Childhood Leukemias [2 ed.] 0521825199, 9780521825191, 9780511471001

Citation preview

Childhood Leukemias

Childhood Leukemias Second edition

Edited by

Ching-Hon Pui St. Jude Children’s Research Hospital, Memphis, Tennessee, USA

cambridge university press Cambridge, New York, Melbourne, Madrid, Cape Town, Singapore, S˜ao Paulo Cambridge University Press The Edinburgh Building, Cambridge CB2 2RU, UK Published in the United States of America by Cambridge University Press, New York www.cambridge.org Information on this title: www.cambridge.org/9780521825191
C Cambridge University Press 2006

This publication is in copyright. Subject to statutory exception and to the provisions of relevant collective licensing agreements, no reproduction of any part may take place without the written permission of Cambridge University Press. First published 2006 Printed in the United Kingdom at the University Press, Cambridge A catalog record for this publication is available from the British Library ISBN-13 978-0-521-82519-1 hardback ISBN-10 0-521-82519-9 hardback

Cambridge University Press has no responsibility for the persistence or accuracy of URLs for external or third-party internet websites referred to in this publication, and does not guarantee that any content on such websites is, or will remain, accurate or appropriate. Every effort has been made in preparing this publication to provide accurate and up-to-date information which is in accord with accepted standards and practice at the time of publication. Although case histories are drawn from actual cases, every effort has been made to disguise the identities of the individuals involved. Nevertheless, the authors, editors and publishers can make no warranties that the information contained herein is totally free from error, not least because clinical standards are constantly changing through research and regulation. The authors, editors and publishers therefore disclaim all liability for direct or consequential damages resulting from the use of material contained in this publication. Readers are strongly advised to pay careful attention to information provided by the manufacturer of any drugs or equipment that they plan to use.

For my parents and all the patients I have been privileged to care for.

Contents

List of contributors Preface

page ix xv

Part I History and general issues 1 Historical perspective Donald Pinkel

3

2 Diagnosis and classification Mihaela Onciu and Ching-Hon Pui

21

3 Epidemiology and etiology Logan G. Spector, Julie A. Ross, Leslie L. Robison, and Smita Bhatia

48

Part II Cell biology and pathobiology 4 Anatomy and physiology of hematopoiesis Connie J. Eaves and Allen C. Eaves 5 Hematopoietic growth factors James N. Ihle 6 Signal transduction in the regulation of hematopoiesis James N. Ihle 7 Immunophenotyping Fred G. Behm 8 Immunoglobulin and T-cell receptor gene rearrangements Jacques J. M. van Dongen and Anton W. Langerak 9 Cytogenetics of acute leukemias Susana C. Raimondi

69

106

125

150

210

235

vii

viii

Contents

10 Molecular genetics of acute lymphoblastic leukemia Adolfo A. Ferrando, Jeffrey E. Rubnitz, and A. Thomas Look

272

24 Acute leukemia in countries with limited resources Raul C. Ribeiro, Scott C. Howard, and Ching-Hon Pui

625

298

25 Antibody-targeted therapy Eric L. Sievers and Irwin D. Bernstein

639

11 Molecular genetics of acute myeloid leukemia Robert B. Lorsbach and James R. Downing

339

26 Adoptive cellular immunotherapy Helen E. Heslop and Cliona M. Rooney

648

12 Apoptosis and chemoresistance Kirsteen H. Maclean and John L. Cleveland

27 Gene transfer: methods and applications Martin Pule´ and Malcolm K. Brenner

661

28 Minimal residual disease Dario Campana, Andrea Biondi, and Jacques J. M. van Dongen

679

13 Heritable predispositions to childhood hematologic malignancies Doan Le, Kevin Shannon, and Beverly J. Lange

362

Part III Evaluation and treatment Part IV Complications and supportive care

14 Pharmacokinetic, pharmacodynamic, and pharmacogenetic considerations Shinji Kishi, William E. Evans, and Mary V. Relling

391

15 Assays and molecular determinants of cellular drug resistance Monique L. den Boer and Rob Pieters

414

29 Acute complications Scott C. Howard, Raul C. Ribeiro, and Ching-Hon Pui

709

30 Late complications after leukemia therapy Melissa M. Hudson

750

31 Therapy-related leukemias Carolyn A. Felix

774

32 Infectious disease complications in leukemia Jeremy A. Franklin and Patricia M. Flynn

805

487

33 Hematologic supportive care Fariba Navid and Victor M. Santana

829

19 Acute myeloid leukemia Jeffrey E. Rubnitz, Bassem I. Razzouk, and Raul C. Ribeiro

499

34 Pain management Alberto J. de Armendi and Doralina L. Anghelescu

850

540

35 Psychosocial issues Raymond K. Mulhern, Sean Phipps, and Vida L. Tyc

858

20 Relapsed acute myeloid leukemia Ursula Creutzig

548

36 Nursing care Pamela S. Hinds, Jami S. Gattuso, and Belinda N. Mandrell

882

21 Myelodysplastic syndrome Henrik Hasle 22 Chronic myeloproliferative disorders Charlotte M. Niemeyer and Franco Locatelli

571

Index

894

23 Hematopoietic stem cell transplantation Rupert Handgretinger, Victoria Turner, and Raymond Barfield

599

16 Acute lymphoblastic leukemia Ching-Hon Pui

439

17 Relapsed acute lymphoblastic leukemia ¨ Gunter Henze and Arend von Stackelberg

473

18 B-cell acute lymphoblastic leukemia and Burkitt lymphoma John T. Sandlund and Ian T. Magrath

The plates are situated between pages 400 and 401.

Contributors

Anghelescu, Doralina L., MD Associate Member Division of Anesthesiology St. Jude Children’s Research Hospital Memphis, TN, USA Barfield, Raymond, MD Assistant Member Department of Hematology/Oncology St. Jude Children’s Research Hospital Memphis, TN, USA Behm, Fred G., MD Associate Member and Vice Chair Department of Pathology St. Jude Children’s Research Hospital Memphis, TN, USA Bernstein, Irwin D., MD Head, Pediatric Oncology Program Fred Hutchinson Cancer Research Center Professor and Head Pediatric Hematology/Oncology Department of Pediatrics University of Washington School of Medicine Seattle, WA, USA Bhatia, Smita, MD, MPH Director, Epidemiology and Outcomes Research Division of Pediatrics City of Hope National Medical Center Duarte, CA, USA

ix

x

List of contributors

Biondi, Andrea, MD Director, M. Tettamanti Research Center Associate Professor Department of Pediatrics University of Milano-Bicocca Monza, Italy Brenner, Malcolm K., MB, BChir, Ph.D., FRCP, FRCPath Director, Center for Cell and Gene Therapy Professor, Departments of Medicine and Pediatrics Baylor College of Medicine Houston, TX, USA Campana, Dario, MD, Ph.D. Member Departments of Hematology/Oncology and Pathology St. Jude Children’s Research Hospital Professor, Department of Pediatrics College of Medicine University of Tennessee Health Science Center Memphis, TN, USA Cleveland, John L., Ph.D. Member, Department of Biochemistry St. Jude Children’s Research Hospital Memphis, TN, USA Creutzig, Ursula, MD Professor, Pediatric Hematology/Oncology ¨ University Children’s Hospital Munster ¨ Munster, Germany De Armendi, Alberto J., MD Chief and Member, Division of Anesthesiology St. Jude Children’s Research Hospital Memphis, TN, USA den Boer, Monique L., Ph.D. Associate Professor, Molecular Pediatric Oncology Head, Research Laboratory Pediatric Oncology Erasmus MC–Sophia Children’s Hospital University Medical Center Rotterdam Department of Pediatric Oncology and Hematology Rotterdam, the Netherlands Downing, James R., MD Member and Chair, Department of Pathology Scientific Director St. Jude Children’s Research Hospital Memphis, TN, USA

Eaves, Allen C., MD, Ph.D. Director Terry Fox Laboratory Vancouver, British Columbia, Canada

Eaves, Connie J., Ph.D. Deputy Director Terry Fox Laboratory Vancouver, British Columbia, Canada

Evans, William E., PharmD Director, St. Jude Children’s Research Hospital Professor, Department of Clinical Pharmacy College of Pharmacy University of Tennessee Health Science Center Memphis, TN, USA

Felix, Carolyn A., MD Associate Professor of Pediatrics University of Pennsylvania School of Medicine Attending Physician The Children’s Hospital of Philadelphia Abramson Research Center Philadelphia, PA, USA

Ferrando, Adolfo, MD, Ph.D. Assistant Professor of Pathology and Pediatrics Institute for Cancer Genetics Columbia University Irving Cancer Research Center New York, NY, USA

Flynn, Patricia M., MD Member, Department of Infectious Diseases Arthur Ashe Chair in Pediatric AIDS Research St. Jude Children’s Research Hospital Professor, Department of Pediatrics and Preventive Medicine University of Tennessee Health Science Center Memphis, TN, USA

Franklin, Jeremy A., MD Assistant Professor, Department of Pediatrics Division of Infectious Diseases Assistant Professor, Department of Pharmacy Practice Texas Tech University Health Sciences Center Amarillo, TX, USA

List of contributors

Gattuso, Jami S., RN, MSN, CPON Nursing Research Specialist Division of Nursing Research St. Jude Children’s Research Hospital Memphis, TN, USA Handgretinger, Rupert, MD, Ph.D. Director, Bone Marrow Transplantation Department of Hematology/Oncology St. Jude Children’s Research Hospital Memphis, TN, USA Hasle, Henrik, MD Associate Professor Department of Pediatrics Skejby Hospital, Aarhus University Aarhus, Denmark ¨ nter, MD Henze, Gu Professor and Director Pediatric Oncology/Hematology Charit´e-Campus Virchow Klinikum Augustenburger Platz Berlin, Germany Heslop, Helen E., MD Director, Adult Stem Cell Transplant Program The Methodist Hospital Professor, Center for Cell and Gene Therapy Baylor College of Medicine Houston, TX, USA Hinds, Pamela S., Ph.D., RN, CS Member and Director Division of Nursing Research St. Jude Children’s Research Hospital Memphis, TN, USA Howard, Scott C., MD Assistant Member Department of Hematology/Oncology St. Jude Children’s Research Hospital Memphis, TN, USA Hudson, Melissa M., MD Member, Department of Hematology/Oncology Director, After Completion of Therapy Clinic St. Jude Children’s Research Hospital Memphis, TN, USA

Ihle, James N., Ph.D. Investigator Howard Hughes Medical Institute Member and Chair Department of Biochemistry St. Jude Children’s Research Hospital Memphis, TN, USA Kishi, Shinji, MD, Ph.D. First Department of Internal Medicine Faculty of Medical Sciences University of Fukui Fukui, Japan Lange, Beverly, J., MD Yetta Dietch Novotny Professor in Clinical Oncology Medical Director of Pediatric Oncology Children’s Hospital of Philadelphia Philadelphia, PA, USA Langerak, Anton W., MD Department of Immunology Erasmus University Rotterdam Rotterdam, the Netherlands Le, Doan, MD, FRCPC Clinical Assistant Professor Alberta Children’s Hospital Calgary, Alberta, Canada Locatelli, Franco, MD Professor of Pediatrics Pediatric Hematology and Oncology IRCCS Policlinico San Matteo Pavia, Italy Look, A. Thomas, MD Professor of Pediatrics Harvard Medical School Vice Chair for Research Department of Pediatric Oncology Dana-Farber Cancer Institute Boston, MA, USA Lorsbach, Robert B., MD, Ph.D. Assistant Member, Department of Pathology St. Jude Children’s Research Hospital Memphis, TN, USA

xi

xii

List of contributors

Maclean, Kirsteen H., Ph.D. Research Fellow Department of Biochemistry St. Jude Children’s Research Hospital Memphis, TN, USA Magrath, Ian T., MB, FRCP, FRCPath Professor of Pediatrics Uniformed Services University of the Health Sciences Bethesda, MD Director International Network for Cancer Treatment and Research at Institut Pasteur Brussels, Belgium Mandrell, Belinda N., RN, MSN Pediatric Nurse Practitioner Patient Care Services St. Jude Children’s Research Hospital Memphis, TN, USA Mulhern, Raymond K., MD∗ Chief, Division of Behavioral Medicine St. Jude Children’s Research Hospital Memphis, TN, USA Navid, Fariba, MD Assistant Member Department of Hematology/Oncology St. Jude Children’s Research Hospital Memphis, TN, USA Niemeyer, Charlotte M., MD Professor of Pediatrics Department of Pediatrics and Adolescent Medicine Division of Pediatric Hematology and Oncology University of Freiburg Freiburg, Germany Onciu, Mihaela, MD Assistant Member, Department of Pathology Director, Hematology and Special Hematology Laboratories St. Jude Children’s Research Hospital Memphis, TN, USA Phipps, Sean, Ph.D. Associate Member Division of Behavioral Medicine St. Jude Children’s Research Hospital Memphis, TN, USA

∗

Deceased

Pieters, Rob, MD, MS, Ph.D. Professor and Head Department of Pediatric Oncology and Hematology Erasmus MC–Sophia Children’s Hospital University Medical Center Rotterdam Rotterdam, the Netherlands Pinkel, Donald, MD Adjunct Professor Biological Sciences Department California Polytechnic State University San Luis Obispo, CA, USA Pui, Ching-Hon, MD Member and Director Leukemia/Lymphoma Division St. Jude Children’s Research Hospital American Cancer Society–F.M. Kirby Clinical Research Professor Professor, Department of Pediatrics College of Medicine University of Tennessee Health Science Center Memphis, TN, USA Pul´e, Martin, MB, MRCP Postdoctoral Research Fellow Center for Cell and Gene Therapy Baylor College of Medicine Houston, TX, USA Raimondi, Susana C., MD Member and Director of Cytogenetics Department of Pathology St. Jude Children’s Research Hospital Memphis, TN, USA Razzouk, Bassem I., MD Associate Member Department of Hematology/Oncology St. Jude Children’s Research Hospital Memphis, TN, USA Relling, Mary V., PharmD Member and Chair Department of Pharmaceutical Sciences St. Jude Children’s Research Hospital Professor, Department of Clinical Pharmacy College of Pharmacy University of Tennessee Health Science Center Memphis, TN, USA

List of contributors

Ribeiro, Raul C., MD Member Department of Hematology/Oncology Director, International Outreach Program St. Jude Children’s Research Hospital Memphis, TN, USA Robison, Leslie L., MPH, Ph.D. Professor Division of Pediatric Epidemiology & Clinical Research Department of Pediatrics University of Minnesota Medical School Minneapolis, MN, USA Rooney, Cliona M., Ph.D. Professor Departments of Pediatrics and Molecular Virology and Microbiology Center for Cell and Gene Therapy Baylor College of Medicine Houston, TX, USA Ross, Julie A., MPH, Ph.D. Professor Division of Pediatric Epidemiology & Clinical Research Department of Pediatrics University of Minnesota Medical School Minneapolis, MN, USA Rubnitz, Jeffrey E., MD Associate Member Department of Hematology/Oncology Director of Fellowship Program St. Jude Children’s Research Hospital Memphis, TN, USA Sandlund, John T., MD Member Department of Hematology/Oncology St. Jude Children’s Research Hospital Memphis, TN, USA Santana, Victor M., MD Member and Director, Solid Tumor Division Department of Hematology/Oncology

