insufficiency. Less commonly, there is a preceding solid lymphomatous deposit. The characteristic histological changes are seen most often in the skin, brain and adrenal gland but may be surprisingly widespread. There is distension of capillaries venules and small arteries by dyscohesive masses of large pleomorphic lymphoma cells. Only rarely do the lymphoma cells invade surrounding tissues. Most cases have exhibited a B cell phenotype but a few T cell cases have been reported. Angiotropic large cell lymphoma is an aggressive disease with a median survival of only 5 months.
Burkitt's lymphoma Burkitt's lymphoma84 provides a graphic illustration of the biological relevance of the site of origin of lymphoma. The mechanisms underlying the predilection of Burkitt's lymphoma for discrete anatomical sites remain unknown but potentially could provide useful insights into treatment.
EXTRANODAL LYMPHOMA ASSOCIATED WITH IMMUNODEFICIENCY The increased risk of developing lymphoma in patients with congenital immunodeficiency has long been
References 87
phoma' for this group of EBV-associated lesions evolved for two reasons. Despite the classical histological features of a high-grade B cell lymphoma, molecular studies of some cases failed to show monoclonal immunoglobulin gene rearrangement and in other cases, occurring in immunosuppressed organ transplant recipients, the 'lymphoma' regressed completely following decrease in the dose of immunosuppressive drugs and restoration of immunocompetence. Currently, there are no objective criteria by which such cases can be recognized and, in any event, in patients whose immunodeficiency is due to AIDS, the immunodeficiency cannot, as yet, be reversed. With the exception of those few cases that respond to immunological manipulation, polymorphic immunoblastic lymphoproliferation is rapidly progressive.
Burkitt-like lymphoma This type of lymphoma is named for its resemblance to Burkitt's lymphoma from which it shows subtle cytological and molecular genetic differences. Only 20-30 per cent of cases are characterized by the presence of EBV in the neoplastic cells. All cases are monoclonal and reversal of the immunosuppression has no effect in this disease. As to be expected, the prognosis is poor.92
REFERENCES Figure 8.16
Angiotropic large cell lymphoma of the brain.
recognized. The association of immunodeficiency and lymphoma has recently been highlighted by the increasing use of therapeutic immunosuppression in organ transplant recipients, and the advent of the acquired immunodeficiency syndrome (AIDS) epidemic. Although there are subtle differences between the lymphomas associated with particular causes of immunodeficiency, there are many more similarities, which are sufficient to suggest that immunodeficiency associated lymphomas as a group share a common pathogenesis and manifest common biological features. These include a strong tendency to arise extranodally, similar histology and the frequent presence of EBV in the tumor cells. Lymphomas associated with immunodeficiency can arise at any extranodal or nodal site amongst which the gastrointestinal tract and brain are the commonest. The histological types include Hodgkin's disease,89 which arises in lymph nodes, and is therefore strictly beyond the brief of this chapter, polymorphic immunoblastic B cell lymphoproliferations90 and Burkitt-like lymphoma.91 Occasional cases have been characterized as T cell lymphoma.
Polymorphic immunoblastic B cell lymphoproliferations Use of the term 'lymphoproliferation' rather than 'lym-
1. Freeman C, Berg JW, Cutler SJ. Occurrence and prognosis of extranodal lymphomas. Cancer 1972; 29: 252-60. 2. Otter R, Bieger R, Kluin PM, et al. Primary gastrointestinal non-Hodgkin's lymphoma in a population-based registry. BrJ Cancer 1989; 60: 745-50. 3. Isaacson PG, Norton AJ. Extranodal lymphomas. Edinburgh: Churchill Livingstone, 1994. 4. Lukes RJ, Collins RD. New approaches to the classification of the lymphomata. BrJ Cancer 1975; 31(suppl): 1-28. 5. Lennert K, Feller A. The Kiel classification. Histopathology of non-Hodgkin's lymphomas, 2nd edn (based on the Updated Kiel Classification). Berlin: Springer-Verlag, 1990: 13-52. 6. Isaacson P, Wright DH. Malignant lymphoma of mucosaassociated lymphoid tissue. A distinctive type of B-cell lymphoma. Cancer 1983; 52:1410-16. 7. Harris NL, Jaffe ES, Stein H, et al. A revised European-American Classification of lymphoid neoplasms: a proposal from the International Lymphoma Study Group. Blood 1994; 84:1361-92. 8. Isaacson P, Wright DH. Extranodal malignant lymphoma arising from mucosa-associated lymphoid tissue. Cancer 1984; 53: 2515-24. 9. Gowans JL, Knight EJ. The route of re-circulation of lymphocytes in the rat. Proc R Soc Lond Biol 1964; 159: 257-82. 10. Wotherspoon AC, Ortiz-Hidalgo C, Falzon MR, Isaacson
88 Extranodal lymphomas PG. Helicobacter pylori-associated gastritis and primary B-cell gastric lymphoma. Lancet 1991; 338:1175-6. 11. Hyjek E, Smith WJ, Isaacson PG. Primary B-cell lymphoma of salivary glands and its relationship to myoepithelial sialadenitis. Human Pathol 1988; 19: 766-76. 12. Hyjek E, Isaacson PG. Primary B cell lymphoma of the thyroid and its relationship to Hashimoto's thyroiditis. Human Pathol 1988; 19:1315-26. 13. Musshoff K. Klinische Stadieneinteilung der NichtHodgkin-Lymphome. Strahlentherapie 1977; 153: 218-21. 14. Wotherspoon AC, Finn TM, Isaacson PG. Trisomy 3 in lowgrade B-cell lymphomas of mucosa-associated lymphoid tissue. Blood; 1995; 85: 2000-4. 15. Willis TGJadayel DM, Du M-Q, etal. BcllO is involved in t(1;14)(p22;p32) of MALT B cell lymphoma and mutated in multiple tumour types. Cell 1999; 96: 35-45. 15a. Dielamin J, Baens M, Wlodarska I, et al. The Apoptous inhibitor gene API 2 and a novel 18q gene, MLT, are recurrently rearranged in thet(11;18)(q21;q21) associated with mucosa-associated lymphoid tissue lymphomas. Blood 1999; 93: 3601-9. 16. Cogliatti SB, Schmid U, Schumacher U, etal. Primary B-cell gastric lymphoma: a clinicopathological study of 145 patients. GastroenterologyWVI; 101:1159-70. 17. Chan JKC, Ng CS, Isaacson PG. Relationship between high-grade lymphoma and low-grade B-cell mucosaassociated lymphoid tissue lymphoma (MALToma) of the stomach. Am7 Potf/o/1990; 136:1153-64. 18. Joensuu H, Soderstrom KO, Klemi PJ, Eevola E. Nuclear DNA content and its prognostic value in lymphoma of the stomach. Cancer 1987; 60: 3042-8. 19. Genta RM, Robason GO, Graham DY. Simultaneous visualization of Helicobacter pylori and gastric morphology: a new stain. Human Pathol 1994; 25: 221-6. 20. Parsonnet J, Hansen S, Rodriquez L, et al. Helicobacter pylori infection and gastric lymphoma. N EnglJ Med 1994; 330:1267-71. 21. Zucca E, Bertoni F, Roggero E, etal. Molecular analysis of the progression from helicobacter pylori-associated chronic gastritis to mucosa-associated lymphoid-tissue lymphoma of the stomach. N EnglJ Med 1998; 338: 804-10. 22. Neubauer A, Thiede C, Morgner A, et al. Cure of helicobacter pylori infection and duration of remission of low-grade gastric mucosa-associated lymphoid tissue lymphoma. J Natl Cancer Inst 1997; 89:1350-5. 23. Pinotti G, Zucca E, Roggero E, et al. Clinical features, treatment and ourtcome in a series of 93 patients with low-grade gastric MALT lymphoma. Leak Lymphoma 1997; 26: 527-37. 24. Hussell T, Isaacson PG, Crabtree J, Spencer J. The response of cells from low-grade B-cell gastric lymphomas of mucosa-associated lymphoid tissue to Helicobacter pylori. Lancet 1993; 342: 571-4.
25. Wotherspoon AC, Doglioni C, Diss TC, et al. Regression of primary low-grade B-cell gastric lymphoma of mucosaassociated lymphoid tissue type after eradication of Helicobacter pylori. Lancet 1993; 342: 575-7. 26. Roggero E, Zucca E, Pinotti G. Eradication of Helicobacter pylori infection in primary low-grade gastric lymphoma of mucosa-associated lymphoid tissue. Ann Intern Med 1995; 122: 767-9. 27. Bayerdorffer E, Neubauer A, Rudolph B. Regression of primary gastric lymphoma of mucosa-associated lymphoid tissue type after cure of Helicobacter pylori infection. Lancet 1995; 345:1591-4. 27a.Zucca E, Roggero E, Delchier J, etal. Interim evaluation of gastric MALT lymphoma response to antibiotics in the ongoing LYO 3 randomized cooperative trial of observation vs chlorambucil after anti-helicobacter therapy. ProcAm SocClin Oncol200Q; 19: 5a (abstract 12). 28. Domizio P, Owen RA, Shepherd NA, etal. Primary lymphoma of the small intestine: a clinicopathological study of 119 cases. AmJSurg Pathol 1993; 17: 429-32. 29. Galian A, Lecestre MJ, Scotto J, et al. Pathological study of alpha-chain disease, with special emphasis on evolution. Cancer 1977; 39: 2081 -101. 30. Price SK. Immunoproliferative small intestinal disease: a study of 13 cases with alpha heavy-chain disease. Histopathology 1990; 17: 7-17. 31. Ben-Ayed F, Halphen M, Najjar T, et al. Treatment of alpha chain disease. Results of a prospective study in 21 Tunisian patients by the Tunisian-French Intestinal Lymphoma Study Group. Cancer 1989; 63:1251-6. 32. O'Briain DS, Kennedy MJ, Daly PA, et al. Multiple lymphomatous polyposis of the gastrointestinal tract. A dinicopathologically distinctive form of non-Hodgkin's lymphoma of B-cell centrocytic type. AmJSurg Pathol 1989;13:691-9. 33. Ruskone-Fourmestraux A, Aergester P, Delmer A, et al. Primary digestive tract lymphoma: a prospective multicentric study of 91 patients. Groupe d'Etude des Lymphomes Digestifs. Gastroenterology 1993; 105: 1662-71. 34. Anaissie E, Geha S, Allam C, et al. Burkitt's lymphoma in the Middle East: a study of 34 cases. Cancer 1985; 56: 2539-43. 35. Isaacson PG, O'Connor NTJ, Spencer J. Malignant histiocytosis of the intestine: a T cell lymphoma. Lancet 1985; 2: 688-91. 36. Spencer J, Cerf-Bensussan N, Jarry A, et al. Enteropathy associated T cell lymphoma (malignant histiocytosis of the intestine) is recognized by a monoclonal antibody (HML-1) that defines a membrane molecular on human mucosal lymphocytes. Am} Pathol 1988; 132: 1-5. 37. Ashton-Key M, Diss TC, Pan LX, et al. Molecular analysis of T-cell clonality in ulcerative jejunitisand enteropathy associated T-cell lymphoma. Am J Pathol 1997; 151: 493-8.
References 89 38. Sciubba JJ, Auclair PL, Ellis GL Malignant lymphomas. In: Ellis CL, Auclair PL, Gnepp DR, eds Surgical pathology of the salivary glands. Philadelphia: WB Saunders. 1991: 528^3. 39. Nizze H, Cogliatti SB, Von Schilling C, et al. Monocytoid B-cell lymphoma: morphological variants and relationship to low-grade B-cell lymphoma of the mucosa-associated lymphoid tissue. Histopathology 1991; 18:403-14. 40. Diss TC, Wotherspoon AC, Speight P, Pan LX, Isaacson PG. B-cell monoclonality, Epstein Barr virus, and t(14;18) in myoepithelial sialadenitisand low-grade B-cell MALT lymphoma of the parotid gland. AmJ Surg Pathol 1995; 19: 531-6. 41. Gould S, Isaacson PG. Bronchus associated lymphoid tissue (BALT) in human fetal and infant lung.7 Pathol 1993; 169: 229-34. 42. Li G, Hansmann ML, ZwingersT, Lennert K. Primary lymphomas of the lung: morphological, immunohistochemical and clinical features. Histopathology 1990; 16: 519-31. 43. Liebow AA, Carrington RB, Friedman PJ. Lymphomatoid granulomatosis. Human Pathol 1972; 3: 457-8. 44. Guinee D, Jaffe E, Kingma N. Pulmonary lymphoid granulomatosis. AmJ Surg Pathol 1994; 18: 754-64. 45. Koss MN. Pulmonary lymphoid disorders. Semin Diag Pathol 1995; 12:158-171. 46. luchi K, Ichimiya A, Akashi A, et al. Non-Hodgkin's lymphoma of the pleural cavity developing from longstanding pyothorax. Cancer WSJ; 60:1771-5. 47. Martin A, Capron F, Liguory-Brunaud MD. Epstein-Barr virus-associated primary malignant lymphomas of the pleural cavity occurring in longstanding pleural chronic inflammation. Human Pathol 1994; 25:1314. 48. Wotherspoon AC, Hardman-Lea S, Isaacson PG. Mucosaassociated lymphoid tissue (MALT) in the human conjunctiva. J Pathol 1994; 174: 33-7. 49. Picker LJ, Michie SA, Rott LS, Butcher EC. A unique phenotype of skin-associated lymphocytes in humans. Preferential expression of the HECA-452 epitope by benign and malignant T cells at cutaneous sites. Am J Pathol 1990; 136:1053-68. 50. Stein H, Mason DY, Gerdes J, et al. The expression of the Hodgkin's disease-associated antigen Ki-1 in reactive and neoplastic lymphoid tissue: evidence that Reed-Stern berg cells and histiocytic malignancies are derived from activated lymphoid cells. Blood 1985; 66: 848-58. 51. Kaudewitz P, Stein H, Dallenbach F, et al. Primary and secondary cutaenous Ki-1+ (CD30+) anaplastic large cell lymphomas. Morphologic, immunohistologic, and clinical characteristics. AmJ Pathol 1989; 135: 359-67. 52. Bitter MA, Franklin WA, Larson RA, et al. Morphology in Ki-1 (CD30)-positive non-Hodgkin's lymphoma is correlated with clinical features and the presence of a unique chromosomal abnormality, t(2;5)(p23;q35). AmJ Surg Pathol 1990; 14: 305-16.
53. Willemze R, Meijer CJLM, van Vloten WA, Scheffer E. The clinical and histological spectrum of lymphomatoid papulosis. BrJ Dermatol 1982; 107:131-44. 54. Parks JD, Synovec MS, Masih AS, et al. Immunophenotypic and genotypic characterization of lymphomatoid papulosis. J Am Acad Dermatol 1992; 26: 968-75. 55. Whittaker S, Smith N, Russell Jones R, Luzzatto L. Analysis of p, y, 8 T-cell receptor genes in lymphomatoid papulosis: cellular basis of two distinct histologic subsets. J Invest Dermatol 1991; 96: 786-91. 56. Santucci M, Pimpinelli N, Arganini L. Primary cutaneous B-cell lymphoma: a unique type of low-grade lymphoma. Clinicopathologic and immunologic study of 83 cases. Cancer 1991; 67: 2311 -26. 57. Handlers JP, Howell RE, Abrams AM, Melrose RJ. Extranodal oral lymphoma. Part 1. A morphologic and immunoperoxidase study of 34 cases. Oral Surg Oral Med Oral Pathol 1986; 61: 362-7. 58. Barton JH, Osborne BM, Butler JJ, et al. Non-Hodgkin's lymphoma of the tonsil. A clinico-pathological study of 55 cases. Cancer 1984; 53: 86-95. 59. Falbaum Chr, Hansmann ML, Lannert K. Malignant lymphomas of the nasal cavity and paranasal sinuses. Virchows Arch A Pathol Anat 1989; 414: 399-405. 60. Eisenbud L, Sciubba J, Mir R, Sachs SA. Oral presentations in non-Hodgkin's lymphoma: a review of thirty-one cases. Part 1. Data analysis. Oral Surg Oral Med Oral Pof/70/1983;56:151-6. 61. Fukuda Y, Ishida T, Fujimoto M, et al. Malignant lymphoma of the oral cavity: Clinicopathologic analysis of 20 cases. 7 Oral Pathol 1987; 16: 8-12. 62. Ng CS, Chan JKC, Lo STH. Expression of natural killer cell markers in non-Hodgkin's lymphomas. Human Pathol 1987; 18:1257-62. 63. Arber DA, Weiss LM, Albujar PF, Chen Y-Y, Jaffe ES. Nasal lymphomas in Peru. High incidence of T-cell immunophenotype and Epstein-Barr virus infection. Am J Surg Pathol 1993; 17: 392-9. 64. Van Gorp J, De Bruin PC, Sie-Go DMDS. Nasal T-cell lymphoma: a clinicopathological and immunophenotypic analysis of 13 cases. Histopathology 1995; 27: 139-48. 65. David DL, Hoel D, FoxJ, Lopez A. International trends in cancer mortality in France, West Germany, Italy, Japan, England and Wales, and in the USA. Lancet 1990; 336: 474-81. 66. Hochberg FH, Miller DC. Primary central nervous system lymphoma. J Neurosurg 1988; 68: 835-53. 67. Isaacson PG, Norton AJ, Addis BJ. The human thymus contains a novel population of B lymphocytes. Lancet 1987; ii: 1488-91. 68. Isaacson PG, Chan JKC, Tang C, Addis BJ. Low-grade B-cell lymphoma of mucosa-associated lymphoid tissue arising in the thymus. A thymic lymphoma mimicking myoepithelial sialadenitis. Am J Surg Pathol 1990; 14: 342-51.
90 Extranodal lymphomas 69. Takagi N, Nakamura S, Yamamoto K, et al. Malignant lymphoma of mucosa-associated lymphoid tissue arising in the thymus of a patient with Sjogren's syndrome. A morphologic, phenotypic and genotypic study. Cancer 1992; 69:1347-55. 70. Lichtenstein AK, Levine A, Taylor CR, et al. Primary mediastinal lymphoma in adults. Am J Med 1980; 68: 509-14. 71. Addis BJ, Isaacson PG. Large cell lymphoma of the mediastinum: a B-cell tumour of probable thymic origin. Histopathology 1986; 10: 379-90. 72. Moller P, Moldenhauer G, Momburg F, et al. Mediastinal lymphoma of clear cell type is a tumor corresponding to terminal steps of B cell differentiation. Blood 1987; 69: 1087-95. 73. Schmid C, Kirkham N, DissT, Isaacson PG. Splenic marginal zone cell lymphoma. AmJ Surg Pathol 1992; 16: 455-66. 74. Isaacson PG, Matutes E, Burke M, Catovsky D. The histopathology of splenic lymphoma with villous lymphocytes. Blood 1994; 11: 3828-34. 75. Melo JV, Hedge U, Parreira A, et al. Splenic B cell lymphoma with circulating villous lymphocytes: differential diagnosis of B cell leukaemias with large spleens. J Clin Pathol 1987; 40: 642. 76. Mulligan SP, Matutes E, Dearden C, Catovsky D. Splenic lymphoma with villous lymphocytes: natural history and response to therapy in 50 cases. BrJ Haematol 1991; 78: 206-9. 77. Gaulard P, Bourquelot P, Kanavaros P, et al. Expression of the alpha-beta and gamma-delta T-cell receptors in 57 cases of peripheral T-cell lymphomas. Identification of a subset of y/8 T-cell lymphomas. AmJ Pathol 1990; 137: 617-28. 78. Paladugu RR, Bearman RM, Rappaport H. Malignant lymphoma with primary manifestation in the gonad. A clinicopathologic study of 38 patients. Cancer 1980; 45: 561-71. 79. Pawade J, Banerjee SS, Harris M, Isaacson PG, Wright D. Lymphomas of mucosa-associated lymphoid tissue arising in the urinary bladder. Histopathology 1993; 23:147-51. 80. Ostrowski ML, Unni KK, Banks PM, etal. Malignant lymphoma of bone. Cancer 1986; 58: 2646-55.
81. Bobrow LG, Richards MA, Happerfield LC, et al. Breast lymphomas: a clinicopathologic review. Human Pathol 1993; 24: 274-8. 82. Osborne BM, Butler JJ, Guarda LA. Primary lymphoma of the liver. Ten cases and a review of the literature. Cancer 1985;56:2902-10. 83. Isaacson PG, Banks PM, Best PV. Primary low-grade hepatic B-cell lymphoma of mucosa-associated lymphoid tissue (MALT)-type. AmJ Surg Pathol 1995; 19: 571-5. 84. Burkitt D. A sarcoma involving the jaws in African children. BrJ Surg, 1958; 46: 218-33. 85. Burkitt DP, Wright DH, eds. Burkitt's lymphoma. Edinburgh: E & S Livingstone, 1970. 86. Dorfman RF. Childhood lymphosarcoma in St Louis, Missouri, clinically and histologically resembling Burkitt's tumor. Cancer 1965; 18: 418-30. 87. Domizio P, Hall PA, Cotter F, etal. (1989) Angiotropic large cell lymphoma (ALCL): morphological, immunohistochemical and genotypic studies with analysis of previous reports. Hematol Oncol 1989; 7: 195-206. 88. Sheibani K, Battifora H, Winberg CD, etal. Further evidence that 'malignant angioendotheliomatosis' is an angiotropic large-cell lymphoma. N EnglJ Med 1986; 314: 943-8. 89. Monfardini S, Tirelli U, Vaccher E, et al. Hodgkin's disease in 63 intravenous drug users infected with human immunodeficiency virus. Gruppo Italiano Cooperative AIDS&Tumori (GICATMw) Onco/1991; 2(suppl 2): 201-5. 90. Frizzera G, Hanto DW, Gajl-Peczalska KJ, etal. Polymorphic diffuse B-cell hyperplasias and lymphomas in renal transplant recipients. Cancer Res 1981; 41: 4262-79. 91. ZieglerJL, DrewWL, Miner RC, etal. Outbreak of Burkitt's-1 ike lymphoma in homosexual men. Lancet 1982; H: 631-4. 92. Roithmann S, Toledano M, Tourami JM, et al. HIVassociated non-Hodgkin's lymphomas: clinical characteristics and outcome. The experience of the French Registry of HIV-associated tumours. Ann Oncol 1991; 2: 289-95.
9 Cytogenetics WG SANGER, BJ DAVE AND MR BISHOP
Introduction Methods Hodgkin's disease Non-Hodgkin's lymphoma
91 91 92 93
INTRODUCTION Investigation into the role of cytogenetic abnormalities in tumorigenesis is one of the most rapidly progressing areas in cancer research. As early as 1914, Boveri, hypothesized that chromosome abnormalities were responsible for the initiation of malignancies.1 The first direct evidence of this hypothesis was forwarded by Nowell and Hungerford in I9602 when they described the Philadelphia chromosome, which appeared to be a deletion of a portion of the long arm of one G-group chromosome in certain patients with chronic myelogenous leukemia (CML). The Philadelphia chromosome was later shown to involve chromosome 22 by chromosome banding.3 With further banding techniques development, the Philadelphia chromosome was shown to be a reciprocal translocation between chromosomes 9 and 22, and found in 90-95 per cent of CML cases.4 It is now known that all true CML cases involve a classical t(9;22), a variant or cytogenetically cryptic chromosome rearrangement. Overall, developments in cancer cytogenetics between 1960 and 1980 were comparatively slow to the present pace, primarily because of the use of cytogenetic procedures, which were traditionally used for the study of constitutional abnormalities, thereby leading to the inadvertent study of non-malignant cells in some patients with malignancies. Additionally, the procedures utilized before 1980 resulted in chromosomes that were very contracted and poorly banded, resulting in subtle rearrangements being undetectable. Cytogenetic procedures and banding quality have continually improved from that period until the present, and have led to the discovery of specific chromosome abnormalities associated
Clinical and prognostic correlations Future studies References
99 99 100
with various subtypes of leukemia and lymphoma, as well as numerous soft tissue and solid tumors. The association of specific chromosome rearrangements in the leukemias have been the most exhaustively studied and specific rearrangements have been associated with the various subtypes of acute myelogenous leukemia (AML), CML, acute lymphocytic leukemia (ALL) and chronic lymphocytic leukemia (CLL). Compared to the leukemias, which are frequently characterized by only one or a few chromosome changes, the lymphomas are more difficult to study because they are commonly associated with numerous chromosome rearrangements and numerical abnormalities, thus making it more of a challenge to ascertain which events are critical in tumorigenesis and which ones are secondary. Nevertheless, there have been numerous series of lymphoma studies carried out.5-12 A summary of the known primary chromosome abnormalities is represented in Table 9.1 and a listing of the secondary acquired abnormalities with their clinical correlations are summarized in Table 9.2. A review of lymphoma cytogenetic studies and the implications of these findings to diagnosis and prognosis are the objectives of this chapter.
METHODS It is imperative that tumor cells (and not non-malignant cells) are analysed cytogenetically when attempting to characterize chromosome changes in malignant disease. To accomplish this, it is necessary that the cytogenetics laboratory receive tumor tissue, and that appropriate culture methods are utilized to culture tumor cells and not normal stromal cells in mitosis for cytogenetic analysis.
92 Cytogenetics Table 9.1 'Primary' acquired chromosome abnormalities in lymphoma
Multiple numerical abnormalities (often pseudotetraploid state)
?
t(8;14)(q24;q32) IgH with c-myc or variants t(2;8)(p21;q24) and t(8;22)(q24;q11) (Ig with c-myc) t(14;18)(q32;q21) IgH with bcl-2 t(3;14)(q27;q32) bcl-6 with IgH t(11;14)(q13;q32) bc\A with IgH t(9;14)(p13;q32) PAX-5 with IgH
Reed-Stern berg
Hodgkin's disease
Bcell
Small non-cleaved (Burkitt's)
Bcell Bcell Bcell Bcell
Follicular (70-90%)/high-grade (20-40%) Diffuse with large cell component Centrocytic (intermediate) lymphoma Small lymphocytotic with plasmacytoid differentiation Low-grade mucosa associated lymphoid tissue (MALT) type lymphoma Anaplastic large cell lymphoma Tcell lymphoma
t(11;18)(q21;q21)
7
Bcell
t(2;5)(p23;q35) inv(14)(q11?)ort(14;?)(q11;?)
ALK gene with NPM T cell receptor deregulation
Tcell Tcell
NPM = nucleophosmin.
Additionally, cell synchronization methodologies are sometimes used to accomplish a higher resolution study. In the case of the lymphomas, lymph node tissue is typically processed by our laboratories within 1 hour after the biopsy procedure, followed by direct harvest or following 24-hour and/or 48-hour culture without mitogens, and with or without synchronization. Following Colcemid exposure, hypotonic and fixation steps are accomplished prior to slide preparation. The slides are generally aged for a minimum of 24-hours and G-banded with Wright's stain, with or without trypsin pretreatment or with Giemsa following enzyme pretreatment. Cytogenetic analysis is accomplished by microscopic analysis of at least 20 cells, when available, followed by karyotyping at least two cells from each abnormal clone according to the International System for Human Cytogenetic Nomenclature.13 An abnormal clone is defined as either two or more cells with the same structural abnormality or the same extra
chromosome, or three or more cells with the same missing chromosome. If a metaphase cannot be analysed confidently, the cell is not included in the analysis. If certain chromosome rearrangements cannot be confirmed utilizing G-banding methods, other banding methods are utilized to clarify the abnormality. Fluorescence in situ hybridization (FISH) procedures, utilizing some of the commercially available ex-satellite, cosmid and paint probes are useful in clarifying and confirming certain complex chromosome rearrangements. These latter methodologies are being used with increasing frequency by most cytogenetics laboratories performing lymphoma studies.
HODGKIN'S DISEASE Compared to the non-Hodgkin's lymphomas (NHLs), relatively few Cytogenetic studies have been performed
Table 9.2 Probable 'secondary' acquired chromosome abnormalities in lymphoma and possible clinical significance
1p22 1q21-22
2p +3 +5, +6, +18 rea(5), or rea(14q11) rea(6q13-16) +7witht(14;18) 10q22-24
+12 13q21-24 +17ori(17q) +der(18)t(14;18) Many abnormalities and no normal cells Normal cells present with abnormal clone
Large cell diffuse lymphoma Poor survival in intermediate/high-grade lymphoma Skin infiltration High-grade lymphoma Shorter survival Shorter survival B cell disease and short survival Diffuse histology Shorter survival Large cell diffuse lymphoma Bulky disease Diffuse histology and shorter survival Disease progression Rapid progression and short survival Better prognosis than those cases without normal cells
Non-Hodgkin's lymphoma 93
on Hodgkin's disease. A primary restraint in the successful study of Hodgkin's disease is that there is typically a low number of mitotic cells, many of which are nonmalignant, cytogenetically normal cells. For example, aneuploid clones were found in only 6 of 20 cases,1413 of 29 cases15 and 13 of 70 cases.16 A probable explanation of this is that cytogenetically normal reactive cells are undergoing mitosis in this disease rather than neoplastic cells. To date, there has been no single structural chromosomal abnormality correlated with any of the histopathological Hodgkin's disease subtypes.14"22 Simultaneous application of FISH and immunophenotyping are being used to define chromosome abnormalities in Hodgkin-Reed Sternberg cells.23-25 Generally speaking, clonal abnormalities in Hodgkin's disease consistently involved numerical abnormalities, quite often reaching the near tetraploid range. The most common numerical abnormalities include trisomies for chromosomes 1, 3, 7, 8 and 21.14-16 A broad spectrum of structural abnormalities which are also common in NHL, were reported among these studies and frequently involved Ip36, 6ql5, 6q21, 7q22, 7q32, 8q24, Ilq23, 12q24, 13pll, 14pll, 14q32, and 15pll rearrangements.14-22 Even though there are some cytogenetic similarities between Hodgkin's disease and NHL, the number of chromosomes observed in Hodgkin's disease cases is frequently in the triploid to tetraploid range, which is in contrast to most other NHLs. The frequency of polyploidy in Hodgkin's disease would be consistent with the theory that these are derived from the Reed-Sternberg cell, which is quite often a polyploid cell.26
NON-HODGKIN'S LYMPHOMA In contrast to Hodgkin's disease, the non-Hodgkin's lymphomas demonstrate abnormal clones in 75-100 per
cent of cases studied.5'12 Compared to the leukemias, the frequency of cytogenetic abnormalities is higher in NHL and the chromosome rearrangements are much more numerous and complex. NHLs have shown one of the widest varieties of recurrent (reported in 3 cases or more) chromosomal changes and these are not frequently seen as sole chromosomal abnormality.27 In attempting to correlate cytogenetic abnormalities with histological classification, most groups utilize the International Working Formulation (IWF)28 and, using this formulation, the correlations between cytogenetic abnormalities and histology have been somewhat successful. In NHL, it has been shown that many oncogenes are activated via translocation of chromosomal material causing a juxtaposition to an enhancer of a lymphoid immunoglobulin gene.29 Among all subtypes of lymphoma, the single most common chromosome abnormality, which is found in at least two-thirds of all NHL cases, is that of a rearrangement involving band 32 on the long arm of chromosome 14 (14q32). The 14q32 region is the immunoglobulin heavy chain (IgH) gene locus and, when this is juxtaposed with other oncogenes, there is an etiological correlation with B cell NHL. A summary of some of the primary chromosome rearrangements in NHL follow.
Translocation 8/14 - t(8;14)(q24;q32) (Fig. 9.1) The t(8;14)(q24;q32) was first described by Manolov and Manolova in 197230 as being associated with Burkitt's lymphoma. This translocation juxtaposes the IgH locus at 14q32 with the c-myc oncogene located on 8q24. At least 60 per cent of patients with Burkitt's lymphoma have the t(8;14) and another 20 per cent have a variant translocation demonstrating either a t(8;22)(q24;qll) or a t(2;8)(p21;q24). All three of these
Figure 9.1
Translocation 8/14 -
t(8;14)(q24;q32). Ideogram showing critical chromosome breakpoints and associated genes, which, when rearranged, have been implicated in Burkitt's lymphoma and certain other highgrade B cell lymphomas and leukemias. c-myc - oncogene, IgH = immunoglobulin heavy-chain gene.
94 Cytogenetics
chromosome rearrangements are believed to result in deregulation of the cellular oncogene, c-myc, thereby resulting in Burkitt's lymphoma. Generally, patients with one of the variant translocations have poorer survival rates than those with the t(8;14).31 In addition to Burkitt's lymphoma, the t(8;14) has also been observed in patients with acute lymphocytic leukemia (L3). Additionally, the t(8;14) or other variant translocations have been observed in other lymphomas, particularly in high-grade lymphomas.5,32 With only rare exception, all cases with a t(8;14), a t(8;22), or a t(2;8) have a B cell phenotype.
Translocation 14/18 - t(14;18)(q32;q21) (
^:^^
The t(14;18) reciprocal translocation juxtaposes the IgH locus with the bd-2 oncogene during early B cell development. The 18q21 region includes major (mbr) and minor (mcr) breakpoint regions, which recombine with the IgH locus at 14q32.33 The mbr and mcr are involved in approximately 66 and 16 per cent, respectively, of those cases with a t(14;18)(q32;q21) by cytogenetics,34 with the remaining cases apparently involving other nearby breakpoints. The t(14;18) occurs in 70-90 per cent of follicular lymphomas and in approximately 30 per cent of diffuse large-cell NHL cases.12 The resulting bd-2 gene product of this translocation apparently inhibits programmed cell death or apoptosis, and extends cell survival rather than causing increased proliferation. The t(14;18) is believed to represent a premalignant change, which, after a second genetic change, results in the development of follicular lymphoma. The course of the disease is generally slow, lasting for several years; however, it is incurable with conventional therapy and is ultimately fatal. Although the observation of the t(14;18) in high-grade lymphomas represents transformation in some cases, the presence of additional or secondary chromosome abnormalities are more likely to be
responsible for the aggressive nature of these tumors.8'11 With only extremely rare exceptions, the presence of a t( 14; 18) is associated with a B cell phenotype.
Translocation 3/14 - t(3;14)(q27;q32)
!!!iL:!^L___ ______ The t(3;14) results in the juxtaposition of the site which codes for a zinc-finger protein (bd-6 or LAZ-3) located at 3q27 with the IgH region of 14q32.35 A variant of this translocation involves the bd-6 region of 3q27 with the immunoglobulin light-chain region of chromosome 22qll.36,37 Offit et a/.38 described the rearrangement of bd-6 in 23/102 cases of diffuse lymphomas with large cell components. Horsman et a/.39 also reported that in 5 of their 22 t(3;14) cases, there was a second lymphoma specific translocation - three cases of t( 14; 18) and two cases of t(8;14). The presence of the t(3;14) is commonly associated with extranodal involvement and is associated with B cell markers. Patients with the t(3;14) have a higher rate of complete remissions with conventional therapy and have a longer disease-free survival compared to other cases of diffuse lymphomas with a large cell component.38,40
Translocation 11/14-t(11;14)(q13;q32)
fig. 9.4
!— ___
_______
The t(ll;14) results in the juxtaposition of the bd-l proto-oncogene with the IgH region of 14q32 and is commonly found in mantle cell (intermediate lymphocytic) lymphomas.41'42 This rearrangement in this intermediate grade of lymphoma is thought to deregulate the nearby CCND1(BCL1/PRAD1) proto-oncogene, which is a member of the cyclin family.43'44 Median survival ranges from 2 to 5 years following diagnosis. Responses to chemotherapy are seen in half of the patients and long-term survival is uncommon. Patients with the t(ll;14) rearrangement frequently develop a
Figure 9.2 Translocation 14/18 t(!4;18)(q32;q21). Ideogram showing critical chromosome breakpoints and associated genes, which, when rearranged, have been implicated in follicular lymphoma and in some highgrade NHLs. IgH = immunoglobulin heavy-chain gene, bd-2 - bd-2 proto-oncogene.
Non-Hodgkin's lymphoma 95 Figure 9.3 Translation 3/14 t(3;14)(q27;q32). Ideogram showing critical chromosome breakpoints and associated genes, which, when rearranged, have been implicated in diffuse large cell lymphoma. IgH = immunoglobulin heavy-chain gene, bcl-6 = bcl-6 oncogene.
Figure 9.4 Translocation 11/14-t(l1;l4)(q13;q32). Ideogram showing critical chromosome breakpoints and associated genes, which, when rearranged, have been implicated in centrocytic or mantle cell lymphoma. IgH = immunoglobulin heavy-chain gene, bcl-3 - bcl-3 oncogene.
96 Cytogenetics
leukemic phase. Nearly all lymphomas with a t(ll;14) are B cell diseases.
Translocation 9/14 - t(9;14)(p13;q32)
(FigJ).^ The t(9;14) is associated with small lymphocytic lymphoma with plasmacytoid differentiation, according to the Working Formulation (lympho plasmacytoid lymphoma (LPL) according to Revised European-American Classification of Lymphoid Neoplasms [REAL]), a subtype of B cell NHL.45,46 The PAX-5 gene which encodes a B cell specific transcription factor involved in the control of B cell proliferation and differentiation is the target of the t(9;14) in LPL whereby its expression may be deregulated by juxtaposition to IgH regulatory elements, thus contributing to lymphomagenesis.47 The disease follows an indolent clinical course.48
Translocation 2/5 - t(2;5)(p23;q35) (Fig. 9.7) The t(2;5) results in the fusion of the amino terminus of the nucleophosmin gene (NPM) on chromosome 5q35 with the anaplastic lymphoma kinase (ALK) gene on chromosome 2p2356 and leads to the constitutive expression of a truncated form of ALK protein p80. This is frequently associated with Ki-1 (CD30)-positive anaplastic large cell lymphoma.56'57 Of all CD30+ anaplastic large cell lymphomas, approximately 36 per cent have a detectable t(2;5)(p23;q35) with a higher incidence in childhood cases and cases with a non-B cell phenotype.58 This is a group of morphologically and immunotypically heterogeneous high-grade large cell lymphomas.58-63 Ki-1-positive lymphomas associated with the t(2;5) rearrangement are almost exclusively T cell diseases. Patients with Ki-l+ anaplastic large cell lymphoma with the t(2;5), have a favorable prognosis and a higher 5-year survival rate, compared to anaplastic large cell cases without the t(2;5).64-65
Translocation 11/18 - t(11;18)(q21;q21) 14q11 clpnal rearrangements-t(14)(q11)
fig 9.6—__________lesions (Fig. 9.8) The t(ll;18) is a frequent and specific chromosomal translocation in low-grade but not high-grade malignant NHLs of the mucosa associated lymphoid tissue (MALT) type.49"52 The molecular equivalent of the t( 11; 18) (q21 ;q21) is yet to be determined. No rearrangements of bd-2 gene, localized in 18q21, have been demonstrated in MALT lymphomas thus far53'55 or in cases cytogenetically positive for the t( 11;18). The primary extra-nodal malignant NHL arising from MALT represents a subtype of B cell lymphoid malignancies and is often associated with a favorable prognosis.
The t(ll;14)(pl3;qll), the t(8;14)(q24;qll), the inv(14)(qllql2) and other 14qll rearrangements result in a rearrangement of the T cell alpha- and delta-chain receptor genes (TCRA and TCRD) at the 14qll region, which deregulates the T cell receptor gene, resulting in a T cell neoplasm. The T cell receptor genes are highly and specifically expressed only in T cells. There are also other chromosomal regions on chromosomes 1, 7, 8, 10, 11 and 14, which, when translocated to the T cell receptor region at 14ql 1, result in a T cell lymphoma process.5,9,10,12
Figure 9.5 Translocation 9/14 t(9;14)(p!3;q32). Ideogram showing critical chromosome breakpoints and associated genes, which, when rearranged, have been implicated in small lymphocytic lymphoma with plasmacytoid differentiation [lympho plasmacytoid lymphoma (LPL)]. PAX5 = paired homeobox-5 gene, IgH = Immunoglobulin heavy chain gene.
Non-Hodgkin's lymphoma 97 Figure 9.6 Translocation 77/78t(11;18)(q21;q21). Ideogram showing critical chromosome breakpoints which, when rearranged, have been implicated in low-grade B-cell lymphomas of the (MALT mucosa associated lymphoid tissue) type.
Additional chromosome abnormalities in NHLs Compared to the leukemias, where there are one or a very few chromosome lesions in the neoplastic clone, lymphomas commonly involve numerous chromosome
rearrangements and other abnormalities. This increases the challenge for cytogeneticists to determine which chromosome rearrangements are primary, and which are secondary or progressive changes. The chromosome rearrangements summarized in Table 9.1 are thought to be primary events in lymphomagenesis; however, there
Figure 9.7 Translocation 2/5 t(2;5)(p23;q35). Ideogram showing critical chromosome breakpoints and associated genes, which, when rearranged, have been implicated in high-grade large cell lymphoma. ALK = anaplastic lymphoma kinase gene, NPM = amino-terminus of nucleophosmin gene.
98 Cytogenetics
Figure 9.8
14q1l clonal
rearrangements - rea(T4)(qll). Ideogram showing critical chromosome breakpoints and associated genes, which, when rearranged, have been implicated in Tcell lymphoma. TCRA - T cell alpha-chain receptor gene, TCRD - Tcell delta-chain receptor gene.
are undoubtedly others involved as primary events in this disease process. It is thought that the reciprocal (balanced) translocations or inversions represent primary chromosome abnormalities of pathogenic importance. In contrast, the unbalanced translocations resulting into trisomies or monosomies of chromosomal regions may constitute secondary or progression-related
abnormalities in NHL. There have been numerous relatively large series of cytogenetic studies on lymphomas.5'12 Some of the common lymphoma breakpoints and oncogene locations are summarized in Fig. 9.9. Among these series, some of the more common structural chromosome abnormalities, in addition to those previously mentioned, include rearrangements at lq21-23, lp32-36, 6pll, 6q21-25, 9p21-23, 10q22-24, llq!3-23, 19pl3, an additional chromosome 17 or isochromosome for the long-arm of chromosome 17, as well as trisomy 3, 7,12,20 per cent and 21. Some investigators have emphasized the correlation of trisomies of certain chromosomes or deletions of chromosomal regions with specific histologic subsets of NHLs,66"70 however, these have been frequently observed as secondary changes in addition to other lymphoma associated chromosomal translocation. Deletion 7q32 is associated with a subset of small lymphocytic lymphoma with plasmocytoid features.71 Deletion of 6q21-23 is associated with a subset of small lymphocytic NHL,72 and deletion llq is associated with small lymphocytic lymphoma.73 Isochromosome 7q and trisomy 8 have been reported in hepatosplenic T gamma/delta peripheral T cell
Figure 9.9 Summary of common lymphoma breakpoints and oncogene locations. The numbers on the left represent approximate number of cases with rearrangements/500 total cases studied at University of Nebraska Medical Center [this does not include cases with t(2;5), t(3;14), t(8;14), t(11;14) or t(14;!8) rearrangements]; - = loci of described proto-oncogenes, oncogenes and critical regions associated with lymphomagenesis.
Future studies 99
lymphoma and isochromosome 7q is thought to be primary nonrandom abnormality.74'75 There is also a relatively large group of patients (approximately 10 per cent) with additional chromosome material translocated to the IgH region of 14q32, which does not involve chromosomes 2, 3, 8, 11 or 18, and most likely represents other groups of primary chromosome rearrangements, which have not yet been adequately studied.12,76,77 Additionally, some of the Ip and 6q rearrangements represent additional primary/secondary chromosome lesions that have been studied cytogenetically or molecularly.78"83 Chromosome regions lp35-36,5qll, 6q21,9p24,12ql3, 13qll, 15pll, 15q21-24 have been reported (in at least two different cases) in reciprocal or balanced translocations with 14q32 or 22ql 1, the regions containing the IgH and IgK loci, respectively.84 These regions may be sites of putative oncogenes, which are activated through juxtaposition to Ig loci and may be potentially important in lymphomagenesis. Thirty per cent of cases in a UK series of lymphomas have also been shown to contain additional X chromosomes.85 Although structural changes affecting the X chromosome are infrequent, the regions most commonly involved are X p22 and X q28.86 Investigation of secondary cytogenetic abnormalities has been addressed by performing serial biopsies on NHL patients.11'87'88 Another approach has been that of analysing chromosome rearrangements that have occurred in conjunction with those chromosomal lesions known to be a primary event.7 These studies have revealed that some of the more common secondary numerical abnormalities include trisomy for chromosomes 3, 7, 8,9,12,17,18, 20 and 21. Some of the more common probable secondary structural abnormalities include isochromosomes Iq, 6p and 17q89 and rearrangements involving lp21-22, lq21-22, Ip36, 2q21, 6ql3, 6q21, 10q22-24.
CLINICAL AND PROGNOSTIC CORRELATIONS A number of studies have attempted to correlate specific chromosome abnormalities with clinical and prognostic factors7,11,68,90-93 (Table 9.2). Although the t(14;18) is the most common abnormality in NHL, it is more frequently seen in follicular histology than in diffuse histology, but most studies have been unable to agree upon whether a correlation with clinical outcome exists. However, other secondary abnormalities, occurring with the t(14;18), have revealed that either an additional chromosome 21 or isochromosome of the large arm of chromosome 17 is associated with a shorter survival.94 Trisomy 7, in conjunction with a t( 14;18), is associated with a diffuse histological pattern, rather than a follicular pattern. Also, an additional der(18)t(14;18) and trisomy 12 are associated with disease progression. On the other hand, absence of the t(14;18) but presence of 6q rearrangements is associ-
ated with shorter survivals.11,95 Involvement of lq21-22 was also significantly associated with poor survival in patients with an intermediate or high-grade disease. Patients with Ki-l+ anaplastic large cell lymphoma with the t(2;5) have a favorable prognosis compared to anaplastic large cell cases without the t(2;5).64 Shorter survivals were also correlated with trisomy 5, 6, 18, and any rearrangement of chromosome 5 or a breakpoint at 14qll-12.8 The 13ql4 deletions have been correlated with low complete remission rate and poor survival.96,97,98 Also, in general, the more complex the karyotype, the poorer the prognosis.99 Additionally, patients with normal metaphases tend to survive longer than those with only abnormal cells, with the possible exception of Hodgkin's disease.15 If the cytogenetically normal cells represent non-neoplastic cells, the frequency of abnormal cells compared to normal cells may imply an immune response to the tumor or may reflect a smaller tumor burden than was presumed and thereby reflect a longer survival.
FUTURE STUDIES The role of cytogenetic changes in lymphomagenesis has been addressed by numerous studies and investigators. Specific primary chromosome abnormalities have been described in the lymphomagenesis process and numerous secondary abnormalities are known to be associated with prognosis and clinical outcome. Cytogenetic elucidation of the lymphomas has been somewhat hampered by the complexity of chromosome abnormalities in lymphomas compared to other hematological neoplastic processes, by frequent less than ideal chromosome morphology and, to some extent, by inconsistency of histological and cytogenetic classification by different research groups. It is imperative that there is a marriage between cytogenetics and molecular biology to elucidate the chromosome rearrangements that have proved to be important in lymphomagenesis and lymphoma progression. It is important that cytogeneticists and molecular biologists communicate closely and collaborate in studying the genetic role of lymphomagenesis and lymphoma progression as well as the association of specific chromosome rearrangements with potential treatment modalities. Chromosome rearrangements that previously could not be defined by classic cytogenetics can now be classified in view of the relatively recent fluorescent FISH methodologies that are utilized increasingly by cytogenetics laboratories.76,78-80,100-104 Plate 60 depicts the presence of a t(14;18) in a metaphase preparation utilizing FISH procedures incorporating a two-color whole chromosome painting and telomere probe assay. Also, in the large number of patients with translocations of unidentifiable material to 14q32 and to other critical chromosome regions, the previously unknown chromosome material can now be ascertained by utilizing some
100 Cytogenetics
of the new FISH technologies,76,78,100,102 including comparative genomic hybridization (CGH) and multicolor FISH.105-107 Plate 61 depicits multicolor FISH analysis of an NHL case which could not be clearly resolved by cytogenetics alone. Likewise, with the development of additional probes, some of the chromosome rearrangements can be more meticulously elucidated by the polymerase chain reaction and other molecular techniques. This must, of course, be correlated and compared with cytogenetic observations. It is quite likely, as we begin to better understand the nature of chromosome rearrangements that cause the loss of cell control in the lymphoma process, that specific and effective treatment modalities be considered for each different type of primary chromosomal rearrangement associated with the lymphoma process. It is becoming more apparent, with the development of non-conventional therapeutics, that specific treatment modalities might be developed for each subtype of lymphoma, based on the specific cytogenetic and/or molecular lesion present.108 In order for us to better understand this spectrum of lymphoma diseases, and to improve disease management, it is mandatory that clinicians, pathologists, molecular biologists, cytogeneticists and researchers in other disciplines collaborate closely, communicate and pursue appropriate clinical trials based on solid biological logic.
5. Bloomfield CD, Arthur DC, Frizzera G, Levine EG, Peterson BA, Gajl-Peczalska KJ. Nonrandom chromosome abnormalities in lymphoma. Cancer Res 1983;43:2975-84. 6. Yunis JJ, Oken MM, Theologides A, Howe RB, Kaplan ME. Recurrent chromosomal defects are found in most patients with non-Hodgkin's lymphoma. Cancer Genet Cytogenet 1984; 13:17-28. 7. Armitage JO, Sanger WG, Weisenburger DD, et al. Correlation of secondary cytogenetic abnormalities with histologic appearances in non-Hodgkin's lymphomas bearing t(14;18)(q32;q21).y Natl Cancer Inst 1988; 80: 576-80. 8. Schouten HC, Sanger WG, Weisenburger DD, Armitage JO. Chromosomal abnormalities in untreated patients with non-Hodgkin's lymphoma: associations with histology, clinical characteristics and treatment outcome. B/ood1990; 75:1841-7. 9. Juneja S, Lukeis R, Tan L, et al. Cytogenetic analysis of 147 cases of non-Hodgkin's lymphoma: non-random chromosomal abnormalities and histological correlations. BrJ Haematol 1990; 76: 231-7. 10. Offit K, Wong G, Filippa DA, Tao Y, Chaganti RS. Cytogenetic analysis of 434 consecutively ascertained specimens of non-Hodgkin's lymphoma: clinical correlations. Blood 1991; 77:1508-15. 11. Whang-Peng J, Knutsen T, Jeffe ES, et al. Sequential
ACKNOWLEDGEMENTS
analysis of 43 patients with non-Hodgkin's lymphoma: clinical correlations with cytogenetic, histologic, immunophenotyping, and molecular studies. Blood
The authors gratefully acknowledge the grant support from the National Cancer Institute (USPHS CA36727), the Department of Health and Human Services, Bethesda, MD, USA, and the Leukemia and Lymphoma Society of America, USA (LSA 6032-99). Thanks to Kathy Olin for her skills in the typing and editing of this chapter, as well as special thanks for the assistance of Michelle Hess, Diane Pickering and Christine Higgins for their help with some of the illustrations.
1995;85:203-16. 12. Sanger WG, and the Nebraska Lymphoma Group. Current Nebraska Lymphoma Study Group database including 1200 lymphoma cases. 1995. 13. Mitelman F (ed). ISCN: An International System for Human Cytogenetic Nomenclature. Basel: Karger, 1995. 14. Slavutsky I, de Vinuesa ML, Esteuev ME, Sen L, de Salum SB. Cytogenetic and immunologic phenotype findings in Hodgkin's disease. Cancer Genet Cytogenet 1985; 114: 123-8. 15. Schouten HC, Sanger WG, Duggan M, Weisenburger DD,
REFERENCES 1. Boveri, T. Zur Frage de Entstehung Maligner Tumoren. Jena: Fischer, 1914. 2. Nowell P, Hungerford DA. A minute chromosome in human chronic granulocytic leukemia. Science 1960; 132:1497. 3. Caspersson T, Cahrton G, Lindsten J, Zech L Identification of the Philadelphia chromosome as a number 22 by quinacrine mustard fluorescence analysis. Exp Cell Res 1970; 63: 238-44. 4. Rowley JD. A new consistent chromosomal abnormality in chronic myelogenous leukemia identified by
MacLennan KA, Armitage JO. Chromosomal abnormalities in Hodgkin's disease. Blood 1989; 73: 2149-54. 16. Ladanyi M, Parsa NZ, Offit K, Wachtel MS, Filippa DA, Jhanwar, SC. Clonal cytogenetic abnormalities in Hodgkin's disease. Genes Chromosomes Cancer 1991; 3: 294-9. 17. Tilly H, Bastard C, Delastre T, et al. Cytogenetic studies in untreated Hodgkin's disease. Blood 1991; 77: 1298-304. 18. Dohner H, Bloomfield CD, Frizzera G, et al. Recurring chromosome abnormalities in Hodgkin's disease. Genes Chromosomes Cancer 1992; 5: 392-8. 19. Koduru PRK, Susin, M, Schulman P, et al. Phenotypic
quinicrine fluorescence and Giemsa banding. Nature
and genotypic characterization of Hodgkin's disease.
1973; 243: 290-2.
AmJHematol 1993; 44:117-24.
References 101
20. Atkin NB. Cytogenetics of Hodgkin's disease. Cytogenet Cell Genet 1998; 80:23-7.
35. Ye BH, Lista F, Coco FL, et al. Alterations of zinc finger-
21. Pedersen RK, Sorensen AG, Pedersen NT, etal. Chromosome aberrations in adult Hodgkin disease in a
Science 1993; 262: 747-50. 36. Bastard C, Tilly H, Lenormand B, et al. Translocations
Danish population-based study. Cancer Genet Cytogenet 1999; 110:128-32.
involving band 3q27 and Ig gene regions in nonHodgkin's lymphoma. Blood 1992; 79: 2527-31.
22. Falzetti D, Crescenzi B, Matteucci C, etal. Genomic instability and recurrent breakpoints are main cytogenetic findings in Hodgkin's disease. Haematologica 1999; 84: 298-305. 23. Schlegelberger B, Weber-Matthiesen K, Himmler A, et al. Cytogenetic findings and results of combined immunophenotypingand karyotyping in Hodgkin's disease. Leukemia 1994; 8: 72-80. 24. Weber-Matthiesen J, DeerbergJ, Poetsch M, etal. Clarification of dubious karyotypes in Hodgkin's disease
encoding gene, BCL-6, in diffuse large-cell lymphoma.
37. Baron BW, Nucifora G, McCabe N, Espinosa R III, Le Beau MM, McKeithan TW. Identification of the gene associated with the recurring chromosomal translocations t(3;14)(q27;q32) and t(3;22)(q27;q11) in B-cell lymphomas. Proc Natl Acad Sci USA 1993; 90: 5262-6. 38. Of fit K, Lo Coco F, Louie DC, et al. Rearrangement of the bcl-6 gene as a prognostic marker in diffuse large-cell
by simultaneous fluorescence immunophenotyping and
lymphoma. N EnglJ Med 1994; 331: 74-80. 39. Horsman DE, McNeil BK, Anderson M, SehenkierT, Gascoyne RD. Frequent association of t(3;14) or variant with other lymphoma-specific translocations. BrJ
interphase cytogenetics (FICTION). Cytogenet Cell Genet 1995; 70: 243-5.
Hematol 1995; 59: 569-75. 40. LoCoco F, Ye BH, Lista F, et al. Rearrangements of the
25. Pringle JH, Shaw JA, Gillies A, Lauder I. Numerical chromosomal aberrations in Hodgkin's disease detected by in situ hybridisation on routine paraffin sections.) Clin Pathol 1997; 50: 553-8. 26. Cabanillas F, Pathak S, Trujillo J, et al. Cytogenetic features of Hodgkin's disease suggest possible origin from a lymphocyte. Blood 1988; 71:1615-17. 27. Mitelman F, Mertens F, Johansson B. A breakpoint map of recurrent chromosomal rearrangements in human neoplasia. Nature Genet 1997; 15: 417-74. 28. The Non-Hodgkin's Lymphoma Pathologic Classification Project. National Cancer Institute sponsored study of classification of non-Hodgkin's lymphomas: summary and description of a working formulation for clinical usage. Cancer 1982; 49: 2112-35. 29. Nowell PC, Croce CM. Chromosome translocations and oncogenes in human lymphoid tumors. Am J Clin Pathol 1990;94:229-37. 30. Manolov G, Manolova Y. Marker band in one chromosome 14 from Burkitt's lymphoma. Nature 1972; 237: 33-4. 31. Kornblau SM, Goodacre A, Cabanillas F. Chromosomal abnormalities in adult non-endemic Burkitt's lymphoma and leukemia: 22 new reports and a review of 148 cases from the literature. Hematol Oncol 1991; 9: 63-78. 32. YunisJJ, Oken MM, Kaplan ME, Ensrud KM, Howe RR,
bcl-6 gene in diffuse large cell non-Hodgkin's lymphoma. Blood 1994; 83:1757-9. 41. Williams ME, Swerdlow SH, Meeker TC. Chromosome t(11;14) (q13;q32) breakpoints in centrocytic lymphoma are highly localized at the bcl-1 major translocation cluster. Leukemia 1993; 7:1437-40. 42. Weisenburger D, Sanger WG, Armitage JO, Purtilo DT. Intermediate lymphocytic lymphoma: Immunophenotypic and cytogenetic find ings. Blood 1987; 69:1617-21. 43. Rimokh R, Berger F, Delsol G, et al. Detection of the chromosomal translocation t(11;14) by polymerase chain reaction in mantle cell lymphomas. Blood 1994; 83: 1871-5. 44. Rosenberg CL, Wong E, Petty EM, et al. PRAD-1, a candidate BCL-1 oncogene: mapping and expression in centrocytic lymphoma. Proc Natl Acad Sci, USA 1991; 88: 9638-12. 45. Offit K, Parsa NZ, Filippa D, et al. t(9;14)(p13;q32) denotes a subset of low-grade non-Hodgkin's lymphoma with plasmocytoid differentiation. Blood 1992; 80: 2594-9. 46. Harris NL, Jaff ES, Stein H, et al. A revised European-American classification of lymphoid neoplasms: a proposal from the international lymphoma study group. Blood 1994; 84:1361-92. 47. lida S, Rao PH, Nallasivam P, et al. The t(9;14)(p13;q32)
Theoglides A. Distinctive chromosomal abnormalities in
chromosomal translocation associated with
histologic subtypes on non-Hodgkin's lymphomas. N EnglJ Med 1982; 307:1231-6.
gene. Blood 1996; 88: 4110-17.
33. Jaeger U, Karth GD, Friedl J, Laczika K, Kusec R. Molecular mechanism of the t(14;18) - a model for lymphoid-specific chromosomal translocations. Leuk Lymphoma 1994; 14:197-202. 34. Klefstrom J, Franssila K, Peltomaki P, Kaartinen M, Solin ML, Knuutila S. Major and minor breakpoint sites of chromosomal translocation t(14;18) in subtypes of nonHodgkin's lymphomas. Leuk Res 1994; 18: 245-50.
lymphoplasmacytoid lymphoma involves the PAX-5 48. Offit K, Parsa NZ, Jhanwar SC, et al. Clusters of chromosome 9 aberrations are associated with clinicopathologic subsets of non Hodgkin's lymphoma. Genes Chromosomes Cancer 1993; 7:1-7. 49. Griffin CA, Zehnbauer BA, Beschorner WE, et al. t(11;18)(q21;q21) is a recurrent chromosome abnormality in small lymphocytic lymphoma. Genes Chromosomes Cancer 1992; 4:153-7.
102 Cytogenetics 50. Leroux D, Seite P, Million J, etal. t(11;18)(q21;q21) may delineate a spectrum of diffuse small B-cell lymphoma with extranodal involvement. Genes Chromosomes Cancer 1993; 7: 54-6. 51. Horsman D, Gascoyne R, Klasa R, Coupland R. t(11;18)(q21;q21): a recurring translocation in lymphoid tissue (MALT)? Genes Chromosomes Cancer 1992; 4:183-7. 52. Ott G, Katzenberger T, Greiner A, et al. The t(11;18)(q21;q21) chromosome translocation is a frequent and specific aberration in low-grade but not high-grade malignant non-Hodgkin's lymphomas of the mucosa-associated lymphoid tissue (MALT-) type. Cancer Res 1997; 57: 3944-8. 53. Wotherspoon AC, Pan L, Diss TC, Issacson PG. A genotypic study of low-grade B-cell lymphomas, including lymphomas of mucosa associated lymphoid tissue (MALT). 7 Pathol 1992; 162:135-40. 54. Clark HM, Jones DB, Wright DH. Cytogenetic and molecular genetic studies of the t(14;18) and t(14;19) in nodal and extranodal B-cell lymphoma. J Pathol 1992; 166:129-37. 55. Dierlamm J, Pittaluga S, Wlodarska I, etal. Marginal zone B-cell lymphomas of different sites share similar cytogenetic and morphologic features. Blood 1996; 87: 299-307. 56. Morris SW, Kirstein MW, Valentine MB, et al. Fusion of a kinase gene, ALK, to a nucleolar protein gene, in nonHodgkin's lymphoma. Science 1994; 263:1281-4. 57. Knuutila S, Lakkala T, Teerenhovi L, Peltomaki P, Kovanen R, Franssila K. t(2;5)(p23;q35) - a specific chromosome abnormality in large cell anaplastic (Ki-1) lymphoma. Leuk Lymphoma 1990; 3: 53-9. 58. Elmberger PG, Lozano M, Weisenburger D, Sanger WG, Chan W. Transcripts of the NPM-ALK fusion gene in anaplastic large cell lymphoma, Hodgkin's disease and reactive lymphoid lesions. Blood 1995; 86: 3517-21. 59. Bullrich F, Morris SW, Hummel M, Pileri S, Stein H, Croce CM. Nucleophosmin (NPM) gene rearrangements in Ki1-positive lymphomas. Cancer Res 1994; 54: 2873-7. 60. Weisenburger DD, Gordon BG, Vose JM, et al. Occurrence of the (2;5)(p23;q35) in non-Hodgkin's lymphoma. Blood 1996; 87: 3860-8. 61. Gordon BG, Weisenburger DD, Warkentin PI, etal. Peripheral T-cell lymphoma in childhood and adolescence. Cancer 1993; 71: 257-63. 62. Lopategui JR, Sun LH, Chan JK, et al. Low frequency association of the t(2;5)(p23;q35) chromosomal translocation with CD30+ lymphomas from American and Asian patients. Am J Pathol 1995; 146: 323-8. 63. Romaguera JE, ManningJT, Tornos CS, etal. Long-term prognostic importance of primary Ki-1 (CD30) antigen expression and anaplastic morphology in adult patients with diffuse large-cell lymphoma. Ann Oncol 1994; 5: 317-22. 64. Pirc-Danoewinata H, Chott A, Onderka E, et al. Karyotype and prognosis in non-Hodgkin's lymphoma. Leukemia 1994; 8:1929-39.
65. Shiota M, Nakamura S, Ichinohasama R, etal. Anaplastic large cell lymphomas expressing the novel chimeric protein p80 NPM/ALK: a distinct clinicopathologic entity. Blood 1995; 86:1954-60. 66. Wotherspoon AC, Finn TM, Isaacson PG. Trisomy 3 in low-grade B-cell lymphomas of mucosa-associated lymphoid tissue. Blood 1995; 85: 2000-4. 67. Brynes RK, Almaguer PD, Leathery KE, et al. Numerical cytogenetic abnormalities of chromosomes 3, 7 and 12 in marginal zone B-cell lymphomas. Modem Pathol 1996; 9: 995-1000. 68. Offit K, Jhanwar SC, Ladanyi M, et al. Cytogenetic analysis of 434 consecutively ascertained specimens of non-Hodgkin's lymphoma: correlations between recurrent aberrations, histology, and exposure to cytotoxic treatment. Genes Chromosomes Cancer 1991; 3:189-201. 69. Cabanillas F, Pathak S, Trujillo J, et al. Frequent nonrandom chromosome abnormalities in 27 patients with untreated large cell lymphoma and immunoblastic lymphoma. Cancer Res 1988; 48: 5557-64. 70. Fifth International Workshop on Chromosomes in Leukemia-Lymphoma. Correlation of chromosome abnormalities with histologicand immunologic characteristics in non-Hodgkin's lymphoma and adult T-cell leukemia-lymphoma. Blood ^37, 70:1554-64. 71. Offit K, Louie DC, Parsa NZ, etal. Del(7)(q32) is associated with a subset of small lymphocytic lymphoma with plasmacytoid features. Blood 1995; 86: 2365-70. 72. Offit K, Louie DC, Parsa NZ, etal. Clinical and morphologic features of B-cell small lymphocytic lymphoma with del(6)(q21q23). Blood 1994; 83: 2611-8. 73. Mrozek K, Bloomfield CD. Cytogenetics of indolent lymphomas. Seminars in Oncol 1993; 20: 47-57. 74. Wang CC, Tien HF, Lin MT, et al. Consistent presence of isochromosome7q in hepatosplenicT gamma/delta lymphoma: a new cytogenetic-clinicopathologic entity. Genes Chromosomes Cancer 1995; 12:161^. 75. Yao M, Tien HF, Lin MT, et al. Clinical and hematological characteristics of hepatosplenic T gamma/delta lymphoma with isochromosome for long arm of chromosome 7. Leuk Lymphoma 1996; 22: 495-500. 76. Arcaroli JA, Dave BJ, Pickering DL, et al. Is duplication 14q32 a new recurrent alteration in non-Hodgkin's lymphoma? Cancer Genet Cytogenet 1999; 113:19-24. 77. Nakamine H, Masih AS, Chan WC, Sanger WG, Armitage JO, Weisenburger DD. Oncogene rearrangement in nonHodgkin's lymphoma with a 14q+ chromosome of unknown origin. Lymphoma Leuk 1993; 10: 79-88. 78. Dave BJ, Hess MM, Pickering DL, etal. Rearrangements of chromosome band 1p36 in non-Hodgkin's lymphoma. Clinical Cancer Res 1999; 5:1401-9. 79. Dave BJ Pickering DL, Hess MM, etal. Deletion of Cell Division Cycle 2 Like 1 gene locus on chromosome 1p36 in non-Hodgkin's lymphoma. Cancer Genet Cytogenet 1999; 108:120-6.
References 103 80. Zhang Q, Siebert R, Van M, et al. Inactivating mutations and overexpression of BCL10, a caspase recruitment domain-containing gene, in MALT lymphoma with t(1;14)(p22;q32). Nature Genet 1999; 22: 63-8. 81. Guan XY, Horsman D, Zhang HE, etal. Localization by chromosome microdissection of a recurrent breakpoint region on chromosome 6 in human B-cell lymphoma. Blood 1996; 88:1418-22. 82. Vitolo U, Gaidano G, Botto B, et al. Rearrangements of bcl-6, bcl-2, c-myc and 6q deletion in B-diffuse largecell lymphoma: clinical relevance in 71 patients. Ann Oncol 1998; 9: 55-61. 83. Zhang Y, Matthiesen P, Harder S, et al. A 3-cM commonly deleted region in 6q21 in leukemias and lymphomas delineated by fluorescence in situ hybridization. Genes Chromosomes Cancer 2000; 27: 52-8. 84. Mrozek K, Bloomfield CD. New recurrent structural chromosome aberrations in non-Hodgkin's lymphoma: a review of the published literature. Int J Oncol 1996; 8: 851-8. 85. Hammond DW, Goepel JR, Aitken M, Hancock BW, Potter AM, Goyns MH. Cytogenetic analysis of a United Kingdom series of non-Hodgkin's lymphomas. Cancer Genet Cytogenet 1992; 61: 31-8. 86. Goyns MH, Hammond DW, Harrison CJ, Menasce LP, Ross F, Hancock BW. Structural abnormalities of the X chromosome in non-Hodgkin's lymphoma. Leukemia 1993; 7: 848-52. 87. Sanger WG, Armitage JO, Bridge JA. Weisenburger DD, Fordyce R, and Purtilo DT. Initial and subsequent cytogenetic studies in malignant lymphoma. Cancer 1987; 60: 3014-19. 88. Levine EG, Juneja S, Arthur D, et al. Sequentional karyotypes in non-Hodgkin's lymphoma: their nature and significance. Genes Chromosomes Cancer 1990; 1: 270-80. 89. Mertens F, Johansson B, Mitelman F. Isochromosomes in neoplasia. Genes Chromosomes Cancer 1994; 10: 221-30. 90. Offit K, Chaganti RS. Chromosomal aberrations in nonHodgkin's lymphoma. Biological and clinical correlations. Hematol Oncol Clinic North America 1991; 5: 853-69. 91. Levine EG, Arthur DC, Frizzera G, et al. Cytogenetic abnormalities predict clinical outcome in non-Hodgkin's lymphoma. Ann Intern Med 1988; 108:14-20. 92. Cabanillas F, Pathak S, Grant G, et al. Refractoriness to chemotherapy and poor survival related to abnormalities of chromosomes 17 and 7 in lymphoma. Am J Med 1989; 87:167-72. 93. Johansson B, Mertens f, Mitelman F. Cytogenetic evolution patterns in non-Hodgkin's lymphoma. Blood 1995; 86: 3905-14. 94. Weisenburger DD, Gascoyne RD, Bierman PJ, et al. Clinical significance of the t(14;18) and BCL2 overexpression in follicular large cell lymphoma. Leuk Lymphoma 2000; 36: 513-23.
95. Schouten H, Sanger WG, Weisenburger DD, Armitage JO. Abnormalities involving chromosome 6 in newly diagnosed patients with non-Hodgkin's lymphoma. Cancer Genet Cytogenet 1990; 47: 73-89. 96. Cuneo A, Bigoni, R, Rigolin GM,et al. 13q14 deletion in non-Hodgkin's lymphoma: correlation with clinicopathologic features. Haematologica 1999; 84: 589-93. 97. Wada M, Okamura T, Okada M, etal. Frequent chromosome arm 13q deletion in aggressive nonHodgkin's lymphoma. Leukemia 1999; 13: 792-8. 98. Rosenwald A, Ott G, Krumdiek AK, et al. A biological role for deletions in chromosomal band 13q14 in mantle cell and peripheral t-cell lymphomas? Genes Chromosomes Cflncer 1999; 26: 210-4. 99. Tilly H, Rossi A, Stamatoullas A, et al. Prognostic value of chromosomal abnormalities in follicular lymphoma. Blood 1994; 84:1043-9. 100. Bajalica S, Sorenson AG, Pederson NT, Heim S, Brodrum-Nielson K. Chromosome painting as a supplement to cytogenetic banding analysis in nonHodgkin's lymphoma. Genes Chromosomes Cancer 1993; 7:231-9. 101. Hammond DW, Hancock BW, Goyns MH. Chromosome analysis of non-Hodgkin's lymphomas by fluorescence in-situ hybridization. Ann Oncol 1994; 5 (suppl 1): 51-4. 102. Hammond DW, Hancock BW, Goyns MH. Analysis of 14q+ derivative chromosomes in non-Hodgkin's lymphomas by fluorescent in-situ hybridization. Leuk Lymphoma 1995; 20:111-7. 103. Zhang Y, Weber-Matthiesen K, Siebert R, et al. Frequent deletions of 6q23-24 in B-cell non-Hodgkin's lymphomas detected by fluorescence in situ hybridization. Genes Chromosomes Cancer 1997; 18: 310-3. 104. Mathew P, Sanger WG, Weisenburger DD, et al. Detection of the t(2;5)(p23;q35) and NPM-ALK fusion in nonHodgkin's lymphoma by two color fluorescence in situ hybridization. Blood 1997; 89:1678-85. 105. Peters K, Zettl A, Starostik P, et al. Genetic imbalances in primary gastric diffuse large B-cell lymphomas: comparison of comparative genomic hybridization, microsatellite, and cytogenetic analysis. Diagn Mol Pathol 2000; 9: 58-65. 106. Alqahtani MH, Hammond DW, Goepel JR, Goyns MH. Cytogenetic analysis of non-Hodgkin's lymphomas by ratio-painting and comparative genomic hybridization reveals unsuspected chromosomal abnormalities. Leuk Lymphoma 1999; 35: 325-37. 107. Hilgenfeld E, Padilla-Nash H, Schrock E, Reid T. Analysis of B-cell neoplasias by special karotyping (SKY). Curr Top Microbiol Immunol 1999; 246:169-74. 108. Hammond DW, Hancock BW, Goyns MH. Clinical implications of molecular and cytogenetic studies of non-Hodgkin's lymphomas. Cancer Treat Rev 1998; 24: 157-63.
This page intentionally left blank
PART
Pathogenesis
Hodgkin's disease Viruses and malignant lymphoma Molecular biology
107 115 133
2
This page intentionally left blank
10 Hodgkin's disease VDIEHL AND J WOLF
Introduction Hodgkin's disease: a malignant proliferation of Hodgkin/Reed-Sternberg cells
107 107
INTRODUCTION There is still a discrepancy between the impressive success of radiation therapy and/or chemotherapy in the treatment of Hodgkin's lymphoma and the lack of understanding of the pathogenetic essential events of this disease, which unusually combines the features of a neoplasm with those of an inflammatory/infectious process. The main reason for this is that methodological problems occur, owing to the scarcity of the Hodgkin/Reed-Sternberg (H-RS) cells in affected lymphatic tissue and their poor outgrowth in vitro and in vivo. Isolation of single H-RS cells from frozen lymph node sections or cytospin preparations followed by enzymatic amplification of their nucleic acids (H-RS single-cell polymerase chain reaction; PCR) provides an opportunity to perform specific genetic analyses of the H-RS cells as well as of their bystander cells. Using this new method, the monoclonal B cell origin of H-RS cells could be demonstrated, at least in the majority of cases analysed. Although Hodgkin's disease can be understood now as a monoclonal B cell disorder, the mechanism of transformation is still unclear. New cytogenetic techniques, such as combined immunophenotyping and fluorescence in situ hybridization (FISH), show that all H-RS cells have structural and/or numerical, but not consistent, chromosomal aberrations. Cytogenetic instability might thus represent one pathogenetic factor in Hodgkin's disease. Epstein-Barr virus (EBV) infection of H-RS cells has now been confirmed in about 50 per cent of cases in industrialized countries and in up to 95 per cent in developing countries. EBV-positive H-RS cells express the viral latent protein LMP with transforming as well as (T cell) immunogenic functions. Activated T helper cells represent the majority of lymphocytes
Hodgkin's disease: an atypical immune reaction
110
Conclusions
111
References
112
surrounding H-RS cells. The H-RS cells themselves express all accessory molecules necessary for efficient T cell recruitment. Thus, in the initial stages of Hodgkin's disease a pronounced but inefficient T cell reaction against a target antigen (LMP in EBV-positive cases?) expressed on H-RS cells seems to take place.
HODGKIN'S DISEASE: A MALIGNANT PROLIFERATION OF HODGKIN/REED-STERNBERG CELLS Lineage origin and cionality Lymph nodes affected by Hodgkin's disease (HD) consist of a heterogeneous mixture of lymphocytes, histiocytes, eosinophils, plasma cells, fibroblasts and other cells. The mononuclear Hodgkin cells and their polynucleated counterparts, the Reed-Sternberg cells, are considered to represent the malignant substrate of classical HD, which summarizes the nodular sclerosing subtype (NSHD), the mixed cellularity subtype (MCHD) and the lymphocyte-depleted subtype (LDHD). Hodgkin/ Reed-Sternberg cells represent only 0.1-1 per cent of the entire cell population in classical HD.1 In the lymphocyte-predominant subtype of HD (LPHD), the lymphocytic and histiocytic (L & H) cells represent the malignant population and consistently express B cell specific surface antigens (CD 19, CD 20). In classical HD, H-RS cells in the majority of cases express the activation markers Ki-1 (CD 30), the Leu-Mi antigen (CD 15), the interleukin-2 receptor (CD 25), the transferrin receptor (CD 71) and HLA class II molecules (HLA-DR), but not surface antigens, which might help unequivocally to determine their cell of origin (Table
108 Hodgkin's disease
Table 10.1 Immunophenotyping of H-RS cells in biopsies:" frequency of cases (per cent) with expression of a given marker dependent on histological subtype
Activation markers CD 30 (KM) CD15(LeuM1) CD25(IL2-R) CD 71 (transferrin-R) HLA-DR
84-91 76-85 82-86 64-94 95-97
32 28 100 89 83
13-30 0(LD)-19
77 88
17-34 13-21 24-33 1-6
5 12 5 6
B cell markers CD 20 CD 75
T cell markers CD 2 CD 3 CD 4
CDS
Monocytes/macrophages CD11b 0 2-14 CD 14
0 0
a
See also Drexler et al. (1992). NS = nodular sclerosis, LP = lymphocyte depleted, MC = mixed cellularity, LP = lymphocyte predominant.