St. Jude Children’s Research Hospital Memphis, TN, USA Shannon, Kevin, MD Roma and Marvin Auerback Distinguished Professor of Molecular Oncology Department of Pediatrics Program Leader, Hematopoietic Malignancies UCSF Comprehensive Cancer Center University of California San Francisco San Francisco, CA, USA Sievers, Eric L., MD Medical Director ZymoGenetics, Inc. Seattle, WA, USA Spector, Logan G., Ph.D. Assistant Professor Division of Pediatric Epidemiology & Clinical Research Department of Pediatrics University of Minnesota Medical School Minneapolis, MN, USA Turner, Victoria E., Ph.D., D (ABHI) Director, HLA Laboratory Department of Pathology St. Jude Children’s Research Hospital Memphis, TN, USA Tyc, Vida L., Ph.D. Associate Member Division of Behavioral Medicine St. Jude Children’s Research Hospital Memphis, TN, USA van Dongen, Jacques J. M., MD, Ph.D. Professor Department of Immunology Erasmus University Rotterdam Rotterdam, the Netherlands von Stackelberg, Arend, MD Pediatric Oncology/Hematology Charit´e-Campus Virchow Klinikum Berlin, Germany

xiii

Preface

The first edition of Childhood Leukemias was produced to meet a growing need for a comprehensive reference that would probe more deeply into important topics only touched upon by general or pediatric oncology texts. In the six years since its publication, I have been gratified by the positive responses of readers from around the globe, who have enthusiastically endorsed the book and have inquired about plans for a second edition. I was therefore pleased when Cambridge University Press asked me to undertake that task. In keeping with the basic editorial premise of the first edition – that readers will benefit most from a judicious blend of clinical and biological topics – this revised text includes eight new chapters. Three deal with the origins and challenges of chemoresistant leukemia, topics that are rapidly moving to the forefront of leukemia research. Two other chapters cover recent advances in therapy-related leukemias and in pain management, while the remaining three add much-needed discussions of the genetic syndromes predisposing to leukemia, treatment in countries with limited resources, and antibody-targeted therapy. The myelodysplastic syndromes, which represent a large fraction of the preleukemic diseases that continue to resist treatment, now occupy a single chapter, as do the chronic myeloproliferative disorders. As you will notice, the emphasis on molecular mechanisms that characterized the first edition has been retained, reflecting my conviction that the path to better patient management lies in serious inquiry into the pathobiology of specific diseases. Nowhere is this more evident than in the acute leukemias of childhood. As skillfully discussed in the chapters on the molecular genetics of acute lymphoblastic leukemia and acute myeloid leukemia, as well as host pharmacogenetics, improved methods of analysis have revealed previously unrecognized genotypes that could be used to increase the power and precision of clinical trials, and eventually to target therapy to tyrosine kinases

xv

xvi

Preface

Donald Pinkel, MD, first director of St. Jude Children’s Research Hospital.

and other key signaling molecules that lie upstream of pathological transcription programs. The development of imatinib for leukemias expressing the BCR-ABL chimeric protein provides a prototype for this direction in treatment innovation. One of the most difficult challenges was to present topics in a way that would appeal to students and trainees as well as experienced oncologists. Thus, all contributors were urged to strike a balance between a scholarly and a practical approach to their topics, and to take particular care in selecting illustrations that would adroitly highlight their major points of emphasis. They were also asked to avoid

technical jargon whenever possible, in favor of terms that would be meaningful to a general biomedical audience. I hope you will agree with me that this challenge has been successfully met, and that Childhood Leukemias is indeed accessible to a diverse readership. During the editing of the chapters, a concerted effort was also made to lessen inconsistencies in coverage. However, as with most books of this type, the final product reflects the unique perspectives of authors working in rapidly evolving fields, so that differences in opinion and interpretation are both frequent and legitimate. Publication of the second edition of Childhood Leukemias would not have been possible without the generous support of St. Jude Children’s Research Hospital, the American Lebanese Syrian Associated Charities (ALSAC), the American Cancer Society (FM Kirby Clinical Research Professorship), the Angel Grant Award from the National Cancer Coalition, and the National Cancer Institute (Cancer Center Core Grant CA-21765). I owe a debt of gratitude to John Gilbert for expert editorial consultation thoughout the preparation of this volume, to Julie Groff for superb scientific drawings, and to Kimberly Meshun Gayden for her excellent word-processing skills. Finally, my heartfelt thanks go to all the contributors – leaders in their respective fields who graciously donated their time so that readers could benefit from their vast knowledge and experience. The pediatric oncology community suffered a tragic loss with the recent untimely death of Dr. Ray Mulhern, whose groundbreaking contributions to the neuropsychology of children with cancer helped to propel this field into the modern era of clinical investigation. Ray was a brilliant author and a valued colleague and friend; I will miss him very much. One of the authors, Dr. Donald Pinkel, deserves special recognition for his pioneering contributions to the treatment of childhood leukemia, which taught us that acute lymphoblastic leukemia is not a death sentence but an imminently curable disease. His inspirational Total Therapy program at St. Jude greatly influenced my decision to pursue a career in leukemia research and ultimately to accept the challenges posed by the editorship of Childhood Leukemias.

Plate 2.2 ALL, L1 (FAB). Small blasts with indistinct nucleoli, with an admixture of some larger blasts. This spectrum of small and larger blasts is common in ALL. (Wright-Giemsa, ×1000.)

Plate 2.4 B-ALL (FAB ALL, L3) with the t(8;14). Blasts are characterized by intensely basophilic cytoplasm, regular nuclear features, prominent nucleoli, and cytoplasmic vacuolization. (Wright-Giemsa, ×1000.)

Plate 2.6 ALL with cytoplasmic granules. Fuchsia-colored granules are present in the cytoplasm of numerous blasts. Such granules may lead to a mistaken diagnosis of AML, but the granules are negative for MPO and myeloid-pattern SBB. Immunophenotyping will confirm a diagnosis of ALL, usually of precursor B-cell lineage. Granular ALL may display granular positivity for esterase stains. (Wright-Giemsa, ×1000.)

Plate 2.3 ALL, L2 (FAB). Blasts with prominent nucleoli and moderate amounts of cytoplasm, with an admixture of smaller blasts. Such cases overlap morphologically with AML and emphasize the importance of ancillary studies to assign the correct lineage in acute leukemia. (Wright-Giemsa, ×1000.)

Plate 2.5 ALL, L1 with prominent cytoplasmic vacuoles. Note the scant, lightly basophilic cytoplasm, and inconspicuous nucleoli (by comparison with Fig. 2.4). Such cases may be mistaken for B-ALL. Vacuolation is not unique to ALL, L3 (Burkitt) leukemia and other cytologic features have to be considered when making this diagnosis. (Wright-Giemsa, ×600.)

Plate 2.7 AML with minimal granulocytic differentiation (FAB M1). Blasts are large and somewhat irregular, with moderate amounts of cytoplasm but little cytoplasmic differentiation. (Wright- Giemsa, ×1000.)

Plate 2.8 Myeloperoxidase positivity in AML, demonstrated by yellow staining against a Romanowsky-stained background. (o-tolidine stain with dilute Giemsa counterstain, ×1000.)

Plate 2.10 Hypergranular acute promyelocytic leukemia (FAB M3, sometimes designated M3h). The neoplastic cells are abnormal hypergranular promyelocytes with reddish granules and occasional clefted nuclei. Several promyelocytes contain multiple Auer rods (so-called “faggot cells”). (Wright-Giemsa, ×1000.)

Plate 2.12 Acute myelomonocytic leukemia (FAB M4). The leukemic cell population includes large blasts, with irregular and reniform nuclei, promonocytes and monocytes. Esterase staining is often positive in such cases. (Wright-Giemsa, ×1000.)

Plate 2.9 AML with granulocytic differentiation (FAB M2). Differentiating granulocyte precursors are admixed with myeloblasts. Several blasts contain Auer rods (arrows). (Wright-Giemsa, ×1000.)

Plate 2.11 Microgranular acute promyelocytic leukemia (FAB M3v). The leukemic process is characterized by cells with bilobed and grooved nuclei, and sparse cytoplasmic granulation. (Wright-Giemsa, ×1000.)

Plate 2.13 Acute monoblastic leukemia (FAB M5). Blasts are large and uniform, with abundant blue-gray cytoplasm, and may have cytoplasmic vacuolation and amphophilic granules. (Wright-Giemsa, ×1000.)

Plate 2.14 Alpha naphthyl butyrate esterase reactivity in acute monoblastic leukemia. ANB positivity is characterized by intense, diffuse reddish-brown cytoplasmic staining, typical of monoblastic leukemia. (ANB stain with hematoxylin counterstain, ×1000.)

Plate 2.15 AML with predominant erythroid differentiation (FAB M6, also termed M6a). An infiltrate of myeloblasts is present admixed with dysplastic erythroid precursors. (Wright-Giemsa, ×1000.)

Plate 2.16 Acute erythroblastic leukemia (also termed FAB M6b). The blasts are large with basophilic cytoplasm, resembling normal erythroblasts, and may show vacuolization. Immunophenotypic analysis confirms erythroid differentiation of the blasts. (Wright-Giemsa, ×1000.)

Plate 2.17 Acute megakaryoblastic leukemia (FAB M7). The blasts have prominent surface blebs, bi- or multinucleation, and may occasionally form cohesive clusters, mimicking metastatic tumor. (Wright-Giemsa, ×1000.)

Plate 2.18 AML with t(8;21). Blasts are large, with basophilic cytoplasm and single needle-shaped Auer rods. Many dysplastic granulocyte precursors are present, some showing diffuse salmon-pink cytoplasmic staining. (Wright-Giemsa, ×600.)

Plate 2.19 AML with inv(16) (FAB AML M4Eo). Acute myelomonocytic leukemia with numerous dysplastic eosinophils that contain coarse basophilic cytoplasmic granules. (WrightGiemsa, ×600.)

Plate 2.20 Dimorphic T/myeloid acute biphenotypic leukemia. A dual leukemic cell population is present, including small blasts with scant cytoplasm, indistinct nucleoli, ‘hand-mirror’ shape, and a small number of larger blasts with apparent myeloid differentiation. Immunophenotypic analysis shows that the blasts uniformly express T-cell-associated (CD2, CD3, CD7) and myeloid-associated (CD13, CD33, MPO) antigens. (Wright- Giemsa, ×1000.)

Plate 2.21 Bilineal acute leukemia. This leukemic process consists of two morphologically and immunophenotypically distinct blast populations: a lymphoid population resembling ALL, L1 and a monoblastic population resembling AML, M4/M5. (Wright-Giemsa, ×600.)

Genes for class distinction (n=588)

Diagnostic BM samples (n=132)

E2APBX1 −3SD

MLL

T-ALL

HD>50

BCRABL

TEL-AML1

+3SD

Plate 2.22 Expression profile of pediatric ALL diagnostic bone marrow blasts. Two-dimensional hierarchical cluster of 132 pediatric ALL diagnostic bone marrow samples; the normalized expression value of each gene is indicated by a color (red, expression above the mean; green, expression below the mean). (From Ross et al.221 Copyright American Society of Hematology, used with permission.)

Plate 9.1 Spectral karyotyping (SKY) analysis of leukemic cells from a patient with AML-M1. G-banded karyotype of bone marrow showed the following abnormalities: 47,XX,add(5)(q33), add(7)(p22), +20[18]/46,XX[2]. SKY established two unbalanced translocations: 47,XX,der(5)t(5;6)(q33;q21), der(7)t(7;13)(p22;q14),+20. Top panel, spectral image; bottom panel, classified image.

Plate 9.2 FISH detection of an MLL rearrangement in a pediatric patient with AML and a t(9;11)(p22;q23). Chromosomes were hybridized with Spectrum Orange- and Spectrum Green-labeled DNA probes homologous to sequences lying telomeric and centromeric to the MLL gene. In the nuclei of normal cells (not shown), hybridization of these two probes produces signals that either overlap (yellow) or are close to one another. In metaphase chromosome spread (A) and interphase nuclei (B) in which the MLL rearrangement is present, the target sequences are “split” (i.e. distant) because of their translocation to different chromosomal positions (C). (D) Partial karyotype. This translocation is predominantly seen in primary and therapy-related AML M5; it is rarely seen in ALL.

Plate 9.3 Partial karyotype of B-lineage ALL blast cells in which the t(12;21)(p13.3;q22) is present. The metaphase chromosomes were stained with 4 , 6-diamidino-2-phenylindole (DAPI). This chromosomal abnormality is the most common recurrent translocation in patients with ALL and is readily detected by FISH (A) and RT-PCR but not by conventional cytogenetic methods. The ETV6 gene on chromosome 12 (green) and the CBFA2 gene on chromosome 21 (red) probes used in FISH detected ETV6-CBFA2 on the der(21)t(12;21) (B, right) and residual of CBFA2 on the der(12)t(12;21) (B, left).

Plate 16.2 Leukemic involvement of the anterior segment with hypopyon in a 9-year-old boy with relapsed ALL.

Plate 16.1 Leukemic retinopathy manifested by disc edema and white-center hemorrhage in an 11-year-old girl with t(4;11)positive, early pre-B ALL and a presenting leukocyte count of 1512 × 109 /L.

Plate 16.4 Enlargement of right parotid gland due to infiltration of leukemic cells at diagnosis in a 5-year-old boy with near-haploid early pre-B ALL.

Plate 16.3 Left testicular relapse in a 12-year-old boy with T-cell ALL.

Plate 16.8 Results of leukapheresis in a 16-year-old boy with T-cell ALL and a presenting leukocyte count of 350 × 109 /L. The cylinders shown contain 2.8 × 1012 leukemia cells, illustrating well the definition of the term leukemia (white blood); the pink tint represents red cells that have not settled out.

Plate 16.6 Telangiectasis on bullar conjunctiva in a 12-year-old girl with ataxia telangiectasia and B-cell ALL.

Plate 16.10 Giant pronormoblast (50 m in diameter) with deeply basophilic cytoplasm, fine chromatin, and a prominent large nucleolus in a 6-year-old girl with ALL and parvovirus B19 infection.

Plate 16.9 Facial maculopapular erythematous skin rash during continuation treatment with mercaptopurine and methotrexate.

Plate 18.1 Histologic section of Burkitt lymphoma. Photo kindly provided by Frederick G. Behm.

Plate 18.3 African child with endemic Burkitt lymphoma of the jaw. Plate 19.4 Gingival hypertrophy in a patient with monoblastic leukemia.

Plate 19.5 Leukemia cutis in congenital myeloid leukemia with the t(9;11).

Plate 21.2 Cytological features of myelodysplasia. Courtesy of Irith Baumann.

Plate 21.4 Transient abnormal myelopoiesis with skin infiltration in Down syndrome. The boy presented at 1 day of age with a WBC of 110 × 109 /L. The blood was dominated by megakaryoblasts. During the following 2 weeks there was increasing hepatomegaly and skin infiltration by myeloid cells. Cytarabine (75 mg/m2 per day) was administered subcutaneously for 4 days. The skin infiltrate disappeared and the WBC normalized within a week.

Plate 23.7 Graft-versus-host disease tends to appear first as a pruritus or erythema on palms (a), soles or ears. Next, a maculopapular rash (b, c) may progress to a total-body erythroderma (d, e).

Plate 23.8 Extensive chronic GVHD in a patient after matchedsibling transplantation. The skin is the most frequently involved organ with hyper- or hypogigmentation, desquamation and a picture similar to scleroderma, including joint contractures. Other features include dysphagia and indolent weight loss. (Photo kindly provided by Dr. G. Vogelsang, John-Hopkins University, Baltimore, MD, USA.)

Plate 29.4 Linear streaks of precipitated uric acid (arrows) in the renal medulla of a 4-year-old boy who died of massive tumor lysis syndrome.

Plate 29.7 Superior vena cava syndrome with venous engorgement of the neck and arms and development of collateral blood vessels in the trunk of a 10-year-old boy with T-cell ALL and a mediastinal mass.

Plate 29.11 Leukoencephalopathy. These magnetic resonance images of a 3.5-year-old girl with acute lymphoblastic leukemia show normal T2 -weighted, fluid attenuation inversion recovery (FLAIR), and color-mapped images at the end of remission induction therapy (upper left, middle, and right panels, respectively). Marked white matter changes (arrows) are seen after four cycles of high-dose methotrexate (lower panels). The patient had no symptoms and a normal neuropsychological evaluation at the time of these studies. She is now 7 years old and has mild deficits in reading comprehension and mathematics, but functions at grade level and has no neurologic deficits. The color-mapped images were generated with digital imaging processing software.340 Yellow denotes normal gray matter; blue denotes cerebrospinal fluid; green denotes normal white matter; and red denotes abnormal white matter. (Courtesy of Dr. Gene Reddick.)

B

A

C

D

Plate 32.1 (A) Target lesions in Pseudomonas aeruginosa sepsis. (B) Ecthyma gangrenosum due to P. aeruginosa. (C) Cutaneous lesions associated with molluscum contagiosum. (D) Cutaneous lesions associated with disseminated candidiasis.

Part I History and general issues

1 Historical perspective Donald Pinkel

Introduction Since its initial recognition 150 years ago, leukemia has been the focus of remarkable research activity and consequent progress. The drama of its manifestations, its frequency in children, its commercial importance in animal husbandry, its usefulness in understanding hematopoiesis, and its ready adaptability as a model for other human cancers are among the reasons for this attention. But perhaps more important for the current generation of its students was the discovery 30 years ago that the most common variety of leukemia could be cured in approximately one-half of children, the first generalized cancer to be cured and the first autologous cancer to be cured with chemicals.1 This chapter summarizes the history of the study of leukemia, particularly childhood leukemia, with regard to description, causation, and treatment. It concludes with comments about the lessons taught by this history.

Description of leukemia Although the first description of a patient with leukemia was published in 1827,2 it was not until 1845 that Virchow3 in Germany (Fig. 1.1) and Bennett4 and Craigie5 in Scotland, in separate case reports, recognized it as a distinct disease, “white blood.” Two years later, Virchow introduced the term “leukemia” for this entity and proceeded on a series of investigations that were summarized in 1856.6 He distinguished leukemia from leukocytosis and described two types: splenic, associated with splenomegaly, and

lymphatic, associated with large lymph nodes and cells in the blood resembling those in the lymph nodes. He also proposed his cellular theory of the origin of leukemia, a concept basic to current understanding of the disease. The following year, acute leukemia was described by Friedreich,7 and in 1878 Neumann8 established the existence of myelogenous leukemia. The close relation between lymphomas and leukemias was defined by Turk9 in 1903. Ehrlich’s introduction of staining methods in 1891 allowed the differentiation of leukocytes and identification of leukemia cell types.10 Splenic and myelogenous leukemias were soon recognized as the same disease, originating from a myeloid precursor. Eventually the leukemic myeloblast, monoblast, and erythroblast were identified. It also became apparent that some acute leukemias were marked only by abnormal leukocytes in the blood, not leukocytosis. By 1913, leukemia could be classified as chronic lymphocytic, chronic myelogenous, acute lymphocytic, myeloblastic or monocytic, or as erythroleukemia.11 Not only did these advances result in refined classification of leukemia, but they shed light on the nature of normal hematopoiesis as well. The prevalence of acute leukemia during childhood, especially between ages 1 and 5 years, was noted in 1917.12 Progress in the description of leukemia has continued to parallel the development of new technologies, such as special staining, electron microscopy, chromosomal analysis, immunophenotyping, and molecular genotyping. With use of electron microscopy, platelet peroxidase staining, and monoclonal antibody reactivity to a platelet glycoprotein, CD41, acute megakaryocytic leukemia became a welldefined entity.13 Although some hematologists and many

C Cambridge University Press 2006. Childhood Leukemias, ed. Ching-Hon Pui. Published by Cambridge University Press.


3

4

Donald Pinkel

Fig. 1.2 Luis Borella identified T-cell leukemia, introduced immunophenotyping of leukemia, and initiated its classification by biological function in addition to morphology. Fig. 1.1 Rudolf Virchow, the father of leukemia research, established leukemia as a medical entity in the years 1845 to 1856. He also classified leukemia by its pathologic anatomy and cell morphology and postulated its cellular origin.

chemotherapists lumped all childhood acute leukemias into one category as late as the 1960s, the discovery that acute lymphoid and acute myeloid leukemias (ALL and AML) responded differently to prednisone and methotrexate made it necessary to use the new technologies to clearly distinguish them. After the discovery in 1960 of the Philadelphia chromosome in adult chronic myeloid leukemia, and the later introduction of banding techniques, many nonrandom chromosomal abnormalities were found to be associated with specific types of acute leukemia.14,15 Application of DNA probing and amplification methods resulted in molecular genotyping of leukemias, both for diagnosis and for detection of residual cells of the leukemia clone.16 It also became possible to use archived neonatal Guthrie blood spots to trace back the fetal origin of many childhood leukemias.17–24

In 1973, Borella and Sen25 (Fig. 1.2) demonstrated that in some children with acute lymphoid leukemia, the leukemic lymphoblasts were of thymic origin. They further showed that T-cell leukemia was clinically as well as biologically unique.26 As monoclonal antibodies to leukocyte cell surface antigens were developed, further immunophenotypic classification of leukemia cell populations became possible.27 Currently, leukemia is classified as acute or chronic, lymphoid or myeloid, as in the 19th century (see Chapter 2). However, the morphology of acute leukemia is subclassified into three lymphoid varieties and eight myeloid. Myelodysplastic syndromes such as monosomy 7 syndrome and juvenile myelomonocytic leukemia are also recognized. Immunophenotyping of leukemia cells with monoclonal antibodies separates the lymphoid lineage into early and late B-precursor, B-cell, and T-cell (see Chapter 7). It also helps to distinguish anaplastic lymphoid from myeloid cell types and to classify the eight myeloid types, and contributes to identifying the rare biphenotypic variety. Genotypic classification by chromosomal analysis, fluorescent in situ hybridization, DNA probing, and

Historical perspective

polymerase chain reaction techniques allows molecular genetic definition of leukemias (see Chapters 9, 10, and 11). Because leukemia is now recognized as a molecular genetic disorder, and the most effective acute leukemia drugs disrupt molecular genetic processes, this approach to cell characterization may be the ultimate descriptive method. With use of recent technology, it has become clear that the most frequent form of acute leukemia in children is B-precursor cell, often with excessive chromosomes or expression of novel hybrid genes such as ETV6-CBFA2 (TEL-AML1), E2A-PBX1, or BCR-ABL (190 kb) and, in young infants, often demonstrating rearrangement of the MLL (HRX) gene.28–30 Recently, the World Health Organization published a new classification of leukemia based on the advice of numerous experts.31,32 Whether its complexity will be justified by more precise diagnosis, better understanding and improved prognosis is uncertain. During the past 30 years, the importance of describing the leukemia host has also become more apparent. Not only such features as age, gender, and disease extent, but also ethnicity, nutrition, socioeconomic status, and accompanying syndromes and diseases, have been correlated with type of leukemia and outcome of treatment.33–39 For example, children with trisomy 21 (Down) syndrome have a high incidence of leukemia, especially acute megakaryocytic leukemia.38 They also have twice the cure rate of other children with acute myeloid leukemia when treated with chemotherapy.39 The extra 21 chromosome introduces not only increased vulnerability but also better curability. Hispanic youngsters have a high frequency of acute promyelocytic leukemia.40 Host genetic polymorphisms with regard to enzymes such as thiopurine methyltransferase that make available, activate, or detoxify antileukemic drugs are important.41,42 Genetic polymorphisms may also play a role in susceptibility to leukemia among persons exposed to environmental leukemogens or prone to dietary deficiency of folic acid.43,44 Malnutrition, poverty, and underprivileged ethnicity are associated with low cure rates.33–37 In summary, the history of the past 150 years illustrates that progress in the comprehension of leukemia has paralleled the continued application of new ideas and technology to this disease by creative, industrious, and practical clinical investigators.

Causation of leukemia The search for the causation of leukemia has followed several approaches: infectious, genetic, physical, and chemical. Pursuit has been vigorous and often marked by heated controversy. Over time it has become apparent that all

approaches may be correct and that leukemia results from numerous causes, often interacting and varying from cell type to cell type and from one patient to another. Recent studies suggest that childhood leukemia is initiated during fetal life. Rearrangements of either leukemia-associated genes or immunoglobulin heavy-chain genes in childhood leukemia cells have been identified retrospectively in stored neonatal Guthrie blood spots.17–24 However, the frequency of leukemia-associated gene rearrangments, such as TEL-AML1, in surveys of blood spots far exceeds the incidence of childhood leukemia. This indicates that the gene rearrangement alone is insufficient to cause leukemia. Other factors must be contributory.

Infectious causes When “white blood” was identified, some observers considered it the result of severe inflammation, but the new technology of blood microscopy revealed that the white cells of leukemic leukocytosis appeared different from those of inflammatory leukocytosis. However, interest continued in an infectious etiology. Ellerman and Bang’s45 transmission of fowl leukemia by cell-free extracts in 1908, suggesting a viral causation, was a landmark finding that led to extensive searches for the virus etiology of all leukemias, both in animals and humans, throughout the 20th century. In 1951, a mammalian leukemia virus was first demonstrated by Gross46 (Fig. 1.3) by injection of newborn mice with cell-free filtrates from leukemic mice. Subsequently, several leukemia-producing viruses were isolated from cats, cattle, gibbon apes, and humans with adult-type T-cell leukemia.47–50 All were characterized as retroviruses. These single-stranded RNA viruses produce DNA polymerase and integrase, which reverse transcribe the viral RNA genome to DNA and integrate it into the cellular genome. This can result in neoplastic transformation of the cell with or without virus production. In addition, two large DNA viruses of the herpes group were associated with leukemia: Marek disease virus in birds and Epstein–Barr virus (EBV) in B-cell lymphoma/leukemia of African children (Burkitt lymphoma).51,52 Since both EBV-positive and EBV-negative B-cell lymphoma/leukemia have comparable gene rearrangements and postulated mechanisms of leukemogenesis, it is doubtful that the virus is causative.53 Extensive attempts to identify leukemia viruses in children with B-precursor, T-cell, myeloid, and temperate zone Bcell leukemia have been unsuccessful.54 However, the critical experiments that led to identification of murine and feline leukemia viruses, injection of newborn of the same species, cannot be performed.

5

6

Donald Pinkel

Northumberland and Durham, United Kingdom,57–60 but study elsewhere has failed to confirm significant aggregation or other evidence of communicability.61,62 Also cited to support the infection hypothesis was the lower incidence and younger age of acute leukemia in children of lower income families.57 It was speculated that this could fit the pattern of infectious diseases such as paralytic poliomyelitis, in which early exposure and maternal immunity contribute to the appearance of disease at an earlier age and less frequently in underprivileged children. More recently, Kinlen and colleagues63 described excessive leukemia and non-Hodgkin lymphoma rates in children living near large rural construction sites. They suggested that the high risk was related to unaccustomed mixing of rural and urban people and was evidence for an infectious process. Greaves and associates64,65 have further modified and expanded Kellett’s hypothesis based on a newer understanding of the biology of childhood leukemia and international epidemiologic data. In summary, infectious causation of childhood leukemia remains only a hypothesis.

Physical causes

Fig. 1.3 Ludwik Gross described the first mammalian leukemia virus in 1951, initiating research efforts that led to study of the molecular pathology of leukemia.

Despite the failure to identify causative leukemia viruses in children with leukemia, some epidemiologic characteristics have been interpreted in favor of an infectious cause. In 1917, Ward12 reviewed 1457 cases of acute leukemia and concluded that the weight of evidence was against infection. In 1942, Cooke55 collected information on children with acute leukemia from 33 American pediatric services (a harbinger of pediatric cooperative studies) and demonstrated a sharp peak in incidence between ages 2 to 5 years, paralleling peaks in measles and diphtheria incidence. He concluded that acute infections were a factor in causing childhood leukemia. Lending weight to an infection hypothesis was the report by Kellett56 in 1937 of a concentration of cases in Ashington, England. He suggested that an infection, possibly widespread but of low infectivity, might be the causative agent. Subsequent instances of temporospatial proximity of children with leukemia were reported from Erie County, New York; Niles, Illinois; and

Although ionizing radiation probably induced leukemia in Marie Curie, its leukemogenic effects in radiologists only became quantified in 1944.66 In 1952, studies of Japanese children who survived atomic bombing demonstrated a marked increase in acute leukemia, both lymphoid and myeloid.67 Subsequently, Simpson et al.68 reported that children who received neonatal thymic irradiation had an increased risk of thymic lymphoma and acute leukemia as well as thyroid carcinoma. Numerous subsequent studies of prenatal and childhood exposure to diagnostic radiography and medical radiation for benign disease yielded evidence that low-dose radiation can be a factor in the causation of childhood leukemia.69,70 The most recent evidence suggests that low-dose radiation induces a transmissable genetic instability in hematopoietic stem cells.71 This results in diverse chromosomal aberrations in their progeny many cell divisions later. Action was taken in the 1960s and 1970s to reduce fetal, neonatal, and childhood exposure to ionizing radiation. Medical radiation for neonatal thymus, tinea capitis, acne, benign tumors, and even some malignancies was eliminated. Shoe store fluoroscopes were removed, medical and dental radiology equipment and protection upgraded, and diagnostic radiography, especially by fluoroscope, was reduced or replaced with ultrasound imaging. However, as long as nuclear weapons continue to exist, radiation remains a potential cause of leukemia.

Historical perspective

Chemical causes In 1928, Delore and Borgomano72 reported a patient with acute leukemia associated with benzene intoxication. Subsequently, numerous reports confirmed that benzene can produce myelodysplasia and acute myeloid leukemia.73,74 A dose–response relationship was recently found in China.75 Although the hazards have been occupational and the victims adults, the significant yield of benzene in cigarette smoke – three times greater in sidestream than in mainstream smoke – and in automobile exhaust raises the question of whether parental smoking and automobiles are causative factors of leukemia in children.76 Smith has proposed that the phenolic metabolites of benzene are converted to quinones that produce DNA strand breaks, topoisomerase II inhibition and mitotic spindle damage in hematopoietic cells.77 In recent years folic acid deficiency has become associated with the causation of childhood leukemia. An unconfirmed case control study in Australia78 suggested a protective effect of maternal folate supplementation against the risk of childhood B-precursor ALL. In both children and adults, genetic polymorphism of 5,10– methylenetetrahydrofolate reductase, resulting in loss of this enzyme’s activity, appears to reduce the risk of some forms of ALL.44 The suggested mechanism is the increased availability of methyl groups from the folate cycle for conversion of uracil to thymine. This reduces the possibility of uridine incorporation into DNA and consequent genomic instability. Transfer of methyl groups by way of the folic acid cycle is essential to purine synthesis and the suppression of untimely gene expression as well as the methylation of uracil to form thymine. Defects in the folic acid cycle produced by dietary deficiency, impaired absorption or transport, antifolate agents, genetic polymorphism or exposure to nonphysiologic methylating agents, such as the pesticide methyl bromide, might contribute to the pathogenesis of leukemia. The advent of cancer chemotherapy in the 1950s and its extension in the 1960s and 1970s led to the appearance of secondary leukemia both in children and adults. Alkylating agents and drugs that bind topoisomerase II, especially etoposide and teniposide, were found to be leukemogenic in children, most often producing acute leukemia characterized by MLL gene fusions.79,80 This observation of the role of topoisomerase binding is consistent with the Smith hypothesis77 for the mechanism of benzene leukemogenesis. A recent study demonstrated that children who had acute leukemia with MLL fusion genes were more likely to have low function of an enzyme that detoxifies quinones.43 Another study revealed an association between

this leukemia genotype and maternal exposure to certain drugs and pesticides.81 These data suggest that both maternal exposure to potential leukemogens and fetal genetic polymorphisms might contribute to the induction of childhood leukemia.