10.1).2'3 Expression of B cell specific and of T cell specific antigens as well as the absence of both have been described. Similiarly, analysis for immunoglobulin (Ig) gene or T cell receptor (TCR) gene rearrangements using Southern blotting has revealed heterogeneous results concerning the cell of origin and clonality.4'6 A methodological breakthrough in the genetic analysis of H-RS cells represents micromanipulation of single H-RS cells from frozen biopsy sections or cytospin preparations, and subsequent enzymatic amplification of their DMA (H-RS single-cell PCR).7'8 Kiippers et al. isolated single H-RS cells from frozen lymph-node sections (Plate 62).7 In all of three cases analysed (nodular sclerosis, lymphocyte predominant and mixed cellularity) clonal rearrangements of genes encoding for Ig heavy chains (VH) could be detected; in addition, in one of these cases a light-chain gene rearrangement was found. Recently, these studies were extended on 11 further cases of HD. In 10 of these cases, clonal Ig gene rearrangements also could be detected.9'10 Using the same method, Hummel et al. studied 12 cases of HD with the B cell immunophenotype for Ig gene heavychain rearrangements.11 In three cases identical (monoclonal) rearrangements were found, in six cases heterogenous (polyclonal) rearrangements and in three cases monoclonal as well as polyclonal rearrangements. However, a reanalysis of four of the polyclonal cases revealed a clonal H-RS cell population in two cases.12
Delabie et al. analysed single cells resuspended from formalin-fixed, paraffin-embedded lymph-node specimens from patients with LPHD. In four cases polyclonal Ig gene rearrangements were described.13 Similarly, only polyclonal Ig gene rearrangements were found by this group in three cases of NSHD with a B cell immunophenotype, while in three cases without expression of lineage-specific antigens (null-phenotype), no Ig gene rearrangement could be amplified.14 Since it could not be ruled out, however, that the specific method used for isolating the tumor cells (see earlier) led to nondetection of clonal Ig gene rearrangements, this group performed another study with micromanipulation of single cells from frozen tissue sections. In five cases of LPHD clonally related Ig gene rearrangements were detected in L & H cells.15 The finding of clonal L & H cell populations was confirmed in further cases of LPHD by others.4'16 In summary, the results of single-cell analysis obtained so far suggest that both H-RS cells in classical HD and L & H cells in LPHD are clonal B cells. The view of HD as a clonal B cell disorder is further supported by the detection of clonally identical H-RS cells in the first diagnosis and disseminated relapse of a patient with MCHD17'18 as well as by the detection of clonal EBV genomes19 and of clonal cytogenetic aberrations20 in H-RS cells obtained from different lymph nodes of the same patient, and in different nodules in one node of NSHD.21 DNA sequence analysis of the rearranged Ig genes in LPHD revealed ongoing mutations in 14 out of 21 cases.4,15,16 These observations suggest that L & H cells are germinal-center-derived B cells and that their growth is dependent on antigen binding and selection, comparable to follicular lymphoma cells. In contrast, H-RS cells in classical HD accumulate 'crippling' somatic mutations within their rearranged Ig genes, which prevent antibody expression.10 Thus, most probably, H-RS cells also derive from germinal center B cells. However, they seem to proliferate independently from antigen selection and antibody expression. Physiologically, germinal center B cells with 'crippling' mutations are eliminated by apoptosis. A still unknown mechanism seems to prevent apoptotic death of H-RS cells. Initially studies suggested that H-RS cells showed no telomerase activity implying that they might require novel mechanisms to remain immortal.22 Further work shows that they maintain telomerase but that ribonucleases released by eosinophils degrade the telomerase message leading to misleading negative results.23
Genetic aberrations The questions regarding lineage origin and clonality of H-RS cells are not the only ones to remain unanswered for a long time. Similarly, no consistent cytogenetic aberration with specific alteration of gene expression could
Malignant proliferation of H-RS cells 109
be identified, for instance in Burkitt's lymphoma, follicular lymphoma or chronic myeloid leukemia. Even analysis of gene expression in H-RS cells using the single cell technique revealed no specific pattern.8 In 1995, the presence of npm-alk fusion transcripts was described in H-RS cells.24 These transcripts, which are translated into a protein with tyrosine kinase activity, result from a chromosomal translocation (2;5) and are consistently found in anaplastic large cell lymphoma (ALCL). Other groups, however, could not confirm these results in HD, either by detection of npm-alk transcripts using reverse transcription PCR (RT-PCR) or by immunostaining of the resulting fusion protein.25-27 Conventional karyotype analysis of H-RS cells is hampered by their paucity in affected tissue and their low proliferation rate. Thus, only in a proportion of the cases analysed have clonal chromosomal aberrations been described in H-RS cells.28-30 Tilly et al. karyotyped 60 lymph nodes from untreated HD patients.31 In 82 per cent of the analysable cases, metaphases were obtained, and in 55 per cent numerical and/or structural aberrations were detected. Although some aberrations occurred more frequently than others (e.g. loss of chromosome 13), no consistent change was found. It also remains an open question whether all of these aberrations have been derived from H-RS cells or whether they might also be present in the surrounding bystander cells. The latter possibility may be underlined by transplantation of lymph node and liver specimens affected by HD under the subrenal capsule of T and B cell-deficient severe combined immunodeficient (SCID) mice. In these experiments EBV-positive B cell tumors grew out. These B lymphoblasts had a high frequency of numerical and structural chromosomal aberrations, which are not found in 'normal' EBV-immortalized B cells.32 In situ hybridization studies have suggested chromosomal abnormalities in morphologically normal cells in HD next to H-RS cells.33 Furthermore, cells with the morphology of small lymphocytes in chronic lymphocytic leukemia (CLL) can evolve to H-RS-like cells in Richter's syndrome ('Hodgkin transformation of CLL') and display identical genotypes.34 These observations point to the presence of karyotypically abnormal (semimalignant?) B cells in the surroundings of the H-RS cells, which gain a growth advantage after transplantation into SCID mice. HD-specific problems of conventional cytogenetics may be partly solved by simultaneous FISH and immunophenotyping. This method allows the specific analysis of chromosomal aberrations in CD 30-positive cells. Using this method, Weber-Matthiesen et al35 analysed 30 HD cases. While in 21 of these cases no aberrations could be demonstrated with conventional karyotyping, in all cases numerical chromosomal aberrations could be detected by combined immunophenotyping and FISH. The variability of the described
aberrations, furthermore, confirms the model of genetic instability as one pathogenetic factor in HD (see also later text).
Cell lines and animal models In numerous human neoplasias, the establishment of an animal model allowed the biological and genetic characterization of the tumor cell population. In contrast, outgrowth of a permanent cell line represents an extremely rare event in HD. So far, only 15 cell lines have been established, which may be regarded as HD-derived. Analysis of immunophenotype, karyotype, Ig or TCR gene rearrangements of these cell lines revealed heterogeneous results analogous to analysis of primary tissue. Expression of CD 30, CD 15 and CD 71 is also found on the cell lines, but there is no consistent feature to define the cell of origin of HD. Moreover, results obtained with these cell lines were considered controversial since their derivation from H-RS cells could not be determined unequivocally.36,37 A novel cell line (L1236) was established from the peripheral blood of a patient with advanced HD of the mixedcellularity subtype (Plate 63). These cells are EBV-negative and express the HD-associated antigens CD 30, CD 15 and CD 71. They have a near triploid karyotype with numerous structural and numerical chromosomal aberrations. By Southern blot analysis a biallelic rearrangement of the Ig gene heavy chains and a monoallelic kappa light-chain rearrangement were detected.38 Using H-RS single-cell PCR, Kanzler et al. have shown that the genomic sequences of the Ig gene rearrangements of the H-RS cells in the patient's bone marrow were identical to those detected in L1236 cells.9 This cell line thus represents the first HD-derived cell line where the H-RS cell origin of the cells has been proved at the molecular level. It is a further example for the monoclonal B cell origin of H-RS cells and thus represents a valid model for biological studies on B cell Hodgkin's disease. Neither HD-derived cell lines or biopsy material could be grown reproducibly in thymus-aplastic T celldeficient nude mice. In contrast, most of the HDderived cell lines form progressively growing tumors after subcutaneous inoculation into SCID mice.39 Owing to a genetic recombinase defect, these animals lack functional T and B cells. The HD-derived cell lines L540,40 HD-MyZ41 and L123642 have been shown to disseminate intralymphatically after inoculation into SCID mice. This experimental model for the in vivo growth of H-RS cells has been used successfully for the preclinical testing of immunotoxins43 and bispecific antibodies.44 In contrast to numerous non-Hodgkin's lymphomas, no reproducible growth of primary H-RS cells has been observed after transplantation of biopsy material in SCID mice.32
110 Hodgkin's disease
nuclear antigen 1 (EBNA 1) and latent membrane protein (IMP) 1 and 2.57'58 This pattern is identical to that found in nasopharyngeal carcinoma endemic in the South West of China, but differs from other EBVassociated neoplasias, such as endemic Burkitt's lymphoma and immunoblastic B cell non-Hodgkin's lymphoma of immunocompromised patients (Table 10.2). Except for EBNA 1, all latent viral proteins represent targets for cytotoxic T lymphocytes.59 Thus, EBVinfected cells either express the complete set of latent viral genes in an immunocompromised host (immunoblastic NHL) or they downregulate these proteins, except EBNA 1 (Burkitt's lymphoma), possibly to escape the host's immune response. So far, it remains unclear how the specific latent viral gene expression pattern in HD (EBNA 2-, LMP+) and the pronounced T cell proliferation in affected lymph nodes relate to each other. The functional relevance of LMP expression in H-RS cells is not understood. LMP has a transforming potential: transformation of epithelial cells after transfection of LMP has been described and mice transgenic for LMP 1 develop lymphomas.60 In lymphocytes, apoptoses can be prevented by LMP via upregulation of the bcl-2 gene.61 LMP is also a target for cytotoxic T cells. In addition, it upregulates (partly in cooperation with EBNA 2) numerous cellular genes, e.g. activation-associated antigens (CD 23, CD 30, CD 39) and adhesion molecules (ICAM-1, LFA-3). Thus it may render a cell indirectly more immunogenic.62 Knecht et al. described mutations in the carboxy-terminal part of the LMP 1 gene identical to those previously reported in LMP-isolates from Chinese nasopharyngeal carcinoma in some HD cases.63 These authors discussed an association of these mutations with a clinically more aggressive HD phenotype. Jarrett et al. detected EBV in H-RS cells significantly more often in patients under 15 years and over 50 years of age compared to young adults, which in the majority were EBV-negative.40 Possibly infection with another still undefined tumor virus occurs in young adults; alternatively, EBV is lost after initial infection. Recently, in a Burkitt's lymphoma hybrid cell model, it has been shown that integration of EBV into the host cell genome can lead to a region of enhanced chromosomal instability (fragile site) leading to loss of the virus together with adjacent chromosomal fragments.64 It remains to be
HODGKIN'S DISEASE: AN ATYPICAL IMMUNE REACTION
Epidemiology Hodgkin's lymphoma presents with a bimodal age/ incidence distribution dependent on demographic features. Three epidemiological patterns can be distinguished: Type I in developing countries shows a first peak in early childhood, a low incidence in young adults and a second peak in the older age groups; and Type III in industrialized countries is characterized by a low incidence in children, a first peak in young adults and a second peak in the elderly. In rural areas of industrialized countries, an intermediate third type has been described.45'46 As in infection with paralytic poliovirus, children with a low socioeconomic status and young adults with a high socioeconomic status have an increased risk of developing Hodgkin's disease.47 These observations have led to the hypothesis of an infectious agent underlying HD.48
EBV infection Individuals suffering from infectious mononucleosis have a 2-3-fold increased risk of developing HD.46 Elevated IgG and IgA titers against the viral capsid antigen (VGA) have also been shown to be correlated with an increased risk for HD.49 Weiss et al. were the first to detect EBV DNA in lymph nodes affected by HD using Southern blotting.50 Subsequently, different groups using different detection methods (immunohistology, Southern blotting, PCR) have reported different, partly controversial results for the association of HD and EBV. In situ hybridization with Epstein-Barr early RNA (EBER)-specific RNA probes (EBER-ISH) combines high sensitivity with high specificity. EBER 1 and 2 RNAs are small EBV-encoded, non-polyadenylated transcripts of high abundancy (about 106 copies/viral genome).51 With EBER-ISH, about 50 per cent of the HD cases in industrialized countries and up to 95 per cent in developing countries were found to harbor the virus in the H-RS cells.52-56 H-RS cells show a specific expression pattern of the latent viral genes with expression of Epstein-Barr
Table 10.2 Pattern of Epstein-Barr virus latent sene expression in EBV-associated neoplasias
Endemic Burkitt's lymphoma Immunoblastic NHL in immunocompromised patients Hodgkin's lymphoma Endemic nasopharyngeal carcinoma a
+ + + +
+ -
About 50% of cases analysed. EBNA = Epstein-Barr nuclear antigen, LMP = latent membrane protein, NHL = non-Hodgkin's lymphoma.
+ -
+ +a +
Conclusions 111
established whether such a hit-and-run mechanism also occurs in 'EBV-negative' HD of young adults.
H-RS cells as antigen-presenting cells Clinically HD often presents with pronounced signs of an inflammatory reaction, such as night sweats, fever, monocytosis. Interleukin-2 (IL-2) and tumor necrosis factor (TNF) receptor levels in sera are elevated, and associated with aggressive disease in some cases.65'66 In affected lymph nodes the rare H-RS cells are surrounded by a majority of non-malignant bystander cells. The (deregulated?) expression of cytokines in H-RS - and bystander - cells may explain part of the complex interaction between these cells. Eosinophilia in HD, for instance, is caused by IL-5,67 and fibrosis may be triggered by IL-1 and TGF-p.6869 The interleukins 1,6 and 9, which are produced by H-RS cells, might act as autocrine growth factors and as paracrine growth stimulators for T cells.70-72 The T cells themselves could stimulate H-RS cells via IL-2 and IL-6. The CD 30/CD 30 ligand interaction probably has a central role in this complex network. After cloning and sequencing the gene, the CD 30 molecule was identified as a member of the TNF receptor superfamily.73 The CD 30 ligand has also been cloned.74 While the CD 30 antigen is consistently expressed on cells of classical HD and on activated bystander cells, the CD 30 ligand is only expressed on the bystanders (T cells, monocytes, granulocytes), and not on the H-RS cells. In tissue culture, some biological functions of the CD 30/CD 30 ligand interaction have now been characterized:75 mitogenic stimulation of H-RS cells, enhanced cytokine production of H-RS cells and activation of antigen expression of H-RS cells. What is the underlying mechanism of the complex interaction between H-RS cells and bystander cells? The majority of lymphocytes in affected lymphatic tissue are activated T helper cells (CD 4+, CD 45RO+, CD 45RB).76'77 These lymphocytes, not the malignant H-RS cells, represent the population with the highest mitotic index in affected lymph nodes.78 The lymph nodes often grow slowly and show fluctuations in their size in early disease stages. These observations point to a cellular immune reaction as primary reason for the lymph node enlargement in HD. In contrast to benign lymphoproliferative leasons, e.g. reactive lymph nodes, the immune reaction in HD is not self-limiting. One reason therefore might be the inability of the immune system to eliminate the malignant cells expressing the target antigen.79 The recently established HD-derived cell line L1236 (see above) expresses HLA class I and II molecules, B7.1 and B7.2 (CD 80, CD 86), and the adhesion molecules ICAM-1 (CD 54) and LFA-3 (CD 58).38 All these molecules are essential for efficient T cell recruitment (accessory molecules). Expression of HLA antigens76 and of the B7.1 molecule80 has also been
described on H-RS cells in biopsies and on other HDderived cell lines (L428, L540; unpublished observation). The presence of appropriate transporter protein (TAP) and functional EBV antigen presentation has been shown in vitro for HD cell lines.81 The expression pattern of the surface antigens of H-RS cells and their function thus may well be compatible with an antigenpresenting function.
CONCLUSIONS More than 150 years after the first description of Hodgkin's disease there are still more open questions than answers. Micromanipulation of H-RS cells from primary tissue material and subsequent single-cell PCR now makes it possible to characterize these cells specifically at the molecular level. First results are the demonstration of their monoclonal B cell origin, at least in the majority of cases analysed, as well as the demonstration of the H-RS cell origin of a novel cell line (L1236). Further applications of this new technical approach may be the detection of minimal residual disease after therapy, the analysis of H-RS cell-specific genetic markers and the cloning of H-RS cell-specific immunoglobulin gene idiotypes for idiotype vaccination. The ability to fractionate viable H-RS cells from fresh tissue by CD 30based magnetic sorting may now become the basis of further advances.82 The pronounced T cell reaction in affected tissue is still not understood. Is this observation caused by the deregulated cytokine expression of the H-RS cells or does it represent a specific, but inefficient immune reaction against a still-undefined target antigen expressed on H-S cells? Is the inability of the immune system to eliminate the H-RS cells caused by an alteration of the immune system itself or by the stepwise transformation of the H-RS cells resulting in an immune escape mechanism? Is the putative target antigen represented by a virally encoded protein (e.g. LMP) or by a cellular protein (autoimmune disease)? The association between infection with human papillomaviruses and cervical carcinoma, as well as between infection with hepatitis B virus and hepatocellular carcinoma demonstrates that initial infective genesis of a disease and malignant progression do not exclude each other. Hodgkin's disease might be understood as the unsuccessful attempt of the organism to eliminate a growth-deregulated cell expressing a still-undefined (cellular or viral) target antigen. In the course of the disease the inability of the immune system to eliminate this antigen expressing cell coincides with a stepwise transformation of the cell, most probably due to an inherited genetic instability. This process might then lead to aggressive, therapy-refractary Hodgkin's disease in the late stages of the disease (Fig. 10.1).
112 Hodgkin's disease
Figure 10.1 A possible pathway for viral genesis of Germinal center B cell Early stage: Pronounced T cell response without elimination of H-RS cells
Hodgkin's disease and malignant progression.
HD-specific (viral) antigen? Immune defect?
Advanced stage/relapse: Increasing malignant potential of H-RS cells Genetic instability Inactivation of suppressor genes? Activation of oncogenes? Deregulated cytokine/cytokine receptor interaction?
REFERENCES 1. Kaplan HS. In: Hodgkin's disease, 2nd edn. Cambridge, MA: Harvard University Press. 1980: 52-115. 2. Drexler HG. Recent results on the biology of Hodgkin and Reed-Stern berg cells. I. Biopsy material. Leuk Lymphoma 1992; 8: 283. 3. Haluska FG, Brufsky AM, Canellos GP. The cellular biology of the Reed-Stern berg cell. Blood 1994; 84:1005. 4. Braeuninger A, Kiippers R, Strickler JG, el al. Hodgkin-Reed Sternberg cells in lymphocyte predominant Hodgkin's disease represent clonal populations of germinal center-derived tumor cells. Proc NatlAcad Sci USA (in press). 5. O'Connor NTJ, Crick JA, Gatter KC, et al. Cell lineage in Hodgkin's disease. Lancet 1987; 1:158. 6. Sundeen J, Lipford E, Uppenkamp M, et al. Rearranged antigen receptor genes in Hodgkin's disease. Blood 1987; 70: 96. 7. Kiippers R, Rajewsky K, Zhao M, et al. Hodgkin disease:
derived from (crippled) germinal center B cells.) Exp Med 1996; 184:1495. 11. Hummel M, Ziemann K, Lammert H, etal. Hodgkin's disease with monoclonal and polydonal populations of Reed-Stern berg cells. N EnglJ Med 1995; 333: 901. 12. Hummel M, Marafioti T, Stein H. Immunglobulin V genes in Reed-Sternberg cells (letter to the editor). N EnglJ Med 1996; 334: 405. 13. Delabie J, Tierens A, Wu G, et al. Lymphocyte predominance Hodgkin's disease: lineage and clonality determination using a single-cell assay. Blood 1994; 84: 3291. 14. Delabie J, Tierens A, Gavrilli T, et al. Phenotype, genotype and clonality of Reed-Sternberg cells in nodular sclerosis Hodgkin's disease: results of a single-cell study. BrJ Haematol 1996; 94:198. 15. Ohno T, Stribley JA, Wu, et al. Clonality in nodular lymphocyte predominant Hodgkin's disease. N EnglJ Med 1997; 337: 459.
Hodgkin and Reed-Stern berg cells picked from histological sections show clonal immunoglobulin gene rearrangements and appear to be derived from B cells at
16. Marafioti T, Hummel M, Agnastopoulos I, et al. Origin of nodular lymphocyte predominant Hodgkin's disease from a clonal expansion of highly mutated germinal center B cells. N EnglJ Med 1997; 337: 453.
various stages of development. Proc Natl Acad Sci USA
17. Jox A, Zander T, Diehl V, Wolf J. Clonal relapse in Hodkin's
1994;91:10962. 8. Triimper L, Brady G, Bagg A, et al. Single-cell analysis of
disease (letter to the editor). N EnglJ Med 1997; 337: 449. 18. Vockerodt M, Soares M, Kanzler H, et al. Detection of
Hodgkin and Reed-Stern berg cells: molecular
clonal Hodgkin and Reed-Sternberg cells with identical
heterogeneity of gene expression and p53 mutations.
somatically mutated and rearranged VH genes in
Blood 1993; 81: 3097. 9. Kanzler H, Hansmann ML, Kapp U, et al. Molecular single cell analysis formally demonstrates the derivation of a peripheral blood-derived cell line (L1236) from the
different biopsies in relapsed Hodgkin's disease. Blood 1998; 92: 2899-907. 19. Brousset P, Schlaifer D, Meggetto F, et al. Persistence of
Hodgkin/Reed-Sternberg cells of a Hodgkin's lymphoma patient. Blood 1996; 87: 3429. 10. Kanzler H, Kuppers R, Hansmann ML, Rajewsky K.
the same viral strain in early and late relapse of Epstein-Barr virus associated Hodgkin's disease. Blood 1994; 8: 2447. 20. Inghirami G, Macri L, Rosati S, et al. The Reed-Sternberg
Hodgkin and Reed Sternberg cells in Hodgkin's disease
cells of Hodgkin's disease are clonal. Proc Natl Acad Sci
represent the outgrowth of a dominant tumor clone
USA 1994; 91: 9842.
References 113
21. Izban KF, Nawrocki JF, Alkan S, Hsi ED. Monoclonal IgH gene rearrangement in microdissected nodules from nodular sclerosis Hodgkin disease. AmJ Clin Pathol 1998; 110:599-606. 22. Brousset P, Chaouche N, Al Saati T, Zenou RC, Delsol G.
37. Drexler HG: Recent results on the biology of Hodgkin and Reed-Stern berg cells. II. Continous cell lines. Leuk Lymphoma 1993; 9:1. 38. Wolf J, Kapp U, Bohlen H, etal. Peripheral blood mononuclear cells of a patient with advanced Hodgkin's
Telomerase activity in Hodgkin's disease. Leak Lymphoma 1998;30:189-92.
lymphoma give rise to permanently growing Hodgkin-Reed Stern berg cells. Blood 1996; 87: 3418.
23. Norrback KF, Enblad G, Erlanson M, Sundstrom C, Roos G.
39. Von Kalle C, Wolf J, Becker A, et al. Growth of Hodgkin
Telomerase activity in Hodgkin's disease. Blood 1998; 92: 567-73. 24. Orscheschek K, Merz H, Hell J, et al. Large cell anaplastic lymphoma-specifictranslocation (t(2;5)(p23;q35)) in Hodgkin's disease: indication of a common pathogenesis? Lancet 1995; 345: 87-90. 25. Elmberger PG, Lozano MD, Weisenburger DD, et al. Transcripts of the npm-alk fusion gene in anaplastic large cell lymphoma, Hodgkin's disease, and reactive lymphoid lesions. Blood 1995; 86: 3517. 26. Herbst H, Anagnostoupoulos, Heinze B, et al. ALK gene products in anaplastic large cell lymphomas and Hodgkin's disease. Blood 1995; 86:1694. 27. Wellmann A, Otsuki T, Vogelbruch M, et al. Analysis of the t(2;5)(p23;q35) translocation by reverse transcription-
cell lines in severely combined immunodeficient mice. IntJ Cancer 1992; 52:887. 40. Jarrett RF, Gallagher A, Jones DB, etal. Detection of EBV genomes in Hodgkin's disease: association with age. 7 C/mP<7?/;o/1991;44:844. 41. Bargou RC, Mapara MY, Zugck C, etal. Characterization of a novel Hodgkin cell line, HD-MyZ, with monocytoid features mimicking Hodgkin's disease in severe combined immunodeficient mice.) Exp Med 1993; 177: 1257. 42. Kapp U, Dux A, Schell-Frederick E, et al. Disseminated growth of Hodgkin derived cell lines L540 and L540cy in immune deficient SCID mice. Ann Oncol 1994; 5 (suppl 1): 121. 43. Winkler U, Gottstein C, Schbn G, et al. Successful
polymerase chain reaction in CD30+ anaplastic large-cell
treatment of disseminated human Hodgkin's disease in
lymphomas, in other non-Hodgkin's lymphomas of T-cell phenotype, and in Hodgkin's disease. Blood 1995; 86: 2232.
SCID mice with deglycosylated ricin A-chain immunotoxins. Blood 1994; 83: 466.
28. Fonatsch C, Diehl V, Schaadt M, et al. Cytogenetic investigations in Hodgkin's disease: I. Involvement of specific chromosomes in marker formation. Cancer Genet Cytogenet1986;2Q:39.
44. Hombach A, Jung W, Pohl C, et al. A CD16/CD30 bispecific monoclonal antibody induces lysis of Hodgkin cells by unstimulated natural killer cells in vitro and in vivo. IntJ
29. Rowley JD. Chromosomes in Hodgkin's disease. Cancer Treat Rep 1982; 66: 639.
Cancer 1993; 55: 830. 45. Correa P, O'Connor GT. Epidemiologic patterns of Hodgkin's disease. IntJ Cancer 1971; 8:192. 46. Gutensohn N, Cole P. Epidemiology of Hodgkin's disease.
30. Atkin NB. Cytogenetics of Hodgkin's disease. Cytogenet Cell Genet 1998; 80: 23-7.
Semin Oncol *\98Q; 7: 92. 47. Gutensohn N, Cole P. Childhood social environment and
31. Tilly H, Bastard C, Delastre T, et al. Cytogenetic studies in untreated Hodgkin's disease. Blood 1991; 77:1298. 32. Kapp U, Wolf J, Hummel M, et al. Hodgkin's lymphoma
Hodgkin's disease. N EnglJ Med 1981; 304:135. 48. MacMahon B. Epidemiology of Hodgkin's disease. Cancer Res 1966; 26:1189.
derived tissue serially transplanted into severe
49. Mueller N, Evans A, Harris NL, et al. Hodgkin's disease
combined immunodeficient (SCID)-mice. Blood 1993; 82: 1247.
and Epstein-Barr virus. Altered antibody a pattern before diagnosis. N EnglJ Med 1989; 32: 689. 50. Weiss LM, Movahed LA, Warnke RA, etal. Detection of Epstein-Barr viral genomes in Reed-Stern berg cells of
33. Jansen MP, Hopman AH, Haesevoets AM, etal. Chromosomal abnormalities in Hodgkin's disease are not restricted to Hodgkin/Reed-Sternberg cells.y Pathol 1998; 185:145-52. 34. Ohno T, Smir BN, Weisenburger DD, Gascoyne RD, Hinrichs SD, Chan WC. Origin of the Hodgkin/Reed-Sternberg cells in chronic lymphocytic leukemia with 'Hodgkin's transformation'. Blood 1998; 91:1757-61. 35. Weber-Matthiesen K, DeerbergJ, Poetsch M, etal. Numerical chromosome aberrations are present within the CD30+ Hodgkin and Reed-Stern berg cells in 100% of analyzed cases of Hodgkin's disease. Blood 1995; 86: 1464. 36. Diehl V, von Kalle C, Fonatsch C, et al. The cell of origin of Hodgkin's disease. Semin Oncol 1990; 17: 660.
Hodgkin's disease. N EnglJ Med 1989; 320: 502. 51. Glickman JN, Howe JG, Steitz JA. Structural analysis of EBER1 and EBER2 ribonucleoprotein particles present in Epstein-Barr virus-infected cells. J V/Vo/1988; 52: 902. 52. Herbst H, Steinbrecher E, Niedobitek G, et al. Distribution and phenotype of Epstein-Barr virus-harboring cells in Hodgkin's disease. Blood 1992; 80: 484. 53. Weiss LM, Chen YY, Liu XF, et al. Epstein-Barr virus and Hodgkin's disease. A correlative in situ hybridization and polymerase chain reaction study. AmJ Pathol 1991; 139: 1259. 54. Wu TC, Mann RB, Charache P, etal. Detection of EBV gene expression in Reed-Stern berg cells of Hodgkin's disease. IntJ Cancer 1990; 46: 801.
114 Hodgkin's disease 55. Dolcetti R, Boiocchi M. Epstein-Barr virus in the pathogenesis of Hodgkin's disease. Biomed Pharmacother 1998; 52:13-25. 56. Liu SM, Chow KC, Chiu CF, Tzeng CH. Expression of Epstein-Barr virus in patients with Hodgkin's disease in Taiwan. Cancer 1998; 83: 367-71. 57. Herbst H, Dallenbach F, Hummel M, etal. Epstein-Barr virus latent membrane protein expression in Hodgkin and Reed-Stern berg cells. Proc Natl Acad Scl USA 1991; 88: 4766. 58. Young LS, Rowe M. Epstein-Barr virus, lymphomas and Hodgkin's disease. Semin Cancer Biol 1992; 3: 273. 59. Klein G. Viral latency and transformation: the strategy of Epstein-Barr virus. Cell 1989; 58: 5. 60. Wang D, Liebowitz D, Kieff E. An EBV membrane protein expressed in immortalized lymphocytes transforms established rodent cells. Cell 1985; 43: 831. 61. Gregory CD, Dive C, Henderson S, et al. Activation of Epstein-Barr virus latent genes protects human B cells from death by apoptosis. Nature 1991; 349: 612. 62. Diehl V, Bohlen H, Wolf J. CD30: Cytokine-receptor, differentiation marker or a target molecule for specific immune response? Ann Oncol 1994; 5: 300. 63. Knecht H, Bachmann E, Brousset P, etal. Deletions within the LMP1 oncogene of Epstein-Barr virus are clustered in Hodgkin's disease and identical to those observed in nasopharyngeal carcinoma. Blood 1993; 82: 2937. 64. Wolf J, Jox A, Skarbek H, et al. Selective loss of integrated Epstein Barr virus genomes after long term cultivation of Burkitt's lymphoma x B-lymphoblastoid cell hybrids due to an increased chromatin instability at the integration site. Virology 1995; 212:179-85. 65. Gause A, Roscshansky V, Tschiersch A, et al. Low serum interleukin-2 receptor levels correlate with a good prognosis in patients with Hodgkin's lymphoma. Ann Onto/1991; 2: 43. 66. Warzocha K, Bienvenu J, Ribeiro P, et al. Plasma levels of tumour necrosis factor and its soluble receptors correlate with clinical features and outcome of Hodgkin's disease patients. BrJ Cancer 1998; 77: 2357-62. 67. Samoszuk M, Nansen L. Detection of interleukin-5 messenger RNA in Reed-Stern berg cells of Hodgkin's disease with eosinophiliea. Blood 1990; 75:13. 68. Kadin M, Agnarrson B, Ellingswoth L, etal. Immunohistochemical evidence of a role for transforming growth factor beta in the pathogenesis of
69.
70.
71.
72.
73.
74.
75.
76. 77.
78.
79. 80.
81.
82.
nodular sclerosing Hodgkin's disease. AmJPathol 1990; 136:1209. Xerri L, Birg F, Giugou V, et al. In situ expression of the IL-1-alpha and TNF-alpha genes by Reed-Stern berg eel Is in Hodgkin's disease. Int J Cancer 1992; 50: 689. Gruss HJ, Brach M, Drexler HG, et al. Interleukin 9 is expressed by primary and cultured Hodgkin and Reed-Stern berg cells. Cancer Res 1992; 52:1026. Jiicker M, Abts H, Ki W, etal. Expression of interleukin-6 and interleukin-6 receptor in Hodgkin's disease. Blood 1991; 77: 2413. Merz H, Houssiau F, Orscheschek K, etal. lnterleukin-9 expression in human malignant lymphomas: unique association with Hodgkin's disease and large cell anaplastic lymphoma. Blood 1991; 78:1311. Durkop H, Latza U, Hummel M, et al. Molecular cloning and expression of a new member of the nerve growth receptor family that is characteristic for Hodgkin's disease. Cell 1992; 69: 421. Smith CA, Gruss HJ, Davis T, et al. CD30 antigen, a marker for Hodgkin's lymphoma, is a receptor whose ligand defines an emerging family of cytokines with homology toTNF.Ce//1993; 73:1349. Gruss HJ, Dower SK. Tumor necrosis factor ligand superfamily: involvement in the pathology of malignant lymphomas. Blood 1995; 85: 3378. Poppema S, Kaleta J, Hepperle B, et al. Biology of Hodgkin's disease. Ann Oncol 1992; 3(suppl 4): 5. Pinto A, Gattei V, Zagonel V et al. Hodgkin's disease: a disorder of dysregulated cellular cross-talk. Biotherapy 1998; 10: 309-20. Peckham MJ, Cooper EH. The cell proliferation characteristic of the various classes of cells in Hodgkin's disease. Cancer 1969; 24:135. Wolf J, Diehl V. Is Hodgkin's disease an infectious disease? Ann Onco/1994; 5(suppl 1): 105. Delabie J, Ceuppens JC, Vandenberghe P, et al. The B7 antigen is expressed by Reed/Sternberg cells of Hodgkin's disease and contributes to the stimulating capacity of Hodgkin derived cell lines. Blood 1993; 82: 2845. Lee SP, Constandinou CM, Thomas WA, et al. Antigen presenting phenotype of Hodgkin Reed-Stern berg cells: analysis of the HLA class I processing pathway and the effects of interleukin-10 on Epstein-Barr virus-specific cytotoxic T-cell recognition. Blood 1998; 92:1020-30. Irsch J, Nitsch S, Hansmann ML, et al. Isolation of viable Hodgkin and Reed-Stern berg cells from Hodgkin disease tissues. Proc Natl Acad Sci USA 1998; 95:10117-22.
11 Viruses and malignant lymphoma LM WEISS AND KL CHANG
Introduction Human T cell leukemia type 1 Kaposi's sarcoma-associated herpes virus Epstein-Barr virus Burkitt's lymphoma B cell lymphomas in immunodeficient patients
115 115 116 117 118 119
INTRODUCTION Over the past several decades, there have been great efforts to implicate several viruses in the etiology of malignant lymphoma. Although there is abundant evidence to indicate that human T cell leukemia virus type 1 (HTLV-1) has an etiologic role in adult T cell leukemia/lymphoma (ATLL), no other viruses have yet been shown to be causative in any of the other malignant lymphomas. Nevertheless, a great deal of data have accumulated over the last several decades to indicate an association of the Epstein-Barr virus (EBV) with the malignant lymphomas including Hodgkin's disease. It is likely that EBV plays an important role in the pathogenesis of a significant subset of these lymphomas. In addition to HTLV-1 and EBV, this chapter will discuss other viruses that may also play a role in the etiology of malignant lymphomas, including the recently discovered Kaposi's sarcoma-associated herpes virus (KSHV, or HHV-8) human herpes virus type 6 (HHV-6) and hepatitis C. A useful additional review is given by Lyons and Liebowitz.1 HUMAN T CELL LEUKEMIA TYPE 1 HTLV-1 is a retrovirus that was first isolated from two patients who had what were thought to be aggressive forms of mycosis fungoides/Sezary syndrome.2'3 The HTLV-1 has been entirely sequenced and shows much more complexity than many of the animal retroviruses.4 In general, an intact retrovirus consists of a diploid RNA
Other B cell lymphomas T/natural killer cell lymphomas Hodgkin's disease Human herpes virus 6 Hepatitis C References
120 121 122 123 124 124
genome covered by a capsid and outer envelope. When the retrovirus enters the cytoplasm of a cell, the RNA is converted into DNA by reverse transcriptase and the DNA integrates into the host genome as a linear provirus. The point of integration is not fixed and varies in different cases,5 but is monoclonal within a given case. The HTLV-1 provirus consists of long terminal repetitive sequences at each end, with gag (coding for core structural protein), pol (coding for DNA polymerase) and env (coding for envelope glycoproteins) genes followed by a long sequence called the X gene region. The X gene includes at least three other genes, including the tax gene, which codes for a 40 kDa protein. The exact mechanism of how HTLV-1 causes lymphoma is not yet known. However, HTLV-1 has been shown to infect and immortalize CD 4-positive T cells in vitro.6 HTLV-1 differs from acutely transforming retroviruses in that it lacks its own viral oncogene and it also differs from chronic leukemogenic retroviruses in that transformation does not depend on integration adjacent to a cellular oncogene. The tax gene product has been shown to trans-activate the long terminal repeats (LTR), which can act as a viral promoter.7 In addition, the tax protein can trans-activate a variety of cellular genes, including the interleukin-2 receptor alpha gene, c-fos, c-jun and the gene encoding parathyroid hormonerelated protein.8 Regardless of the precise lymphomagenic mechanism of HTLV-1, there is a large body of epidemiologic, serologic and molecular biological evidence supporting an etiologic role for the virus in ATLL, an aggressive T cell neoplasm with characteristic clinical features (Fig. 11.1).
116 Viruses and malignant lymphoma
role for HTLV-1 in a subset of cases of mycosis fungoides/Sezary syndrome has been postulated by some.14 Although these patients are usually seronegative for HTLV-1 and HTLV-2, evidence of the HTLV-1 pol and/or tax. gene has been found by polymerase chain reaction (PCR) in lesional tissues or peripheral blood mononuclear cells from mycosis fungoides patients in 0-68 per cent of cases. Overall, about 15 per cent of cases have been positive.15'20 In addition, Hall and colleagues found the presence of monoclonally integrated HTLV-1 provirus with a 5.5 kb deletion in one case.19 In cases shown to contain the tax or pol gene by PCR, there is no evidence of virus capsid p24 protein or reverse transcriptase activity, leading to the hypothesis that the proviruses may be defective.20 Finally, one study of CD 30-positive cutaneous large cell lymphoma has revealed evidence of HTLV-1 by Southern blotting or PCR in six of ten patients with CD 30-positive anaplastic large cell lymphoma. Again, none of the samples contained a full-length proviral DNA, suggesting that the proviruses may contain deletions.21 The demonstration of the role of HTLV-1 in human lymphoma is now the basis of new approaches to therapy.22"24
Figure 11.1 Adult T cell leukemia/lymphoma from Japan. These cases may be indistinguishable from peripheral Tcell lymphomas not associated with HTLV-1, but the former are generally more pleomorphic.
Most cases of ATLL occur in only a few regions of the world, most notably in southwestern Japan, but also in the Caribbean, southeastern USA, equatorial Africa and southern Italy.9-11 These areas of high incidence of ATLL have also been shown to have a high prevalence of HTLV-1 infection, as determined by serological studies. In addition, all patients with ATL have elevated levels of antibodies against HTLV-1. Finally, direct molecular evidence of HTLV-1 can be identified in all cases of ATLL. Antibodies to HTLV-1 have been detected in over one million people. However, only approximately 1000 new cases occur annually; therefore, the large majority of HTLV-1 carriers do not develop lymphoma.9 Approximately one in every 900 male carriers and one in every 2000 female carriers older than 40 years are estimated to develop ATLL each year.12 HTLV-1 may be transmitted by sexual intercourse from male to female, by blood transfusion, by intravenous drug abuse or by breast feeding.10 Human T lymphotropic virus-like particles have been found by ultrastructural analysis of cultures of peripheral blood lymphocytes from 90 per cent of patients with mycosis fungoides, raising the possibility that a high percentage of cases may be associated with retroviruses.13 A
KAPOSI'S SARCOMA-ASSOCIATED HERPES VIRUS In 1994, evidence was reported for a viral etiology for Kaposi's sarcoma.25 Preliminary sequencing data suggest that this candidate virus, called Kaposi's sarcomaassociated herpes virus (KSHV) or human herpes virus-8 (HHV-8), has partial homology to two gammaherpesviruses, the EBV and the animal virus herpesvirus saimiri. The latter virus usually infects squirrel monkeys but also causes fulminant T cell lymphomas in New World primates other than its natural hosts, and also can transform human T lymphocytes. Cesarman and colleagues initially identified KSHV sequences in eight of 193 lymphomas.26 The eight lymphomas occurred in human immunodeficiency virus (HIV)-infected patients, were predominantly body cavity-based, and were characterized by pleural, pericardial or peritoneal lymphomatous effusions (Fig. 11.2). All eight lymphomas were CD 45 (panleukocyte) positive and showed variable expression of activationassociated markers, but, with the exception of one case, lacked B lineage, T lineage, myeloid and monocyte markers. However, all eight lymphomas had clonal immunoglobulin heavy-chain gene rearrangements, consistent with a B cell lineage. Interestingly, all eight lymphomas were also EBV positive. Since that initial report, KSHV sequences have been identified in rare body cavity-based lymphomas in nonHlV-infected patients.27 These studies have helped define
Epstein-Barr virus 117
EPSTEIN-BARR VIRUS
Figure 11.2
Body cavity-based AIDS-associated malignant
lymphoma. Note the marked pleomorphism. This case contained both KSHVand EBV. Case courtesy of Dr Ethel Cesarman, Cornell University Medical College.
a new clinicopathological entity, termed primary effusion lymphoma, as a distinctive type of lymphoma associated with KSHV.28 This new lymphoma can be distinguished from small non-cleaved cell (Burkitt-like) lymphoma involving body cavities because the latter lymphoma is associated with c-myc gene rearrangements and is not associated with KSHV. Primary effusion lymphomas can also be distinguished from pyothoraxassociated lymphomas that may develop after longstanding inflammation; the latter are usually associated with a tumor mass, and lack KSHV sequences.29'30 KSHV sequences have also been consistently identified in multicentric Castleman's disease, both HIV- and nonHIV-associated.31,32 Early studies of KSHV identified the presence of two potential oncogenes, including a cellular-type cyclin D similar to the PRAD 1 oncogene involved in mantle cell lymphoma, a gene homologous to the cellular G proteincoupled receptor family of proteins,33 as well as a functional bd-2 homologue.34-36 KSHV-infected cell lines are now available and KSHV-infected cells from primary effusion lymphoma have been grown in BNX tripleimmunodeficient mice, with the development of malignant ascites as well as solid tumor masses.37,38
Epstein-Barr virus is a double-stranded DNA virus and is a member of the gammaherpesvirus group. It consists of approximately 172 kilobase pairs, making it the largest human tumor virus. An association between EBV and human neoplasia was suspected since the virus was first identified in electron micrographs of a cell line derived from a case of African Burkitt's lymphoma.39 The virus is endemic world-wide. It is found universally in developing countries and is present in 90-95 per cent of individuals in developed countries, casting doubt on any significant role in the etiology or pathogenesis of neoplasms. However, interest in EBV has been rekindled in the past decade, with the demonstration of a significant association between EBV and several types of malignant lymphomas, including B cell lymphomas, T cell lymphomas and Hodgkin's disease. When EBV infection occurs in early childhood, it is usually asymptomatic. When EBV infection occurs at a later time, it results in a benign, self-limited lymphoproliferation, which is clinically recognized as acute infectious mononucleosis. The virus causes a lytic (active) infection of epithelial cells and lymphocytes that is rapidly converted to a latent infection of lymphocytes once the individual mounts an immune response. In the lytic phase, the virus has a linear form, and expresses most of its genes (approximately 100 genes), including the early antigens and components of the viral capsid and envelope. When the infection converts to the latent phase, the genome circularizes into an episome by fusion of two terminal repeat regions; a small proportion of the genomes may also become randomly integrated into the host DNA. In the latent phase, only a few genes are expressed. These include genes encoding six nuclear antigens, including Epstein-Barr nuclear antigen (EBNA)-l, EBNA-2, EBNA-3A, EBNA-3B and EBNA-3C, and leader protein, and three membrane proteins, including latent membrane protein (LMP)-l, LMP-2A and LMP-2B. In addition, two non-translated small RNAs, Epstein-Barr early RNA (EBER)-l and EBER-2, are transcribed. EBNA-1 binds to specific genomic sequences and is essential for the establishment and maintenance of EBV latency. EBNA-2 is needed for transformation, but is not essential for the maintenance of EBV latent infection. EBNA-2 also can upregulate several host and viral genes, including LMP-1 and the activation antigen CD 23 in B cells. EBNA-3A and 3C are also required for immortalization. LMP-1 acts as an oncogene in transfection studies, and causes upregulation of a variety of cellular genes. In B cells, it induces CD 23, the adhesion molecules LFA-1 and LFA-3, and CD 54, interleukin-10 and bd-2. Induction of bd-2 may prevent EBV-infected cells from undergoing programmed cell death. A characteristic 30 bp deletion near the 3' end of the LMP-1 gene occasionally may be found, and in some studies, the LMP
118 Viruses and malignant lymphoma
deletion variant has been found to be associated with more clinically aggressive behavior: LMP-1 transgenic mice develop lymphomas.40 However, the function of the EBERs is not yet known. However, because they are produced in great abundance (up to 107 copies per cell, in all latently infected cells), they are an attractive target for detection by molecular methods. Not all latently infected cells express all of the latent gene proteins. Three general patterns of EBV latency gene expression are recognized. In latency pattern I, only EBNA-1 is expressed. Since EBNA-1 is the only latency antigen that does not elicit a cytotoxic T cell response, this latency pattern may provide a useful form of 'immune escape' in hosts that have a relatively intact immune system. An example of this latency pattern is found in cases of EBV-associated Burkitt's lymphoma. In latency pattern II, LMP-1, LMP-2A and LMP-2B are expressed in addition to EBNA-1. An example of this latency pattern is found in EBV-associated cases of Hodgkin's disease. Finally, in latency pattern III, all of the EBNAs and LMPs are expressed. In this latency pattern, the promoter used for EBNA-1 is different from that used in latency patterns I and II. This latency pattern may be preferentially used in EBV-infected cells in patients lacking a competent immune system, since immunocompetent individuals might be expected to mount an immune response to and eliminate cells expressing multiple EBNAs and LMPs. An example of this latency pattern is found in many cases of EBV-associated cases of post-transplantation lymphoproliferative disorder. Latency pattern III has also been identified in most EBV-infected lymphoblastoid cell lines, where the phenomenon of EBV latency has been best studied. There are two types of EBV (types A and B, or types 1 and 2), which differ in their EBNA-2 and EBNA-3A-C gene sequences.41 Type A EBV is globally ubiquitous, can readily transform B cells, and is the type usually identified in the oropharynx and peripheral blood of healthy seropositive individuals. Type B EBV has a more restricted geographic distribution (most commonly found in equatorial Africa), does not readily transform B cells and is generally identified in peripheral blood only from immunocompromised patients.42 When evaluating studies putatively identifying a significant association between EBV and a specific neoplasm, one needs to analyse the EBV detection methods critically. Many individuals harbor latent EBV and so, demonstrating EBV exposure by serologic studies or demonstrating EBV within involved tissues by PCR is not sufficient evidence for a true association. One must clearly show that the EBV is actually present in the tumor cells, and that EBV infection is a relatively early event and not merely a superinfection of an already established neoplasm. Currently, the best method for detecting EBV within neoplastic cells is in situ hybridization studies using probes directed against the highly abundant EBERs.43 The EBER in situ hybridization
technique is highly sensitive and highly practical, as it is easy to perform in routinely processed paraffinembedded sections. The demonstration of EBV in all or nearly all tumor cells is indirect evidence that infection occurred early in the neoplastic process, while the presence of EBV in only a subset of the cells implies little causative role or only a role in disease progression. Paraffin section immunohistochemistry for LMP-1 antigen is also practical,44 but only useful in neoplasms with latency patterns II and III, since latency pattern I neoplasms fail to express LMP-1. However, monoclonal antibodies against EBNA-1 that react in paraffin sections have recently become available.45 Southern blotting hybridization studies are not useful for determining which individual cells harbor EBV but may be used to determine the clonality of the EBVpositive cells.46 As mentioned above, the linear EBV genome has highly polymorphic terminal repeat regions at each of its ends that fuse to form the episomal genome. These terminal repeat regions consist of variable numbers of tandemly repeated units approximately 500 bp in length. In the episomal form, the number of repeats is stable over time as well as among progeny virus derived from an EBV-infected cell. Thus, hybridization probes against the terminal repeat regions, and adjacent regions in Southern blotting studies can be used to determine whether a solitary terminal repeat region is present (indicating EBV in a clonal proliferation of cells) or whether multiple terminal repeat regions are present (indicating EBV in a polyclonal population). In addition, by determining the molecular weight of the terminal repeat region, one can distinguish a latent infection (longer fragments due to fusion of the two terminal repeat regions) from a lytic infection (smaller fragments).46'47 PCR is too sensitive to use to determine whether EBV is actually associated with a tumor rather than merely present in host cells, but it can be used to type the EBV strain as well as to analyse the LMP-1 gene for the characteristic 30-bp LMP-1 deletion.48,49
BURKITT'S LYMPHOMA Burkitt's lymphoma was the first malignancy shown to be significantly associated with EBV. Approximately 95 per cent of cases of Burkitt's lymphoma in equatorial Africa, a region of high incidence of Burkitt's lymphoma (endemic Burkitt's lymphoma), are associated with EBV.50 A low percentage of EBV positivity, approximately 5-20 per cent, is seen in Burkitt's lymphoma from Western countries.51'52 An intermediate rate is seen in Burkitt's lymphoma cases from South America and Asia;51,56 however, there may be an endemic region of EBV-associated Burkitt's lymphoma in Bahia, a tropical region of Northeast Brazil that is populated predominantly by Mestizoes, individuals with mixed European,
B cell lymphomas in immunodeficient patients 119
African, and/or Indian ancestry.57 In the EBV-positive cases, the virus is detected in all or virtually all of the tumor cells by in situ hybridization (Fig. 11.3). PCR studies have demonstrated endemic cases to contain either type A or type B EBV, while sporadic cases are almost always type A. Southern blotting studies using probes to the EBV termini have shown that the EBV is present in a monoclonal population. In Brazilian Burkitt's lymphoma, a high incidence of the characteristic 30-bp LMP-1 gene deletion has been reported, similar to the high incidence of the deletion found in EBV-derived non-neoplastic tissues in this population.58 These data are in contrast to another study which failed to identify the same IMP-1 gene deletion in eight EBVassociated cases of Burkitt's lymphoma occurring in Pakistan.59 Despite many years of investigation, the precise role of the virus in the etiology and pathogenesis of Burkitt's lymphoma is not known. Endemic Burkitt's lymphoma was observed early on to occur in areas endemic for malaria, suggesting that chronic immunosuppression induced by chronic malarial infection predisposed an individual to EBV-associated Burkitt's lymphoma. However, a high association with EBV is also found in
Figure 11.3 EBV-positive Burkitt's lymphoma. This EBER RNA in situ hybridization study labels all of the Burkitt cell nuclei positively. The non-staining 'holes' represent the tingible-body macrophages, which fail to demonstrate evidence of EBV.
Burkitt's lymphoma occurring in Egypt, an area outside the malarial belt, suggesting that EBV positivity may correlate more closely with overall socioeconomic status.60 Since EBV can be identified only in a subset of cases of Burkitt's lymphoma, particularly in the sporadic form in Western countries, it is clear that EBV must be of much lesser importance than the nearly constant chromosomal translocations involving the c-myc gene on chromosome 8. There is no correlation between the presence of EBV and the specific breakpoint location at c-myc, providing evidence against a direct role for EBV in the induction of the translocation. Perhaps infection of B lymphocytes by EBV results in stimulation and a polyclonal proliferation, resulting in an increased pool of B cells, each with an increased probability of chromosomal translocation.
B CELL LYMPHOMAS IN IMMUNODEFICIENT PATIENTS Immunocompromised patients have an increased risk for developing a variety of malignant neoplasms, most notably non-Hodgkin's lymphomas. These lymphomas are usually of B lineage. Approximately 90-95 per cent of post-transplantation lymphoproliferations are associated with EBV.61-63 In these cases, DNA in situ hybridization studies have demonstrated EBV in all or virtually all of the tumor cells, generally in large amounts.63 In many cases, a heterogeneous signal is seen from cell to cell, suggesting a lytic infection superimposed on a latent infection. This interpretation is supported by in situ hybridization studies using probes detecting EBV genes expressed in lytic infections and immunohistochemical studies for the BZLF protein, which is involved in the switch from a latent to a lytic infection.64 In situ hybridization studies indicate that increased numbers of EBV-positive cells are often present prior to the development of lymphoma.65 PCR studies have demonstrated that EBV type A, and not type B, is almost always present in post-transplant lymphoproliferations,49 despite the fact that peripheral blood lymphocytes maybe infected with either type.66 Southern blot studies using probes to the EBV termini generally parallel similar studies of the immunoglobulin heavy- and lightchain genes, demonstrating that EBV is present in a polyclonal population in some cases (particularly those cases at the benign end of the histologic spectrum), an oligoclonal population in other cases and a monoclonal population in many cases.47,61,62,67 Different monoclonal populations may be present at different sites of disease within the same patient, again similar to Southern blot studies of the immunoglobulin heavy- and light-chain genes, suggesting that these lymphomas are 'multiclonal'.61'67 In addition, the proliferating cells of monomorphous lymphomas usually contain oncogene alterations in addition to EBV.62
120 Viruses and malignant lymphoma
These studies suggest a important role for EBV in the pathogenesis of immunodeficiency-related lymphoproliferations. In the setting of immunosuppression, greater than normal numbers of EBV-infected lymphocytes are present. In theory, the EBV drives a polyclonal lymphoproliferation, which progresses to an oligoclonal proliferation, followed by a monoclonal proliferation. The presence of lytically infected tumor cells may serve as an additional source of EBV, suggesting a possible clinical role for acyclovir, a drug active only in lytically infected cells. Early in the process, the proliferation may spontaneously reverse itself, if immunosuppression is reduced, giving the immune system a chance to control the EBVinfected cells. Eventually, additional oncogene alterations occur, leading to an autonomous and irreversible malignant lymphoma.68 Lymphomas may also occur in patients with collagen vascular disease who use immunosuppressive agents, such as azathioprine, cyclophosphamide, methotrexate, cyclosporin and steroids. In one study, EBV was found in one-third of lymphomas occurring in patients with connective tissue disease.69 There are anecdotal reports of spontaneous lymphoma regression following cessation of the immunosuppression in these patients.70 EBVassociated lymphomas have also been reported in patients with congenital immunodeficiencies, most commonly in patients with X-linked lymphoproliferative (Duncan's) syndrome, a congenital immunodeficiency in which patients have defective handling of EBV infection.71 Patients with Duncan's disease who survive their first exposure to EBV often subsequently develop an EBV-positive malignant lymphoma. Patients with HIV infection have abnormally high numbers of EBV-positive lymphocytes, both in their peripheral blood and their lymph nodes.72'74 PCR studies have identified EBV types A and B.42 Approximately 40-70 per cent of acquired immunodeficiency syndrome (AIDS)-associated lymphomas are associated with EBV.75-78 In contrast to post-transplantation lymphomas, either type A or type B EBV can be identified in individual cases.79 The presence or absence of EBV in the tumor correlates with the histologic subtype as well as the involved site.76,80 Large cell immunoblastic lymphoma is the most commonly EBV-associated AIDS tumor; approximately 80 per cent of these lymphomas are EBVpositive. A lower percentage of EBV positivity is seen in cases of Burkitt's lymphoma and large cell lymphoma, approximately 30 and 20 per cent, respectively. A high percentage of cases of AIDS-associated anaplastic large cell lymphomas have been reported to be EBV positive.81 Virtually all primary central nervous system lymphomas are EBV positive as are all body cavity-based lymphomas (the latter also contain KSHV), while node-based lymphomas have a much lower association with EBV.26,76,82,83 The EBV latency pattern also varies with the different histological types of lymphoma. Similar to endemic Burkitt's lymphoma, latency type I is most common in
the HIV-associated, small, non-cleaved lymphomas, while latency type III is most commonly seen in the large cell lymphomas.84 Latency type II, a pattern of latency otherwise only seen in Hodgkin's disease among EBVassociated lymphoid neoplasms, can be found in large cell and Burkitt-type lymphomas. Approximately onehalf of EBV-positive lymphomas contain evidence for a lytic infection, as detected by in situ hybridization using probes directed against the EBV NotI gene or immunohistochemical studies using antibodies against ZEBRA protein.85 Almost all AIDS-associated lymphomas have been reported to be monoclonal by immunoglobulin gene rearrangement or EBV termini studies.75'86 One group has reported a subset of cases with a polyclonal proliferation; in these cases, EBV was identified only occasionally.87 Another group reported three cases without detectable clonal immunoglobulin gene rearrangements; EBV was present in two of the cases and was found to be clonal.80 The role of EBV in the etiology of AIDS-associated lymphomas is less clear than its role in the post-transplant setting because a significant subset of AIDS-related lymphomas lack evidence of EBV. Nevertheless, the presence of EBV in about one-half of cases, and particularly its consistent identification in a monoclonal population, even in rare cases that lack detectable clonal immunoglobulin gene rearrangements,88 suggests a strong role in pathogenesis in those. EBV-positive cases. Understanding of the role of EBV in lymphoproliferation and lymphoma following transplant has led to novel approaches to immunomerapy using EBV-specific cytotoxic T lymphocytes (CTL) with some early encouraging results.89
OTHER B CELL LYMPHOMAS Most B cell lymphomas are not associated with EBV. In fact, less than 5 per cent of sporadic B cell lymphomas are EBV positive, particularly those with lymphoplasmacytic features, or those occurring in the stomach or head and neck region.90,91 There have been no reports of EBV positivity in cases of follicular lymphoma and the only cases of small lymphocytic lymphoma/chronic lymphocytic leukemia reported to be EBV-positive have been rare cases associated with Reed-Sternberg-like cells (see below).92 Hairy cell leukemia was once thought to be EBV associated, but the early results probably represented technical artefact, since EBER in situ hybridization studies have been uniformly negative.93 In contrast, the rare pyothorax-associated pleural B cell lymphoma, which occurs at the site of chronic inflammation and fibrosis 20-50 years after active tuberculosis, does have a significant and fairly uniform association with EBV.29,30 These cases show a type III latency pattern.94 Finally, lymphomatoid granulomatosis, a rare entity long
T/natural killer cell lymphomas 121
thought to be a T cell lymphoproliferation, is now thought to represent an EBV-positive monoclonal B cell lymphoma with a prominent T cell reaction.95
T/NATURAL KILLER CELL LYMPHOMAS Infection of B cells by EBV has been recognized for years. B cells represent the primary EBV-infected cell population in a latent infection, with the virus usually entering the cell via the CD 21 (C3d) complement receptor. However, EBV has recently been shown to also infect T cells in vitro96 and EBV-infected T cells have been shown to represent another reservoir of latent infection in seropositive individuals.97 What receptor the virus uses to enter T cells is not definitely known, although fetal T cells express CD 21 and even adult T cells may express CD 21 in low levels.98,99 EBV was first identified in T cell lymphoma in a small series of cases of peripheral T cell lymphoma arising in patients with a chronic mononucleosis-like syndrome.100 Since that report, EBV has been identified in 'sporadic' peripheral T cell lymphomas in apparently healthy individuals. Sporadic peripheral T cell lymphomas may, in fact, be more likely to show an EBV association than sporadic B cell lymphomas,88 especially in Asian or Latin populations.101-104 The detection of EBV in non-lymphoblastic T cell lymphoma has been correlated with a poor prognosis.105 Nasal T/natural killer (NK) cell lymphoma is the T/NK cell lymphoma most strongly associated with EBV (Fig. 11.4).106 The tumor is more frequently seen in the Orient than in Western countries, where it is extremely rare. The neoplasm often has a polymorphous cell population; the tumor cells usually permeate vascular structures. Thus, in the past, many of these cases had been diagnosed as polymorphic reticulosis or lymphomatoid granulomatosis. The cells have an unusual phenotype, expressing the T lineage markers CD 2 and CD 3 (the latter in paraffin sections only), as well the NK marker CD 56. However, other T lineage and NK markers are rarely or not expressed, thus, the true lineage of these neoplasms is still not clear. Molecular studies usually show a germline configuration of the beta T cell receptor gene. Similar neoplasms may occur in other locations, usually elsewhere in the head and neck. Using a variety of detection methods, EBV has been consistently identified in approximately 90 per cent of cases of nasal T/NK cell lymphoma in Western, Asian and South American populations.90,101,103,107-110 In EBVpositive cases, double-labeling immunohistochemical/m situ hybridization studies have localized EBV within all or nearly all of the neoplastic cells. Southern blotting studies using probes to the EBV termini have demonstrated EBV in a clonal population of cells, even in cases that are germline for the beta T cell receptor gene.111 Approximately 50 per cent of cases have EBV type A and
Figu re 11.4
EB V-positive T/NK cell nasalT cell lymphoma.
This EBER RNA in situ hybridization study labels all of the neoplastic nuclei black, indicating the presence of EBV. Note how the neoplastic cells tend to cluster around blood vessels.
the other 50 per cent of cases are positive for EBV type B.112 LMP-1 protein has not been consistently demonstrated in paraffin sections but it has been identified when frozen sections are available, suggesting low levels of the protein.110 EBNA-2 is usually not found; thus, the latency pattern is type II in these neoplasms. EBV has also been found in cases of NK/large granular lymphocyte leukemia and peripheral T cell lymphoma associated with hemophagocytosis, especially in Asian populations.113,114EBV has been found in approximately 20 per cent of cases of the HTLV-1-associated lymphoma ATLL.115 Researchers have hypothesized that EBV and HTD7-1 may coinfect the same T cells early in life, leading to an increased risk of subsequent lymphomagenesis. EBV has also been associated with the rare T cell lymphomas occurring in immunosuppressed populations, including T cell post-transplantation lymphomas and T cell AIDS-associated lymphomas.116,117 In the AIDS population, T cell lymphomas involving skin have been reported to have a high incidence of EBV positivity.117 Many other T cell lymphomas have been shown to be EBV positive in a subset of the neoplastic population. This pattern has been identified in many peripheral T
122 Viruses and malignant lymphoma
cell lymphomas, including angioimmunoblastic lymphadenopathy (AILD)-like T cell lymphoma and Lennert's lymphoma.118'122 Although it is unlikely that the EBV is of etiologic significance in these cases, the virus may contribute to disease progression. In addition to EBV-infected T cells in T cell lymphomas, increased numbers of EBV-infected B cells may also be present, particularly in AILD-like lymphoma. These latter cells may explain the paradoxical occurrence of B cell lymphomas complicating T cell lymphomas; some of these cases have been found to be EBV associated.123,124
HODGKIN'S DISEASE The epidemiology of Hodgkin's disease had long suggested a viral etiology in at least a subset of cases,125,126 particularly the pediatric and young adult cases. Investigators hypothesized that the pediatric cases developed in lower socioeconomic groups as a result of early life exposure to a virus, while the young adult cases resulted from later exposure to the same or different virus.126 EBV had been suspected to be that virus, because patients with Hodgkin's disease had a higher incidence of a history of acute infectious monoucleosis over controls.126-128 In addition, patients with Hodgkin's disease were known to have a higher incidence of abnormally high titers to EBV, both before and after diagnosis.129,130 In the last decade, numerous molecular studies have supplemented the previous epidemiologic and serologic data. EBV genomes were first identified in Hodgkin's disease tissues in 1987,131 and in 1989 EBV genomes were localized to Reed-Sternberg cells in a subset of cases.132 EBER in situ hybridization studies have shown EBV in Reed-Sternberg cells and variants in approximately 40-50 per cent of cases in Western populations (Fig. 11.5).133,134 It has been estimated that the EBV is amplified at least 50-fold in Reed-Sternberg cells.135 More recently, single-cell PCR studies have confirmed these findings and have definitively demonstrated that EBV-negative cases do occur.136 Study of the EBV terminus reveals the EBV to be in a monoclonal population in involved sites, with the same clonal population present when multiple sites from one case are examined.137,138 EBV type A is identified in almost all cases occurring in immunocompetent individuals.139,140 In HIV-infected patients, almost all cases of Hodgkin's disease are EBV associated, with either EBV type A or type B identified.141,142 EBV latency type II is found in the Reed-Sternberg cells of cases of EBV-associated Hodgkin's disease.143-145 In most cases, only a latent infection is present. A minority of cases show evidence of an abortive lytic infection146 and only rare cases show evidence of active viral replication.147 LMP-1 is strongly expressed in the Reed-Sternberg cells and variants (Fig. 11.6).143,144 In a subset of cases, a characteristic 30-bp deletion in the
Figu re 11.5
EB V-positive Hodgkin 's disease. This EBER RNA i n
situ hybridization study labels all or nearly all the Reed-Sternberg nuclei black, indicating the presence of EBV. Note that the vast majority of the small lymphocytes are negative.
LMP-1 gene is detected, particularly those cases with necrosis and numerous or atypical Reed-Sternberg cells.148 This 30-bp deletion is the same deletion reported in some cases of nasopharyngeal carcinoma associated with aggressive behavior when transplanted into nude mice. EBV positivity of Hodgkin's disease is most often seen in the mixed cellularity and lymphocyte depletion subtypes (approximately 80 per cent EBV positive) and infrequently seen in the nodular sclerosing subtype (approximately 20 per cent positive).133,134,143,149 Cases of the nodular, lymphocytic and histiocytic (L & H) lymphocyte predominance subtype, now thought by most investigators to represent a disease distinct from the other subtypes of Hodgkin's disease, are universally EBV negative.133,143,149,150 There is little correlation between EBV status and the cell lineage of the Reed-Sternberg cells, HLA-A2 positivity,151 CD 23 expression,152 cytokine profile153 or p53.154 One study reported a correlation with bd2 expression155 but this has not been confirmed in other reports.152,156,157 Several studies have shown an increased incidence of EBV positivity in pediatric Hodgkin's disease in various
Human herpes virus 6 123
or pathogenesis. The presence of EBV in a significant percentage of cases does suggest some role, but its absence in a large percentage of cases implies that it is not the sole etiologic factor. The presence of LMP-1 deletions in cases with 'aggressive' morphologic features suggest a role for EBV in disease progression. However, the lack of correlation of EBV with stage or patient outcome, at least in preliminary studies, suggests that the impact of EBV is not of major significance to the individual patient. Rare cases of small lymphocytic lymphoma/chronic lymphocytic leukemia contain Reed-Sternberg-like cells.92 In some cases, the Reed-Sternberg cells have the phenotype of the underlying B cell malignancy. In other cases, the Reed-Sternberg cells have the phenotype typical for Hodgkin's disease. Finally, yet other cases have a transitional phenotype. A proportion of these rare cases have progressed to overt Hodgkin's disease. In a study of these rare malignancies, we demonstrated that the Reed-Sternberg-like cells, but not the small lymphocytes, were EBV positive in 12 of 13 cases, as were the Reed-Sternberg cells in one case of Hodgkin's disease complicating this process.92 We hypothesized that EBV superinfected the low-grade B cell lymphoma and transformed the infected cell into a Reed-Sternberg-like cell. We believe that these rare cases, when further analysed, may provide interesting clues to the pathogenesis of Hodgkin's disease. Figure 11.6 EBV-positive Hodgkin's disease. This immunohistochemical study for LMP-1 demonstrates strong cytoplasmic and membrane positivity in the Reed-Stern berg cells. Note that the small lymphocytes are unstained.
populations,158-160 and one study has suggested an increased incidence of EBV positivity in patients older than 50 years, when compared to patients between 15 and 50 years old.158 There is no correlation between EBV status and patient serology,161 stage of disease or patient prognosis.162-164 In stage I patients, EBV positivity has been associated with presentation with neck lymph nodes, which interestingly are the sites of lymphatic draining from most primary EBV infections.165 There has been great interest in correlating EBV positivity with socioeconomic status and geographical, cultural and genetic factors, features that are often difficult to separate from one another. A slightly higher incidence of EBV positivity has been found in Hodgkin's disease occurring in Asian populations,166,167 and a much higher incidence has been found in South America,168 although pediatric patients comprise a relatively high percentage of cases of Hodgkin's disease in the latter population. One study showed an association between the presence of EBV in Hodgkin's disease and Hispanic ethnicity, in a series of cases from the USA, Mexico and Costa Rica.135 Although much progress has been made in elucidating the molecular epidemiology of EBV in Hodgkin's disease, there is not much known about its role in etiology
HUMAN HERPES VIRUS 6 Human herpes virus 6 was originally isolated (as human B lymphotropic virus; HBLV) in patients with a variety of lymphoproliferative disorders,169 leading to speculation that this virus may contribute to the etiology or pathogenesis of the malignant lymphomas. The virus has a wide cellular tropism, infecting primarily T, but also B, lymphocytes, as well as a variety of other cell types.170,171 The HHV-6 genome is a linear, doublestranded DNA of approximately 170 kilobases that has a 66 per cent nucleotide identity to cytomegalovirus (CMV).172 Almost all individuals are seropositive by 2-3 years of age, with liters gradually falling throughout adulthood.173,174 HHV-6 is the etiologic agent of exanthema subitum (roseola infantum),175 and maybe a cause of pneumonitis and other adverse manifestations in bone marrow transplant recipients.171 HHV-6 has been linked to cases of non-EBV/nonCMV-associated acute infectious mononucleosis.176 The virus also has been reported in a variety of other benign and malignant lymphoproliferative disorders,177-180 but many of these studies are difficult to interpret. Some of the findings have been based on assessment of HHV-6 antibody prevalence. Much of the results may be attributed to HHV-6 carrier state reactivations, secondary to
124 Viruses and malignant lymphoma
impaired cellular immune responses associated with the primary disease processes. In one large and carefully performed study, evidence of HHV-6 by Southern blotting studies was found in only three of 104 cases of non-Hodgkin's lymphoma, and none of 61 other benign and malignant lymphoid disorders (including 0/8 cases of Hodgkin's disease).177 Other reports document HHV-6 in occasional cases of non-Hodgkin's lymphoma and 'atypical polyclonal lymphoproliferation'.178 Torelli and colleagues, using PCR methodology, found HHV-6 sequences in 0/41 cases of non-Hodgkin's lymphoma, and in 3/25 cases of Hodgkin's disease. All three positive cases had similar clinical features.179 One study has reported HHV-6 in lesions of sinus histiocytosis with massive lymphadenopathy (RosaiDorfman disease) by in situ hybridization,181 while another group has reported HHV-6 in cases of histiocytic necrotizing lymphadenitis (Kikuchi's disease).182,183 Both reports need confirmation. In this regard, we were unable to demonstrate evidence of HHV-6 in Kikuchi's disease in a study using PCR methodology.
2. Poiesz BJ, Ruscetti FW, Gazdar AF. Detection and isolation of type C retrovirus particles from fresh and cultured lymphocytes of a patient with cutaneous T-cell lymphoma. Proc Natl Acad Sci USA 1980; 77: 7415-19. 3. Poiesz BJ, Ruscetti FW, Reitz MS, Kalyanaraman VS, Gallo RC. Isolation of a new type-C retrovirus (HTLV) in primary uncultured cells of a patient with Sezary T-cell leukaemia. Nature 1981; 294: 268-71. 4. Reitz JMS. Human T-cell leukemia virus, type 1, and human leukemia and lymphoma. In: Cossman J, ed. Molecular genetics in cancer diagnosis. New York: Elsevier, 1990:163-78. 5. Seiki M, Eddy R, Shows TB, Yoshida M. Nonspecific integration of the HTLV provirus into adult T-cell leukemia cells. Nature 1984; 309: 640-2. 6. Myoshi I, Kubonishi I, Yoshimoto S, et al. Type C virus particles in a cord blood T-cell line derived by cocultivating normal human cord leukocytes and human leukaemicT-cells. Nature 1981; 294: 770-2. 7. Seiki M, Inoue J, Takeda T, Yoshida M. Direct evidence that p40X of human T-cell leukemia virus type 1 is a trans-acting transcriptional activator. EMBOJ 1986; 5: 561-5.
HEPATITIS C Hepatitis C (HCV) is a linear, single-stranded RNA virus most similar to the Flaviviridae family. It has long been known to be a significant factor in the development of hepatocellular carcinoma. Recent evidence links this virus to lymphoproliferative diseases. HCV RNA has been detected in peripheral blood, bone marrow as well as in lymph nodes.184 A strong association between HCV infection and mixed cryoglobulinemia has been reported.185 In some populations, particularly in Italy, an increased prevalence of anti-HCV antibodies has been detected in patients with B-lineage non-Hodgkin's lymphomas. These are predominantly marginal zone lymphomas, including lymphomas of mucosa-associated lymphoid tissue, and immunocytoma.186'187 As with other viruses, the identification of a significant association does not necessarily imply a role in the etiology of the lymphoma. The HCV genome does not contain known oncogenes or reverse transcriptase that would lead to viral integration into the host's cellular DNA. Nonetheless, it is possible that HCV infection may induce a chronic B-cell proliferation that may act as a cofactor in lymphomagenesis.
8. Yoshida M, Fujisawa J. Positive and negative regulation of HTLV-1 gene expression and their roles in leukemiogenesis in ATL. In: Takatsuki K, Hinuma Y, Yoshida M, eds. Advances in adult T-cell leukemia and HTLV-I research, Vol. 39. Tokyo: Japanese Scientific Society Press, 1992; 217-35. 9. Yamaguchi K. Human T-lymphotropic virus type I in Japan. Lancet 1994; 343: 213-16. 10. Blattner WA, Blayney DW, Robert-Guroff M, et al. Epidemiology of human T-cell leukemia/lymphoma virus (HTLV).y Infect Dis 1983; 147: 406-16. 11. Catovsky D, Greaves MF, Rose M, et al. Adult T-cell lymphoma-leukaemia in Blacks from the West Indies. Lancet! 982; 1:639-43. 12. Tajima K, Kuroishi T. Estimation of incidence rate of ATL among ATLV(ATLV-I) carriers in Kyushu, Japan. JapJ Clin Oncol 1985; 15: 423-30. 13. Zucker-Franklin D, Coutavas EE, Rush MG, Zouzias DC. Detection of human T-lymphotropic virus-like particles in cultures of peripheral blood lymphocytes from patients with mycosis fungoides. Proc Natl Acad Sci USA 1991;88:7630-4. 14. Hall WW. Human T cell lymphotyropic virus type I and cutaneous T cell leukemia/lymphoma. J Exp Med 1994; 180:1581-5. 15. Whittaker SJ, Luzzatto L. HTLV-1 provirus and mycosis fungoides (letter). Science 1993; 259:1470-1.
REFERENCES
16. Pancake BA, Zucker-Franklin D. HTLV tax and mycosis fungoides. N EnglJ Med 1993; 329: 580. 17. Capesius C, Macro FSE, Bazarbachi A, et al. No evidence
1. Lyons SF, Liebowitz DN. The roles of human viruses in
for HTLV-I infection in 24 cases of French and
the pathogenesis of lymphoma. Semin Oncol 1998; 25:
Portuguese mycosis fungoides and Sezary syndrome (as
461-75.
seen in France). Leukemia 1991; 5: 416-19.
References 125 18. Srivastava Bl, Banki K, Perl A. Human T-cell leukemia virus type I or a related retrovirus in patients with mycosis fungoides/Sezary syndrome and Kaposi's sarcoma. Cancer Res 1992; 52: 4391-5. 19. Hall WW, Liu CR, Schneewind 0, etal. Deleted HTLV-I provirus in blood and cutaneous lesions of patients with mycosis fungoides. Science 1991; 253: 317-20. 20. Manca N, Piacentini E, Gelmi M, et al. Persistence of human T cell lymphotropic virus type 1 HTLV-1 sequences in peripheral blood mononuclear cells from patients from mycosis fungoides. J Exp Med 1994; 180: 1973-8. 21. Anagnostopoulos I, Hummel M, Kaudewitz P, Herbst H, Braun-Falco 0, Stein H. Detection of HTLV-I proviral sequences in CD30 positive large cell cutaneous T cell lymphomas. AmJPathol 1990; 137:1317-22. 22. Waldmann TA, White JD, Carrasquillo JA, et al. Radioimmunotherapy of interleukin-2R alpha-expressing adult T-cell leukemia with Yttrium-90-labeled anti-Tac. Blood 1995; 86:1063-75. 23. Hermine 0, Bouscary D, Gessain A, et al. Brief report: Treatment of adult T-cell leukemia-lymphoma with zidovudine and interferon alfa. N EnglJ Med 1995; 332: 1749-51. 24. Gill PS, Harrington W Jr, Kaplan MH, et al. Treatment of adult T-cell leukemia-lymphoma with a combination of interferon alfa and zidovudine. N EnglJ Med 1995; 332: 1744-48. 25. ChangY, Cesarman E, Pessin MS, etal. Identification of herpesvirus-like DMA sequences in AIDS-associated Kaposi's sarcoma. Science 1994; 266:1865-9. 26. Cesarman E, Chang Y, Moore PS, Said JW, Knowles DM. Kaposi's sarcoma-associated herpesvirus-like DNA sequences in AIDS-related body-cavity-based lymphomas. N EnglJ Med 1995; 332:1186-91. 27. Nador RG, Cesarman E, Knowles DM, Said JW. Herpeslike DNA sequences in a body-cavity-based lymphoma in an HIV-negative patient. N EnglJ Med 1995; 333: 943. 28. Nador RG, Cesarman E, Chadburn A, et al. Primary effusion lymphoma: a distinct clinicopathologic entity associated with the Kaposi's sarcoma-associated herpes virus. Blood 1996; 88: 645-56. 29. Martin A, Capron F, Liguory-Braunaud M-D, Frejacques C, Pluot M, Diebold J. Epstein-Barr virus-associated primary malignant lymphomas of the pleural cavity occurring in longstanding pleural chronic inflammation. Human Pathol 1994; 25:1314-18. 30. Fukayama M, Ibuka T, Hayashi Y, Ooba T, Koike M, Mizutani S. Epstein-Barr virus in pyothorax-associated pleural lymphoma. Am J Pathol 1993; 143:1044-9. 31. Soulier J, Grollet L, Oksenhendler E, etal. Kaposi's sacoma-associated herpesvirus-like DNA sequences in multicentric Castleman's disease. Blood 1995; 86: 1276-80. 32. Luppi M, Barozzi P, Maiorana A, etal. Human herpesvirus-8 DNA sequences in human
33.
34.
35.
36.
37.
38.
39.
40.
41.
42.
43.
44.
immunodeficiency virus-negative angioimmunoblastic lymphadenopathyand benign lymphadenopathy with giant germinal center hyperplasia and increased vascularity. Blood 1996; 87: 3903-9. Said JW. Body cavity-based (primary effusion) lymphoma: a new lymphoma subtype associated with Kaposi's sarcoma herpesvirus (human herpesvirus 8). AdvAnat Pathol 1996; 3: 254-8. Sarid R, Sato T, Bohenzky RA, Russo JJ, Chang Y. Kaposi's sarcoma-associated herpesvirus encodes a functional bcl-2 homologue. Nat Med 1997; 3: 293-8. Cheng EH, Nicholas J, Bellows DS, et al. A Bcl-2 homolog encoded by Kaposi sarcoma-associated virus, human herpesvirus 8, inhibits apoptosis but does not heterodimerize with Bax or Bak. Proc Natl Acad Sci USA 1997;94:690-4. Arvanitakis L, Geras Raaka E, Varma A, et al. Human herpesvirus KSHV encodes a constitutively active Gprotein-coupled receptor linked to cell proliferation. M7fure1997;385:347-50. Said JW, Chien K, Tasaka T, et al. Kaposi's sarcoma associated herpesvirus (KSHV or HHV8) in primary effusion lymphomas: ultrastructural demonstration of herpesvirus in lymphoma cell lines. Blood 1996; 87: 4937-43. Cesarman E, Moore PS, Rao PH, Inghirami G, Knowles DM, Chang Y. In vitro establishment and characterization of two acquired immunodeficiency syndrome-related lymphoma cell lines (BC-1 and BC-2) containing Kaposi's sarcoma-associated herpesvirus-like (KSHV) DNA sequences. Blood 1995; 86: 2708-14. Epstein MA, Achong BG, Barr YM. Virus particles in cultured lymphoblastsfrom Burkitt's lymphoma. Lancet 1964; 1:702-3. Kulwichit W, Edwards RH, Davenport EM, Baskar JF, Godfrey V, Raab-Traub N. Expression of the Epstein-Barr virus latent membrane protein 1 induces B cell lymphoma in transgenic mice. Proc Natl Acad Sci USA 1998; 95:11963-8. Sample J, Young L, Martin B, Chatman T, Kieff E, Rickinson A. Epstein-Barr virus types 1 and 2 differ in the EBNA-3A, EBNA-3B, and EBNA-3C genes.7 Virol 1990;64:4084-92. Sculley TB, Apollini A, Hurren L, Moss DE, Cooper D. Coinfection with A- and B-type Epstein-Barr virus in human immunodeficiency virus-positive subjects. J Infect Dis 1990; 162: 643-8. Chang KL, Chen Y-Y, Shibata D, Weiss LM. In situ hybridization methodology for the detection of EBV EBER-1 RNA in paraffin-embedded tissues, as applied to normal and neoplastic tissues. Diagn Mol Pathol 1992; 1:246-55. Murray PG, Young LS, Rowe M, Crocker J. Immunohistochemical demonstration of the Epstein-Barr virus-encoded latent membrane protein in paraffin sections of Hodgkin's disease.) Pathol 1992; 166: 1-5.
126 Viruses and malignant lymphoma 45. Grasser FA, Murray PG, Kremmer E, etal. Monoclonal antibodies directed against the Epstein-Barr virusencoded nuclear antigen 1 (EBNA1): immunohistologic detection of EBNA1 in the malignant cells of Hodgkin's disease. Blood 1994; 84: 3792-8. 46. Raab-Traub N, Flynn K. The structure of the termini of the Epstein-Barr virus as a marker of clonal cellular proliferation. Cell 1986; 47: 883-9. 47. Katz BZ, Raab-Traub N, Miller G. Latent and replicating forms of Epstein-Barr virus DMA in lymphomas and lymphoproliferative disease. 7 Infect Dis 1989; 160: 589-98. 48. Knecht H, Bachmann E, Joske DJL, et al. Molecular analysis of the IMP (latent membrane protein) oncogene in Hodgkin's disease. Leukemia 1993; 7: 580-5. 49. Frank D, Cesarman E, Liu YF, Michler RE, Knowles DM. Posttransplantation lymphoproliferative disorders frequently contain type A and not type B Epstein-Barr virus. Blood 1995; 85:1396-403. 50. Lenoir GM, Philip T, Sohier R. Burkitt-type lymphoma: EBV association and cytogenetic markers in cases from various geographic locations. In: Magrath I, O'Conor GT, Ramot B, eds. Pathogenesis of leukemias and lymphomas: environmental influences. New York: Raven Press, 1984:283. 51. Hummel M, Anagnostopoulos I, Korbjuhn P, Stein H. Epstein-Barr virus in B-cell non-Hodgkin's lymphomas: unexpected infection patterns and different infection incidence in low- and high-grade types. J Pathol 1995; 175:263-71. 52. Shiramizu B, Barriga F, Neequaye J, etal. Patterns of chromosomal breakpoint locations in Burkitt's lymphoma: relevance to geography and Epstein-Barr virus association. Blood 1991; 77:1516-26. 53. Bacchi MM, Bacchi CE, Alvarenga M, Miranda R, Chen Y-Y, Weiss LM. Burkitt's lymphoma in Brazil: strong association with Epstein-Barr virus. Mod Pathol 1995; 9: 63-7. 54. Drut RM, Day S, Drut R, Meisner L Demonstration of Epstein-Barr viral DNA in paraffin-embedded tissues of Burkitt's lymphoma from Argentina using the polymerase chain reaction and in situ hybridization. Pediatr Pathol 1994; 14:101-9. 55. Chan JKC, Tsang WYW, Ng CS, Wong CSC, Lo ESF. A study of the association of Epstein-Barr virus with Burkitt's lymphoma occurring in a Chinese population. Histopathology 1995; 26: 239-45. 56. Gutierrez Ml, Bhatia K, Barriga F, etal. Molecular epidemiology of Burkitt's lymphoma from South America: differences in breakpoint location and Epstein-Barr virus association from tumors in other world regions. Blood 1992; 79: 3261-6. 57. Araujo I, Foss H-D, Bittencourt A, etal. Expression of Epstein-Barr virus gene products in Burkitt's lymphoma in Northeast Brazil. Blood 1996; 87: 5279-86. 58. Chen W-G, Chen Y-Y, Bacchi MM, Bacchi CE, Alvarenga
59.
60.
61.
62.
63.
64.
65.
66.
67.
68.
69.
70.
M, Weiss LM. Genotyping of Epstein-Barr virus in Brazilian Burkitt's lymphoma and reactive lymphoid tissue: type A with a high prevalence of deletions within the latent membrane protein gene. Am) Pathol 1996; 148:17-23. Mansoor A, Stetler-Stevenson M, Li RZ, etal. Prevalence of Epstein-Barr viral sequences and EBV LMP1 concogene deletions in Burkitt's lymphoma from Pakistan: epidemiological correlations. Human Pathol 1997; 28: 283-8. Anwar N, Kingma DW, Bloch AR, etal. The investigation of Epstein-Barr viral sequences in 41 cases of Burkitt's lymphomas from Egypt: epidemiological correlations. Cancer Molec Biol 1994; 1: 213-22. Cleary ML, Nalesnik MA, Shearer WT, Sklar J. Clonal analysis of transplant-associated lymphoproliferations based on the structure of the genomic termini of the Epstein-Barr virus. Blood 1988; 72: 349-52. Knowles DN, Cesarman E, Chadburn A, et al. Correlative morphologic and molecular genetic analysis demonstrates three distinct categories of posttransplantation lymphoproliferative disorders. Blood 1995; 85: 552-65. Randhawa PS, Jaffe R, Demetris AJ, et al. The systemic distribution of Epstein-Barr virus genomes in fatal posttransplant lymphoproliferative disorders: an in situ hybridization study. Am7 Pathol 1991; 138:1027-33. Montone KT, Friedman H, Hodinka RL, Hicks DG, Kant JA, Tomaszewski JE. In situ hybridization for Epstein-Barr virus Notl repeats in posttransplant lymphoproliferative disorder. Mod Pathol 1992; 5: 292-302. Randhawa PS, Jaffe R, Demetris AJ, et al. Expression of Epstein-Barr virus-encoded small RNA (by the EBER-1 gene) in liver specimens with post-transplant lymphoproliferative disease. N Engl J Med 1992; 327: 710-14. Kyaw MT, Hurren L, Evans L, et al. Expression of B-type Epstein-Barr virus in HIV-infected patients and cardiac transplant recipients. AIDS Res 1992; 8:1869-74. Kaplan MA, Ferry JA, Harris NL, Jacobson JO. Clonal analysis of posttransplant lymphoproliferative disorders, using both episomal Epstein-Barr virus and immunoglobulin genes as markers. Am) Clin Pathol 1994; 101: 590-6. Liebowitz D. Epstein-Barr virus and a cellular signaling pathway in lymphomas from immunosuppressed patients. N Engl J Med 1998; 338:1413-21. Kamel OW, van de Rijn M, LeBrun DP, Weiss LM, Warnke RA, Dorfman RF. Lymphoproliferative lesions in patients with rheumatoid arthritis and dermatomyositis: frequency of Epstein-Barr virus and other features associated with immunosupression. Human Pathol 1994; 25: 638-43. Kamel OW, van de Rijn M, Weiss LM, et al. Reversible EBV-associated lymphomas occurring in the setting of methotrexate therapy for rheumatoid arthritis and dermatomyositis. N Engl J Med 1993; 328:1317-21.
References 127 71. Grierson H, Purtilo DT. Epstein-Barr virus infections in males with the X-linked lymphoproliferative syndrome. Ann Intern Med 1987; 106: 538^5. 72. Birx D, Redfield RR, Tosato G. Defective regulation of Epstein-Barr virus infection in patients with acquired immunodeficiency syndrome (AIDS) or AIDS related disorders. N EnglJ Med 1986; 314: 874-9. 73. Arber DA, Shibata D, Chen Y-Y, Weiss LM. Characterization of the topography of Epstein-Barr virus infection in human immunodeficiency virus-associated lymphoid tissues. Mod Pathol 1992; 5: 559-66. 74. Shibata D, Weiss LM, Nathwani BN, Brynes RK, Levine AM. Epstein-Barr virus in benign lymph node biopsies from individuals infected with the human immunodeficiency virus is associated with concurrent or subsequent development of non-Hodgkin's lymphoma. 6W1991; 77:1527-33. 75. Shibata D, Weiss LM, Hernandez AM, Nathwani BN, Bernstein L, Levine AM. Epstein-Barr virus-associated non-Hodgkin's lymphoma in patients infected with the human immunodeficiency virus. Blood 1993; 81: 2102-9. 76. Hamilton-Dutoit SJ, Raphael M, Audouin J, et al. In situ demonstration of Epstein-Barr virus small RNAs (EBER1) in acquired immunodeficiency syndrome-related lymphomas: correlation with tumor morphology and primary site. Blood 1993; 82: 619-24. 77. Schlaifer D, Brousset P, Attal M, et al. bcl-2 protooncogeneand Epstein-Barr virus latent membrane protein-1 expression in AIDS-related lymphoma. Histopathology 1994; 25: 77-82. 78. Pederson C, Gerstoft J, Lundgren JD, et al. HIVassociated lymphoma: histopathology and association with Epstein-Barr virus genome related to clinical, immunological and prognostic features. EurJ Cancer 1991; 27: 1416-23. 79. Borisch B, Finke J, Hennig I, et al. Distribution and localization of Epstein-Barr virus subtypes A and B in AIDS-related lymphomas and lymphatic tissue of HIVpositive patients. J Pathol 1992; 168: 229-36. 80. Raphael MM, Audouin J, Lamine M, et al. Immunophenotypic and genotypic analysis of acquired immunodeficiency syndrome-related non-Hodgkin's lymphomas. Correlation with histologic features in 36 cases. AmJClin Pathol 1994; 101: 773-82. 81. Carbone A, Gloghini A, Zanette I, Canal B, Volpe R. Demonstration of Epstein-Barr viral genomes by in situ hybridization in acquired immune deficiency syndrome-related high grade and anaplastic large cell CD30+ lymphomas. AmJClin Pathol 1993; 99: 289-97. 82. MacMahon E, Glass J, Hayward S, et al. Epstein-Barr virus in AIDS-related primary central nervous system lymphoma. Lancet 1991; 338: 969-73. 83. Chang KL, Flaris N, Hickey WF, Johnson RM, Meyer P, Weiss LM. Brain lymphomas of immunocompetent and immunocompromised patients: study of the association with Epstein-Barr virus. Mod Pathol 1993; 6: 427-32.
84. Hamilton-Dutoit SJ, Rea D, Raphael M, et al. Epstein-Barr virus-latent gene expression and tumor cell phenotype in acquired immunodeficiency syndromerelated non-Hodgkin's lymphoma: correlation of lymphoma phenotype with three distinct patterns of viral latency. Am J Pathol 1993; 143:1072-85. 85. Brousset P, Drouet E, Schlaifer D, et al. Epstein-Barr virus (EBV) replicative gene expression in tumor cells of AIDS-related non-Hodgkin's lymphoma in relation to CD4cell number and antibody titres to EBV. AIDS 1994; 8: 583-90. 86. Neri A, Barriga F, Inghirami G, et al. Epstein-Barr virus infection precedes clonal expansion in Burkitt'sand AIDS-associated lymphoma. Blood 1991; 77:1092-5. 87. Herndier BG, Shiramizu BT, McGrath MS. AIDS associated non-Hodgkin's lymphomas represent a broad spectrum of monoclonal and polyclonal lymphoproliferative processes. Curr Topics Microbiol Immunol 1992; 182: 385-94. 88. Hamilton-Dutoit S, Pallesen G. A survey of Epstein-Barr virus gene expression in sporadic non-Hodgkin's lymphomas. Detection of Epstein-Barr virus in a subset of peripheral T cell lymphomas. Am) Pathol 1992; 140: 1315-25. 89. Rooney CM, Roskrow MA, Smith CA, Brenner MK, Heslop HE. Immunotherapy for Epstein-Barr virus-associated cancers. J Natl Cancer Inst Monogr 1998; 23: 89-93. 90. Weiss LM, Gaffey MJ, Chen Y-Y, Frierson HF. Frequency of Epstein-Barr viral DMA in 'Western' sinonasal and Waldeyer's ring lymphoma. Am J Surg Pathol 1992; 16: 156-62. 91. Hui PK, TokunagaM, Chan WY, NgCS, Chow J, LeeJCK. Epstein-Barr virus-associated gastric lymphoma in Hong Kong Chinese. Human Pathol 1994; 25: 947-52. 92. Momose H, Jaffe ES, Shin SS, Chen Y-Y, Weiss LM. Chronic lymphocytic leukemia/small lymphocytic lymphoma with Reed-Stern berg-like cells and possible transformation to Hodgkin's disease. Mediation by Epstein-Barr virus. Am J Surg Pathol 1992; 16: 859-67. 93. Chang KL, Chen Y-Y, Weiss LM. Lack of evidence of Epstein-Barr virus in hairy cell leukemia and monocytoid B-cell lymphoma. Human Pathol 1993; 24: 58-61. 94. Sasajima Y, Yamabe H, Kobashi Y, Hirai K, Mori S. High expression of the Epstein-Barr virus latent protein EB nuclear antigen-2 on pyothorax-associated lymphomas. AmJ Pathol 1993; 143:1280-5. 95. Guinee D, Jaffe E, Kingma D, et al. Pulmonary lymphomatoid granulomatosis: evidence for a proliferation of Epstein-Barr virus infected Blymphocytes with a prominent T-cell component and vasculitis. AmJ Surg Pathol 1994; 18: 753-64. 96. Paterson RLK, Kelleher C, Amankonah TD, et al. Model of Epstein-Barr virus infection of human thymocytes: expression of viral genome and impact on cellular receptor expression in the T-lymphoblastic cell line. HPB-ALL. Blood 1995; 85: 456-64.
128 Viruses and malignant lymphoma 97. Deamant FD, Albujar PF, Chen Y-Y, Weiss LM. Epstein-Barr virus distribution in non-neoplastic lymph nodes. Mod Pathol 1993; 6: 729-32. 98. Tsoukas CD, Lambris JD. Expression of CR2/EBV receptors on human thymocytes detected by monoclonal antibodies. EurJ Immunol 1988; 18: 1299-302. 99. Fischer E, Delibrias C, Kazatchkine MD. Expression of the CR2 (the C3dg/EBV receptor, CD21) on normal human peripheral blood T lymphocytes.) Immunol 1991; 146: 865-9. 100. Jones J, Shurin S, Abramowsky C, et al. T-cell lymphomas containing Epstein-Barr viral DMA in patients with chronic Epstein-Barr virus infections. N Englj Med 1988; 318: 733-41. 101. Harabuchi Y, Yamanaka N, Kataura A, etal. Epstein-Barr virus in nasal T-cell lymphoma in patients with lethal midline granuloma. Lancet 1990; 335: 128-30. 102. Su I-J, Hsieh H-C, Lin K-H, etal. Aggressive peripheral Tcell lymphomas containing Epstein-Barr viral DMA: a clinicopathologicand molecular analysis. B/oorf1991; 77: 799-808. 103. Chan JKC, Yip TTC, Tsang WYW, et al. Detection of Epstein-Barr viral RNA in malignant lymphomas of the upper aerodigestive tract. AmJ Surg Pathol 1994; 18: 938-46. 104. Quintanilla-Martinez L, Lome-Maldonado C, Ott G, et al. Primary non-Hodgkin's lymphoma of the intestine: high prevalence of Epstein-Barr virus in Mexican lymphomas as compared to European cases. Blood 1997; 89: 644-51. 105. d'Amore F, Johansen P, Houmand A, Weisenburger DD, Mortensen LS. Epstein-Barr virus genome in nonHodgkin's lymphomas occurring in immunocompetent patients: highest prevalence in nonlymphoblastic T-cell lymphoma and correlation with a poor prognosis. Blood 1996;87:1045-55. 106. Weiss LM, Arber DA, Strickler JG. Nasal T cell lymphoma. Ann Oncol 1994; 5: (suppl 1): 39-42. 107. Arber DA, Weiss LM, Albujar PF, Chen Y-Y, Jaffe ES. Nasal lymphomas in Peru: high incidence of T-cell immunophenotype and Epstein-Barr virus infection. AmJ Surg Pathol 1993; 17: 659-65. 108. Strickler JG, Meneses M, Habermann TM, etal. 'Polymorphic reticulosis': a reappraisal. Human Pathol 1994; 25: 659-65. 109. Mishima K, Horiuchi K, Kojya S, Takahashi H, Ohsawa M, Aozasa K. Epstein-Barr virus in patients with polymorphic reticulosis (lethal midline granuloma) from China and Japan. Cancer 1994; 73: 3041-6. 110. Kanavaros P, Lescs M-C, Briere J, et al. Nasal T-cell lymphoma: a clinicopathologic entity associated with peculiar phenotype and with Epstein-Barr virus. Blood 1993;81:2688-95. 111. Medeiros LJ, Peiper SC, Elwood L, Yano T, Raffeld M, Jaffe ES. Angiocentric immunoproliferative lesions: a
112.
113.
114.
115.
116.
117.
118.
119.
120.
121.
122.
123.
124.
molecular analysis of eight cases. Human Pathol 1991; 22:1150-7. Borisch B, Hennig I, Laeng RH, Waelti ER, Kraft R, Laissue J. Association of the subtype 2 and the Epstein-Barr virus with T-cell non-Hodgkin's lymphoma of the midline granuloma type. Blood 1993; 82: 858-64. Su IJ, HsuYH, Lin MT, Cheng AL, Wang CH, Weiss LM, Epstein-Barr containing T cell lymphoma presents with hemophagocytic syndrome mimicking malignant histiocytosis. Cancer 1993; 72: 2019-27. Kawa-Ha K, Ishihara S, Ninomiya T, etal. CDS-negative lymphoproliferative disease of granular lymphocytes containing Epstein-Barr viral DMA. 7 Clin Invest 1989; 84:51-5. Tokunaga M, Imai S, Uemura Y, TokudomeT, Osato T, Sato E. Epstein-Barr virus in adult T-cell leukemia/lymphoma. AmJ Pathol 1993; 143:1263-9. Kumar S, Kumar D, Kingma DW, Jaffe ES. Epstein-Barr virus-associated T-cell lymphoma in a renal transplant patient. AmJ Surg Pathol 1993; 17:1046-53. Kerschmann RL, Berger TG, Weiss LM, et al. Cutaneous presentations of lymphoma in human immunodeficiency virus disease: predominance of T-cell lineage with two clinicopathologic presentations. Arch Derm 1995; 131:1281 -8. Zhou XG, Hamilton-Dutoit SJ, Yan QH, Pallesen G. High frequency of Epstein-Barr virus in Chinese peripheral T-cell lymphoma. Histopathology 1994; 24:115-22. Weiss LM, Jaffe ES, Liu X, Chen Y, Shibata D, Medeiros LJ. Detection and localization of Epstein-Barr viral genomes in angioimmunoblastic lymphadenopathy and angioimmunoblastic lymphadenopathy-like lymphomas. Blood 1992; 79:1789-95. Anagnostopoulos I, Hummel M, Finn T, et al. Heterogeneous Epstein-Burr virus infection patterns in peripheral T-cell lymphoma of angioimmunoblastic lymphadenopathy type. Blood 1992; 80:1804-12. Korbjuhn P, Anagnostopoulos I, Hummel M, etal. Frequent latent Epstein-Barr virus infection of neoplastic T cells and bystander B cells in HIV-negative European peripheral pleomorphic T-cell lymphomas. Blood 1993; 82: 217-23. Anagnostopoulos I, Hummel M, Tiemann M, Korbjuhn P, Parwaresch MR, Stein H. Frequent presence of latent Epstein-Barr virus infection in lymphoepithelioid cell lymphoma (Lennert's lymphoma). Histopathology 1994; 25:331-7. Abruzzo LV, Schmidt K, Weiss LM, etal. B cell lymphoma evolving from AILD: a case with oligoclonal gene rearrangements associated with Epstein-Barr virus. B/oorf1993;82:241-6. Knecht H, Martius F, Bachmann E, et al. A deletion mutant of the LMP1 oncogene of Epstein-Barr virus is associated with evolution of angioimmunoblastic lymphadenopathy into B immunoblastic lymphoma. Leukemia 1995; 9: 458-65.
References 129 125. MacMahon B. Epidemiology of Hodgkin's disease. Cancer Res 1966; 26:1189-200. 126. Gutensohn N, Cole P. Childhood social environment and Hodgkin's disease. N EnglJ Med 1981; 304:135-40. 127. Kvale G, H0iby E, Pederson E. Hodgkin's disease in patients with previous infectious mononucleosis. Int J Cancer 1979; 23: 593-7. 128. Rosdahl N, Larsen S, Clemmesen J. Hodgkin's disease in patients with previous infectious mononucleosis: 30 years' experience. BrMedJ 1974; 2: 253-6. 129. Mueller N, Evans A, Harris NL, et al. Hodgkin's disease and Epstein-Barr virus: altered antibody pattern before diagnosis. N EnglJ Med 1989; 320: 689-92. 130. Evans AS, Gutensohn NM. A population-based casecontrol study of EBV and other viral antibodies among persons with Hodgkin's disease and their siblings. IntJ Cancer 1984; 34:149-57. 131. Weiss LM, Strickler JG, Warnke RA, Purtilo DT, Sklar J. Epstein-Barr viral DMA in tissues of Hodgkin's disease. AmJ Pathol 1987; 129: 86-91. 132. Weiss LM, Movahed LA, Warnke RA, Sklar J. Detection of Epstein-Barr viral genomes in Reed-Stern berg cells of Hodgkin's disease. N EnglJ Med 1989; 320: 502-6. 133. Weiss LM, Chen Y-Y, Liu X-F, Shibata D. Epstein-Barr virus and Hodgkin's disease: a correlative in situ hybridization and polymerase chain reaction study. Am J Pathol 1991; 139:1259-65. 134. Herbst H, Steinbrecher E, Niedobitek G, et al. Distribution and phenotype of Epstein-Barr virusharboring cells in Hodgkin's disease. Blood 1992; 80: 484-91. 135. Gulley ML, Eagan PA, Quintanilla-Martinez L, et al. Epstein-Barr virus DMA is abundant and monoclonal in the Reed-Stern berg cells of Hodgkin's disease: association with mixed cellularity subtype and Hispanic American ethnicity. Blood 1994; 83:1595-602. 136. Roth J, Daus H, Gause A, Trumper L, Pfreundschuh M. Detection of Epstein-Barr virus DNA in Hodgkin- and Reed-Stern berg-eel Is by single cell PCR. Leuk Lymphoma 1994; 13: 137-42. 137. Boiocchi M, Dolcetti R, Re VD, Gloghini A, Carbone A. Demonstration of a unique Epstein-Barr virus-positive cellular clone in metachronous multiple localizations of Hodgkin's disease. Am J Pathol 1993; 142: 33-8. 138. Brousset P, Schlaifer D, Meggetto F, et al. Persistence of the same viral strain in early and late relapses of Epstein-Barr virus-associated Hodgkin's disease. Blood 1994;84:2447-51. 139. Gledhill S, Gallagher A, Jones DB, et al. Viral involvement in Hodgkin's disease: detection of clonal type A Epstein-Barr virus genomes in tumour samples. BrJ Cancer 1991; 64: 227-32. 140. Lin JC, Lin SC, De BK, Chan WP, Evatt BL Precision of genotyping of Epstein-Barr virus by polymerase chain reaction using three gene loci (EBNA-2, EBNA-3C, and EBER): predominance of type A virus associated with Hodgkin's disease. Blood 1993; 81: 3372-81.
141. Re VD, Boiocchi M, Vita SD, et al. Subtypes of Epstein-Barr virus in HIV-associated and HIV-unrelated Hodgkin's disease cases. IntJ Cancer 1993; 54: 895-8. 142. Boyle MJ, Vasak E, Tschuchnigg M, et al. Subtypes of Epstein-Barr virus (EBV) in Hodgkin's disease: association between B-type EBV and immunocompromise. Blood 1993; 81: 468-74. 143. Pallesen G, Hamilton-Dutoit SJ, Rowe M, Young S. Expression of Epstein-Barr virus latent gene products in tumour cells of Hodgkin's disease. Lancet 1991; 337: 320-2. 144. Herbst H, Dallenbach F, Hummel M, et al. Epstein-Barr virus latent membrane protein expression in Hodgkin and Reed-Stern berg cells. Proc Natl Acad Sci USA 1991; 88: 4766-70. 145. Deacon EM, Pallesen G, Niedobitek G, et al. Epstein-Barr virus and Hodgkin's disease: transcriptional analysis of virus latency in the malignant cells. J Exp Med 1993; 177: 339-49. 146. Pallesen G, Sandvej K, Hamilton-Dutoit SJ, Rowe M, Young LS. Activation of Epstein-Barr virus replication in Hodgkin's and Reed-Stern berg cells. Blood 1991; 78: 1162-5. 147. Bibeau F, Brousset P, Knecht H, et al. Epstein-Barr virus replication in Hodgkin's disease. Bull Cancer 1994; 81: 114-88. 148. Knecht H, Bachmann E, Brousset P, et al. Deletions within the LMP1 oncogene of Epstein-Barr virus are clustered in Hodgkin's disease and identical to those observed in nasopharyngeal carcinoma. Blood 1993; 82: 2937-42. 149. Brousset P, Chittal S, Schlaifer D, et al. Detection of Epstein-Barr virus messenger RNA in Reed-Sternberg cells of Hodgkin's disease by in situ hybridization with biotinylated probes on specially processed modified acetone methyl benzoate xylene (ModAMeX) sections. B/oo1991;77:1781-6. 150. Stoler MH, Nichols GE, Symbula M, Weiss LM. Nodular L&H lymphocyte predominance Hodgkin's disease: evidence fora kappa light chain-restricted monotypic B cell neoplasm. AmJ Pathol 1995; 146:812-18. 151. Poppema S. Epstein-Barr virus positivity in Hodgkin's disease does not correlate with an HLA A2-negative phenotype. Cancer 1994; 73: 3059-63. 152. Armstrong AA, Gallagher A, Krajewski AS, et al. The expression of the EBV latent membrane protein (LMP-1) is independent of CD23 and bcl-2 in Reed-Sternberg cells in Hodgkin's disease. Histopathology 1992; 21: 72-3. 153. Foss H-D, Herbst H, Oelmann E, et al. Lymphotoxin, tumour necrosis factor and interleukin-6 gene transcripts are present in Hodgkin and Reed-Sternberg cells of most Hodgkin's disease cases. BrJ Haematol 1993; 84: 627-35. 154. Niedobitek G, Rowlands DC, Young LS, et al. Overexpression of p53 in Hodgkin's disease: lack of correlation with Epstein-Barr virus infection. J Pathol 1993; 169: 207-12.
130 Viruses and malignant lymphoma 155. Khan G, Gupta RK, Coates PJ, Slavin G. Epstein-Barr virus infection and bcl-2 proto-oncogene expression. Separate events in the pathogenesis of Hodgkin's disease. Am J Pathol 1993; 143:1270-4. 156. Bhagat SK, Medeiros LJ, Weiss LM, Wang J, Raffeld M, Stetler-Stevenson M. bcl-2 expression in Hodgkin's disease. Correlation with the t(14;18) translocation and Epstein-Barr virus. Am J Clin Pathol 1993; 99: 604-8. 157. LeBrun DP, Ngan BY, Weiss LM, Huie P, Warnke RA, Cleary ML. The bcl-2 oncogene in Hodgkin's disease arising in the setting of follicular non-Hodgkin's lymphoma. Blood 1994; 83: 223-30. 158. Jarrett RF, Gallagher A, Jones DB, et al. Detection of Epstein-Barr virus genomes in Hodgkin's disease: relation to age. 7 Clin Pathol 1991; 44: 844-8. 159. Armstrong AA, Alexander FE, Paes RP, et al. Association of Epstein-Barr virus with pediatric Hodgkin's disease. Am J Pathol 1993; 142:1683-8. 160. Ambinder RF, Browning PJ, Lorenzana I, et al. Epstein-Barr virus and childhood Hodgkin's disease in Honduras and the United States. Blood 1993; 81: 462-7. 161. Levine PH, Pallesen G, Ebbesen P, Harris N, Evans AS, Mueller N. Evaluation of Epstein-Barr virus antibody patterns and detection of viral markers in the biopsies of patients with Hodgkin's disease. Int J Cancer 1994; 59: 48-50. 162. Fellbaum C, Hansmann M-L, Niedermeyer H. Influence of Epstein-Barr virus genomes on patient survival in Hodgkin's disease. Am J Clin Pathol 1992; 98: 319-23. 163. Armstrong AA, Lennard A, Alexander FE, et al. Prognostic significance of Epstein-Barr virus association in Hodgkin's disease (letter). EurJ Cancer 1994; 30A: 1045-6. 164. Vestlev PM, Pallesen G, Sandvej K, Hamilton-Dutoit SJ, Bendtzen SM. Prognosis of Hodgkin's disease is not influenced by Epstein-Barr virus latent membrane antigen. IntJ Cancer 1991; 50: 670-1. 165. O'GradyJ, Stewart S, Elton RA, Krajewski AS. Epstein-Barr virus in Hodgkin's disease and site of origin of tumour. Lancet 1994; 343: 265-6. 166. Zhou X-G, Hamilton-Dutoit SJ, Van Q-H, Pallesen G. The association between Epstein-Barr virus and Chinese Hodgkin's disease. Int J Cancer 1993; 55: 359-63. 167. Chan JKC, Yip TTC, Tsang WYW, Lau W-H, Wong CSC, Ma VWS. Detection of Epstein-Barr virus in Hodgkin's disease occurring in an Oriental population. Human Pathol 1995; 26: 314-18. 168. Chang KL, Albujar PF, Chen Y-Y, Johnson RM, Weiss LM. High prevalence of Epstein-Barr virus in the Reed-Stern berg cells of Hodgkin's disease occurring in Peru. Blood 1993; 81: 496-502. 169. Salahuddin SZ, Ablashi DV, Markham PD, et al. Isolation of a new virus, HBLV, in patients with lymphoproliferative disorders. Science 1986; 234: 596-601.
170. Horwitz C, Beneke J. Human herpesvirus-6 revisited (editorial). Am J Clin Pathol 1993; 99: 533-5. 171. Caserta MT, Hall CB. Human herpesvirus-6. Ann Rev Med 1993; 44: 377-83. 172. Lawrence GL, Chee M, Craxton MA, Honess RW, et al. Human herpesvirus 6 is closely related to human cytomegalovirus.y Virol 1990; 64: 287-99. 173. Yoshikawa T, Suga S, Asano Y, Yazaki T, Ozaki T. Neutralizing antibodies to human herpesvirus-6 in healthy individuals. Pediatr Infect DisJ 1990; 9: 589-90. 174. Brown NA, Sumaya CV, Liu C, Ench Y, Kovacs A, et al. Fall in human herpesvirus 6 seropositivity with age. Lancet 1988; 2: 396. 175. Yamanishi K, Toshiomi 0, Kimiyasu S, Takahashi M, Kondo T, Asano Y. Identification of human herpesvirus 6 as a causal agent for exanthema subitum. Lancet 1988; 1:1065-7. 176. Steeper TA, Horwitz CA, Ablashi AV, Salahuddin SZ, Saxinger C, Saltzman R. The spectrum of clinical and laboratory findings resulting from human herpesvirus-6 (HHV-6) in patients with mononucleosis-like illness not resulting from Epstein-Barr virus or cytoegalovirus. Am J Clin Pathol 1990; 93: 776-83. 177. Josephs SF, Buchbinder A, Streicher HZ, et al. Detection of human B-lymphotropic virus (human herpesvirus 6) sequences in B cell lymphoma tissues of three patients. Leukemia 1988; 2:132-5. 178. Kruger GRF, Manak M, Bourgeois N, et al. Persistent active herpesvirus infection associated with atypical polyclonal lymphoproliferation (APL) and malignant lymphoma. Anticancer Res 1989; 9:1457-76. 179. Torelli G, Marasca R, Luppi M, etal. Human herpesvirus6 in human lymphomas: identification of specific sequences in Hodgkin's lymphomas by polymerase chain reaction. Blood 1991; 77: 2251-8. 180. Borisch B, Ellinger K, Neipel F, et al. Lymphadenitis and lymphoproliferative lesions associated with the human herpes virus-6 (HHV-6). VirchowsArch B Cell Pathol 1991; 61:179-87. 181. Levine PH, Jahan N, Murari P,etal. Detection of HHV-6 in tissues involved by sinus histiocytosis with massive lymphadenopathy (Rosai-Dorfman disease).) Infect Dis 1992; 166: 291-5. 182. Hoffman A, Kirn E, Kuerten A, etal. Active human herpesvirus-6 (HHV-6) infection associated with Kikuchi-Fujimoto disease and systemic lupus erythematosus (SLE). In Vivo 1991; 5: 265-9. 183. Hollingsworth HC, Peiper SC, Weiss LM, Raffeld M, Jaffe ES. An investigation of the viral pathogenesis of Kikuchi-Fujimoto disease (KFD): lack of evidence for Epstein-Barr virus or human herpesvirus-6 as the causative agents. Arch Pathol Lab Med 1994; 118: 134-40. 184. Sansonno D, Vita SD, Cornacchiulo V, Carbone A, Boiocchi M, Dammacco F. Detection and distribution of hepatitis C virus-related proteins in lymph nodes of patients with type II mixed cryoglobulinemia and
References 131 neoplastic or non-neoplastic lymphoproliferation. Blood 1996; 88: 4638-45. 185. Ferri C, Civita LL, Longombardo G, F FG, Bombardier! S. Hepatitis C virus and mixed cryglobulinaemia. EurJ Clin Invest 1993; 23: 399-405. 186. Silvestri F, Pipan C, Barillari G, etal. Prevalence of
hepatitis C virus infection in patients with lymphoproliferative disorders. Blood 1996; 87: 4296-301. 187. Luppi M, Longo G, Ferrara MG, etal. Additional neoplasms and HCV infection in low-grade lymphoma of MALT type. BrJ Haematol W96; 94: 373-5.
This page intentionally left blank
12 Molecular biology VI PAPPA AND BD YOUNG
Introduction Burkitt's lymphoma and the c-myc oncogene bc/-1/PRAD-1 gene rearrangement t(14;18) and the bcl-2 oncogene bd-6/laz-3 gene rearrangements t(2;5)(p23;q35) translocation 11q23 abnormalities in lymphoma
133 134 135 137 140 141 142
INTRODUCTION Malignant lymphoma cells usually have clonal chromosomal abnormalities and a number of these are associated with clinical and cellular phenotypes. These are now commonly helpful in diagnosis and prognosis. Recent advances in molecular genetics have led to the identification of a series of genes affected by chromosomal translocations in lymphoma. It is now clear that the proteins encoded by these genes play important roles in the control of apoptosis and cell-cycle progression. The timing and frequency of tumors in animals made transgenic for a single event, such as a translocation, has highlighted the requirement for subsequent genetic changes that act in a co-operative manner to create the fully transformed cell. It is now clear that the control mechanisms of both apoptosis and cell-cycle progression involve a complex series of dynamic protein-protein interactions. Disruptions to these interactions, whether by translocation, mutation or gene deletion can be expected to have profound effects on the behavior of cells. The complex nature of these interactions suggests that it may be possible to reach full tumorigenicity by several different routes. Thus, although there are clear associations between certain chromosomal translocations and lymphoma subgroups (Table 12.1), there may well exist underlying co-operative genetic changes that provide full tumorigenicity1 and we are only beginning to understand how specific translocations become associated with specific lineages.2 Molecular analysis of translocations in lymphoma,
fo/-10 and MALT lymphoma
142
Gene mutations in lymphoma
142
Gene amplification, the REL/NFKB transcription factors and bcl-3
144
Conclusion
144
References
144
Table 12.1
Chromosome translocations in lymphoma
t(8;14) t(2;8), t(8;22)
c-myc/IgH c-myc/\gk, IgK
Burkitt's
t(14;18)
bcl-2l\$\
Follicular
t(2;5)
npm/alk
Anaplastic large cell
t(3;14)
bc/-6/lgH + others
Wide range
t(11;14)
(fo/-1) cyclin D1/lgH
Mantle cell
t(l;H)
bd-W
MALT lymphoma
t(14;19)
bd-3/REl, NFKB
Bcell NHLandCLL
t(9;14)
PAX 5/lgH lymphoma
Lymphoplasmacytoid
Ig = immunoglobulin, MALT = mucosa-associated lymphoid tissue, NHL = non-Hodgkin's lymphoma, CLL = chronic lymphocytic leukemia.
leukemia and certain solid tumors3 has shown that they fall into general classes. The first type involves the activation of a proto-oncogene by juxtaposition into the immunoglobulin (Ig) or T cell receptor (TCR) genes, usually resulting in aberrant expression from the coding exons, which are normally intact. The second type involves breakage and rejoining of an intronic sequence in such a way that an in-frame fused mRNA is expressed, resulting in a fused chimaeric protein. In a third type, translocation is associated with mutation and loss of function. All types of event are found in lymphomas.
134 Molecular biology
Cytogenetically, lymphomas frequently reveal few or only one abnormalities and there is little evidence of microsatellite instability or a general defect in DNA repair mechanisms.4 The identification of the molecular basis of these events has provided a wide range of investigative techniques, including the polymerase chain reaction for the junctional sequences and immunohistochemical staining for the proteins encoded by genes involved in translocations. The addition of these approaches to conventional chromosome analysis has helped to clarify the contribution of these events to the various pathological subtypes of lymphomas.
BURKITT'S LYMPHOMA AND THE c-myc ONCOGENE Burkitt's lymphoma has characteristic chromosomal translocations that involve a recombination between the Ig heavy-chain locus and the c-myc oncogene.5"11 The human Ig heavy-chain locus is oriented on chromosome 14 with the variable regions telomeric to the constant regions. The breakpoint on chromosome 14 is often in the switch region located 5' to the C|l constant region gene. The breakpoints on chromosome 8 are usually in or around the first non-coding exon of the c-myc gene. The usual result of this translocation is to create a fusion sequence with c-myc joined with the S(l region in opposite transcriptional directions. The two variant chromosome translocations, t(2;8) and t(8;22), observed in about 10 per cent of Burkitt's lymphoma are due to a recombination between the c-myc locus and the kappa light-chain gene (IgK)12 on chromosome 2 or the lambda light-chain gene (IgX)13 on chromosome 22. In contrast to the heavy-chain locus, both light-chain loci are oriented with their variable regions centromeric to their constant regions and the breakpoints usually lie 5' to the joining region segments, thus leaving the } region enhancer sequences intact. The corresponding breakpoints on chromosome 8 occur 3' to the c-myc gene at a distance that can range from 400 bp to over 100 kb from the c-myc polyadenylation site.14-16 Occasionally the distance can be sufficient to result in a Cytogenetically different breakpoint on chromosome 817 and it remains possible that a different gene may be involved. Indeed, the analogous variant translocation in murine plasmacytomas involves a locus other than c-myc, named pvt-\ (plasmacytoma variant translocation).18 Myc-related translocations are also a feature of acquired immunodeficiency syndrome (AIDS)-related B cell lymphomas,19'20 where it is believed that they represent a stage in the progression toward EBV-independent proliferation. A further consequence of translocation of the c-myc locus can be the introduction of point mutations into its promoter region, the first non-coding exon21,22 or either
of the two coding exons.23 Additional evidence of somatic mutation, particularly in the endemic form of Burkitt's lymphoma has been demonstrated by restriction enzyme mapping. PvuII restriction enzyme digestion of DNA from 13 endemic lymphomas revealed 10 abnormally sized c-myc alleles indicating a high level of point mutation around the site of transcription termination.22 Somatic mutation is a common means of generating antibody diversity during normal VDJ rearrangement and similar mechanisms may be directed at the translocated c-myc allele. Since the other c-myc allele on chromosome 8 is usually in a germline configuration, it can be used as an internal control to evaluate the deregulation of the translocated allele. Most studies indicate that the translocated allele continues to be expressed, whereas the normal allele is switched off.24"26 The extinction of the normal allele and expression of the translocated allele appears to happen independently of whether the breakpoint is within or 5' to the c-myc gene.27'28 A similar analysis of a variant t(2;8) translocation indicated that the rearranged allele continued to be expressed.12 Although this deregulation leads to more c-myc mRNA2930 and more c-Myc protein31'32 in Burkitt's lymphoma cells than other lymphoblastoid cell lines, such quantitative changes may be less important than the inability of the rearranged gene to respond normally to regulatory factors. There is epidemiological and clinical evidence that there are two distinct forms of Burkitt's lymphoma, endemic and sporadic.33 Although both forms have the typical chromosome translocations, there appear to be subtle differences in the mode of activation, with the sporadic form often involving deletion of c-myc regulatory sequences, whereas the endemic form frequently involves point mutations or insertions.22 It has been suggested that cells from the endemic cases may be derived from the germinal center of lymph nodes and that cells from the sporadic cases may be derived from the bone marrow.33 It can be speculated that the endemic form is derived from cells at an earlier stage in the B cell lineage than the sporadic form. Thus the endemic translocations, which have a proportion of light-chain Ig gene involvement and a high rate of mutation, could represent errors in VDJ recombination, an event which takes place early in B cell development.34 The t(8;14) translocation found in AIDS-associated Burkitt's lymphoma appears to resemble more closely that found in the endemic form.35,36 Rearrangements of the c-myc gene have also been reported in other types of high-grade lymphoma, but with a much lower incidence than in Burkitt's.37,38 Direct evidence that deregulated c-myc genes may have a causative role in Burkitt's lymphoma has come from the study of transgenic mice. Animals transgenic for c-myc linked to a steroid inducible promoter often have breast carcinomas,39 whereas mice transgenic for c-myc linked to an Ig enhancer frequently develop
£c/-1/PRAD-1 gene rearrangement 135
lymphomas.40 It is interesting that there is a definite latency period for tumor development, during which further genetic changes take place, implying that c-myc activation alone is insufficient for full tumorigenicity. The c-myc gene encodes a 47-kDa protein, which is concentrated in the nucleus and which is known to have DNA binding properties, albeit at a high concentration.41 It is highly conserved between mouse and man, and is part of a gene family which includes N-rayc and L-myc. The latter have so far been implicated in only one case of lymphoma, where expression of N-myc was detected in an unclassified high-grade T cell lymphoma, which was also observed to have a breakpoint at the N-rayc locus on chromosome 2.42 Many studies have shown that c-Myc promotes cell cycle progression and inhibits differentiation (see review43). Withdrawal of growth factors from normal cells results in the down regulation of c-Myc and the accumulation of cells at the Gl/S boundary.44,45 Enforced c-Myc expression can overcome such cell-cycle arrest and drive cells into S phase.46 Ablation of c-Myc by antisense oligonucleotides or expression constructs blocks entry into S phase.47,48 c-Myc protein may therefore be regarded as a positive regulator of cell-cycle progression with its function being essential for progression through Gl into S phase. The presence of two sequence motifs in c-Myc that mediate protein-protein interactions implies that c-Myc functions as a transcription factor. In particular, the heptad repeat of leucine residues (leucine zipper or LZ) and the helix-loop-helix (HLH) domain are both features known to mediate protein-protein interactions between transcription factors. c-Myc also contains an adjacent basic region (b region), which makes direct sequencespecific interactions with DNA. Thus c-Myc contains an extended bHLH-LZ region of a type also found in a group of closely related transcription factors, including USF,49 TFE3,50 TFEB,51 AP-452 and the c-Myc dimerization partner Max.53,54 c-Myc protein is known to form heterodimers with Max,55 an interaction which greatly enhances the sequence-specific binding of c-Myc to DNA.56 Max can homodimerize56 and heterodimerize with two additional bHLH-LZ proteins, Mxil57 and Mad58 forming complexes that can bind to DNA. Although the c-Myc/Max interaction is central to c-Myc function, other interactions with c-Myc have been demonstrated. For example, c-Myc has been shown to interact with the retinoblastoma-related protein P107 (a suppressor of cell growth).59,60 Interactions between c-Myc and PI07 require the N-terminus of c-Myc and, in some Burkitt's lymphomas, mutations in the N-terminus of c-Myc have been shown to render c-Myc resistant to P107-mediated suppression.59 Thus events that free c-Myc from the negative regulation of PI07 may contribute to tumorigenesis in some lymphomas. A number of experimental systems have been used to demonstrate that the enforced expression of c-Myc not
only results in cell-cycle progression but can also induce apoptosis. For example, enforced c-Myc expression in 32D.3 myeloid progenitor cells induces apoptosis in the absence of interleukin-3-(IL-3).44 Upon withdrawal of IL-3 these cells normally down regulate c-Myc and accumulate at the GO/Gl boundary. These and other experiments have led to the idea that c-Myc expression, in circumstances when the cell would normally be quiescent, leads to apoptosis.43,61 It may seem paradoxical that activation of c-Myc, which can lead to apoptosis, is so strongly linked to tumors such as Burkitt's lymphoma. However, it is clear from c-Myc transgenic mice that additional events are required before tumors develop. Additionally, a series of genes (bd-2,pim-l, bmi-l, raf-l, ras and abl) has been identified as co-operating with c-Myc to accelerate tumorigenesis.43 Some of these genes (bd-2,pim-l, raf-l) can suppress apoptosis and it would seem likely therefore that c-Myc-mediated tumorigenesis is due to c-Myc-induced cell-cycle progression with concomitant suppression of apoptosis provided by other genetic events.
/7C/-1/PRAD-1 GENE REARRANGEMENT The t(l I;14)(ql3;q32) was identified as a recurring cytogenetic abnormality in the lymphoproliferative diseases,62 and subsequently shown to be associated with a subset of diffuse small B cell non-Hodgkin's lymphomas variously described as centrocytic lymphoma, mantle zone lymphoma, intermediate differentiated lymphocytic lymphoma and presently under the term mantle cell lymphoma.63-66 The t(ll;14) translocation has also been found in 2-6 per cent of B cell chronic lymphocytic leukemias (B-CLLs), in about 18 per cent of B cell prolymphocytic leukemias (B-PLLs), multiple myeloma67 and in approximately 15 per cent of splenic lymphomas with villous lymphocytes.68 The breakpoints were cloned in 1984 from cases reported to be CLL and diffuse large cell lymphoma.69'70 The breakpoints on chromosome Ilql3 showed tight clustering in a region called bcl-l (B cell lymphoma leukemia I).71 Most of them involve a 2 kb locus called MTC (major translocation cluster).72,74 The breakpoints on chromosome 14 involve the immunoglobulin heavychain joining regions (J H ) and it was hypothesized that the juxtaposition to enhancer elements associated with the IgH locus would affect the transcription of a gene near the bd-1 locus. However, despite intensive searches surrounding the breakpoint and substantial chromosomal walks, no deregulated transcriptional unit could be readily identified. A number of parathyroid adenomas are characterized by clonal rearrangements of the parathyroid hormone gene at Ilpl5. This proved to be an interchromosomal translocation with Ilql3 that overexpressed a newly
136 Molecular biology
identified gene located in this region, designated PRAD-1 (parathyroid adenomatosis) gene.75 PRAD-1 encodes a protein of 295 amino acids with sequence similarities to cyclins, and which can complex with p34cdc2 and induce its kinase activity.76 The PRAD-1 gene was considered a good candidate for the bd-1 linked oncogene, especially as the same gene had been identified in two other systems. A p36cyl has been observed in mouse macrophages and is induced by CSF-1 in the Gl phase of the cell cycle.77 p36cyl is the murine equivalent of PRAD-1 and also associates with a cdcl-related polypeptide. Additionally, genes from a human glioblastoma introduced into a budding yeast strain, mutant in all three of the known yeast Gl cyclins, complemented the defect and revealed a new subclass, cyclin Dl.78 Cyclin Dl is abundant in human glioblastoma and is identical to PRAD-1. The PRAD-1/cyclinDl (CCND1) gene on Ilql3 is located 120 kb telomeric from thebcl-1/MTC locus,76,79-81 although there are two minor translocation clusters (mTCs) that are less frequently involved. mTCl is localized 22 kb telomeric of MTC82 and mTC2 maps to the 5' flanking region of cyclin D1 gene.83 Southern blots using probes spanning the 110 kb distance between the bd-l MTC and the PRAD-1 gene, can detect bd-l rearrangements in up to 73 per cent of cases of mantle cell lymphomas.84'85 The relatively tight clustering of bd-1 MTC breakpoints suggests that the polymerase chain reaction (PCR) may be a suitable technique for the detection of the t(ll;14) in clinical samples, providing a diagnostic tool for the differential diagnosis of mantle cell lymphomas.86 Cyclin Dl mRNA is overexpressed in B cell malignancies87,88 with Ilql3 abnormalities and in the vast majority of mantle cell lymphomas analysed so far.79,89,90 Recent evidence suggests that the t(ll;14)(ql3;q32) identifies a cytologically atypical subset of B-CLL, characterized by a frequent cytological and cytogenetic evolution and by a distinct immunophenotype with a bright-staining pattern for surface immunoglobulins (SIg), CD 5 positivity and rare CD 23 expression, sharing some biological features with mantle cell lymphoma.91 Moreover, the cyclin Dl protein has been shown by monoclonal and polyclonal cyclin Dl antibodies to be overexpressed in all cases of mantle cell lymphomas examined with no expression in control tissue.92 The cyclin Dl gene is normally silent in T and B lymphocytes.93 The tight linkage of the cyclin Dl gene with bd-l, without intervening CpG islands, and its overexpression in B cell lymphomas with bd-l rearrangement and cell lines with the t(ll;14)(ql3;q32), provide strong evidence that the cyclin D1 gene is the bd-1 oncogene. The fact that cyclin D1 overexpression has been shown in almost all cases of mantle cell lymphomas79,88,89,94 even in the absence of the t(ll;14)(ql3;q32) raises questions about the mechanisms of cyclin Dl overexpression in these cases. In some examples, the cyclin Dl transcript is
truncated at its 3' untranslated region, resulting in loss of AUUUA sequences involved in mRNA stability. The consequence is the production of more stable transcripts, implying a mechanism of post-transcriptional derangement in the activation of cyclin Dl.88,95,96 This mechanism may account in part for the overexpression of the cyclin Dl mRNA in the absence of chromosomal translocations, which is observed in some B cell tumors and breast cancers. Furthermore, overexpressed cyclin Dl transcripts seem to contain no mutations.97 Overexpression of cyclin Dl has also been reported in numerous solid tumors, including breast cancer,98-100 oesophageal carcinoma,101 colon cancer102 and hepatocarcinoma.103 It is therefore likely that overexpression of cyclin Dl is associated with the pathogenesis of these diseases. Overexpression of cyclin Dl in rat embryo fibroblasts by retroviral transduction, resulted in a decrease in the duration of the Gl phase and decreased cell size. These cells produced tumors when injected into nude mice, but with a long latency period and perturbed the expression of several cellular growth-related genes, including c-myc, c-jun and cyclin A, but not cyclin D3.104 Furthermore, transfection of a Gl cyclin into fibroblasts or myeloid cell lines results in a shortening of Gl phase, but the total length of the cell cycle remains unchanged because of a compensatory prolongation of the S phase.105'106 Moreover, microinjection of cyclin Dl antibodies into fibroblasts arrests the cell at the Gl phase.107 The transforming activity of cyclin Dl has been shown in primary rat embryonic fibroblasts by co-transfection with c-myc or Ha-ras genes. Cyclin Dl, together with Ha-ras has been shown to transform primary cells and enable them to grow fibrosarcomas in nude mice.108 Cyclin Dl can also transform BRK cells in the presence of activated Ha-ras and an altered E1A protein.109 Strong evidence was provided, using osteosarcoma cells deficient for the retinoblastoma suppressing protein pRB, that all three D-type cyclins may exert their function as parts of a multiprotein complex involving interaction with pRB.110-113 It has been suggested that cyclin Dl functions by inactivating the inhibitory effect of Rb on cell-cycle progression. There is conflicting evidence, however, on the ability of cyclin Dl to phosphorylate Rb in vivo, in mammalian cells.110-111 The observation that Dtype cyclins contain the C-X-L-X-E motif, also found in the DNA viral oncoproteins SV40 T antigen, adenovirus El A and human papilloma virus E7, may provide a clue to the mechanism of interaction of cyclin Dl with Rb.114 This motif is required for the binding of cyclin Dl to Rb.110 The same regions of the Rb protein interact with the E2F transcription factor, which regulates the expression of several cellular genes involved in progression through the cell cycle.115 Cyclin Dl might affect the equilibration between the Rb-E2F complex and free E2F, acting as a dominant transcription represser.116 Overexpression of cyclin Dl would increase the amount of free E2F or decrease the amount of Rb-E2f complex,
t(14;18)and the fo/-2 oncogene 137
which might enhance the expression of several genes like c-myCy c-jun and cyclin A. This hypothesis is supported by the observation that transient overexpression of cyclin Dl decreases the amount of Rb-E2F complex in mammalian and insect cell systems.111,113 There is little evidence for alterations in the Rb gene in lymphomas. A comprehensive analysis of the Rb gene in a series of lowand high-grade lymphomas failed to reveal any mutations, although there was evidence for reduced levels of Rb protein in a few cases.117 The oncogenic potential of the cyclin Dl gene has been demonstrated in E|Ll-cyclin Dl transgenic mice that overexpress cyclin Dl, in B and T lymphocytes.118,119 The cell-cycle activity of these lymphocytes, their size and mitogen responsiveness are normal, but young transgenic animals contain fewer mature B and T cells. Although spontaneous tumors are infrequent, lymphomagenesis was more rapid in mice overexpressing cyclin Dl with c-Myc compared to mice expressing either transgene alone. Moreover, the spontaneous lymphomas of c-myc transgenic animals often expressed the endogenous cyclin Dl.118 Similarly, in another study, the cyclin Dl/N-myc or L-myc double transgenic mice developed clonal pre-B and B cell lymphomas. Furthermore, the crossing of cyclin Dl transgenic mice, with Eu/L-myc transgenics that overexpress L-myc in B and T cell populations, but predominantly develop T cell tumors, leads in double transgenics to B cell neoplasia. These findings establish cyclin Dl, as a proto-oncogene whose activity depends on a specific cell type as well as on a specific co-operating partner.119
t(14;18) AND THE bcl-2 ONCOGENE The t(14;18)(q32;q21) translocation, the most common translocation within human lymphoid malignancies,120,121 characterizes approximately 85 per cent of follicular and 20 per cent of diffuse B cell lymphomas.1,121,122 Molecular cloning of the t(14;18) breakpoint revealed a putative proto-oncogene, bcl-2 on chromosome 18q21,123~125 juxtaposed to one of the six immunoglobulin heavy-chain joining regions at 14q32. The bcl-2 gene consists of three exons. The first intron is 220 bp and a second intron of 370 kb separates the two coding exons II and III. Two proteins are potentially encoded from the 5.5 kb and the 3.5 kb mRNA transcripts: one encoding a 29 kDa 239 amino-acid bd-2a protein and the other a 205 aminoacid bd-2b protein that lacks a hydrophobic carboxyl tail, although this protein has never been seen in vivo in any appreciable amounts.126 The region of 19 hydrophobic amino-acids near the carboxyl terminus is followed by two charged residues that may serve to anchor the protein in membranes.127 The bcl-2 protein resides in the nuclear envelope, parts of the endoplasmic reticulum and outer mitochondrial membrane, but not in a variety
of other intracellular membrane compartments including the plasma membrane.127,128 In 60 per cent of cases the breakpoint on chromosome 18 falls in a 500 bp area in the 3' untranslated portion of the third exon of bcl-2 known as mbr (major breakpoint region)129,130 and in 25 per cent of cases it occurs 20 kb downstream in an untranscribed region known as mcr (minor cluster region).131,132 The breakpoints on chromosome 14 occur adjacent to the joining region of the immunoglobulin heavy-chain gene. Although the open reading frame of bcl-2 remains intact during the translocation, somatic mutations within it have been described.133 The t(14;18) is not confined to follicular lymphomas as it has also been reported in some highgrade lymphomas which appear to have transformed from low-grade follicular lymphomas,134 in a small percentage of B-CLL with breakpoints mapping between mbr and mcr,135 in de novo ALL of L2 or L3 FAB subtype,136'138 and in a number of cases of Hodgkin's disease.139-141 There are also conflicting reports about the detection of the t(14;18) in cases of benign follicular hyperplasia.142'143 Variant t(2;18)(pll;q21) and t(18;22)(q21;qll) translocations have also been described in B-CLLs144 and in follicular lymphomas.145'146 Molecular studies of these variant translocations have shown that the chromosomal breakpoints consistently map to the 5' region of the bd2 gene (known as the VCR region for variant cluster region) on chromosome 18 and within the K or A, lightchain loci on chromosomes 2 and 22, respectively.144'147'148 A variant t(2;18)(pll;q21) translocation was described in a follicular lymphoma resulting in the juxtaposition of a JK segment to a chromosome 18 transcriptional unit located 10 kb upstream of bcl-2, termed the FVT-1 (for follicular variant translocation) gene.149 Moreover, evidence has been found for multiple rearrangements affecting the mbr, mcr and VCR regions.150'151 The reciprocal partner on chromosome 14, the immunoglobulin heavy-chain locus (IgH) on the derivative 14 chromosome has been demonstrated to bear deletions in 76 per cent of follicular lymphomas, probably as a consequence of non-physiological activation of the recombinases involved in class switching.152 The lack of Ig detection, despite the mature origin of follicular lymphomas, may be explained in some of the cases by the inactivation of both IgH chain genes due to translocation of one allele in combination with deletions or defective rearrangements of the other allele.153 The tight clustering of breakpoints on chromosomes 14 and 18 during the translocation has allowed the use of the PCR technique for the detection of t(14;18)-bearing cells using genomic tumor DNA. The level of sensitivity obtained (up to 106) make this approach valuable for the detection of minimal residual disease.154-159 The presence of the translocation can be demonstrated also by Southern blotting using probes to various bcl-2 breakpoints, the main limitation being the requirement of
138 Molecular biology
frozen tissue for high molecular weight DNA extraction and the reduced sensitivity compared to PCR.159'160 Since PCR examines shorter lengths of DNA than Southern analysis, PCR-based tests may be more susceptible to microheterogeneity in breakpoint location.160 Most studies indicate the use of a combination of cytogenetics, Southern analysis and PCR for the most reliable detection of the t(14;18).160'162 Additionally, pulsed-field gel electrophoresis has proved to be the most informative process compared with standard methodology.163 The PCR technique has allowed the detection of t(14;18)-translocation-bearing cells in the peripheral blood or bone marrow in patients with localized (stage I and II) disease at diagnosis and in long-term remission after radiation therapy.164'165 Moreover, the presence of the t(14;18) translocation has been demonstrated at the completion of treatment as well as in long-term remission of advanced stage disease.166-169 The prognostic significance of the t(14;18) translocation has been the subject of controversy due probably to differences in the levels of sensitivity obtained by PCR. In some studies the presence of the t(14;18) translocation has been associated with a poor response to treatment and short survival,121 although for others the t(14;18) translocation and high bcl-2 protein expression did not correlate with clinical outcome.168-174 Other cytogenetic abnormalities such as chromosomal breaks higher than 6, abnormalities of chromosome regions lp21-22, 6q23—26 or 17p have been associated recently with short survival.174 The detection of the t(14;18) translocation in the reinfused bone marrow autograft in patients with follicular lymphoma treated with highdose treatment, is associated with shorter remission duration,175-177 and its presence at follow-up but probably not early after transplantation, is associated with a higher risk of relapse.177 The mechanisms by which these abnormalities influence survival in patients is unclear and may result from alterations in drug sensitivity and resistance due to kc/-2.178'180 It has been postulated that the t(14;18)(q32;q21) translocation event takes place in early B cells at the time of the D to J rearrangement. At the junction on both der!4 and derlS chromosomes, short segments of random insertions, up to 24 nucleotides in length, have been observed and considered to represent N insertions, normally found at the VD and DJ junctions after IgH rearrangement.130 This supports the hypothesis that VDI recombinase is involved in the generation of the breakpoints on chromosome 14.123 However the bcl-2 regions on chromosome 18 possess no combining heptamerspacer-nonamer motifs surrounding these breakpoints. The mbr region contains several sequence elements that could potentially infer genetic instability or facilitate homologous recombination. A polypurine-polypyrimidine stretch, with a potential to form alternative DNA structures (H-DNA or triplicates) is present in both the mbr and mcr regions. These polypurine tracts can act as
targets for mammalian endonucleases and are found around other genetic hot spots.181'183 Similar repeats of G-rich tetranucleotides have been implicated in immunoglobulin class switching.184'185 The mbr region contains a Chi-like octamer similar to the recombination signal in Escerichia coli and the human minisatellite core sequence, which could activate a Rec BCD-like mechanism.186"188 More recently, an SI nuclease-sensitive site was found that can be the target of an endogenous nuclease present in early B cells. A 45 kDa protein has been identified which binds to the Chi-like polypurine-polypyrimidine tract within mbr and mcr as well as to corresponding Ig sequences, and may play a role in homologous site-specific recombination.189 The juxtaposition of the bcl-2 oncogene with powerful enhancer elements within the IgH locus, markedly deregulates the gene by altering both its transcription and the efficiency of RNA processing, resulting in elevated amounts of bd-2/Ig chimaeric RNA, giving rise to overexpression of the bcl-2 protein.190'192 In normal lymph nodes the bcl-2 protein is most abundant in the long-lived recirculating B cells of the follicular mantle.192'194 bcl-2 is absent from the centroblasts and centrocytes destined to die within the dark zone and basal light zone. In parallel, surviving T cells within the medulla of the thymus demonstrate substantial bcl-2 staining while those in the cortex are negative.193'194 Overexpression of the bcl-2 protein has also been shown in follicular and diffuse B cell lymphomas without concurrent bcl-2 rearrangement in small lymphocytic, mantle zone lymphomas, CLL, plasma cell dyscrasias, chronic and acute myelogenous leukemia, and breast cancer, suggesting other mechanisms of deregulation of bcl-2.195~m As recently shown, the distribution of bcl-2 mRNA is roughly reciprocal to that of the protein with intense hybridization signal in germinal centers and almost absent in mantle zones. Discordant bcl-2 RNA and protein levels were also observed in tonsillar epithelial cells and cortical thymocytes. Importantly, follicular lymphomas and cell lines with the t(14;18) have concordant and abundant bcl-2 mRNA and protein expression, suggesting disruption of translational control mechanisms in follicular lymphomas.195 Overexpression of bcl-2 protein results in prevention of apoptosis or programmed cell death in selected hematopoietic cell lines following deprivation of IL-3, IL4 or granulocyte-macrophage colony stimulating factor (GM-CSF).199'203 Moreover studies of bcl-2 function in cultured postmitotic neurons have established a role for this gene in the suppression of apoptosis in complete absence of cell proliferation. Microinjection of bcl-2 expression plasmids into nerve growth factor (NGF)dependent neurons has been shown to delay markedly the rate of cell death that occurs upon removal of the neurotrophic factors from cultures.204 Overexpression of bcl-2 can block apoptosis induced by y-irradiation, glucocorticoids and a variety of chemotherapeutic
t(14;18)and the bc/-2 oncogene 139 agents.205 208 bcl-2 also blocks apoptosis that accompanies terminal differentiation of myeloid leukemic cell lines without affecting the differentiation process.209 bcl-2 has also been implicated in positive selection of thymocytes and in maintenance of B cell memory through its effect on survival at specific stages of lymphoid development and differentiation.210-212 A distinct biological effect of the bcl-2 protein in different cell types has been demonstrated by a paradoxical inhibition of cell growth in solid tumor cell lines.213 bcl-2 has also been implicated in the development of viral persistence and the pathogenesis of virus-associated malignancies as shown through induction of bcl-2 expression by the Epstein-Barr virus latent membrane protein 1, resulting in protection from apoptosis of B cells.214 Genetic analysis of apoptosis in the nematode worm C. elegans has identified three genes that regulate cell death, ced-3, ced-4 and ced-9. ced-9 suppresses apoptosis and has sequence similarity to bcl-2, and human bcl-2 can rescue ced-9 deficient worms.215'216 Thus some functions of bcl-2 are extremely well conserved throughout phylogeny. In contrast to ced-9 and bcl-2, ced-3 and ced-4 are involved in the initiation or execution of cell death, as loss of function mutations in these genes results in adult worms with additional cells.217 The oncogenic potential of the t(14;18) translocation and resulting bcl-2 overexpression have been demonstrated by transfection of bcl-2 constructs into NIH3T3 fibroblasts, which develop tumors when injected into mice with a latency period of approximately 4 weeks.218 Moreover &c/-2/IgH transgenic mice develop a polyclonal follicular lymphoproliferation of small resting IgM/IgD cells, which progress to immunoblastic lymphoma, by additional genetic alterations involving c-myc rearrangement.219-220 Most of the lymphomas developing in the E\Ji/bcl-2 transgene involve predominantly the B cell lineage and their immunophenotype (Sca-1, CD 4, Thy-1, CD 34, CD 45) is consistent with an origin very early in B lymphoid development.221 These observations suggest that the t(14;18) forms a basis for lymphomagenesis but is not sufficient itself to trigger the neoplastic process. In support of this hypothesis is the observation that the t(14;18) translocation can also be found in blood B cells of normal individuals, and that the frequency increases in the spleen and the blood with age, suggesting a multistep process of lymphomagenesis involving the t(14;18).222-223 In a bcl-2 transgenic line expressing high levels of bcl-2 protein in both cortical and medullary thymocytes, the immature thymocytes are resistant to apoptotic stimuli and have prolonged survival. It was also found that a proportion of thymocytes and peripheral T cells escape the process of negative antigenic selection, eliminating normally autoreactive T cells during thymocyte maturation.224 T cells in the peripheral lymphoid tissues are moderately elevated despite enhanced survival in vitro and there is a relatively high proportion of thymocytes with a mature
phenotype.225 Moreover, a bcl-2 transgene driven by the kk promoter results in mice that overexpress bcl-2 within the thymus and in the peripheral T cells, and which develop peripheral T cell lymphomas predominantly of diffuse large cell type at a mean age of 18 months.226 These data emphasize the tumorigenic potential of repression of cell death in multiple lineages. The role of bcl-2 in the normal control of tissue homeostasis was assessed in bcl-2 knockout mice, which complete embryonic development, but display growth retardation and early postnatal mortality. As expected differentiation of the lymphoid lineage is initially normal but thymus and spleen undergo massive apoptotic involution, resulting in lymphopenia. The animals die as a consequence of renal failure, the result of severe polycystic kidney disease and display hypopigmented hair, indicating a defect in redox-regulated melanin synthesis. The mechanism through which bcl-2 exerts its antiapoptotic function is not known but several possibilities have been explored. The localization of bcl-2 protein in the membranes of nuclear envelope and endoplasmic reticulum (ER) compartments in patches is highly reminiscent of nuclear pore complexes (NPCs),228 raising the possibility of a role for bcl-2 in some aspect of nuclear transport, NPC formation or nuclear envelope assembly and maintenance. The relevance of oxidative phosphorylation to the antiapoptotic function of bcl-2 has been assessed in human fibroblast cell lines lacking mitochondrial DNA. These experiments showed no requirement for oxidative phosphorylation for the induction of apoptosis or for the death represser activity of bcl-2.229 bcl-2 overproduction in a rat pheochromocytoma cell line, PC 12, did not correlate with the levels of ATP or oxygen consumption despite their resistance to apoptosis.230 However, bcl-2 overexpression and increased mitochondrial activity are closely related properties of cell lines resistant to apoptosis induced by glucocorticoids.231 Another important aspect is the role of Ca2+ in apoptosis, since Ca2+-dependent endonucleases may be involved in the internucleosomal DNA digestion typical of apoptotic cells, bcl-2 blocks Ca2+-ionophore-induced apoptosis in thymocytes, T cell leukemia lines and PC 12 cells. Overproduction of bcl-2, however, does not prevent rises in intracellular Ca2+, suggesting that bcl-2 blocks apoptosis downstream of this event. Moreover, lymphokine withdrawal, in IL-3-dependent cell lines results in gradual loss of Ca2+ from the ER and rise in the Ca2+ in the mitochondria. Overexpression of bcl-2 reverses these effects, suggesting that bcl-2 can influence Ca2+ partitioning.232 In favor of this hypothesis are the observations that glucocorticoid treatment of T cells results in massive loss of Ca2+ from the ER prior to apoptosis.233 and overexpression of an ER Ca2+-binding protein, calbindin-D, in glucocorticoid-sensitive T cells delays apoptosis, presumably by allowing ER to retain its Ca2+.234 Another attractive hypothesis is the role of oxidative injury in the
140 Molecular biology
induction of cell death, and the finding that bd-2 blocks the accumulation of lipid peroxides and possibly other reactive oxygen species in at least some settings.235 The concept of bd-2 regulating apoptosis through a dynamic process of positive and negative interactions with other proteins led to a search for other proteins with a structural similarity to bd-2. Viral bd-2 homologs, like the BHRF1 in the EBV236 and the LMW5HL in the African swine fever virus,237 can both prevent apoptosis, perhaps explaining the latency and persistence of some viral infections. However, cDNAs have been cloned recently for several novel human genes, revealing a family of bcl/-2-related members. One of these, called bax, has a six-exon structure and demonstrates a complex pattern of alternative RNA splicing that would encode several membrane and cytosolic proteins. Bax protein has been shown to homodimerize and heterodimerize with bd-2 in vivo. Bax overexpression accelerates apoptosis induced by cytokine deprivation in IL-3-dependent cell lines and counters the antiapoptotic function of bd-2. The ratio of bd-2 to Bax determines survival or death upon apoptotic stimuli.238 Another member of the family termed bd-x produces, by alternative splicing, two distinct bd-x mRNAs and encodes a 241 amino-acid protein with 74 per cent homology to bd-2. The protein product of the large mRNA, bd-xL, stably transfected into IL-3-dependent cell lines, inhibits cell death as effectively as bd-2. The second mRNA species, bd-xS encodes a protein that inhibits the ability of bd-2 to enhance the survival of growth-factordeprived cells.239 Other members of the family include the md-l gene and the Al gene isolated from cDNA libraries, derived from myeloid leukemic and normal cells, respectively.240-242 Additionally, the bad gene encodes a 204 amino-acid protein, which forms heterodimers with bd-2 and bd-xL. Its overexpression in IL-3dependent cells abolishes the protective effect of bd-xL upon IL-3 deprivation and promotes apoptosis, probably by competing with Bax for bd-xL, resulting in reduction of the bd-xL/Eax heterodimers and an increase in the amount of the death-accelerating protein Bax.243 Another new member of the bd-2 family is the bok gene, encoding a protein of 211 aminoacids, 25 per cent identical to bd-2. Three closely related bak genes exist: bak mapping to chromosome 6, bak-2 to chromosome 20 and bak-3 to chromosome 11. Like Bax, the Bak gene product enhances apoptosis following apoptotic stimuli and, unlike Bax, it inhibits cell death in an Epstein-Barr virus (EBV)-transformed cell line. The Bak protein can also bind to an apoptosis-inhibiting adenovirus protein E1B19K.244-246 In the emerging family of foc/-2-related proteins, two domains, termed bd-2 homology 1 and 2 (BH1 and BH2), have been identified.247 Site-specific mutagenesis of bd-2 established these two domains as novel dimerization motifs. Substitution of Gly 145 in the BH1 domain or Trp 188 in BH2 completely abrogates the
death represser activity of bd-2 upon apoptotic stimuli, such as IL-3 deprivation, y-irradiation and glucocorticoids. Mutations that affected the function of bd-2 also disrupted its heterodimerization with Bax while preserving its ability to form homodimers, suggesting that these domains are functionally important and that bd-2 exerts its function through heterodimerization with Bax.247
&c/-6//az-3 GENE REARRANGEMENTS The reciprocal translocation t(3;22)(q27;qll) was first identified248 in nine examples from a large lymphoma series (187 specimens). Most of these tumors had in common a diffuse morphology and predominantly a large cell type. Subsequent cytogenetic analysis confirmed249 and extended250 these results to include t(3;14)(q27;q32) and t(2;3)(p!2;q32) translocations. Molecular analysis with DNA probes from the IGH locus251 resulted in the identification of a gene at the breakpoint on chromosome 3 by four independent groups. This gene, which has been named bd-6252'253 kz-3254 and bd-52K contains a Kruppel-type COOHterminal Cys2-His2 zinc-finger region. This region has been shown to bind DNA and a consensus DNA binding sequence has been identified256 as (T/A)NCTTTCNAGG(A/G)AT, a sequence which can be found in the 5' region of certain genes, bd-6 also contains a NH2terminal region with significant homology252 to other zinc-finger transcription factors such as the ZFPJS protein, which regulates the major histocompatibility complex II promoter, the Tramtrack (ttk) and Broadcomplex (Br-c) proteins in Drosophila that regulate developmental transcription,257 the human KUP protein258 and the human PLZF protein, which is involved in the t( 11;17) variant translocation in acute promyelocytic leukemia.259 This evolutionarily conserved feature has been named the BTB domain and may be present in up to 40 different Drosophila genes.260 bd-6 mRNA has been detected in cell lines derived from mature B cells, but not from pro-B cells or plasma cells, T cells, other hemopoietic lineages or other tissues. This restricted range of normal expression has been taken252 to suggest that bd-6 is a transcription factor involved in the control of normal B cell differentiation and lymphoid organ development,261 and in the formation of germinal centers.262 The genomic structure of bd-6 encompasses about 26 kb263 and consists of nine exons. bd-6 rearrangements with IgH,251'264 IgK265 and IGX255 have now been molecularly analysed, and it is clear that all these events consistently break bd-6 in the same cluster region of about 10 kb at the 5' end of the gene. Breakpoints are clustered around the first exon and the putative regulatory region of the bd-6 gene is removed during translocation, leading to overexpression of the presumably intact coding region of the gene. Evidence has been found that
t(2;5)(p23;q35) translocation 141
biallellic rearrangements of bd-6 can occur.266 The fact that the IgH and IgA, rearrangements occur in the same region of bd-6 contrasts with the rearrangements affecting bd-2 and c-myc. IgH rearrangements occur 5' to c-myc and 3' to bd-2, whereas IgA rearrangements occur 3' to c-myc and 5' to bcl-2. Thus the configuration of both bd-2 and c-myc in relation to the Ig loci seems to be important, whereas the removal of the 5' untranslated region appears to be the important feature of bd-6 activation. It is now clear, however, that the bd-6 gene can be rearranged with not only the immunoglobulin loci but with a range of other chromosomal sites (Table 1221 '>267>268 Molecular analysis of the t(3;ll)(q27;q23) in a cell line has shown273 that this translocation results in a fusion between bd-6 and BOB1/OBF1, the B cell specific co-activator of octamer binding transcription factors. The regulatory regions upstream of the non-coding exon 2 of bd-6 are replaced by those of BOB1/OBF1, leaving the coding sequence of the bd-6 intact. Similar analysis of the t(3;4)(q27;qll) translocation in a cell line identified the gene product TTF267a fused to bd-6. TTF, which is only expressed in hemopoietic cells, has homology to the RAS superfamily and may define a new subgroup of RHO-like proteins. Sequence analysis of the bd-6 gene coding sequence in both rearranged and non-rearranged cases found no evidence that mutation plays a role in bd6 activation.274 A series of large-scale investigations of the incidence and clinical significance of bd-6 rearrangements in lymphoma has been reported.269,271,275 Approximately 15-20 per cent of all lymphomas were found to have bd-6 rearrangements, assessed by Southern analysis. As expected from previous cytogenetic studies, the greatest incidence (30 per cent,269 45 per cent275) was found in tumors classified as diffuse large cell lymphomas (DLCL). However, rearrangements have also been found in follicular lymphomas269,275 and in a case of CLL.271 In a study of 102 patients with DLLC, 23 cases with bd-6 rearrangements and 21 cases with bd-2 rearrangements Table 12.2
14q32 22q11 2p12 2q23 1q21 4p11 6p21 7p12 8q24 11q13 11q23 12q11 15q21
Chromosomal sites found rearranged with bd-6
IgH IgA IgK
TTF
BOB1
252, 269 252, 269 252, 269 252 252 269, 270 269, 271 269, 272 269 252 273 252 269
were found.276 From the clinical follow-up, it was concluded that rearrangement of bd-6 may be a favorable prognostic marker when compared to rearrangement of bd-2 in this series. The incidence of bd-6 rearrangements in AIDS-related lymphomas has also been investigated277 and these data showed that about 20 per cent of AIDS-DLCL carried rearrangements to bd-6.
t(2;5)(p23;q35) TRANSLOCATION Early cytogenetic investigations identified the t(2;5) translocation as a recurring event278 in tumors referred to as malignant histiocytosis.279 Subsequent analysis280"282 has shown that these tumors should be reinterpreted as large cell anaplastic lymphomas. A particular characteristic of this group is the presence of the Ki-1 (CD 30) surface antigen and it is now clear that a proportion of this group carry the t(2;5) translocation,283 particularly where there is evidence of T cell origin.284,285 Alternative cytogenetic abnormalities in CD 30-positive ALCL include t(2;2).286 Molecular analysis has resulted in the identification of the genes involved in the t(2;5) translocation. It has been shown287 that the translocation results in the in-frame fusion of a gene encoding a protein known as nucleophosmin (NPM) on chromosome 5 to a previously unknown tyrosine kinase gene (ALK) on chromosome 2. The normal ALK protein appears to be a membranespanning tyrosine kinase receptor, the fusion resulting in a chimeric NPM-ALK protein from which the membrane and putative extracellular domains have been lost. Reverse transcription PCR (RT-PCR) analysis of cells known to contain the t(2;5) translocation, including three cell lines and four examples of anaplastic large cell lymphomas (ALCLs), indicated that breakpoints occur in the same introns of the npm and alk genes, resulting in identical fusion junctions in the mRNA. The normal NPM protein is a nucleolar phosphoprotein, involved in the assembly of the small and large ribosomal subunits. NPM expression is cell-cycle regulated with a peak before entry into S phase. It has been postulated that, since the alk gene is normally silent in lymphoid cells, a critical consequence of its fusion to npm is the upregulation of an activated ALK tyrosine kinase domain.287 The availability of DNA probes and an RT-PCR assay for the t(2;5) translocation has facilitated molecular studies of its incidence in large cell anaplastic lymphomas and in related disorders.288 A DNA probe 5' to the npm gene was used289 in Southern analysis of a series of lymphomas. Rearrangements were found in 1/12 T cell ALCL and in 1/2 B cell ALCL. These data contrast with previous cytogenetic results which have suggested that most T cell ALCLs carry the t(2;5) translocation.290 Using RT-PCR291 a similar low incidence has been reported in a study in which a total of 6 of 37 cases of
142 Molecular biology
Ki-1 ALCLs were positive for the t(2;5) translocation. Within this study". 5/17 T cell ALCLs and 1/15 B cell ALCLs were positive. It was noted that the t(2;5) translocation may be strongly associated with pediatric Ki-1 lymphomas since two out of three in this study were positive. However, both studies taken together do not support the idea that there is a strong relationship between Ki-1-positive lymphomas and the t(2;5).291 It should be borne in mind that both strategies (RT-PCR and Southern blotting) may fail to detect rearrangements if breakpoints occur in different regions of either gene. Howevers the current evidence287 suggests that all breakpoints' occur at similar positions in both genes. The similarities between Ki-1-positive ALCL and Hodgkin's disease has prompted investigations of the occurrence of the t(2;5) translocation in the latter disease. Although an RT-PCR study292 indicated the presence of the t(2;5) in 11 of 13 patients with Hodgkin's disease, regardless of subtype, subsequent studies have failed to confirm this association.293294 Antibodies to the ALK protein have been developed and are useful diagnostically.295 These may be added to molecular probes in assessments of the prognostic significance of ALK protein and the NPM-ALK chimera. Current evidence suggests that these may be predictors of a good prognosis.296,297
11q23 ABNORMALITIES IN LYMPHOMA Abnormalities at chromosome band Ilq23 have been recorded in previous studies of lymphoma karyotypes. In a recent survey of 43 cases of non-Hodgkin's lymphomas, 3 out of 21 examples which lacked the t(14;18) translocation had abnormalities affecting Ilq23.298 Previous studies have found 2 out of 27 cases of large cell lymphoma with Ilq23 abnormalities299 and 2 out of 94 cases with the translocation t(ll;14)(q23;q32),300 which has also been noted in a single case report301 and in a lymphoma-derived cell line.302 Two examples of lymphoma subsequent to polycythemia vera have been reported to have Ilq23 abnormalities.303 A relatively high incidence of Ilq23 abnormalities has also been noted in Hodgkin's disease (6 out of 18 cases with abnormal metaphases),304 although subsequent studies have found different clonal abnormalities in Hodgkin's disease. Recently a cell line has been established from a pleural effusion of Hodgkin's disease and shown to contain Ilq23 abnormalities.305 The rck gene, known to be the target of the t(ll;14) translocation in the RCK-8 cell line306 has not yet been reported to be involved in other primary abnormalities at Ilq23. The hrx/mll gene, the target of chromosome translocation in acute leukemia, was investigated307 in 20 lymphoma samples with Ilq23 abnormalities and found to be rearranged in only three. It may therefore be concluded that a low but consistent
proportion of lymphomas have abnormalities to this cytogenetic region and that these events may target a gene other than rck or hrx/mll. Recently, two new genes have been identified as potential targets for translocation at Ilq23 in lymphomas. As discussed above, the bobl gene has been identified273 in a t(3;ll) translocation in a lymphoma-derived cell line resulting in a rearrangement with bcl-6. An example of the t(ll;14)(q23;q32) translocation has been molecularly cloned308 and the gene at Ilq23 identified as a new member of the proprotein convertase gene family (Ipc). The breakpoint was shown to lie within the 3' untranslated tail of the Ipc gene and therefore the breaks do not disrupt the coding region of this gene. It remains uncertain which of these genes, Ipc, bobl or rck are responsible for the majority of the Ilq23 abnormalities in lymphomas.
be/-10 AND MALT LYMPHOMA The molecular mechanisms that underlie lymphomatous progression at chronically inflamed sites, which results in mucosa-associated (MALT) lymphomas, remain poorly understood. However, cytogenetics identified abnormalities on chromosome Ip22 and a translocation t(l;14) as recurrent events309 and associations with more aggressive behavior in vivo and growth in vitro.310'311 The breakpoint of the t(l;14) translocation was cloned and found to exhibit a caspase recruitment domain suggesting a role in apoptosis and proapoptotic activity was seen in vitro.3n MALT lymphoma cells showed a frameshift truncating mutation, which lacked proapoptotic effects. This gene has been designated bcl-W and shows mutations in other cancers (colonic cancer, mesothelioma, germ cell tumors) and may have a role in carcinogenesis beyond that in this rare subtype of lymphomas.