Genetic causes A genetic cause of leukemia was first suggested in 1876 by Hartenstein,82 who observed lymphoid leukemia in a cow and its mother and speculated that it was hereditary. In 1931, strains of mice with high frequencies of leukemia/lymphoma were identified,83 and by 1935 an inbred strain with a 90% incidence of lymphoid leukemia was produced.84 Extrinsic nonhereditary factors were postulated to explain the 10% failure of this inbred strain to develop leukemia. The evidence for a possible genetic basis of murine leukemia led to studies of the familial incidence of human leukemia. A 1937 report85 of three families with multiple cases was followed by a large study by Videbaek86 in Denmark comparing families of patients with leukemia and families of healthy persons. A significant difference was found and a genetic hypothesis proposed. An institution-based study in Boston in 195787 did not support Videbaek’s findings, but the author acknowledged three families with multiple cases of acute leukemia, two with parental consanguinity, and suggested a recessive gene in these families. Although leukemia in twins was described in 1928,88 the high concordance rates for leukemia in likesex and monozygous twins were uncovered in 1964 by MacMahon and Levy.89 Recent studies by Ford et al.18 using genetic markers indicate that twin concordance probably results from intrauterine metastases from fetus to fetus. In addition to increased familial incidence and twin concordance, the increased risk of leukemia in children with constitutional chromosome abnormalities further supported a genetic hypothesis. The report of a child with Down syndrome and acute lymphoid leukemia in 193090 and subsequent similar reports led to a national survey in 1957 by Krivit and Good38 that demonstrated the high incidence of leukemia in this trisomy disorder. Over the past 40 years, childhood leukemia has become associated with numerous constitutional genetic disorders, including primary immunodeficiency diseases, chromosome instabilities, and inherited cancer syndromes.91 Observation of the distinct Philadelphia chromosome associated with chronic myeloid leukemia by Nowell and Hungerford14 in 1960, and Rowley’s discovery15 that it resulted from a 9;22 chromosomal translocation in 1973,

7

8

Donald Pinkel

were followed by identification of numerous nonrandom chromosomal abnormalities associated with biologically distinct leukemias and hybrid genes. In 1982, the human homologue of the Abelson murine leukemia virus protooncogene abl was found to be relocated from chromosome 9 to 22 in chronic myeloid leukemia, to form its characteristic hybrid gene, BCR-ABL.92 In the same year the human homologue of an avian leukemia oncogene (MYC) was identified on the region of chromosome 8 that is translocated in B-cell lymphoma/leukemia of children.93 By the mid-1980s, there was a clear consensus that leukemia was a somatic genetic disorder of hematopoiesis.94 More important, these translocations became models of the two general mechanisms of leukemogenesis by chromosome/gene rearrangements. The BCR-ABL hybrid gene gives rise to a BCR-ABL fusion protein with excessive and promiscuous tyrosine kinase activity.95 This leads to the activation of myriads of proteins along several signaling pathways and reduced cell adhesion, increased mitoses and inhibition of apoptosis – conditions favorable to leukemogenesis, either chronic myeloid or acute lymphoblastic. The second mechanism is exemplified by the translocation of the MYC oncogene of chromosome 8 to the immunoglobulin heavy-chain region of chromosome 14.96 The consequence is remarkably increased expression of the MYC gene, whose translation product dimerizes with the normal MAX protein. This drives cell replication at the expense of differentiation. B-cell lymphoma and/or leukemia results. Although the ultimate causation of most childhood leukemias remains unknown, the establishment of a genetic mechanism, recognition of the role of homologues of animal leukemia virus oncogenes in human leukemia cells, and the knowledge that ionizing radiation and chemical leukemogens modify genetic DNA appear to reconcile the four historical approaches to causation. The more recent insights about genetic polymorphisms, folic acid and the consequences of leukemia-associated gene rearrangements have introduced new potentials for the prevention and treatment of childhood leukemias.

Treatment Palliative treatment Because of the diffuse nature of leukemia and its catastrophic manifestations, physicians began to treat patients with chemicals shortly after it became recognized as a disease entity. In 1865, Lissauer97 reported a patient with leukemia whose disease remitted after she received Fowler solution (arsenious oxide); arsenicals became a standard but marginally useful palliation. With the discovery of

Fig. 1.4 Sidney Farber and his colleagues discovered that a synthetic antifolate, 4-amino-pteroylglutamic acid, produced remissions of childhood leukemia. This introduced antimetabolite chemotherapy and began the research leading to a cure for many children with leukemia.

roentgen rays in 1896, interest turned to their clinical application in cancer therapy. In 1903, Senn98 reported the response of leukemia to irradiation, and this modality, applied most often to the spleen, largely replaced arsenious oxide as a palliative measure, especially in chronic leukemia. When radioactive nuclides became available in 1940, radioactive phosphorus came into use for chronic myelogenous leukemia and polycythemia vera.99 Based on pathology reports of hematosuppression in mustard gas victims on the Western Front in World War I100 and at the Bari Harbor disaster in World War II,101 nitrogen mustard was synthesized and tested in animals and then patients with lymphoma and leukemia in 1943.102,103 Temporary partial remissions were produced, but toxicity was considerable, especially in patients with acute leukemia. The chemical identification of folic acid in 1941104 as an essential vitamin, its synthesis in 1946,105 and the reversal of megaloblastosis by its administration106 raised the question of whether it might be useful in the treatment of acute leukemia. In 1947, Farber (Fig. 1.4)107,108 and

Historical perspective

use in sequential and combination chemotherapy with a corticosteroid (usually prednisone) and methotrexate, the 4-amino-N10 -methy1-folate analogue that succeeded aminopterin.108 The enthusiasm generated by the discovery of three effective drugs for childhood acute leukemia in 5 years was dampened, however, by the realization that virtually all of the patients eventually died of resistant leukemia or its complications.108 This led to a fixed notion among most pediatricians and hematologists that temporary remissions and prolongation of survival in comfort were the most one could expect from leukemia chemotherapy. In 1959, a prodrug analogue of nitrogen mustard, cyclophosphamide, with less toxicity for platelet production, was introduced and later shown to have value in lymphoid leukemia.113 In 1962, vincristine, an alkaloid from the periwinkle plant with a unique mode of action, was shown to induce complete remissions of childhood lymphoid leukemia resistant to other agents.114 But, as with all the other agents, remissions were temporary and relapse with resistant leukemia ensued.

Fig. 1.5 Gertrude Elion, working with George Hitchings, used an understanding of purine metabolism to develop three drugs important to children with leukemia: mercaptopurine, allopurinol, and acyclovir.

colleagues gave folic acid (pteroylglutamic acid) to children with acute leukemia and were impressed that it might have produced acceleration of the leukemia. Subsequently, a 4-amino antimetabolite of folic acid, aminopterin, synthesized by Seeger et al.,109 was provided to Farber for use in children with acute leukemia. Many of them achieved complete clinical and hematologic remissions that lasted for several months.107 The era of specific leukemia therapy had begun! A year after the report of remissions with aminopterin, a 1949 conference on the newly isolated adrenocorticotrophic hormone (ACTH) revealed that it produced prompt although brief remissions of acute lymphoid leukemia.110 Cortisone and its synthetic analogue, prednisone, had similar activity and soon replaced ACTH. Unlike the folate antagonists, the purine antimetabolites 6-mercaptopurine and thioguanine resulted from a lengthy study of purine metabolism, purine analogue synthesis, and structure-activity relationships by Elion and Hitchings111 (Fig. 1.5) in the 1940s and early 1950s. In 1953, a report by Burchenal and associates112 that 6mercaptopurine produced remissions in patients with acute leukemia, especially children, promptly led to its

Curative therapy The first cure of leukemia was described in 1930 by Gloor,115 who treated an adult with arsenious oxide, mesothorium, irradiation, and blood transfusions from two siblings (presaging current myeloblation and peripheral blood stem cell transplantation?). In 1964, Burchenal and Murphy116 collected 36 cases of 5-year cures of treated childhood acute leukemia by a questionnaire survey of hematologists. Zuelzer117 reported a 3% 5-year cure rate in children with ALL who received cyclic chemotherapy with prednisone, methotrexate, and mercaptopurine. A 5% 5-year cure rate was reported by Krivit et al.118 for sequential or cyclic chemotherapy of ALL with these agents in a Children’s Cancer Group study. Stimulated by the studies of Skipper et al.119 and Goldin et al.120 in treating mouse leukemia with chemotherapy, Leukemia Study Group B121–123 used two-drug combinations and National Cancer Institute investigators124,125 used four-drug combinations that yielded similar low cure rates in patients with ALL. The failure to achieve a significant cure rate in these courageous attempts reinforced the prevailing pessimism about leukemia therapy. Persons who continued to advocate anything beyond palliation were looked upon with skepticism, if not scorn, into the early in 1970s. In 1962, St. Jude Children’s Research Hospital was opened in Memphis, Tennessee, with a mandate to seek prevention or cure of childhood leukemia. The St. Jude investigators defined several specific obstacles to the cure of

9

10

Donald Pinkel

childhood acute leukemia.94 First was drug resistance: initial, as demonstrated by the high proportion of patients who failed to experience remission on single-drug treatment; and acquired, as indicated by eventual relapse in most children despite continued drug administration. The second obstacle was clinically isolated meningeal relapse that occurred with increasing frequency as systemic chemotherapy became more effective and hematologic remissions lasted longer. Meningeal relapse was thought to be due to the inadequate diffusion of methotrexate and mercaptopurine through the blood–cerebrospinal fluid barrier with consequent proliferation of leukemia cells in the leptomeninges. The third obstacle was the overlapping toxicity of antileukemic drugs, especially hematosuppression, immunosuppression, and mucositis, and thus the dilemma of limiting dosage or risking treatment-related death. However, the greatest obstacle was a pessimism that inhibited thoughts of curing patients with leukemia. A curative approach to children with ALL was initiated in 1962. It consisted of four treatment phases: remission induction, intensification or consolidation, preventive meningeal treatment, and prolonged continuation therapy.94,126–128 The main features were the administration of combination chemotherapy for induction, intensification and continuation chemotherapy, the use of different drug combinations for induction and continuation, preemptive irradiation of the cranial or craniospinal meninges, elective cessation of chemotherapy after 2 to 3 years, and most important, the objective of cure rather than palliation. The pilot studies from 1962 to 1965 were fraught with considerable difficulty, including the emergence of Pneumocystis carinii pneumonia due to immunosuppression and the inadequacy of low-dose craniospinal irradiation to prevent meningeal relapse.126–128 However, longer complete remissions were achieved than previously and 7 of 41 children became long-term leukemia-free survivors after cessation of therapy, a higher rate than previously reported, justifying the notion that acute leukemia could no longer be considered incurable. A fourth study129 compared full versus half-dosage continuation chemotherapy and demonstrated that, despite its toxicity, full dosage was required to achieve longer remission. It was clear from this experience that more capability in prevention and control of infection, especially with Pneumocystis carinii and the herpesviruses, was required. With this information, another pilot study1 was inaugurated in December 1967, in which the intensity of continuation chemotherapy was increased and higher-dose cranial irradiation combined with intrathecal methotrexate was used to treat the leptomeninges. Within 6 months, the superiority of this regimen was apparent, and a randomized comparative study of meningeal irradiation was

initiated.130 Both the pilot study and the subsequent randomized study demonstrated a 50% cure rate for children with ALL who had received multiple-agent chemotherapy and effective preventive meningeal therapy. Since 1970, many institutional and collaborative groups throughout the world, using the same four phases of treatment but with modifications of drug selection and dosage schedules, have confirmed the curability of ALL in children.28 Intrathecal methotrexate alone failed to prevent meningeal leukemia in one study.131 However, Sullivan and associates132 demonstrated that repeated administration of three drugs intrathecally during remission induction and continuation therapy was equivalent to meningeal irradiation for this purpose. Radiotherapy and its adverse sequelae could be avoided in most patients. In the 1980s and 1990s, improved cure rates of up to 75% were reported.28,133 National surveys in the United States and United Kingdom demonstrated marked reduction in childhood leukemia mortality.134,135 Much of this improvement was related to more positive attitudes and greater clinical skill with experience, a remarkable increase in hematology-oncology medical and nursing specialists, better means of prevention and treatment of infection, more availability and use of blood components, earlier diagnosis and treatment, increased governmental and private health insurance coverage, improved childhood nutrition, and, in some instances, patient selection. But the discovery and judicious introduction into treatment of additional antileukemic drugs was also important. These included cytarabine, a synthetic pyrimidine antimetabolite (1968),136,137 daunorubicin, a natural DNA-intercalating anthracycline antibiotic (1968),138 asparaginase, an enzyme synthesized by bacteria that lyses the essential amino acid asparagine (1967),139 and the epipodophyllotoxins etoposide and teniposide, topoisomerase-binding agents derived from the mandrake root.140 Modification of drug schedules, such as the intravenous administration of methotrexate in high dosages with delayed leucovorin rescue, was another factor.141 The definition of subtypes of ALL and the successful targeting of specifically designed chemotherapy in children with T-cell or B-cell leukemia or those otherwise at high risk of relapse with B-precursor leukemia have been important also.142,143 From the beginning of leukemia chemotherapy, the morphologic differences in response to chemotherapy were apparent. Although occasional patients with AML experienced remissions with 6-mercaptopurine or thioguanine, a 50% remission rate was first achieved in 1967 when thioguanine was combined with cytarabine.144 Further improvement followed the introduction and inclusion of daunorubicin and etoposide. By intensive administration of these drugs, accompanied by considerable supportive

Historical perspective

therapy, it became possible in the 1980s to cure approximately 25% to 30% of unselected children with AML.145 More recent reports are more optimistic.146,147 In 1957, Barnes and Loutit148 administered lethal doses (LD98) of total-body irradiation to leukemic mice with or without subsequent homologous bone marrow transplants. The mice that received marrow homografts tended to survive without leukemia but died of a wasting disease; those that did not receive grafts had recurrence of leukemia. This led the investigators to suggest that the grafts had an antileukemic effect and stimulated similar experiments in humans. With the introduction of human leukocyte antigen (HLA) typing and matching,149 Thomas and colleagues150 achieved successful treatment of leukemia by myeloablation with total-body irradiation and chemotherapy and subsequent marrow transplantation from an HLA-compatible sibling. Evaluation of the efficacy of this procedure relative to intensive chemotherapy alone for acute leukemia has been hindered by patient selection and lack of randomized comparative studies.151 Also, the sequelae of the procedure in children, such as chronic graft-versus-host disease, multiorgan impairment, and growth failure, often preclude true cure (i.e. restoration of the capacity for normal growth, development, and health as well as freedom from leukemia). On the other hand, experience demonstrated that some types of leukemia were not curable by chemotherapy alone. Treatment with very high dosage chemotherapy and radiotherapy and histocompatible hematopoietic transplant was often successful in eliminating chronic myeloid leukemia152 that otherwise was only palliated by chemotherapy with myleran153 or hydroxyurea.154 Success was reported in some cases of juvenile myelomonocytic leukemia, myelodysplasia/myeloid leukemia associated with chromosomal monosomy 7, and AML that failed to respond to intensive chemotherapy or relapsed despite it.155–157 Evidence, again from nonrandomized comparisons, was reported that implied an advantage of hematopoietic transplantation in eliminating leukemia from children with ALL who develop hematologic relapse during chemotherapy.158 However, recent comparisons employing more acceptable analysis of results indicate no advantage over aggressive chemotherapy in children with ALL in first relapse and children with ALL that demonstrates rearrangements of the 11q23 chromosomal region.159–161 For children with newly diagnosed AML 6-year event-free survival is similar whether treated with transplant or chemotherapy.147,162 In recent years the original concept of hematosuppression and transplant proposed by Barnes and Loutit148 has been rediscovered. Transplants are viewed as immunotherapy and success dependent on graft versus leukemia reaction, not myeloablation.163 Moderate chemotherapy with-