GENE MUTATIONS IN LYMPHOMA Frequent and non-random chromosome 17p alterations and p53 mutations have been observed in patients with lymphoproliferative diseases, suggesting an important role in lymphomagenesis and disease progression. Thirty-seven per cent of patients with large cell lymphoma have been reported to have either loss of whole chromosome 17 or of part of 17 (17p-) or all of the short arm (i!7q).299 Other investigators have reported chromosome 17 abnormalities in 31-44 per cent of patients with large cell lymphoma300312 and such abnormalities were correlated with a grave prognosis and a short survival.313'314 Overexpression of p53 has been shown by immunostaining in 20-50 per cent of diffuse high-grade
Gene mutations in lymphoma 143
lymphomas.315 32° The incidence of mutations of the p53 gene varies from 20 to 30 per cent among different studies,320'323 but as recently shown, the percentage of p53 mutations in the group of high-grade non-Hodgkin's lymphomas is higher (45 per cent) if the analysis includes exons 2-11 and not only the 'hot spot' region (exons 5-8) where most mutations are known to occur.324 Moreover, staining for p53 is not predictive of mutations since cases overexpressing p53 in the absence of mutations have been described.324,325 Within the group of high-grade B cell lymphomas, the reported frequency of p53 mutations in Burkitt's lymphomas, varies between 35 and 45 per cent.325,326 Almost all cases of Burkitt's lymphoma have a translocation involving the c-myc gene, and therefore activation of both p53 and c-Myc may be critical in the development of these tumors. The coexistence of c-Myc and p53 alterations has also been shown in 3/8 and 1/11 diffuse large cell lymphoma cell lines and tumors, respectively.327 In Burkitt's lymphoma cell lines, a high frequency of p53 alterations (up to 60 per cent) has also been described.321,328-330 The majority of these cell lines contained only mutant versions of the p53 gene, and most of the mutations were missense and occasionally nonsense and frameshift. There was no correlation with the type of translocation, the ethnic origin of the patients or the EBV status. The expression of p53 in a group of 119 patients with high-grade B cell NHL has been shown to be an independent poor prognostic factor for survival. Simultaneous expression of bd-2 and p53 was associated with a poorer prognosis than p53 alone, particularly in the subgroup of nodal lymphomas.331 Detection of p53 mutation in bone marrow have been associated with histologic progression and transformation.332 A high frequency of p53 mutations has also been observed in patients with AIDS-related lymphomas. Although these patients have an increased incidence of B cell immunoblastic lymphoma, the incidence or type of p53 mutation does not differ when compared with B cell immunoblastic lymphomas in individuals without human immunodeficiency virus (HIV) infection.333 Interestingly, in other studies, overexpression of p53 was observed in 10 out of 45 of cases analysed, mainly clustering in the small non-cleaved cell and Ki-1 anaplastic large cell subtypes. A diffuse or clustered pattern of p53-positive neoplastic cells was seen in the small non-cleaved lymphomas, consequent upon p53 mutations. In contrast, in Ki-1 anaplastic large cell lymphomas, p53 immunohistochemical reactivity was limited to scattered tumor cells but no p53 gene alterations could be detected.334'335 Low-grade non-Hodgkin's lymphomas rarely have p53 alterations. Higher frequencies of p53 mutations have recently been shown in the splenic B cell leukemia/lymphoma of possible marginal origin, where p53 gene alterations involving exons 5, 6 and 8 were found in 40 per cent of cases examined. These mutations were missense or frameshift and the wild-type sequence
at the mutations site was barely visible, implying the loss of the normal p53 allele in leukemic cells.336 Most of the p53 alterations in the group of low-grade non-Hodgkin's lymphomas have been associated with progression to high-grade lymphoma. Serial biopsies of patients with follicular lymphoma who underwent histological transformation showed that a p53 mutation was observed in one-third of the transformed samples and this was not detected at the follicular stage of the disease.337 Another study revealed that four of five cases of follicular lymphoma transformed to diffuse large cell lymphoma were associated with p53 mutations.338 Interestingly, in one of these positive cases, the same mutation was also present in the pretransformation biopsy, correlating with the presence of diffuse-type areas within a predominantly follicular pattern.339 p53 immunoreactivity was observed in 50 per cent of low-grade non-Hodgkin's lymphomas and was higher in tumors of T cell origin.339 There was a positive association between p53 staining and the proliferation state as expressed by proliferating cell nuclear antigen (PCNA). Another study, however, revealed variable p53 immunostaining in follicular lymphomas, between biopsies and between individual follicles within the same tumor, with no correlation with the state of cell proliferation.340 Peripheral T cell lymphomas seem to have a low incidence of p53 mutations, ranging from 3 to 9 per cent of cases examined.321'341 p53 overexpression was observed in 50 per cent of the cases and did not correlate with cell proliferation as assessed by Ki-67 expression.341 However, within the group of peripheral T cell lymphomas, the adult T cell leukemia/lymphoma associated with human T cell leukemia virus type I (HTLV-1) infection, seems to have a much higher incidence of p53 gene alterations, ranging from 30 to 50 per cent in different studies.342'345 Based on the evaluation of at least 51 patients, it was postulated that p53 alterations represents one of the genetic changes responsible for the progression of the disease.342^345 The mutations found were mainly missense, nonsense, silent and frameshift, and the majority of those found in patients with the acute type of disease, occurred in highly conserved regions of p53. Moreover, cells carrying p53 mutations showed loss of the other p53 allele.342-345 p53 mutations correlated with an altered pattern of p53 expression as assessed by immunohistochemical staining.344,345 The frequency of p53 gene alterations in the acute phase of adult T cell leukemia/ lymphoma was significantly higher than that in the chronic type, suggesting a possible involvement of p53 gene alterations in the disease progression.346 Overexpression of p53 protein as assessed by immunohistochemical staining has been detected in about 60-80 per cent of cases with mixed cellularity and nodular sclerosing type of Hodgkin's disease.315,317,319,347 Immunoreactivity is localized to the nuclei of the Reed-Sternberg (RS) cells or its mononuclear variants, and the number of positive cells varies between 10 and 60 per cent of RS cells.
144 Molecular biology
Mutations of p53 have been detected in enriched RS cell preparations.348'349 No correlation has been found between EBV infection and p53 reactivity in RS cells. The background of small lymphocytes, plasma cells, eosinophils and histiocytes in Hodgkin's samples are unstained for p53. This finding supports the idea that the RS cell is the neoplastic component of Hodgkin's disease. Moreover, no p53 staining was observed in lymphocyte-predominant Hodgkin's disease, suggesting that this disease may be a form of B cell lymphoma rather than a subtype of Hodgkin's disease. Although the occurrence of p53 mutations in lymphoma has been the focus of many investigations, it remains possible that alterations to other genes may be equally important in lymphomagenesis. The murine double minute 2 (MDM2) gene encodes a protein that binds to and inactivates p53. Reports of MDM2 overexpression in lymphomas are contradictory. A Japanese study indicated that overexpression occurred mainly in low-grade lymphomas,350 whereas a UK study reported elevated MDM2 expression was primarily associated with highgrade disease.351 In the former case, the quantification was carried out by comparison with actin expression, but the expression of this gene is known to vary considerably between lymphoma samples.352 It has been suggested that accumulation of wild type (wt) p53 could promote the overexpression of the MDM2 gene product and that the ratio of MDM2/P53 could play a critical role in lymphoma progression.353 p53 is known to mediate expression of the P21 (p21wafl/cipl) gene whose protein product is involved in growth arrest. The coding sequence of P21 has been investigated for mutations in Burkitt's lymphoma samples. No mutations were found in the clinical samples but a mutation in Burkitt's cell line was found and shown to result in a loss of function of p21,354 The CDKN2 gene located on chromosome 9p21 encodes the cyclin-dependent kinase 4 inhibitor pi6. This gene is a putative tumor supressor because of its frequent alteration in many kinds of tumor cells. Several studies have investigated the prevalence of CDKN2 alterations and between 6 and 14 per cent of NHL samples have been shown to have acquired alterations.355'359 Selective hypermethylation of the gene is described.360 It is not yet possible to determine whether CDKN2 alterations are particularly associated with one subgroup or grade of lymphoma, although a high frequency of alteration (35 per cent) to CDKN2, and the neighboring gene MTS2 has been noted in T-ALL/lymphoblastic lymphoma361 and in low-grade lymphomas that have transformed to high grade.358,362
Mutations are seen in these factors in less than 10 per cent of lymphomas.1,364 and maybe pathogenetically significant in some cases.365 An example of gene amplification in lymphoma is described for the REL family in primary extranodal lymphomas and mediastinal lymphomas. Twenty-six out of 111 diffuse large cell lymphomas showed amplification to more than four copies and 19 of them were primary extranodal.366,367 Gene amplification, in general, in lymphoma is so far only found in a minority of cases.368 bd-3, the translocation breakpoint of the t(14;19) found in B-CLL and some B cell lymphomas, is a family member with positive regulation properties. It may be involved in upregulation of kappa B responsive genes,
CONCLUSION The molecular genetics of lymphoma has been one of the most exciting areas of biological advance in oncology during the 1990s. Basic mechanisms of lymphomagenesis have been partially explained and the knowledge acquired has already been applied in diagnosis and prognosis. Further advances and applications for classification, pathology and choice of therapy, and the identification of new targets for therapy are likely in the coming decade.
REFERENCES 1. Ong ST, Le Beau MM. Chromosomal abnormalities and molecular genetics of non-Hodgkin's lymphoma. Semin Oncol 1998; 25: 447-60. 2. Neves H, Ramos C, Gomes da Silva M, Parreira A, Parreira L The nuclear topography of ABL, BCR, PML, and RARa genes: evidence for gene proximity in specific phases of the cell cycle and stages of hematopoietic differentiation. Blood 1999; 93:1197-207. 3. Rabbits TH. Chromosomal translocations in human cancer. Nature 1994; 372:143-9. 4. Gamberi B, Gaidano G, Parsa N, etal. Microsatellite instability is rare in B-cell non-Hodgkin's lymphomas. Blood 1997; 89: 975-9. 5. Taub R, Kirsch I, Morton C, et al. Translocation of the cmyc gene into the immunoglobulin heavy chain locus in human Burkitt lymphoma and murine plasmacytoma cells. Proc NatlAcadSci USA 1982; 79: 7837-41.
GENE AMPLIFICATION, THE REL/NFk B TRANSCRIPTION FACTORS AND bcl-1 REL/NF kappa B transcription factors regulate many genes involved in the immune response and inflammation.363
6. Adams JM, Gerondakis S, Webb E, Corcoran LM, Cory S. Cellular myc oncogene is altered by chromosome translocation to an immunoglobulin locus in murine plasmacytomas and is rearranged similarly in human Burkitt lymphomas. Proc NatlAcadSci USA 1983; 80: 2146-50.
References 145 7. Erikson J, Ar-Rushdi A, Drwinga HL, Nowell PC, Croce CM. Transcriptional activation of the translocated c-myc oncogene in burkitt lymphoma. Proc Natl Acad Sci USA 1983;80:1707-11. 8. Hamlyn PH, Rabbitts TH. Translocation joins c-myc and immunoglobulin gamma 1 genes in a Burkitt lymphoma revealing a third exon in the c-myc oncogene. Nature 1983; 304:172-4. 9. Dalla-Favera R, Westin E, Gelmann EP, etal. The human one gene c-myc: structure, expression, and amplification in the human promyelocytic leukemia cell line HL-60. Nature 1983; 304:135-9. 10. BatteyJ, Moulding C, Taub R, etal. The human c-myc oncogene: structural consequences of translocation into the IgH locus in Burkitt lymphoma. Cell 1983; 34: 779-87. 11. Croce CM, Tsujimoto Y, Erikson J, Nowell P. Chromosome translocations and B cell neoplasia. Nature 1984; 310: 655-60. 12. Emanuel BS, Selden JR, Chaganti RS, Jhanwar S, Nowell PC, Croce CM. The 2p breakpoint of a 2;8 translocation in Burkitt lymphoma interrupts the V kappa locus. Proc Natl Acad Sci USA 1984; 81: 2718-22. 13. HollisGF, Mitchell KF, BatteyJ, etal. A variant translocation places the lambda immunoglobulin genes 3' to the c-myc oncogene in Burkitt's lymphoma. Nature 1984;307:752-5. 14. Rappold GA, Hameister H, Cremer T, et al. The myc proteins are not associated with chromatin in mitotic cells. EMfiOyi984;3:2951-5. 15. Davis M, Malcolm S, Rabbitts TH. A variant translocation places the lambda immunoglobulin genes 3' to the cmyc oncogene in Burkitt's lymphoma. Nature 1984; 308: 286-8. 16. Malcolm S, Davis M, Rabbitts TH. Breakage on chromosome 2 brings the CK gene to a region 3' of cmyc in a Burkitt's lymphoma line carrying a (2;8) translocation. Cytogenet Cell Genet 1985; 39:168-72. 17. Manolov G, Manolova Y, Klein G, Lenoir G, Levan A. Alternative involvement of two cytogenetically distinguishable breakpoints on chromosome 8 in Burkitt's lymphoma associated translocations. Cancer Genet Cytogenet 1986; 20: 95-9. 18. Cory S, Graham M, Webb E, Corcoran L, Adams JM. Variant (6;15) translocation in murine plasmacytomas involves a chromosome 15 locus at least 72 Kb from the c-myc oncogene. EMBOJ 1985; 4: 675-81. 19. Gaidano G, Pastore C, Lanza C, etal. Molecular pathology of AIDS-related lymphomas. Biologic aspects and clinicopathologic heterogenetics. Ann Hematol 1994; 69: 281-90. 20. Ballerini P, Gaidano G, Gong JZ, et al. Multiple genetic lesions in acquired immunodeficiency syndromerelated non-Hodgkin's lymphoma. Blood 1993; 26: 166-76. 21. Rabbitts TH, Forster A, Hamlyn P, Baer R. Effect of somatic mutation within translocated c-myc genes in Burkitt lymphoma cells. Nature 1984; 309: 592-7.
22. Pelicci PG, Knowles D, Magrath I, Dalla-Favera R. Chromosomal breakpoints and structural alterations of the c-myc locus differ in endemic and sporadic forms of Burkitt lymphoma. Proc Natl Acad Sci USA 1986; 83: 2984-8. 23. Murphy W, Sarid J, Taub R, et al. A translocated human c-myc oncogene is altered in a conserved coding sequence. Proc Natl Acad Sci USA 1986; 83: 2939-43. 24. Ar-Rushdi A, Nishikura K, Erikson J, Watt R, Rovera G, Croce CM. Differential expression of the translocated and the untranslocated c-myc oncogene in Burkitt lymphoma. Science 1983; 222: 390-3. 25. Taub R, Moulding C, Battey J, et al. Activation and somatic mutation of the translocated c-myc gene in burkitt lymphoma cells.7 Exp Med 1984; 159: 276-91. 26. Wiman KG, Clarkson B, Hayday AC, Saito H, Tonegawa S, Hayward WS. Activation of a translocated c-myc gene: role of structural alterations in the upstream region. Proc Natl Acad Sci USA 1984; 81: 6798-802. 27. Nishikura K, Ar-Rushdi A, Erikson J, Watt R, Rovera G, Croce CM. Differential expression of the normal and of the translocated human c-myc oncogenes in B cells. Proc Natl Acad Sci USA 1983; 80: 5107-11. 28. Croce CM, Erikson J, Ar-Rushdi A, Aden D, Nishikura K. Translocated c-myc oncogene of Burkitt lymphoma is transcribed in plasma cells and repressed in lymphoblastoid cells. Proc Natl Acad Sci USA 1984; 81: 3547-51. 29. Maguire RT, Robins TS, Thorgeirsson SS, Heilman CA. Expression of cellular myc and mos genes in undifferentiated B cell lymphomas of Burkitt and nonBurkitt types. Proc Natl Acad Sci USA 1983; 80:1982-6. 30. Reed JC, Nowell PC, Hoover RG. Regulation of c-myc mRNA levels in normal human lymphocytes by modulators of cell proliferation. Proc Natl Acad Sci USA 1985;82:4221-4. 31. Hahn SR, Eisenman RN. Proteins encoded by the human c-myc oncogene: differential expression in neoplastic cells. Molec Cell Biol 1984; 4: 2486-97. 32. Ramsay G, Evan Gl, Bishop JM. The protein encoded by the human proto-oncogene c-myc. Proc Natl Acad Sci USA 1984; 81: 7742-6. 33. Favrot MC, Philip I, Philip T, DoreJF, Lenoir GM. Possible duality in Burkitt's lymphoma origin. Lancet 1984; 1:745-7. 34. Morse B, South VJ, Rothberg PG, Astrin SM. Somatic mutation and transcriptional deregulation of myc in endemic Burkitt's lymphoma disease: heptamer-nonamer recognition mistakes? Molec Cell Biol 1989; 9: 74-82. 35. Haluska FG, Russo G, Andreeff M, Croce CM. Molecular analysis of an AIDS-associated Burkitt's lymphoma: near-identity with endemic cases. Curr Topics Microbiol Immunol 1988; 141: 75-9. 36. Haluska FG, Russo G, Kant J, Andreef M, Croce CM. Molecular resemblance of an AIDS-associated lymphoma and endemic Burkitt lymphomas:
146 Molecular biology
37.
38.
39.
40.
41.
42.
43. 44.
45. 46.
47.
48.
49.
50.
51.
52.
implications for their pathogenesis. Proc NatlAcad Sd 175/41989; 86: 8907-11. Chenevix-Trench G, Behm FG, Westin EH. Somatic rearrangements of the c-myc oncogene in primary human diffuse large-cell lymphoma. IntJ Cancer 1986; 38: 513-16. Ladanyi M, Offit K, Jhanwar SC, Filippa DA, Chaganti RSK. MYC rearrangements and translocations involving band 8q24 in diffuse large cell lymphomas. Blood 1991; 77:1057-63. Leder A, Pattengale PK, Kuo A, Stewart TA, Leder P. Consequences of widespread deregulation of the c-myc gene in transgenic mice: multiple neoplasms and normal development. Cell 1986; 45: 485-95. Adams JM, Harris AW, Pinkert CA, et al. The c-myc oncogene driven by immunoglobulin enhancers induces lymphoid malignancy in transgenic mice. Mm/re 1985; 318: 533-8. Persson H, Leder P. Nuclear localisation and DMA binding of a protein expressed by human c-myc oncogene. Science 1984; 225: 718-21. Finnegan MCM, Hammond DW, Hancock BW, Goyns MH. Activation of MYC-N in a case of non-Hodgkin's lymphoma. Leuk Lymphoma 1995; 18: 511-14. Packham G, Cleveland JL c-Myc and apoptosis. Biochim BiophysActa 1995; 1242:11-28. Spencer CA, Groudine M. Control of c-myc regulation in normal and neoplastic cells. Adv Cancer Res 1991; 56: 1-48. Marcu KB, Bossone SA, Patel AJ. myc function and regulation. Annu Rev Biochem 1992; 61: 809-60. Askew DS, Ashmun RA, Simmons BC, Cleveland JL Constitutive c-myc expression in an IL-3-dependent myeloid cell line suppresses cell cycle arrest and accelerates apoptosis. Oncogene 1991; 6:1915-22. Heikkila R, Schwab G, Wickstrom E, et al. A c-myc antisense oligodeoxynucleotide inhibits entry into S phase but not progress from GO to G1. Nature 1987; 328:445-9. Prochownik EV, Kukowska J, Rodgers C. c-myc antisense transcripts accelerate differentiation and inhibit G1 progression in murine erythroleukemia cells. MolecCell 8/o/1988;8:3683-95. Gregor PD, Sawadogo M, Roeder RG. The adenovirus major late transcription factor USF is a member of the helix-loop-helix group of regulatory proteins and binds to DMA as a dimer. Genes Dev 1990; 4:1730-40. Beckmann H, Su LK, Kadesch T. TFE3: a helix-loop-helix protein that activates transcription through the immunoglobulin enhancer muE3 motif. Genes Dev 1990; 4:167-79. Carr CS, Sharp PA. A helix-loop-helix protein related to the immunoglobulin E boxbinding proteins. MolecCell Biol 1990; 10: 4384-8. Hu YF, Luscher B, Admon A, Mermod N, Tjian R. Transcription factor AP-4 contains multiple dimerization domains that regulate dimer specificity. Genes Dev 1990; 4:1741-52.
53. Blackwood EM, Eisenman RN. Max: a helix-loop-helix zipper protein that forms a sequence-specific DMAbinding complex with Myc. Science 1991; 251:1211-7. 54. Prendergast GC, Lawe D, Ziff EB. Association of Myn, the murine homolog of max, with c-Myc stimulates methylation-sensitive DMA binding and ras cotransformation. Cell 1991; 65: 395-407. 55. Littlewood TD, Amati B, Land H, Evan Gl. Max and cMyc/Max DMA-binding activities in cell extracts. Oncogene 1992; 7:1783-92. 56. Berberich SJ, Cole MD. Casein kinase II inhibits the DMAbinding activity of Max homodimers but not Myc/Max heterodimers. Genes Dev 1992; 6:166-76. 57. Zervos AS, Gyuris J, Brent R, Mxi1, a protein that specifically interacts with Max to bind Myc-Max recognition sites [published erratum appears in Cell 1994 Oct 21: 79(2): following 388]. Cell 1993; 72: 223-32. 58. Ayer DE, Kretzner L, Eisenman RN. Mad: a heterodimeric partner for Max that antagonizes Myc transcriptional activity. Cell 1993; 72: 211-22. 59. Schneider JW, Gu W, Zhu L, Mahdavi V, Nadal-Ginard B. Reversal of terminal differentiation mediated by p107 in Rb-/- muscle cells. Science 1994; 264:1467-71. 60. Beijersbergen RL, Hijmans EM, Zhu L, Bernards R. Interaction of c-Myc with the pRb-related protein p107 results in inhibition of c-Myc-mediated transactivation. EMBO J 1994; 13: 4080-6. 61. Harrington EA, Fanidi A, Evan Gl. Oncogenes and cell death. Curr Opin Genet Dev 1994; 4:120-9. 62. Van den Berghe H, ParloirC, David G, MichauxJL, Sokal G. A new characteristic karyotypic anomaly in lymphoproliferative disorders. Cancer 1979; 44: 188-95. 63. Leroux D, Le-Marc'Hadour F, Gressin R, et al. NonHodgkin's lymphomas with t(11;14)(q13;q32): a subset of mantle zone/intermediate lymphocytic lymphoma? BrJ Haematol 1991; 77: 346-53. 64. Vandenberghe E, De-Wolf-Peeters C, Wlodarska I, et al. Chromosome 11q rearrangements in B non Hodgkin's lymphoma. BrJ Haematol 1992; 81: 212-7. 65. Weisenburger DD, Chan WC. Lymphomas of follicles. Mantle cell and follicle center cell lymphomas. Am J Clin PathoS 1993; 99: 409-20. 66. Shivdasani RA, Hess JL, Skarin AT, Pinkus GS. Intermediate lymphocytic lymphoma: clinical and pathologic features of a recently characterized subtype of non-Hodgkin's lymphoma.7 Clin Oncol 1993; 11: 802-111. 67. Brito-Babapulle V, Ellis J, Matutes E, et al. Translocation t(11;14)(q13;q32) in chronic lymphoid disorders. Genes Chromosomes Cancer 1992; 5:158-65. 68. Oscier DG, Matutes E, Gardiner A, et al. Cytogenetic studies in splenic lymphoma with villous lymphocytes. BrJ Haematol 1993; 85: 487-91. 69. Tsujimoto Y, Yunis J, Onorato-Showe L, Erikson J, Nowell PC, Croce CM. Molecular cloning of the chromosomal
References 147
70.
71.
72. 73.
74.
75.
76.
77.
78. 79.
80.
81.
82.
83.
84.
breakpoint of B-cell lymphomas and leukemias with the t(11;14) chromosome translocation. Science 1984; 224: 1403-6. Tsujimoto Y, Jaffe E, Cossman J, Gorham J, Nowell PC, Croce CM. Clustering of breakpoints on chromosome 11 in human B-cell neoplasms with the t(11;14) chromosome translocation. Nature 1985; 315: 340-3. Erikson J, Finan J, Tsujimoto Y, Nowell PC, Croce CM. The chromosome 14 breakpoint in neoplastic B cells with the t(11 ;14) translocation involves the immunoglobulin heavy chain locus. Proc Natl Acad Sci tyS/41984;81:4144-8. Ince C, Blick M, Lee M, etal. Bcl-1 gene rearrangements in B cell lymphoma. Leukemia 1988; 2: 343-6. Medeiros LJ, Van-Krieken JH, Jaffe ES, Raffeld M. Association of bcl-1 rearrangements with lymphocytic lymphoma of intermediate differentiation. Blood 1990; 76: 2086-90. Williams ME, Westermann CD, Swerdlow SH. Genotypic characterization of centrocytic lymphoma: frequent rearrangement of the chromosome 11 bcl-1 locus. Blood 1990; 76:1387-91. Arnold A, Kim HG, Gaz RD, et al. Molecular cloning and chromosomal mapping of DNA rearranged with the parathyroid hormone gene in a parathyroid adenoma. J Clin Invest 1989; 83: 2034-40. Motokura T, Bloom T, Kim HG, Juppner H, Ruderman JV, Kronenberg HM, Arnold A. A novel cyclin encoded by a bell-linked candidate oncogene. Nature 1991; 350: 512-5. Matsushime H, Roussel MF, Ashmun RA, Sherr CJ. Colonystimulating factor 1 regulates novel cyclins during the G1 phase of the cell cycle. Cell 1991; 65: 701-13. Xiong Y, Connolly T, Futcher B, Beach D. Human D-type cyclin. Cell 1991; 65: 691 -9. Rosenberg CL, Wong E, Petty EM, et al. PRAD1, a candidate BCL1 oncogene: mapping and expression in centrocytic lymphoma. Proc Natl Acad Sci USA 1991; 88: 9638-42. Withers DA, Harvey RC, Faust JB, Melnyk 0, Carey K, Meeker TC. Characterization of a candidate bcl-1 gene. Molec Cell Biol 1991; 11: 4846-53. Williams ME, Swerdlow SH, Rosenberg CL, Arnold A. Chromosome 11 translocation breakpoints at the PRAD1/cyclin D1 gene locus in centrocytic lymphoma. /.ewfert7/01993;7:241-5. Meeker TC, Sellers W, Harvey R, et al. Cloning of the t(11;14)(q13;q32) translocation breakpoints from two human leukemia cell lines. Leukemia 1991; 5: 733-7. Rimokh R, Berger F, Delsol G, et al. Rearrangement and overexpression of the BCL-1/PRAD-1 gene in intermediate lymphocytic lymphomas and in t(11q13)bearing leukemias. Blood 1993; 81: 3063-7. Williams ME, Swerdlow SH, Rosenberg CL, Arnold A. Characterization of chromosome 11 translocation breakpoints at the bcl-1 and PRAD1 loci in centrocytic lymphoma. Cancer Res 1992; 52: 5541s-5544s.
85. Williams ME, Meeker TC, Swerdlow SH. Rearrangement of the chromosome 11 bcl-1 locus in centrocytic lymphoma: analysis with multiple breakpoint probes. Blood 1991; 78: 493-8. 86. Molot RJ, Meeker TC, Wittwer a, et al. Antigen expression and polymerase chain reaction amplification of mantle cell lymphomas. Blood 1994; 83:1626-31. 87. Delmer A, Ajchenbaum-Cymbalista F, Tang R, et al. Over-expression of cyclin D1 in chronic B-cell malignancies with abnormality of chromosome 11q13. BrJ Haematol 1995; 89: 798-804. 88. Seto M, Yamamoto K, lida S, etal. Gene rearrangement and overexpression of PRAD1 in lymphoid malignancy with t(11;14)(q13;q32) translocation. Oncogene 1992; 7: 1401-6. 89. Raynaud SD, Bekri S, Leroux D, et al. Expanded range of 11q13 breakpoints with differing patterns of cyclin D1 expression in B-cell malignancies. Genes Chromosome Cancer 1993; 8: 80-7. 90. Atkin NB, Baker MC. Diffuse large cell (Kiel-1) lymphoma with a t(9;11)(p21-22;q13) and a missing Y as the only chromosome changes. Cancer Genet Cytogenet 1998; 101: 72-4. 91. CuneoA, Balboni M, Piva N, etal. Atypical chronic lymphocytic leukaemia with t(11;14)(q13;q32): karyotype evolution and prolymphocytic transformation. BrJ Haematol 1995; 90: 409-16. 92. Yang Wl, Zukerberg LR, Motokura T, Arnold A, Harris NL. Cyclin D1 (Bcl-1, PRAD1) protein expression in low-grade B-cell lymphomas and reactive hyperplasia./4m./Paf/?o/ 1994; 145: 86-96. 93. Palmero I, Holder A, Sinclair AJ, Dickson C, Peters G. Cyclins D1 and D2 are differentially expressed in human B-lymphoid cell lines. Oncogene 1993; 8:1049-54. 94. Bosch F, Jares P, Campo E, et al. PRAD-1/cyclin D1 gene overexpression in chronic lymphoproliferative disorders: a highly specific marker of mantle cell lymphoma. Blood 1994; 84: 2726-32. 95. Lebwohl DE, Muise-Helmericks R, Sepp-Lorenzino L, et al. A truncated cyclin D1 gene encodes a stable mRNA in a human breast cancer cell line. Oncogene 1994; 9: 1925-9. 96. Rimokh R, Berger F, Bastard C, et al. Rearrangement of CCND1 (BCL1/PRAD1) 3' untranslated region in mantlecell lymphomas and t(11q13)-associated leukemias. Blood 1994; 83: 3689-96. 97. Rosenberg CL, Motokura T, Kronenberg HM, Arnold A. Coding sequence of the overexpressed transcript of the putative oncogene PRAD1/cyclin D1 in two primary human tumors. Oncogene 1993; 8: 519-21. 98. Buckley MF, Sweeney KJ, Hamilton JA, et al. Expression and amplification of cyclin genes in human breast cancer. Oncogene 1993; 8: 2127-33. 99. Keyomarsi K, Pardee AB. Redundant cyclin overexpression and gene amplification in breast cancer cells. Proc Natl Acad Sci USA 1993; 90:1112-6.
148 Molecular biology 100. Gillett C, Fantl V, Smith R, et al. Amplification and overexpression of cyclin D1 in breast cancer detected by immunohistochemical staining. Cancer Res 1994; 54: 1812-17. 101. Tsuruta H, Sakamoto H, Onda M, Terada M. Amplification and overexpression of EXP1 and EXP2/Cyclin D1 genes in human esophageal carcinomas. Biochem Biophys Res Commun 1993; 196:1529-36. 102. Leach FS, Elledge SJ, Sherr CJ, et al. Amplification of cyclin genes in colorectal carcinomas. Cancer Res 1993; 53:1986-9. 103. Zhang YJ, Jiang W, Chen Q,etal. Amplification and overexpression of cyclin D1 in human hepatocellular carcinoma. Biochem Biophys Res Commun 1993; 196: 1010-16. 104. Jiang W, Kahn SM, Zhou P, et al. Overexpression of cyclin D1 in rat fibroblasts causes abnormalities in growth control, cell cycle progression and gene expression. Oncogene 1993; 8: 3447-57. 105. Quelle DE, Ashmum RA, Shurtleff SA, etal. Overexpression of mouse D-type cyclins accelerates G1 phase in rodent fibroblasts. Genes Dev 1993; 7: 1559-71. 106. Resnitzky D, Gossen M, Bujard H, Reed SI. Acceleration of the G1/S phase transition by expression of cyclins D1 and E with an inducible system. Molec Cell Biol 1994; 14:1669-79. 107. Baldin V, LukasJ, Marcote M\,etal. Cyclin D1 is a nuclear protein required for cell cycle progression in G1. Genes Dev 1993; 7:1010-16. 108. Lovec H, Sewing A, Lucibello FC, Muller R, MoroyT. Oncogenic activity of cyclin D1 revealed through cooperation with Ha-ras: link between cell cycle control and malignant transformation. Oncogene 1994; 9: 323-6. 109. Hinds PW, Dowdy SF, Eaton EN, Arnold A, Weinberg RA. Function of a human cyclin gene as an oncogene. Proc Natl Acad Sci USA 1994; 91: 709-13. 110. Dowdy SF, Hinds PW, Louie K, Reed SI, Arnold A, Weinberg RA. Physical interaction of the retinoblastoma protein with human D cyclins. Cell 1993; 73: 499-511. 111. Ewen ME, Sluss HK, Sherr CJ, Matsushime H, KatoJ, Livingston DM. Functional interactions of the retinoblastoma protein with mammalian D-type cyclins. Cell 1993; 73: 487-97. 112. Hall FL, Williams RT, Wu L, etal. Two potentially oncogenic cyclins, cyclin A and cyclin D1, share common properties of subunit configuration, tyrosine phosphorylation and physical association with the Rb protein. Oncogene 1993; 8:1377-84. 113. KatoJ, Matsushime H, HiebertSW, Ewen ME, Sherr CJ. Direct binding of cyclin D to the retinoblastoma gene product (pRb) and pRb phosphorylation by the cyclin Ddependent kinase CDK4. Genes Dev 1993; 7: 331-42. 114. Weinberg RA. Tumor suppressor genes. Science 1991; 254:1138-46. 115. NevinsJR. E2F: a link between the Rb tumor suppressor
116.
117.
118.
119.
120.
121.
122.
123.
124.
125.
126.
127.
128.
129.
protein and viral oncoproteins. Science 1992; 258: 424-9. Weintraub SJ, Prater CA, Dean DC. Retinoblastoma protein switches the E2F site from positive to negative element. Nature 1992; 358: 259-61. Grierson AJ, Hodgkins MA, Hancock BW, RoydsJ, Goepel JR, Goyns MH. Investigation of the RB-1 tumour suppressor gene in a United Kingdom series of nonHodgkin's lymphomas. Leuk Lymphoma 1996; 23: 353-63. BodrugSE, Warner BJ, Bath ML, Lindeman GJ, Harris AW, Adams JM. Cyclin D1 transgene impedes lymphocyte maturation and collaborates in lymphomagenesis with the myc gene. EMBOJ 1994; 13: 2124-30. Lovec H, Grzeschiczek A, Kowalski MB, Moroy T. Cyclin D1/bcl-1 cooperates with myc genes in the generation of B-cell lymphoma in transgenic mice. EMBOJ 1994; 13: 3487-95. Offit K, Wong G, Filippa DA, Tao Y, Chaganti RS. Cytogenetic analysis of 434 consecutively ascertained specimens of non-Hodgkin's lymphoma: clinical correlations. Blood 1991; 77:1508-15. Yunis JJ, Frizzera G, Oken MM, McKenna J, Theologides A, Arnesen M. Multiple recurrent genomic defects in follicular lymphoma. A possible model for cancer. N EnglJ Med 1987; 316: 79-84. Fukuhara S, Rowley JD, Variakojis D, Golomb HM. Chromosome abnormalities in poorly differentiated lymphocytic lymphoma. Cancer Res 1979; 39: 3119-28. Tsujimoto Y, Gorham J, Cossman J, Jaffe E, Croce CM. The t(14;18) chromosome translocations involved in Bcell neoplasms result from mistakes in VDJ joining. Science 1985; 229:1390-3. Bakhshi A, Jensen JP, Goldman P, et al. Cloning the chromosomal breakpoint of t(14;18) human lymphomas: clustering around JH on chromosome 14 and near a transcriptional unit on 18. Cell 1985; 41: 899-906. Cleary ML, Sklar J. Nucleotide sequence of a t(14;18) chromosomal breakpoint in follicular lymphoma and demonstration of a breakpoint-cluster region near a transcriptionally active locus on chromosome 18. Proc Natl Acad Sci USA 1985; 82: 7439-43. Tsujimoto Y, Croce CM. Analysis of the structure, transcripts, and protein products of bcl-2, the gene involved in human follicular lymphoma. Proc Natl Acad Sc/t;SA1986;83:5214-8. Cazals-Hatem DL, Louie DC, Tanaka S, Reed JC. Molecular cloning and DNA sequence analysis of cDNA encoding chicken homologue of the Bcl-2 oncoprotein. Biochim Biophys Acta 1992; 1132:109-13. Monaghan P, Robertson D, Amos TA, Dyer MJ, Mason DY, Greaves MF. Ultrastructural localization of bcl-2 protein. J Histochem Cytochem 1992; 40:1819-25. Cleary ML, Smith SD, Sklar J. Cloning and structural analysis of cDNAs for bcl-2 and a hybrid bcl2/immunoglobulin transcript resulting from thet(14;18) translocation. Cell 1986; 47:19-28.
References 149 130. Bakhshi A, Wright JJ, Graninger W, et al. Mechanism of the t(14;18) chromosomal translocation: structural analysis of both derivative 14 and 18 reciprocal partners. Proc NatlAcadSci USA 1987; 84: 2396-400. 131. Cleary ML, Galili N, Sklar J. Detection of a second t(14;18) breakpoint cluster region in human follicular lymphomas.y Exp Med 1986; 164: 315-20. 132. Ngan BY, Nourse J, Cleary ML. Detection of chromosomal translocation t(14;18) within the minor cluster region of bcl-2 by polymerase chain reaction and direct genomic sequencing of the enzymatically amplified DMA in follicular lymphomas. Blood 1989; 73: 1759-62. 133. Tanaka S, Louie DC, Kant JA, Reed JC. Frequent incidence of somatic mutations in translocated BCL2 oncogenes of non-Hodgkin's lymphomas. Blood 1992; 79:229-37. 134. Lee KA, Goepel JR, Winfield DA, Hancock BW, Goyns MH. Investigation of BCL-2 gene rearrangements in a UK series of low and high grade non-Hodgkin's lymphomas. Leuk Lymphoma 1993; 11: 91-8. 135. Dyer MJ, Zani VJ, Lu WZ, etal. BCL2 translocations in leukemias of mature B cells. Blood 1994; 83: 3682-8. 136. Mufti GJ, Hamblin TJ, Oscier DG, Johnson S. Common ALL with pre-B-cell features showing (8;14) and (14;18) chromosome translocations. Blood 1983; 62:1142-6. 137. Islington DM, Monard S, Johnson PW, etal. The t(14;18) in a patient with de novo acute lymphoblastic leukemia is associated with t(8;9). Leukemia 1994; 8: 560-3. 138. Kramer MH, Raghoebier S, Beverstock GC, de Jong D, Kluin PM, Kluin-NelemansJC. De novo acute B-cell leukemia with translocation t(14;18): an entity with a poor prognosis. Leukemia 1991; 5: 473-8. 139. LeBrun DP, Ngan BY, Weiss LM, Huie P, Warnke RA, Cleary ML. The bcl-2 oncogene in Hodgkin's disease arising in the setting of follicular non-Hodgkin's lymphoma. Blood 1994; 83: 223-30. 140. Gupta RK, Lister TA, Bodmer JG. Thet(14;18) chromosomal translocation and Bcl-2 protein expression in Hodgkin's disease. Leukemia 1994; 8: 1337-41. 141. Stetler-Stevenson M, Crush-Stanton S, Cossman J. Involvement of the bcl-2 gene in Hodgkin's disease. J Natl Cancer Inst 1990; 82: 855-8. 142. Limpens J, de Jong D, van Krieken JH, et al. Bcl-2/JH rearrangements in benign lymphoid tissues with follicular hyperplasia. Oncogene 1991; 6: 2271-6. 143. Segal GH, Scott M, Jorgensen T, Braylan RC. Standard polymerase chain reaction analysis does not detect t(14;18) in reactive lymphoid hyperplasia. Arch Pathol /.flbM«/1994;118:791-4. 144. Adachi M, Tefferi A, Greipp PR, Kipps TJ, Tsujimoto Y. Preferential linkage of bcl-2 to immunoglobulin light chain gene in chronic lymphocytic leukemia. J Exp Med 1990; 171: 559-64. 145. Aventin A, Mecucci C, Guanyabens C, et al. Variant t(2;18) translocation in a Burkitt conversion of follicular lymphoma. BrJ Haematol 1990; 74: 367-9.
146. Leroux D, Monteil M, Sotto JJ, et al. Variant t(2;18) translocation in a follicular lymphoma. BrJ Haematol 1990; 75: 290-2. 147. Adachi M, Tsujimoto Y. Juxtaposition of human bcl-2 and immunoglobulin lambda light chain gene in chronic lymphocytic leukemia is the result of a reciprocal chromosome translocation between chromosome 18 and 22. Oncogene 1989; 4:1073-5. 148. Hillion J, Mecucci C, Aventin A, et al. A variant translocation t(2;18) in follicular lymphoma involves the 5' end of bcl-2 and Ig kappa light chain gene. Oncogene 1991; 6:169-72. 149. Rimokh R, Gadoux M, Bertheas MF, etal. FVT-1, a novel human transcription unit affected by variant translocation t(2;18)p11;q21) of follicular lymphoma. Blood 1993; 81:136-42. 150. Mikraki V, Ladanyi M, Chaganti RS. Structural alterations in the 5' region of the BCL2 gene in follicular lymphomas with BCL2-MBR or BCL2-MCR rearrangements. Genes Chromosom Cancer 1991; 3: 117-21. 151. Seite P, Hillion J, d'Agay MF, Berger R, Larsen CJ. BCL2 complex rearrangement in follicular lymphoma: translocation mbr/JH and deletion in the vcr region of the same BCL2 allele. Oncogene 1993; 8: 3073-80. 152. Zelenetz AD, Cleary ML, Levy R. Asubmicroscopic interstitial deletion of chromosome 14 frequently occurs adjacent to the t(14;18) translocation breakpoint in human follicular lymphoma. Genes Chromosomes Cancer 1993; 6:140-50. 153. de Jong D, Voetdijk BM, van OmmenGJB, KluinNelemansJC, Beverstock GC, Kluin PM. Translocation t(14;18) in B cell lymphomas as a cause for defective immunoglobulin production.y Exp Med 1989; 169: 613-24. 154. Lee MS, Chang KS, Cabanillas F, Freireich EJ, Trujillo JM, Stass SA. Detection of minimal residual cells carrying the t(14;18) by DNA sequence amplification. Science 1987; 237:175-8. 155. Stetler-Stevenson M, Raffeld M, Cohen P, Cossman J. Detection of occult follicular lymphoma by specific DNA amplification. Blood 1988; 72:1822-5. 156. Crescenzi M, Seto M, Herzig GP, Weiss PD, Griffith RC, Korsmeyer SJ. Thermostable DNA polymerase chain amplification of t(14;18) chromosome breakpoints and detection of minimal residual disease. Proc Natl Acad Sci USA 1988; 85: 4869-73. 157. Cotter F, Price C, Zucca E, Young BD. Direct sequence analysis of the 14q+ and 18q-chromosome junctions in follicular lymphoma. Blood 1990; 76:131-5. 158. Corbally N, Grogan L, Dervan PA, Carney DN. The detection of specific gene rearrangements in nonHodgkin's lymphoma using the polymerase chain reaction. BrJ Cancer 1992; 66: 805-9. 159. Cotter FE, Price C, Meerabux J, Zucca E, Young BD. Direct sequence analysis of 14q+ and 18q- chromosome junctions at the MBR and MCR revealing clustering
150 Molecular biology
160.
161.
162.
163.
164.
165.
166.
167.
168.
169.
170.
171.
within the MBR in follicular lymphoma. Ann Oncol 1991; 2: 93-7. Pezzella F, Ralfkiaer E, Gatter KC, Mason DY. The 14;18 translocation in European cases of follicular lymphoma: comparison of Southern blotting and the polymerase chain reaction. BrJ Haematol 1990; 76: 58-64. Ladanyi M, Wang S. Detection of rearrangements of the BCL2 major breakpoint region in follicular lymphomas. Correlation of polymerase chain reaction results with Southern blot analysis. Diagn Molec Pathol 1992; 1: 31-5. Benitez J, Robledo M, Santon A, Santos M, Rivas C, Echezarreta G, Martinez-Castro P. Correlation between cytogenetic and molecular analysis of t(14;18) in follicular lymphomas. Cancer Genet Cytogenet 1992; 59: 68-72. Zelenetz AD, Chu G, Galili N, etal. Enhanced detection of the t(14;18) translocation in malignant lymphoma using pulsed-field gel electrophoresis. Blood 1991; 78: 1552-60. Lambrechts AC, Hupkes PE, Dorssers LC, van't-Veer MB. Translocation (14;18)-positive cells are present in the circulation of the majority of patients with localized (stage I and II) follicular non-Hodgkin's lymphoma. fi/ood1993;82:2510-6. Finke J, Slanina J, Lange W, Dolken G. Persistence of circulating t(14;18)-positive cells in long-term remission after radiation therapy for localized-stage follicular lymphoma. 7 C/m Oncol 1993; 11:1668-73. Price CG, Meerabux J, Murtagh S, etal. The significance of circulating cells carrying t(14;18) in long remission from follicular lymphoma. J Clin Oncol 1991; 9: 1527-32. Gribben JG, Freedman A, Woo SD, etal. All advanced stage non-Hodgkin's lymphomas with a polymerase chain reaction amplifiable breakpoint of bcl-2 have residual cells containing the bcl-2 rearrangement at evaluation and after treatment. Blood 1991; 78: 3275-80. Lambrechts AC, Hupkes PE, Dorssers LC, van't-Veer MB. Clinical significance of t(14;18)-positive cells in the circulation of patients with stage III or IV follicular nonHodgkin's lymphoma during first remission.) Clin Oncol 1994; 12:1541-6. Berinstein NL, Jamal HH, Kuzniar B, Klock RJ, Reis MD. Sensitive and reproducible detection of occult disease in patients with follicular lymphoma by PCR amplification of t(14;18) both pre- and post-treatment. Leukemia 1993; 7: M3-9. Tang SC, Visser L, Hepperie B, etal. Clinical significance of /7C/-2-MBR gene rearrangement and protein expression in diffuse large-cell non-Hodgkin's lymphoma: an analysis of 83 cases.) Clin Oncol 1994; 12:149-54. Kramer MH, Hermans], Parker J, etal. Clinical significance of bc!2 and p53 protein expression in diffuse large B-cell lymphoma: a population-based study. 7 Clin Oncol 1996; 14: 2131-8.
172. Levine EG, Arthur DC, Frizzera G, Peterson BA, Hurd DD, Bloomfield CD. Cytogenetic abnormalities predict clinical outcome in non-Hodgkin lymphoma. Ann Intern Med 1988; 108:14-20. 173. Pezzella F, Jones M, Ralfkiaer E, Ersboll J, Gatter KC, Mason DY. Evaluation of bcl-2 protein expression and 14;18 translocation as prognostic markers in follicular lymphoma. BrJ Cancer 1992; 65: 87-9. 174. Tilly H, Rossi A, Stamatoullas A, et al. Prognostic value of chromosomal abnormalities in follicular lymphoma. fi/oorf1994;84: 1043-9. 175. Gribben JG, Freedman AS, Neuberg D, et al. Immunologic purging of marrow assessed by PCR before autologous bone marrow transplantation for Bcell lymphoma. N Engl J Med 1991; 325:1525-33. 176. Johnson PW, Price CG, Smith T, et al. Detection of cells bearing the t(14;18) translocation following myeloablative treatment and autologous bone marrow transplantation for follicular lymphoma. J Clin Oncol 1994; 12: 798-805. 177. Gribben JG, Neuberg D, Freedman AS, et al. Detection by polymerase chain reaction of residual cells with the bcl-2 translocation is associated with increased risk of relapse after autologous bone marrow transplantation for B-cell lymphoma. Blood 1993; 81: 3449-57. 178. Miyashita T, Reed JC: Bcl-2 oncoprotein blocks chemotherapy-induced apoptosis in a human leukemia cell line. Blood 1993; 81:151-7. 179. Reed JC, Kitada S, Takayama S, et al. Regulation of chemoresistance by the bcl-2 oncoprotein in nonHodgkin's-lymphoma and lymphocytic leukemia cell lines. Ann Oncol 1994; 1: 61-5. 180. Walton Ml, Whysong D, O'Connor PM, etal. Constitutive expression of human Bcl-2 modulates nitrogen mustard and camptothecin induced apoptosis. Cancer Res 1993; 53:1853-61. 181. Conrad MN, Wright JH, Wolf AJ, Zakian VA. RAP1 protein interacts with yeast telomeres in vivo: overproduction alters telomere structure and decreases chromosome stability. Cell 1990; 63: 739-50. 182. Wahls WP, Wallace LJ, Moore PD. Hypervariable minisatellite DNA is a hotspot for homologous recombination in human cells. Cell 1990; 60: 95-103. 183. Boehm T, Mengle-Gaw L, Kees UR, et al. Alternating purine-pyrimidine tracts may promote chromosomal translocations seen in a variety of human lymphoid tumours. EMBOJ 1989; 8: 2621-31. 184. Nikaido T, Yamawaki-Kataoka Y, Honjo T. Nucleotide sequences of switch regions of immunoglobulin C epsilon and C gamma genes and their comparison.7 Biol Chem 1982; 257: 7322-9. 185. Gerstein RM, Frankel WN, Hsieh CL, etal. Isotype switching of an immunoglobulin heavy chain transgene occurs by DNA recombination between different chromosomes. Cell 1990; 63: 537-48. 186. Jeffreys AJ, Wilson V, Thein SL Hypervariable 'minisatellite' regions in human DNA. Nature 1985; 314: 67-73.
References 151 187. Krowczynska AM, Rudders RA, KrontirisTG. The human minisatellite consensus at breakpoints of oncogene translocations. Nucleic Acids Res 1990; 18:1121-7. 188. Wyatt RT, Rudders RA, Zelenetz A, Delellis RA, Krontiris TG. BCL2 oncogene translocation is mediated by a chilike consensus. J Exp Med 1992; 175:1575-88. 189. Jaeger U, Purtscher B, Karth GD, Knapp S, Mannhalter C, Lechner K. Mechanism of the chromosomal translocation t(14;18) in lymphoma: detection of a 45Kd breakpoint binding protein. Blood 1993; 81: 1833-40. 190. Ngan BY, Chen LZ, Weiss LM, Warnke RA, Clearly ML Expression in non-Hodgkin's lymphoma of the bcl-2 protein associated with the t(14;18) chromosomal translocation. N EnglJ Med 1988; 318:1638-44. 191. Gaulard P, d'Agay MF, Peuchmaur M, etal. Expression of the bcl-2 gene product in follicular lymphoma. AmJ Pathol 1992; 140:1089-95. 192. Seite P, Million J, d'Agay MF, et al. BCL2 gene activation and protein expression in follicular lymphoma: a report on 64 cases. Leukemia 1993; 7: 410-7. 193. Pezzella F, Tse AG, Cordell JL, Pulford KA, Gatter KC, Mason DY. Expression of the bcl-2 oncogene protein is not specific for the 14;18 chromosomal translocation. AmJ Pathol 1990; 137: 225-32. 194. Hockenbery DM, Zutter M, Mickey W, Nahm M, Korsmeyer SJ. BCL2 protein is topographically restricted in tissues characterized by apoptotic cell death. Proc NatlAcadSci USA 1991; 88: 6961-5. 195. Chleq-Deschamps CM, LeBrun DP, Huie P, etal. Topographical dissociation of BCL-2 messenger RNA and protein expression in human lymphoid tissues. Blood 1993; 81: 293-8. 196. Zutter M, Hockenbery D, Silverman GA, Korsmeyer SJ. Immunolocalization of the Bcl-2 protein within hematopoietic neoplasms. Blood 1991; 78:1062-8. 197. Pettersson M, Jernberg-Wiklund H, Larsson LG, etal. Expression of the bcl-2 gene in human multiple myeloma cell lines and normal plasma cells. Blood 1992; 79: 495-502. 198. Silvestrini R, Veneroni S, Daidone MG, etal. The Bcl-2 protein: a prognositic indicator strongly related to p53 protein in lymph node-negative breast cancer patients. J Natl Cancer Inst 1994; 86: 499-504. 199. Vaux DL, Cory S, Adams JM. Bcl-2 gene promotes haemopoietic cell survival and cooperates with c-myc to immortalize pre-B cells. Nature 1988; 335: 440-2. 200. Nunez G, London L, Hockenbery D, Alexander M, McKearn JP, Korsmeyer SJ. Deregulated Bcl-2 gene expression selectively prolongs survival of growth factor-deprived hemopoietic cell lines. 7 Immunol 1990; 144: 3602-10. 201. Lotem J, Sachs L. Induction of dependence on hematopoietic proteins for viability and receptor upregulation in differentiating myeloid leukemic cells. Blood 1989; 74: 579-85. 202. Hockenbery D, Nunez G, Milliman C, Schreiber RD,
203.
204.
205.
206.
207.
208.
209.
210.
211.
212.
213.
214.
215.
216.
Korsmeyer SJ. Bcl-2 is an inner mitochondrial membrane protein that blocks programmed cell death. Nature 1990; 348: 334-6. Campana D, Coustan-Smith E, Manabe A, et al. Prolonged survival of B-lineage acute lymphoblastic leukemia cells is accompanied by overexpression of bcl-2 protein. Blood-\993; 81:1025-31. Allsopp TE, Wyatt S, Paterson HF, Davies AM. The protooncogene bcl-2 can selectively rescue neurotrophic factor-dependent neurons from apoptosis. Cell 1993; 73: 295-307. Miyashita T, Reed JC. bcl-2 gene transfer increases relative resistance of S49.1 and WEH17.2 lymphoid cells to cell death and DNA fragmentation induced by glucocorticoids and multiple chemotherapeutic drugs. Cancer Res 1992; 52: 5407-11. Lotem J, Sachs L. Regulation by bcl-2, c-myc, and p53 of susceptibility to induction of apoptosis by heat shock and cancer chemotherapy compounds in differentiation-competent and -defective myeloid leukemic cells. Cell Growth Differ 1993; 4: 41-7. Reed JC, Kitada S, Takayama S, Miyashita T. Regulation of chemoresistance by the bcl-2 oncoprotein in nonHodgkin's lymphoma and lymphocytic leukemia cell lines. Ann Onco/ 1994; 1: 61-5. Miyashita T, Reed JC. Bcl-2 oncoprotein blocks chemotherapy-induced apoptosis in a human leukemia cell line. Blood 1993; 81:151_7. Naumovski L, Clearly ML. Bcl-2 inhibits apoptosis associated with terminal differentiation of HL-60 myeloid leukemia cells. Blood 1994; 83: 2261-7. Strasser A, Whittingham S, Vaux DL, etal. Enforced BCL2 expression in B-lymphoid cells prolongs antibody responses and elicits autoimmune disease. Proc Natl Acad Sci USA 1991; 88: 8661 -5. Sentman CL, Shutter JR, Hockenbery D, Kanagawa 0, Korsmeyer SJ. bcl-2 inhibits multiple forms of apoptosis but not negative selection in thymocytes. Cell 1991; 67: 879-88. Nunez G, Hockenbery D, McDonnell TJ, Sorensen CM, Korsmeyer SJ. Bcl-2 maintains B cell memory. Nature 1991; 353: 71-3. Pietenpol JA, Papadopoulos N, Markowitz S, Willson JK, Kinzler KW, Vogelstein B. Paradoxical inhibition of solid tumor cell growth by bc!2. Cancer Res 1994; 54: 3714-7. Henderson S, Rowe M, Gregory C, et al. Induction of bcl2 expression by Epstein-Barr virus latent membrane protein 1 protects infected B cells from programmed cell death. Ce//1991; 65:1107-15. Vaux DL, Weissman IL, Kim SK. Prevention of programmed cell death in Caenorhabditis elegans by human bcl-2. Science 1992; 258:1955-7. Yuan J, Horvitz HR. The Caenorhabditis elegans cell death gene ced-4 encodes a novel protein and is expressed during the period of extensive programmed cell death. Development 1992; 116: 309-20.
152 Molecular biology 217. Ellis HM, Horvitz HR. Genetic control of programmed cell death in the nematode C. elegans. Cell 1986; 44: 817-29. 218. Reed JC, Cuddy M, SlabiakT, Croce CM, Nowell PC. Oncogenic potential of bcl-2 demonstrated by gene transfer. Nature 1988; 336: 259-61. 219. McDonnell TJ, Deane N, Platt FM, etal. bcl-2-immunoglobulin transgenic mice demonstrate extended B cell survival and follicular lymphoproliferation. Cell 1989; 57: 79-88. 220. McDonnell TJ, Korsmeyer SJ. Progression from lymphoid hyperplasia to high-grade malignant lymphoma in mice transgenic for the t(14;18). Nature 1991; 349: 254-6. 221. Strasser A, Harris AW, Cory S. E mu-bcl-2 transgene facilitates spontaneous transformation of early pre-B and immunoglobulin-secreting cells but not T cells. Oncogene 1993; 8:1-9. 222. Limpens J, Stad R, Vos C, et al. Lymphoma-associated translocation t(14;18) in blood B cells of normal individuals. Blood 1995; 85: 2528-36. 223. Liu Y, Hernandez AM, Shibata D, Cortopassi GA. BCL2 translocation frequency rises with age in humans. Proc NatlAcadSci USA 1994; 91: 8910-4. 224. Siegel RM, Katsumata M, Miyashita T, Louie DC, Greene Ml, Reed JC. Inhibition of thymocyte apoptosis and negative antigenic selection in bcl-2 transgenic mice. Proc NatlAcad Sci USA 1992; 89: 7003-7. 225. Katsumata M, Siegel RM, Louie DC, et al. Differential effects of Bcl-2 on T and B cells in transgenic mice. Proc NatlAcadSci USA 1992; 89:11376-80. 226. Linette GP, Hess JL, Sentman CL, Korsmeyer SJ. Peripheral T-cell lymphoma in lck(pr)-bcl-2 transgenic mice. Blood 1995; 86:1255-60. 227. Veis DJ, Sorenson CM, Shutter JR, Korsmeyer SJ. Bcl-2deficient mice demonstrate fulminant lymphoid apoptosis, polycystic kidneys, and hypopigmented hair. Cell 1993; 75: 229-40. 228. Krajewski S, Tanaka S, Takayama S, Schibler MJ, Fenton W, Reed JC. Investigation of the subcellular distribution of the bcl-2 oncoprotein: residence in the nuclear envelope, endoplasmic reticulum, and outer mitochondria! membranes. Cancer Res 1993; 53: 4701-14. 229. Jacobson MD, Burne JF, King MP, Miyashita T, Reed JC, Raff MC. Bcl-2 blocks apoptosis in cells lacking mitochondrial DMA. Nature 1993; 361: 365-9. 230. Mah SP, Zhong LT, Liu Y, Roghani A, Edwards RH, Bredesen DE. The protooncogene bcl-2 inhibits apoptosis in PC12 cells.; Neurochem 1993; 60:1183-6. 231. Smets LA, Van-den-BergJ, Acton D, Top B, Van-Rooij H, Verwijs-Janssen M. BCL-2 expression and mitochondrial activity in leukemic cells with different sensitivity to glucocorticoid-induced apoptosis. Blood 1994; 84: 1613-9. 232. Baffy G, Miyashita T, Williamson JR, Reed JC. Apoptosis induced by withdrawal of interleukin-3 (IL-3) from an IL-3-dependent hematopoietic cell line is associated
233.
234.
235.
236.
237.
238.
239.
240.
241.
242.
243.
244.
245.
246.
with repartitioning of intracellular calcium and is blocked by enforced Bcl-2 oncoprotein production. 7 BiolChem 1993; 268: 6511-9. Lam M, Dubyak G, Distelhorst CW. Effect of glucocorticosteroid treatment on intracellular calcium homeostasis in mouse lymphoma cells. Molec Endocrinol 1993; 7: 686-93. Dowd DR, MacDonald PN, Komm BS, Haussler MR, Miesfeld RL. Stable expression of the calbindin-D28K complementary DMA interferes with the apoptotic pathway in lymphocytes. Molec Endocrinol 1992; 6: 1843-8. Kane DJ, Sarafian TA, Anton R, et al. Bcl-2 inhibition of neural death: Decreased generation of reactive oxygen species. Science 1993; 262:1274-7. Henderson S, Huen D, Rowe M, Dawson C, Johnson G, Rickinson A. Epstein-Barr virus-coded BHRF1 protein, a viral homologue of Bcl-2, protects human B cells from programmed cell death. Proc NatlAcad Sci USA 1993; 90: 8479-83. Neilan JG, Lu Z, Afonso CL, Kutish GF, Sussman MD, Rock DL. An African swine fever virus gene will similarity to the proto-oncogene bcl-2 and the Esptein-Barr virus gene BHRF1.7 Virol 1993; 67: 4391-4. Oltvai ZN, Milliman CL, Korsmeyer SJ. Bcl-2 heterodimerizes in vivo with a conserved homolog, Bax, that accelerates programmed cell death. Cell 1993; 74: 609-19. Boise LH, Gonzalez-Garcia M, Postema CE, etal. bcl-x, a bcl-2-related gene that functions as a dominant regulator of apoptotic cell death. Cell 1993; 74: 597-608. Kozopas KM, Yang T, Buchan HL, Zhou P, Craig RW. MCL1, a gene expressed in programmed myeloid cell differentiation, has sequence similarity to BCL2. Proc NatlAcadSci USA 1993; 90: 3516-20. Craig RW, Jabs EW, Zhou P, et al. Human and mouse chromosomal mapping of the myeloid cell leukemia-1 gene: MCL1 maps to human chromosome 1q21, a region that is frequently altered in preneoplastic and neoplastic disease. Genomics 1994; 23: 457-63. Lin EY, Orlofsky A, Berger MS, Prystowsky MB. Characterization of A1, a novel hemopoietic-specific early-response gene with sequence similarity to bcl-2. J /mmuno/1993; 151:1979-88. Yang E, Zha J, Jockel J, Boise LH, Thompson CB, Korsmeyer SJ. Bad, a heterodimeric partner for Bel-XL and Bcl-2, displaces Bax and promotes cell death. Cell 1995;80:285-91. Farrow SN, White JH, Martinou I, et al. Cloning of a bcl-2 homologue by interaction with adenovirus E1B 19K. Nature 1995;374:731-3. Chittenden T, Harrington EA, O'Connor R, etal. Induction of apoptosis by the bcl-2 homologue Bak. Nature 1995; 374: 733-6. Kiefer MC, Brauer MJ, Powers VC, et al. Modulation of apoptosis by the widely distributed bcl-2 homologue Bak. Nature 1995; 374: 736-9.
References 153 247. Yin XM, Oltvai ZN, Korsmeyer SJ. BH1 and BH2 domains of Bcl-2 are required for inhibition of apoptosis and heterodimerization with Bax. Nature 1994; 369: 321-3. 248. Offit K, Jhanwar S, Ebrahim SA, Filippa D, Clarkson BD, Chaganti RS. t(3;22)(q27;q11): a novel translocation associated with diffuse non-Hodgkin's lymphoma. Blood 1989; 74:1876-9. 249. Leroux D, Stul M, Sotto JJ, et al. Translocation t(3;22)(q23;q11) in three patients with diffuse large B cell lymphoma. Leukemia 1990; 4: 373-6. 250. Bastard C, Tilly H, Lenormand B, et al. Translocations involving band 3q27 and Ig gene regions in nonHodgkin's lymphoma. Blood 1992; 79: 2527-31. 251. Ye BH, Rao PH, Chaganti RS, Dalla-Favera R. Cloning of bcl-6, the locus involved in chromosome translocations affecting band 3q27 in B-cell lymphoma. Cancer Res 1993;53:2732-5. 252. Ye BH, Lista F, Lo Coco F, et al. Alterations of a zinc finger-encoding gene, BCL-6, in diffuse large-cell lymphoma. Science 1993; 262: 747-50. 253. Baron BW, Nucifora G, McCabe N, Espinosa R3d, Le Beau MM, McKeithan TW. Identification of the gene associated with the recurring chromosomal translocations t(3;14)(q27;q32) and t(3;22)(q27;q11) in B-cell lymphomas. Proc Natl AcadSci USA 1993; 90: 5262-6. 254. Kerckaert JP, Deweindt C, Tilly H, Quief S, Lecocq G, Bastard C. LAZ3, a novel zinc-finger encoding gene, is disrupted by recurring chromosome 3q27 translocations in human lymphomas. Nat Genet 1993; 5: 66-70. 255. Miki T, Kawamata N, Arai A, et al. Molecular cloning of the breakpoint for 3q27 translocation in B-cell lymphomas and leukemias. Blood 1994; 83: 217-22. 256. Kawamata N, Miki T, Ohashi K, et al. Recognition DMA sequence of a novel putative transcription factor, BCL6. Biochem Biophys Res Commun 1994; 204: 366-74. 257. Harrison SD, Travers AA. The tramtrack gene encodes a Drosophila finger protein that interacts with the ftz transcriptional regulatory region and shows a novel embryonic expression pattern. EMBOJ1990; 9: 207-16. 258. Chardin P, Courtois G, Mattei MG, Gisselbrecht S. The KUPgene, located on human chromosome 14, encodes a protein with two distant zinc fingers. Nucleic Acids Res 1991;19:1431-6. 259. Chen Z, Brand NJ, Chen A, et al. Fusion between a novel Kruppel-like zinc finger gene and the retinoic acid receptor-alpha locus due to a variant t(11;17) translocation associated with acute promyelocytic leukaemia. EMBOJ 1993; 12:1161-7. 260. Zollman S, Godt D, Prive GG, Couderc JL, Laski FA. The BTB domain, found primarily in zinc finger proteins, defines an evolutionary conserved family that includes several developmentally regulated genes in Drosophila. Proc Natl Acad Sci USA 1994; 91:10717-21. 261. Dent AL, Shaffer AL, Yu X, et al. Control of inflammation, cytokine expression, and germinal center formation by bcl-6. Science 1997; 276: 589-92.
262. Chen W, lida S, Louie DC, Dalla-Favera R, Chaganti RS. Heterologous promoters fused to BCL6 by chromosomal translocations affecting band 3q27 cause its deregulated expression during B-cell differentiation. Blood 1998; 91: 603-7. 263. Kawamata N, Miki T, Fukuda T, Hirosawa S, Aoki N. The organization of the BCL6 gene. Leukemia 1994; 8: 1327-30. 264. Deweindt C, Kerckaert JP, Tilly H, Quief S, Nguyen VC, Bastard C. Cloning of a breakpoint cluster region at band 3q27 involved in human non-Hodgkin's lymphoma. Genes Chromosom Cancer 1993; 8:149-54. 265. Suzuki K, Miki T, Kawamata N, etal. Variant translocation of the BCL6 gene to immunoglobulin kappa light chain gene in B-cell lymphoma. Jap J Cancer Res 1994; 85: 911-7. 266. Nakamura Y, Miki T, Kawamata N, et al. DNA rearrangements and deletions within the BCL-6 gene in B-cell non-Hodgkin's lymphoma. BrJ Haematol 1995; 90: 404-8. 267. Ichinohasama R, Miura I, Shishido T, et al. Translocation (3;16)(q27;p11) in a patient with diffuse large B-cell lymphoma associated with the BCL-6 gene rearrangement. Cancer Genet Cytogenet 1998; 103: 133-9. 268. Lai JL, Daudignon A, Kerchaert JP, et al. Translocation (3;13)(q27;q14): a nonrandom and probably secondary structural change in non-Hodgkin lymphomas. Cancer Genet Cytogenet 1998; 103:140-3. 269. Bastard C, Deweindt C, Kerckaert JP, et al. LAZ3 rearrangements in non-Hodgkin's lymphoma: correlation with histology, immunophenotype, karyotype, and clinical outcome in 217 patients. Blood 1994; 83: 2423-7. 270. Dallery E, Galiegue-Zouitina S, Collyn-d'Hooghe M, et al. TTF, a gene encoding a novel small G protein, fuses to the lymphoma-associated LAZ3 gene by t(3;4) chromosomal translocation. Oncogene 1995; 10: 2171-8. 271. Ohno H, Kerckaert JP, Bastard C, Fukuhara S. Heterogeneity in B-cell neoplasms associated with rearrangement of the LAZ3 gene on chromosome band 3q27.JapJ Cancer Res 1994; 85: 592-600. 272. Ichinohasama R, Miura I, Funato T, et al. A recurrent nonrandom translocation (3;7)(q27;p12) associated with BCL-6 gene rearrangement in B-cell diffuse large cell lymphoma. Cancer Genet Cytogenet 1998; 104:19-27. 273. Galiegue-Zouitina S, Quief S, Hildebrand MP, et al. The B-cell transcriptional coactivator BOB1/OBF1 gene fuses to the LAZ3/BCL6 gene by t(3;11)(q27;q23.1) chromosomal translocation in a B-cell leukaemia line (Karpas231). Blood 1995; 86: 38a. 274. Otsuki T, Yano T, Clark HM, et al. Analysis of LAZ3 (BCL6) status in B-cell non-Hodgkin's lymphomas: results of rearrangement and gene expression studies and a mutational analysis of coding region sequences. Blood 1995; 85: 2877-84.
154 Molecular biology 275. Lo Coco F, Ye BH, Lista F, et al. Rearrangements of the BCL6 gene in diffuse large cell non-Hodgkin's lymphoma. Blood 1994; 83:1757-9. 276. Offit K, Lo Coco F, Louie DC, et al. Rearrangement of the bcl-6 gene as a prognostic marker in diffuse large-cell lymphoma. N EnglJ Med 1994; 331: 74-80. 277. Gaidano G, Lo Coco F, Ye BH, et al. Rearrangements of the BCL-6 gene in acquired immunodeficiency syndrome-associated non-Hodgkin's lymphoma: association with diffuse large-cell subtype. Blood 1994; 84: 397-402. 278. Morgan R, Hecht BK, Sandberg AA, Hecht F, Smith SD. Chromosome 5q35 breakpoint in malignant histiocytosis. N EnglJ Med 1986; 314:1322. 279. Benz-Lemoine E, Brizard A, Huret JL, et al. Malignant histiocytosis: a specific t(2;5)(p23;q35) translocation? Review of the literature. Blood 1988; 72:1045-7. 280. Rimokh R, Magaud JP, Berger F, et al. A translocation involving a specific breakpoint (q35) on chromosome 5 is characteristic of anaplastic large cell lymphoma ('Ki-1 lymphoma'). Brj Haematol 1989; 71: 31-6. 281. Mason DY, Bastard C, Rimokh R, et al. CD30-positive large cell lymphomas ('Ki-1 lymphoma') are associated with a chromosomal translocation involving 5q35. Brj W0e/mrto/1990;74:161-8. 282. Kaneko Y, Frizzera G, Edamura S, et al. A novel translocation, t(2;5)(p23;q35), in childhood phagocytic large T-cell lymphoma mimicking malignant histiocytosis. Blood 1989; 73: 806-13. 283. Le Beau MM, Bitter MA, Larson RA, etal. The t(2;5)(p23;q35): a recurring chromosomal abnormality in Ki-1-positive anaplastic large cell lymphoma. Leukemia 1989; 3: 866-70. 284. Fischer P, Nacheva E, Mason DY, et al. A Ki-1 (CD30)positive human cell line (Karpas 299) established from a high-grade non-Hodgkin's lymphoma, showing a 2;5 translocation and rearrangement of the T-cell receptor beta-chain gene. Blood 1988; 72: 234-40. 285. Tilly H, Gaulard P, Lepage E, et al. Primary anaplastic large-cell lymphoma in adults: clinical presentation, immunophenotype, and outcome. Blood 1997; 90: 3727-34. 286. Mitev L, Christova S, Hadjiev E, et al. A new variant chromosomal translocation t(2;2)(p23;q23) in CD30+/Ki-1+ anaplastic large cell lymphoma. Leuk Lymphoma 1998; 28: 613-16. 287. Morris SW, Kirstein MN, Valentine MB, et al. Fusion of a kinase gene, ALK, to a nucleolar protein gene, NPM, in non-Hodgkin's lymphoma. Science 1994; 263:1281-4. 288. Sarris AH, Luthra R, Cabanillas F, Morris SW, Pugh WC. Genomic DMA amplification and the detection of t(2;5)(p23;q35) in lymphoid neoplasms. Leuk Lymphoma 1998; 29: 507-14. 289. Bullrich F, Morris SW, Hummel M, Pileri S, Stein H, Croce CM. Nucleophosmin (npm) gene rearrangements in ki-1positive lymphomas. Cancer Res 1994; 54: 2873-7. 290. Stein H, Dallenbach F. In: Knowles DM, ed. Neoplastic
291.
292.
293.
294.
295.
296.
297. 298.
299.
300.
301.
302.
303.
304.
hematology. Baltimore: Williams and Wilkins, 1992; 675-714. Lopategui JR, Sun LH, Chan JK, etal. Low frequency association of the t(2;5)(p23;q35) chromosomal translocation with CD30+ lymphomas from American and Asian patients. A reverse transcriptase-polymerase chain reaction study. AmJ Pathol 1995; 146: 323-8. Orscheschek K, Merz H, Hell J, Binder T, Bartels H, Feller AC. Large-cell anaplastic lymphoma-specific translocation (t[2;5] [p23;q35]) in Hodgkin's disease: indication of a common pathogenesis? Lancet 1995; 345: 87-90. Ladanyi M, Cavalchire G, Morris SW, Downing], Filippa DA. Reverse transcriptase polymerase chain reaction for the Ki-1 anaplastic large cell lymphoma-associated t(2;5) translocation in Hodgkin's disease. Am J Pathol 1994; 145: 1296-300. Weiss LM, Lopategui JR, Sun L-H, Kamel OW, Koo CH, Glackin C. Absence of the t(2;5) in Hodgkin's disease. Blood 1995; 85: 2845-7. Pulford K, Lamant L, Morris SW etal. Detection of anaplastic lymphoma kinase (ALK) and nucleolar protein nudeophosmin (NPM)-ALK proteins in normal and neoplastic cells with the monoclonal antibody ALKi. Blood 1997; 89: 1394-404. Benharroch D, Meguerian-Bedoyan Z, Lamant L, et al. ALK-positive lymphoma: a single disease with a broad spectrum of morphology. Blood 1998; 91: 2076-84. Kadin ME, Morris SW. The t(2;5) in human lymphomas. Leuk Lymphoma 1998; 29: 249-56. Whang-Peng J, Knutsen T, Jaffe ES, et al. Sequential analysis of 43 patients with non-Hodgkin's lymphoma: clinical correlations with cytogenetic, histologic, immunophenotyping, and molecular studies. Blood 1995; 85: 203-16. Cabanillas F, Pathak S, Trujillo J, et al. Frequent nonrandom chromosome abnormalities in 27 patients with untreated large cell lymphoma and immunoblastic lymphoma. Cancer Res 1988; 48: 5557-64. Bloomfield CD, Arthur CD, Frizzera G, Levine EG, Peterson BA, Gajl-Peczalska KJ. Nonrandom chromosome abnormalities in lymphoma. Cancer Res 1983;43:2975-84. Cimino MC, Roth DG, Golomb HM, Rowley JD. A chromosome marker for B-cell cancers. N EnglJ Med 1978;298:1422. Kubonishi I, Niiya K, Yamashita M, etal. Characterization of a new human lymphoma cell line (RC-K8) with t(11;14) chromosome abnormality. Cancer 1986; 58:1453-60. Schlaifer D, Dastugue N, Brousset P, et al. B-cell lymphoma following polycythemia vera: evidence for the involvement of two different clones. Leukemia 1994; 8: 895-6. Cabanillas F, Pathak S, Trujillo J, et al. Cytogenetic features of Hodgkin's disease suggest possible origin from a lymphocyte. Blood 1988; 71:1615-17.
References 155 305. Decoteau JF, Reis MD, Griesser H, Lorenzana A, Alhashmi I, Hawley RG. Sbh-1, a novel Reed-Sternberg cell-line capable of inducing tumors in scid mice immunophenotypic, cytogenetic and cytokine expression profiles. Blood 1995; 85: 2828-38. 306. Akao Y, Seto M, Yamamoto K, et al. The RCK gene associated with t(11;14) translocation is distinct from the MLL/ALL-1 gene with t(4;11) and t(11;19) translocations. Cancer Res 1992; 52: 6083-7. 307. Thirman MJ, Gill HJ, Burnett RC, etal. Rearrangement of the MIL gene in acute lymphoblastic and acute myeloid leukemias with 11q23 chromosomal translocations. N EnglJ Med 1993; 329: 909-14. 308. MeerabuxJM, Cotter FE, Kearney L, etal. Molecular cloning of a novel 11q23 breakpoint associated with non-Hodgkin's lymphoma. Oncogene 1994; 9: 893-8. 309. Wotherspoon AC, Pan L, Diss TC, Isaacson PG. Cytogenetic study of B-cell lymphoma of mucosa-associated lymphoma tissue. Cancer Genet Cytogenet 1992; 58: 35-8. 310. Hussell T, Isaacson PG, Spencer J. Proliferation and differentiation of tumour cells from B-cell lymphoma of mucosa-associated lymphoid tissue in vitro. J Pathol 1993; 169: 221-7. 311. WillisTGJadayel DM, Du MQ,etal. Bel 10 is involved in t(1 ;14)(p22;q32) of MALT B cell lymphomas and mutated in multiple tumour types. Cell 1999; 96: 35-45. 312. YunisJJ, Oken MM, Kaplan ME, etal. Distinctive chromosomal abnormalities in histologic subtypes of non-Hodgkin's lymphoma p53 and bcl-2 expression in high-grade B-cell lymphomas: correlation with survival time. N EnglJ Med 1982; 307: 337-41. 313. Schouten HC, Sanger WG, Weisenburger DD, Anderson J, ArmitageJO. Chromosomal abnormalities in untreated patients with non-Hodgkin's lymphoma: associations with histology, clinical characteristics, and treatment outcome. The Nebraska Lymphoma Study Group. Blood 1990; 75: 1841-7. 314. Cabanillas F, Pathak S, Grant G, et al. Refractoriness to chemotherapy and poor survival related to abnormalities of chromosomes 17 and 7 in lymphoma. Am J Med 1989; 87:167-72. 315. Doglioni C, Pelosio P, Mombello A, Scarpa A, Chilosi M. Immunohistochemical evidence of abnormal expression of the antioncogene-encoded p53 phosphoprotein in Hodgkin's disease and CD30+ anaplastic lymphomas. Hematol Pathol 1991; 5: 67-73. 316. Pezzella F, Morrison H, Jones M, etal. Immunohistochemical detection of p53 and bcl-2 proteins in non-Hodgkin's lymphoma. Histopathology 1993; 22: 39-44. 317. Said JW, Barrera R, Shintaku IP, Nakamura H, Koeffler HP. Immunohistochemical analysis of p53 expression in malignant lymphomas. AmJ Pathol 1992; 141:1343-8. 318. Soini Y, Paakko P, Alavaikko M, Vahakangas K. p53 expression in lymphatic malignancies.7 Clin Pathol 1992; 45:1011-4.
319. Villuendas R, Piris MA, Orradre JL, et al. P53 protein expression in lymphomas and reactive lymphoid tissue. J Pathol 1992; 166: 235-41. 320. Lee KA, Finnegan MCM, Sheridan E, et al. Analysis of the P53 gene, its expression and protein stabilization in non-Hodgkin's lymphomas. Ann Oncol 1994; 5: (suppl 1): 85-8. 321. GaidanoG, Ballerini P, GongJZ, et al. p53 mutations in human lymphoid malignancies: association with Burkitt lymphoma and chronic lymphocytic leukemia. Proc NatlAcadSci USA 1991; 88: 5413-7. 322. Ichikawa A, Hotta T, Takagi N, et al. Mutations of p53 gene and their relation to disease progression in B-cell lymphoma. Blood 1992; 79: 2701-7. 323. Shiraishi Y. p53 mutation in fresh lymphocytes, Blymphoblastoid cell lines and their transformed cell lines originating from Bloom syndrome patients. Cancer Genet Cytogenet 1993; 68: 70-3. 324. Kocialkowski S, Pezzella F, Morrison H, etal. Mutations in the p53 gene are not limited to classic 'hot spots' and are not predictive of p53 protein expression in highgrade non-Hodgkin's lymphoma. BrJ Haematol 1995; 89: 55-60. 325. Villuendas R, Piris MA, Algara P, et al. The expression of p53 protein in non-Hodgkin's lymphomas is not always dependent on p53 gene mutations. Blood 1993; 82: 3151-6. 326. Bhatia KG, Gutierrez Ml, Huppi K, Siwarski D, Magrath IT. The pattern of p53 mutations in Burkitt's lymphoma differs from that of solid tumors. Cancer Res 1992; 52: 4273-6. 327. Farrugia MM, Duan LJ, Reis MD, Ngan BY, Berinstein NL. Alterations of the p53 tumor suppressor gene in diffuse large cell lymphomas with translocations of the c-MYC and BCL-2 proto-oncogenes. Blood 1994; 83:191-8. 328. Wiman KG, Magnusson KP, RamqvistT, Klein G. Mutant p53 detected in a majority of Burkitt lymphoma cell lines by monoclonal antibody PAb240. Oncogene 1991; 6:1633-9. 329. Farrell PJ, Allan GJ, Shanahan F, Vousden KH, Crook T. p53 is frequently mutated in Burkitt's lymphoma cell lines. EMBOJ1991; 10: 2879-87. 330. Duthu A, Debuire B, Romano J, et al. p53 mutations in Raji cells: characterization and localization relative to other Burkitt's lymphomas. Oncogene 1992; 7: 2161-7. 331. Piris MA, Pezella F, Martinez-Montero JC, et al. p53 and bcl-2 expression in high-grade B-cell lymphomas: correlation with survival time. BrJ Cancer 1994; 69: 337-41. 332. Jhanwar-Uniyal M, Gulati SC. p53 gene mutation in the bone-marrow of a patient with diffuse mixed cell type lymphoma at diagnosis predicting eventual progression to large cell lymphoma. Leak Lymphoma 1998; 29: 415-21. 333. Nakamura H, Said JW, Miller CW, Koeffler HP. Mutation and protein expression of p53 in acquired immunodeficiency syndrome-related lymphomas. Blood 1993; 82:920-6.
156 Molecular biology 334. De-Re V, Carbone A, De-Vita S, et al. p53 protein overexpression and p53 gene abnormalities in HIV-1-related non-Hodgkin's lymphomas. IntJ Cancer 1994; 56: 662-7. 335. Ballerini P, Gaidano G, Gong JZ, et al. Multiple genetic lesions in acquired immunodeficiency syndromerelated non-Hodgkin's lymphoma. Blood 1993; 81: 166-76. 336. Baldini L, Fracchiolla NS, Cro LM, et al. Frequent p53 gene involvement in splenic B-cell leukemia/lymphomas of possible marginal zone origin. Blood 1994; 84: 270-8. 337. Sander CA, Yano T, Clark HM, et al. p53 mutation is associated with progression in follicular lymphomas. Blood 1993; 82:1994-2004. 338. Lo Coco F, Gaidano G, Louie DC, Offit K, Chaganti RS, Dalla-Favera R. p53 mutations are associated with histologic transformation of follicular lymphoma. Blood 1993;82:2289-95. 339. Korkolopoulou P, Dates J, Kittas C, Crocker J. p53, c-myc p62 and proliferating cell nuclear antigen (PCNA) expression in non-Hodgkin's lymphomas.7 Clin Pathol 1994; 47: 9-14. 340. Barrans S, Randerson J, Evans P, et al. Heterogeneity in cell proliferation and expression of p53 and bcl-2 during the indolent phase of germinal centre cell lymphoma: an explanation for clinical variability. BrJ Haematol 1995; 90: 830-6. 341. Matsushima AY, Cesarman E, Chadburn A, Knowles DM. Post-thymic T cell lymphomas frequently overexpress p53 protein but infrequently exhibit p53 gene mutations. AmJ Pathol 1994; 144: 573-84. 342. Sakashita A, Hattori T, Miller CW, et al. Mutations of the p53 gene in adult T-cell leukemia. Blood 1992; 79: 477-80. 343. Sugito S, Yamato K, Sameshima Y, Yokota J, Yano S, Miyoshi I. Adult T-cell leukemia: structures and expression of the p53 gene. IntJ Cancer 1991; 49: 880-5. 344. Nagai H, Kinoshita T, Imamura J, et al. Genetic alteration of p53 in some patients with adult T-cell leukemia. Jap J Cancer Res 1991; 82:1421-7. 345. Cesarman E, Chadburn A, Inghirami G, Gaidano G, Knowles DM. Structural and functional analysis of oncogenes and tumor suppressor genes in adult T-cell leukemia/lymphoma shows frequent p53 mutations. Blood 1992; 80: 3205-16. 346. Nishimura S, Asou N, Suzushima H, et al. p53 gene mutation and loss of heterozygosity are associated with increased risk of disease progression in adult T cell leukemia. Leukemia 1995; 9: 598-604. 347. Gupta RK, Norton AJ, Lister TA, Bodmer JG. p53 protein expression in Reed-Stern berg cells of Hodgkin's disease. Leukemia 1993;7:S31-3. 348. Trumper LH, Brady G, Bagg A, et al. Single-cell analysis of Hodgkin and Reed-Stern bergcells:molecular heterogeneity of gene expression and p53 mutations. Blood 1993; 81: 3097-115.
349. Gupta RK, Patel K, Bodmer WF, Bodmer JG. Mutation of p53 in primary biopsy material and cell lines from Hodgkin disease. Proc NatlAcadSci USA 1993; 90: 2817-21. 350. WatanabeT, Hotta T, Ichikawa A, etal. The MDM2 oncogene overexpression in chronic lymphocytic leukemia and low-grade lymphoma of B-cell origin. Blood 1994; 84: 3158-65. 351. Finnegan MCM, Goepel JR, RoydsJ, Hancock BW, Goyns MH. Elevated levels of MDM-2 and P53 expression are associated with high grade non-Hodgkin's lymphomas. Cancer Lett 1994; 86: 215-21. 352. Finnegan MCM, Goepel JR, Hancock BW, Goyns MH. Investigation of the expression of housekeeping genes in non-Hodgkin's lymphomas. Leuk Lymphoma 1993; 10: 387-93. 353. Maestro R, Gloghini A, Doglioni C, et al. MDM2 overexpression does not account for stabilization of wild-type p53 protein in non-Hodgkin's lymphomas. Blood 1995; 85: 3239-46. 354. Bhatia K, Fan S, Spangler G, Weintraub M, O'Connor PM, Judde JG, Magrath I. A mutant p21 cydin-dependent kinase inhibitor isolated from a Burkitt's lymphoma. Cancer/?«1995; 55:1431-5. 355. Ogawa S, Hangaishi A, Miyawaki S, et al. Loss of the cyclin-dependent kinase 4-inhibitor (p16;MTS1) gene is frequent in and highly specific to lymphoid tumors in primary human hematopoietic malignancies. Blood 1995; 86:1548-56. 356. Uchida T, Watanabe T, Kinoshita T, Murate T, Saito H, Hotta T. Mutational analysis of the CDKN2 (MTS1/p16ink4A) gene in primary B-cell lymphomas. Blood 1995; 86: 2724-31. 357. Duro D, Flexor MA, Bernard 0, d'Agay MF, Berger R, Larsen CJ. Alterations of the putative tumor suppressor gene p16/MTS1 in human hematological malignancies. C R Acad Sci III 1994; 317: 913-9. 358. Dreyling MH, Roulston D, Bohlander SK, Vardiman J, Olopade Ol. Codeletion of CDKN2 and MTAP genes in a subset of non-Hodgkin's lymphoma may be associated with histologic transformation from low-grade to diffuse large-cell lymphoma. Genes Chromosomes Cancer 1998; 22: 72-8. 359. Nakahara Y, Nagai H, Kinoshita T, etal. Mutational analysis of the PTEN/MMAC1 gene in non-Hodgkin's lymphoma. Leukemia 1998; 12:1277-80. 360. Martinez-Delgado B, Robledo M, Arranz E, et al. Hypermethylation of p15/ink4b/MTS2 gene is differentially implicated among non-Hodgkin's lymphomas. Leukemia 1998; 12: 937-41. 361. Otsuki T, Clark HM, Wellmann A, Jaffe ES, Raffeld M. Involvement of CDKN2 (p16INK4A/MTS1) and p15INK4B/MTS2 in human leukemiasand lymphomas. Cancer Res 1995; 55:1436-40. 362. Elenitoba-Johnson KS, Gascoyne RD, Lim MS, Chhanabai M, Jaffe ES, Raffeld M. Homozygous deletions at chromosome 9p21 involving p16 and p15 are associated
References 157
363.
364.
365.
366.
367.
with histologic progression in follicle center lymphoma. Blood 1998; 91: 4677-85. Verma IM, Stevenson JK, Schwarz EM, etal. Rel/NFkappa B/l kappa B family: intimate tales of association and dissociation. Genes Develop 1995; 9: 2723-35. Fracchiolla NS, Lombard! L, Salina M, etal. Structural alterations of the NF-kB transcription factor lyt-10 in lymphoid malignancies. Oncogene 1993; 8: 2839-45. Chang CC, ZhangJ, Lombardi L, etal. Rearranged NFKB2 genes in lymphoid neoplasms code for constitutively active nuclear transactivators. Molec Cell Biol 1995; 15: 5180-7. Houldsworth J, Mathew S, Rao PH, et al: REL protooncogene is frequently amplified in extranodal diffuse large cell lymphoma. Blood 1996; 87: 25-9. Joos S, Otano JM, Ziegler S, et al. Primary mediastinal (thymic) B-cell lymphoma is characterized by gains of
368.
369.
370.
371.
chromosomal material including 9p and amplification of the REL gene. Blood 1996; 87:1571-8. Rao PH, Houldsworth J, Dyomina K, et al. Chromosomal and gene amplification in diffuse large B-cell lymphoma. Blood 1998; 92: 234-40. Ohno H, Doi S, Yabumoto K, et al. Molecular characterization of the t(14;19) (q32;q13) translocation in chronic lymphocytic leukemia. Leukemia 1993; 7: 2057-63. Michaux L, Dierlamm J, Wlodarska I,etal. t(14;19)/fiO3 rearrangements in lymphoproliferative disorders: a review of 23 cases. Cancer Genet Cytogenet 1997; 94: 36-43. McKeithan TW, Takimoto GS, Ohno H, et al. BCL3 rearrangements and t(14;19) in chronic lymphocytic leukemia and other B-cell malignancies: a molecular and cytogenetic study. Genes Chromosomes Cancer 1997; 20: 64-72.
This page intentionally left blank
PART
Epidemiology
Hodgkin's disease
161
Non-Hodgkin's lymphoma
169
3
This page intentionally left blank
13 Hodgkin's disease NE MUELLER
Introduction Descriptive epidemiology Analytic epidemiology
161 161 163
INTRODUCTION Hodgkin's disease (HD) has been the focus of a substantial amount of epidemiologic research. This interest reflects the uniqueness of the disease in terms of its demographic patterns of occurrence, its risk factors suggestive of an infectious etiology, and its intimate relationship with the Epstein-Barr virus (EBV). This chapter reviews the features of the distribution of the disease within and between populations and the current knowledge of its risk factors. DESCRIPTIVE EPIDEMIOLOGY The incidence of HD is less than one-quarter that of the non-Hodgkin's lymphomas (NHLs), and among younger adults displays substantial international variation in magnitude (Fig. 13.1). Overall, the incidence rate generally increases with level of economic development. The disease is more common among men than women and, in affluent populations, generally more common in whites than among minority groups. The rates are lowest in Asian populations, regardless of level of economic development. In the USA in 1986-90, the age-adjusted incidence rate was 2.8 per 100 000 person-years with the age-adjusted mortality rate being only 0.6 per 100 000 person-years, reflecting the efficacy of treatment.1 Space-time clustering has been observed in HD.2'3 The relevance of such studies to possible infective etiologies is still uncertain. The distinguishing epidemiological feature of HD is its bimodal age-incidence curve (Fig. 13.2). This is most striking in economically advantaged populations. In such populations there are very few cases occurring
Summary References
165 166
among children, a rapid increase of incidence among teenagers peaking about age 25 and then a decrease to a plateau through middle age; after about age 55 years, rates increase with advancing age to the second peak. There is generally a male excess in both age peaks. Clemmesen4 has noted the existence of a 'secondary peak' among middle-aged males in economically advantaged populations. In a landmark study published in 1971, Correa and O'Connor5 noted that, among economically disadvantaged populations, a different age pattern is evident. In this case, there is an initial peak in childhood only for boys, relatively low rates among young adults followed by the late peak among those of advanced age. They noted that, with time, in parallel to economic development, there is a decrease in disease occurrence among children and a reciprocal increase among young adults. This evolution of the young adult peak is evident in birth cohort analysis of mortality from HD among British men in this century (Fig. 13.3).6 Within the same population, the age pattern will differ between population subgroups with major differences in socioeconomic level. At present, essentially all majority populations in Europe and North America have a well-defined 'developed' pattern of HD incidence. The height of peak occurrence in young adulthood varies within this set of countries, being higher in Canada, the USA, Switzerland and France, and lower in southern Europe. The pattern within the Eastern Block countries is variable. In general, the pattern in Asia and Africa is 'intermediate' or 'developing'.7'8 In contrast to the substantial variation of the magnitude of incidence in the younger adult peak in accord with socioeconomic conditions, the rates among older persons appear to be quite stable over time and between
162 Hodgkin's disease
Figure 13.1
Estimated
cumulative rate of Hodgkin's disease for age 15-54 years by sex: (a) women, (b) men. For description of method, see Chapter 9[7].
populations. This is more evident in recent data in which there is less misclassification between HD and NHL among older persons.9 A general observation is that HD cases occurring in economically developing populations and among lower social class groups in developed populations10 are predominantly of the mixed cellularity and lymphocyte depletion subtypes, indicative of more advanced disease. The explanation for this age-specific variation in geographic patterns was offered in 1957 by MacMahon, who proposed that the bimodality results from the overlap of
two disease distributions with differing age peaks. Specifically, he proposed that among young adults, HD is caused by a biological agent of low infectivity and that, among the elderly, the cause is probably similar to those of the other lymphomas.11 From social class-specific incidence data, there is evidence that rates of HD are elevated among young adults from higher social class. Henderson et al.12 computed histologic-specific incidence rates for HD in Los Angeles County from 1972 to 1975 by social class. This group found that the incidence of nodular sclerosis HD was
Analytic epidemiology 163
Figure 13.2 Age-specific incidence rate of Hodgkin's disease for men for 5-year age groups from 1983 to 1987.7
directly related to social class, but no consistent association for the other histological types was evident. This finding was corroborated by Cozen et a/.13 in the same population in a later time period and by Glaser14 in USA incidence data from 1969 to 1980. As part of a large population-based leukemia/lymphoma registry covering about half of the UK over a 5year period, Alexander and colleagues evaluated the characteristics of over 1800 HD cases by area-based socioeconomic and population density indices. They reported that, of the 486 cases diagnosed under age 25, incidence rates were significantly somewhat greater in the high socioeconomic areas, relative risk (RR) =1.2, and there was a significant trend in increase in areas closer to 'built-up areas' with mutual adjustment.15
Image Not Available
Figure 13.3 Age-specific mortality rates of Hodgkin's disease for successive 10-year birth cohorts of males in England and Wales. Reprinted by permission of Wiley-Liss, Inc., a subsidiary of John Wiley & Sons, Inc., © 1977 from Gutensohn N, Cole P. Epidemiology of Hodgkin's disease in the young. Int J Cancer 7977; 19: 595-604.
The social class differences that are seen in the histologic presentation of HD may reflect a mix of agedependent host responses related to environmental exposures. These observations are consistent with an extension of the MacMahon hypothesis: that HD in young adults may develop as a rare consequence of a common lymphotropic infection with risk increased if infection is delayed until adolescence or young adulthood owing to better living conditions and reduced infectious exposures.6 With more clinically severe infection, which often occurs when 'childhood' infections are experienced in adulthood, the immunological control of a latent virus may be altered. In parallel, the disease seen among children reflects the effect of severe early infection. An interesting feature of the descriptive epidemiology of HD is the variation in sex ratio by age as discussed by Glaser.16 She notes that there appears to be a deficit of cases among women in their late thirties and forties in recent data from the USA and elsewhere. She proposes that this may represent a protective effect from childbearing, perhaps mediated by estrogens. However, a case-control study of 917 cases in the Swedish Cancer Register showed no significant difference between parous and non-parous women, and no link to the number of children. There was some evidence of a link between risk and later first birth.17
ANALYTIC EPIDEMIOLOGY
Socioeconomic status Analytic epidemiologic studies have confirmed that social class factors related to age of infection vary with age of diagnosis of HD through the first two-thirds of life.18 It appears that young children with HD tend to come from lower socioeconomic class and larger families, consistent with early infectious exposure. A shift in social class factors is then seen among young adult and middle-aged cases (aged 15-54 years), those who comprise the first major peak of disease. In this age group, the risk of HD has been found to be associated with a variety of risk factors associated with higher social class. As we have found in a population-based case-control study conducted in eastern Massachusetts (USA) in the 1970s, these factors include higher education, higher parental education (particularly maternal) and less dense housing in childhood.19 An exception to this statement is the observation of Alexander et al.20 in their large study in England, who reported that, while for cases aged less than 35 years at diagnosis there was a significantly positive association for social class, there was an insignificant negative association found for cases aged 35-49. A characteristic of HD occurring among young and middle-aged persons is its inverse association with sibship size, i.e. cases are more likely to come from small
164 Hodgkin's disease
families, and those from large families from the early birth orders. All of these factors foster susceptibility to late infections with the common childhood infections. For the middle-aged patients, this infectious exposure may come from their own children. Among the oldest persons, those in the second peak of HD incidence, the relationship between social class factors and risk of HD is not clear. In the Massachusetts population noted above, risk was not directly associated with social class.19 If anything, patients came from a somewhat lower social class than their controls. In the English study, there was a significant negative association with social class residence in cases aged 50-79 years.20 However, within an Israeli population, older cases appeared to come from somewhat higher social class as evidenced by level of education (RR = 2.1) and by the presence of a flush toilet in their childhood home (RR = 1.6).21 Whether these latter observations are confounded by the apparent generally increased risk of HD among adult Jews (see later) is unknown. In summary, children - particularly boys - living under relatively poor living conditions are at higher risk of HD than children elsewhere. For both young adult and middle-aged persons, there is evidence that HD may be the result of an age-related host response to a common infection. Among older persons, risk is not consistently related to social environment risk factors and shows minimal geographic variation.
Epstein-Barr virus The major candidate as an etiological infectious agent for HD is the ubiquitous herpesvirus, the Epstein-Barr virus.22 This association was first suggested by the virus's ability to transform B lymphocytes, the finding of Reed-Sternberg (RS) cells in lymphoid tissue of patients with infectious mononucleosis, and the consistent association found between history of infectious mononucleosis and HD.6 Multiple serologic studies revealed an invariable pattern of altered antibody responses to the EBV among HD patients in comparison to controls. In studies of more than 1900 HD cases of all ages, the proportion who have antibodies indicative of prior infection has been quite similar to that of controls. However, the cases consistently have higher mean immunoglobulin G (IgG) antibody titers to the EBV viral capsid antigen (VGA) than controls. Further, more of the cases have antibodies (as well as higher titers) against the early antigen (EA) of EBV, indicative of viral replication.23 Since these results were based on blood specimens collected after the diagnosis of the disease, the findings may simply reflect the reactivation of latent EBV as a result of the immune dysfunction characteristically seen in HD. However, these findings have been confirmed in two nested case-control studies that used blood specimens obtained from populations followed prospectively for
the diagnosis of HD.24,25 In both studies, the risk of HD was also associated with high antibody titers against the EBV nuclear antigen complex (EBNA). These observations imply that, for a subset of cases, the development of HD is preceded over an extended period of time by an increased level of EBV activation. Of note, the joint pattern of elevated EBV antibodies of both those of replication (anti-VGA and anti-EA) and that of latency (anti-EBNA) is not seen in normal populations.26 With the advent of highly sensitive molecular probes, striking evidence that EBV is directly related to the pathogenesis of HD has been found. These data began with the initial observation of Weiss et a/.27,28 that monoclonal episomal EBV was detectable in HD tissue and localized to the RS cells in 4 out of 21 specimens tested. This discovery has been confirmed in a large number of subsequent reports and the detection rate increased with the use of more sensitive methods.18 Overall, the findings suggest that about one-third to one-half of HD cases are EBV-genome positive. Pallesen et a/.29 further demonstrated that the EBV genome present in RS express a restricted latent phenotype of LMP1+/EBNA2-, both viral gene products normally being coexpressed in latently infected lymphocytes. This important finding has also been confirmed by many subsequent investigators.18 The consistency of the finding of clonal episomal EBV of a unique latent phenotype expressing LMP1 with its oncogene-like properties in a substantial proportion of HD cases in many patient populations throughout the world argues strongly for a causal role of the virus in these cases.30,31 The EBV-genome positivity rate is higher in mixedcell and lymphocyte-depleted HD, the subtypes associated with more advanced disease, and lower in nodular sclerosis; that for lymphocyte predominance appears to be none or extremely low, although the data are sparse and somewhat inconsistent.3 In a combined analysis of data from 14 studies involving more than 1500 HD cases, Glaser et al. evaluated the characteristics of patients with EBV-positive tumors. They found that having EBVpositive HD was associated with being male; being Hispanic; and among children, coming from less economically developed populations.32 These observations suggest that the factors affecting or reflecting immune competency are associated with EBV-genome status. In this light, a report by Frisan et al. suggests that tumor-associated suppression of EBVspecific T cell responses may play a role in EBV-positive HD. These investigators assessed the presence of HLA class I restricted EBV-specific reactivities within the tumor-infiltrating lymphocytes. Of six EBV-positive tumors tested, none had EBV-specific cytotoxic T lymphocytes (CTLs) compared to 3 of 3 EBV-negative tumors [P=0.003].33 These paradoxical findings raise a number of questions. Is the EBV involved in the pathogenesis of all HD, but the viral episome is somehow lost in persons with strong EBV-specific CTLs?34 Is
Summary 165
EBV-negative HD due to infection with another, unidentified virus?35
Immune dysfunction There is some evidence that risk of HD is increased among patients with certain primary immunodeficiencies. These cases account for about 9 per cent of all malignancies reported among these patients. HD does not appear to occur excessively among organ transplant recipients.36 It is of interest that HD is now recognized to be a part of the spectrum of opportunistic malignancies occurring in the natural history of human immunodeficiency virus-I (HIV-I) infection,37 particularly among intravenous drug abusers.38 Hessol et al. have reported a significant excess of HD in a cohort of homosexual men with HIV infection (RR = 5.0).39 Among these acquired immunodeficiency syndrome (AIDS)-related cases, the proportion that is EBV-genome positive is extremely high. In general, these HIV-I-infected patients present with advanced HD and exhibit a poor prognosis. In many, HD appears to spread non-contiguously - without mediastinal or splenic involvement.40 This alteration in the clinical behavior of HD in AIDS has been attributed to the loss of T-helper cells.
Genetic factors Several investigators have dealt with the question of whether genetic factors play a role in the etiology of HD as reviewed by Grufferman and Delzell.41 As in other hematopoietic malignancies, there is evidence for increased risk among relatives of cases. Two unusual observations warrant attention. The first is that of Grufferman et al42 who found 13 sibling cases of a total of 1577 cases in a population-based incidence survey. They reported that siblings of a young adult with HD have, overall, a roughly seven times higher risk than do members of the general population. Of note, their findings suggest that the excess risk among siblings is concentrated among those of the same sex as the first affected. That is, if a male is affected, his brothers have about a nine-fold increased risk, while his sisters have a five-fold increased risk. If a female is affected, the reverse applies. This finding was confirmed in a literature review.43 This observation argues against the theory that genetic factors play a major role in the etiology of HD, since it may be explained by childhood environmental exposures that are more likely to be shared by siblings of the same sex. In contrast, Mack et al. have recently reported that, among 366 sets of twins in which at least one was diagnosed with HD, the risk of HD in both twins was many times greater among the 179 monozygotic than among the 187 dizygotic sets.44
An unexplained observation is the consistent finding of increased risk among Jews.11,13,45 We found this also in our population-based case-control study in eastern Massachusetts. This association was not explained by social class correlates, such as family size. With control for all such factors, the increased risk was about twofold.19
Other exposures Concerning occupational exposures as risk factors for HD, only two have been found to be related with the disease in a number of studies: wood and chemicals; for neither are the data consistent. Grufferman and Delzell41 have been reviewed these reports in detail. In our study in the Boston-Worcester area (Matte and Mueller, unpublished) no association with occupation was found overall. There was an association with self-reported exposure to dust or sawdust, which increased with age (for subjects aged 15-39 years: RR = 1.1; 40-54 years, 2.1; > 55 years, 2.7). However, no dose response was evident. Turning to occupational exposure to chemicals, the findings are also quite varied. Early reports of an association of HD with herbicide exposure (primarily chlorophenols) have not been supported in a large casecontrol study in Kansas (USA).46 There are general findings of increased risk with employment in the rubber, plastics or synthetics industry and paper mill workers, but no specific exposures are evident.18 It is of interest that HD is one of the handful of malignancies that is not related to radiation exposure.47,48 This observation underscores the notion that the pathogenesis of HD is unique among the malignancies.
SUMMARY Much of the epidemiological and molecular biological evidence points to dysfunctional immunological control of latent EBV infection. The epidemiological evidence points to age at infection as an important modifier of risk. However, the current evidence does not fit together easily. What is needed is epidemiological research that concurrently integrates the molecular, serological and risk factor data. The possibility that a second virus is involved warrants consideration. Smithers49 has postulated a mechanism by which chronic antigenic stimulation could act in the pathogenesis of HD. He commented that'... we are bound to look at the evidence for the effect of prolonged pressures on the cell-mediated arm of the immune system and for feed-back failure of restraint in influencing the development of this disease.' It may be that, in HD, an alteration in gene expression occurs as a consequence of continuing antigenic stimulation from a chronically expressed
166 Hodgkin's disease EBV infection. However, in this case, the gene involved is not one controlling proliferation, but rather one controlling the expression of the normal, functionally active mediators released by antigen-stimulated cells. The alteration of gene expression may be quantitative change, resulting in a greatly amplified message. Alternatively, it may be that the gene product that normally shuts down the messages by feedback inhibition is underexpressed or dysfunctional. The result would be an immune system that is continually 'turned on', that is, an immune system that is perpetually mobilized in response to a chronic antigen and thus unable to respond to others. This hypothesis can be used to explain the biological paradox of HD: the malignant properties reflect the underlying genetic changes, the histological features reflect the response of normal immune cells to the perpetual stimulation, and the immune defects reflect the resulting imbalance in the immune response system.50
REFERENCES 1. Miller BA, Ries LAG, Hankey BF et al., eds. SEER cancer statistics review: 1973-1990. Washington DC: National Cancer Institute, NIH Publ. no. 93-2789,1993. 2. Evans AR, Hancock BW, Brown MJ, Richmond J. A small cluster of Hodgkin's disease. Br Med J 1977; 1: 1056-7. 3. Mueller N, Grufferman S. The epidemiology of Hodgkin's disease. In: Mauch P, Armitage JO, Diehl V, Hoppe RT, Weiss LM, eds Hodgkin's disease. Philadelphia: Lippencott Williams &Wilkins 1999; 61-78. 4. Clemmesen J. To the epidemiology of Hodgkin's lymphogranulomatosis.y Beige Radiol 1981; 3: 263-71. 5. Correa P, O'Connor GT. Epidemiologic patterns of Hodgkin's disease. Int J Cancer 1971; 8:192-201. 6. Gutensohn (Mueller) N, Cole P. Epidemiology of Hodgkin's disease in the young. Int J Cancer 1977; 19: 595-604. 7. Parkin DM, Muir CS, Whelan SL et al., eds. Cancer incidence in five continents, Vol. VI. IARC publication no. 120. Lyon: IARC, 1992. 8. Stiller CA. What causes Hodgkin's disease in children? Eur J Cancer 1998; 34: 523-8. 9. Glaser SL, Swatz WG. Time trends in Hodgkin's disease incidence: the role of diagnostic accuracy. Cancer 1990; 66: 2196-204. 10. Hu E, Hufford S, Lukes R, etal. Third-world Hodgkin's disease at Los Angeles County-University of Southern California Medical Center. J Clin Oncol 1988; 6:1285-92. 11. MacMahon B. Epidemiology of Hodgkin's disease. Cancer Res 1966; 26: 1189-2000. 12. Henderson BE, Dworsky R, Pike MC, et al. Risk factors for nodular sclerosis and other types of Hodgkin's disease. Cancer Res 1979; 39: 4507-11. 13. Cozen W, Katz J, Mack T. Risk patterns of Hodgkin's disease in Los Angeles vary by cell type. Cancer Epidemiol Prevent 1992; 1:261-8.
14. Glaser SL. Regional variation in Hodgkin's disease incidence by histologic subtype in the US. Cancer 1987; 60: 2841-7. 15. Alexander FE, Ricketts TJ, McKinney PA, et al. Community lifestyle characteristics and incidence of Hodgkin's disease in young people. IntJ Cancer 1991; 48: 10-14. 16. Glaser SL Reproductive factors in Hodgkin's disease in women: a review. Am J Epidemiol 1994; 139: 237^6. 17. Lambe M, Hsieh CC, Tsaih SW, Adami J, Glimelius B, Adami HO. Childbearing and the risk of Hodgkin's disease. Cancer Epidemiol Biomarkers Prev 1998; 7: 831-4. 18. Mueller N. Hodgkin's disease. In: Schottenfeld D, Fraumeni J Jr, eds Cancer epidemiology and prevention, 2nd edn. New York: Oxford University Press, 1996: 893919. 19. Gutensohn (Mueller) N. Social class and age at diagnosis of Hodgkin's disease: new epidemiologic evidence on the 'two-disease' hypothesis. Cancer Treatment Rep 1982; 66: 689-95. 20. Alexander FE, McKinney PA, Williams J, et al. Epidemiological evidence for the 'two-disease hypothesis' in Hodgkin's disease. IntJ Epidemiol 1991; 20: 354-61. 21. Abramson JH, Pridan H, Sacks Ml, etal. A case-control study of Hodgkin's disease in Israel.) Natl Cancer Inst 1978;61:307-14. 22. Evans AS. The spectrum of infections with Epstein-Barr virus: a hypothesis. J Infect Dis 1971; 124: 330-7. 23. Evans AS, Gutensohn (Mueller) N. A population-based case-control study of EBV and other viral antibodies among persons with Hodgkin's disease and their siblings. Int J Cancer 1984; 34:149-57. 24. Mueller N, Evans A, Harris NL, et al. Hodgkin's disease and Epstein-Barr virus: altered antibody pattern before diagnosis. N Engl J Med 1989; 320: 689-95. 25. Lehtinen T, Lumio J, Dillner J, et al. Increased risk of malignant lymphoma indicated by elevated Epstein-Barr virus antibodies - a prospective study. Cancer Causes Control 1993; 4:187-93. 26. Rocchi G, Tosato G, Papa G, et al. Antibodies to Epstein-Barr virus-associated nuclear antigen and to other viral and non-viral antigens in Hodgkin's disease. /Art7 Cancer 1975; 16: 323-8. 27. Weiss LM, Strickler JG, Warnke RA, et al. Epstein-Barr viral DNA in tissue of Hodgkin's disease. Am J Pathol 1987; 129:86-91. 28. Weiss LM, Movahed LA, Warnke RA, etal. Detection of Epstein-Barr viral genomes in Reed-Stern berg cells of Hodgkin's disease. N Engl J Med 1989; 320: 502-6. 29. Pallesen G, Hamilton-Dutoit SJ, Rowe M, etal. Expression of Epstein-Barr virus latent gene products in tumour cells of Hodgkin's disease. Lancet 1991; 337: 320-2. 30. Knecht H, Brousset P, Bachmann E, etal. Latent membrane protein 1: a key oncogene in EBV-related carciogenesis?/4rto Hematol 1993; 90: 167-71. 31. Herbst H, Niedobitek G. Epstein-Barr virus and Hodgkin's disease. IntJ Clin Lab Res 1993; 23:13-16.
References 167
32. Glaser SL, Lin RJ, Steward SL, et al. Epstein-Barr virus-
42. Grufferman S, Cole P, Smith PG, et al. Hodgkin's disease
associated Hodgkin's disease: epidemiologic characteristics in international data. Int J Cancer 1997: 70: 375-86. 33. Frisan T, SjbbergJ, Dolcetti R, etal. Local suppression of
in siblings. N Engl J Med 1977; 296: 248-50. 43. Grufferman S, Barton JW III, Eby NL. Increased sex concordance of sibling pairs with Behcet's disease, Hodgkin's disease, multiple sclerosis and sarcoidosis. Am
Epstein-Barr virus (EBV)-specific cytotoxity in biopsies of EGV-positive Hodgkin's disease. Blood 1995; 86:1493-501. 34. Mueller NE. Epstein-Barr virus and Hodgkin's disease: an epidemiological paradox. Epstein-Barr Virus Rep 1997; 4: 1-2. 35. Jarrett RF. Epstein-Barr virus and Hodgkin's disease. Epstein-Barr Virus Rep 1998; 5: 77-85. 36. Mueller N. Overview of the epidemiology of malignancy in immune deficiency. JAIDS 1999; 21: S5-S10. 37. Goedert JJ, Cote TR, Virgo P, et al. Spectrum of AIDSassociated malignant disorders. Lancet 1998; 351:1833-9. 38. Roithmann S, Tourani J-M, Andrieu J-M. Hodgkin's disease in HIV-infected intravenous drug abusers. N Eng J MedWW; 323:275-6.
J Epidemiol 1987; 126: 365-9. 44. Mack TM, Cozen W, Shibata DK, et al. Concordance for Hodgkin's disease in identical twins suggesting genetic susceptibility to the young-adult form of the disease. N EnglJ Med 1995; 332: 413-18. 45. Bernard SM, Cartwright RA, Darwin CM, et al. Hodgkin's disease: case control epidemiological study in Yorkshire. BrJ Cancer W87; 55:85-90. 46. Hoar SK, Blair A, Holmes EF, et al. Agricultural herbicide use and risk of lymphoma and soft-tissue sarcoma.y/AM/4 1986; 256: 1141-7. 47. Hainan, KE. Failure to substantiate two cases of alleged occupational radiation carcinogenesis. Lancet 1988; 1: 639.
39. Hessol NA, Katz MH, Liu JY, et al. Increased incidence of
48. Boice JD Jr, Land CE, Preston DL. Radiation. In:
Hodgkin's disease in homosexual men with HIV infection. Ann Intern Med 1992; 117: 309-11. 40. Knowles DE, Chamulak GA, Subar M, et al. Lymphoid neoplasia associated with the acquired immunodefi-
prevention 2nd edn. New York: Oxford University Press 1996; 319-54. 49. Smithers D. On some general concepts in oncology with
ciency syndrome (AIDS): the New York University Medical Center experience with 105 patients (1981-1986). Ann Intern Med 1988; 108: 744-53. 41. Grufferman S, Delzell E. Epidemiology of Hodgkin's disease. Epidemiol Rev 1984; 6: 76-106.
Schottenfeld D, Fraumeni J, eds Cancer epidemiology and
special references to Hodgkin's disease. Int J Radial Oncol Biol Phys 1983; 9: 731-8. 50. Mueller NE. The epidemiology of Hodgkin's disease. In: Selby D, McElwain TJ, eds Hodgkin's disease Oxford: Blackwell Scientific Publications, 1987; 68-93.
This page intentionally left blank
14 Non-Hodgkin's lymphoma RACARTWRIGHT
Introduction Descriptive epidemiology Analytic epidemiology
169 170
The future References
174 174
172
INTRODUCTION The epidemiology of non-Hodgkin's lymphoma (NHL) presents serious difficulties. As a result there are still many gaps in our knowledge of the distribution and causes of these conditions. There are several reasons for this. The classification systems have been in flux for many years. As opinions about the disease have changed, so a series of classifications have been produced. This had led to problems for epidemiologists who collect routine and often historical information from cancer registries or from other special studies. All the various classifications, dating back over 30 years, have been used by epidemiologists at one time or another, many of which are regarded as out of date by contemporary histopathologists. Two aspects of these classifications should be noted: what is included or excluded as 'NHL', and the internal subtypes of the disease. Both have changed and these changes vary from country to country. Thus comparisons over time and/or internationally are fraught with difficulties of interpretation. Classification errors might account for 10-15 per cent of wrongly associated cases.1 This has led many epidemiologists simply to avoid the issue of the need to account for or incorporate subtypes of NHL in their studies. They have assumed NHL to be one cohesive pathological entity. This is nowadays at odds with the concepts of the histopathologists whose classification strategies have gradually shifted from those which purely reflected disease outcome to those more truly in line with the recognized pathogenic processes. In addition, most epidemiologists have tended to ignore the site of origin of the NHL. A further noteworthy event has been that the use of cell-surface markers has led to more undifferentiated tumors being classified as NHL.
A related difficulty lies in the pathological status of Hodgkin's disease (HD). Originally regarded as distinct entities, the pathological boundaries between HD and NHL have become increasingly blurred. This is particularly true in the older, non-nodular sclerosing cases where a decline in case numbers has been interpreted as a possible diagnostic artefact of a tendency to classify more such cases as NHL. Further, many risk factors for HD, particularly, but not exclusively, in older cases, are similar to those for NHL. The boundaries between chronic lymphocytic leukemia (CLL) and multiple myeloma (MM) are also blurred from an epidemiological viewpoint, although they are relatively clear diagnostically. The relatively rare lymphoblastic lymphoma is regarded as part of the acute lymphoblastic leukemia (ALL) spectrum and is not referred to in this chapter, along with CLL and MM. In summary, because of the chaos of classification and the limitation of many earlier epidemiological studies, much of the descriptive epidemiology of NHL is difficult to interpret, and could be full of spatial and temporal artefacts. Some of the earlier epidemiological attempts to find causal links may also be seriously flawed as a consequence of the pooling of all NHL cases together or the inappropriate use of systems of classification. This could mean that highly significant risk factors have been diluted or completely lost. Although most of these challenges are now being addressed by epidemiologists, the data from this new generation of studies are just becoming available. However, the reader must still interpret epidemiology data with caution. For example, Anderson et al. found substantial differences in the geographical distribution of NHL subtypes (by the Revised European-American Classification).2
170 Non-Hodgkin's lymphoma
DESCRIPTIVE EPIDEMIOLOGY
Rates by classification of NHL by all type and all sites In view of the difficulties noted in the introduction, only the recently available data are supplied. Table 14.1 gives incidence rate data by age and sex for recently acquired information from parts of the UK, collected as a result of a special study spanning 1984-1993.3 These rates are typical of many European populations in that they are greater in males at all ages and show a steadily increasing incidence in both sexes with the highest rate being in the elderly.
Internationally, however, there are greater differences and some selected populations are given in Table 14.2 to reflect this. These are taken from data computed by the International Agency for Research on Cancer4 and roughly represent cases occurring between the years 1982 and 1989. Higher childhood rates are seen in certain countries, such as in parts of sub-Saharan Africa, where it is largely due to the Burkitt-type lymphoma.5 Asian children living in the UK may have a higher NHL rate than in the white population.6 This is also true in some Arab children.7 Generally, Middle Eastern and Caucasian populations have the highest rates of NHL in all older age groups. There is some slight evidence to suggest rates in Europe are higher in the north and central parts than in the far south.4
Table 14.1 Age-specific incidence rate (x per 100 000 per year) for cases of NHL collected from geographically defined areas of the UK 1984-1993 inclusive based on 13 901 cases
0-4 10-14 20-24 30-34 40-44 50-54 60-64 70-74 80-84 90+
0.61 1.07 1.56 2.93 7.22 1449 23.34 41.85 48.02 47.92
0.46 0.35 0.82 2.21 5.03 9.54 16.97 28.07 31.26 20.00
Rates by all histological types by site until recently mere nas been very little systematic data reported by the site of origin of the NHL. A recent special collection from the UK3 has allowed this to be rectified. Table 14.3 gives the relative proportion of the various sites based on the first histopathology report in each case. The 'skin' site includes mycosis fungoides and Sezary's syndrome. Table 14.4 gives some age-specific incidence rates for these sites. The rarity of some sites is noteworthy as is the brief span of risk for others, e.g. the testes.
0.54 0.72 1.20 2.58 6.13 12.01 20.01 33.98 36.79 25.55
Table 14.2 Age-specific incidence rates in selected populations
South America Columbia
1.0
0.7
2.1
0.9
1.7
2.9
10.1
9.6
19.0
32.9
North America Connecticut whites Connecticut blacks
0.8 1.4
1.5 0.0
0.8 1.5
0.6 1.6
4.3 3.4
2.4 4.5
16.0 11.2
14.2 6.7
79.0 44.2
58.0 21.6
Asia India - Bombay
0.8
0.3
1.1
0.3
1.5
1.0
5.9
5.6
19.9
15.5
Israel Jews
1.6
0.9
1.9
0.3
3.3
3.6
15.5
18.3
5.6
Japan Miyagi
0.7
0.7
3.0
1.4
1.8
1.2
5.0
3.3
33.3
19.5
Europe Finland Belarus Soab- Basque
1.3 1.3 1.1
0.6 0.9 2.3
1.8 1.3 1.6
0.5 0.9 3.4
3.1 1.6 2.6
1.7 1.1 1.3
14.9 4.2 8.0
10.0 2.2 6.6
48.5 12.5 38.8
38.2 6.4 19.4
Australasia New South Wales
1.2
0.6
1.0
0.5
4.5
2.3
18.0
14.7
63.1
46.5
44.4
Descriptive epidemiology 171
Table 14.3 Site of origin of non-Hodgkin's lymphoma in cases aged 0-79 based on 12 033 cases collected 1984-1993
Lymph node, marrow, blood, tonsils Gastrointestinal tract Skin Bone, orbit, jaw Spleen Liver Central nervous system Testes Breast Other
80.00 8.35 6.12 1.37 1.15 0.98 0.11 0.07 0.05 1.80
Seascale near British Nuclear Fuels Ltd, Sellafield.17 The inclusion in such studies of NHL in both children and young adults (under age 25) is controversial, as they are pooled with ALL cases. It is unlikely that the NHL cases have all been of the lymphoblastic type. The major type of 'clustering' occurs within families. Secular trends
Variation in rates in the UK by all histological types and sites The use of panel-reviewed data has enabled local variation to be examined in the UK at various geographical levels from that of administrative county to the small electoral wards. This highlights various features, particularly a non-random distribution of cases between both counties and administrative districts, with an excess of cases in the south of England compared to the north.3 This has been confirmed in a mortality study.8 Further, the data confirm an excess of cases in predominantly rural areas, an observation made also in North America.9 At the electoral ward level, a non-random and unusual case distribution has been described in parts of the UK,10,11 suggesting local rural excesses. Local case clusters have also been described for the disease in Africa, owing to Burkitt's lymphoma and its links with the EpsteinBarr virus (EBV) and malaria.12 Post hoc clusters in neighborhoods have been described occasionally in various countries.13-16 However, the number of reports are limited compared to HD or leukemia, for example. It should also be noted that NHL in young people has been associated with the 'cluster' reports relating to nuclear power plants of various kinds in the UK, e.g.
One of the most remarkable features of NHL epidemiology is the increase in recorded rates of the condition over the years. This increase is true for most age groups over 30 years of age and for both sexes. No study has shown any increasing trend in childhood NHL.18 This is particularly so for Caucasian populations, and has been recorded from Europe, North America and Australasia. The increase is also seen in black populations in the USA but at a lower rate, and is not seen in African or southwest Asian peoples.19 The same trends have also been seen in UK populations.20 In a series from Yorkshire utilizing a reviewed histopathology panel dataset for 1977-1992, the increase is a remarkable 5 per cent per annum21 but this has been exceeded by a series from France.22 NHL sited in the central nervous system (CNS) has been shown to be on the increase in London.23 However, a Danish study has shown no increase in NHL of the gastrointestinal tract.24 There is no doubt that the trends exist. The controversy revolves around the reasons for it. Because of the problems outlined in the introduction, it has long been argued that the increases are artefactual and the result of increased diagnostic acumen, an increased biopsy rate and the improved ability to distinguish NHL from other tumors. This could, for example, account for some increase in CNS lymphomas due to the advent of better imaging techniques.23,25 To investigate one aspect of this, two series of nodal tumors were abstracted from the same geographical area some years apart and all were reviewed again by a panel. The results showed a uniform increase in incidence in all ages over 30 for NHL, whilst HD only showed an increase in the elderly. The latter
Table 14.4 Age-specific incidence (x per 100000 per year) of NHL non-Hodgkin's lymphomas of all histopathological types by site (sexes pooled where appropriate)
Lymph node, marrow, blood, tonsils Gastrointestinal tract Skin Bone, orbit, jaw Spleen Liver Central nervous system
Testes Breast
0.28 0 0.02 0 0 0.2 0 0 0
0.37 0.07 0.01 0 0 0.01 0.01 0.04 0
1.65 0.12 0.19 0.05 0 0.01 0.02 0 0.09
9.44 0.98 0.83 0.12 0.08 0.11 0.14 0.08 0.05
28.22 2.77 2.22 0.38 0.44 0.33 0.33 0.61 0.25
172 Non-Hodgkin's lymphoma
could be due to increased or more thorough investigations, whilst the former argues for a biological and not a spurious phenomenon.26 Nowadays the consensus is for the existence of a real 'biological' increase but of unknown causation (see later). Some attempts have been made to examine which, if any, subtypes are increasing. Some studies have suggested a general increase using various classification systems but one has implied that large cell disease is responsible.27 If the rates of increase continue, varying between 3 and 10 per cent per annum, NHL will be amongst the commonest cancers in Europe in the next few years.28
ANALYTIC EPIDEMIOLOGY Given the caveats noted in the introduction, piecing together the jigsaw of published studies can be difficult, in some cases, to the point of speculation. However, it is apparent when reviewing the literature that the attention of research workers has fallen into three broad categories: aspects of altered immunity (using its broadest interpretation), occupational studies, and a ragbag of 'lifestyle' studies. The most plausible area of research is that dealing with aspects of immune competence.
Studies of altered immunity INHERITED SYNDROMES AND SUSCEPTIBILITY VIA INDIVIDUAL INHERITED GENETIC SEQUENCES
Several studies have suggested that a slight excess of lymphoma cases occur in blood relatives. The significance of this is not yet known; however, there are a range of generally rare, but well-defined and usually simply inherited conditions that have an excess of NHL as part of a syndrome. Such conditions are typified by ataxia telangiectasia. This is an autosomal recessive condition due to a gene deficit on chromosome 1 lq.29 The syndrome gives a progressive ataxia, many infections and progeria. Immunologically, there are a complex array of deficits, including impaired synthesis of immunoglobulin A (IgA).30 A wide range of cancers occur in these individuals but these include NHL probably more frequently than any other.31 Although such a condition is very rare, theoretically the unaffected heterozygotes could occur quite commonly in the population and it is thus a question of great significance to learn of the health of such individuals. So far only studies of the parents of cases (who are necessarily heterozygotes) have been published, with a suggestion of excess cancers but with such small case numbers that the results are difficult to interpret.32 However, the same argument would be true of all the other autosomal recessive and sex-linked conditions of impaired immunity.
These would include Chediak-Higashi syndrome,33 Wiskott-Aldrich syndrome,34 Bloom's syndrome,35 common variable immunodeficiency, the various agammaglobulinaemias and the severe combined immunodeficiency disorder.36 If the Hardy-Weinberg law of population genetics holds for all these very rare homozygous or sex-linked conditions, the frequency of the heterozygotes in the general population would be quite common. It will be necessary to identify the molecular characteristics of the gene sites in order to identify the carriers and then to determine the status of such people as possibly susceptible to NHL. The potential underlying genetic defects could represent a major initiating cause of NHL, although it would be unlikely to be the only step required to manifest the disease in these heterozygotes. IMMUNODEFICIENCY DUE TO PAST MEDICAL HISTORY
There is an excess of various cancers in patients following organ transplantation, consistent with the intensity of the immunosuppressive regimen.37 The risk of NHL in such patients can be very high, between 20- and 60-fold, but varying from study to study. One other consistently common secondary malignancy is skin carcinoma.38 Some series suggest that extranodal NHL of the CNS is particularly common. Individuals receiving immunosuppressive therapy without transplantation also have a risk of NHL;39 for rheumatoid arthritis cases the risk is roughly ten-fold. There is, however, a problem with all these studies. This is highlighted by more recent reports, which recognize the existence of post-transplant lymphoproliferative disease,40 the diagnosis of which may have been confused in earlier studies, with the contemporary understanding of NHL. Even if this is the case, some underlying risk must exist for NHL in those immunosuppressed by drug therapy. A further complexity in interpreting this outcome for conditions such as rheumatoid arthritis is that the condition itself, not necessarily with the use of immunosuppressive therapy, may have an excess of NHL,41-43 although only a 2-4-fold excess has been observed. This may also apply to other chronic conditions associated with altered immunity such as systemic lupus.44 VIRAL IMMUNOSUPPRESSION
The role of EBV in African Burkitt's lymphoma and in nasopharyngeal carcinoma, in concert with other chronic parasitic infections, is well recognized.45,46 The role of EBV in other NHL types is less well understood, although a small number of cases do have active EBV involvement. However, in most instances, EBV does not appear to be a significant pathogenic factor in the majority of NHL cases. The role of human immunodeficiency
Analytic epidemiology 173
virus (HIV) as an immune depleting agent, is much clearer, however, both from studies of HIV in the general public and in hemophiliacs given HIV-positive Factor VIII preparations. As far as acquired immunodeficiency syndrome (AIDS) cases are concerned, roughly 3 per cent acquire some type of NHL. This equates to a 60-fold risk when cases are contrasted with the general unaffected population.47 Of the NHL in a series from the USA, nearly 60 per cent were described as 'immunoblastic', approximately 20 per cent as primary CNS tumors and approximately 20 per cent as 'Burkitt' in type. The risk in a series of purely HIV-infected hemophiliacs was half that of the AIDS study noted above. Almost all such cases were described as extranodal but for a very wide range of sites, including gastrointestinal tract, skin and CNS.48 The relationship between human T-cell lymphotropic virus type 1 (HTLV-1) and the development of acute T cell leukemia/lymphoma is now much clearer in the endemic areas of Japan and the Caribbean. In Jamaica the risk of NHL in those infected with HTLV-1 is roughly 10-fold.49 Most HTLV-1 infections are a result of vertical transmission and thus the condition lends itself to public health measures. Other HTLV-1 endemic areas have been identified in Africa50 and the USA51 and South America.52 The role of needle transmission and the status of HTLV-2 are still under investigation. The newly identified herpes-like agent associated with Kaposi's sarcoma may also have a role in NHL.53 Known or unknown viruses have also been postulated to account for the excess of NHL found in blood transfusion recipients.54 CHEMICALLY INDUCED IMMUNOSUPPRESSION
Many of the chemicals possibly linked to NHL are thought to have immunosuppressive qualities.55
for in the calculation. These could be the mundane explanation as to why NHL excesses occur quite commonly in this type of literature. The links with occupations have been extensively reviewed elsewhere56 and the salient results are listed below. AGRICULTURE
Various occupations related to agriculture have been associated with an excess risk of NHL.57 Overall, at least 20 studies have seen statistical excesses in farmers, horticulturists and contract applicators.56 In addition, a lowgrade NHL excess is seen in farm animal breeding workers.58 However, a large study of herbicide production workers showed no NHL excess,59 nor has any study of dioxin-exposed workers shown any NHL excess, although the general population of Seveso does show a statistical excess but only for males.60 Attention has been given to the particular exposures of farmers themselves and what substances, if any, might be responsible. One review suggested that 2;4D is significantly involved,61 while another study has incriminated atrazine.62 In a more recent workers-exposure study, no significant excess of NHL emerged for any subtype of phenoxyherbicides, chlorophenols or dioxins with only a minor possible effect of usage of the insecticide lindane.63'64 No firm biological evidence has so far been found to support causal links between NHL and any such substances. It has been clearly demonstrated that veterans of the Vietnam war show an excess of NHL.65 This remains unexplained but it has been suggested that the phenomenon may be a result of exposure to defoliants. However, close examination of the data show this excess is confined to naval veterans. Studies of possible links between Agent Orange and contaminants with NHL and sarcoma are ongoing.66
Occupational links PETROCHEMICAL INDUSTRY
The risks associated with NHL's links with immune alterations range from four-fold rising to roughly 60fold. The occupational risks, however, are very different, with risks being four-fold at most but usually far smaller. Despite this, the literature on this aspect of etiology is very large. This should not be taken as a sign that occupations are deemed to be particularly significant in the pathogenesis of these conditions. Occupational cohort and other studies 'turn up' statistically significant results, a proportion at random, and there are large numbers of such studies. Secondly, the calculation of'expectated case numbers' in these studies is made on a basis of known national rates. In the case of NHL, these may be too low, owing to the problems of basic ascertainment (see 'Introduction'), that is, the comparison between a wellinvestigated cohort and national rates are inappropriate. Also the increasing trends in incidence are not accounted
At least four studies have shown a statistical excess of NHL with the petrochemical industry.56 However, two cohort studies were negative.67,68 This indicates the risk may be weak or of uncertain significance. OTHER INDUSTRIES
A wide variety of other industries have been named sporadically as having an excess risk of NHL. These include, firefighters,69 flour industry workers,70 asbestos exposed workers,71,72 dry cleaners,73 nickel refinery workers,74 carpenters, painters, plasterers and others in the wood and building trades.75-77 All the risks are either very low or based on very small numbers and, with a few exceptions, have little support from biological studies.78 Most of these observations are unsupported by independent studies.
174 Non-Hodgkin's lymphoma
Lifestyle and other exposures
REFERENCES
Most studies have shown little statistically significant relationship between NHL and cigarette smoking. There are, however, studies showing, albeit weak, risks and some evidence of a dose response relationship and a relationship of risk to age.79-81 Attempts to analyse the relationship between smoking and different subtypes of NHL are fraught with difficulty, the association with follicular lymphomas possibly being the strongest.82 The evidence for a relationship with exposure to ionizing irradiation is equally sparse, amounting to very little risk consequential on the atomic bomb explosions83 or from diagnostic X-rays,84 but the results from alphairradiation via the use of diagnostic thorotrast demonstrate a definite risk, although this is difficult to quantify.85 Despite this, there are no unequivocal data available to suggest a simple relationship in risk between NHL and these two acknowledged human carcinogens. This maybe due to the protean nature of NHL, with possibly a strong risk with some subtypes of NHL only, or perhaps because lymphomagenesis is solely dependent on chronic immune interference. As to other exposures, there is in general no convincing aspect of dietary exposure linked to risk of NHL,86-87 including non-alcoholic beverages88 and alcohol.89 In some populations, data suggested a possible link to meat intake.90 It has been suggested that the use of hair colorants gives a risk of NHL91 but this has been challenged.92 Attempts have been made to associate residential proximity to industrial sites with the risk of NHL in adults or children, with varying success and little by way of convincing results.93-95
1. Palackdharry CS. The epidemiology of non-Hodgkin's lymphoma: why the increased incidence? Oncology 1994; 8: 67-75. 2. Anderson JR, Armitage JO, Weisenburger DD, for the NonHodgkin's Lymphoma Classification Project. Epidemiology of the non-Hodgkin's lymphomas: distribution of the major subtypes differ by geographical locations. Ann Oncol 1998; 9: 717-20. 3. Cartwright RA, McNally RJQ, Rowland DJ, etal. The descriptive epidemiology of leukaemia and related conditions in parts of the United Kingdom 1984-1993. London: Leukaemia Research Fund, 1997. 4. Parkin DM, Muir CS, Whelan SL, eds. Cancer incidence in five continents, Vol. VI. IARC Scientific Publications 120. Lyon: IARC, 1992. 5. Stiller CA, Parkin DM. International variations in the incidence of childhood lymphomas. Paediat Perinatal Epidemiol 1990; 4: 303-24. 6. Powell J, Parkes SE, Cameron AH, et al. Is the risk of cancer increased in Asians living in the UK? Arch Disease Childhood 1994; 71: 398-403. 7. Revesz T, Mpofu C, Oyejide C. Ethnic differences in the lymphoid malignancies of children in the United Arab Emirates. A clue to aetiology? Leukemia 1995; 9:189-93. 8. Swerdlow A, Silva Ide S. Atlas of cancer incidence in England and Wales 1968-85. Cancer Research Campaign. Oxford: Oxford University Press, 1993. 9. Pickle LW, Mason TJ, Howard N, et al. Atlas of US cancer mortality among whites: 1950-1980. DHHS Publ No. (NIH) 87-2900. Washington, DC: US Government Printing Office, 1987. 10. Eddington GM. The Burkitt lymphoma in the Northern Savannah of Nigeria. Prog Clin Biol Res 1981; 53: 133-49. 11. Barnes N, Cartwright RA, O'Brien C, et al. Variation in lymphoma incidence within Yorkshire health region. BrJ Cancer W87; 56:169-72. 12. Smith PG. Current assessment of 'case clustering' of lymphomas and leukaemias. Cancer 1978; 42:1026-34. 13. Schimpff SC, Schimpff CR, Brager DM, et al. Leukaemia and lymphoma patients interlinked by prior social contact. Lancet 1975; i: 124-9. 14. Dowsett EG. Leukaemia in Kingston, Surrey, 1958-64: an epidemiological study. BrJ Cancer 1966; 20:16-31. 15. Mainwaring D, Martin J. Leukaemia and reticuloses. Br Med J 1968;ii:702. 16. Kolandaivelu G. A cluster of non-Hodgkin's lymphoma. Indian Paediat 1988; 25: 583. 17. Draper GJ, Stiller CA, Cartwright RA, et al. Cancer in Cumbria and in the vicinity of the Sellafield nuclear installation, 1963-1990. BrMedJ 1993; 306: 89-94. 18. Blair V, Birch JM. Patterns and temporal trends in the incidence of malignant disease in children: I. Leukaemia and lymphoma. EurJ Cancer 1994; 30A: 1490-9. 19. Devesa SS, Fears T. Non-Hodgkin's lymphoma time
THE FUTURE A close reading of the preceding chapter will have suggested ways in which epidemiological studies of NHL can be improved and indeed some such studies are now underway, with better diagnostic and exposure markers. The problems as to how to subdivide the disease and why the condition is increasing so much and in such a widespread fashion, however, still remain. Common sense would suggest that the pathways of greatest influence upon lymphomagenesis are through the many areas of immune depletion. Unless major selective pressures exist, the gene pool of immune deficiency syndrome heterozygotes is unlikely to be increasing disproportionately to the total population. Further, whatever is causing an increase must be very common. It has been argued that this could be due to the immune modifying effects of sunlight,28 but other models based on antibiotic usage and traffic pollution also exist. Given the future potential burden of the disease on society, studies to address these and related issues are necessary.
References 175
20.
21.
22.
23. 24.
25. 26.
27.
28.
29.
30. 31.
32.
33.
34.
35.
36.
trends: United States and international data. Cancer Res 1992; 52 (suppl.): 5432s-40s. Cartwright RA, McKinney PA, Barnes N. Epidemiology of the lymphomas in the United Kingdom: recent developments. Bailliere's Clin Haematol 1987; 1: 59-76. McNally RJQ, Alexander FE, Staines A, et al. Effect of age, period and cohort on the incidence of non-Hodgkin's lymphoma in Yorkshire, UK, 1978-1991. IntJ Epidemiol 1997; 26: 32-46. Warren HW, Anderson JH, O'Gorman P, et al. A phase II study of regional 5-flourouracil infusion with intravenous folinic acid for colorectal liver metastases. BrJ Cancer 1994; 70: 677-80. Lutz JM, Coleman MP. Trends in primary cerebral lymphoma. BrJ Cancer 1994; 70: 716-8. d'Amore F, Brincker H, Gronbaek K, et al. Non-Hodgkin's lymphoma of the gastrointestinal tract: a populationbased analysis of incidence, geographic distribution, clinicopathologic presentation features and prognosis.J Clin Oncol 1994; 12:1673-84. Grant JW, Isaacson PG. Primary central nervous system lymphoma. Brain Pathol 1992; 2: 97-109. Barnes N, Cartwright RA, O'Brien C, et al. Rising incidence of lymphoid malignancy; true or false? BrJ Cancer 1988; 53: 393-8. Weisenburger DD. Epidemiology of non-Hodgkin's lymphoma: recent findings regarding an emerging epidemic. Ann Oncol 1994; 5: S19-S24. Cartwright RA, McNally R, Staines A. The increasing incidence of non-Hodgkin's lymphoma (NHL): the possible role of sunlight. Leuk Lymphoma 1994; 14: 387-94. Gatti R, Berkel I, Boder E, et al. Localisation of an ataxiatelangiectasia gene to chromosome 11q22-23. Nature 1989; 336: 577-80. Buckley RH. Immunodeficiency diseases. JAMA 1987; 258:2841-50. Morrell D, Cromartie E, Swift M. Mortality and cancer incidence in 263 patients with ataxia-telangiectasia.y Nat Cancer Instl 986; 77: 89-92. Swift M, Reitnauer P, Morrell D, et al. Breast and other cancers in families with ataxia-telangiectasia. N EnglJ Med 1987; 316:1289-94. Merino F, Henle W, Duque-Ramirez P. Chronic active Epstein Barr virus infection in patients with Chediak-Higashi syndrome.y Clin Immunol 1986; 6: 299-305. Perry GS, Spector B, Schuman LM, et al. The Wiskott-Aldrich syndrome in USA and in Canada (1892-1979). J Pediat 1980; 97: 72-4. Aldrich RA, Steinberg AG, Campbell DC. Pedigree demonstrating a sex-linked recessive condition characterised by draining ears, eczematoid dermatitis and bloody diarrhea. Pediatrics 1954; 13:133-9. Strigini P, Carobbi S, Sansone R, et al. Molecular epidemiology of cancer in immune deficiency. Cancer Detect Prevent 1991; 15:115-27.
37. Kinlen LJ. Immunosuppressive therapy and cancer. Cancer Surveys 1982; 1: 565-83. 38. Kinlen LJ, Sheil AGR, Peto J, et al. Collaborative United Kingdom-Australasian study of cancer in patients treated with immunosuppressive drugs. BrMedJ 1979; 2: 1461-6. 39. Kinlen LJ. Incidence of cancer in rheumatoid arthritis and other disorders after immunosuppressive treatment. AmJ Med 1985; 78(suppl 1A): 44-9. 40. Morgan G, Superina RA. Lymphoproliferative disease after pediatric liver transplanation.y PediatSurg 1994; 29:1192-6. 41. Tennis P, Andrews E, Bombardier C, et al. Record linkage to conduct an epidemiologic study on the association of rheumatoid arthritis and lymphoma in the province of Saskatchewan, Canada.) Clin Epidemiol 1993; 46: 685-95. 42. Gridley G, McLaughlin JK, Ekbom A, et al. Incidence of cancer among patients with rheumatoid arthritis. J Nat Cancer Inst 1993; 85: 307-11. 43. Isomaki HA, Hakulinen T, Joutsenlahti U. Excess risk of lymphomas, leukemia and myeloma in patients with rheumatoid arthritis.) ChronicDis 1978; 31: 691-6. 44. Dupla ML, Khamashta M, Garcia VP, et al. Malignancy in systemic lupus erythematosus: a report of five cases in a series of 96 patients. Lupus 1993; 2: 377-80. 45. de The G. Virus-associated lymphomas, leukaemias and immunodeficiencies in Africa. IARC Scientific Publication 63. Lyon: IARC, 1984; 727-44. 46. de The G. Epstein-Barr virus behaviour in different populations and implications for control of Epstein-Barr virus-associated tumours. Cancer Res 1976; 36: 692-5. 47. Beral V, Peterman T, Berkelman R, et al. AIDS-associated non-Hodgkin's lymphoma. Lancet 1991; 337: 805-9. 48. Ragni MV, Belle SH, Jaffe RA, et al. Acquired immunodeficiency syndrome-associated non-Hodgkin's lymphomas and other malignancies in patients with hemophilia. Blood 1993; 81:1889-97. 49. Mann A, Cleghorn FR, Falk RT, et al. Role of HTLV-I in development of non-Hodgkin's lymphoma in Jamaica and Trinidad and Tobago. Lancet 1993; 342:1447-50. 50. Muller N. The epidemiology of HTLV-I infection. Cancer Causes Control 1991; 2: 37-52. 51. MacMahon B. Epidemiological evidence on the nature of Hodgkin's disease. Cancer Res 1966; 26:1189-2000. 52. Gerard Y, Lepere JF, Pradinaud R, et al. Clustering and clinical diversity of adult T-cell leukemia/lymphoma associated with HTLV-I in a remote black population of French Guiana. IntJ Cancer 1995; 60: 773-6. 53. Moore PS, Chang Y. Detection of herpesvirus-like DNA sequences in Kaposi's sarcoma in patients with and those without HIV infection. N EnglJ Med 1995; 332:1181-91. 54. Blomberg J, Moller T, Olsson H, et al. Cancer morbidity in blood recipients- results of a cohort study. EurJ Cancer 1993;29A:2101-5. 55. Hardell L, Axelson 0. Environmental and occupational aspects on the etiology of non-Hodgkin's lymphoma. Onto/fta 1998; 10:1-5.