Fig. 1.6 Zhen Yi Wang and his team developed successful therapy with the vitamin tretinoin in acute promyelocytic leukemia, the first effective differentiation agent and gene-targeted drug in cancer treatment.

out radiotherapy is often used instead of “megatherapy.” This reduces treatment-related mortality and morbidity and may improve eventual outcome. In the 1980s, a new class of agents, biological response modifiers, became available. One of them, alpha interferon, was shown by Talpaz and colleagues164 in 1986 to produce remissions of chronic myeloid leukemia, some complete, both hematologic and cytogenetic, and enduring.165 Children with adult-type chronic myeloid leukemia had similar responses.166 This offered an alternative to myeloablation and marrow transplantation. The conclusion in the 1980s that leukemia was a genetic disorder and observations that drugs effective in curing leukemia modified DNA suggested that chemotherapy might focus on genetic targeting.94,167 In 1988, Wang and colleagues (Fig 1.6).168 reported the differentiation of acute promyelocytic leukemia with resultant complete remission after administration of all-trans-retinoic acid (tretinoin). Subsequently, the genetic defect in acute

11

12

Donald Pinkel

promyelocytic leukemia was linked with an abnormal intranuclear retinoic acid receptor.169 When tretinoin was combined with conventional cytotoxic chemotherapy, the cure rate was significantly increased.170 This was the first instance of successful differentiation-inducing therapy for a human cancer, the first successful use of a vitamin to treat a human cancer, and the first specific targeting of a therapeutic agent to a cancer-associated gene rearrangement. This discovery was a major stimulant to searching for other methods of genetic targeting in the leukemias associated with specific gene rearrangements. With the introduction of molecular diagnostic technology in the 1990s, it became possible to classify most childhood leukemias genetically.28–30 For example, TEL-AML1 + leukemia resulting from a t(12;21) translocation can only be identified by molecular technology in most cases.29 The advantage of genetic classification quickly became clear when Druker and colleagues171 showed that BCRABL leukemia, whether myeloid or lymphoid, could be effectively treated by blocking the tyrosine kinase activity of the BCR-ABL fusion protein. The agent currently used, imatinib mesylate, has replaced hematopoietic transplantation and alpha interferon as initial therapy for chronic myelocytic leukemia.172 It is also included in the treatment of BCR-ABL + ALL. Although Southern blotting, the polymerase chain reaction and fluorescent in situ hybridization have been the mainstays of molecular genetic analysis of leukemia, the introduction of microarray techniques has been an important recent advance.173 With this method, one can predict the likely response to chemotherapy as well. In summary, the past 40 years of clinical investigation to identify curative treatment of childhood leukemia have met with mixed success, as demonstrated by the wide variation in cure rates. This lack of uniformity reflects not only differences in leukemia cell biology and the extent of leukemia, but also the economic status, ethnicity, residence, nutrition and constitutional genetics of the patients. The cost and complexity of curative leukemia therapy severely limit its usefulness, placing it beyond the reach of the majority of the world’s children who need it.174 Another and perhaps increasing problem are the serious adverse late sequelae of treatment with alkylating agents, anthracyclines, epipodophyllotoxins, radiotherapy, and allogeneic transplantation of hematopoietic cells, discussed elsewhere in this text (see Chapters 30 and 31).

Supportive therapy During the 100 years between Virchow’s establishment of leukemia as an entity and the advent of alkylating agents, comforting the patient with narcotics and human empathy

was the first consideration. When ionizing radiation was introduced in 1903, it became an important palliative agent for relieving local bone pain and obstructive masses as well as reducing white blood cell counts.98 Since chemotherapy was introduced in the 1940s, radiation has remained important for palliation of painful lesions as well as for curative therapy in management of extramedullary relapse in the meninges and testes and in myeloablation prior to hematopoietic transplantation.150,175,176 In 1828, Blundell177 reported a successful direct blood transfusion in a woman with postpartum hemorrhage. However, severe reactions discouraged further use. Landsteiner’s178 identification of human blood groups in 1901 enabled safer blood transfusion. During World War I, Rous and Turner179 discovered that a citrate dextrose solution and cold would preserve red blood cells. Robertson,180 an American Army surgeon who had recently worked with Rous,181 used this solution and packing boxes containing ice to preserve human red blood cells for prompt transfusion of wounded soldiers near the battlefront. For children with acute leukemia, the introduction of the hospital blood bank in 1937 was the first step in prolonging their lives.182 By the late 1940s, blood transfusions together with the newly available antibacterial agents became generally accepted as a way of maintaining life while families tried to adapt to the prognosis and begin their grieving. In 1954, with the advent of plastic blood transfusion and transfer bags and the use of the refrigerated centrifuge, platelet transfusions became available to control thrombocytopenic bleeding.183,184 This resulted in a remarkable reduction in hemorrhage as a cause of death. Platelet transfusions also provided time for antileukemic drugs to produce remission, especially in patients with AML, leading to increased rates of remission induction. Finally, the availability of platelet transfusions allowed administration of higher or more prolonged dosages of hematosuppressive agents because one could tide patients through periods of drug-induced thrombocytopenia. When effective chemotherapy was first employed in acute leukemia, rapid lysis of leukemic cells often resulted in serious and occasionally fatal metabolic disturbances, especially in florid leukemia with high white blood cell counts or massive organ involvement. The introduction of allopurinol, a synthetic inhibitor of xanthine oxidase, together with skillful fluid and electrolyte therapy, did much to solve this problem.185 More recently, recombinant urate oxidase (rasburicase) was developed as a more potent drug than allopurinol in the prevention and treatment of hyperuricemia.186 As children survived longer in remission, the immunosuppression caused by chemotherapy was more evident.

Historical perspective

Varicella became a major problem, particularly with prednisone therapy.187,188 Many children died of severe disseminated varicella, while others had treatment interrupted for long periods with consequent increased risk of relapse. With recognition that varicella and herpes zoster were caused by the same virus, plasma from adults convalescing from zoster was used both for treatment and for prevention in recently exposed children. After convalescent plasma was found effective for prevention or modification, varicella-zoster immune globulin (VZIG) was prepared and demonstrated to be effective also.189 The availability of VZIG and the education of parents and teachers about the hazard of varicella zoster infection were a major advance in reducing mortality, morbidity, and treatment interruption in exposed children. However, the third contribution of Gertrude Elion to children with leukemia, the introduction of acyclovir in 1980, was perhaps more important.190,191 Shortly after intensive multiagent therapy was introduced for acute leukemia at St. Jude Children’s Research Hospital, a peculiar pneumonia began to appear in many of the children. At first it was called “St. Jude pneumonia” and thought to be related to drug toxicity, viral infection, or both. However, postmortem study of the lungs and pulmonary needle aspiration in patients and methenamine silver nitrate staining revealed Pneumocystis carinii organisms.192 An institutional epidemiologic study performed in collaboration with the federal Communicable Disease Center (CDC) indicated that the disease was solely related to immunosuppression of the patients and not to contagion.193 Again, this disease became a major limiting factor in treating children with acute leukemia because of its occurrence during remission, its mortality and morbidity, and the consequent interruption of chemotherapy, especially in the critical early months of treatment. Pentamidine isethionate was used to treat infantile Pneumocystis pneumonia in Europe, but it was unavailable in the United States.194 It had to be imported with Food and Drug Administration approval for each diagnosed case. Subsequently, the CDC obtained an investigational new drug permit that not only expedited therapy, but eventually was the mechanism by which the acquired immunodeficiency disease syndrome was recognized. Finally, the brilliant studies of Hughes (Fig. 1.7) and colleagues,195 first in rats and then in patients, demonstrated the value of trimethoprim and sulfamethoxazole (cotrimoxazole) not only in treatment but, more important, in prevention of the disease. Early in the combination therapy of acute leukemia, severe and sometimes fatal bacteremia, particularly with gram-negative bacteria, especially Pseudomonas aeruginosa, was a major obstacle.196 Bodey and associates197

Fig. 1.7 Walter Hughes pioneered infectious disease control in children with leukemia; his work virtually eliminated the threat of Pneumocystis pneumonia as a cause of death or interruption of therapy.

showed that neutropenia was the major reason for these infections, although mucositis was an important contributor. They identified critical levels of neutrophils for control of the infections and demonstrated the need for prompt initiation of appropriate antibiotics in patients with fever and severe neutropenia. As effective aminoglycoside antibiotics became available in the 1960s and were used appropriately, mortality and morbidity due to gram-negative bacteremia declined, resulting again in better survival of children with acute leukemia. Infections with resistant gram-positive cocci have become a problem in the past 25 years, prompting the greater use of vancomycin in patients with staphylococcal or enterococcal infections and neutropenia.198 The immunosuppression and mucositis due to chemotherapy, radiation, and poor nutrition in children with leukemia also encouraged serious and sometimes fatal mycoses.199 The introduction of amphotericin B in 1958200 and of fluconazole in 1990201 represented significant

13

14

Donald Pinkel

advances in controlling these infections. However, some mycoses such as aspergillosis and mucormycosis remain resistant to treatment and are major causes of mortality, especially in children with prolonged neutropenia who are receiving extensive antibiotic therapy (see Chapter 32). Psychosocial issues became more important as children began to survive longer. Farber and associates108 recognized early the need for “total care” of children with acute leukemia. In 1964, Vernick and Karon202 introduced truthfulness in communicating with the children. Anticipating the significance of survival quality, Soni and colleagues203 pioneered longitudinal study of the neuropsychological consequences of acute leukemia and its treatment. Other late effects have also been studied extensively with the goal of defining the human cost/benefit ratio for each element of leukemia therapy (Chapter 30).

Lessons from the history of leukemia The value of history is not just in savoring the past but in appreciating how it illuminates the present and guides us into the future. Several lessons can be learned from the study of the history of leukemia, particularly childhood leukemia. One is the importance of heeding new facts and listening to new ideas and hypotheses. At each point in the history of leukemia, there have been instances of lost time and opportunity because of unreasoned resistance to innovation. Ten years after Virchow’s description of leukemia and its verification by others, its existence was still denied by many. In 1958, 8 years after his pivotal discovery, Gross was still criticized for describing the viral etiology of a mouse leukemia. Twenty years elapsed between the establishment of a battlefront blood bank and the first blood bank in an American hospital. When antifolate and antipurine drugs were first introduced, many hematologists and pediatricians refused to prescribe them because they were “too toxic.” Into the 1960s some parents were advised and medical students taught to withhold chemotherapy from childhood leukemia patients: “let the children die in peace.”204 It is important for physicians and scientists to be open to new thinking that challenges conventional wisdom and ways. Another lesson is the significance of the case report describing a patient and what the patient taught the physician. Virchow’s case report of leukemia in 1845, Lissauer’s description of a patient whose leukemia responded to arsenious oxide, Brewster and Cannon’s observation of leukemia in a child with Down syndrome, and Gloor’s patient who was cured of leukemia after arsenious oxide,

mesothorium, irradiation, and sibling blood transfusions eventually led to important knowledge of leukemia biology and treatment. A third lesson is the need to encourage rather than dampen speculation in spoken and printed discussion. Kellett’s idea that the residential aggregation of leukemia cases in Ashington might reflect an infectious agent, widespread but of low infectivity, remains viable, although statistical significance of time-space clustering is dubious. Equally important, however, is the need to clearly identify speculation and to require adequately controlled, scientifically sound investigations before drawing conclusions. Many children with acute leukemia were subjected to BCG injection on the basis of an uncontrolled study before appropriate investigations demonstrated its lack of efficacy.205–207 The relative lack of value and unfavorable risk/benefit ratio of hematopoietic transplantation for children with most types of acute leukemia has taken decades to clarify because proper comparison with optimal treatment omitting transplantation was not performed initially. The most important lesson is the need to encourage original investigator-initiated research of leukemias by clinicians and scientists working together, exchanging ideas and coordinating clinical observations with biological experimentation. For example, after Gross heard a lecture by Gilbert Dalldorf on the use of newborn mice to identify Coxsackie virus, he switched to newborn mice as subjects of his experiments and discovered the first mammalian leukemia virus. Farber’s impression that folic acid accelerated leukemia encouraged development of antifolates and the first effective treatment for childhood leukemia. Robertson’s knowledge of red blood cell preservation gained at the Rockefeller Institute enabled him to initiate blood banking on a Belgian battlefront. Borella’s observation that children with thymomegaly had a more aggressive lymphoid leukemia and his identification of thymic cell leukemia as a distinct entity led to immunophenotyping and initiated classification of leukemia by biological function. It is also important that clinical and laboratory researchers be free to think independently and to pursue goals as they see fit with minimal intervention by managers and committees. The long-term advantage of scientific freedom often exceeds the short-term gain of tightly restricted research. The late Robert Guthrie illustrates this. Assigned to provide microbiological assays of experimental antileukemic drugs, he deviated when he conceived the notion of using such an assay to screen heel-stick blood spots of newborn for high phenylalanine levels. His purpose was early detection of phenylpyruvic oligophrenia so that

Historical perspective

mental retardation could be prevented by dietary deletion of phenylalanine.17 In order to continue this research, Dr. Guthrie was compelled to resign his position for a lesser one elsewhere. Not only did his work result in today’s highly successful neonatal screening programs, but 45 years later stored “Guthrie spots” are used to track fetal origins of leukemia. Good research benefits all eventually. There is an anecdote that an accomplished senior leukemia researcher was asked by a site visit committee for his 5-year plan. He is said to have responded: “Five years? I don’t know what I will do this afternoon. I haven’t looked at my mice today.”

REFERENCES 1 Aur, R. J. A., Simone, J. V., Hustu, H. O., et al. Central nervous system therapy and combination chemotherapy of childhood lymphocytic leukemia. Blood, 1971; 37: 272–81. 2 Velpeau, A. Sur la resorption du pusaet sur l’alteration du sang dans les maladies clinique de persection nenemant. Premier observation. Rev Med, 1827; 2: 216. 3 Virchow, R. Weisses blut. Notiz Geb Natur Heilk, 1845; 36: 152–6. 4 Bennett, J. H. Case of hypertrophy of the spleen and liver in which death took place from suppuration of the blood. Edinburgh Med Surg J, 1845; 64: 413–23. 5 Craigie, D. Case of disease of the spleen in which death took place in consequence of the presence of purulent matter in the blood. Edinburgh Med Surg J, 1845; 64: 400–13. 6 Virchow, R. Die leuk¨amie. In R. Virchow, ed., Gesammelte abhandlungen zur wissenschaft lichen medizin (Frankfurt, Germany: Meidinger, 1856), pp. 190–211. 7 Friedreich, N. Ein neuer fall von leuk¨amie. Virchow’s Arch Pathol Anat, 1857; 12: 37–58. 8 Neumann, E. Ueber myelogene leuk¨amie. Berl Klin Wochenschr, 1878; 15: 69–72. 9 Turk, W. Ein system der lymphomatosen. Wien Klin Wochenschr, 1903; 16: 1073–85. 10 Ehrlich, P. Farbenanalytische untersuchungen zur histologie und klinick des blutes (Berlin: Hirschwald, 1891). 11 Reschad, H. & Schilling-Torgau, V. Ueber eine neue leuk¨amie durch echte uebergangsformen und ihre bedeutung fur dies selbst¨andigkeit dieser zellen. Munch Med Wochenschr, 1913; 60: 1981–4. 12 Ward, G. The infective theory of acute leukemia. Br J Child Dis, 1917; 14: 10–20. 13 Bennett, J. M., Catovsky, D., Daniel, M.-T., et al. Criteria for the diagnosis of acute leukemia of megakaryocytic lineage (M7). A report of the French–American–British Cooperative Group. Ann Intern Med, 1985; 103: 460–2. 14 Nowell, P. C. & Hungerford, D. A. A minute chromosome in human chronic granulocytic leukemia. Science, 1960; 132: 1497.