176 Non-Hodgkin's lymphoma 56. McNally RJQ, Cartwright RA. Epidemiology of nonHodgkin's lymphoma. In: Burnett A, Armitage J, Newland A, Keating A, eds Cambridge Medical Review. London: 1994; 3:1-34. 57. Blair A, Linos A, Stewart PA, et al. Comments on occupational and environmental factors in the origin of non-Hodgkin's lymphoma. Cancer Res 1992; 52: 5501 s-2s. 58. Amadori D, Nanni 0, Falcini F, et al. Chronic lymphocytic leukaemias and non-Hodgkin's lymphomas by histological type in farming-animal breed ing workers: a population case-control study based on job titles. Occupat Environ Med 1995; 52: 374-9. 59. Saracci R, Kogevinas M, Bertazzi P-A, et al. Cancer mortality in workers exposed to chlorophenoxy herbicides and chlorophenols. Lancet 1991; 338: 1027-32. 60. PA Pesatori, Consonni D, et al. Cancer incidence in a population accidentally exposed to 2,3,7,8tetrachlorodibenzo-para-dioxin. Epidemiology 1993; 4: 398-406. 61. Zahm SH, Blair A. Pesticides and non-Hodgkin's lymphoma. Cancer Res 1992; 52: 5485s-8s. 62. Zahm SH, Weisenburger DD, Cantor KP. Role of the herbicide atrazine in the development of non-Hodgkin's lymphoma. ScandJ Work Environ Health 1993; 19: 108-14. 63. Blair A, Cantor KP, Zahm SH. Non-Hodgkin's lymphoma and agricultural use of the insecticide lindane. AmJ Indust Med 1998; 33: 82-7. 64. Kogevinas M, Kauppinen T, Winkelmann R, et al. Soft tissue sarcoma and non-Hodgkin's lymphoma in workers exposed to phenoxy herbicides, chlorophenols and dioxins: two nested case-control studies. Epidemiology 1995; 6: 396-402. 65. Namboodiri KK, Harris RE. Hematopoietic and lymphoproliferative cancer among male veterans using the veterans administration medical system. Cancer 1991;68:1123-30. 66. Kramarova E, Kogevinas M, Anh CT, et al. Exposure to Agent Orange and occurrence of soft-tissue sarcomas or non-Hodgkin lymphomas: an ongoing study in Vietnam. Environ Health Perspect 1998; 106(suppl 2): 671-8. 67. Ott MG, Olson RA, Cook RR, et al. Cohort mortality study of chemical workers with potential exposure to the higher chlorinated dioxins. J Occupat Med 1987; 29: 422-9. 68. Christie D, Robinson K, Gordon I, et al. A prospective study in the Australian petroleum industry. II. Incidence of cancer. BrJ Indust Med 1991; 48: 511-14. 69. Sama SR, Martin TR, Davis LK, et al. Cancer incidence among Massachusetts firefighters, 1982-1986. AmJ Indust Med 1990; 18: 47-54. 70. Alavanja MCR, Blair A, Masters MN. Cancer mortality in the US flour industry, y Natl Cancer Inst 1990; 82: 840-8. 71. Lieben J, Harrisburg PA. Malignancies in asbestos workers. Arch Environ Health 1966; 13: 619-21.
72. Gerber MA. Asbestosis and neoplastic disorders of the hematopoietic system. AmJClin Pathol 1969; 53: 204-8. 73. Blair A, Stewart PA, Tolbert PE, et al. Cancer and other causes of death among a cohort of dry cleaners. BrJ Indust Med 1990; 47:162-8. 74. Egedahl RD, Carpenter M, Homik R. An update of an epidemiology study at a hydrometallurgical nickel refinery in Fort Saskatchewan, Alberta. Health Rep 1993; 5:291-302. 75. Edling C, Jarvholm B, Andersson L, et al. Mortality and cancer incidence among workers in an abrasive manufacturing industry. BrJ Indust Med 1987; 44: 57-9. 76. Scherr PA, Hutchison GB, Neiman RS. Non-Hodgkin's lymphoma and occupational exposure. Cancer Res 1992; 52: 5503s-9s. 77. Perrson B, Dahlander A-M, Fredriksson M, et al. Malignant lymphomas and occupational exposures. BrJ Indust Med 1989; 46: 516-20. 78. Garry VF, Danzl TJ, Tarone R, et al. Chromosome rearrangements in fumigant appliers: possible relationship to non-Hodgkin's lymphoma risk. Cancer Epidemiol Biomarkers Prevent 1992; 1: 287-91. 79. Brown LM, Everett GD, Gibson R, et al. Smoking and risk of non-Hodgkin's lymphoma and multiple myeloma. Cancer Causes Control 1992; 3: 49-55. 80. Linet MS, McLaughlin JK, Hsing AW, et al. Is cigarette smoking a risk factor for non-Hodgkin's lymphoma or multiple myeloma? Results from the Lutheran Brotherhood cohort study. Leak Res 1992; 16: 621-4. 81. Freedman DS, Tolbert PE, Coates R, Brann EA, Kjeldsberg CR. Relation of cigarette smoking to non-Hodgkin's lymphoma among middle-aged men. Am J Epidemiol 1998;148:833-41. 82. Herrinton LJ, Friedman GD. Cigarette smoking and risk of non-Hodgkin's lymphoma subtypes. Cancer Epidemiol Biomarkers Prev 1998; 7: 25-8. 83. Nishiyama H, Anderson RE, Ishimaru T, et al. The incidence of malignant lymphoma and multiple myeloma in Hiroshima and Nagasaki atomic bomb survivors. Cancer 1973; 32:1301-9. 84. Boice JD, Morin MM, Glass AG. Diagnostic X-ray procedures and risk of leukemia, lymphoma and multiple myeloma. JAMA 1991; 265:1290-4. 85. Visfeldt J, Andersson M. Pathoanatomical aspects of malignant haematological disorders among Danish patients exposed to thorium dioxide. Acta Pathol Microbiol Immunol Scand 1995; 103: 29-36. 86. Ward MH, Zahm SH, Wisenburger DD, et al. Dietary factors and non-Hodgkin's lymphoma in Nebraska (United States). Cancer Causes Control 1994; 5: 422-32. 87. Scott D. Nutritional factors and the development of nonHodgkin's lymphoma: a review of the evidence. Cancer Res 1992;52:5492s-5s. 88. Tavani A, Negri E, Franceschi S, et al. Coffee consumption and risk of non-Hodgkin's lymphoma. EurJ Cancer Prevent 1994; 3: 351 -6.
References 177
89. Brown LM, Gibson R, Burmeister LF, et at. Alcohol consumption and risk of leukemia, non-Hodgkin's lymphoma and multiple myeloma. Leuk Res 1992; 16: 979-84. 90. De Stefani E, Fierro L, Barrios E, Ronco A. Tobacco, alcohol, diet and risk of non-Hodgkin's lymphoma: a
study of permanent hair dye use and hematopoietic cancer. 7 Natl Cancer Inst 1994; 86:1466-70. 93. Linos A, Blair A, Gibson RW, et al. Leukemia and nonHodgkin's lymphoma and residential proximity to industrial plants. Arch Environ Health 1991; 46: 70-4. 94. Lyons RA, Monaghan SP, Heaven M, et al. Incidence of
case-control study in Uruguay. Leuk Res 1998; 22:
leukaemia and lymphoma in young people in the
445-52.
vicinity of the petrochemical plant at Baglan Bay, South
91. Zahm SH, Weisenburger DD, Babbitt PA, et al. Use of hair coloring products and the risk of lymphoma, multiple myeloma and chronic lymphocytic leukemia. Am J Public Health 1992; 82: 990-7. 92. Grodstein F, Hennekens CH, Colditz GA, et al. A prospective
Wales, 1974 to 1991. Occupat Environ Med 1995; 52: 225-8. 95. Sans S, Elliott P, Kleinschmidt I, et al. Cancer incidence and mortality near the Baglan Bay petrochemical works, South Wales. Occupat Environ Med 1995; 52: 217-24.
This page intentionally left blank
PART
Clinical Management
Hodgkin's disease: clinical features Imaging of lymphoma Localized Hodgkin's disease Localized non-Hodgkin's lymphoma Advanced Hodgkin's disease Aggressive non-Hodgkin's lymphoma Lymphoblastic lymphoma in adults Follicular lymphoma Other low-grade non-Hodgkin's lymphomas High-dose therapy AIDS-related lymphoma Cutaneous lymphomas Pediatric lymphomas Lymphoma in the elderly Infections Long-term problems The way forward
181 205 221 247 269 287 299 309 325 331 351 359 371 385 399 421 437
4
This page intentionally left blank
15 Hodgkin's disease: clinical features PWM JOHNSON, PJ SELBY AND BW HANCOCK
Introduction
181
Prognostic factors
Presenting features of Hodgkin's disease
181
Diagnosis and staging
184
Patterns of recurrence Hodgkin's disease in special situations
Paraneoplastic manifestations of Hodgkin's disease
188
Contrasts with non-Hodgkin's lymphoma
Clinical approach to the patient with Hodgkin's disease
188
References
INTRODUCTION Thomas Hodgkin first described the 'morbid appearances of the absorbant glands and spleen' in six patients with lymph node enlargement at Guy's Hospital in London in January 1832.1 Samuel Wilks in 1865 first used the term 'Hodgkin's disease' in describing a condition involving bulky lymphadenopathy and 'a deposit of a morbid kind in internal viscera, more especially in the spleen'.2 Sternberg3 and Reed4 later identified the characteristic giant cells now given their names. Such cells had also been mentioned by Olivier and Ranvier,5 Tuckwell,6 Langans,7 Greenfield8 and Gowers.9 Treatment with radiotherapy began with Pusey10 in 1902 and was developed by Gilbert in 1939," leading to the work of Peters in Toronto,12 and major contributions at Stanford University by Henry Kaplan.13,14 Single-agent chemotherapy began in the 1940s,15 and this was followed 20 years later by useful combination chemotherapy16 and then effective quadruple combination chemotherapy.17,18 It is perhaps surprising that with such a long history there still seems to be a great deal to learn about the clinical features of Hodgkin's disease. Understanding of these features, their prognostic significance and their biological basis continues to develop. The subject is still in evolution, and thinking around the clinical features is adapted to changing concepts of histopathology and pathogenesis. Treatment decisions continue to be based predominantly upon anatomical staging of the disease and simple indicators of its severity, but there is some dissatisfaction with the lack of predictive power in this approach, particularly its inability to pick out those for
191 194 195 196 198
whom conventional therapy is likely to fail. It is to be hoped that novel methods of prognostic division may emerge in the next few years to overcome this deficiency.
PRESENTING FEATURES OF HODGKIN'S DISEASE The clinical features of Hodgkin's disease have probably changed little despite changing interpretations. The epidemiology of the disease has been discussed in Chapter 13 but certain aspects are relevant to the patterns of presentation. The illness is more common in males, particularly among cases occurring in childhood. The well-known bimodal age incidence varies according to the geographical location, with more childhood and adolescent patients in less developed countries, and a generally higher peak age at presentation in the Western world.19 The clinical pattern of the disease at presentation also varies between units and hospitals and trials groups. Patients with localized disease are seen more often in radiotherapy centers, and advanced systemic disease may lead to a referral to a medical oncology unit. This largely explains the wide variations between reported series. Table 15.1 contains the clinical features of patients who presented to three different research groups in the UK during the last three decades: the Royal Marsden Hospital, the United Kingdom Central Lymphoma Group and the British National Lymphoma Investigation. These show considerable variation in the gender ratio, presence of systemic symptoms, clinical
182 Hodgkin's disease: clinical features
Table 15.1 Clinical features of Hodgkin's disease at presentation
Sex Male Female
77 152
33.6 66.4
148 79
65.2 34.8
182 122
59.9 40.1
172 117
59.5 40.5
117 112
51.1 48.9
79 148
34.8 65.2
84 220
27.6 72.4
87 202
30.1 69.9
27 89 64 49
11.8 38.9 27.9 21.4
13 78 77 54
5.9 35.1 34.7 24.3
10 71 121 93
3.3 23.4 42.6 30.7
8 58 135 88
2.8 20.1 46.7 30.4
10 12 70 137
4.4 5.2 30.6 59.8
11 20 50 135
5.1 9.3 23.2 62.5
5 2 47 250
1.6 0.7 15.5 82.2
2 6 54 227
0.7 2.1 18.7 78.6
193 36
84.3 15.7
198 20
90.8 9.2
273 31
89.8 10.2
250 39
86.5 13.5
139 90
60.7 39.3
172 47
78.5 21.5
246 58
80.9 19.1
220 69
76.1 23.9
117 112
51.1 48.9
81 145
35.8 64.2
103 198
34.2 65.8
95 186
33.8 66.2
171 58
74.7 25.3
193 33
85.4 14.6
281 20
93.4 6.6
252 29
89.7 10.3
197 32
86.0 14.0
230 23
89.8 10.2
220 9
96.1 3.9
173 20
89.6 10.4
199 30
86.9 13.1
191 36
84.1 15.9
B symptoms
No Yes Clinical stage
I II III IV Histology
LD LP MC NS Involvement of: Liver
No Yes Spleen
No Yes Mediastinum
No Yes Bulky mediastinum
No Yes Lung
No Yes Marrow
No Yes Inguinal nodes
No Yes
The authors are grateful to Dr Gillian Vaughan Hudson (BNLI) and Dr Michael Cullen (CLG) for provision of the data included in the table from Hancock et al.20 and Cullen et al.21 BNLI = British National Lymphoma Investigation, CLG = Central Lymphoma Group, LD = lymphocyte depleted, LP = lymphocyte predominant, MC = mixed cellularity, NS = nodular sclerosing.
stage, histologic subtypes, and involvement of various visceral and lymph node sites. This reflects selective referral patterns and differences in investigation policies. The National Survey of Patterns of Care for Hodgkin's Disease of the Commission on Cancer of the American College of Surgeons reported findings in 9482 patients from 611 geographically scattered hospitals (Fig. 15.1).
The authors estimated that this series included some 45 per cent of the US annual incidence during the study period, and it represents a useful unselected description of the clinical features of the disease.22 Painless enlargement of lymph nodes, usually in the neck, but occasionally in the axilla or inguinal region, is the most common presentation. Spontaneous waxing
Presenting features 183 Table 15.3 Stage distribution at presentation24
Image Not Available
IA IB AIM
11 0.8 12
20
HA IIB IIAE IIBE
29 10 5 2.8 47
} 4 I 26
All II IMA IIIB IIIAE IIIBE
All III Figure 15.1 Clinical examination findings at presentation. Reprinted by permission of Wiley-Liss, Inc., a subsidiary of John Wiley & Sons, Inc., from Kennedy BJ, et a I. National survey of patterns of care for Hodgkin's disease. Cancer 7985; 56: 2547-56. © 7985 American Cancer Society.
and waning of nodes is well recognized. Systemic symptoms, such as, fever, night sweats, weight loss or itching, occur in about one-third of patients initially. In a few patients pain is experienced in the nodes, often in the chest, after drinking alcohol - a symptom usually associated with nodular sclerotic histology.23 Table 15.2 gives an overall impression of the pattern of Hodgkin's disease at presentation from across the literature. There are still few examples of detailed studies that add much to this general understanding. The outstanding early work at Stanford24 is summarized in Table 15.3. It is derived from Table 15.2 Sites of involvement with Hodgkin's disease at presentation
Right neck nodes Left neck nodes Mediastinum Axillary nodes Hilar nodes Para-aortic nodes Iliac nodes Inguinal and femoral nodes Mesenteric nodes Splenic hilar, celiac, portal nodes Spleen Liver Lung Bone Bone marrow
55-60 60-70 60-65 20-25 10-25 25-35
10-15 5-15 5-10 10-20 30
5-15 10-20 5-15 5-15
5 10 2 5 <1 <1 <1 <5 <1 <1 <1 <1 <1 <1 <1
IVA IVB
16 11 1.5 2.3 30
I J
1-7
20
All IV
3.2 7.4 11
13
Not given
0
21
PS = pathological stage.
a consecutive series of unselected patients all of whom underwent staging laparotomy and bone marrow biopsy. There is a close association between stage and systemic symptoms that was also shown in the later Stanford series25 (Fig. 15.2). Involvement of particular lymph node groups in the absence of disease elsewhere may carry special prognostic significance. Isolated high cervical lymph nodes are a preferential site of involvement by the lymphocytepredominant histologic type and usually carry a favorable prognosis.26 By contrast, infradiaphragmatic Hodgkin's disease presenting with inguinal lymphadenopathy alone carries an unfavorable prognosis.27
Figure 15.2 symptoms.
Relationship of stage to occurrence of systemic
184 Hodgkin's disease: clinical features
This pattern accounts for about 10 per cent of patients who present with localized disease, and is associated with male gender, more advanced age, a higher incidence of systemic symptoms and with lymphocyte-depleted and mixed cellularity histology.28,29 Nodular sclerosis histologic types tend to present with central lymph node enlargement, usually in the chest, whereas the less favorable histologic types more often involve the abdomen and extranodal sites. Splenic involvement is usually in association with lymphadenopathy on both sides of the diaphragm, and is the sole site of abdominal disease in less than 10 per cent of those with thoracic disease. Isolated extranodal presentations are rare: less than 5 per cent of cases. Pulmonary and endobronchial lesions, liver infiltration, cortical bony deposits, gastrointestinal tumors, bone marrow and meningeal deposits are all described as the sole site of disease in a few cases, but all are very uncommon. When Hodgkin's disease is found in an isolated extranodal site, the diagnosis requires careful and critical review; non-Hodgkin's lymphomas are much more often the cause of such presentations. Extranodal involvement in association with disseminated disease is found in 10-20 per cent of cases. The most commonly affected sites are the lung, liver, cortical bone and bone marrow. Definite diagnosis of extranodal involvement is not always straightforward. Pleural effusions may occur due to either lymphatic obstruction in the mediastinum or direct pleural infiltration. Similarly liver blood tests may be abnormal in the absence of radiologic liver lesions or histologic involvement. Bone marrow infiltration may be inhomogeneous. A variety of paraneoplastic phenomena occur and may occasionally be the cause of presentation. These include: anemia, which may be normochromic/normocytic or due to autoimmune hemolysis; nephrosis, which may also be due to immune complexes or to amyloidosis; skin manifestations, such as, erythema multiforme, icthyosis, hyperpigmentation or urticaria, and hypercalcemia.
DIAGNOSIS AND STAGING In over 90 per cent of all cases the diagnosis is made by lymph node biopsy.22 Needle aspiration cytology may have a minor role in the diagnosis of recurrence30 but is no substitute for adequate surgical biopsy, preferably of a whole node, at the time of presentation. The early work of Kaplan and colleagues31 in California and of Sir David Smithers32'33 at the Royal Marsden Hospital examined the location of nodal and extranodal involvement by Hodgkin's disease, and emphasized the distinctly non-random pattern of spread. Smithers postulated that there might be a continuous movement of Hodgkin cells through lymph nodes and through the bloodstream, but that there may
be intrinsic differences in the susceptibility of different lymph node chains; Kaplan believed that spread was contiguous. The Harvard group have recently analysed their cases to examine patterns of spread34 and certain trends are discernible: supraclavicular nodes are associated with mediastinal disease much more than high neck nodes; mediastinal and hilar nodes are associated; and abdominal nodes are associated with splenic involvement. It is upon the relatively orderly anatomic pattern of disease progression that the various staging systems have been based, influenced also by the diagnostic and therapeutic approaches available at the time. The Hodgkin's Disease Staging Committee meeting at Ann Arbor established the basis of the current staging system35 (Table 15.4). The limitations of the Ann Arbor system were recognized to be the lack of information on several prognostically important criteria, in particular the number of lymph node sites and the bulk of the tumor, particularly in the mediastinum.36 It also ignored recent advances in imaging, including computed tomography (CT) and magnetic resonance imaging (MRI). This led to modification of the Ann Arbor staging at a meeting in the Cotswolds in England, which has subsequently become the basis of the Cotswold staging system37 (Table 15.5). In both staging systems, classifications apply only to the patients at the time of disease presentation and prior to definitive therapy. Each stage is subdivided into A and B categories - A for those without defined systemic symptoms, B for those with one or more of the following: (a) unexplained weight loss of more than 10 per cent of the body weight in the 6 months previous to presentation, (b) unexplained fever with temperatures above Table 15.4
Ann Arbor staging system for Hodgkin's disease
Stage I Involvement of a single lymph region (I) or of a single extralymphatic organ or site (IJ. Stage II Involvement of two or more lymph node regions on the same side of the diaphragm (II) or localized involvement of an extralymphatic organ or site and of one or more lymph node regions on the same side of the diaphragm (llE). An optional recommendation is that the numbers of node regions involved be indicated by a subscript (e.g. II3). Stage III Involvement of lymph node regions on both sides of the diaphragm (III), which may also be accompanied by localized involvement of an extralymphatic organ or site (IIIE), or by involvement of the spleen (Ills) or both (IIIJ. Stage IV Diffuse or disseminated involvement of one or more extralymphatic organs or tissues with or without associated lymph node enlargement. The reason for classifying the patient as stage IV should be identified further by defining site by symbols.
Diagnosis and staging 185
Ta ble 15.5 The Cotswold staging classification for Hodgkin 's disease Stage I Involvement of a single lymph region or a lymphoid structure (e.g. spleen, thymus, Waldeyer's ring). Stage II Involvement of two or more lymph node regions on the same side of the diaphragm (i.e. the mediastinum is a single site, hilar lymph nodes are lateralized). The number of anatomic sites should be indicated by a subscript (e.g. II2). Stage III Involvement of lymph node regions or structures on both sides of the diaphragm: III, with or without splenic hilar, celiac or portal nodes III2 with para-aortic, iliac, mesenteric nodes Stage IV Involvement of extranodal site(s) beyond that designated E: A B X
No symptoms Fever, drenching sweats, weight loss Bulky disease > 1/3 the width of the mediastinum > 10 cm maximal dimension of nodal mass E Involvement of a single extranodal site, contiguous or proximal to a known nodal site CS Clinical stage PS Pathological stage
The following notation is used in further description of pathologic stage: N+ or NH+ or HS+ or SL+ or LM+ or MP+ or P0+ or 0D+ or D-
For other lymph node, positive for disease or negative by biopsy For liver, positive or negative by liver biopsy For spleen, positive or negative following splenectomy For lung, positive or negative by biopsy For marrow, positive or negative by biopsy or smear For pleura involved, or negative by biopsy or cytological examination For osseous involvement or negative by biopsy For skin involvement or negative by biopsy
38°C; and (c) drenching night sweats. Neither pruritus alone nor a short, febrile illness associated with a known infection qualifies for B classification. A distinction is made in both staging systems between clinical and pathologic staging. Clinical staging (CS) is determined by history, physical examination, radiological studies, isotopic scans, laboratory tests of urine and blood, and the initial biopsy results. Whilst the clinical stage derives from the clinical findings and non-invasive tests, the pathologic stage (PS) is based upon further biopsies taken subsequent to the initial diagnosis, either at laparotomy or by some other invasive procedure, such as liver or bone marrow biopsy. Some care must be exercised in the use of the term
'pathological staging'. Comprehensive pathological staging involves laparotomy and splenectomy, but the term is allowed for less extensive biopsy procedures, such as percutaneous liver biopsy. The use of the PS notation does not by itself indicate that a splenectomy has been carried out. Lymphatic Hodgkin's disease The lymphatic structures are defined as the lymph nodes, spleen, thymus, Waldeyer's ring, appendix and Peyer's patches. The lymph node regions accepted as single groups are shown in Table 15.6. Figure 15.3 shows the characteristic appearances of Hodgkin's disease in the mediastinum. Radiologic enlargement alone is insufficient to diagnose splenic involvement; clinically detectable splenomegaly or focal defects on scans are required (Fig. 15.4). In patients subjected to staging laparotomies, only 60 per cent of spleens that were enlarged clinically or radiographically were found to be involved on histological examination.24,38,39 This figure rose to 70-80 per cent in Table 15.6 Lymph node groups classified separately for staging Cervical, supraclavicular, occipital and preauricular (all one) Infraclavicular Axillary and pectoral Mediastinal, including thymus Hilar Para-aortic Mesenteric Iliac Inguinal and femoral Splenic Popliteal Epitrochlear Waldeyer's ring
Figure 15.3
Computed tomographic scan showing mediastinal
involvement by Hodgkin's disease.
186 Hodgkin's disease: clinical features
Figure 15.4
Figure 15.5 Computed tomographic scan showing pulmonary involvement by Hodgkin's disease.
Computed tomographic scan showing
involvement of the liver and spleen by Hodgkin's disease. 25
the presence of involved infradiaphragmatic nodes. Conversely, 30 per cent of spleens that were thought to be normal were found to be involved. Clinical assessment is therefore misleading. The main factors that indicate an increased probability of splenic involvement are mixed cellularity histologic subtype, para-aortic lymph node involvement, systemic symptoms and a large number of nodal sites above the diaphragm. Epitrochlear lymphadenopathy is uncommon, even when Hodgkin's disease is widespread, and is probably found in less than 2 per cent of cases.40,41 There are a few cases of epitrochlear disease as the presenting feature of Hodgkin's disease in the absence of any other disease on full investigation.42 Involvement of Waldeyer's ring is very uncommon and was found in 3.7 per cent of 750 patients in one series.43 It is usually found in the tonsil or nasopharynx, and is associated with upper cervical or pre-auricular nodal disease in around half the cases.44 Mixed-cellularity histology is over-represented at this site.
Extranodal Hodgkin's disease Table 15.2 gives an indication of frequency of involvement at extranodal sites. LUNG
About 10-20 per cent of patients have lung disease at presentation45 (Fig. 15.5). Older series recorded that up to 40 per cent involved the lung at some phase of the illness,46,47 but this has become less common with increasingly successful treatment. Over 50 per cent may have involvement at autopsy.48 Hodgkin's disease in the lung is usually associated with lymphadenopathy in the mediastinum or hilar regions, and is most common with nodular sclerosing histology. Relapse in lung may occur in the absence of regrowth of the lymph nodes in patients who have been treated for intrathoracic lymphadenopathy by radiotherapy. Primary intrapulmonary Hodgkin's disease without
lymphadenopathy has been described in occasional case reports (see reviews49,50). The commonest pattern of lung involvement is direct spread of tumor from the mediastinum into the lung parenchyma. The X-rays may, on the other hand, show large lung nodules, multiple small nodules, or coarse and fine reticular shadows. PLEURA
Pleural effusion in Hodgkin's disease may be the result of tumor invasion of the pleura or may be secondary to obstruction of lymphatic drainage by mediastinal disease. The presence of an effusion is not necessarily indicative of extranodal Hodgkin's disease. Distinguishing between these two processes is of some clinical importance and may be difficult. Malignant cells may be seen on cytological examination of the pleural fluid or on needle biopsy of the pleura. If the pleural surfaces are involved with tumor the protein content of the fluid will usually be high, but this is not invariably the case. When radiotherapy was the only curative treatment available, patients were subjected to thoracotomy in order to distinguish the cause of pleural effusion. In more modern practice, since even non-malignant effusions are usually associated with bulky mediastinal lymph nodes, which are best treated by chemotherapy, such invasive methods are no longer required.51 ENDOBRONCHIAL DISEASE
Although involvement of the distal bronchiole is not uncommonly found at autopsy in patients with advanced pulmonary Hodgkin's disease, it is unusual to find Hodgkin's disease as an endobronchial lesion. Harper et al. have reviewed 13 such cases, most of which had lobulated masses on bronchoscopy.52 SUPERIOR VENA CAVAL OBSTRUCTION
Although Hodgkin's disease commonly involves the mediastinum, superior vena caval obstruction is rare, and less frequent, for instance, than in small cell lung
Diagnosis and staging 187
cancer or aggressive non-Hodgkins lymphoma.53 Its clinical and radiological features are not specific to Hodgkin's disease and it is usually associated with nodal disease in the neck, biopsy of which will yield the diagnosis. Dexamethasone allows substantial relief of symptoms and some regression of the mass. Since the disease is usually bulky, combination chemotherapy is the treatment of choice, and works as quickly as radiotherapy. LIVER
The reported incidence of liver involvement at presentation ranges from 3 to 24 per cent, reflecting the selection of patients and the extent of investigation.51,54 The mixedcellularity and lymphocyte-depleted histologic types are the most common in this site. The diagnosis of hepatic Hodgkin's disease may present several problems. Jaundice is usually only found in very advanced cases, and there is little correlation between liver size and the presence of Hodgkin's disease histologically. Liver function tests may be deranged in Hodgkin's disease in the absence of histological involvement of the liver.55,57 Imaging scans of the liver lack sensitivity, whether by CT, isotope imaging scans, ultrasound or any of the newer methods now available. In a study of 215 patients, Fabian et al.58 found 24 per cent positive liver biopsies (laparotomy, peritonoscopy or percutaneous) in patients with stage IIIA-IVB disease with an increasing rate in higher stages (15 per cent in CSIIIA; 48 per cent in CSIVB). Twenty-nine patients had a negative percutaneous biopsy followed by a laparotomy at which 8/29 had liver disease on wedge biopsy. Nine patients had negative peritonoscopy and only 1/9 had a positive wedge biopsy. In the absence of a satisfactory diagnostic method for liver Hodgkin's disease, the presence of two substantially abnormal liver function tests, or one abnormal test plus hepatomegaly, or an abnormal liver scan plus one abnormal enzyme were previously used as criteria (Ann Arbor criteria35). Unfortunately these criteria were not accurate: in the study of Fabian et al.,58 56 per cent of patients with a positive biopsy were negative by Ann Arbor criteria and 21 per cent of patients with normal liver biopsies had Ann Arbor criteria liver disease. The Cotswold staging system therefore dispensed with the liver blood tests as a means of determining stage, relying instead upon radiology and specialized scans. The finding of focal defects in the liver on CT scan is taken as evidence of involvement provided it is confirmed by a second imaging technique, such as ultrasound or isotope scanning. Hepatitis, hepatic failure, cholestatic jaundice and granulomatous hepatitis have all been described.59-63Liver granulomas are found in 5-10 per cent of patients at staging laparotomy for known Hodgkin's disease.60 However, in a small number of cases, hepatic granulomas may precede the diagnosis of Hodgkin's disease by several years, and be associated with hepatosplenomegaly and fever.61 Hodgkin's disease should be considered in the differential
diagnosis of'idiopathic' hepatic granulomas. Cholestatic jaundice may occur in the absence of detectable liver involvement59,63 either with advanced disease or localized supradiaphragmatic lymph node disease. In that case, the jaundice may resolve after mantle radiotherapy.63 BONE
As with other visceral sites of disease, there is a very wide range of reported findings from different centers (6-19 per cent in series after 1960), but if these are taken together the average is 12 per cent in 2991 patients.51 Bone disease may or may not be painful, but can be associated with alcohol-induced pain. It may be lytic or sclerotic, and may result from hematogenous spread or direct extension into bone, for example, the lumbar spinal vertebrae from para-aortic nodes or the sternum from mediastinal nodes. The spine is the commonest site of involvement.23,64 SPINAL CORD COMPRESSION
Extradural spinal cord compression is an unusual manifestation of advanced Hodgkin's disease65,66 or a rare presentation of localized bony disease. There may be associated bony destruction or it may occur in the absence of obvious bony abnormality, by extension from nodal disease through the intervertebral foramina or from meningeal deposits. The clinical features of cord compression in Hodgkin's disease are similar to those from other causes. Unlike other malignant processes that cause cord compression, Hodgkin's disease is usually very sensitive to treatment by radiotherapy and chemotherapy. The results are best when treatment is given immediately and patients who have good neurological function at the time of treatment usually make excellent recoveries. Spinal laminectomy may be required to make the diagnosis in a few patients.51 CENTRAL NERVOUS SYSTEM
Intracranial Hodgkin's disease is uncommon.67,68 The disease tends to involve midline structures at the base of the brain, including brain stem and cerebellar vermis, and causes cranial nerve palsies, weakness, headache, and seizures. It carries a grave prognosis, particularly when associated with resistant disease elsewhere, and median survival is less than 6 months, with only about 20 per cent of patients surviving 2 years. Leptomeningeal Hodgkin's disease has been the subject of sporadic case reports,69,70 and Hodgkin's disease is one of the least frequent causes of meningopathy. When encountered, it maybe treated with intrathecal methotrexate and whole-brain irradiation, and occasionally useful remissions may result.70 BONE MARROW
Involvement of bone marrow is found in between 5 and 15 per cent of patients. It is closely associated with symptoms
188 Hodgkin's disease: clinical features
and advanced stage, being found in less than 2 per cent of IA and IIA patients, and it is less common in nodular sclerotic subtypes.71 Alkaline phosphatase is commonly raised and marrow involvement confers a relatively poor prognosis, even within patients with advanced disease.71,72 HODGKIN-CELL LEUKEMIA
Hodgkin-cell leukemia is rare and is usually associated with very advanced disease.73 In such cases, the distinction from lymphoid and myeloid leukemia may be difficult. Linch et al. have analysed one such case, found the cells to be derived from the B cell lineage and have postulated an association with the lymphocyte-depleted form of nodular sclerotic Hodgkin's disease.74 PERICARDIAL DISEASE
Paracardiac nodal disease is visible in 2 per cent of patients at presentation75 but may be more important at relapse after mantle irradiation, which may not include nodes in the cardiophrenic angle.76 Pericardial effusions are rare at the time of presentation but may be a feature of very advanced mediastinal disease. SKIN AND SUBCUTANEOUS DISEASE
The cutaneous manifestations of Hodgkin's disease are pruritus, urticaria, erythema multiforme, eczema, herpetic infections, ichthyosis, alopecia, oedema, erythema nodosum, toxic epidermal necrolysis and infiltration. The findings at involved sites are usually multiple dermal or subcutaneous erythematous nodules and extension from involved lymph nodes may be a feature. GASTROINTESTINAL TRACT
Gastrointestinal involvement with Hodgkin's disease is distinctly rare, although sporadic case reports have been made for most parts of the gastrointestinal tract.25,51 There are no specific clinical syndromes, and the diagnosis in any of these sites should be regarded with suspicion, since non-Hodgkin's lymphomas are a much commoner cause of gut involvement. OTHER SITES
Rare cases of Hodgkin's disease are described in extranodal sites, including thyroid, presenting as a 'cold' nodule,77 larynx,78 kidney,79 bladder, adrenal, and ovary and endometrium.25'80 Thyroid involvement usually results from direct spread from contiguous nodes.
PARANEOPLASTIC MANIFESTATIONS OF HODGKIN'S DISEASE Skin manifestations, such as erythema multiforme or nodosum, are recognized paraneoplastic phenomena in
Hodgkin's disease, which may give rise to several different cutaneous effects as listed above. Nephrotic syndrome has been described in cases without perirenal or intraabdominal Hodgkin's disease.81,82 This manifestation has been associated with immune complexes in the glomeruli and improves on treatment of the underlying Hodgkin's disease. Hypercalcemia is described in Hodgkin's disease but is uncommon. Intrathoracic Hodgkin's disease may be associated with hypertrophic osteoarthropathy.83 Neurologic syndromes, including progressive multifocal leukoencephalopathy, subacute motor neuropathy, cerebellar degeneration, myelopathy, anterior horn cell degeneration and diffuse reticular cerebral infiltration, have all been described occasionally in association with Hodgkin's disease. The occurrence of any neurological syndrome in association with Hodgkin's disease is an immediate indication for detailed neurological examination and investigation. CT and MPJ scanning and careful examination of the cerebrospinal fluid are all required. In addition to specific involvement by Hodgkin's cells and to the various paraneoplastic syndromes, a careful search should be made for evidence of opportunist infection.
CLINICAL APPROACH TO THE PATIENT WITH HODGKIN'S DISEASE History As with all lymphoma patients, the initial approach to those with Hodgkin's disease involves a careful history, seeking evidence of systemic symptoms. The performance status of the patient may constitute useful prognostic information and should be recorded. There is a spectrum of severity in systemic symptoms. Sweats may be drenching and obvious with patients frequently changing their night attire or sheets, or they may be less prominent. Fever may be a high swinging fever, sometimes in the Pel-Epstein pattern, which is intermittent occurring every 3 or 4 days. Weight loss may be profound or minor. To be characterized as a significant symptom, the weight loss must be unexplained and of more than 10 per cent of the body weight in 6 months. Fever has to be more than 38°C; night sweats are not defined in severity but unexplained night sweats sufficient to cause a change of night attire is a useful criterion to use. Careful evaluation of a patient's emotional state is necessary, as well as a social and family history, which will influence the impact of treatment. It is important to know whether a patient has completed their family in order to judge the need for gamete cryopreservation. Physical examination Examination will involve all lymph node areas, including Waldeyer's ring, attention to hepatosplenomegaly and a
Clinical approach to the patient 189
general examination of cardiovascular, respiratory and neurologic systems.
Investigations
Table 15.7 Investigations for newly diagnosed patients with Hodgkin's disease Blood tests Blood count: Hemoglobin White blood cell count Lymphocyte count Platelet count Erythrocyte sedimentation rate Biochemistry Electrolytes Urea Creatinine Urate Calcium Liver blood tests (bilirubin, alkaline phosphatase, transaminases) Lactate dehydrogenase Albumin [32 microglobulin
Laboratory investigations in all cases involve a full blood count and differential. The absolute lymphocyte count may have prognostic significance and should be recorded. The erythrocyte sedimentation rate (ESR) should be measured. Biochemical tests will include assessments of hepatic, renal and bone metabolism. Liver tests to be measured include particularly alkaline phosphatase, transaminases, lactate dehydrogenase, bilirubin and albumin. The renal function tests should include serum urate concentration and the serum calcium should be recorded.b2 microglobulin is increasingly reported as a prognostic factor and should also be measured prospectively.84 All patients require chest X-rays, and CT scans of the chest, abdomen and pelvis with oral and intravenous contrast. Bipedal lymphangiography, which was widely used for staging in the past, is now largely supplanted by the less invasive procedure of CT scanning. The latter investigation has the additional advantage of visualizing the celiac axis, mesenteric, portal and splenic hilar nodes, which are poorly demonstrated if at all by lymphangiography. Involvement of lymph nodes that are of normal size cannot be detected by CT scanning, ultrasonography or MRI at present. Investigations that may contribute more information include gallium scanning85-88 and positron emission tomography.89 The place of each of these investigations is reviewed by Sandrasegaran et al. in Chapter 16. Bone marrow biopsy has been widely used in the staging of Hodgkin's disease, mainly because of its recognition as an unfavorable prognostic marker.90 More recently it has been shown that bone marrow findings rarely alter the clinical management or survival, so that the procedure may no longer be recommended as part of the routine staging.91 If there is a case to be made for bone marrow biopsy, it might be for those patients with stage I/IIA, where a positive biopsy would mandate systemic rather than local therapy;92 however, since the positive yield in this group is less than 1 per cent this is difficult to justify clinically. The recommended staging investigations for newly diagnosed patients are given in Table 15.7.
marrow infiltration.95 Coomb's positive hemolytic anemia is rare and associated with advanced symptomatic disease.96,97 Leucocytosis with a total white cell count exceeding 10 x 109/1 is found in about a quarter of patients, whereas leucopenia with white cell counts of less than 5 x 109/1 occurs in less than 10 per cent. Lymphopenia with counts less than 1.0 x 109/1 is seen in about 20 per cent of cases and is associated with advanced stage.25 Eosinophilia is occasionally present but is probably of no special significance. In a few patients, thrombocytopenia has been the first manifestation of the disease.98,99 Most commonly, this appears to be due to platelet consumption by immune mechanisms, although the alternative explanation of heavy bone marrow infiltration has to be considered. Autoimmune neutropenia has been described.100
HEMATOLOGY
BIOCHEMISTRY
Hematological tests may be normal. An elevation of the ESR is commonly found in patients with advanced disease or systemic systems, and confers a relatively poor prognosis.93,94 Severe anemia (<9 g/dl) at diagnosis is uncommon and may be multifactorial, owing to hemodilution, hemolysis, iron utilization block and occasionally
Abnormalities in liver function tests may be seen in patients with Hodgkin's disease in the absence of liver infiltration56,57 (see later) or because of liver involvement. The lactate dehydrogenase level has emerged as an important prognostic factor and is often elevated in advanced disease.101 Hypercalcemia is rare and associated
Radiology Chest X-ray Computed tomography (chest, abdomen, pelvis) Bone marrow biopsy Optional Under special circumstances Ultrasound scanning Magnetic resonance imaging Gallium scanning Technetium bone scanning Computed tomography scanning of the brain
190 Hodgkin's disease: clinical features
with extensive involvement of bones. In occasional patients, abnormalities of vitamin D metabolism have been demonstrated.102 Renal function must be checked, because of the risk of renal damage by hyperuricemia or infection. Urinalysis occasionally reveals proteinuria. ALTERNATIVE MARKERS OF THE SEVERITY OF HODGKIN'S DISEASE
Although the staging procedures described above yield useful prognostic information for the whole population of patients with Hodgkin's disease, the relatively crude techniques of anatomical localization ignore the heterogeneity of behavior of the illness in different patients. The different histopathologic subtypes may give some help in this respect, but they remain broad categories within which differing patterns of disease are observed. Because of this, different assays have been tested in the hope of identifying measurements that might better reflect the course of the illness. b2 microglobulin has already been mentioned as one such measurement, but a variety of others are under active investigation. Among these, cytokines and their receptors, and adhesion molecules appear most promising. There is now emerging evidence from the literature that measurement of several cytokines may correlate with the severity of Hodgkin's disease. This is in spite of the fact that many studies are based on small numbers of patients with only a few cytokines examined in each and it is often difficult to compare results between studies owing to differences in the assays used. Conflicting results have come from studies relating interleukin-6 (IL-6) concentrations to survival and the presence of B symptoms. IL-6 has been found to be elevated in between 50 and 75 per cent of patients at presentation compared with controls.103-106 In some studies, IL-6 concentration did not relate to histological subtype, stage or the presence of B symptoms, and was no longer detectable in the majority of patients following therapy, even in the presence of progressive disease.103-105 However, a strong correlation between serum IL-6 concentrations and the presence of B symptoms has been found in later studies.106-107 In addition, higher concentrations of IL-6 were related to poorer survival.106-107 The reasons for these discrepancies are unclear but may be related to assay differences in some cases, different detection limits, and the mixture of untreated and treated patients in some studies. More recently, IL-10 has been identified as another potential prognostic marker. A study of 85 patients at the MD Anderson Cancer Center showed IL-10 and serum lactate dehydrogenase levels to be independent predictors of failure-free survival in a multivariate analysis: among 54 patients with normal levels of both markers, failure-free survival (FFS) was 78 per cent whilst among 23 patients with elevated levels of both, FFS was only 22 per cent.108
Elevated soluble IL-2 receptor (sCD 25) concentrations are seen in the great majority of patients - 93 per cent in one study.109 Levels have been related to disease stage, B symptoms and disease status106,110,111 with lower levels relating to better survival.111,112 A multivariate analysis of sCD 25 levels among 127 previously untreated patients demonstrated an independent correlation with survival.113 A further study used a panel of soluble antigens as biomarkers: the combination of elevated sCD 8, sCD 25, sCD 30 and sCD 54 at diagnosis in 18 of 80 patients was found to predict a group with only 33 per cent disease-free survival at 5 years.109 The same study found that 37 per cent of patients acheiving clinical complete remission retained at least one elevated marker, and that this also correlated with the probability of recurrence. Between 22 and 87 per cent of newly diagnosed patients have been reported as having elevated levels of sCD 30111,114,115 with the wide range ascribed to differing assay sensitivity. Correlations with the presence of B symptoms, stage and tumor burden have all been reported. In addition, sCD 30 was of prognostic use with higher levels being associated with poorer survival111,114,116 and correlated with disease activity and treatment response suggesting a possible use in longitudinal monitoring. Significantly elevated concentrations of soluble p55 TNF receptors have been found in Hodgkin's disease compared with controls117,118 with correlations to disease stage, B symptoms and complete response rate. Significantly higher sICAM-1 concentrations have been found in patients with Hodgkin's disease compared to controls.116,119,121 Levels were related to stage of disease, B symptoms and response rate. sICAM-1 appears to offer prognostic information for disease-free and overall survival, but these findings require amplification in larger series. STAGING LAPAROTOMY
Prior to the development of accurate cross-sectional imaging and the analysis of clinical data to predict intraabdominal involvement, complete staging of Hodgkin's disease apparently restricted to the upper body required a staging laparotomy with sampling of multiple node groups, liver biopsy and splenectomy. A series of studies in the late 1970s demonstrated the correlation between clinical features, such as number of sites of involvement, size of the largest tumor mass, pathologic subtype and ESR, and the finding of abdominal disease at laparotomy.122,124 These findings allowed application of chemotherapy for patients with adverse prognostic features. This, together with the success of 'salvage' therapy for patients with progressive abdominal disease following extended field radiotherapy to the chest, and concerns about the higher numbers of patients developing secondary leukemia following splenectomy,125 all
Prognostic factors 191
contributed to the decline of staging laparotomy. A randomized trial demonstrated the equivalent survival of patients staged by laparotomy or clinically when subtotal nodal irradiation was used,126 and such results appear to have been sustained in the long term.127 The staging laparotomy is therefore no longer used in the great majority of centers worldwide.
PROGNOSTIC FACTORS Many attempts have been made to predict the outcome of treatment for patients with Hodgkin's disease. A wide range of descriptive studies have identified features of the illness that predict for remission, freedom from relapse and overall survival, along the lines of the factors described above. These fall broadly into three groups, with obvious areas of overlap: 1 Factors relating to the extent of the disease, such as stage, number of sites of involvement, bulk of tumor mass, lactate dehydrogenase (LDH), b2 microglobulin and cytokine levels. 2 Factors reflecting the state of the patient and the effect of the illness, such as performance status, age, gender, albumin, hemoglobin and white blood cell count. 3 Factors reflecting the aggression of the malignancy such as histologic type, systemic symptoms and ESR.
and most studies have identified age as prognostic in both univariate and multivariate analyses (see also Chapter 28). GENDER
Women with Hodgkin's disease are more likely than men to survive. Attention has been drawn to this over many years25 and a survival advantage overall of about 10 per cent after 5 years for all patients is estimated. This remains true in the International Database study, the effect being most prominent in patients with advanced stage disease. The causes of this difference are not known. STAGE Stage is an important predictor of survival in population-based studies22 (Fig. 15.7). The more recent and extensive collection of data into the international Hodgkin's disease database also confirms that stage is an important predictor of freedom from failure, which was the principal end point of that analysis.130 The presence of B symptoms also has a highly significant impact on prognosis.25,130,131 The impact of B symptoms is probably less in modern series of patients treated with state-of-the-art chemotherapy than in early series. Although it remains the custom to consider systemic symptoms to be of prognostic significance according to the criteria set out in the Ann Arbor staging system
This latter category has appeared to lose some significance as the efficacy of therapy has increased. The different factors have been used in analyses of increasing complexity, usually multivariate models of proportional hazards128 or multiple regression. The most pressing application for such models is to identify in advance the minority of patients poorly served by conventional therapy who might be most likely to benefit from alternative approaches such as early high-dose chemotherapy. The largest such analysis to date was that of the International Database on Hodgkin's Disease, which included over 5000 patients with advanced disease from 22 study groups in Europe and North America.129 Despite this, it was not possible to identify a large subgroup with a 5-year progression-free survival of less than 50 per cent: the failure of therapy was distributed across prognostic categories in too diffuse a manner. This work continues but so far it has proved a difficult challenge and no good predictive model exists. Individual prognostic factors
AGE Age is a major determinant of survival even when deaths due to causes other than Hodgkin's disease are excluded from the analysis. The population-based study of cases in the USA22 very clearly demonstrates this (Fig. 15.6),
Figure 15.6 Age as a prognostic factor in cause-specific survival. Reprinted by permission of Wiley-Liss, Inc., a subsidiary of John Wiley & Sons, Inc., from Kennedy BJ, et a I. National survey of patterns of care for Hodgkin's disease. Cancer 7985; 56. 2547-56. © 1985 American Cancer Society.
192 Hodgkin's disease: clinical features
paragranuloma patients (10 per cent only). Patients with lymphocyte-rich classical Hodgkin's disease did not differ significantly in clinical features from those with nodular sclerotic histology. Patients with nodular paragranuloma showed a particular pattern of outcome. Overall survival was favorable at 95 per cent at 10 years. However, there was a continuing pattern of recurrence with only 50 per cent of patients free from disease after 15 years. Late recurrence was common up to 20 years (Fig. 15.8). Some patients had multiple recurrences extending over 15 years but still a good overall outcome in terms of survival. There were 103 deaths among 478 patients in the study as a whole. Among 83 for whom the cause of death was known, 41 were due to Hodgkin's disease, four were treatment related, 19 were due to secondary neoplasia including ten non-Hodgkin's lymphomas, six solid tumors and three acute leukemias. Deaths from Hodgkin's disease occurred relatively early with 50 per cent between 2 and 4 years from diagnosis. Deaths from non-Hodgkin's lymphomas and acute leukemia occurred between 2 and 8 years, and from solid tumors between 4 and 14 years.54 NUMBER OF NODAL SITES
HISTOLOGICTYPE
In patients presenting with supradiaphragmatic Hodgkin's disease, an increasing number of involved nodal sites is known to indicate an increased risk of infradiaphragmatic disease. Accordingly there is an increased probability of abdominal relapse if such supradiaphragmatic disease is treated with radiation alone. In the Harvard Joint Centre for Radiotherapy series124 the number of nodal sites predicted powerfully for pathological upstaging after laparotomy. Seventeen per cent of patients with one lymph node site had their stage increased after laparotomy whereas 27 per cent of patients with two or more lymph node sites were similarly upstaged (P < 0.01). The presence of inguinal lymphadenopathy predicts for worse progression-free survival on both univariate and multivariate analysis.130
The association between histology and prognosis in Hodgkin's disease is described in detail in Chapter 2. In the context of clinical features, it is possible to identify a distinctive pattern of disease, particularly for the lymphocyte-predominant type. A recent international collaborative study examined the characteristics of 201 cases in which the pathology was reviewed. Of these, 54.7 per cent were reclassified into nodular paragranuloma (nodular lymphocyte-predominant Hodgkin's disease) and 28.4 per cent into lymphocyte-rich classical Hodgkin's disease. The patients with nodular paragranuloma differed from patients with classical nodular sclerosing or mixed cellularity Hodgkin's disease in a number of important respects. Nodular paragranuloma patients were younger (average 30 years vs 37 years), more often male (74 vs 49 per cent), and more often had early-stage disease (74 vs 57 per cent). Mediastinal disease was less frequent with nodular
Figure 15.8 Survival (SV) and freedom from treatment failure (FFTF) in lymphocyte-predominant Hodgkin's disease.
Figure 15.7 Stage as a prognostic factor. Reprinted by permission of Wiley-Liss, Inc., a subsidiary of John Wiley & Sons, Inc., from Kennedy BJ, et al. National survey of patterns of care for Hodgkin's disease. Cancer 7985; 56:2547-56. © 7985 American Cancer Society.
(weight loss, night sweats and fever), more detailed analysis132 suggests that weight loss, significant fevers and severe pruritus are associated with a poorer prognosis on univariate and multivariate analysis, whereas night sweats, minor fever, minor degrees of weight loss and mild pruritus had no influence.
Prognostic factors 193
MEDIASTINAL DISEASE
Mediastinal involvement is associated with a relatively young population of patients and with nodular sclerotic histology. Small-volume mediastinal disease probably has little influence on survival or on probability of relapse after radiotherapy. However, when a mediastinal mass is bulky (mass diameter to thoracic width ratio of > 33 per cent), most authors find a higher risk of relapse after radiotherapy than for patients without such bulk.133-137 There are, however, some analyses that do not show this effect, probably as a result of differences in radiotherapy technique.138-139 The high relapse rate, which is seen in many centers after radiotherapy alone, is reduced by the addition of combination chemotherapy.140 However, presence of a very bulky (> 45 per cent thoracic diameter) mediastinal mass remains a significant predictor of treatment failure even after modern chemotherapy.130 FAMILIAL LONGEVITY
Bjorkholm et al 141 showed that parental longevity influenced the outcome of 98 patients with Hodgkin's disease. Although patients aged less than 50 at the time of diagnosis were not affected, patients more than 50 years old had a better outlook if their parents had been long lived. There was a significant excess of death from tuberculosis in the parents of patients who did badly with Hodgkin's disease. The authors postulated that this observation might be explained by a familial deficiency in T cell immunity. PERSONALITY
A single study has been reported, which investigated the relation between personality and outcome in malignant lymphoma. Sixty-three newly diagnosed patients were graded by the Personality Inventory questionnaire at the time of diagnosis and the L-score was found to be an independent predictor for survival in multivariate analysis, as was the Hospital Anxiety and Depression score.142
Multivariate analysis and models Many different prognostic models and indices have been proposed for Hodgkin's disease. Some of these are
described in Table 15.8. Four of these models were evaluated in a study of 344 patients with advanced Hogdkin's disease included in the Groupe d'Etudes des Lymphomes de 1'Adulte (GELA) study.145 Patients with at least three adverse factors according to the Memorial Sloan-Kettering Cancer Centre or the European Bone Marrow Transplant criteria had a higher risk of failing with conventional treatment; however, based on survival rate, no very high risk group could be identified. For such models, three features can be used to describe their utility. These are the statistical significance of the factors included, the fit of the data to the chosen model and the predictive power of the model. Most reports of prognostic models describe their statistical significance and goodness of fit. Until recently, no good method was available to evaluate predictive power in survival time models. The predictive power of a set of prognostic factors is their ability on average to make accurate predictions of what will happen to patients. If a model is fitted well by the data, it does not necessarily imply that, for an individual patient, the prediction will be precise. Without quantification of the predictive power, highly significant P values of covariates in an adequate fitting model may give a misleading impression of accuracy and utility.146 As an example of this problem, data have been analysed from over 1000 patients treated in the UK in the Royal Marsden Hospital series, the United Kingdom Central Lymphoma Group and the British National Lymphoma Investigation.20,21,147 The data were used to evaluate the meaning of prognostic factors for complete remission, recurrence and survival. The data on log rank results for the entire database are shown in (Table 15.9). After Cox proportional hazards analysis independent prognostic factors were age, clinical stage IV, mediastinal bulk and albumin. In this analysis, choice of chemotherapy was also significant, with poorer results for patients receiving only alkylating agent-based combinations compared to those who received alternating regimens including doxorubicin and etoposide. In this analysis, the ability of the prognostic factors to predict patient outcome was examined. The proportion of variance that was explained by the best prognostic factor analysis was surprisingly small for the prediction of complete remission (7 per cent of variance explained), survival to 5 years (9 per cent of variance explained) and estimated survival probability (10-14
Table 15.8 Prognostic models in Hodgkin's disease factors found predictive of outcome in multivariate analysis 132
Gobbi (1988) Wagstaff (1988)143 Proctor (1991 )144 Hasenclever(1996)129 Sarris (1996)108
586 301 92 validated in 455
1618 155 validated in 226
Age, stage, gender, histology, erythrocyte sedimentation rate, albumin Age, stage, gender, lymphocyte count Age, stage, hemoglobin, lymphocyte count Age, stage, gender, white blood cell count, hemoglobin, albumin Age, lactatedehydrogenase, (32 microglobulin
194 Hodgkin's disease: clinical features Table 15.9 Analysis of prognostic factors in a combined UK series of patients
Sex
Male
654
73.9
Female
395
73.3
0.1
0.814
367 682
79.2 70.7
4.8
0.28
I, II, III
759
77.8
IV
284
62.4
27.7
< 0.001
Involvement of: Liver No 914 Yes 126
75.3 60.6
15.9
< 0.001
Spleen No Yes
74.8 69.7
3.7
0.53
B symptoms
No Yes Clinical stage
Mediastinum No Yes
777 264
PATTERNS OF RECURRENCE 396 641
73.7 73.4
0.8
0.362
Bulky mediastinum No 897 Yes 140
74.4 67.9
1.5
0.222
Lung No Yes
77.3 64.3
2.8
0.094
Marrow No
400 55
393
76.2
29
61.2
3.1
0.078
Inguinal nodes No Yes
390 66
78.5 59.8
10.7
0.001
Age < 45 >45
790 259
78.4 59.0
59.2
< 0.001
463 319
82.0 62.3
33.5
< 0.001
585 229
78.1 62.9
15.?
< 0.001
Yes
Albumin
> 36 <36 Hemoglobin
>11 <11
possible in 64 per cent of cases; with the addition of the best statistical model derived from the patient database, the proportion of patients with the correct prediction rises to only 67 per cent. For prediction of survival at 5 years, in the absence of a model, prediction would be correct in 74 per cent of cases and, with the best model (excluding information on remission status), this rises to 76 per cent. Even when the retrospective information that complete remission is attained is included in the model, the proportion of correct predictions rises only to 81 per cent. It is apparent from this analysis that the prognostic factors currently available are very limited in their discriminatory power. The implication of this is that there is a need for new prognostic factors that might more accurately reflect the behavior of the illness. This implies a better understanding of the biological factors predicting for outcome and, when possible, for evaluation of chemosensitivity.
per cent variance explained). Further analyses have considered the ability of prognostic factors to predict the outcome of individual patients. (If the probability of remission or survival was estimated by the model to be > 50 per cent, this was scored as a prediction.) These analyses are also disappointing. Without using the statistical model, prediction of complete remission is
Following radiotherapy for localized disease, the factors predicting risk of recurrence are well documented. When patients are carefully selected, surgically staged and treated with extended-field radiotherapy, the recurrence rate is less than 20 per cent. However, unselected clinical stage I and II patients may have relapse rates of over 50 per cent.148 Most such recurrences occur in lymph nodes lying outside the radiotherapy field. Recurrence after chemotherapy occurs in 20-40 per cent of cases and in the great majority this is at sites of previous disease. In the remainder it is generally in closely related sites.149 Most recurrences occur within 3 years of the end of treatment. However, late recurrences are well recognized up to 20 years.150 In one large series of 1360 patients from Stanford University, the actuarial risk of recurrence after 3 years was 12.9 per cent, and 52 patients suffered recurrences between 3 and 15 years.151 Most recurrences of Hodgkin's disease are diagnosed clinically or on chest X-ray. In the Stanford series it is notable that only 13 patients of 52 called attention to their relapse (three by palpating their own nodes; ten by reporting systemic symptoms). The remainder were diagnosed by physicians at follow-up. In the Christie Hospital, Manchester, UK, among 135 relapses, 113 patients reported lumps or systemic symptoms themselves. However, in 22 cases the recurrence was detected by physical examination or tests (most commonly chest X-ray).152 Most recurrences will be detected if patients are asked to return if they have symptoms. A few more will be found by examination and investigations. Frequent, detailed follow-up is unlikely to have a large impact on outcomes.
Hodgkin's disease in special situations 195
HODGKIN'S DISEASE IN SPECIAL SITUATIONS
Hodgkin's disease in pregnancy (see also Chapter 19)
Fertility in Hodgkin's disease
Being a condition predominantly affecting young adults, Hodgkin's disease is occasionally diagnosed during pregnancy, at a rate of between 1 in 1000 and 1 in 6000.165 In fact, some large-scale epidemiologic studies have suggested a protective effect with increasing parity, although it is possible that confounding factors may be responsible for the association.166 In such cases, decisions regarding investigation and therapy are clearly complicated.167,168 The balance between effective treatment of the mother and the potential effects upon the fetus will differ considerably according to the maturity of the gestation and the extent of the disease. While it is necessary to consider each individual case with great care, certain general statements may be made. The first of these is that the prognosis of Hodgkin's disease diagnosed during pregnancy does not appear to differ greatly from that diagnosed at other times, in retrospective series.165,169 The pattern of disease is similar and metastatic involvement of the placenta or fetus is extremely rare.170 The need to limit radiation exposure to the fetus throughout gestation restricts the staging investigations that may be used. Whilst chest X-rays may be performed in the third trimester, the use of CT results in too high an exposure and is not employed. Ultrasound examination or MRI may provide alternative methods, but often staging is incomplete and treatment must be given on a pragmatic basis with limited information. For early pregnancy (up to 20 weeks), limited radiotherapy to sites in the chest may be used, although a full mantle field is thought too likely to result in fetal damage, particularly neurological. Chemotherapy is very rarely used during the first trimester, owing to the very high incidence of malformation or abortion.171 For those mothers presenting during early pregnancy with advanced disease requiring treatment, termination of pregnancy may be the only option offering a reasonable prospect of a cure. In later pregnancy, the options are greater; in some cases, it may be possible to defer full staging and therapy until the fetus reaches adequate maturity and can be delivered. For mothers with localized disease, radiotherapy is an option and the risks of mantle irradiation appear to be acceptable.169 For those with advanced disease there is some information to suggest that chemotherapy may be given relatively safely,171 although growth retardation is commonly seen and the long-term sequelae have not been characterized. Theoretical considerations of germ cell damage make the use of alkylating agents inadvisable.
In males, oligozoospermia, poor motility and an increased proportion of abnormal forms are commonly found at presentation, particularly in patients with advanced disease,153,154 and histopathological changes may be seen in the testes of such patients. In females, although the effects are less well documented, disturbances of menstrual function may occur. Chemotherapy, particularly with alkylating agents, can directly damage spermatogenesis and this damage may be permanent.155,58 Early in the course of treatment with MOPP-like regimens, irreversible azoospermia is seen; this also applies to alternating and hybrid regimens (e.g. MOPP/ABVD, LOPP/EVAP, ChlVPP/EVA, ChlVPP/PABlOE, see Chapter 19). Where alkylating agents are avoided (e.g. in anthracycline-based regimens, such as ABVD) fertility is usually preserved.159 In females, chemotherapy progressively damages and irreversibly destroys ovarian follicles; this manifests as oligomenorrhea or amenorrhea and a reduced reproductive span. These changes occur more often in older women and are again particularly a feature of alkylating agent-based regimens.158,,60-163
Exposure of both male and female gonads to therapeutic radiotherapy results in severe damage to these organs; radiation effects are dose related but the threshold dose for damage may be as low as 0.5 Gy. A dose of 4 Gy will produce permanent infertility in a third of even young women and prolonged, possibly irreversible, sterility in men. The likely effects of treatment with chemotherapy or radiotherapy should be discussed with patients. For males, sperm banking should be offered whenever possible. For females, effective techniques for oocyte or ovarian tissue preservation are desirable; as yet, however, these available remain experimental and not generally available. Embryo storage is generally precluded by the need to stimulate the ovaries artificially for some weeks prior to harvesting of oocytes and subsequent fertilization, thus causing unacceptable delays in starting treatment. For patients remaining fertile after treatment, it is prudent to delay starting a family for at least a year; not least because it is during this time that the risk of relapse is highest but also because of the theoretical risk of mutagenesis. However, children successfully conceived after treatment do not appear to have an increased incidence of congenital or later-onset medical problems.164 Males remaining sterile after chemotherapy do not suffer androgen lack. In females, estrogen production is affected and, where premature menopause is confirmed, hormone replacement therapy should be given.
HIV-related Hodgkin's disease Epidemiologic studies have indicated an increased incidence of Hodgkin's disease among the HIV-positive
196 Hodgkin's disease: clinical features
population. One such study yielded an odds ratio of 36.6 (95 per cent confidence limits 0-78.12).172 Several series of cases have been reported in the literature and, in all, the clinical characteristics among HIV-positive patients show a bias towards unfavorable features. These include a higher frequency of mixed cellularity histology, systemic symptoms, stage IV disease and extranodal involvement, particularly of the bone marrow.173,176 The finding of Epstein-Barr virus involvement is universal174,175177,178 and the prognosis is generally reported as poor, with median survival between 1 and 2 years.175,179 As for non-Hodgkin's lymphomas, the prognosis correlates to some extent with the severity of HIV-disease, being worse in those with a low CD 4+ count. In one series no patient with initial CD 4 count below 300/ml survived 2 years.173
CONTRASTS WITH NON-HODGKIN'S LYMPHOMA The clinical features and investigation of nonHodgkin's lymphomas (NHLs) are described fully in Chapters 18, 20-23. However, some general comments can be made to highlight important differences from Hodgkin's disease. NHLs are commoner than Hodgkin's disease in the UK in the ratio of 3.5 to I.180 Their incidence is increasing, whereas that of Hodgkin's disease seems stable. The NHLs are a heterogeneous group of malignancies whose behavior varies from indolent to rapidly lethal; high-grade tumors arise in all age groups although their incidence increases with age, while low-grade tumors are generally confined to later life. They are only marginally commoner in men than women and there is a marked geographical variation in incidence. Their presentation is variable; localized lymphadenopathy particularly in the neck is common, as in Hodgkin's disease, but in contrast unusual nodes may be involved, or there may be more generalized bulky and 'centrifugal' lymphadenopathy. Over one-quarter start outside the classic lymph node areas - most often the gut and Waldeyer's ring but almost any tissue can be affected (Table 15.10) - an extreme rarity in Hodgkin's disease. Knowing the stage of the patient's disease is not so important nowadays for determining treatment but it still has a bearing on prognosis. The Ann Arbor criteria are still useful, although sometimes not entirely appropriate because of the protean patterns of nodal and extranodal spread. For example, NHL starting in the Waldeyer's ring is nowadays generally described as being 'extranodal', which is at variance with the original Ann Arbor definitions. This may in part account for the different recorded incidence rates for 'extranodal' lymphoma. In addition, some series report all cases where
the primary presentation is believed to be extranodal, including disseminated disease. The extent of investigation will be determined largely by the clinical status of the patient and the histopathological subtype. In general, full hematological and biochemical investigation is necessary, and radiological investigations must include, at the very least, chest X-ray to detect mediastinal and hilar lymphadenophathy and some investigation to establish the presence or not of abdominal disease. In the past this was often lymphangiography; current 'state of the art' CT scanning has made this redundant, particularly since it is now evident that, unless the lymphoma is truly localized (an uncommon situation), chemotherapy is likely to be needed as part of the planned treatment. Clinical staging tends to underestimate the incidence of systemic disease. Bone marrow biopsy should be assessed, since marrow involvement will be found in over 30 per cent of cases and, in certain forms with a poor prognosis, particularly of the lymphoblastic type, cerebrospinal fluid should be cytologically examined. The clinicopathological heterogeneity of the NHLs makes it very difficult to define prognostic factors. Perhaps the best researched group is that of the diffuse large cell lymphomas. For example, major centers treating this subtype, which accounts for over a third of NHLs, have collaborated to develop an International Prognostic Index (IPI). Based on data from over 3000 patients, this index clearly distinguishes subgroups that differ in terms of tumor response, relapse and overall survival.186 The IPI is based on straightforward clinical and biological data (age, Ann Arbor stage, serum LDH level, performance status and number of extranodal sites of disease). In its age-adjusted simplified version, it enables clinicians to make appropriate choices of treatment, and can be used to select and stratify patients entering prospective clinical trials. Such factors may vary during the course of the disease and in a study of such variations (involving 1271 patients on GELA trials), it was concluded that, while IPI factors are relevant to short-term follow-up, only performance status was predictive of patients' ability to tolerate induction chemotherapy and only tumor stage was predictive of long-term survival.187 Unfortunately, with several other putative prognostic indices, the number of patients assessed is not large enough to give statistically relevant guidelines. Relatively simple predictors of adverse prognosis which merit further study include the following: • • • •
T cell immunophenotype;188 low serum albumin;189,190 raised ferritin;191 raised b2 microglobulin.192
Other more sophisticated but generally expensive and time-consuming investigations take into account the intrinsic biological characteristics of lymphoma cells.193
Contrasts with non-Hodgkin's lymphoma 197
Table 15.10 The incidence of extranodal lymphoma
All NHL Extranodal NHL
8767 1467(17%)
Gastrointestinal Stomach 346 (23.5%) Small intestine 110(7.5%) Large intestine 82 (5.5%) Head and neck Tonsil Nasopharynx Tongue Nose Salivary gland Thyroid Other Skin
142 (9.5%) 37 (2.5%) 20(1.0%) 33 (2.0%) 69 (4.5%) 36 (2.5%) 48 (3.5%) 110(7.5%)
1257 463(37%)a,b
580 236(41 %)a
832 299 (36%)a
2007 394(20%)
87(17.0%) 41 (8.0%) 11 (2.0%)
54 (23.0%) 13(5.5%) 16(6.5%)
46(15.5%)' 56(18.5%)f 16(5.5%)'
53(13.5%) 25 (6.5%) 16(4.0%)
d
25(10.5%) 11 (4.5%) 5 (2.0%) 3(1.0%) 3(1.0%) 4(1.5%) 2(1.0%)
28 (9.5%) 26 (8.5%)
2 (0.5%)
47 (12.0%) 15(4.0%) 5(1.5%) 20 (5.0%) 16(4.0%) 27 (7.0%) 13 (3.5%)
12 (4.0%)
33 (8.5%)
d d
17(3.5%) 17(3.5%) 25 (5.0%) 3 (0.5%)
e
4(1.5%)
56(11.0%)
4(1.5%)
Connective tissues 90 (6.0%)
12 (2.5%)
5 (2.0%)
6 (2.0%)
11 (3.0%)
Bone
69 (4.5%)
41 (8.0%)
6 (2.5%)
19(6.5%)
19(5.0%)
Lung
53 (3.5%)
24 (5.0%)
11 (4.5%)
7 (2.5%)
3(1.0%)
Breast
33 (2.0%)
6(1.0%)
5 (2.0%)
—
5(1.5%)
Testis
23(1.5%)
15(3.0%)
4(1.5%)
1 (0.5%)
14 (3.5%)
Orbit (and eye)
32 (2.0%)
4(1.0%)
7 (3.0%)
7 (2.5%)
31 (8.0%)
c
8
Central nervous system
23(1.5%)
33 (6.5%)
15(6.5%)
3(1.0%)
34 (9.0%)
Marrow
—
89(17.5%)
33 (14.0%)
—
—
Liver
6 (0.5%)
27 (5.0%)
—
—
1 (0.5%)
Female genital
16(1.0%)
6(1.0%)
1 (0.5%)
2(0.5)
2(0.5)
Others
89 (6.0%)
29 (5.5%)
9 (4.0%)
64(21.5%)
4(1.0%)
a b c d e f g
Including disseminated disease at presentation. Including cases with more than one presenting site. Including orbital connective tissue. Excluding Waldeyer's ring. Excluding mycosis fungoides. Excluding 47 cases of gastrointestinal presentation where the origin is uncertain or multiple. Brain only.
Although numbers of patients studied are small, the following predictive factors have been reported: • • •
•
genetic abnormalities - a number of these have been described (see Chapter 9); abnormalities in cell cycle regulation - in general high proliferative activity adversely affects survival;194-197 pattern of expression of adhesion molecules - variably reported to be predictive of outcome, but CD 44 (hyaluronate receptor) expression in particular seems to be an adverse marker;198-199 cytokine levels - poor outcome has been described
•
with elevated levels of IL-6, IL-10 and tumor necrosis factor oc (TNFa); 200-202 lymphoma cell chemosensitivity indicators - MDR and bcl-2 expression at diagnosis are unfavorable prognostic factors.203,204
The shortcomings of the Ann Arbor staging classification are particularly well demonstrated with the follicular lymphomas. In a study of 398 patients entered into British National Lymphoma Investigation trials between 1974 and 1980, the Ann Arbor system fared poorly as a prognostic indicator; most powerfully significant factors
198 Hodgkin's disease: clinical features
for cause-specific survival were the number of lymph node regions involved, splenomegaly, constitutional symptoms and age. With the exception of splenomegaly, other studies have shown similar findings.205,208 The factors described above for aggressive NHLs186 may also be relevant to the follicular lymphomas,209 and likewise molecular and cytogenetic markers may be of prognostic value. However, it is still difficult to pick out a group of patients with such adverse features as to warrant more intensive (or experimental) treatment at presentation.
16.
17.
18.
19.
REFERENCES 1. Hodgkin T. On some morbid appearances of the absorbent glands and spleen. Med Chir Trans 1832; 17: 68-114. 2. Wilks Sir S. Cases of enlargement of the lymphatic glands and spleen (or Hodgkin's disease) with remarks. , Guy's Hospital Rep 1865; 11: 56-67. 3. Sternberg C. Uber eine eigenartige unter dem Bilde der Pseudoleukamie verlaufende Tuberculose des lymphatischen Apparates. Z Heilk 1898; 19: 21-90. 4. Reed DM. On the pathological changes in Hodgkin's disease, with especial reference to its relation to tuberculosis. Johns Hopkins Hasp Rep 1902; 10:133-96. 5. Olivier A, Ranvier L. Observation pour sevir a I'histoire de I'adenie Memoires lus a la societe de biologie, 1867: 99-112. 6. Tuckwell HM. Enlargement of lymph glands in the abdomen with formation of peculiar morbid growths in the spleen and peritoneum. Trans Path Soc. London, 1870; 21: 362-5. 7. Langans T. Das maligne lymphosarkom. VirchowsArch PathAnat 1872; 54: 509-37. 8. Greenfield S. Specimens illustrative of the pathology of lymphadenoma and leucocythaemia. Trans Path Soc. London,1878; 29:272-304. 9. GowersWR. Hodgkin's disease. In: Reynolds JRed./l system of medicine. London: Macmillan, 1878: 306-52. 10. Pusey WA. Cases of sarcoma and of Hodgkin's disease treated by exposure to x rays: a preliminary report. JAMA 1902; 38:166-9. 11. Gilbert R. Radiotherapy in Hodgkin's disease (malignant granulomatosis): anatomic and clinical foundations; governing principles; results. Am Roentgenol 1939; 41: 198-241. 12. Peters MV. A study of survivals in Hodgkin's disease treated radiologically. Am J Roentgenol 1950; 63: 299-311. 13. Kaplan HS. The radical radiotherapy of regionally localised disease. Radiology 1962; 78: 553-61. 14. Kaplan HS. Long term results of palliative and radical radiotherapy of Hodgkin's disease. Cancer Res 1966; 26: 1250-2. 15. Goodman LS, Wintrobe MM, Dameshek W, Goodman MJ,
20.
21.
22.
23.
24.
25. 26.
27.
28.
29.
30.
31.
Culman AZ, McLennon MT. Nitrogen mustard therapy.7 Am Med Ass 1946; 132:126-32. Lacher ML), DurantJR. Combined vinblastine and chlorambucil therapy of Hodgkin's disease. Ann Intern Med 1965; 62: 468-76. De Vita VT, Serpick A, Carbone PP. Combination chemotherapy in the treatment of advanced Hodgkin's disease. Ann Intern Med 1970; 73: 881-95. De Vita VT, Simon RM, et al. The curability of advanced Hodgkin's disease with chemotherapy. Ann Intern Med 1980; 92: 587-95. Correa P, O'Conor GT. Epidemiologic patterns of Hodgkin's disease. IntJ Cancer 1971; 8:192-201. Hancock BW, Vaughan HG, Vaughan HB, Bennett MH. LOPP alternating with EVAP is superior to LOPP alone in the initial treatment of advanced Hodgkin's disease: results of a British National Lymphoma Investigation trial. 7 Clin Oncol 1992; 10:1252-8. Cullen MH, Stuart NS, Woodroffe C, Murphy A, Fletcher J, Blackledge GR. ChlVPP/PABlOE and radiotherapy in advanced Hodgkin's disease.J Clin Oncol 1994; 12: 779-87. Kennedy BJ, Loeb V, Peterson VM, et al. National survey of patterns of care for Hodgkin's disease. Cancer 1985; 56: 2547-56. Atkinson MK, McElwain TJ, Peckham MJ, Thomas PRM. Hypertrophic pulmonary osteoarthropathy in Hodgkin's disease; reversal with chemotherapy. Cancer 1976; 38: 1729-34. Selby P, McElwain T. Clinical features of Hodgkin's disease. Hodgkin's disease. Oxford: Blackwell Science, 1987; 94-125. Kaplan HS. Hodgkin's disease. 2nd edn. Cambridge, MA: Harvard University Press, 1980. Magrini SM, Cellai E, Papi MG, Pertid M, Lastrucci L, Ponticelli P. Therapeutic implications of the peculiar clinical characteristics of Hodgkin's disease in clinical stages l-ll with isolated neck presentation. Radiol Med. Torino, 1995; 90:108-12. Enrici RM, Osti MF, Anselmo AP, Banelli E, Cartoni C, Sbarbati S. Hodgkin's disease stage I and II with exclusive subdiaphragmatic presentation. The experience of the Departments of Radiation Oncology and Hematology, University 'La Sapienza' of Rome. Tumori 1996; 82: 48-52. Barrett A, Gregor A, McElwain TJ, Peckham MJ. Infradiaphragmatic presentation of Hodgkin's disease. Clin Radiol 1981; 32: 221-4. Krikorian JG, Portlock CS, Rosenberg SA, Kaplan HS. Hodgkin's disease, stages I and II, occurring below the diaphragm. Cancer 1979; 43:1866-71. Dimitrovsky E, Martin SE, et al. Lymph node aspiration in the management of Hodgkin's disease.) Clin Oncol 1986; 4: 306-10. Rosenberg SA, Kaplan HS. Evidence for an orderly progression in the spread of Hodgkin's disease. Cancer Res 1966; 26:1225-31.
References 199
32. Smithers DW. Spread of Hodgkin's disease. Lancet 1970; i: 1262-7.
52. Harper PG, Fischer C, MacLennan K, Souhami RL Presentation of Hodgkin's disease as an endobronchial
33. Smithers DW. Modes of spread. In: Smithers DW, ed. Hodgkin's disease. Edinburgh: Churchill Livingstone, 1973:107-17.
lesion. Cancer 1984; 53:147-50. 53. Perez-Soler R, McLaughlin P, Velasquez WS, et al.
34. Mauch PM, Kalish LA, Kadin M, Coleman CN, Osteen R, Hellman S. Patterns of presentation of Hodgkin disease: implications for etiology and pathogenesis. Cancer 1993; 71: 2062-71. 35. Carbone PP, Kaplan HS, Musshoff K, Smithers DW, Tubiana M. Report of the committee on Hodgkin's disease staging. Cancer Res 1971; 31:1860-1. 36. Specht L, Nordentoft AM, Cold S, Clausen NT, Nissen Nl.
Clinical features and results of management of superior vena cava syndrome secondary to lymphoma. J Clin Oncol 1984; 2: 260-6. 54. Sextro M, Diehl V, Franklin J, Hansmann ML, Anagnostopoulos I. Lymphocyte predominant Hodgkin's disease - a workshop report. European Task Force on Lymphoma. Ann Oncol 1996; 4: 61-5. 55. Aisenberg AC, Kaplan MM, Rieder SG, Goldman JM.
Tumor burden as the most important prognostic factor in early stage. Cancer 1988; 61:1719-27.
Serum alkaline phosphatase at the onset of Hodgkin's disease. Cancer 1970; 26: 318-26. 56. Glees JP, Barr LC, McElwain TJ, Peckham MJ, Gazet J-C.
37. Lister TA, Crowther D, Sutcliffe SB, Glatstein E, Canellos
The changing role of staging laparotomy in Hodgkin's
GP. Report of a committee convened to discuss the evaluation and staging of Hodgkins disease, J Clin Oncol 1989; 7:1630-6.
disease: a personal series of 310 patients. BrJ Surg 1982; 69:181-7. 57. Thyss A, Schieder M, Caldani C, Viot M, Bourry J. Re-
38. Glatstein E, Guernsey JM, Rosenberg SA, Kaplan HS. The
evaluation of alkaline phosphatase measurement
value of laparotomy and splenectomy in the staging of Hodgkin's disease. Cancer 1969; 24: 709-18. 39. Glatstein E, Trueblood HW, Enright LP, Rosenberg SA, Kaplan HS. Surgical staging of abdominal involvement
during Hodgkin's disease by electrophoretic isoenzyme separation. BrJ Cancer 1985; 52:183-7. 58. Fabian CJ, Mansfield CM, Dixon DO, et al. Incidence of liver involvement and correlation of biopsy results with
in unselected patients with Hodgkin's disease. Radiology 1970; 97: 425-32. 40. Dorfman RF. Relationship of histology to site in Hodgkin's disease. Cancer Res 1971; 31:1786-93. 41. LandbergT, Larsson L-E. Hodgkin's diseaseretrospective clinico-pathologic study in 149 patients. Acta Radiol 1969; 8: 390-414. 42. Yu A, Steinfeld AD. Hodgkin's disease presenting in epitrochlear nodes. Med Pediat Oncol 1984; 12: 244-6. 43. Cionini L, Bastiani P, Biti G, et al. Waldeyer's ring involvement in Hodgkin's disease. Radiother Oncol 1985; 3:299-302. 44. Kapadia SB, Roman LN, Kingma DW, Jaffe ES, Frizzera G. Hodgkin's disease of Waldeyer's ring. Clinical and histoimmunophenotypic findings and association with Epstein-Barr virus in 16 cases. Am J Surg Pathol 1995; 19:1431-9. 45. Colby TV, Hoppe RT, Warnke RA. Hodgkin's disease: a clinicopathologic study of 659 cases. Cancer 1981; 49: 1848-58. 46. McDonald JS. Lung involvement in Hodgkin's disease. Thorax 1977; 32: 664-7. 47. Whitcomb ME, Schwartz Ml, Keller AR, et al. Hodgkin's disease of the lung. Am Rev Resp Dis 1972; 106: 79-85. 48. Colby TV, Hoppe RT, Warnke RA. Hodgkin's disease at autopsy: 1972-1977. Cancer 1981; 47:1852-62. 49. Homan R, Lechner K, Neumann E, et al. Primary Hodgkin's disease of the lung. Blut 1983; 47: 231-5.
Ann Arbor clinical criteria and response to chemotherapy in advanced Hodgkin's disease: a SWOG study. In: Cavalli F, Bonadonna G, Rosencweig M, eds. Malignant lymphoma and Hodgkin's disease: experimental and therapeutic advances. The Hague: Martinus Nijhoff, 1985: 409-16. 59. Trewby PN, Portmann B, Brinkley DM, Williams R. Liver disease as the presenting manifestation of Hodgkin's disease. Q J Med 1979; 189:137-50. 60. Sacks E, Donaldson SS, Gordon J, et al. Epitheloid granulomas associated with Hodgkin's disease. Cancer 1979; 41: 562-7. 61. Aderka D, Kraus M, Avido I, etal. Hodgkin's disease and non-Hodgkin's lymphoma masquerading as idiopathic liver granulomas. Am J Gastroenterol 1984; 79: 642-4. 62. Trier H, Christensen M. Hodgkin's disease in sarcoidosis. Acta Med Scand 1985; 218:137-40. 63. Perera DR, Greene ML, Fenster LF. Cholestasis associated with extrabiliary Hodgkin's disease; report of three cases and review of four others. Gastroenterology 1974; 67: 680-5. 64. Peck B. Hypertrophic osteoarthropathy with Hodgkin's disease in the mediastinum.y/4M/4 1976; 238:1400-1. 65. Friedman M, Kim TH, Panahon AM. Spinal cord compression in malignant lymphoma. Treatment and results. Cancer 1976; 37:1485-91. 66. Higgins SA, Peschel RE. Hodgkin's disease with spinal cord compression. A case report and a review of the
50. Yousem SA, Weiss LM, Colby TV. Primary pulmonary
literature. Cancer 1995; 75: 94-8. 67. Sapozink MD, Kaplan HS. Intracranial Hodgkin's disease.
Hodgkin's disease. Cancer 1986; 57:1217-24. 51. Selby P, McElwain T. Hodgkin's disease. Oxford:
68. Vera R, Enriquez R, Papac R. Hodgkin's disease,
Blackwell Scientific Publications, 1987.
Cancer 1983; 52:1301-7. intracranial involvement. Amy Clin Oncol 1985; 8: 73-6.
200 Hodgkin's disease: clinical features 69. Dillman RO, Much J, Greco CM, et al. Leptomeningeal Hodgkin's disease. Ann Intern Med 1980; 92: 714-15. 70. Orlowski EP, Hansen RM, Anderson T, et al. Hodgkin's disease with leptomeningeal involvement. Cancer 1984; 53:1833-5. 71. Bartl R, Frisch B, Burkhardt R, Huhn D, Pappenberger R. Assessment of bone marrow histology in Hodgkin's disease: correlation with clinical factors. BrJ Haematol 1982;51:345-60. 72. Carde P, MacKintosh FR, Rosenberg SA. A dose and time response analysis of the treatment of Hodgkin's disease with MOPP chemotherapy.) Clin Oncol 1983; 1: 146-53. 73. Hayhoe FGJ, Burns GF, Cawley JC, Stewart JW. Cytochemical, ultrastructural and immunological studies of circulating Reed-Sternberg cells. BrJ Haematol 1978; 38: 485-90. 74. Linch DC, Berliner N, O'Flynn, et al. Hodgkin cell leukaemia of B cell origin. Lancet 1985; i: 78-80. 75. Filly R, Blank N, Castellino RA. Radiographic distribution of intrathoracic disease in previously untreated patients with Hodgkin's disease and nonHodgkin's lymphoma. Radiology 1976; 120: 277-81. 76. Huber SC, Webb WR, Kuykendall JD, Meier WL Abnormal cardiac silhouette in a patient with Hodgkin's disease. JAMA 1978; 239: 2785-6. 77. ReiffersJ, Pars J B, HuguesA, Hoerni B. Hodgkin's disease localised in the thyroid gland. Nouv Presse Med 1976; 5:1590. 78. Al Kutoubi MA, Patep PR, Coulter C. Laryngeal involvement in Hodgkin's disease. BrJ Radiol 1985; 58: 184-6. 79. Champion A, Coup AJ, Hancock BW. Hodgkin's disease and chronic renal failure. Cancer 1976; 38:1867-8. 80. Hung LHY, Kartz DM. Hodgkin's disease of the endometrium. Arch Pathol Lab Med 1983; 109: 952-3. 81. Routledge RC, Haun IM, Morris Jones PH. Hodgkin's disease complicated by the nephrotic syndrome. Cancer 1976; 38:1735-40. 82. Shapiro CM, VanderLaan BF, Wellington J, Sloan DR. Nephrotic syndrome in two patients with cured Hodgkin's disease. Cancer 1985; 55:1799-804. 83. Hancock BW, Richmond J, Powell T, Emmanuel IG. Intrathoracic Hodgkin's disease presenting as hypertrophic osteoarthropathy. BrJ Radiol 1976; 49: 647-9. 84. Dimopoulos MA, Cabanillas F, Lee JJ, Swan F, Fuller L, Allen PK. Prognostic role of serum beta 2-microglobulin in Hodgkin's disease.) Clin Oncol 1993; 11:1108-11. 85. Jochelson MS, Herman TS, Stomper PC, Mauch PM, Kaplan WD. Planning mantle radiation therapy in patients with Hodgkin disease. Am J Roentgenol 1988; 151:1229-31. 86. Hagemeister FB, Purugganan R, Podoloff DA, Hess M, Rodriguez MA. The gallium scan predicts relapse in patients with Hodgkin's disease. Ann Oncol 1994; 2: 59-63.
87. Salloum E, Brandt DS, Caride VJ, Cornelius E, Zelterman D, Cooper DL. Gallium scans in the management of patients with Hodgkin's disease: a study of 101 cases. Blood 1996; 10(suppl 1): 229a. 88. Devizzi L, Gasparini M, Santoro A, et al. The role of gallium-67 scan in staging and restaging patients with Hodgkin's disease. Ann Oncol 1996; 7(suppl 3): 35. 89. Bangerter M, Griesshammer M, Binder T, Hafner M, Heimpel H, Reske SN. New diagnostic imaging procedures in Hodgkin's disease. Ann Oncol 1996; 4: 55-9. 90. Rosenberg SA. Hodgkin's disease of the bone marrow. Cancer Res 1971; 31:1733-6. 91. Macintyre EA, Vaughan Hudson B, Linch DC, Vaughan Hudson G, Jelliffe AM. The value of staging bone marrow trephine biopsy in Hodgkin's disease. EurJ Haematol 1987; 39: 66-70. 92. Munker R, Hasenclever D, Brosteanu 0, Hiller E, Diehl V. Bone marrow involvement in Hodgkin's disease: an analysis of 135 consecutive cases, y Clin Oncol 1995; 13: 403-9. 93. Haybittle JL, Hayhoe FG, Easterling MJ, Jelliffe AM, Bennett MH. Review of British National Lymphoma Investigation studies of Hodgkin's disease and development of prognostic index. Lancet 1985; 1: 967-72. 94. Henry-Amar M, Friedman S, Hayat M, Somers R, Meerwaldt JH, Carde P. Erythrocyte sedimentation rate predicts early relapse and survival in early-stage Hodgkin disease. Ann Intern Med 1991; 114: 361-5. 95. Wrigley PFM, Hamilton Fairley G, Matthias JQ. Anaemia and bone marrow involvement in Hodgkin's disease. In: Smithers DW, ed. Hodgkin's disease. Edinburgh: Churchill Livingstone, 1973:154-63. 96. Eisner E, Ley AB, Meyer K. Coomb's positive haemolytic anaemia in Hodgkin's disease. Ann Intern Med 1967; 66: 258-73. 97. Levine AM, Thornton P, Forman SJ, et al. Positive Coomb's test in Hodgkin's disease: significance and implications. Blood 1980; 55: 607-10. 98. Waddell CC, Cino PL. Immune thrombocytopenia purpura occurring in Hodgkin's disease after splenectomy. Am J Hematol 1979; 7: 381-7. 99. Murphy WG, Perry DJ, Allan NC, Stockdill G. Hodgkin's disease presenting as idiopathic thrombocytopenic purpura. Postgrad Med j 1984; 60: 614-15. 100. Weitberg AB, Harmon DC. Autoimmune neutropenia, haemolytic anaemia and reticulocytopenia in Hodgkin's disease. Ann Intern Med 1984; 100: 702-4. 101. Garcia R, HernandezJM, Caballero MD, Gonzalez M, Galende J. Serum lactate dehydrogenase level as a prognostic factor in Hodgkin's disease. BrJ Cancer 1993; 68:1227-31. 102. Zaloga GP, Eil C, Medbury CA. Humoral hypercalcaemia in Hodgkin's disease. Arch Intern Med 1985; 145:155-6. 103. Gause A, Scholz R, Klein S, et al. Increased levels of circulating interleukin-6 in patients with Hodgkin's disease. Haematol Onco/1991; 9: 307-13.
References 201 104. Denizot Y, Fixe P, Liozon E, Brigaudeau C, Praloran V. Serum IL-6 concentrations in lymphomas. BrJ Haematol 1995; 90: 731-2. 105. Blay J-Y, Farcet J-P, Lavaud A, Radoux D, Chouaib S. Serum concentrations of cytokines in patients with Hodgkin's disease. EurJ Cancer Part A1994; 30A: 321-4. 106. Gorschluter M, Bohlen H, Hasenclever D, Diehl V, Tesch H. Serum cytokine levels correlate with clinical parameters in Hodgkin's disease. Ann Oncol 1995; 6: 477-82. 107. KurzrockR, Redman J, Cabanillas F, Jones D, RothbergJ, Talpaz M. Serum interleukin levels are elevated in lymphoma patients and correlate with survival in advanced Hodgkin's disease and with B symptoms. Cancer Res 1993; 53: 2118-22. 108. Sarris AH, Straus D, Preti A, et al. A prognostic model for advanced Hodgkin's disease at MD Anderson validated with an independent set of patients treated at Memorial Sloan-Kettering. Blood 1996; 88(suppl 1): 226a. 109. Sarris AH, Kliche KO, Pethambaram P, etal. Interleukin10 levels are often elevated in serum of adults with Hodgkin's disease and are associated with inferior failure-free survival. Ann Oncol 1999; 10: 433-40. 110. Hamon MD, Unal E, Macdonald I, Shamim F, Boesen E, Prentice HG. Plasma soluble interleukin 2 receptor levels in patients with malignant lymphoma are correlated with disease activity but not cellular immunosuppression. Leak Lymphoma 1993; 10:111-5. 111. Gause A, Jung W, Schmits R et al. Soluble CD8, CD25 and CD30 antigens as prognostic markers in patients with untreated Hodgkin's lymphoma. Ann Oncol 1992; 3(suppl 4): 49-52. 112. Gnant M, Mader RM, Djavanmard M, Jakesz R, Steger GG. Soluble interleukin-2 receptor is a powerful prognostic marker in Hodgkin's disease. Proceedings of the Annual Meeting of the American Society of Clinical Oncology 1995: 648a. 113. Enblad G, Sundstrom C, Gronowitz S, Glimelius B. Serum levels of interleukin-2 receptor (CD25) in patients with Hodgkin's disease, with special reference to age and prognosis. Ann Oncol 1995; 6: 65-70. 114. Nadali G, Vinante F, Ambrosetti A, et al. Serum levels of soluble CD30 are elevated in the majority of untreated patients with Hodgkin's disease and correlate with clinical features and prognosis. J Clin Oncol 1994; 12: 793-7. 115. Gause A, Pohl C, Tschiersch A, et al. Clinical significance of soluble CD30 antigen in the sera of patients with untreated Hodgkin's disease. Blood 1991; 77:1983-8. 116. Christiansen I, Enblad G, Kalkner KM, Gidlof C, Glimelius B, Totterman TH. Soluble ICAM-1 in Hodgkin's disease: a promising independent predictive marker for survival. Leuk Lymphoma 1995; 19: 243-51. 117. Gruss H-J, Dolken G, Brach MA, Mertelsmann R, Herrmann F. The significance of serum levels of soluble 6- kDa receptors for tumor necrosis factor in patients with Hodgkin's disease. Leukemia 1993; 7:1339-43.
118. DenzH, Prth B, Weiss G, et al. Serum soluble tumour necrosis factor receptor 55 is increased in patients with haematological neoplasiasand is associated with immune activation and weight loss. EurJ Cancer 1993; 29A: 2232-5. 119. Gruss H-J, Dolken G, Brach MA, Mertelsmann R, Herrmann F. Serum levels of circulating ICAM-1 are increased in Hodgkin's disease. Leukemia 1993; 7: 1245-9. 120. Nadali G, Vinante F, Rigo A, et al. Correlation between clinical features and circulating levels of soluble intercellular adhesion molecule-1 in Hodgkin's disease. Int J Clin Lab Res 1995; 25: 84-7. 121. Pizzolo G, Vinante F, Nadali G,et al. ICAM-1 tissue overexpression associated with increased serum levels of its soluble form in Hodgkin's disease. BrJ Haematol 1993;84:161-2. 122. Tubiana M, Henry-Amar M, Burgers MV, et al. Prognostic significance of erythrocyte sedimentation rate in clinical stage ll-lll of Hodgkin's disease. J Clin Oncol 1984; 2:194-200. 123. Carde P, Burgers JM, Henry AM, Hayat M, Sizoo W. Clinical stages I and II Hodgkin's disease: a specifically tailored therapy according to prognostic factors. J Clin Oncol 1988; 6: 239-52. 124. Mauch P, Larson D, Osteen R, et al. Prognostic factors for positive surgical staging in patients with Hodgkin's disease. J Clin Oncol 1990; 8: 257-65. 125. Kaldor JM, Day NE, Clarke EA, Van LF, Henry AM. Leukemia following Hodgkin's disease. N EnglJ Med 1990; 322: 7-13. 126. Carde P, Hagenbeek A, Hayat M, et al. Clinical staging versus laparotomyand combined modality with MOPP versus ABVD in early-stage Hodgkin's disease: the H6 twin randomized trials from the European Organization for Research and Treatment of Cancer Lymphoma Cooperative Group. J Clin Oncol 1993; 11: 2258-72. 127. Abrahamsen AF, Hannisdal E, Nome 0, et al. Clinical stage I and II Hodgkin's disease: long-term results of therapy. Ann Oncol 1996; 7:145-50. 128. Cox DR. Regression models and life tables. J R Stat Soc 1972; 34: 187. 129. Hasenclever D, Diehl V for the International Prognostic Factors Project on Advanced Hodgkin's Disease. A prognostic score for advanced Hodgkin's disease. N Engl J Med 1998; 339:1506-14. 130. Hasenclever D, Diehl V. A numerical index to predict tumour control in advanced Hodgkin's disease. Blood 1996; 88(suppl 1): 673a. 131. Dady PJ, McElwain TJ, Austin DE, et al. Five years experience with ChlVPP: effective low-toxicity combination chemotherapy for Hodgkin's disease. BrJ Cancer 1982; 45: 851-9. 132. Gobbi PG, Cavalli C, Federico M, Bertoloni D, Di PU, Rossi A. Hodgkin's disease prognosis: a directly predictive equation. Lancet 1988; 1: 675-9.
202 Hodgkin's disease: clinical features 133. Mauch P, Hellman S. Mediastinal Hodgkin's disease. Haematol Oncol 1984; 2: 69-71. 134. Mauch P, Tarbell N, Weinstein H, et al. Stage IA and MA supradiaphragmatic Hodgkin's disease: prognostic factors in surgically staged patients treated with mantle and para-aortic irradiation.J Clin Oncol 1988; 6: 1576-83. 135. TharTL, Million RR, Housner RJ, et al. Hodgkin's disease stage I and II, relationship of recurrence to size, radiation and dose and number of sites involved. Cancer 1979; 43:1101-5. 136. Lee CK, Bloomfield CD, Goldman Al, et al. Prognostic significance of mediastinal involvement in Hodgkin's disease treated with curative radiotherapy. Cancer 1980; 46: 2403-9. 137. Liew KH, Easton D, Horwich A, Barrett A, Peckham MJ. Bulky mediastinal Hodgkin's disease, management and prognosis. Haematol Oncol 1984; 2: 45-59. 138. Cosset JM, Henry-Amar M, Carde P, et al. The prognostic significance of large mediastinal masses in the treatment of Hodgkin's disease. The experience of the Institute Gustave-Roussy. Haematol Oncol 1984; 2: 33-43. 139. Zucali R, Zanini M, Banfi A. Significance of mediastinal involvement in early Hodgkin's disease. Haematol Oncol 1984; 2: 72-4. 140. Radford JA, Cowan RA, Ryder WDJ, et al. Four weeks of neo-adjuvant chemotherapy significantly reduces the progression rate in patients treated with limited field radiotherapy for clinical stage IA/IIA Hodgkin's disease. Results of a randomised pilot study. Ann Oncol 1996; 7(suppl 3): 10. 141. Bjorkholm M, Wedelin C, Holm G, Essy-Ehsing B. Familial longevity and prognosis in Hodgkin's disease. Cancer 1984; 54:108-92. 142. Walker LG, Ratcliffe MA, Dawson AA. Does personality affect survival in lymphoma patients? Psychol Oncol 1995; 4: 98. 143. Wagstaff J, Gregory WM, Swindell R, Crowther D, Lister PA. Prognostic factors for survival in stage NIB and IV Hodgkin's disease: a multivariate analysis comparing two specialist treatment centres. BrJ Cancer 1988; 58: 487-92. 144. Proctor SJ, Taylor P, Donnan P, Boys R, Lennard A, Prescott RJ. A numerical prognostic index for clinical use in identification of poor-risk patients with Hodgkin's disease at diagnosis. EurJ Cancer 1991; 27: 624-9. 145. Ferme C, Bastion Y, Brice P, et al. for the Groupe d'Etudes des Lymphomes de'Adulte. Prognosis of patients with advanced Hodgkin's disease. Cancer 1997; 80:1124-33. 146. Korn EL, Simon R. Measures of explained variation for survival data. Stat Med 1990; 9: 487-503. 147. Selby P, Patel P, Milan S, et al. ChlVPP combination chemotherapy for Hodgkin's disease: long-term results. BrJ Cancer 1990; 62: 279-85.
148. Tubiana M, Henry-Amar M, Hayat M. Prognostic significance of the number of involved areas in the early stages of Hodgkin's disease. Cancer 1984a; 54: 885-94. 149. Young RC, Canellos GP, Chabner BA, Hubbard SM, De Vita VT Jr. Patterns of relapse in advanced Hodgkin's disease treated with combination chemotherapy. Cancer 1978;42:1001-7. 150. Oza AM, Ganesan TS, Leahy M, et al. Patterns of survival in patients with Hodgkin's disease: long follow up. Ann Oncol 1993; 4: 385-92. 151. Herman TS, Hoppe RT, Donaldson SS, et al. Late relapse among patients treated for Hodgkin's disease. Ann Intern Med 1985; 102: 292-7. 152. Radford JA, Eardley A, Woodman C, Crowther D. Follow up policy after treatment for Hodgkin's disease: too many clinic visits and routine tests? A review of hospital records. BrMedJ 1997; 314: 343-6. 153. Vigersky R, Chapman RM, BerenbergJ, et al. Testicular dysfunction in untreated Hodgkin's disease. Am J Med 1982; 73: 482-6. 154. Hendry WF, Stedronska J, Jones CR, Blackmore CA, Barrett A, Peckham MJ. Semen analysis in testicular cancer and Hodgkin's disease: pre-and post-treament findings and implications for cryopreservation. BrJ Urol 1983; 55: 769-73. 155. Chapman RM, Sutcliffe SB, Rees LH, et al. Cyclical combination chemotherapy and gonadal function. Lancet 1979; 1:285-9. 156. Waxman JHX, Terry YA, Wrigley PFM, et al. Gonadal function in Hodgkin's disease: long-term follow-up of chemotherapy. BrMedJ 1982; 285:1612-13. 157. Whitehead E, Shalet SM, Blackledge G, et al. The effects of Hodgkin's disease and combination chemotherapy on gonadal function in the adult male. Cancer 1982; 49: 418-22. 158. Clark ST, Radford JA, Crowther D, Swindell R, Shalet SM. Gonadal function following chemotherapy for Hodgkin's disease: a comparative study of MVPP and a seven-drug hybrid regimen. J Clin Oncol 1995; 13:134-9. 159. Viviani S, Santoro A, Ragni G, et al. Gonadal toxicity after combination chemotherapy for Hodgkin's disease. Comparative results of MOPP vs ABVD. EurJ Cancer Clin Oncol 1985; 21: 601-5. 160. Chapman RM, Sutcliffe SB, Malpas JS. Cytotoxic-induced ovarian failure in women with Hodgkin's disease. JAMA 1979; 242:1877-81. 161. Horning SJ, Hoppe RT, Kaplan HS, Rosenberg SA. Female reproductive potential after treatment for Hodgkin's disease. N EnglJ Med 1981; 304:1377-82. 162. Schilsky RL, Sherins RJ, Hubbard SM, Wesley MN, Young RC, DeVita VT. Long term follow-up of ovarian function in women treated with MOPP chemotherapy for Hodgkin's disease. 4m7 Med 1981; 71: 552-6. 163. Whitehead E, Shalet SM, Blackledge G, Todd I, Crowther D, Beardwell CG. The effect of combination chemotherapy on ovarian function in women treated for Hodgkin's disease. Cancer 1983; 52: 988-93.
References 203 164. Swerdlow AJ, Jacobs PA, Marks A, et al. Fertility, reproductive outcomes, and health of offspring, of patients treated for Hodgkin's disease: an investigation including chromosome examinations. BrJ Cancer 1996; 74: 291-6. 165. Lishner M, Zemlickis D, Degendorfer P, Panzarella T, Sutcliffe SB, Koren G. Maternal and foetal outcome following Hodgkin's disease in pregnancy. BrJ Cancer 1992; 65:114-17. 166. Kravdal 0, Hansen S. Hodgkin's disease: the protective effect of childbearing. IntJ Cancer 1993; 55: 909-14. 167. Sutcliffe SB. Treatment of neoplastic disease during pregnancy: maternal and fetal effects. Clin Invest Med 1985; 8: 333-8. 168. Doll DC, Ringenberg QS, Yarbro JW. Management of cancer during pregnancy. Arch Intern Med 1988; 148: 2058-64. 169. Woo SY, Fuller LM, Cundiff JH, Bondy ML, Hagemeister FB, Mclaughlin P. Radiotherapy during pregnancy for clinical stages IA-IIA Hodgkin's disease. IntJ Radial Oncol Biol Phys 1992; 23: 407-12. 170. Rothman LA, Cohen CJ, Astarloa J. Placental and fetal involvement by maternal malignancy: a report of rectal carcinoma and review of the literature. AmJObstet Gynecol 1973; 116:1023-34. 171. Yahalom J. Treatment options for Hodgkin's disease during pregnancy. Leuk Lymphoma 1990; 2: 151-61. 172. Serraino D, Pezzotti P, Cozzi-Lepri A, Grigoletti E, Tirelli U, Rezza G. Incidence of Hodgkin's disease in a cohort of HIV seroconverters. Proc ASC01996.15: a847. 173. Levy R, Colonna P, Tourani JM, Gastaut JA, Brice P, Raphael M. Human immunodeficiency virus associated Hodgkin's disease: report of 45 cases from the French registry of HIV-associated tumors. Leuk Lymphoma 1995; 16:451-6. 174. Bellas C, Santon A, Manzanal A, et al. Pathological, immunological, and molecular features of Hodgkin's disease associated with HIV infection: comparison with ordinary Hodgkin's disease. AmJSurg Pathol 1996; 20: 1520-4. 175. Levine AM. HIV-associated Hodgkin's disease: biologic and clinical aspects. Hematol Oncol Clin North Am 1996; 10:1135-48. 176. Errante D, Zagonel V, Vaccher E, Serraino D, Bernardi D, Sorio R. Hodgkin's disease in patients with HIV infection and in the general population: comparison of clinicopathological features and survival.Ann Oncol 1994; 2: 37-40. 177. Gloghini A, Boiocchi M, De RV, etal. High incidence of monoclonal EBV episomes in Hodgkin's disease and anaplastic large-cell Ki-1-positive lymphomas in HIV-1positive patients. IntJ Cancer 1993; 54: 53-9. 178. Herndier BG, Sanchez HC, Chang KL, Chen YY, Weiss LM. High prevalence of Epstein-Barr virus in the Reed-Stern berg cells of HIV- associated Hodgkin's disease. Am J Pathol 1993; 142:1073-9.
179. Tirelli U, Errante D, Dolcetti R, Gloghini A, Serraino D, Vaccher E. Hodgkin's disease and human immunodeficiency virus infection: clinicopathologic and virologic features of 114 patients from the Italian Cooperative Group on AIDS and Tumors. J Clin Oncol 1995; 13:1758-67. 180. Leukaemia Research Fund Centre for Clinical Epidemiology. An Atlas of Leukaemia and Lymphoma. London: Leukaemia Research Fund, 1990: 81-9. 181. Freeman C, Berg JW, Cutler SJ. Occurrence and Prognosis of Extranodal Lymphomas. Cancer 1972; 29: 252-60. 182. d'Amore F, Christensen BE, Brincker H, et al. Clinicopathological features and prognostic factors in extranodal non-Hodgkin lymphomas. EurJ Cancer 1991; 27:1201-8. 183. Otter R, Gerrits WBJ, Sandt MMVD, et al. Primary extranodal and nodal non-Hodgkin's lymphoma. A survey of a population-based registry. EurJ Cancer Clin Oncol 1989; 25:1203-10. 184. Modan B, Shani M, Goldman B, et al. Nodal and extranodal malignant lymphoma in Israel: an epidemiological study. BrJ Haematol 1969; 16: 53-9. 185. Dobson LS, Hancock H, Bright N, Robinson MH, Hancock BW. Localised non-Hodgkin's lymphoma: the Sheffield Lymphoma Group experience (1970-1995). Int JOncol1998;13:1313-18. 186. The International Non-Hodgkin's Lymphoma Prognostic Factors Project. A predictive model for aggressive nonHodgkin's lymphoma. N EnglJ Med 1993; 329: 987-94. 187. Mounier N, Morel P, Haioun C, et al. fortheGroupe d'Etudes des Lymphomes de'Adulte: a multivariate analysis of the survival of patients with aggressive lymphoma. Cancer 1998; 82:1952-62. 188. Harris NL, Jaffe ES, Stein H, et al. A revised European-American classification of lymphoid neoplasms. A proposal from the International Lymphoma Study Group. Blood 1994; 84:1361-92. 189. Cowan RA, Jones M, Harris M, et al. Prognostic factors in high and intermediate grade non-Hodgkin's lymphoma. BrJ Cancer 1989; 59: 276-82. 190. Coiffier B, Gisselbrecht C, Vose JM, et al. Prognostic factors in aggressive malignant lymphomas. Description and validation of a prognostic index that could identify patients requiring a more intensive therapy. J Clin Oncol 1991; 9: 211-19. 191. Hancock BW, May K, Bruce L, et al. Haematological and immunological markers in malignant lymphoma. TumourDiagWSQ; 1:140-4. 192. Johnson PWM, Whelan J, Longhurst S, et al. b-2 Microglobulin: a prognostic factor in diffuse aggressive non-Hodgkin's lymhomas. BrJ Cancer 1993; 67: 792-7. 193. Salles G. Towards new prognostic factors in diffuse large cell non-Hodgkin's lymphoma. Ann Oncol 1996; 7: 993-6. 194. Cowan RA, Harris M, Jones M, et al. DNA content in high and intermediate grade non-Hodgkin's lymphoma. Prognostic significance and clinicopathological correlations. BrJ Cancer 1989; 60: 904-10.
204 Hodgkin's disease: clinical features 195. Akerman M, Brandt L, Johnson A, et al. Mitotic activity in non-Hodgkin's lymphoma. Relation to the Kiel classification and to prognosis. BrJ Cancer 1987; 55: 219-23. 196. Joensuu H, Ristamaki R, Sbderstrom KO, et al. Effect of treatment on the prognostic value of S-phase fraction in non-Hodgkin's lymphoma. J Clin Oncol 1994; 12: 2167-75. 197. Miller TP, Grogan TM, DahlbergS, et al. Prognostic significance of the Ki-67 associated proliferative antigen in aggressive non-Hodgkin's lymphomas: a prospective Southwest Oncology Group trial. Blood 1994; 83: 1460-6. 198. Jalkanen S, Joensuu H, Soderstrom KO, et al. Lymphocyte homing and clinical behaviour of nonHodgkin's lymphoma.y Clin Invest 1991; 87:1835-40. 199. Salles G, Zain M, Jiang WM, et al. Alternatively spliced CD44 transcripts in diffuse large-cell lymphomas: characterization and comparison with normal activated B-cells and epithelial malignancies. Blood 1993; 82: 3539-47. 200. Seymour JF, Talpaz M, Cabanillas F, et al. Serum interleukin-6 levels correlate with prognosis in diffuse large-cell lymphoma.J Clin Oncol 1995; 13: 575-82. 201. BlayJ-Y, Burdin N, Rousset F, et al. Serum interleukin-10 in non-Hodgkin's lymphoma: a prognostic factor. Blood 1993;82:2169-74. 202. Salles G, Bienvenu J, Bastion Y, et al. Elevated circulating alpha-TNF and soluble TNF-receptor levels
203. 204.
205.
206.
207.
208. 209.
are associated with adverse prognostic factors and poor outcome in lymphoma patients. BrJ Haematol 1996; 93: 352-9. Yuen AR, Sikic Bl. Multidrug resistance in lymphomas. J Clin Oncol 1994; 12: 2453-9. Hill ME, Maclennan KA, Cunningham DC, et al. Prognostic significance of BCL-2 expression and bcl-2 major breakpoint region rearrangement in diffuse large cell non-Hodgkin's lymphoma-a British National Lymphoma Investigation study. Blood 1996; 88: 1046-51. Boiocchi M, De RV, Bloomfield CD, Goldman A, Dick F, et al. Multivariate analysis of prognosis factors in the nonHodgkin's malignant lymphomas. Cancer 1974; 33: 870-9. Gospodarowicz MK, Bush RS, Brown TC, et al. Prognostic factors in nodular lymphomas: a multivariate analysis based on the Princess Margaret Hospital experience. Int J Radial Oncol Biol Phys 1984; 10: 489-97. Romaguera JE, McLaughlin P, North L, et al. Multivariate analysis of prognostic factors in stage IV follicular lowgrade lymphoma: a risk model.J Clin Oncol 1991; 9: 762. Coiffier B, Bastion Y, Berger F, et al. Prognostic factors in follicular lymphomas. Semin Oncol 1993; 20(suppl 5): 89. Lopez-Guillermo A, Montserrat E, Bosch F, et al. Applicability of the international index for aggressive lymphomas to patients with low-grade lymphoma.y Clin Oncol 1994; 12:1343-8.
16 Imaging of lymphoma K SANDRASEGARAN, PJ ROBINSON AND A SPRIGG
Introduction Imaging characteristics of nodes Chest Abdomen and pelvis Skeletal system Head and neck Central nervous system
205 205 206 207 210 211 213
INTRODUCTION Hodgkin's disease (HD) is primarily a nodal disease, with extranodal involvement being rare at presentation. Extension of nodal involvement is typically by a contiguous pattern, for example, supraclavicular to anterior mediastinal nodes. The Ann Arbor classification, which is described elsewhere, is used for staging disease extent and for grading prognosis. Non-Hodgkin's lymphoma (NHL) presents with extranodal disease in 20-40 per cent of cases. Nodal involvement is typically more bulky than in HD, with more widespread and non-contiguous disease being common at presentation. The Ann Arbor classification is less useful in NHL, with disease bulk, histology and biochemical markers being important prognostic factors. Differences between HD and NHL in pathology, clinical presentation and in the natural history of the disease are sufficiently great to require different imaging strategies during staging and follow-up. In this chapter, we will first discuss the imaging of nodal and extranodal disease in various sites. Specific requirements for the imaging of pediatric andacquired immunodeficiency syndrome (AIDS)-related lymphomas are described, and strategies are suggested for the selection and timing of follow-up imaging. IMAGING CHARACTERISTICS OF NODES The recognition of disease in nodes depends on their enlargement; involvement of normal-sized nodes cannot
214 214 216 216 217 218
Breast Lymphoma in patients with AIDS Lymphoma in children Imaging in the follow-up of lymphoma New imaging techniques in lymphoma References
be detected by computed tomography (CT), ultrasonography (US) or magnetic resonance imaging (MRI) at present. Discrimination between minor nodal enlargement by lymphoma, and reactive or inflammatory enlargement is also impossible by current imaging methods. Nodes are readily distinguished from surrounding fat on CT and appear as masses of uniform attenuation similar to that of muscle. Groups of nodes appear lobulated, and necrosis or calcification is rare unless there has been previous treatment. Nodes typically show little enhancement after contrast medium injection, although moderate enhancement may be seen in nodular sclerosing HD. They rarely cause invasion of surrounding tissue, but often displace or compress adjacent structures. There is considerable normal variation in the size of nodes. Table 16.1 indicates the size criteria for enlargement of nodes in various anatomic sites.1,2 Subcarinal nodes measuring 10 mm are usually normal Table 16.1 Maximum normal sizes of lymph nodes on computed tomography
Mediastinal and hilar Axillary Paracardiac Para-aortic Retrocrural Portocaval Mesenteric Iliac and inguinal
10 10 8 10 6 8 8 10
206 Imaging of lymphoma
unless there is also involvement of tracheobronchial nodes. In the abdomen the maximum size of normal nodes increases with distance from the diaphragm and decreases with age. Normal nodes cannot usually be differentiated from surrounding fat on US. Enlarged reactive and lymphomatous nodes are echopoor without distal enhancement. Malignant nodes tend to be more rounded but it is not usually possible to distinguish them from benign enlargement. On Tl-weighted magnetic resonance (MR) images, lymph nodes have low signal and contrast well against the high signal of surrounding fat. On Tl-weighted images, nodes have a homogeneously high signal. A mixed high and intermediate signal on T2-weighted images is occasionally seen in untreated nodular sclerosing HD, but is typical after treatment. Gallium localizes in viable tumor cells to a degree, which varies with histologic grading and with disease activity. Gallium uptake is highest in HD and high-grade NHL. Since gallium is taken up by normal liver and bone marrow, and is partly excreted via the bowel, interpretation of planar images of the abdomen is more difficult than in the chest. With optimum technique, gallium-67 scanning has been claimed to have greater than 90 per cent sensitivity, specificity and accuracy in detecting active disease in the thorax.3
CHEST Intrathoracic involvement at the time of initial presentation is found in 65-85 per cent of patients with HD and 25-40 per cent of patients with NHL.4,5 In HD, mediastinal lymphadenopathy is common and typically affects paratracheal and anterior mediastinal nodes (Fig. 16.1). Posterior mediastinal nodes are rarely involved. In the absence of mediastinal disease, hilar adenopathy is uncommon and involvement of the lung
Figure 16.1 CT scan at the level of the aortic arch showing enlarged axillary, anterior mediastinal and azygous nodes in Hodgkin's disease.
parenchyma rare in HD, unless there has been previous radiotherapy to the mediastinum. The pattern is less predictable in NHL and isolated pulmonary disease may occur. The thymus is considered to be a lymphatic organ and so there is little need to determine whether an anterior mediastinal mass is of thymic or nodal origin. Thymic enlargement is seen on CT in 30-50 per cent of patients with HD.6 The thymus remains enlarged in about a third of cases after treatment. Lung manifestations include direct extension from involved nodes, pulmonary nodules with or without cavitation, collapse from bronchial infiltration by tumor or obstruction by nodes and, rarely, interstitial infiltration. CT is the staging method of choice. When the above changes are seen in the untreated patient, pulmonary involvement can be diagnosed with confidence.7 In the treated patient, however, biopsy may be needed to differentiate pulmonary relapse from infection, radiation pneumonitis or drug-induced lung disease. Pleural effusions are seen in 7-10 per cent of untreated adult patients. They are not of prognostic importance, unless associated with a pleural mass, since they rarely contain malignant cells and usually resolve on treatment of adenopathy. Invasion of the chest wall may occur in NHL with involvement of internal mammary nodes and the thoracic spine may be involved by extension from posterior mediastinal nodes. Cardiophrenic nodes (Fig. 16.2) may produce diaphragmatic invasion (Fig. 16.3). Histologic evidence of pericardial invasion is rarely obtained but about 5 per cent of patients have pericardial effusions on CT at the time of presentation.7 Myocardial involvement is rare and has a poor prognosis.
Figure 16.2 CT scan at the level of the right hemidiaphragm showing enlarged nodes in the right cardiophrenic angle and adjacent to the apex of the heart. Histology: NHL
Abdomen and pelvis 207
value in initial diagnosis or staging. The role of positron emission tomography (PET) in staging lymphoma appears promising but still needs further investigation (see later).
ABDOMEN AND PELVIS
Figure 16.3 CT through the superior part of the liver showing a large non-enhancing mass of lymphoma arising in the left hemidiaphragm infiltrating the left lobe of the liver and other adjacent viscera. Histology: NHL
Staging of thoracic lymphoma In addition to initial chest X-ray, thoracic CT should be routinely performed in all lymphoma patients. Scans should be examined with lung windows as well as soft tissue windows. The routine use of intravenous contrast enhancement is not necessary in adults but may be useful in cases where there is difficulty in interpretation, e.g. in patients with aberrant mediastinal vessels or in the absence of intrathoracic fat. If the results of CT are doubtful or inconclusive, MRI may be helpful to improve the definition of chest wall or mediastinal disease (Fig. 16.4). Where cardiac involvement is suspected, echocardiography can be added to detect small pericardial effusions. Gallium-67 single photon emission computed tomography (SPECT) should detect the presence of active disease in patients with residual masses after treatment of mediastinal HD,8 but has little
Figure 16.4 ~[2-weighted MRI scan through mid-thorax showing a multinodular mass of fairly high signal intensity surrounding the great vessels (areas of signal void) and invading the right anterior chest wall. A large right pleural effusion is present. Histology: NHL
Lymphoma should be considered in the diagnosis of any unusual mass within the abdomen. Lymphoma can involve nodes in the retroperitoneum, mesentery, porta hepatis, splenic hilum, around the pancreas or the major vascular bundles. CT is the first-line imaging technique (Fig. 16.5). Where there are no superficial nodes available for excision biopsy, CT- or US-guided fine-needle aspiration is usually adequate to arrive at a diagnosis of NHL. In HD, a core specimen is normally required. The relatively rare condition of primary splenic lymphoma typically presents with single or multiple masses in the spleen.9 Much more frequently the spleen is involved concurrently with nodal lesions elsewhere; in such cases, splenic disease is typically diffuse with only a small minority of cases showing nodules larger than 1 cm in size. The problem of detecting secondary splenic involvement in lymphoma is still largely unsolved. Staging laparotomy has shown that, at the time of the presentation, the spleen is infiltrated in about 30-40 per cent of patients with Hodgkin's disease and in 10-40 per cent of those with NHL.10~12 In 10 per cent of patients presenting with thoracic HD, splenic involvement is the sole site of infradiaphragmatic disease13 so detection of splenic disease changes both staging and treatment. In patients with disease above and below the diaphragm, the additional detection of splenic disease does not often change staging but may indicate a different treatment plan. The size of the spleen is not much help, since diffuse infiltration may be present in spleens of normal size, while mild to moderate reactive splenomegaly occurs in
Figure 16.5 CT at the level of L1/2 showing multiple enlarged para-aortic and mesenteric nodes surrounding the inferior vena cava, aorta and left renal vein.
208 Imaging of lymphoma
about 30 per cent of patients with HD and up to 70 per cent of those with NHL; however, marked splenomegaly in NHL almost always indicates infiltration. Diffuse infiltration of the spleen cannot be reliably detected by ultrasound, by CT, or by unenhanced or gadoliniumenhanced MRI. Discrete nodules, when large enough to be visible, are hypoechoic on ultrasound and show low attenuation with reduced contrast enhancement compared to normal splenic tissue on CT (Fig. 16.6). On MRI, the lesions are hypointense or isointense on Tlweighted MRI images, hyperintense on 77-weighted images and show reduced enhancement after gadolinium compared to normal spleen (Fig. 16.7). The presence of enlarged nodes at the splenic hilum may be taken as an indicator of parenchymal involvement of the spleen. Sensitivity of ultrasound and CT in detecting splenic disease is about 35 per cent.14 CT following intravenous or intra-arterial iodized oil has been reported to improve the detection of diffuse hepatosplenic lymphoma15 but the side effects of the contrast agent have prohibited its widespread adoption. Superparamagnetic iron oxide particles are ingested by
Figure 16.6 Unenhanced CTscan through the upper abdomen showing multiple low-attenuation lesions in both liver and spleen. Note also the enlarged retrocrural lymph nodes. History: Hodgkin's disease.
Figure 16.7 Abdominal MRI in NHL (a) Transverse T2weighted scan through upper abdomen showing multiple lesions of increased signal intensity in liver and spleen, (b) Coronal STIR images showing one of the larger liver lesions with high signal and also multiple bone deposits showing areas of high signal affecting most of the vertebral bodies, (c) Transverse T2-weighted scan through the renal areas showing multiple splenic lesions of increased signal, together with numerous enlarged para-aortic nodes. (Courtesy of DrAlan Chalmers.)
Abdomen and pelvis 209
normal reticuloendothelial cells and produce a marked reduction in signal intensity on both Tl- and Unweighted MRI. Initial trials suggest that diffuse lymphomatous involvement of the liver or spleen might be detected even in organs of normal size.16 Further experience with this technique is needed. Primary hepatic lymphoma is a very rare tumor of middle-aged whites that is occasionally associated with cyclosporin treatment. It presents as a solitary mass involving Kupffer cells, sometimes with lymph nodes at the porta hepatis. Secondary hepatic lymphoma is associated with lymph node disease and at presentation is seen in 5-10 per cent of NHL patients and 10-20 per cent of HD patients.11,12,15 The liver is never involved in the absence of splenic disease (except possibly in some patients with AIDS), whereas the spleen is often involved without the liver being affected. As with splenic disease, involvement is diffuse in 90 per cent of cases17 and not reliably detectable on US, CT or MRI. Nodular disease in the liver looks similar to that in spleen (Fig. 16.6 and 16.7). Hepatomegaly in NHL is likely to signify liver involvement but is non-specific in HD. Disease of the bile ducts and gall bladder is rare but has been described in AIDS-related lymphoma. Parenchymal involvement of the pancreas occurs in 0.5-2 per cent of NHL patients12,18 and develops in association with disease in adjacent nodes. The incidence is higher in patients with high-grade histology, particularly Burkitt's lymphoma. Appearances on US, CT and MRI may be indistinguishable from those of pancreatic carcinoma except that dilatation of the pancreatic and biliary ducts is relatively uncommon. It may be difficult to differentiate peripancreatic lymphadenopathy from pancreatic disease. Percutaneous needle biopsy has been reported to be diagnostic in 95 per cent of cases.18 Renal lymphoma is usually an incidental finding on CT. It is often a late manifestation, being seen at initial staging in only 3-6 per cent of cases. Bilateral disease is rare initially but commonly develops in the late stages.19,20 Intravenous urography is frequently normal, whereas US usually shows single or multiple hypoechoic foci without distal acoustic enhancement.20 Occasionally echogenic nodules or multiple cystic lesions are seen. The typical appearance on CT is that of multiple intrarenal nodules with attenuation similar to or less than that of normal renal cortex. The nodules show less enhancement than normal renal tissue after intravenous contrast injection. Most cases are associated with lesions in the retroperitoneal nodes, liver or spleen. When renal lymphoma is suspected, contrast-enhanced CT is required for staging. When there is evidence of lymphoma at other sites, typical CT appearances may be taken to indicate renal involvement (therefore Stage IV disease) but, in the minority of cases presenting with primary renal lymphoma, a guided biopsy is required. Ultrasound may be used for following up those lesions that are clearly visible by this technique.
Adrenal involvement occurs in 1-4 per cent of patients with lymphoma21,22 and in the majority of cases is associated with retroperitoneal lymphadenopathy. The usual appearance is that of homogenous enlargement of one or both glands, which typically resolves with treatment. A solitary adrenal mass without retroperitoneal lymphadenopathy in a lymphoma patient does not necessarily signify adrenal infiltration since incidental non-functioning adenomas are relatively common, particularly in older patients. Bladder involvement is usually asymptomatic and is most often picked up on staging CT. The appearances are those of a large multinodular intramural mass or widespread thickening of the bladder wall. Most cases of prostatic lymphoma are intermediate or high-grade NHL. There is usually extensive involvement of the whole gland and adjacent tissue. NHL of prostate is typically associated with disease in bone, Waldeyer's ring, liver, mesenteric and inguinal nodes. The prognosis is very poor.22 Primary testicular lymphoma accounts for one-third of testicular tumors presenting in patients aged over 50 years. It is always due to intermediate- or high-grade NHL; bilateral disease is present in 10-25 per cent of cases.23 Primary imaging is by ultrasound, which usually shows well-defined homogeneous masses with reduced echogenicity in an enlarged testis. The appearances are characteristic given the age of patient. Staging is crucial since the 2-year survival of Stage I-II patients is 45 per cent compared with 1 per cent for Stage III-IV.24 Testicular lymphoma is typically associated with involvement of the central nervous system (CNS), Waldeyer's ring and the lungs. Staging may include sonography of both the testes, abdominal CT, chest X-ray, lumbar puncture and clinical examination with endoscopic biopsy of the nasopharynx.23 Ovarian lymphoma, which is typically solid homogeneous masses of NHL with low vascularity on ultrasound, usually presents late and has a poor prognosis, whereas lymphoma of the uterus more commonly has an earlier presentation with a 73 per cent 5-year survival for Stage I-II disease.22 Staging should include pelvic ultrasound and abdominopelvic CT. In gastrointestinal (GI) lymphomas, the clinical stage affects the prognosis more than histology or location.23,25 Regional lymph node involvement implies significantly worse prognosis. HD rarely involves the GI tract, whereas in NHL the GI tract is the commonest extranodal site of disease with involvement in 5-15 per cent of patients at the time of presentation.25 The stomach is affected in about half of these cases [including those arising in mucosa-associated lymphoid tissue (MALT) which have similar imaging appearances], the small bowel is affected in a third, and large bowel involvement accounts for the remainder. NHL is the cause of 2-5 per cent of gastric tumors.26 Radiologic appearances include enlarged gastric rugae, ulceration, aneurysmal dilatation
210 Imaging of lymphoma
Figure 16.8 CT through the upper abdomen showing extensive gastric wall thickening. Histology: NHL.
and multiple polyps. It may be impossible to differentiate from adenocarcinoma on CT, but lymphoma is suggested if there is gross thickening of the entire stomach wall, clear separation between stomach wall and adjacent organs, lymphadenopathy above and below the renal hilum, or extension across the pylorus into the duodenum (Fig. 16.8). Because the lesion infiltrates the submucous gastric wall, endoscopic biopsies may miss the pathology in up to 20 per cent of cases.26 Lymphoma is probably the commonest primary tumor of small bowel, accounting for about 20 per cent of cases. Typically it shows a constrictive, nodular or ulcerative appearance on barium studies, and associated mesenteric nodes may be shown by CT. Multiple lesions occur in about 10 per cent of cases.25 The rare primary lymphoma of large bowel usually affects either cecum or rectum. Secondary spread is associated with multiple lesions affecting any part of the colon. The incidence is highest in AIDS patients and in those with poorly differentiated NHL.
Staging of abdominal lymphoma CT is usually the initial mode of investigation. Where only lymph node disease is expected, oral contrast is used and there is no necessity for intravenous contrast. Contrast-enhanced images are useful if extranodal involvement is suspected. Lymphangiography (LAG) and gallium-67 scanning probably make no additional contribution to the staging of abdominal NHL. In early HD, LAG has historically been thought to be superior to CT in staging para-aortic nodes27 but other authors have recently found no advantage.28 MRI can detect retroperitoneal and pelvic nodal disease as sensitively as CT. Mesenteric nodes are not well seen due to the mobility of the bowel. GI lymphoma is usually diagnosed on endoscopy or barium studies, but CT is also required for staging. In patients presenting with lymphoma elsewhere, barium
studies are not routinely required to look for GI involvement, except in those with NHL of Waldeyer's ring, who have a 6-15 per cent probability of asymptomatic GI disease.29 The use of laparotomy has declined considerably in the last decade. It is inappropriate in patients with NHL and Stage III/IV HD. HD patients with Stage I or II and features suggesting a poor prognosis, e.g. bulky tumors or B symptoms, now receive chemotherapy without laparotomy. Those with very good prognosis, i.e. single site, small tumor load and no symptoms, will probably be cured by radiotherapy alone without laparotomy, and even if they relapse will be salvaged by chemotherapy. Consequently, staging laparotomy is reserved for an ever-decreasing group of HD patients with intermediate prognosis. In the absence of axillary or inguinal nodes that can easily be biopsied surgically, CT- and US-guided core biopsies performed with spring-loaded firing devices have an important role in the diagnosis and staging of lymphoma. The role of fine-needle aspiration biopsy (FNAB) in this situation is controversial. The development of ancillary diagnostic techniques, such as immunologic markers and cytogenetics, has increased the specificity and accuracy of FNAB in diagnosis of high-grade NHL. Accuracies of up to 90 per cent have been quoted.30,31 However, the exclusion value of a negative FNAB remains low in patients with HD and intermediate-grade lymphoma.32 FNAB, however, has a useful role in the assessment of residual masses following treatment (see later).
SKELETAL SYSTEM It is important to differentiate between bone marrow and osseous involvement. At presentation, bone marrow involvement is rare in HD but is found in 20-40 per cent of NHL patients;11,12,33 this indicates Stage IV disease and is associated with worse prognosis than involvement of liver, lung or osseous bone. Bone marrow aspirates are relatively late indicators of marrow involvement. In a large study of NHL patients, marrow biopsy increased the staging in 32 per cent of cases, mostly from Stage II to Stage IV.33 The performance of bilateral bone marrow biopsies increases the pick-up rate of positive cases by 10 per cent to 40 per cent compared with single site biopsies.34 Bone marrow involvement in low-grade NHL is typically diffuse but in intermediate- to high-grade NHL and in HD marrow disease is likely to be focal. Bone scintigraphy and CT are not accurate in assessing marrow disease. On MRI, tumor deposits have a low signal on Tl -weighted images and a high signal on T2weighted sequences. Tumor deposits down to 3 mm in size can be detected as having a high signal on short-tau inversion recovery (STIR) sequences.
Head and neck 211
Osseous involvement occurs in 1-2 per cent at presentation and 5-15 per cent during the course of the disease35 in adult lymphoma patients. The incidence is much higher in children. Primary bone lymphoma (also termed reticulum cell sarcoma or histiocytic lymphoma) is another rare form of NHL. By definition there is no lymphadenopathy and a single bone, typically femur or pelvis, is involved, so the lesion is automatically classified as Stage I. Secondary bone lymphoma is seen in NHL and HD, and tends to involve the axial skeleton, especially the spine. Spread may be hematogenous (Stage IV) or from adjacent lymphadenopathy. In either case the prognosis is relatively favorable; the survival with hematogenous bone involvement is 50 per cent at 10 years. Abnormal biochemistry is a poor predictor of bone involvement. Plain radiographs may show a permeative or moth-eaten appearance. Periosteal reaction is more common in HD. Pelvic bone lesions often have soft tissue extension. Sclerotic lesions are more frequent in HD and following treatment. Primary bone lesions usually show reduced signal intensity on Tl -weighted MRI but their appearance on T2-weighted imaging is variably isointense or hypointense.36'37.
Staging of lymphoma of osseous bone and marrow Skeletal radiographs need only be performed in those with bone or joint pain, and in areas of bone disease suspected by CT, chest X-ray or bone scintigraphy. Bone scintigraphy has a sensitivity and accuracy of 95 per cent in detecting bone involvement35 but it is not routinely indicated, since most cases of bone lymphoma are revealed on initial chest X-ray and CT. It is used for patients who have localized skeletal symptoms and those with known bone involvement for staging and followup. CT or MRI is valuable in assessing soft-tissue extension of lymphoma. Currently, iliac crest biopsy to detect marrow disease is routine, although PET using 2-[18F]fluoro-2-deoxy-Dglucose (FDG) may provide a less invasive alternative because it samples the whole skeleton and is probably at least as accurate.38,39 A case can also be made for the use of MRI (Tl-weighted and STIR sequences) of the pelvis, spine and proximal femora in patients with negative marrow biopsies who would otherwise be candidates for radiotherapy, e.g. Stage I high-grade NHL. In such patients a positive MR study may be followed by appropriately sited biopsy.
per cent of all HD cases, including those without throat symptoms or radiological evidence of disease, have positive nasopharyngeal biopsies and some authors advocate nasopharyngeal biopsy as part of routine staging of HD.40 About a third of patients with clinical Stage I or II disease will have infradiaphragmatic disease, so abdominal CT is mandatory. Abdominal disease is particularly likely in those with bulky supraclavicular lymphadenopathy or bilateral neck disease, and in those with mixed cellularity on histology. With NHL, the head and neck is a primary site for involvement in about 10 per cent of patients and about half of these will also have systemic involvement.41,42 NHL is limited to nodes in a third of head and neck cases, with deep lymphatic chains being the most common sites. Extranodal tissue such as Waldeyer's ring and paranasal sinuses account for the majority of head and neck NHL. The nasopharyngeal adenoids together with oropharyngeal, faucal and lingual tonsils make up the Waldeyer's ring of extranodal lymphatic tissue. It is the commonest site of NHL in the head and neck region (Fig. 16.10). In 6-15 per cent of cases, there is coexistent disease of the GI tract (most often stomach) and barium studies are indicated even in the absence of symptoms.43 On CT, head and neck lymphoma typically is homogeneous, has similar attenuation to muscle and shows little enhancement. On MRI it is of low signal on Tl-weighted images and intermediate signal on T2-weighted images. There is moderate enhancement with gadolinium.
HEAD AND NECK HD of the head and neck is typically limited to lymph nodes; extranodal disease is manifest clinically in less than 1 per cent of cases (Fig. 16.9). However, about 20
Figure 16.9
CT scan through the level of the hyoid showing a
lymph node mass compressing the right submandibular gland. Histology: Hodgkin's disease.
212 Imaging of lymphoma
Figure 16.10 C7 scan through the level of the angle of the mandible showing an ill-defined soft tissue mass extending into the oropharynxfrom the right tonsillar fossa. Histology: NHL
Tonsillar lymphoma is frequently associated with widespread nodal disease and prognosis is consequently poor. Recurrence after treatment usually occurs within a year and tends to involve bone marrow, the GI tract or the CNS. After squamous cell carcinoma, NHL is the second
Figure 16.11 Transverse Tl-weighted (a) and T2-weighted (b) MRI scans through the orbits showing an extensive tumor mass with moderately high signal intensity on both T7 and T2 scans enveloping the anterior and lateral aspects of the left globe together with the lacrimal gland. Histology: NHL. (Courtesy of Dr Ashley Guthrie.)
most common tumor of the paranasal sinuses. Extension into the infratemporal fossa and cheek may occur without visible bone destruction. Clinically and radiologically lymphoma can mimic squamous carcinoma, although bone destruction is usually not as extensive as in carcinoma, and lymphadenopathy tends to be bilateral, without calcification or necrosis. Lymphoma is also suggested if nodes are large and in atypical sites, such as retropharyngeal, submental, submandibular or posterior triangle chains. Occasionally it may be difficult to differentiate tumor mass from retained secretions on CT and, in these cases, T2- weigh ted MRI or Tl- weigh ted sequence with gadolinium should allow discrimination. Primary orbital NHL is the commonest orbital malignancy in adults and is bilateral in up to 40 per cent of cases.42 Most cases are associated with systemic disease, which may present up to 5 years later.41 NHL may resemble reactive lymphoid hyperplasia, but homogeneity and bone destruction are indicative of the former, while scleral thickening and infiltration of retro-orbital fat are suggestive of the latter.44 Either CT or MRI can be used to demonstrate disease in the retrobulbar space, around the globe, optic nerves and muscles, and within the lacrimal glands (Fig. 16.11 and 16.12). The examination protocol should include the neck to evaluate cervical nodal disease and, if CT is used, images with bone windows should be included. Intraocular involvement is rare and is almost always associated with CNS involvement. NHL of salivary gland is rare and, although it is usually part of widespread disease, the prognosis is good.43 The parotid gland is the most frequently affected. The tumor can usually be distinguished from adjacent parotid tissue and fat on CT or MRI; the imaging features are similar to those of lymphoma elsewhere in the head and neck region. Single lymphomatous masses can-
Central nervous system 213
studies of the upper GI tract and small bowel. In patients with paranasal sinus lymphoma, ideally both CT and MRI with axial and coronal sections as well as intravenous contrast enhancement are necessary to delineate extension of the tumor. CT shows bone destruction better, while MRI is superior in delineating soft tissue and intracranial extension. These patients also have a high risk of associated meningeal disease so cerebrospinal fluid (CSF) sampling may be helpful.
CENTRAL NERVOUS SYSTEM
Figure 16.12 CT through orbits showing obliteration of the contents of the left orbit by infiltrating tumor. Histology: NHL Note the spread of the tumor into the infratemporal fossa on the left and previous surgery to right globe. (Courtesy of Dr Ashley Guthrie.)
not be distinguished from the more common pleomorphic adenoma. Multiple NHL masses may look similar to Warthin's tumor or metastasis, but bilaterality and the presence of periparotid lymphadenopathy are suggestive of lymphoma. Thyroid NHL occurs more commonly in females, is associated with chronic thyroiditis and is bilateral in half the cases. Ultrasound is sensitive but not specific for diagnosing NHL. A variety of patterns have been reported, including discrete hyperechoic nodules, multiple lesions of complex echogenicity, diffusely hypoechoic goitre or even thyroid enlargement with normal echopattern.45 On CT, the typical appearance is a low attenuation nodular mass with extracapsular spread and adjacent nodal involvement. Unlike anaplastic carcinoma, necrosis and calcification are rare. In the absence of nodal involvement, the prognosis is good (80 per cent - 5-year survival) but it is much worse with nodal disease (30 per cent - 5-year survival).46
Lymphomas account for about 2 per cent of primary brain tumors. Brain involvement occurs in 10-15 per cent of NHL47 at some time but, of those presenting initially with CNS involvement, the vast majority (over 90 per cent) are primary lymphoma.41 Brain involvement in HD is so rare that a space-occupying lesion in the brain of a patient with known HD should suggest a second diagnosis. Primary brain lymphoma has been previously termed reticulum cell sarcoma, microglioma and round cell sarcoma. Most cases present as solitary masses, which, unlike gliomas, tend to affect central sites, particularly the periventricular areas, basal ganglia, corpus callosum and thalamus. With disease progression, 30 per cent of patients will develop multifocal involvement and about 10 per cent will develop diffuse parenchymal infiltration. On CT, about two-thirds of the lesions are hyperdense and less than 10 per cent are hypodense.48 Surrounding edema and central necrosis are much less commonly seen than with gliomas. Homogeneous enhancement with contrast is typical, although patchy or peripheral enhancement may occur. Non-enhancement and the presence of edema are associated with a poorer prognosis. On Tl -weighted MR images, parenchymal lesion are isointense or hypointense compared with gray matter and show enhancement with gadolinium. The lesions are hyperintense or isointense to gray matter on T2weighted images, which also show the presence of edema. Leptomeningeal spread is seen in both NHL and HD (Fig. 16.13).
Staging of head and neck NHL Staging of CNS lymphoma Where extranodal disease is suspected, CT is required for anatomic definition and radiotherapy planning. All patients with head and neck NHL should have abdominal CT because of the high prevalence of coexisting infradiaphragmatic disease. If CT is equivocal, e.g. with dental artefacts and in the difficult supraclavicular region, MRI may be helpful; we would recommend coronal STIR and transverse T2 -weighted sequences. In Stage I and Stage II disease, bone marrow biopsy should be carried out since these procedures result in upstaging in about 15 per cent of cases. Patients with Waldeyer's ring involvement, as already mentioned, should have barium
CT and MR are equally effective in detecting primary brain lymphoma.48 Meningeal deposits are best seen on gadolinium-enhanced Tl-weighted MRI, which is considerably more sensitive than contrast-enhanced CT. Coronal sections are recommended. Lymphoma of the spinal cord and meninges is rare, and both CT and MRI are recommended in staging. Tumor in the spinal cord, meninges, paraspinal soft tissue and vessels is well shown by MRI (Fig. 16.14). CT is superior in showing the extent of cortical bone destruction. In the absence of neurological symptoms, CNS
214 Imaging of lymphoma
2 those who are immunocompromised, especially with AIDS, in whom the risk of CNS involvement is increased by a factor of 500; 3 those with diffuse NHL, and with involvement of testes or bone marrow, who have a 25 per cent risk of developing CNS involvement.49
BREAST
Figure 16.13 Transverse T2-weighted MRI scan through mid-brain and posterior fossa, showing multiple areas of high signal intensity on the surface of the cerebellar hemispheres. Histology: NHL (Courtesy of Dr Keith Blanshard.)
Primary lymphoma accounts for 0.1-0.5 per cent of breast tumors with a disproportionate (6 per cent) of cases occurring in males and on the right side.50 The mammographic appearance is usually that of a welldefined mass (Fig. 16.15) with minimal irregularity of contour and no microcalcification. Fibrous reaction distorting the breast architecture and nipple retraction are rare. Diffuse or localized skin thickening may occur due to lymphatic obstruction. Occasionally there is diffuse increase in breast density. Ipsilateral axillary lymphadenopathy is seen in 35-40 per cent of cases and is associated with poor prognosis. The mammographic appearances overlap with those of carcinoma but the presence of bilateral lymphadenopathy is suggestive of lymphoma. Staging procedures for breast lymphoma should include chest radiograph and thoracoabdominal CT. The roles of ultrasound and MRI are not yet clear.
LYMPHOMA IN PATIENTS WITH AIDS screening is not required in routine staging of lymphoma, except in the following patients, who should have lumbar puncture and, if possible, MRI of brain and spinal cord: 1 those with very high grade histology, i.e. lymphoblastic lymphoma and small cell non-cleaved (Burkitt's and non-Burkitt's) lymphoma;
Figure 16.14 Transverse T2-weighted MRI scan through the mid-thoracic spine showing intradural, but extra medullary infiltration by tumor spreading from the paraspinal muscles on the left and also extending into the left subpleural space. Histology: NHL. (Courtesy of DrAlan Chalmers.)
AIDS-related lymphomas have several features that are distinct from lymphoma in other patients. Peripheral lymphadenopathy is relatively uncommon and the diagnosis of lymphoma often depends on guided biopsy of liver or para-aortic nodes.
Figure 16.15 CT through the mid-thorax showing a large homogenous tumor in the left breast. Histology: NHL.
Lymphoma in patients with AIDS 215
Hodgkin's disease in AIDS patients HD is increasingly recognized as a complication of human immunodeficiency virus (HIV) infection although at present it is not one of the major diagnostic criteria for defining a diagnosis of AIDS in HIV-positive patients. The distinctive features of HD in AIDS patients include the following:51-53 1 most patients present at Stage III or Stage IV; 2 there is a high incidence of bone marrow involvement even in the absence of splenic disease; 3 mesenteric nodes are involved in 20 per cent of cases; 4 nodal disease is often non-contiguous, e.g. pelvic and mediastinal lymphadenopathy occur without paraaortic disease.
Non-Hodgkin's lymphoma in AIDS patients Like HD, NHL in AIDS patients shows many unique features:54-56 1 the signs and symptoms are non-specific, e.g. CNS lymphoma often presents with altered personality, which is also a feature of other CNS complications of AIDS; 2 peripheral lymphadenopathy is absent in half of cases; 3 high-grade B cell type histology occurs in over 50 per cent of cases compared with less than 5 per cent in the general population; 4 involvement of multiple extranodal sites occurs in 75-95 per cent of cases; 5 primary brain lymphoma accounts for about 25 per cent of AIDS-related lymphoma (ARL), whereas only 2 per cent of non-AIDS NHL presents in this way. 6 median survival is very poor, being less than 6 months compared to 40 per cent 5-year survival of Stage III and IV NHL in the general population.
Imaging features of ARL Mediastinal and hilar node lesions are less bulky and less commonly found than in non-AIDS patients. Mediastinal nodes measuring 5-10 mm in diameter must be regarded with suspicion and larger intrathoracic lymph nodes are more likely to be caused by fungal or mycobacterial infection than lymphoma. Fineneedle aspiration biopsy with the appropriate immunochemistry is usually adequate for diagnosis in NHL, whereas in HD, because of the importance of structural elements in the histology, a cutting needle sample or open biopsy is better. CNS lymphoma in AIDS tends to be multifocal with lesions smaller than 2 cm in size. Subarachnoid invasion is seen in a quarter of the cases48 with epidural involvement in 6 per cent. With larger lesions, central necrosis is
common, so a rim enhancement pattern is seen on CT or gadolinium-enhanced MRI. Neither technique can differentiate this appearance from infective complications, such as toxoplasmosis and biopsy may be required for firm diagnosis. However, recent radionuclide SPECT studies using either thallium-201 or technetium-99mlabeled methoxyisobutyl nitrile (MIBI) have made this distinction by showing focal uptake in areas of tumor, whereas infective lesions showed little or no concentration of the tracer.57,58 MRI is more sensitive than CT in detecting parenchymal and leptomeningeal disease, and is the imaging method of choice. AIDS-related NHL occasionally presents with multiple discrete intramedullary spinal masses. About a quarter of AIDS-related lymphomas include bowel lesions,55 typically the mouth, rectum and terminal ileum. The barium findings are similar to those of lymphoma in non-AIDS patients. CT is necessary to demonstrate the full extent of bowel wall thickening and to detect mesenteric disease. It may be difficult to differentiate rectal lymphoma from inflammatory perirectal disease seen in homosexual men. Occasionally lymphoma may present as diffuse peritoneal nodularity without visible lymph node enlargement, mimicking carcinomatosis. Abdominal and pelvic lymphadenopathy in AIDS may be due to several causes. AIDS-related complex (a syndrome seen in HIV-positive patients without overt neoplasia) may cause generalized enlargement of nodes up to 1.5 cm. Nodes larger than this are abnormal, the likely causes being lymphoma, Kaposi's sarcoma, or Mycobacterium avium intracellulare (MAI) infection. Kaposi's sarcoma generally spares mesenteric nodes and causes less bulky lymph node enlargement. MAI typically produces bulky nodes with central low density due to necrosis. Ultrasound or CT-guided biopsy is recommended in order to distinguish these conditions from NHL. Focal lesions in the liver are more common (25-50 per cent) than in lymphomas in other patients (5-10 per cent).51,53 Appearances on CT and ultrasound are similar to those of lymphoma nodules in non-AIDS patients. Ultrasound and CT are both fairly insensitive in detecting diffuse liver and spleen involvement, and needle biopsy is usually required.51 The incidence of renal involvement is about 7-11 per cent.52 The typical appearance is that of multiple focal lesions, which are of low attenuation on CT and are hypoechoic on ultrasound. Diffuse enlargement of both kidneys in AIDS is more likely to be due to AIDS-related glomerular disease, which is a focal and segmental glomerulosclerosis manifest by renal failure and proteinuria.
Staging of ARL Lymphoma affects the chest much less frequently than infections, which may mimic it. Definitive diagnosis
216 Imaging of lymphoma
requires bronchoscopic biopsy or lavage in cases of diffuse disease and percutaneous biopsy with localized disease. CT may be useful in assessing mediastinal lymphadenopathy prior to a transcarinal needle biopsy. Since abdominal node disease tends to be bulky, there is a case for using ultrasound, which is comparable with CT in detecting liver lesions, as the initial screening investigation. If ultrasound is negative, or where involvement of bowel or other extranodal organs (e.g. psoas muscle) is suspected, CT is required. Lumbar puncture is required in all ARL cases. With positive CSF assay or clinical signs of CNS involvement, MRI should be carried out because its sensitivity is greater than that of CT in detecting brain and meningeal lymphoma. Bone marrow aspiration is required routinely owing to the high incidence of marrow lesions.
ographs are required prior to anesthesia, since large anterior mediastinal masses may compress the trachea in a sedated child lying supine. A nasogastric tube may be used to give sedation and oral contrast. At the end of the scan, residual contrast is aspirated from the stomach. We routinely use a clip-on transcutaneous oxygen saturation monitor (pulse oximeter) on sedated children. Mediastinal sonography is useful for follow-up of lymph node masses in the mediastinum or abdomen and planning biopsy. Bone marrow biopsies, preferably following MRI of the pelvis and femurs, are routine. CNS staging is usually performed by assaying CSF. Cranial MRI or CT, with contrast, are reserved for those with appropriate symptoms.
IMAGING IN THE FOLLOW-UP OF LYMPHOMA LYMPHOMA IN CHILDREN The principles of diagnosis and staging of lymphoma in the pediatric population are the same as those for adults. Hodgkin's disease accounts for 40 per cent of childhood lymphomas and has a peak incidence in teenagers. Childhood NHLs tend to be of high-grade histology. T cell histology is found in 30 per cent of pediatric NHL and presents with thoracic disease. Mediastinal lymphadenopathy may be large enough to cause dysphagia or superior vena caval obstruction. On the other hand, B cell NHL presents with extranodal disease in the abdomen or head and neck region. Burkitt's lymphoma is an unusual form of high-grade B cell NHL that occurs in an epidemic form in tropical Africa and New Guinea. Involvement of the maxilla or mandible is found in 60 per cent of these patients and the abdomen is affected in a similar proportion. American or European Burkitt's lymphoma is sporadic and is less strongly associated with the Epstein-Barr virus. In pediatric B cell NHL gastrointestinal disease tends to occur in the ileocecal region and intussusception may be the mode of presentation. Disease of the pancreas, kidneys, mesentery, ovaries and thyroid glands are common. Ascites and occasionally pleural effusion are present. Lymphadenopathy in the thorax or abdomen is unusual and, when present, is confined to the iliac and inguinal regions. There is an increased frequency of involvement of the CNS (20 per cent) and bone marrow (20 per cent). Staging pediatric lymphoma
In all cases CT of the chest and abdomen is performed as a baseline study. Intravenous contrast is routinely used. Good bowel opacification is also vital. Those under 2 years of age usually need sedation. This is best done in the presence of an anesthetist during a dedicated pediatric scanning session. Frontal and lateral chest radi-
Assessment of response to treatment In the phase of active treatment, imaging is used as a guide to therapeutic response. This is particularly true with cyclical chemotherapy, which should be discontinued or changed if the patient fails to show objective evidence of response. If there is visible intrathoracic disease, chest X-ray should be repeated at each clinic visit, which typically is monthly. CT may be repeated after each two cycles of chemotherapy, which usually means a gap of 6-8 weeks between scans. In our institution, chest, abdominal and pelvic scans, without intravenous contrast enhancement, are performed in all follow-up patients referred for CT, although there may be a case for a study limited to the regions showing disease on initial staging in NHL patients treated with chemotherapy.59 Also, a less detailed and frequent follow-up has been advocated for patients with HD (see Chapter 15).