15 Rowley, J. D. A new consistent chromosome abnormality in chronic myelogenous leukemia identified by quinacrine fluorescence and Giemsa staining. Nature, 1973; 243: 290–3. 16 Jurlander, J., Caliguri, M. A., Ruutu, T., et al. Persistence of AML1/ETO fusion transcript in patients treated with allogeneic bone marrow transplantation for t(8;21) leukemia. Blood, 1996; 88: 2183–91. 17 Guthrie, R. Organization of a regional newborn screening laboratory. Neonatal Screening for Inborn Errors of Metabolism (Berlin: Springer, 1980), In H. Bickel, R. Guthrie & G. Hammersen, eds., pp. 259–70. 18 Ford, A. M., Ridge, S. A., Cabrera, M. E., et al. In utero rearrangements in the trithorax-related oncogene in infant leukaemias. Nature, 1993; 363: 358–60. 19 Gale, K. B., Ford, A. M., Repp, R., et al. Backtracking leukemia to birth: identification of clonotypic gene fusion sequences in neonatal blood spots. Proc Natl Acad Sci U S A, 1997; 94: 13 950–4. 20 Wiemels, J. L., Cazzaniga, G., Daniotti, M., et al. Prenatal origin of acute lymphoblastic leukaemia in children. Lancet, 1999; 354: 1499–1503. 21 Yagi, T., Hibi, S., Tabata, Y., et al. Detection of clonotypic IGH and TCR rearrangements in the neonatal blood spots of infants and children with B-cell precursor acute lymphoblastic leukemia. Blood, 2000; 96: 264–8. 22 Taub, J. W., Konrad, M. A., Ge, Y., et al. High frequency of leukemic clones in newborn screening blood samples of children with B-precursor acute lymphoblastic leukemia. Blood, 2002; 99: 2992–6. 23 Wiemels, J. L., Xiao, Z., Buffler, P. A., et al. In utero origin of t(8;21) AML1-ETO translocations in childhood acute myeloid leukemia. Blood, 2002; 99: 3801–5. 24 Greaves, M. Childhood leukaemia. BMJ, 2002; 324: 283– 7. 25 Borella, L. & Sen, L. T cell surface markers on lymphoblasts from acute lymphocytic leukemia. J Immunol, 1973; 111: 1257– 60. 26 Sen, L. & Borella, L. Clinical importance of lymphoblasts with T markers in childhood acute leukemia. N Engl J Med, 1975; 92: 828–32. 27 Ritz, J., Pesando, J. M., Notis-McConarty, J., et al. A monoclonal antibody to human acute lymphoblastic leukemia antigen. Nature, 1980; 283: 583–5. 28 Pui, C. H. & Crist, W. M. Biology and treatment of acute lymphoblastic leukemia. J Pediatr, 1994; 124: 491–503. 29 Romana, S. P., Poirel, H., Leconiat, M., et al. High frequency of t(12;21) in childhood B-lineage acute lymphoblastic leukemia. Blood, 1995; 86: 4263–9. 30 Pinkel, D. Genotypic classification of childhood acute lymphoid leukemia. Leukemia, 1999; 13(Suppl.): S90–1. 31 Jaffe, E. S., Harris, N. L., Stein, H., & Vardiman, J. W., eds. World Health Organization Classification of Tumors. Pathology and Genetics of Tumours of Hematopoietic and Lymphoid Tissues (Lyon, France: IARC Press, 2001).

15

16

Donald Pinkel

32 Vardiman, J. W., Harris, N. L., & Brunning, R. D. The World Health Organization (WHO) classification of the myeloid neoplasms. Blood, 2002; 100: 2292–302. 33 Walters, T. R., Bushore, M., & Simone, J. Poor prognosis in Negro children with acute lymphocytic leukemia. Cancer, 1972; 29: 210–14. 34 Viana, M. B., Murao, M., Ramos, G., et al. Malnutrition as a prognostic factor in lymphoblastic leukemia: a multivariate analysis. Arch Dis Child, 1994; 71: 304–10. 35 Lobato-Mendizabal, E., Ruiz-Arguelles, G. J., & Marin-Lopez, A. Leukemia and nutrition: malnutrition is an adverse prognostic factor in the outcome of treatment of patients with standard risk acute lymphoblastic leukemia. Leuk Res, 1989; 13: 899–906. 36 Lobato-Mendizabal, E., Ruiz-Arguelles, G. J., & Ganci-Cerrud, G. Effects of socioeconomic status on the therapeutic response of children with acute lymphoblastic leukemia of common risk. Neoplasia, 1991; 8: 161–5. 37 Hord, M. H., Smith, T. L., Culbert, S. J., et al. Ethnicity and cure rates of Texas children with acute lymphoid leukemia. Cancer, 1996; 77: 563–9. 38 Krivit, W. & Good, R. A. Simultaneous occurrence of mongolism and leukemia. AMA J Dis Child, 1957; 94: 289–93. 39 Ravindrinath, Y., Abella, E., Krischer, J. P., et al. Acute myeloid leukemia in Down’s syndrome is highly responsive to chemotherapy: experience of Pediatric Oncology Group AML Study 8498. Blood, 1992; 80: 2210–4. 40 Douer, D., Preston-Martin, S., Chang, E. et al. High frequency of acute promyelocytic leukemia among Latinos with acute myeloid leukemia. Blood, 1996; 87: 308–13. 41 Lennard, L., Lilleyman, J. S., Van Loon, J., et al. Genetic variation in response to 6-mercaptopurine for childhood acute lymphoblastic leukemia. Lancet, 1990; 336: 225–9. 42 McLeod, H. L., Relling, M. V., Liu, Q., et al. Polymorphic thiopurine methyltransferase in erythrocytes is indicative of activity in leukemic blasts from children with acute lymphoblastic leukemia. Blood, 1995; 85: 1897–1902. 43 Wiemels, J. L., Pagnamenta, A., Taylor, G. M., et al. A lack of a functional NAD(P)H:quinone oxidoreductase allele is selectively associated with pediatric leukemias that have MLL fusions. United Kingdom Childhood Cancer Study Investigators. Cancer Res, 1999; 59: 4095–9. 44 Wiemels, J. L., Smith, R. N., Taylor, G. M., et al. Methylenetetrahydrofolate reductase (MTHFR) polymorphisms and risk of molecularly defined subtypes of childhood acute leukemia. Proc Natl Acad Sci U S A, 2001; 98: 4004–9. ¨ 45 Ellerman, V. & Bang, O. Experimentelle leuk¨amie bei huhnern. Zentrabl Bakteriol, 1908; 46: 595–609. 46 Gross, L. “Spontaneous” leukemia developing in C3 H mice following inoculation, in infancy, with AK-leukemic extracts or AK-embryos. Proc Soc Exp Biol Med, 1951; 76: 27–32. 47 Rickard, C. G., Post, J. E., Noronha, F., et al. A transmissable virus-induced lymphocytic leukemia of the cat. J Natl Cancer Inst, 1969; 42: 987–1014.

48 Miller, J. M., Miller, L. D., Olson, C. & Gillette, K. G. Virus-like particles in phytohemagglutinin-stimulated lymphocyte cultures with reference to bovine lymphosarcoma. J Natl Cancer Inst, 1969; 43: 1297–1305. 49 Kawakami, T. G., Huff, S. D., Buckley, P. M., et al. C-type virus associated with gibbon lymphosarcoma. Nat New Biol, 1972; 235: 170–1. 50 Poiesz, B. J., Ruscette, F. W., Gagdar, A. F., et al. 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 U S A, 1980; 77: 7415–19. 51 Churchill, A. E. & Biggs, P. M. Agent of Marek disease in tissue culture. Nature, 1967; 215: 528–30. 52 Epstein, M. A., Achong, B. G., & Barr, Y. M. Virus particles in cultured lymphoblasts from Burkitt’s lymphoma. Lancet, 1964; 15: 702–3. 53 Pagano, J. S. Epstein–Barr virus: the first human tumor virus and its role in cancer. Proc Assoc Am Physicians, 1999; 111: 573–80. 54 Smith, J. W., Freeman, A. & Pinkel, D. Search for a human leukemia virus. Archiv Gesamte Virusforschung, 1967; 22: 294– 302. 55 Cooke, J. V. The incidence of acute leukemia in children. JAMA, 1942; 119: 547–50. 56 Kellett, C. E. Acute myeloid leukemia in one of identical twins. Arch Dis Child, 1937; 12: 239–52. 57 Pinkel, D. & Nefzger, D. Some epidemiological features of childhood leukemia in the Buffalo, NY, area. Cancer, 1959; 12: 351–8. 58 Pinkel, D., Dowd, J. E. & Bross, I. D. J. Some epidemiological features of malignant solid tumors of children in the Buffalo, NY, area. Cancer, 1963; 16: 28–33. 59 Heath, C. W. & Hasterlik, R. J. Leukemia among children in a suburban community. Am J Med, 1963; 34: 796–812. 60 Knox, G. Epidemiology of childhood leukemia in Northumberland and Durham. Br J Prev Soc Med, 1964; 18: 17–24. 61 Lock, S. P. & Merrington, M. Leukemia in Lewisham (1957– 1963). Br Med J, 1967; 3: 759–60. 62 Ederer, F., Myers, M. H., Eisenberg, H., et al. Temporal-spatial distribution of leukemia and lymphoma in Connecticut. J Natl Cancer Inst, 1965; 35: 625–9. 63 Kinlen, L. J., Dickson, M., & Stiller, C. A. Childhood leukemia and non-Hodgkin’s lymphoma near large rural construction sites, with a comparison with Sellafield nuclear site. BMJ, 1995; 310: 763–8. 64 Greaves, M. F. & Alexander, F. E. An infectious etiology for common acute lymphoblastic leukemia in childhood? Leukemia, 1993; 7: 349–60. 65 Greaves, M. F., Colman, S. M., Beard, M. E. J., et al. Geographical distribution of acute lymphoblastic leukemia subtypes: second report of the collaborative group study. Leukemia, 1993; 7: 27–34. 66 March, H. C. Leukemia in radiologists. Radiology, 1944; 43: 275–8.

Historical perspective

67 Folley, J. H., Borges, W., & Yamawaki, T. Incidence of leukemia in survivors of the atomic bomb in Hiroshima and Nagasaki, Japan. Am J Med, 1952; 13: 311–21. 68 Simpson, C. L., Hempelman, L. H., & Fuller, L. M. Neoplasia in children treated with x-rays in infancy for thymic enlargement. Radiology, 1955; 64: 840–5. 69 Stewart, A., Webb, J., Gates, D., et al. Malignant disease in childhood and diagnostic irradiation in utero. Lancet, 1956; 2: 447. 70 Ron, E., Modan, B., & Boice, J. D., Jr. Mortality after radiotherapy for ringworm of the scalp. Am J Epidemiol, 1988; 127: 713–25. 71 Kadhim, M. A. & Wright, E. G. Radiation-induced transmissable chromosomal instability in haemopoietic stem cells. Adv Space Res, 1998; 22: 587–96. 72 Delore, P. & Borgomano, C. Leuc´emie aigue au cours de l’intoxication benzenique: sur l’origine toxique de certaines leuc´emies aigues et leurs relations avec les an´emies graves. J M´ed Lyon, 1928; 9: 227–33. 73 Aksoy, M., Erdem, S., & Dincol, G. Leukemia in shoe workers exposed chronically to benzene. Blood, 1974; 44: 837–41. 74 Vigliani, E. C. & Saita, G. Benzene and leukemia. N Engl J Med, 1964; 271: 872–6. 75 Hayes, R. B., Yin, S. N., Dosemeci, M., et al. Mortality among benzene-exposed workers in China. Environ Health Perspect, 1996; 104 (Suppl. 6): 1349–52. 76 Hoffmann, D., Brunnemann, K. D., & Hoffman, I. Significance of benzene in tobacco carcinogenesis. In M. A. Mehlman, ed., Benzene: Occupational and Environmental Hazards: Scientific Update (Princeton, NJ: Princeton Scientific Publications, 1989), pp. 99–112. 77 Smith, M. T. The mechanism of benzene-induced leukemia: a hypothesis and speculations on the causes of leukemia. Environ Health Perspect, 1996; 104(Suppl. 6): 1219–25. 78 Thompson, J. R., Gerald, P. F., Willoughby, L. N., et al. Maternal folate supplementation in pregnancy and protection against acute lymphoblastic leukaemia in childhood: a case-control study. Lancet, 2001; 358: 1935–40. 79 Tucker, M. A., Meadows, A. T., Boice, J. D., et al. Leukemia after therapy with alkylating agents for childhood cancer. J Natl Cancer Inst, 1987; 78: 459–64. 80 Pui, C.-H., Behm, F. G., Raimondi, S. C., et al. Secondary acute myeloid leukemia in children treated for acute lymphoid leukemia. N Engl J Med, 1989; 321: 136–42. 81 Alexander, F. E., Patheal, S. L., Biondi, A., et al. Transplacental chemical exposure and risk of infant leukemia with MLL gene fusion. Cancer Res, 2001; 61: 2542–6. 82 Hartenstein, Ber Veterin¨arw, Sachsen: 1876; 44: 41: As cited by Engelbreth-Holm, J. In Spontaneous and Experimental Leukemia in Animals (Edinburgh, UK: Oliver and Boyd, 1942), p. 130. 83 Slye, M. The relation of heredity to the occurence of spontaneous leukemia, pseudoleukemia, lymphosarcoma and allied diseases in mice. Preliminary report. Am J Cancer, 1931; 15: 1361–86.