Assessment of a residual mass Imaging is also used to confirm remission. The definition of complete remission requires that all previously abnormal investigations, including imaging studies, are repeated and found to be normal. In practice, residual mediastinal masses are common in HD or T cell NHL. Continuing regression of a mass with treatment implies persistence of active disease. Once the mass has reached a stable minimum size, further investigation may be necessary if there is clinical doubt about its activity (Fig. 16.16). Serial CT scans are the usual method of assessing the status of a residual mass and are performed every 2-3 months. Up to 15 per cent of residual masses may show a change in attenuation but an increase in size is highly suspicious of relapse. Masses that remain static after 1 year's follow-up can be considered inactive residue. Using gallium-67 imaging with SPECT, sensitivity and
New imaging techniques in lymphoma 217
Figure 16.16 Resolution of lymph node disease with treatment. Unenhanced CTscans through upper abdomen before (a) and after (b) treatment for abdominal Hodgkin's disease.
specificity of over 80 per cent has been quoted in assessing the activity of residual mediastinal masses.60 In HD, 67 Ga is fairly sensitive in detecting active disease residues after treatment but a negative result is less reliable in defining remission.61 Some histological subtypes of NHL and many necrotic tumors do not take up gallium and the accuracy of this method is lower for abdominal disease. On MRJ carried out 6-12 weeks after treatment, a reduction in the signal on a T2-weighted sequence usually means replacement of tumor by mature fibrosis. However, persistence of high signal is not specific to tumor and can be caused by inflammatory edema, cyst formation or partial volume averaging with fat. A change in signal from low to high on serial T2-weighted MRI is highly suggestive of relapse. There may be a role for MRI in predicting the size of a residual mass by measurement of T2-weighted signal intensities on the pre- and posttreatment scans.62 Fine-needle aspiration biopsy can be performed if there is clinical or radiological doubt of the activity of the residual mass. The accuracy of FNAB in the followup situation is good, and complications are rare even when biopsies of deep mediastinal structures are performed. If FNAB is negative, when there is strong clinical suspicion of relapse, a cutting needle or surgical biopsy is indicated. Detection of late relapse
The third function of follow-up is to detect relapse, which occurs in 10-40 per cent of patients with HD and in up to 60 per cent of NHL patients. Over 85 per cent of HD patients who relapse do so within 3 years of treatment.63 The timing of radiological reassessment depends on the clinical context. If a residual mass of uncertain significance exists, it should be assessed as above. If
remission has been complete, then 6-monthly CT scans during the first 3 years are sufficient. There are no clear rules for the length of long-term follow-up. In most institutions follow-up spans at least 10 years. A shorter follow-up may be reasonable for high-grade NHL, which can be considered to be cured if the patient survives for 5 years. Most late relapses are diagnosed clinically or with simple tests such as blood counts and chest X-rays, so after 3 years of treatment expensive imaging techniques, such as CT scans, should only be performed if there is clinical suspicion of relapse.
NEW IMAGING TECHNIQUES IN LYMPHOMA
Positron emission tomography Several studies have shown that PET with FDG is useful in staging64-66 and predicting response to therapy67 in lymphoma. PET has been shown to be at least as sensitive as CT in detecting disease in lymph nodes and extranodal sites.65-66-68 PET may also be useful in predicting the behavior of a residual mass after treatment.65,69 Further studies have suggested that it may be possible to differentiate cerebral lymphoma from infective lesions in HIV-positive patients.70 FDG uptake is related to histological grading, so PET may be negative in low-grade NHL71 and MALT lymphomas,'68 whilst false-positive results have been found in cases of rebound thymic hyperplasia,72 infection, fractures and other metabolically active benign lesions. PET using 11C-labeled methionine has also been found to be sensitive for detecting lymphoma lesions but less clearly related to histological grading and prognosis than FDG.73
218 Imaging of lymphoma
Somatostatin receptor scintigraphy
12. Chabner BA, Johnson RE, Young RC, et al. Sequential nonsurgical and surgical staging of non-Hodgkin's
The accuracy of somatostatin receptor scintigraphy with 11 'In-labeled octreotide (DTPA-pentetriotide) in staging and restaging of lymphoma has been tested in several trials.74-76 In general, the sensitivity is better for head and neck or thoracic tumors than for disease in the abdomen.74-75 This would in part explain the greater sensitivity of this technique in detecting the tumor site in Hodgkin's disease compared to non-Hodgkin's lymphoma.76 The place of this method in staging of lymphoma remains to be confirmed. Immunoscintigraphy with radiolabeled antibodies is undergoing early trials and results of controlled studies are awaited.
lymphoma. Ann Intern Med 1976; 85:149-54. 13. Reznek H, Richards MA. The radiology of lymphoma. Clin Haematol 1987; 1:77-107. 14. Castellino RA, Hoppe RT, Blank N, et al. Computed tomography, lymphography and staging laparotomy: correlation in initial staging of Hodgkin's disease. Am J Roentgenol 1984; 143: 37-41. 15. Thomas JL, Bernadino ME, Vermess M, et al. EOE-13 in the detection of hepatosplenic lymphoma. Radiology 1982; 145: 629-34. 16. Weisslander R, Elizondo G, Stark DD, et al. The diagnosis of splenic lymphoma by MR imaging: value of superparamagnetic iron oxide. AmJ Roentgenol 1989; 152: 175-80. 17. Zornoza J, Dood GD. Lymphoma of the gastrointestinal
REFERENCES
tract. Semin Roentgenol 1980; 15: 272-87. 18. Webb TH, Lillemoe KD, Pitt HA, Jones RJ, Cameron JL. Pancreatic lymphoma. Is surgery mandatory for
1. Glazer GM, Gross BH, Quint LE, Francis IR, Bookstein FL, Oringer MB. Normal mediastinal lymph nodes: number and size according to American Thoracic Society mapping. Am J Roentgenol 1985; 144: 261-5. 2. Dorfman RE, Alpern MB, Gross BH, Sandier MA. Upper abdominal lymph nodes: criteria for normal size determined with CT. Radiology 1991; 180: 319-22. 3. Mclaughlin AF, Magee MA, Greenough R, et al. Current role of gallium scanning in the management of lymphoma. EurJ Nud Med 1990; 16: 755-71. 4. Filly R, Blank N, Castellino RA. Radiographic distribution of intrathoracic disease in previously untreated patients with Hodgkin's disease and non-Hodgkin's lymphoma. Radiology 1976; 120: 277-81. 5. Castellino RA, Blank N, Hoppe RT, Cho C. Hodgkin's disease: contribution of chest CT in the initial staging evaluation. Radiology 1986; 160: 603-5. 6. Heron CW, Husband JE, Williams MR Hodgkin's disease: CT of the thymus. Radiology 1988; 167: 647-51. 7. Blank N, Castellino RA. The intrathoracic manifestations of the malignant lymphomas and leukaemias. Semin Roentgenol 1980; 15: 227-46. 8. Kostakoglu L, Yeh SD, Portlock C, et al. Validation of gallium-67 citrate SPECT in biopsy confirmed residual Hodgkin's disease in the mediastinum. J Nud Med 1992; 33:345-50. 9. Dachman AH, BuckJL, Krishnan J, Aguilera NS, Buetow PC. Primary non-Hodgkin's splenic lymphoma. Clin Radiol 1998; 53:137-42. 10. Kadin ME, Glatstein EJ, Dorfman RE. Clinicopathologic studies in 117 untreated patients subject to laparotomy for the staging of Hodgkin's disease. Cancer 1977; 27: 1277-94. 11. Castellino RA, Goffinet DR, Blank N, Parker BR, Kaplan HS. The role of radiography in the staging of nonHodgkin's lymphoma with laparotomy correlation. Radiology 1974; 110: 329-38.
diagnosis and treatment? Ann Surg 1989; 209: 25-30. 19. Heiken JP, Gold RP, Schnur MJ, King DL, Bashist B, Glazer HS. Computed tomography of renal lymphoma with ultrasound correlation. J ComputAssist Tomog 1983; 7:245-50. 20. Horii SC, Bosniak MA, Megibow AJ, Raghavendra BN, Subramanium BR, Rothberg M. Correlation of CT and ultrasound in the evaluation of renal lymphoma. Urol Radiol 1983; 5: 69-76. 21. Moulton JS, Moulton JS. CT of adrenal glands. Semin Roentgenol 1988; 23: 288-303. 22. Charnsangavej C. Lymphoma of the genitourinary tract. Radiol Clinics North Am 1990; 28: 865-77. 23. Bragg DG, Colby TH, Ward JH. New concepts in the non-Hodgkin's lymphomas: radiologic implications. Radiology 1986; 159: 289-304. 24. Duncan PR, Gowing NFC, McElwain TJ, Peckham MJ. Extranodal non-Hodgkin's lymphoma presenting in the testicle: a clinical and pathological study of 24 cases. Cancer 1980; 45:1578-84. 25. Lewin KL, Ranchod M, Dorfman RF. Lymphomas of the gastrointestinal tract: a study of 117 cases presenting with gastrointestinal disease. Cancer 1978; 42: 693-707. 26. Dragosick SB, Bauer P, Radaszkiewicz TUS. Primary gastrointestinal non-Hodgkin's lymphomas: a retrospective clinico-pathologic study of 150 cases. Cancer 1984; 152: 291-6. 27. Castellino RA, Hoppe RT, Blank N, et al. Computed tomography, lymphography and staging laparotomy: correlation in initial staging of Hodgkin's disease. AmJ Roentgenol 1984; 143: 37-41. 28. Stomper PC, Cholewinski SP, ParkJ, Bakshi S, Barcos MP. Abdominal staging of thoracic Hodgkin's disease: CTlymphangiography-Ga-67 scanning correlation. Radiology 1993; 187: 381-6. 29. Saul SC, Kapaida SB. Primary lymphoma of Waldeyer's ring. Clinicopathologic study of 68 cases. Cancer 1985; 56:157-66.
References 219 30. Carter TR, Feldman PS, Innes DJ, Frierson HF, Frigy AF. The role of fine needle aspiration cytology in the diagnosis of lymphoma. Acta Cytol 1988; 32: 848-53. 31. Pontifex AH, Klimo P. Application of aspiration biopsy cytology to lymphomas. Cancer 1984; 53: 553-6. 32. Cafferty LL, Katz RL, Ordonez NG, Carrasco CH, Cabanillas FR. Fine needle aspiration diagnosis of intraabdominal and retroperitoneal lymphomas by a morphologic and immunocytochemical approach. Cancer 1990; 65: 72-7. 33. Pond GD, Castellino RA, Horning S, Hoppe RT. NonHodgkin's lymphoma: influence of lymphography, CT and bone marrow biopsy on staging and management. Radiology 1989; 170:159-64. 34. Coller BS, Chabner BA, Gralnick HR. Frequency and patterns of bone marrow infiltration in non-Hodgkin's lymphomas: observations on the value of bilateral biopsies. Am] Haematol 1977; 3:105-19. 35. Anderson KC, Kaplan WD, Leonard RCF, Skarin AT, Canellos GP. Role of 99mtechnetium-methylene diphosphonate bone imaging in the management of lymphoma. Cancer Treat Rep 1985; 69:1347-51. 36. Hermann G, Klein MJ, Abdelwahab IF, Kenan S. MRI appearance of primary non-Hodgkin's lymphoma of bone. Skeletal Radiol 1997; 26: 629-32. 37. White LM, Schweitzer ME, Khalili K, Howarth DJ, Wunder JS, Bell RS. MR imaging of primary lymphoma of bone: variability of T2-weighted signal intensity. AmJ Roentgenoll998; 170:1243-7. 38. Carr R, Barrington SF, Madan B, etal. Detection of lymphoma in bone marrow by whole-body positron emission tomography. Blood 1998; 91: 3340-6. 39. Moog F, Bangerter M, Kotzerke J, Guhlmann A, Frickhofen N, ReskeSN. 18-F-fluorodeoxyglucose-positron emission tomography as a new approach to detect lymphomatous bone marrow. J Clin Oncol 1998; 16: 603-9. 40. Biorklund A, Cavallin-Smith E, Landberg T, Lindberg LG, Akerman M. Biopsy of the nasopharynx as a staging procedure in Hodgkin's disease. Acta Radiol 1976; 15: 387-93. 41. Bragg DG. Radiology of the Lymphomas. Curr Probl Diagnos Radiol 1987; 16:183-206. 42. Harnsberger RH, Bragg DG, Osborn AG, et al. NonHodgkin's lymphoma of the head and neck: CT evaluation of nodal and extranodal sites. Am J Neuroradiol 1987; 8: 673-9. 43. Cobleigh MA, Kennedy JL Non-Hodgkin's lymphoma of the upper aerodigestive tract and salivary glands. Otoloaryngol Clinics North Am 1986; 19: 685-710. 44. Westacott S, Garner A, Moseley IF, Wright JE. Orbital lymphoma versus reactive lymphoid hyperplasia: an analysis of the use of computed tomography in differential diagnosis. BrJ Ophthalmol 1991; 75: 722-5. 45. Chisin R, Weber AL. Imaging of lymphoma manifestations in the extracranial head and neck region. Leuk Lymphoma 1994; 12:177-89. 46. DePena CA, Van Tassel P, Lee Y-Y. Lymphoma of the head and neck. Radiol Clin North Am 1990; 28: 723-43.
47. Herman TS, Hammond N, Jones SE, Butler JJ, Byrne GE, McKelvey EM. Involvement of the CNS by non-Hodgkin's lymphoma. The south-western oncology group experience. Cancer 1979; 43: 390-7. 48. Hochberg FH, Miller DC. Primary central nervous system lymphoma. J Neurosurg 1988; 68: 835-53. 49. Young RC, Howser DM, Anderson T, et al. Central nervous complications of non-Hodgkin's lymphoma: potential role for prophylactic therapy. AmJ Med 1979; 66: 435-43. 50. Schouten JT, Weese JL, Carbonne PP. Lymphoma of the breast. Ann Surg 1981; 194: 749-53. 51. Townsend RR, Laing FC, Jeffrey RB, Bottles K. Abdominal lymphoma in AIDS: evaluation with US. Radiology 1989; 171:719-24. 52. Nyberg DA, Jeffrey RB, Federle MP, Bottles K, Abrams Dl. AIDS-related lymphomas: evaluation by abdominal CT. Radiology 1986; 159: 59-63. 53. Jeffrey RB, Nyberg DA, Bottles K, et al. Abdominal CT in acquired immunodeficiency syndrome. Am J Roentgenol 1986; 146: 7-13. 54. Nyberg DA, Federle MP. AIDS-related Kaposi sarcoma and lymphomas. Semin Roentgenol 1987; 22: 54-65. 55. Ziegler JL, Beckstead JA, Volberding PA, et al. NonHodgkin's lymphoma in 90 homosexual men. N EnglJ Med 1984; 311:565-70. 56. Lowenthal DA, Strauss DJ, Campbell SW, Gold JWM, Clarkson BD, Koziner B. AIDS-related lymphoid neoplasia. Cancer 1988; 61:2325-37. 57. Lorberboym M, Wallach F, Estok L, et al. Thallium-201 retention in focal intracranial lesions for differential diagnosis of primary lymphoma and nonmalignant lesions in AIDS patients. J Nucl Med 1998; 39:1366-9. 58. Naddaf SY, Akisik MF, Aziz M, et al. Comparison between 201 TI-chloride and "Tc(m)-sestamibi SPET brain imaging for differentiating intracranial lymphoma from nonmalignant lesions in AIDS patients. Nucl Med Commun 1998; 19: 47-53. 59. Thomas JL, Barnes PA, Bernadino ME, Hagemeister FB. Limited CT studies in monitoring treatment of lymphoma. AmJ Roentgenol 1982; 138: 537-9. 60. Front D, Israel 0, Epelbaum R, et al. Ga-67 SPECT before and after treatment of lymphoma. Radiology 1990; 175: 515-19. 61. BogartJA, Chung CT, Mariados NF, et al. The value of gallium imaging after therapy for Hodgkin's disease. Cancer 1998; 82: 754-9. 62. Nyman R, Rehn SM, Glimelius BLG, et .al. Residual mediastinal masses in Hodgkin's disease; prediction of size with MR imaging. Radiology 1989; 170: 435-40. 63. Herman TS, Hoppe RT, Donaldson SS. Late relapse among patients treated for Hodgkin's disease. Ann Intern Med 1985;102:292-7. 64. Newman JS, Francis IR, Kaminski MS, Wahl RL. Imaging of lymphoma with PET with 2-[F-18]-fiuoro-2-deoxy-Dglucose: correlation with CT. Radiology 1994; 190: 111-16.
220 Imaging of lymphoma
65. Stumpe KD, Urbinelli M, Steinert HC, Glanzmann C, Buck
71. Goldberg MA, Thrall JH, Alpert NM, Mueller PR, Fischman
A, von Schulthess GK. Whole-body positron emission
AJ, Lee MJ. Fluorodeoxyglucose PET of abdominal and
tomography using fluorodeoxyglucose for staging of lymphoma: effectiveness and comparison with computed tomography. EurJ Nucl Med 1998; 25: 721-8. 66. Rodriguez M. Computed tomography, magnetic resonance imaging and positron emission tomography in non-Hodgkin's lymphoma. Acta Radiol 1998; 417 (suppl): 1-36. 67. Okada J, Arimizu N, Imaseki K, et al. FDG-PET for predicting the prognosis of malignant lymphoma. Ann Nucl Med 1994; 8:187-91. 68. Rodriguez M, Glimelius B, Hagberg H, et al. [18F]FDG PET in gastric non-Hodgkin's lymphoma. Acta Oncol 1997; 36: 577-84. 69. De Wit M, Hossfeld DK, Clausen M, Herbst K, Beyer W, Bumann D. Whole-body positron emission tomography (PET) for diagnosis of residual mass in patients with lymphoma. Ann Oncol 1997; 8 (suppl): 57-60. 70. Villringer K, Schwaiger M, Pfister HW, et al. Differential
pelvic neoplasms: potential role in oncologic imaging. Radiographics 1993; 13:1047-62. 72. Weinblatt ME, Kochen J, Babchyck B, Belakhlef A, Zanzi I. False-positive FDG-PET imaging of the thymus of a child with Hodgkin's disease.) Nucl Med 1997; 38: 888-90. 73. Nuutinen J, Leskinen S, Lindholm P, et al. Use of carbon11 methionine positron emission tomography to assess malignancy grade and predict survival in patients with lymphomas. EurJ Nucl Med 1998; 25: 729-35. 74. Cerulus G, Leonard JP. A comparison of 111ln-octreotide and 67 Ga scintigraphy in malignant lymphoma. Nucl MedCommun 1997; 18: 616-22. 75. Ivencevic V, Emrich D, Hiddemann W, et al. Somatostatin receptor scintigraphy in the staging of lymphomas. Leuk Lymphoma 1997; 26:107-14. 76. Van den Anker-Lugtenburg PJ, Krenning EP, Lamberts SW, Lowenberg B. The relevance of somatostatin receptor
diagnosis of CNS lesions in AIDS patients by FDG-PET. 7
expression in malignant lymphomas. Metabolism 1996;
Comput Assist Tomogr 1995; 19: 532-6.
45 (suppl 1): 96-7.
17 Localized Hodgkin's disease SB SUTCLIFFE,AR TIMOTHY AND MH ROBINSON
Introduction
221
Management of Stage MIA Hodgkin's disease
222
Side effects of radiation therapy Conclusion
Localized disease and the determination of prognostic factors Radiation therapy for localized Hodgkin's disease Systemic therapy for localized Hodgkin's disease
226 234
INTRODUCTION The prognosis for patients with Hodgkin's disease has improved dramatically over the past 30 years. This is well illustrated by examination of incidence and mortality rates for Hodgkin's disease from a population database of approximately 10 million individuals within the Province of Ontario, Canada (Fig. 17.1). Incidence has remained fairly constant with a rate of approximately 2.85 per 100000 persons. It is noteworthy that, unlike non-Hodgkin's lymphoma, the incidence rate is not rising and there has been no identification of predetermining conditions, e.g. congenital, iatrogenic
Figure 17.1
Incidence and mortality rates for Hodgkin's
disease, male and female, Province of Ontario, Canada 1969-1992 (age adjusted to the world standard population). Data kindly provided by the Ontario Cancer Registry.
References
236 237 239 240
or acquired immunodeficiency states, Helicobacter pylori infection and gastric mucosa-associated lymphoid tissue (MALT) lymphoma, chronic immunoproliferative conditions (Sjogren's disease, Hashimoto's disease). Furthermore, the demographics of Hodgkin's disease and the bimodal incidence relative to age remain constant. Mortality, however, has declined substantially from a rate of approximately 1.46 (1965) to 0.47 per 100000 persons in 1992. In this context, mortality should be considered as a 'cause-specific' survival end point inasmuch as the information is derived from death certificate data - deaths from causes other than Hodgkin's disease will not necessarily be included in the analysis of Hodgkin's disease mortality. Important components of interpretation would include the continuous decline in mortality throughout the 30-year period, a mortalityrincidence ratio of approximately 50 per cent prior to the advent of four-drug combination chemotherapy (Fig. 17.1)1 and a current mortaliry:incidence ratio of approximately 17 per cent. Throughout the past 30 years, the prevailing philosophy of management has focused upon the highest overall and cause-specific survival rates, and the highest progression-free or relapse-free rates with individual or combined modality therapy. More recently, this philosophy has been tempered as it is increasingly recognized that: • the most important end point is overall survival, i.e. all causes of death must be considered, given the increasing recognition of treatment-related causes of death offsetting incremental cause-specific survival gains;
222 Localized Hodgkin's disease
• the most effective treatment strategy is that which achieves the highest overall survival with the least amount of therapy and the greatest functional integrity. Several developments have taken place in management of Hodgkin's disease over the last three decades: • the determination of prognostic factors, i.e. the identification of determinants of treatment and outcome, including stage of disease (clinical or pathological); • the role of radiation therapy, i.e. the selection of patients for whom radiation therapy alone results in the highest level of disease control and the definition of late radiation-related morbidity and mortality; • the role of chemotherapy, either alone or in combination with radiation therapy, and the identification of the most effective regimens with the least associated morbidity and mortality. For each of these developments, the extent to which they have contributed to overall survival, i.e. procedure, treatment and disease-related causes of death, is an important consideration with respect to further improvements in patient management. The relationship of treatment efficacy to treatment-related morbidity and mortality, particularly in early-stage Hodgkin's disease, will in large measure determine current and future therapeutic practice.
LOCALIZED DISEASE AND THE DETERMINATION OF PROGNOSTIC FACTORS The definition of early, or localized, Hodgkin's disease has developed along two lines: the description of disease extent, and the definition of prognostic factors.
Description of disease extent
Table 17.1
Ann Arbor classification for staging of Hodgkin's
disease. Reproduced from Sutcliffe SB. Hodgkin's disease: chemotherapy method. In: Raquel RE, ed. Conn's current therapy. Philadelphia: W.B. Saunders Co., 1995
I II III
IV
Fever Night sweats Weight loss
Single lymph node region Two or more lymph node regions on the same side of the diaphragm Lymph nodes on both sides of diaphragm (the spleen is considered a lymph node in this classification) Disseminated involvement of one or more extralymphatic organs or tissue, with or without associated lymph node involvement
>38.4°C; repeat episodes; no documented infectious cause Repeated episodes, sufficient to moisten night attire or bed linens >10+% body weight in 6 months prior to diagnosis, in absence of other identifiable cause
a All stages can be A (asymptomatic) or B (symptomatic). Stage III can be subdivided into III, (involvement of upper abdomen, i.e. spleen, splenic hilar, celiac or porta hepatis nodes) or III2 (involvement of lower abdomen, i.e. para-aortic, mesenteric, pelvic or inguinofemoral nodes). Stage IV, disseminated involvement, is distinguished from E disease (applicable to Stages l-lll), in which extension into an extranodal site occurs contiguous with adjacent nodal disease. The distinction of E disease from Stage IV disease is therapeutically valid only if the extranodal extension can be encompassed with the nodal disease and relevant lymphatic pathways in an appropriate irradiation field to the full therapeutic radiation dose. b If any or all symptoms are present, disease is classified as B; if all symptoms are absent, disease is classified as A. Other relevant symptoms/history include lassitude, anorexia, generalized pruritus, alcohol-induced pain and immunodeficiency state (congenital or acquired, e.g. HIV infection or predisposition; therapeutic immunosuppression).
This is the determination of the anatomic distribution of disease according to the Ann Arbor classification2 (Table 17.1) and subsequent refinements.3 The principal purpose of staging is to establish the most appropriate selection for localized therapy (radiotherapy), systemic therapy (chemotherapy), or both, and to categorize patient groups reproducibly to allow comparison of outcomes between different treatments and treating institutions. The approach to staging has evolved through the following.
resonance imaging and radioscintigraphy (gallium scans, bone scans). Bone marrow aspiration and biopsy evaluation have traditionally been a component of clinical staging, although current practice would probably limit its application to those with advanced clinical stage, adverse prognostic factors or abnormal hematology. Furthermore, a positive study would constitute a pathological stage.
CLINICAL STAGING
SURGICAL (OR PATHOLOGICAL STAGING)
This is the description of disease extent by history, including the evaluation of systemic symptoms, physical examination, hematological and biochemical parameters, imaging studies including plain radiographs, computerized tomography, lymphography, magnetic
Staging laparotomy with splenectomy, liver biopsy(ies) and lymph node mapping was introduced by the Stanford group in the late 1960s as a means of more accurately denning those with localized disease as defined by the Ann Arbor classification, and thereby
Determination of prognostic factors 223
more appropriately defining the extent and configuration of irradiation fields.4 Twenty-five years of experience with surgico-pathological staging has established the following benefits. 1 Staging laparotomy and splenectomy remains the most accurate means of evaluating the abdomen in Hodgkin's disease. In those with supradiaphragmatic clinical Stage I and II disease, it is the only certain way to establish the approximately 25-30 per cent of otherwise unselected patients, or the 15-20 per cent of patients selected by other clinical prognostic factors, who have occult abdominal disease.4-8 Likewise, splenectomy remains the only reliable means of assessing splenic disease below the threshold of clinical detection.9 2 The probability of occult intra-abdominal disease and the patterns of intra-abdominal involvement have been defined as a result of surgical staging. The spleen, the splenic hilar and the celiac axis nodes are me most likely sites of involvement in those with clinical Stage I and II supradiaphragmatic disease. Mesenteric and porta hepatic nodes are very rarely involved, and liver involvement rarely, if ever, occurs without splenic disease and correlates with the extent of splenic involvement. Para-aortic node involvement is commonly associated with splenic involvement. 3 Aspects of the pathobiology of disease have been defined, e.g. the variation in histology from presentation site to lesser areas of involvement defined at laparotomy and the lympho-depleting character of the ongoing natural history of disease. In addition, the origins of the disease in T celldependent areas of lymphoid tissue (the paracortex of lymph nodes and the periarteriolar sheath within the spleen) were defined by examination of grossly uninvolved tissues sampled at laparotomy. 4 The sensitivity and specificity of imaging and percutaneous biopsy procedures have been defined, and their limitation with respect to disease distribution and mass size characterized. 5 Clinical characteristics at presentation and related prognostic factors can be correlated with the probability of abdominal disease, its distribution and the likelihood of single versus multiple sites of involvement.6,7 6 Oophoropexy can be performed for those who would receive lower abdominal or pelvic irradiation. 7 Abdominal radiation fields may be minimized to avoid unnecessary exposure to organs that would otherwise be included in fields determined according to clinical staging practices, e.g. the exposure of the left lower lobe of the lung, upper half of left kidney, and portions of stomach and bowel within fields designed to irradiate the spleen, splenic hilar nodes and lymphatic pathways along the splenic artery.
8 Hematological tolerance to chemotherapy or to total nodal irradiation may be enhanced, although this remains a controversial issue.10,11 9 Laparotomy findings have prognostic relevance beyond documentation of involvement, e.g. pathological Stage III A, versus IIIA2,12 number of sites of involvement in pathological Stage IIIA, and splenic involvement and the probability of liver disease.13 There are also several important adverse issues relating to surgical staging. These are listed below. 1 The procedure has a recognized sampling error, particularly in the upper abdomen in the mid-line between T12 (celiac axis node level and a common site of involvement) and L 2/3 [the root of the mesentery and the superior (L 1/2) and inferior mesenteric (L 2/3) nodal levels]. 2 The spleen must be removed to be properly assessed, thereby conferring the hazards of the postsplenectomy state. 3 The procedure has a small mortality rate (less than 1 per cent), but a significant morbidity rate, including bowel obstruction, deep vein thrombosis, urinary tract infections and wound dehiscence.14 4 Postsplenectomy myocardial infarction possibly related to thrombocythaemia and altered coagulability states. 5 Controversy relating to an increased risk of acute leukemia in patients treated for Hodgkin's disease following splenectomy.15,16 While there appears to be no increased risk of acute leukemia following splenectomy for traumatic rupture and other benign and malignant causes,17 other factors are also of relevance in the Hodgkin's disease population, e.g. immune dysfunction postsplenectomy and possible increased dosages of chemotherapeutic agents in splenectomized versus non-splenectomized patients. 6 Costs and delay in therapy associated with a major surgical procedure and any associated complications. 7 With selection using multiple favorable prognostic factors at presentation, the probability of having abdominal disease defined by laparotomy is less than 20 per cent.7 Accordingly, some 80 per cent of patients are exposed to a major non-therapeutic procedure that is unlikely to influence management or outcome. Given that staging laparotomy and splenectomy is a well-characterized procedure with highly reproducible findings between similarly selected patients within and between institutions, the question as to the impact of surgical versus clinical staging upon overall survival is important. Has any of the improvement in survival for patients with Hodgkin's disease over the past 25 years been due to surgical staging and the increased accuracy of definition of disease extent it offers?
224 Localized Hodgkin's disease
Firstly, would it be expected that improvements in staging technology would influence survival? This has been addressed by Bradford-Hill ('the fallacy of attribution'18 and Bush ('stage migration')19 - the apparent improvement in overall survival by improving the staging methodology. As shown in Fig. 17.2, various subgroups of patients, all equal in number, are characterized by a survival probability that becomes progressively less according to a stage designation shown above the boxes. The overall survival of the group is 50 per cent. A more refined description of stage is employed as shown beneath the boxes. The reclassification into new stages affects survival by stage, even though the overall survival (50 per cent) is unchanged. Expressed more simply, the outcome depends upon the available therapy and not upon the description applied to the disease, unless the staging procedure, in and of itself, modifies the natural history of the disease. Secondly, is there any historical precedent for the view that surgically staged patients fare better than clinically staged patients during the same era of treatment? While this evidence is circumstantial given the retrospective nature of the data, a comparison of survival in a large cohort from Stanford University (predominantly surgically staged) and from Princess Margaret Hospital, Toronto (clinically staged) between 1968 and 1977 revealed no difference in outcome between the two groups.20 Finally, the most compelling evidence derives from the analysis of prospective randomized trials in early-stage Hodgkin's disease conducted by the European Organisation for Research and Treatment of Cancer (EORTC). In the H-2 study, a comparison of two differing radiation techniques according to whether laparotomy and splenectomy was performed revealed no difference in survival rates in the total cohort or the subset receiving no additional maintenance chemotherapy.21 The H-6 study, involved a randomization of'favorable' patients as judged by presentation prognostic factors, between surgical staging and clinical staging (H6-F group). Surgically staged patients were treated with mantle radiation or combined modality therapy, with clinically
staged patients receiving subtotal nodal irradiation encompassing the spleen. No difference in cause-specific survival was noted between the two trial arms, but overall survival by 10 years of follow-up was actually worse in the laparotomy arm due to an excess of both early and late deaths, including a small number of postlaparotomy myocardial infarctions.22 It is also pertinent to ask which groups of patients derive a benefit from laparotomy and splenectomy. By conventional wisdom, it would seem to be those who have a negative procedure (pathological Stage I-II) in whom radiation is deemed appropriate therapy, or those with a positive result (pathological Stage III-IV) in whom chemotherapy is the most appropriate treatment. A theoretical analysis is presented in Fig. 17.3 based upon anticipated results of performing laparotomy and splenectomy in clinical Stage I and II patients, and applying optimal therapeutic outcomes as defined in the literature. The analysis indicates that the proportion of patients with unselected clinical Stage I and II disease who can remain disease-free following radiation therapy alone is approximately 60 per cent. This figure is in accordance with the 61 per cent relapse-free rate reported in a series of 252 patients with clinical Stage I and II Hodgkin's disease treated with radiation therapy alone between 1968 and 1977.23 Thus, the proportion of patients who remain permanently relapse-free following radiation is the same whether laparotomy is employed or not; the radiocurable population exists more as a function of radiation fields employed than as a consequence of surgical staging. The real population deriving benefit from laparotomy are those who are found to have occult intra-abdominal disease and receive chemotherapy. Paradoxically, laparotomy is selecting for benefit from chemotherapy, not from radiation therapy.
Definition of prognostic factors While improvements in therapy have been such as to minimize the impact of prognostic factors defined prior
Figure 17.2 Diagram illustrating how the results in each stage can be improved by 'improving' the staging methods. Reproduced with permission from Bush RS. Malignancies of the ovary, uterus and cervix: the management of malignant disease series; 2. Peckham MJ, Carter RL, gen. eds. London: Edward Arnold, T 979:34.
Determination of prognostic factors 225
Figure 17.3
Theoretical
analysis of treatment outcome for 100 patients with clinical stage I and II Hodgkin's disease based on anticipated surgical staging results. PS = pathological stage, XRT = radiotherapy, CT = chemotherapy.
to therapy as a means of estimating survival, these factors have become increasingly incorporated into decision making with regard to optimal therapy. As many of these factors are inter-related, multivariate analyses are required to establish independent determinants of outcome. Even so, the variation in patient assessment, therapy, prognostic factors examined and the differing methods of assigning values, and the differing types of statistical analyses used render exact comparison and uniform conclusions difficult. Three general questions may be asked of prognostic factors: 1 What are the independent determinants of prognosisf For what endpoints or outcome do they predict? With what disease or treatment-related process do they correlate? Recent reviews have addressed the first two questions.24-25 In the absence of definitive information on the biology of treated and untreated disease, answers to the third question are purely speculative. The principal prognostic factors for localized Hodgkin's disease are shown in Table 17.2. Systemic symptoms are recognized in the Ann Arbor classification (Table 17.1). A recent re-evaluation of systemic symptoms indicates that severe pruritus, a symptom not incorporated in the Ann Arbor classification, also confers an adverse prognosis.26,27 Systemic symptoms confer an adverse prognosis in relation to primary disease control and cause-specific survival28 and are correlated strongly with extent of disease defined by stage or total tumor burden.28,29 The prognostic significance of anatomical stage as described by the Ann Arbor classification has been confirmed repeatedly over the last three decades. However,
clinical experience has indicated that subgroups of patients with widely differing prognoses exist within individual Ann Arbor stages. Specific examples would include: the favorable prognosis of isolated unilateral high cervical nodes or unilateral inguinal or femoral nodes within Stage I disease; the adverse prognosis of bulky mediastinal disease (see section on 'Mediastinal adenopathy' later); the adverse impact of IIIAj versus III A2 disease;12,13 the extent of splenic involvement;30 inguinal adenopathy in advanced stage disease;31 and bone marrow involvement in Stage IV disease.32 Agreement exists that both anatomic stage and volume of disease (total tumor burden), whether considered separately or in conjunction, constitute highly significant independent prognostic determinants.33 The principal
Table 17.2 Prognostic factors in Stage I and II Hodgkin's disease.
Factor
Variables
Stage Symptoms
I v II Ann Arbor classification Fever (>38.5°C), weight loss (>10%), night sweats, severe pruritus LPand NSvsMCand LD NS types I and II > 3 sites vs < 3 Bulk and number of sites None or small vs large > 50 years vs < 50 years Male vs female >50 mm/h vs<50
Histological type Number of sites involved Tumor burden Large mediastinal mass Age Gender Sedimentation rate
LP = Lymphocyte predominant, NS = nodular sclerosis, MC = mixed cellularity, LD = lymphocyte depleted.
226 Localized Hodgkin's disease
impact of stage or tumor burden is upon primary disease control, relapse distant to irradiation fields and causespecific survival. An important exception, however, is bulky mediastinal disease, where the adverse impact upon locoregional control and local relapse-free rate following radiation alone does not appear to have as established adverse impact upon overall or cause-specific survival. Histological subtyping according to the Rye classification34 has proved to be prognostically sensitive, although this is largely expressed in patients with localized disease receiving radiation therapy alone.7,23,35 Patients with mixed-cellularity and lymphocyte-depleted histology have a worse prognosis with a poorer cause-specific survival. Histology is, however, strongly correlated with age, gender and extent of disease, and its independent significance is controversial. The prognostic implications of subdivision of nodular sclerosing histology into lymphocyte-rich (type I) or lymphocyte-depleted (syncytial variant - type II) is also a continuing source of debate.36-39 The independent adverse prognostic impact of older age (>40 years or >50 years) has been defined in numerous reports. This is expressed upon primary disease control, survival after relapse and cause-specific survival.7,23,40-42 While a reduced ability to tolerate treatment in older patients is often cited, this is probably a consequence of more aggressive disease biology. Gender, independent of other prognostic factors, appears to have minimal, if any, impact upon outcome or treatment allocation. A number of hematologic and biochemical parameters have been shown to have an adverse impact upon prognosis, including elevated sedimentation rate,28,40,43,44 lymphocytopenia and severe anemia in advanced disease. In early stage disease, abnormalities of serum copper, ceruloplasmin, ferritin, b2 microglobulin, serum albumin, lactic dehydrogenase and soluble CD 30 are unusual, and their prognostic utility is not established. Characterization of tumor cell populations by phenotypic, flow cytometric, genotypic or karyotypic attributes has provided information relevant to diagnostic accuracy and insights into tumor cell biology, but none has yet been shown to be of prognostic value. Similarly, HLA characterization has provided no clinically applicable prognostic information.39
RADIATION THERAPY FOR LOCALIZED HODGKIN'S DISEASE The evolution of curative radiation therapy for Hodgkin's disease has largely been established on two constructs: the radiation dose required to control disease within the radiation field, and the radiation volume(s)/ [field disposition(s)] required to control both clinically
apparent disease and adjacent regions with a high probability of occult involvement.
Radiation dose, fractionation and tumor control Peters reported in 1966 the difference in survival between 'high'- (>2500 roentgen) and 'low'-dose radiation.1 However, the first definitive analysis of a dose control relationship was presented by Kaplan in 1966.45 This retrospective analysis of data from predominantly kilovoltage therapy proposed a linear relationship with a 22 per cent control rate with <1000 roentgen and a 98.6 per cent control rate at 4400 roentgen. Fletcher and Shukovsky subsequently reinterpreted the Kaplan analysis and presented a sigmoid dose-response relationship with control rates of 93 per cent and 97 per cent at 3000 roentgen, and between 3101 and 4000 roentgen.46 Further analyses incorporating current patient cohorts treated with megavoltage beams have largely confirmed Fletcher and Shukovsky's report. Vijayakumar and Myrianthopoulos cite a 98 per cent in-field control rate for 4117 sites at risk with a tumor dose of 37.5 Gy, and a relationship with tumor bulk (<6 cm; >6 cm) with 98 per cent in-field control rates for subclinical, <6 cm and >6 cm disease at tumor doses of 32.4 Gy, 36.9 Gy and 37.4 Gy.47 A reanalysis of the Vijayakumar data set by Brincker and Bentzen, however, failed to demonstrate any dose-response relationship above 35 Gy, nor was there any evidence of a requirement for doses higher than 32.5 Gy for bulkier disease.48 The absence of an incremental dose control relationship above 30 Gy has also been reported in the Patterns of Care Outcome Studies49 and from the Medical College of Wisconsin data.50 Gospodarowicz et al. reported a 95 per cent infield control rate with 35 Gy.43 The current evidence would appear to be fairly uniform in establishing that doses in excess of 35 Gy result in no gain in local control, that in-field relapse at this dose is low (<5 per cent), the majority of failures are either 'marginal' or distant to the radiation field, and that doses in excess of 35 Gy can only contribute to enhanced treatment morbidity. Indeed, in the German Hodgkin's Study Group's randomized study of extended field radiotherapy of 40 Gy compared with extended field radiotherapy of 30 Gy followed by involved field radiotherapy of 10 Gy, it was concluded that subclinical disease could be sufficiently treated with 30 Gy.51 Data relating to overall treatment time, including the continuous or split course regimens reported by Johnson et al.52 have been reviewed by Brincker and Bentzen.48 No major effect of overall treatment time up to 7 weeks was defined. Despite a paucity of reports of differing fraction size, sensitivity to change in dose per fraction appears to be low within the range of daily fraction size of 1.5-2.0 Gy,48,50 while morbidity is increased with increasing dose per fraction (>2.5 Gy).
Radiation therapy 227
matic and infradiaphragmatic presentation of Stage I and II Hodgkin's disease.
Field size (radiation volume) The importance of including adjacent, clinically uninvolved regions within the radiation field encompassing the involved presentation site was recognized by Finzi53 and demonstrated by Gilbert, Desjardins, Peters and Middlemiss, and Easson and Russell.54-57 However, Specht et al, in their meta-analysis of 1974 patients in eight randomized trials of more versus less expensive radiotherapy, showed that while more extensive fields had a large effect on disease control, there was only a small effect on overall survival.58 Management considerations differ for supradiaphrag-
SUPRADIAPHRAGMATIC DISEASE
The results of radiation therapy for control of Stage I and II Hodgkin's disease are shown in Table 17.3-17.6 according to the use of staging laparotomy and splenectomy and radiation fields confined to one side of the diaphragm or to both mantle and abdominal fields. The data are largely single-arm, consecutive patient experience and encompass a time period of more than two decades, during which management was influenced by surgical staging, radiotherapeutic technique and elabo-
Table 17.3 Relapse rates following extended-field mantle irradiation" for clinical Stage I and II Hodgkin's disease. Reproduced with modifications with permission from Timothy AR, VanDykJ, Sutcliffe SB. Radiation therapy for Hodgkin's disease. In: Selby P, McElwain TJ, eds. Hodgkin's disease. Oxford: Blackwell Scientific Publications, 1987:184-7
a
83
ns
Mantle
53/83 (63%)
Transdiaphragmatic failure rate (62%) Local relapse (14%)
Rubin et al. (1974)59
45
36/4-6
Mantle
17/45(38%)
13 of 15 lymph node extension below diaphragm IB + IIB Mantle and PA field MC + LD
StoffelandCox(1977) 60
32
35/4
Mantle
IA (29%, ii = 15) HA (5896, n = 14) IIB(66%,n = 3)
152
40/4
Mantle ± inverted
88/152(58%)
Local relapse (11%) Non-irradiation area (28%) Extranodal relapse (19%)
Tubiana et al. (1979)62
64
30-38/3-4
Mantle
37/64 (58%)
Local relapse (17%) Transdiaphragm (33%)
Liew et al. (1983)63
130
35/4
Mantle
IA (29%, n - 37) MA (46%, A7 = 84) IB + IIB (33%, i? = 9)
521
35/4
Involved
42.7% (1968-72)'
Mantle
30.3% (1973-7)
Mantle
29.0% (1978-82)
Mantle
20% (1978-86) ,
Timothy et al. (1978)61
Sutcliffe et al. (1985)23
Gospodarowicz(1992)43 OS - 87.2%, 77.6% and 63% @ 5,10 and 20 years CSS - 89.6%, 86.7% and 83.3% @ 5,10 and 20 years RFR - 70.1%, 65.8% and 63.7% @ 5,10 and 20 years Low risk (CS I + MA, < 50 years, LP/NS ESR < 40, no large mediastinal mass no 'E' lesions)
35
40/4
Mantle
0/35
Very favorable subgroup - H-7 total OS -100% @ 3 years FFS - 82% @ 3 years
Noordijk(1994)64
24
35/4 (+ 5/3 boost)
Mantle
9/24
CS IA + B RFR > 60%
Ganesan (1990)6 15 years
Radiation confined to nodal sites on one side of the diaphragm. OS = overall survival, RFR = relapse-free rate, ns = not specified, MC = mixed cellularity, LD = lymphocyte depleted, PA = para-aortic, LP = lymphocyte predominant, NS = nodular sclerosis, ESR = erythrocyte sedimentation rate, CS = clinical stage, CSS = cause-specific survival
228 Localized Hodgkin's disease
ration of prognostic factors predictive for local and distant relapse following radiation. Several general statements can be drawn from the data. 1 In otherwise unselected patients with clinical Stage I and II Hodgkin's disease, the overall progression rate following radiation confined to supradiaphragmatic nodes (mantle) is of the order of 45-60 per cent (Table 17.3). The substantial majority of postradiation failures will be in the non-irradiated area, most commonly in the abdomen in patients treated with mantle irradiation alone for supradiaphragmatic disease. The pattern and magnitude of disease progression risk are not surprising given that staging laparotomy results indicate an advance to pathological Stage III or IV disease in at least 25 per cent of unselected patients with clinical Stage I and II disease. 2 In otherwise unselected patients with pathological Stage I and II disease, the overall progression rate following mantle radiation is commonly in the range of 30-35 per cent (Table 17.4). Again, the most common site of failure is in non-irradiated areas, usually in transdiaphragmatic regions. That this pattern of rate of failure exists despite surgical staging is somewhat surprising; however, staging laparotomy and splenectomy is subject to sampling error, the examination of excised tissue is subject to Table 17.4
sampling variation and observer interpretation, and the pathobiology of early disease may not be adequately represented by surgically derived, histopathologically examined material. 3 The use of mantle and abdominal irradiation in otherwise unselected patients with clinical Stage I and II disease results in progression rates of approximately 30-40 per cent. Failures are commonly distant to the radiation field, but can occur both within and beyond the treated volume. It may reasonably be assumed that some of the risk of 45-60 per cent failure with radiation fields confined to only one side of the diaphragm would be reduced by treating the predictable sites of abdominal disease with radiation in the absence of confirmation by laparotomy. 4 Mantle and abdominal irradiation fields for otherwise unselected patients with pathological Stage I and II disease again show a progression rate of approximately 25-35 per cent (Table 17.6). Failure is usually beyond the radiation field. 5 The results reported above draw attention to the limitation of staging laparotomy and splenectomy to describe disease extent adequately, given that a significant relapse rate exists despite negative abdominal surgical staging, the use of radiation fields that encompass the regions sampled at laparotomy, and the high rate of local disease control within the radiation field.
Relapse rates following extended-field mantle irradiation" for pathological Stage I and II Hodgkin 's disease.
Reproduced with modifications with permission from Timothy AR, VanDykJ, Sutcliffe SB. Radiation therapy for Hodgkin's disease. In: Selby P, McElwain TJ, eds. Hodgkin's disease. Oxford: Blackwell Scientific Publications, 1987:184-7
36/4-6
Mantle
9/17(53%)
39
35/4
Mantle
IA(0%,n = 17) MA (30%, n = 19) IIB (33%, n = 3)
17
40/4
Mantle
4/17(23%)
66
30-38/3-4
Mantle
22/66 (33%)
Liew et al. (1983)63
55
35/4
Mantle
16/55(29%)
Anderson et al. (1984)67
46
40/4
Mantle
5/46 (all patients)
2/23 (prospective study)
66
a
StoffelandCox(1977) 60
17
35/4 (+ 5/3 boost)
Mantle (62) Inverted Y(4)
9/66
IB+IIB MC + LD
Mantle ± PA
Timothy et al. (1978)61
I and II non-mediastinal presentation only
Hagemeister et al. (1982)6
Patients selected by NS and LP, 'A', Mauch (1995)6 disease above carina, negative lap.
OS -100% @ 4 years FFR- 83% @ 4 years
CS IA + B RFR-80%@15years
Ganesan (1990)6
Radiation confined to nodal sites on one side of the diaphragm. MC = mixed cellularity, LD = lymphocyte depleted, NS = nodular sclerosis, LP = lymphocyte predominant, OS = overall survival, FFR = freedom from relapse, RFR = relapse-free rate.
Radiation therapy 229 Table 17.5 Relapse rates following supradiaphragmatic and infradiaphragmatic irradiation for clinical Stage I and II Hodgkin's disease. Reproduced with modifications with permission from Timothy AR, VanDykJ, SutcliffeSB. Radiation therapy for Hodgkin's disease. In: Selby P, McElwain TJ, eds. Hodgkin's disease. Oxford: BlackwelI Scientific Publications, 1987:184-7
32/50 laparotomy staged but outcome Rubin et al. (1974)5 according to clinical stage presented Local relapse (14%) Transdiaphragm (0%) Dissemination/extranodal relapse (14%)
50
ns
14/50 (28%) Total nodal irradiation (no spleen radiotherapy if splenectomy)
24
36/4-6
Mantle PA + spleen
4/24(17%)
IB + IIB Total nodal irradiation MC+ LD
156
40/4
Mantle PA + spleen
44/156(28%)
Patients with MC + LD randomized to Tubiana et al. (1981)21 two chemotherapy schemes (n = 48) Local relapse (10%) Non-radiotherapy area (6%) Extranodal relapse (12%)
51
35/4
Mantle upper abdomen
IA (66%, n = 3) Spleen radiotherapy < 25 Gy in 36/52 IIA(24%, n - 38) patients IB + IIB (50%, n = 10)
128
40/4
STNI + spleen 24/128(19%)
250
35/4
Mantle + UA
28.4% @ 8 years CS I and MA, non-bulky mediastinum and no E lesion
88
35/4
Mantle + UA
Gospodarowicz(1992)43 12.7% @ 8 years Low risk (CS I and IIA, < 50 years, LP/NS ESR < 40, no large mediastinal mass, no E lesion)
130
40/4
STN I + spleen 14/130
H=6 favorable - CS I and CSII2 no bulky mediastinum A and ESR < 50 or B and ESR < 30
Favorable subgroup - H-7 trial OS - 99% @ 3 years FFS-81%@3years
StoffelandCox(1977) 6
Sutcliffe et al. (1985)23
Carde(1993)2
Noordijk(1994)64
ns = not specified, PA = para-aortic, STNI = subtotal nodal irradiation, UA = upper abdomen, MC = mixed cellularity, LD = lymphocyte depleted, ESR = erythrocyte sedimentation rate, LP = lymphocyte predominant, NS = nodular sclerosis, OS = overall survival, FFS = freedom from relapse.
6 Several investigators from the early 1980s onwards established prognostic determinants, other than clinical and surgical stage, to characterize relapse risk following radiation therapy. There is substantial agreement on the nature and predictive role of these prognostic factors that include age, systemic symptoms, Stage I vs II, histology, number of involved nodal sites, mediastinal mass (size or ratio) and sedimentation rate.7,23,40,77,78 In addition, evidence has been presented relating the probability and extent of abdominal involvement with clinically derived prognostic factors.7,6 The impact of prognostic attributes other than stage alone on treatment outcome after radiation is illustrated by relapse-free rates in excess of 80 per cent for favorably selected clinical Stage I and II patients43 treated with mantle or subtotal nodal fields22,43,64 or favorably selected patients with pathological Stage I and II disease treated with mantle irradiation alone.77
7 This experience would indicate that, by patient selection according to clinically determined prognostic parameters with or without surgical staging, relapsefree rates in excess of 80 per cent following radiation can be achieved. The more rigorous application of prognostic parameters for selection of patients results in fewer patients receiving radiation alone, but a higher relapse-free rate and, the logical corollary, more patients with Stage I and II disease receiving chemotherapy as part of initial management. Despite the evolution of an effective strategy for obtaining high relapse-free rates without the need for surgical staging and by the definition of optimum radiation volumes, the following points are salutary. 1 The mortality rate for late effects of treatment wholly or, in part, attributable to radiation now exceeds the mortality rate attributed to Hodgkin's disease in those with clinical Stage I and II disease.68,79 While
230 Localized Hodgkin's disease
Table 17.6 Relapse rates following supradiaphragmatic and infradiaphragmatic irradiation for pathological Stage I and II Hodgkin's disease. Reproduced with modifications with permission from Timothy AR, VanDykJ, SutcliffeSB. Radiation therapy for Hodgkin's disease. In: Selby P, McElwain TJ, eds. Hodgkin's disease. Oxford: Blackwell Scientific Publications, 1987:184-7
59
36/4-6
Mantle
9/59(15%)
IB + IIB
StoffelandCox(1977) 60
MC + LD)TNI 41
40/4-5
Mantle + PA
14/41 (34%)
PA treated only med. +ve
Wierniketo/. (1979)69
65
40/4 involved 30/4 non-involved
Mantle + PA
22/65 (34%)
I and II (all)
Hagemeister ef al. (1982)70
94
40/4
Mantle + PA
27/94 (29%)
109
40-50/4-6
STNI TNI
25/109(23%)
I and II (al
Hoppeefo/. (1982)71
233
36-44/4-6
STNI
(16%)
I and MA only
Mauchefo/. (1982)72
128
37/4-5
STNI TNI
28/128 (22%)
Supradiaphragm only 'E' lesions excluded
Nissen and Nordentoft (1982)73
45
40-44/4
Mantle + periaortic
12/45 (27%)
OS - 93% @ 8 years FFP-76%@8years RFS-70%@8years
fttietal.
51
45/4
STNI
17/51 (35%)
OS - 76% @ 10 years (all) Longo et al. (1991)75 - 85% @ 10 years (bulky med. and IIA1 excluded) DFS - 60% @ 10 years (all) - 67% @ 10 years (bulky med. and IIIA1 excluded)
88
40-44/4
STN1 (85% of patients)
30/83 (36%)
RFS - 69% and 62% @ 5 and Brusamolino (1994)76 10 years
Tubiana et al. (1981)21
(1992)74
MC = mixed cellularity, LD = lymphocyte depleted, STNI = subtotal nodal irradiation, TNI = total nodal irradiation, PA = para-aortic, OS = overall survival, med. = mediastinum, FFP = freedom from progression, RFS = relapse-free survival, DFS = disease-free survival.
modifications to radiation technique and treatment volumes could potentially reduce the incidence of treatment-related second malignancy or heart disease, the magnitude would probably be small (Fig 17.4). 2 Although radiation treatment volume clearly has a bearing on relapse rate, the only circumstance where this has a survival impact is in patients who receive no therapy other than radiotherapy prior to death. In practice, radiation field size has no bearing on overall survival in an era of successful salvage of disease progression by chemotherapy. 3 Given that radiation field size has no bearing on overall survival other than in those treated solely by radiation, and that radiation is a significant contributor to the cumulative incidence of 25 per cent non-disease-related mortality at 20 years of follow-up,80 the benefits of attempting to achieve higher relapse-free rates with radiation are uncertain and not measurable in terms of survival unless they are associated with a reduced treatment-related mortality. 4 Finally, if the evolution of practice is to define an
ever-diminishing population of patients with Stage I and II Hodgkin's disease with a very high relapse-free rate with radiotherapy in whom the only impact upon survival can be the reduction in treatmentrelated mortality, the major question becomes whether any radiation is required. If chemotherapy abrogates the survival impact of varying radiation parameters (or surgical staging), it is a logical question to ask whether any radiation has relevance in an era of effective chemotherapy. It is important to note, however, that this reasoning is based only on consideration of overall survival, i.e. the impacts of radiation therapy or chemotherapy on disease and treatment-related mortality. It does not take into account other aspects of therapeutic choice that might include patient preference, acute side effects, gonadal impacts, psychosocial impacts of relapse probability and therapy, and access and availability to high-quality radiation therapy.81-83 Accordingly, the future role of radiation therapy in the management of early-stage Hodgkin's disease will be determined more by comparative morbidity and mortality impacts (quality of life, incidence and causes of
Radiation therapy 231
Figure 17.4
Cumulative incidence of death - early-stage
Hodgkin's disease (n = 9041). HD - Hodgkin's disease related death (1242 cases). Other causes (715 cases). Reproduced with permission from Treatment strategy in Hodgkin's disease. Proceedings of the Paris International Workshop and Symposiumjwfle 28-30, 7990:409. Somers R, Henri-Amar M, MeerwaldtJH, Carde P, eds. London: John Libbey& Co. Ltd (Eurotext).
death) than by the traditional determinant, relapsefree rate. This rationale underlies the current EORTC H-8 trial and the National Cancer Institute (Canada) HD-6 study.
1 For patients with peripheral clinical Stage I disease, the probability of advancing stage by performing through laparotomy and splenectomy is approximately 10 per cent, and sites would include retroperitoneal nodes (despite negative lymphogram) or spleen. Such patients have commonly received inverted Y fields with or without splenic coverage with a relapse rate not exceeding 20 per cent. Supradiaphragmatic nodal areas constitute the most frequent relapse sites. 2 Patients with symptomatic clinical or pathological Stage II disease have a very high failure rate with extended field radiation. 3 Other than for clinical or pathological Stage IA disease, confined to inguinal or femoral nodes, where extended field radiation is an appropriate therapy with a relapse risk of 20 per cent, patients with infradiaphragmatic disease [clinical stage (CS) + pathological stage (PS) IIA/B and IISA/B] should receive chemotherapy as part of initial management. Given the relative infrequency of patients with infradiaphragmatic disease and the small subset with central abdominal presentation, most investigators have failed to define an adverse prognostic effect of infradiaphragmatic disease on overall survival, although Specht and Nissen90 determined an inferior disease-free survival in this group.
INFRADIAPHRAGMATIC PRESENTATION
Infradiaphragmatic Hodgkin's disease is uncommon and accounts for approximately 10 per cent of clinical Stage I and II presentations.23,84-90 Most frequently, patients present with peripheral disease in the femoral, inguinal or superficial iliac region. Approximately 20 per cent of infradiaphragmatic patients present with deep iliac or abdominal disease. This may be manifest initially as symptoms (fever, sweats, weight loss, etc.), paravertebral mass with pain or nerve root symptomatology, splenomegaly, autoimmune hemolytic anemia, or thrombocytopenia or an abdominal mass. Patients presenting with abdominal, as opposed to peripheral, disease tend to be older, more frequently symptomatic, have bulkier disease, and mixed cellularity or lymphocytedepleted histology.91 Nodular sclerosis histology is less represented in infradiaphragmatic presentations. Patients with lymphogram-negative, clinical Stage I infradiaphragmatic disease rarely have more extensive disease defined at laparotomy. This also applies in the very rare circumstance of bilateral inguinal/femoral disease (clinical Stage I) with a negative lymphogram. A positive lymphogram predicts for more extensive disease with a high correlation with splenic involvement and a lower probability of liver disease.86,88,91 Treatment approaches have varied from involved field irradiation to combined modality therapy based on small numbers of patients treated over a long time period. The following guidelines can be derived.
MEDIASTINAL ADENOPATHY, HILAR ADENOPATHY AND EXTENSION OF MEDIASTINAL DISEASE
Mediastinal adenopathy occurs in approximately 70 per cent of all patients with Hodgkin's disease. Whilst the description of bulk varies, most series define massive mediastinal adenopathy as >l/3 of the transthoracic diameter on a posterio-anterior radiograph or > 10 cm by direct measurement of the maximal mediastinal contour. The importance of the definition of bulky (massive) mediastinal disease derives from the high intrathoracic failure rate with conventionally planned radiation fields.71,75,92-98 Such failures are usually within field, marginal and within the thoracic cavity, including lung, pleural, chest wall, pericardial, pleural or pericardial effusions or any combination of the above. The pattern of failure of massive mediastinal disease treated by conventional radiation technique almost certainly has less to do with tumor bulk than with the lymphatic anatomy of intrathoracic structures, and the limitations posed upon the concept of prophylactic or extended fields by the limiting radiation tolerance of the lungs.98,99 The lymphatic network of the lungs comprises superficial lymphatics draining from the visceral pleural surface and the alveoli to a deep lymphatic plexus following the bronchioles and bronchi to the hilar nodes. Nodes are located at second- and first-order
232 Localized Hodgkin's disease
bronchi, and communicate within pulmonary hilar and mediastinal nodes. Non-encapsulated lymph follicles do not occur distal to the bronchioles but small aggregates of lymphocytes occur in loose connective tissue throughout the lung. The parietal pleura drains through intercostal lymphatics anteriorly to internal mammary nodes and posteriorly to intercostovertebral nodes. The lower portion of the parietal pleura interconnects with diaphragmatic lymphatics. Lymphatics of the pericardium drain inferiorly to the diaphragmatic network, superiorly to the brachiocephalic nodes and the anterior and posterior mediastinal nodes. Extensive mediastinal disease may exist as discrete clusters of enlarged nodes or as conglomerate masses that may result in: • extension to the internal mammary nodes and anterior chest wall with intercostal spread, parietal pleural masses, paravertebral mass formation or pleural effusion; • extension along bronchi resulting in extensive interstitial spread with or without endobronchial disease; • direct lung infiltration; • pneumonic infiltration from bronchovascular spread with consolidation and, less commonly, cavitation; • extension through the pericardium and myocardium with costophrenic adenopathy, pericardial effusion and/or diaphragmatic involvement. The patterns of intrathoracic spread have previously been described from historic clinical series and autopsy reports, but may now be much more clearly appreciated with the routine use of computerized tomography (CT) and magnetic resonance imaging (MRI). With the use of conventionally applied radiation fields, full-thickness lung shields are used to minimize lung dose. Even with such shielding, the lung scatter dose approximates 20 per cent of the tumor dose with allowance for lung density. Lung blocks are also planned according to the radiological mediastinal contour, as defined at simulation, with a margin of approximately 1 cm around soft-tissue contours and with acknowledged shielding of intercostal lymphatics, a large part of the pericardium and most of the diaphragm. Even if'shrinking field' techniques are employed for bulky mediastinal disease to allow redesign of lung blocks as the radiological mediastinal contour recedes, it does not follow that all original disease recedes with the radiologic contour, thereby introducing the risk of dose heterogeneity throughout the tumor volume. Most importantly, disease can change in size but necessarily in anatomic position, e.g. hilar nodes, E-lesions, lung extensions. The problem of massive mediastinal disease is that of including all known disease and lymphatic channels involved by contiguous or retrograde spread within a radiation field that can be treated to a tumoricidal dose given the limiting tolerance of the lung and, to a lesser
extent, the entire pericardium. Support for this hypothesis is evident from the change in overall and intrathoracic recurrence patterns using either whole thoracic radiation with lung dose attenuation100 or combined modality therapy.101 In current practice, combined modality therapy is the standard approach for patients with bulky mediastinal disease. Staging laparotomy adds nothing to the management plan. Chemotherapy is the initial therapy and, given a high probability of response, permits a more optimal radiation volume to be defined. Split-course therapy, choice of chemotherapy regimen, number of cycles of therapy, chemotherapy end-point (partial or complete response), radiation as therapeutic or adjuvant modality, and radiation dose and field arrangement are all variables of the combined modality approach. Notwithstanding these variations, the probability of relapse with radiation of 40-70 per cent, depending on mass size, may be improved to 15-25 per cent with combined modality therapy. Despite this improvement in progression-free rate, the choice of initial therapy has no bearing on overall survival.71,72,75,92,102 An additional consideration for patients with large mediastinal disease is the interpretation of the mediastinal contour following therapy. A persisting abnormal mediastinal contour is quite common.103-105 Important interpretational features include the stability of the radiological appearance, comparison of pre- and post-treatment gallium studies, and CT and MRI characteristics. Alternate pathologies, e.g. thymic cysts, should also be considered.106 While the majority of situations of stable, gallium-negative residual abnormality do not constitute active disease based on longer term follow-up observation, no clear unambiguous management guidelines exist. Given that the further therapeutic options could range from observation to radiation to high-dose therapy with marrow/stem cell transplantation, guided thoracoscopic mediastinal biopsy(ies) may be appropriate. It is also probable that functional or biological evaluation through modalities such as positron-emission tomography may help further in the interpretation of residual mediastinal abnormalities post-therapy.
Technical aspects of radiation therapy The preceding sections on supradiaphragmatic disease and infradiaphragmatic presentation have established that the control of localized Hodgkin's disease is dependent on two factors: a tumoricidal radiation dose, and the inclusion of the known extent of disease and adjacent clinically non-involved areas and their lymphatic connections within the irradiated volume. As noted previously, 'in-field' failure is extremely uncommon when tumoricidal doses of radiation are achieved uniformly throughout the radiation field. The vast majority of radiation 'failures' occur distant to the radiation field and
Radiation therapy 233
reflect the clinical expression of occult disease that is not incorporated into the treatment plan. The important technical aspects of radiation relate mainly to the following: • the achievement of a uniform tumoricidal radiation dose throughout a large treatment volume; • the complicated, irregularly shaped fields resulting from the spatial relationship between the treatment volume and adjacent normal tissues; • the limiting tolerance of critical normal structures within, or close to, the treatment volume; • the variation in external surface contour and internal tissue densities within the irradiated volume, and their impact upon dose homogeneity; • the variable depth of involved tissues within the treatment volume from superficial nodal sites to mid-thoracic or abdominal regions; • the need to achieve a reproducible, accurate and consistent treatment over the duration of a fractionated daily treatment program.107,408 Most commonly 'mantle' and 'inverted Y' fields are treated with a parallel-pair technique using megavoltage photon beams in the range of 4-18 MV. Optimal beam energy is largely determined by maximum patient separation, or thickness, within the treatment field. Depth dose characteristics increase as a function of increasing photon energy - 60Co beams may result in an uneven dose distribution with unacceptably high maximum to mid-plane dose unless used at an extended skin-surface distance. High-energy photon beams (> 10 MV) will have minimal dose variation at a depth beyond the maximum dose (Dmax) but may have a dose 'build-up' zone, which could result in 'underdose' to superficial nodal regions. This 'sparing' of superficial tissues is favorable in terms of skin reaction, but unfavorable if superficial nodes are also 'spared'. In this circumstance, bolus (tissue equivalent material) may be used to bring the maximum dose closer to the skin surface, but with resultant loss of skin sparing and enhanced acute skin reaction. The treatment geometry is dependent on the required field size, the desired penumbra width and the treatment unit characteristics. Given the treatment source-patient distance and the necessity for interposed beammodifying devices (blocks, attenuators, compensators, etc.), rigorous attention to the daily reproduction of treatment geometry is essential. Ideally, isocentric megavoltage units will be used to treat anterior and posterior fields on a daily basis with the patient in a supine or prone position to avoid variation that might otherwise result from supine and prone setups with fixed units. To confirm the suitability of the planned treatment volume in relation to disease distribution and to achieve a technically reproducible setup on a daily basis, patients should undergo simulation prior to therapy. All subsequent procedures are referenced to the simulation process, which includes the validation of treatment vol-
ume, the position of beam-modifying devices, and contour characterization and placement of reference points or tattoos for reproducible positioning. Any simulation errors will be reproduced throughout the treatment, thus emphasizing the need for accurate simulation and verification of daily reproducibility of technique through the use of portal images or treatment check films.109,110 The Patterns of Care Study110 clearly established the unacceptably high in-field and marginal failure rates associated with inadequately planned or unreproducible treatment techniques. Patient immobilization devices, shielding blocks for critical normal tissues, compensators to adjust for surface contour irregularity and attenuators to deliver modified dosage to regions of limited tolerance are required to establish reproducible, uniformly distributed irradiation to the desired treatment volume. Special consideration is required when matching supra and infradiaphragmatic fields to avoid areas of 'overdosage' to critical normal tissue, e.g. spinal cord, or 'underdosage' within areas that may harbor overt or occult disease. The construction of such junctions is dependent on technique and may involve the use of customized shielding, junction wedges or 'moving' field borders. Dose uniformity throughout the radiation field and daily fraction size have an important bearing not only on local tumor control but also on normal tissue morbidity. Of particular importance are: the use of beams of appropriate photon energy to avoid significant maximum dose versus tumor dose/mid-plane dose differences; equally weighted beams from anterior and posterior to avoid inappropriately high anterior or posterior dose within the treatment volume; planning techniques that achieve dose homogeneity across the treatment volume; and fractionation techniques that treat both anterior and posterior portals each treatment day to achieve an individual fraction dose of 175-185 cGy. The importance of irradiation technique cannot be overemphasized. Radiation results in very high levels of local disease control when appropriately planned and executed. Radiation also contributes to late morbidities, which have a mortality rate in excess of the mortality rate of the disease for those presenting with Stage I and II disease.79 The morbidity of radiation, and hence the mortality, has an important relationship with radiation dose - both 'intended' dose and actually delivered dose. It is salutary to note that radiation field size does not influence survival given the availability of effective chemotherapy111 but does influence survival as a function of late, radiation-related mortality.
Radiation therapy for salvage of locally recurrent disease Local in-field recurrence following appropriate radiation therapy is uncommon. Retreatment with radiation with
234 Localized Hodgkin's disease
curative intent is rarely feasible in the original field given the limiting tolerance of normal tissues within the proposed retreatment volume. Chemotherapy is the standard and practical treatment of choice. Localized radiation for recurrence distant to the irradiation field is also rarely attractive from a theoretical consideration, i.e. the concept of accuracy of knowledge of disease extent given recurrence in sites of previously occult involvement despite wide field irradiation, or from a practical consideration, e.g. normal tissue tolerances, extranodal disease, low probability of salvage with radiation, and availability of effective chemotherapy with high potential for cure. Isolated local recurrence may occur, however, in a small proportion of patients treated initially with chemotherapy alone for unfavorable local disease or advanced disease. Experience with radiation therapy for salvage in this setting is limited by its relative rarity, differences in patient identification and selection, literature availability and reporting bias. It is clear, however, that a small number of patients with locally recurrent disease postchemotherapy may achieve long-term disease-free survival after salvage radiation.112-115 Factors that are most likely to have a bearing on the success of salvage radiotherapy are those that relate to the success of radiation as sole therapy, e.g. absence of symptoms, localized Stage I and II with a limited number of nodal sites, age, histology, erythrocyte sedimentation rate (ESR) and non-bulky mediastinum. A disease-free interval between local recurrence and prior chemotherapy of greater than 12 months has also been noted to be associated with a higher probability of salvage by radiation therapy.115 The importance of defining this small and highly selected cohort of patients derives from the desirability of avoiding intensive chemotherapy salvage with autologous marrow or stem-cell transplantation with its associated morbidity and mortality, if lesser therapy with radiation can achieve long-term disease control.
SYSTEMIC THERAPY FOR LOCALIZED HODGKIN'S DISEASE
Combined modality therapy versus radiation therapy The beneficial impact of adjuvant chemotherapy following radiation has long been established in the context of disease-free survival, particularly in populations selected according to unfavorable presentation characteristics. Thus, the evolution of practice from the late 1970s and early 1980s acknowledged the higher recurrence rates following radiation for symptomatic patients with more advanced or bulky localized disease, and established an improved relapse-free rate with subsequent adjuvant chemotherapy. The logical progression was the primary
use of chemotherapy with adjuvant radiation, thereby rendering 'unfavorable' presentations 'favourable' for conventionally applied radiation, or for fields more restricted in size or dose. The two principal questions relating to combined modality therapy versus radiation therapy as optimal initial treatment for early stage Hodgkin's disease - is there a significant improvement in disease-free survival and also in overall survival? - have been addressed in a number of randomized clinical trials. In the EORTC HI trial, otherwise unselected clinical Stage I and IIA+B patients were randomized to mantle irradiation with adjuvant maintenance vinblastine for 2 years for those with mixed cellularity and lymphocyte-depletion histology.62 At 12 years of follow-up, disease free rates of 38 per cent and 58 per cent characterized the radiation-only versus radiation plus chemotherapy arms, respectively, but overall survival rates were 58 per cent and 65 per cent. Nissen and Nordentoft73 compared subtotal nodal irradiation versus mantle plus MOPP chemotherapy in patients with PS I/IIA+B disease. At 6 years of follow-up, disease-free survival rates were 68 per cent and 95 per cent, respectively, but overall survival was no different (92 per cent and 88 per cent). Mantle radiation was compared with mantle and MVPP chemotherapy in PS I/II A+B patients by Anderson et al.67 Disease-free rates at 5 years were 69 per cent and 93 per cent, respectively, with overall survival rates of 94 per cent and 91 per cent. A chemotherapy regimen modified to avoid the gonadal and late second cancer risks of MOPP - vinblastine, bleomycin and methotrexate (VBM) - was combined with involved field radiation and compared with subtotal nodal irradiation in patients with PS I/II A+B and IIIA Hodgkin's disease by Horning et al.116 A significant difference in disease-free survival was noted with the combined modality therapy (97 per cent versus 72 per cent) but overall survival was equivalent (100 per cent and 97 per cent). A single arm confirmatory study of effectiveness of VBM plus involved field radiation has been reported but with reservations regarding treatment morbidity.117 In the EORTC H5 trial, patients with clinical Stage I and II unfavorable disease were randomized between subtotal nodal irradiation versus 'sandwich' MOPP.118 At 9 years of follow-up, disease-free survival differences were 66 per cent and 83 per cent, respectively, with a non-significant difference in overall survival (73 per cent and 88 per cent). The EORTC #7 trial has examined subtotal nodal irradiation compared with EBVP (epirubicin, bleomycin, vinblastine and prednisone) with involved field radiation in 254 patients with favorable clinical Stage I/II Hodgkin's disease. At 3 years of follow-up, failure-free survival rates of 81 per cent and 79 per cent, and overall survival rates of 99 per cent and 100 per cent, respectively, have been recorded.64 The German Hodgkin's Study Group has conducted a randomized study in patients with CS/PS I, II and IIIA disease with risk factors (mediastinal mass, extranodal
Systemic therapy 235
lesions, splenomegaly) comparing low-dose (20 Gy) versus high-dose (40 Gy) radiation in various field dispositions with all patients receiving eight-drug combination therapy. To date, this study, last updated through presentation at the Third International Symposium on Hodgkin's Lymphomas in Cologne 1995, has yielded no differences in overall survival or freedom from treatment failure for any group.119,120 However, interim analysis of their more recent study suggests that two cycles of ABVD plus radiotherapy is better than radiotherapy alone in extending the duration of freedom from treatment failure in early stage (I, II) Hodgkin's disease.121 In an attempt to establish whether the failure of all randomized studies to demonstrate a difference in overall survival related to sample size and statistical power, Specht reported a meta-analysis of all randomized trials, published or unpublished, of radiation therapy and combined modality therapy.33 She concluded that combined modality therapy was associated with a proportional reduction of approximately 15 per cent in the hazard of death, but that this reduction was not statistically significant. There was no additional indication that combined modality therapy differed in impact according to age, stage or presence of symptoms. In a more recent meta-analysis of 1688 patients in 13 trials of radiotherapy plus chemotherapy alone, Specht et al concluded that the addition of chemotherapy to radiotherapy has a large effect on disease control but only a small effect on overall survival.58
Chemotherapy versus radiation therapy Two randomized studies have addressed the issue of chemotherapy versus radiation therapy. The Italian study compared a random allocation to mantle followed by para-aortic radiation with six cycles of MOPP therapy in a total of 89 adult patients with PS I/IIA Hodgkin's disease. Initial disease control was achieved in all radiation patients (n = 45) and 40 of 44 receiving chemotherapy. At an 8-year median follow-up, overall survival was significantly higher in the radiation (93 per cent) compared with the chemotherapy group (56 per cent). Freedom from progression rates (76 per cent and 64 per cent) and relapse-free survival rates (70 per cent and 71 per cent) were not significantly different. Eight patients (of 44 treated with MOPP) died of Hodgkin's disease and three additional MOPP-treated patients died of a second cancer.74 The National Cancer Institute (NCI) study compared radiation alone (predominantly subtotal nodal) with MOPP chemotherapy in patients with Stage I/II A+B and IILAj disease.75 Over the time course of the study initiated in 1978, the majority of patients were surgically staged, those with bulky mediastinal masses were excluded in 1981 and those with Stage IIIA1 disease in 1983. Of the 51 patients receiving radiation, the pro-
jected 10-year disease-free survival and overall survival rates were 60 per cent and 76 per cent compared with 86 per cent and 92 per cent for the 54 patients randomized to MOPP. With exclusion of patients with large mediastinal masses and Stage IIIA1 disease, features recognized to be unfavorable for radiation therapy alone, disease-free and overall survival rates for radiation and chemotherapy were 67 per cent and 82 per cent, and 85 per cent and 90 per cent, respectively, with no significant difference between the two arms. These two studies, apparently with differing conclusions, raise questions relating to the relatively poor performance of the MOPP regimen in the Italian study and the unexpectedly high cause-specific and overall mortality with chemotherapy, and the relatively poor performance of the radiation therapy arm in the NCI study. They would also indicate that the superiority of radiation or chemotherapy, based upon random allocation in appropriately selected cohort Stage I and II Hodgkin's disease patients, remains to be determined and that overall survival remains the single most important end point, given both disease and treatment-related mortality. Non-randomized studies of chemotherapy alone for management of early-stage Hodgkin's disease have been reported, with somewhat variable results. Expected high remission rates with relapse rates up to 30 per cent have been reported by Ziegler et a1., Olweny et al, Lauria et al. and Bubman et al.122-125 Colonna and Andrieu126 reported a low complete remission rate in a small series of clinically staged patients with adverse prognostic factors in Algeria. Al-ldrisse and Ibrahim83 reported a more favorable experience from Saudi Arabia in patients with adverse prognostic factors characteristic of non-Western patient populations. These studies highlight the differing distribution of prognostic factors in patient populations with earlystage disease in Western and non-Western countries, and the practical and pragmatic considerations that influence treatment choice given limited access to radiation therapy appropriate to the technical complexity and quality assurance required for optimal radiation for Hodgkin's disease.81,83
Chemotherapy versus combined modality therapy Chemotherapy alone (cyclophosphamide and vinblastine on day 1, and procarbazine and prednisone on days 1-14 on a 28-day cycle for six cycles - CVPP) was compare