84 MacDowell, E. C. & Richter, M. N. Mouse leukemia. IX. The role of heredity in spontaneous cases. Arch Pathol, 1935; 20: 709–24. 85 Ardashnikov, S. N. The genetics of leukemia in man. J Hyg, 1937; 37: 286–302. 86 Videbaek, A. Heredity in Human Leukemia and its Relation to Cancer: A Genetic and Clinical Study of 209 Probands (London: H K Lewis, 1947). 87 Steinberg, A. G. A genetic and statistical study of acute leukemia in children. In Proceedings of the Third National Cancer Conference (Philadelphia, PA: J. B. Lippincott, 1957), pp. 353–6. 88 Siegel, A. E. Lymphocytic leukemia occurring in twins. Atlantic Med Monthly J, 1928; 31: 748–9. 89 MacMahon, B. & Levy, M. A. Prenatal origin of childhood leukemia. Evidence from twins. N Engl J Med, 1964; 270: 1082–5. 90 Brewster, H. F. & Cannon, H. E. Acute lymphatic leukemia: report of a case in an eleventh month mongolian idiot. New Orleans Med Surg J, 1930; 82: 872–3. 91 Miller, R. W. Persons with an exceptionally high risk of leukemia. Cancer Res, 1967; 27: 2420–3. 92 De Klein, A., Kessel, A. G. van, Grosveld, G., et al. A cellular oncogene is translocated to the Philadelphia chromosome in chronic myelocytic leukemia. Nature, 1982; 300: 765–7. 93 Dalla-Favera, R., Bregni, M., Erikson, J., et al. Human c-myc onc gene is located on the region of chromosome 8 that is translocated in Burkitt lymphoma cells. Proc Natl Acad Sci U S A, 1982; 79: 7824–7. 94 Pinkel, D. Curing children of leukemia. Cancer, 1987; 59: 1683– 91. 95 Deininger, M. W., Goldman, J. M., & Melo, J. V. The molecular biology of chronic myeloid leukemia. Blood, 2000; 96: 3343–56. 96 Blackwood, E. M. & Eisenman, R. N. Max: a helix-loop-helix zipper protein that forms a sequence-specific DNA-binding complex with Myc. Science, 1991; 251: 1211–17. 97 Lissauer, H. Zwei f¨alle von leucaemie. Berl Klin Wochenschr, 1865; 2: 403–4. 98 Senn, N. The therapeutical value of the Roentgen ray in the treatment of pseudoleukemia. N Y Med J, 1903; 77: 665–8. 99 Lawrence, J. H. Nuclear physics and therapy: preliminary report on a new method for the treatment of leukemia and polycythemia. Radiology, 1940; 35: 51–60. 100 Krumbhaar, E. B. & Krumbhaar, H. D. The blood and bone marrow in yellow cross (mustard gas) poisoning. Changes produced in the bone marrow of fatal cases. J Med Res, 1919; 40: 497–506. 101 Alexander, A. F. Medical report of the Bari harbor mustard casualties. Mil Surg, 1947; 101: 1–17. 102 Goodman, L. S., Wintrobe, M. W., Dameshek, W., et al. Nitrogen mustard therapy. Use of methyl-bis (beta-chloroethyl) amine hydrochloride and tris (beta-chloroethyl) amine hydrochloride for Hodgkin’s disease, lymphosarcoma, leukemia, and

17

18

Donald Pinkel

103

104 105

106 107

108 109 110

111

112

113

114

115 116 117

118

119

120

121

certain allied and miscellaneous disorders. JAMA, 1946; 132: 126–132. Karnofsky, D. A. Summary of results obtained with nitrogen mustard in the treatment of neoplastic disease. Ann NY Acad Sci, 1958; 68: 889–914. Mitchell, H. K., Snell, E. E., & Williams, R. J. The concentration of “folic acid”. J Am Chem Soc, 1941; 63: 2284. Angier, R. B., Boothe, J. H., Hutchings, B. L., et al. The structure and synthesis of the liver (L. casei) factor. Science, 1946; 103: 667–9. Spies, T. D. Treatment of macrocytic anemia with folic acid. Lancet, 1946; 1: 225–8. Farber, S., Diamond, L. K., Mercer, R. D., et al. Temporary remissions in acute leukemia in children produced by folic acid antagonist, 4-amino-pteroylglutamic acid (aminopterin). N Engl J Med, 1948; 238: 787–93. Farber, S., Toch, R., Sears, E. M., et al. Advances in chemotherapy of cancer in man. Adv Cancer Res, 1956; 4: 1–71. Seeger, D. R., Smith, J. M., & Hultquist, M. E. Antagonist for pteroylglutamic acid. J Am Chem Soc, 1947; 69: 2567. Farber, S. The effect of ACTH in acute leukemia in childhood. In J. R. Mote, ed., First Clinical ACTH Conference (New York: Blakiston, 1950). Elion, G. B., Hitchings, G. H., & Vanderwerff, H. Antagonists of nucleic acid derivatives. VI. Purines. J Biol Chem, 1951; 192: 505–18. Burchenal, J. H., Murphy, M. L., Ellison, R. R., et al. Clinical evaluation of a new antimetabolite, 6-mercaptopurine, in treatment of leukemia and allied diseases. Blood, 1953; 8: 965–99. Fernbach, D. J., Sutow, W. W., Thurman, W. G., et al. Clinical evaluation of cyclophosphamide. A new agent for the treatment of children with acute leukemia. JAMA, 1962; 182: 30–7. Karon, M. R., Freireich, E. J., & Frei, E., III. A preliminary report on vincristine sulfate: a new active agent for the treatment of acute leukemia. Pediatrics, 1962; 30: 791–6. Gloor, W. Ein fall von geheilter myeloblastenleuk¨amie. Munch Med Wochenschr, 1930; 77: 1096–8. Burchenal, J. H. & Murphy, M. L. Long-term survivors in acute leukemia. Cancer Res, 1965; 25: 1491–4. Zuelzer, W. W. Implications of long-term survival in acute stem cell leukemia of childhood treated with composite cyclic therapy. Blood, 1964; 24: 477–94. Krivit, W., Gilchrist, G., & Beatty, E. The need for chemotherapy after prolonged complete remission in acute leukemia of childhood. J Pediatr, 1970; 76: 138–41. Skipper, H. E., Schabel, F. M., Bell, M., et al. On the curability of experimental neoplasms. I. A-methopterin and mouse leukemias. Cancer Res, 1957; 17: 717–26. Goldin, A., Venditti, J. M., Humphreys, S. R., et al. Influence of the concentration of leukemic inoculum on the effectiveness of treatment. Science, 1956; 123: 840. Frei, E., III, Holland, J. F., Schneiderman, M. A., et al. A comparative study of two regimens of combination chemotherapy in acute leukemia. Blood, 1958; 13: 1126–48.

122 Frei, E., III, Freireich, E. J., Gehan, E., et al. Studies of sequential and combination antimetabolite therapy in acute leukemia. 6-mercaptopurine and methotrexate. Blood, 1961; 18: 431–54. 123 Frei, E., III, Karon, M., Levin, R. H., et al. The effectiveness of combinations of antileukemia agents in inducing and maintaining remission in children with acute leukemia. Blood, 1965; 26: 642–56. 124 Henderson, E. S. Combination chemotherapy of acute lymphocytic leukemia of childhood. Cancer Res, 1967; 27: 2570–2. 125 Henderson, E. S. & Samaha, R. J. Evidence that drugs in multiple combinations have materially advanced the treatment of human malignancies. Cancer Res, 1969; 29: 2272–80. 126 George, P., Hernandez, K., Hustu, O., et al. A study of “total therapy” of acute leukemia in children. J Pediatr, 1968; 72: 399–408. 127 Pinkel, D. Five-year follow-up of “total therapy” of childhood lymphocytic leukemia. JAMA, 1971; 216: 648–52. 128 Simone, J. V. Treatment of children with acute lymphocytic leukemia. Adv Pediatr, 1972; 19: 13–45. 129 Pinkel, D., Hernandez, K., Borella, L., et al. Drug dosage and remission duration in childhood lymphocytic leukemia. Cancer, 1971; 27: 247–56. 130 Aur, R. J. A., Simone, J. V., Hustu, H. O., et al. A comparative study of central nervous system irradiation and intensive chemotherapy early in remission of childhood acute lymphocytic leukemia. Cancer, 1972; 29: 381–91. 131 Jacquillat, C., Weil, M., Gemon, M.-F., et al. Combination therapy in 130 patients with acute lymphoblastic leukemia (Protocol O6 LA 66-Paris). Cancer Res, 1973; 33: 3278–84. 132 Sullivan, M. P., Chen, T., Dyment, P. G., et al. Equivalence of intrathecal chemotherapy and radiotherapy as central nervous system prophylaxis in children with acute lymphatic leukemia. A Pediatric Oncology Group study. Blood, 1982; 60: 948–58. 133 Rivera, G. K., Pinkel, D., Simone, J. V., et al. Treatment of acute lymphoblastic leukemia: 30 years experience at St. Jude Children’s Research Hospital. N Engl J Med, 1993; 329: 1289– 95. 134 Miller, R. W. & McKay, F. W. Decline in US childhood cancer mortality, 1950 through 1980. JAMA, 1984; 251: 1567–70. 135 Birch, J. M., Marsden, H. B., Morris Jones, P. H., et al. Improvements in survival from childhood cancer: results of a population based survey over 30 years. BMJ, 1988; 296: 1372–6. 136 Ellison, R. R, Holland, J. F., Weil, M., et al. Arabinosyl cytosine, a useful agent in the treatment of leukemia in adults. Blood, 1968; 32: 507–23. 137 Howard, J. P., Albo, V., Newton, W. A. Cytosine arabinoside. Results of a cooperative study in acute childhood leukemia. Cancer, 1968; 21: 341–5. 138 Holton, C. P., Lonsdale, D., Nora, A. H., et al. Clinical study of daunomycin in children with acute leukemia. Cancer, 1968; 22: 1014–17. 139 Hill, J. M., Roberts, J., Loeb, E., et al. L-asparaginase therapy for leukemia and other malignant neoplasms. JAMA, 1967; 202: 882–8.

Historical perspective

140 Math´e, G., Schwarzenberg, L., Pouillart, P., et al. Two epipodophyllotoxin derivatives, VM 26 and VP 16213, in the treatment of leukemias, hematosarcomas and lymphomas. Cancer, 1974; 34: 985–92. 141 Djerassi, I., Farber, S., Abir, E., et al. Continuous infusion of methotrexate in children with acute leukemia. Cancer, 1967; 20: 233–42. 142 Lauer, S. J., Pinkel, D., Buchanan, G. R., et al. Cytosine arabinoside/cyclophosphamide pulses during continuation therapy for childhood acute lymphoblastic leukemia. Cancer, 1987; 60: 2366–71. 143 Patte, C., Thierry, P., Chantal, R., et al. High survival rate in advanced-staged B-cell lymphomas and leukemias without CNS involvement with a short intensive polychemotherapy. J Clin Oncol, 1991; 9: 123–32. 144 Gee, T. S., Yu, K.-P., & Clarkson, B. D. Treatment of adult acute leukemia with arabinosylcytosine and thioguanine. Cancer, 1969; 23: 1019–32. 145 Dahl, G. V., Kalwinsky, D. K., Mirro, J. et al. A comparison of cytokinetically based versus intensive chemotherapy for childhood acute myelogenous leukemia. Hematol Blood Transfusion, 1987; 30: 83–7. 146 Perel, Y., Aurvrignon, A., Leblanc, T., et al. Impact of addition of maintenance therapy to intensive induction and consolidation chemotherapy for childhood acute myeloblastic leukemia: results of a prospective randomized trial, LAME 89/91. J Clin Oncol, 2002; 20: 2774–82. 147 Woods, W. G., Neudorf, S., Gold, S., et al. A comparison of allogeneic bone marrow transplantation, autologous bone marrow transplantation, and aggressive chemotherapy in children with acute myeloid leukemia in remission. Blood, 2001; 97: 56–62. 148 Barnes, D. W. H., & Loutit, J. F. Treatment of murine leukemia with x-rays and homologous bone marrow: II. Br J Haematol, 1957; 3: 241–52. 149 Dausset, J. Iso-leuco-anticorps. Acta Haematol, 1958; 20: 156–66. 150 Thomas, E. D., Buckner, C. D., Rudolph, R. H., et al. Allogeneic marrow grafting for hematologic malignancy using HL-A-matched donor recipient sibling pairs. Blood, 1971; 38: 267–87. 151 Pinkel, D. Bone marrow transplantation in children. J Pediatr, 1993; 122: 331–41. 152 Fefer, A., Cheever, M. A., Thomas, E. D., et al. Disappearance of Ph1 -positive cells in four patients with chronic granulocytic leukemia after chemotherapy, irradiation and marrow transplantation from an identical twin. N Engl J Med, 1979; 300: 333–7. 153 Galton, D. A. G. Myleran in chronic myeloid leukemia. Results of treatment. Lancet, 1953; 264: 208–13. 154 Fishbein, W. N., Carbone, P. P., Freireich, E. J., et al. Clinical trials of hydroxyurea in patients with cancer and leukemia. Clin Pharmacol Ther, 1965; 5: 574–80. 155 Sanders, J., Buckner, C., Thomas, E. D., et al. Allogeneic marrow transplantation for children with juvenile chronic myelogenous leukemia. Blood, 1988; 71: 1144–6.

156 Bunin, N., Casper, J., Chitambar, C., et al. Partially matched bone marrow transplantation in patients with myelodysplastic syndromes. J Clin Oncol, 1988; 6: 1851–5. 157 Appelbaum, F. R., Clift, R. A., Buckner, C. D., et al. Allogeneic marrow transplantation for acute nonlymphoblastic leukemia after first relapse. Blood, 1983; 61: 949–53. 158 Dopfer, R., Henze, G., Bender-Gotze, C., et al. Allogeneic bone marrow transplantation for childhood acute lymphoblastic leukemia in second remission after intensive primary and relapse therapy according to the BFM and Co-ALL protocols; results of the German cooperative study. Blood, 1991; 78: 2780–4. 159 Harrison, G., Richards, S., Lawson, S., et al. Comparison of allogeneic transplant versus chemotherapy for relapsed childhood acute lymphoblastic leukaemia in the MRC UKALL R1 trial. Ann Oncol, 2000; 11: 999–1006. 160 Gaynon, P. S., Harris, R. E., Trigg, M. E., et al. Chemotherapy (CT) vs. BMT for children (pts) with acute lymphoblastic leukemia (ALL) and early marrow relapse (MR): CCG-1941. Blood, 2000; 96: 418a. 161 Pui, C. H., Gaynon, P. S., Boyett, J. M., et al. Outcome of treatment in childhood acute lymphoblastic leukaemia with rearrangements of the 11q23 chromosomal region. Lancet, 2002; 359: 1909–15. 162 Pinkel, D. Treatment of children with acute myeloid leukemia. Blood, 2001; 97: 3673. 163 Giralt, S., Estey, E., Albitar, M., et al. Engraftment of allogeneic hematopoietic progenitor cells with purine analog-containing chemotherapy: harnessing graft-versusleukemia without myeloablative therapy. Blood, 1997; 89: 4531–6. 164 Talpaz, M., Kantarjian, H. M., & McCredie, K. Hematologic remission and cytogenetic improvement induced by human interferon alpha in chronic myelogenous leukemia. N Engl J Med, 1986; 314: 1065–9. 165 Talpaz, M., Kantarjian, H., Kurzrock, R., et al. Interferon-alpha produces sustained cytogenetic responses in chronic myelogenous leukemia. Ann Intern Med, 1991; 114: 532–8. 166 Dow, L., Raimondi, S., Culbert, S., et al. Response to alpha-interferon in children with Philadelphia chromosomepositive chronic myelocytic leukemia. Cancer, 1991; 68: 1678–84. 167 Pinkel, D. & Granoff, A., eds. Genetic Targeting in Leukemia. Accomplishments in Oncology, vol. 2 (no. 2) (Philadelphia, PA: J. B. Lippincott, 1988). 168 Huang, M. E., Ye, Y. C., Chen, S. R., et al. Use of all-trans retinoic acid in the treatment of acute promyelocytic leukemia. Blood, 1988; 72: 567–72. 169 De Th´e, H., Lavau, C., Marchio, A., et al. The PML-RAR fusion mRNA generated by the t(15;17) translocation in acute promyelocytic leukemia encodes a functionally altered RAR
. Cell, 1991; 66: 675–84. 170 Fenaux, P., Wattel, E., Archimbaud, E., et al. Prolonged followup confirms that all-trans retinoic acid followed by chemotherapy reduces the risk of relapse in newly diagnosed acute promyelocytic leukemia. Blood, 1994; 84: 666–7.

19

20

Donald Pinkel

171 Druker, B. J. & Lydon, N. B. Lessons learned from the development of an abl tyrosine kinase inhibitor for chronic myelogenous leukemia. J Clin Invest, 2000; 105: 3–7. 172 Mauro, M. J., O’Dwyer, M., Heinrich, M. C., Druker, B. J. STI 571: a paradigm of new agents for cancer therapeutics. J Clin Oncol, 2001; 20: 325–334. 173 Yeoh, E. J., Ross, M. E., Shurtleff, S. A., et al. Classification, subtype discovery, and prediction of outcome in pediatric acute lymphoblastic leukemia by gene expression profiling. Cancer Cell, 2002; 1: 133–43. 174 Chandy, M. Childhood acute lymphoblastic leukemia in India: an approach to management in a three-tier society. Med Pediatr Oncol, 1995; 25: 197–203. 175 Kun, L. E., Camitta, B. M., Mulhern, R. K., et al. Treatment of meningeal relapse in childhood acute lymphoblastic leukemia. I. Results of craniospinal irradiation. J Clin Oncol, 1984; 2: 359–64. 176 Stoffel, T. J., Nesbit, M. E., Levitt, S. H. Extramedullary involvement of the testes in childhood leukemia. Cancer, 1975; 35: 1203–11. 177 Blundell, J. Successful case of transfusion. Lancet, 1828; 1: 431–2. 178 Landsteiner, K. Ueber agglutinationserscheinungen normalen menschlichen blutes. Wien Klin Wochenschr, 1901; 14: 1132–4. 179 Rous, P. & Turner, J. R. The preservation of living red blood cells in vitro I. Method of preservation. J Exp Med, 1916; 23: 219–37. 180 Robertson, O. H. Transfusion with preserved red blood cells. Br Med J, 1918; 1: 691–5. 181 Rous, P. & Robertson, O. H. The normal fate of erythrocytes I. The findings in healthy animals. J Exp Med, 1917; 25: 651–64. 182 Fantus, B. The therapy of the Cook County Hospital: blood preservation. JAMA, 1937; 109: 128–131. 183 Gardner, F. H., Howell, D. H., & Hirsch, E. O. Platelet transfusion utilizing plastic equipment. J Lab Clin Med, 1954; 43: 196–207. 184 McGovern, J. J. Platelet transfusions in pediatrics. New Engl J Med, 1957; 256: 922–7. 185 Rundles, R. W., Wyngarden, J. B., Hitchings, G. H., et al. Effects of the xanthine oxidase inhibitor, allopurinol, on thiopurine metabolism, hyperuricemia and gout. Trans Assoc Am Phy, 1963; 76: 126–40. 186 Pui, C.-H., Mahmond, H. H., Wiley, J. M., et al. Recombinant urate oxidase for the prophylaxis or treatment of hyperuricemia in patients with leukemia or lymphoma. J Clin Oncol, 2001; 19: 697–704. 187 Pinkel, D. Chickenpox and leukemia. J Pediatr, 1961; 58: 729– 37. 188 Feldman, S., Hughes, W. T., & Daniel, C. B. Varicella in children with cancer. Seventy-seven cases. Pediatrics, 1975; 56: 388–97. 189 Zaia, J. A., Levin, M. J., & Preblud, S. R., et al. Evaluation of varicella-zoster immune globulin: protection of immunosuppressed children after household exposure to varicella. J Infect Dis, 1983; 147: 737–43.

190 Biron, K. K. & Elion, G. B. In vitro susceptibility of varicellazoster virus to acyclovir. Antimicrob Agents Chemother, 1980; 18: 443–7. 191 Prober, C. G., Kirk, L. E., & Keeney, R. E. Acyclovir therapy of chickenpox in immunosuppressed children: a collaborative study. J Pediatr, 1982; 101: 622–5. 192 Johnson, H. D. & Johnson, W. W. Pneumocystis carinii pneumonia in children with cancer. Diagnosis and treatment. JAMA, 1970; 214: 1067–73. 193 Perera, D. R., Western, K. A., Johnson, H. D., et al. Pneumocystis carinii pneumonia in a hospital for children. Epidemiologic aspects. JAMA, 1970; 214: 1074–8. 194 Ivady, G. & Paldy, L. A new method of treating interstitial plasma cell pneumonia in premature infants with pentavalent antimony and aromatic diamidines. Mschr Kinderheilk, 1958; 106: 10–14. 195 Hughes, W. T., Kuhn, S., Chaudhary, S., et al. Successful chemoprophylaxis for Pneumocystis carinii pneumonitis. N Engl J Med, 1977; 297: 1419–26. 196 Frei, E., Levin, R. H., Bodey, G. P., et al. The nature and control of infections in patients with acute leukemia. Cancer Res, 1965; 25: 1511–15. 197 Bodey, G. P., Buckley, M., Sathe, Y. S., et al. Quantitative relationships between circulating leucocytes and infection in patients with acute leukemia. Ann Intern Med, 1966; 64: 328– 40. 198 Pizzo, P. A., Ladisch, S., Simon, R. M., et al. Increasing incidence of gram-positive sepsis in cancer patients. Med Pediatr Oncol, 1978; 5: 241–4. 199 Young, R. C., Bennett, J. E., Geelhoed, G. W., et al. Fungemia with compromised host resistance. Ann Intern Med, 1974; 80: 605–12. 200 Procknow, J. J. & Loosli, C. G. Treatment of the deep mycoses. AMA Arch Intern Med, 1958; 101: 765–802. 201 Galgiani, J. N. Fluconazole, a new antifungal agent. Ann Intern Med, 1990; 113: 177–9. 202 Vernick, V. & Karon, M. Who’s afraid of death on a leukemia ward? Am J Dis Child, 1965; 109: 393–7. 203 Soni, S. S., Marten, G. W., Pitner, S. E., et al. Effects of central nervous system irradiation on neuropsychologic functioning of children with acute lymphocytic leukemia. N Engl J Med, 1975; 293: 113–18. 204 Pinkel, D. Selecting treatment for children with acute lymphoblastic leukemia. J Clin Oncol, 1996; 14: 4–6. 205 Math´e, G., Amiel, J. L., Schwarzenberg, L., et al. Active immunotherapy for acute lymphoblastic leukemia. Lancet, 1969; 1: 697–9. 206 Kay, H. Treatment of acute lymphoblastic leukemia. Comparison of immunotherapy (BCG), intermittent methotrexate, and no therapy after a 5 month intensive cytotoxic regimen (Concord trial). Br Med J, 1971; 4: 189–94. 207 Heyn, R. M., Joo, P., Karon, M., et al. BCG in the treatment of acute lymphocytic leukemia. Blood, 1975; 46: 431–42.

2 Diagnosis and classification Mihaela Onciu and Ching-Hon Pui

Introduction Precise diagnosis and classification are essential to the successful treatment and biologic study of the childhood leukemias. In broadest terms, the leukemias are classified as acute versus chronic and as lymphoid versus myeloid. The terms acute and chronic originally referred to the relative durations of survival of patients with these diseases when effective therapy was not available. With improvements in treatment, they have taken on new meanings. Acute currently refers to leukemia characterized by rapid tumor cell proliferation and a predominance of blast cells, while chronic leukemia encompasses a variety of myeloproliferative and lymphoproliferative disorders in which the predominant tumor cells show variable degrees of differentiation beyond the blast stage. The vast majority of childhood leukemia cases are acute, unlike those in adults. The most common subtype, acute lymphoblastic (also termed lymphocytic or lymphoid) leukemia (ALL) accounts for 75% to 80% of all childhood cases, while acute myeloid (also termed myelocytic, myelogenous, or nonlymphoblastic) leukemia (AML) comprises approximately 20%.1 By contrast, chronic myeloid leukemia (CML) represents only approximately 2% of childhood leukemias1,2 and chronic lymphocytic leukemia (CLL) is reported only rarely in children.3–6 Finally, myelodysplastic syndrome (MDS) designates a heterogeneous group of clonal diseases related to a subset of AML.2 MDS is characterized by peripheral blood cytopenias, normocellular or hypercellular but nonproductive bone marrow (inefficient hematopoiesis), and dysmorphic maturation of hematopoietic precursors. It may evolve into frank AML or result in death due to cytopenic complications.

The modern approach to leukemia classification incorporates morphologic findings, immunophenotype, and genetic lesions, in an attempt to delineate homogeneous and clinically and biologically relevant disease categories. This chapter is an overview of the current principles and techniques used for the diagnosis and classification of the childhood leukemias as a basis for treatment assignment and biologic study. Data regarding immunophenotyping, cytogenetics, molecular genetic analysis and gene expression profiling, are introduced briefly in this context and covered in greater detail in Chapters 7, 9, 10, and 11.

Bone marrow sampling Bone marrow examination is essential to establishing the diagnosis of leukemia because as many as 20% of patients with acute leukemia lack circulating blast cells at diagnosis,7 and the morphology of leukemic cells in peripheral blood may differ from that in marrow. Marrow samples are usually obtained by aspiration, supplemented in selected circumstances by biopsy.8 The aspirated material provides cells for morphologic evaluation and biologic studies, while biopsy specimens are useful for the estimation of marrow cellularity, documentation of bone marrow fibrosis, and assessment of marrow involvement by certain types of nonHodgkin lymphoma or solid tumors. Biopsy samples can also serve as surrogates for aspirated samples when the latter are not available (e.g. due to extreme packing of marrow cells, myelofibrosis, or markedly hypocellular marrow). The site chosen for marrow aspiration depends mainly on the age and size of the patient (Fig. 2.1). Whether one performs aspiration or a biopsy, full knowledge of bone anatomy is essential to success. Faulty positioning of the

C Cambridge University Press 2006. Childhood Leukemias, ed. Ching-Hon Pui. Published by Cambridge University Press.


21

22

Mihaela Onciu and Ching-Hon Pui

Fig. 2.1 Common anatomic sites of bone marrow sampling, indicated by arrows (anterior iliac crest, posterior iliac crest, sternum, and anterior tibial tuberosity).

needle (e.g. failure to seat the needle on the posterior iliac crest) is the most common cause of failure. Sliding of the needle point on heavily innervated periosteum due to faulty positioning may cause severe pain, especially if analgesia is insufficient. Aspiration should be done with a needle designed specifically for that purpose; substitution with a biopsy needle should be avoided because it has a much larger dead space, leading to wasteful collection and possible clotting of the specimen. Although an 18-gauge needle and a 1- to 5-ml syringe are adequate for surveillance marrow aspiration, a larger needle (16-gauge) and syringe (20- or 50-ml) ensure better results at diagnosis, when the marrow is densely packed and when a larger sample is needed for characterization of the leukemic process. Most marrow aspirates are obtained from the posterior superior iliac crest. The anterior iliac crest can serve as an alternative site in obese children, or if multiple samples are required, for example, in newly diagnosed cases in which large samples are needed. The anteromedial surface of the tibia is occasionally sampled in infants younger than 12 months old. Rarely, the spinous processes of the most prominent vertebral segments (C7, L1, or L2) or the sternum are used in older adolescents. In some circumstances, needle placement is directed radiographically, usually to access a specific bone lesion.

Although generally a safe procedure, bone marrow aspiration can produce serious complications, such as bone penetration with damage to underlying structures. Sterile technique and sterile (preferably disposable) equipment should be used in all situations. No critical structures lie in close proximity to the posterior superior and anterior iliac crests. However, the heart and the ascending aorta are near the sternum, which is only approximately 1-cm thick, rendering sternal aspiration potentially hazardous8 ; a prudent precaution in sternal aspiration is to use a guard that limits the depth of penetration of the needle. Sternal aspiration is contraindicated in young children, and is only rarely necessary in older children and adolescents. Marrow biopsy is usually performed at the posterior superior iliac crest. Other sites may also be used, with the exception of the sternum, where biopsy is absolutely contraindicated. This procedure should be performed before marrow aspiration, or the biopsy should be redirected away from the aspiration site, to avoid hemorrhage and distortion of the biopsy core. The needle should be directed into the marrow cavity (not tangentially along the marrow cortex) by firmly seating it on the posterior superior iliac crest and aiming toward the anterior crest. The operator must take care to obtain a sufficient quantity of bone marrow for analysis; in infants and young children the needle may initially traverse epiphyseal cartilage, which is of no use for the evaluation of marrow disease. The trochar of the biopsy needle should remain in place until bone is contacted, to avoid contamination of the biopsy with skin, muscle, and connective tissue fragments. To avoid distortion of the biopsy, one should rotate the needle on its long axis as it advances to facilitate cutting rather than crushing the bone. The biopsy specimen should be gently pushed out the butt rather than the cutting edge of the needle, as the latter has a slight inward curve to trap the material. For years, the Jamshidi needle has been used for this procedure.9 A recently developed snare-coil device with an internal capturing coil may minimize postinsertion needle manipulations, and hence pain, and yields intact core specimens.10 Touch preparations should be made from all biopsy specimens, to provide air-dried material for Romanowsky and cytochemical staining in the event that aspiration attempts fail.11

Morphologic and cytochemical analysis Specimen preparation Leukemia diagnosis and classification begins with morphologic analysis of air-dried Romanowsky (Wright,

Diagnosis and classification

¨ Wright-Giemsa, or May-Grunwald-Giemsa)-stained peripheral blood and bone marrow smears and/or biopsy touch preparations. The Wright stain, used widely for analysis of peripheral blood smears, is not satisfactory for bone marrow analysis, as it stains granules poorly and does not allow adequate discrimination of immature cells. Preparation of the bone marrow biopsy material requires several steps, including fixation in formalin or a mercurial fixative (such as B5 or Zenker) and decalcification, followed by paraffin-embedding, sectioning (at 4 or 5 microns) and staining. The hematoxylin-eosin (H&E) stain allows for a general assessment of marrow cellularity, myeloid:erythroid ratio, numbers of megakaryocytes and the presence of abnormal infiltrates such as blasts, lymphoma cells, metastatic tumor or granulomatous inflammation. Additional histochemical staining can further highlight the presence of fibrosis (reticulin and trichrome stains) or the expression of chloroacetate esterase by the tumor cells (Leder stain). Immunohistochemical staining using monoclonal antibodies [e.g. antibodies specific for myeloperoxidase, lysozyme, terminal deoxynucleotidyl transferase (TdT), or CD3, CD10, or CD79a] can aid in lineage assignment when there is insufficient aspirate material to perform flow cytometric analysis. Other types of tissue preparation, such as plastic-embedded sections and electron microscopy are expensive, seldom provide information beyond simpler techniques, and hence are not widely used.

Morphologic diagnosis The morphologic diagnosis of leukemia consists essentially of two steps: establishing a diagnosis of leukemia and classifying the leukemic process according to lineage and degree of differentiation. Establishing a diagnosis of leukemia Correlation of the findings in peripheral blood and bone marrow samples is often required to establish a diagnosis of leukemia. The peripheral blood counts are variably abnormal, depending on the type of leukemia. In acute leukemias bone marrow infiltration by the leukemic process often results in anemia and thrombocytopenia, while the leukocyte counts may be decreased or variably increased with a predominance of blasts. Even in the setting of marked leukopenia, a rare circulating blast may still be encountered on the peripheral smear. However, occasional cases of acute leukemia may present with profound cytopenia and no circulating blasts. The presence of dysplastic features is typically associated with AML and MDS, and therefore such a finding makes ALL less likely.

Chronic leukemias, by contrast, are invariably associated with variable degrees of leukocytosis. In chronic myeloproliferative disorders, the leukocytosis is usually due to marked neutrophilia with associated increase in immature myeloid precursors, including myeloblasts. Depending on the subtype of disease, there may be associated monocytosis, basophilia, and eosinophilia. In chronic lymphoproliferative disorders (such as CLL or leukemic presentation of non-Hodgkin lymphoma), circulating atypical lymphocytes usually account for most of the increase in leukocyte counts. The bone marrow examination should include, at a minimum, an assessment of the myeloid-to-erythroid cell ratio, the presence and percentage of blasts, the percentage of monocytes, and the morphologic features of all cell lines. Acute leukemias are by definition characterized by variable replacement of the marrow cellularity by blasts or abnormal promyelocytes (the latter being characteristic of acute promyelocytic leukemia). The minimum percentage of marrow blasts required to establish a diagnosis of acute leukemia varies with the lineage of leukemia and the classification system applied. Hence, one must first establish the blast lineage (lymphoid versus myeloid) before proceeding with the evaluation (see details below). For ALL, the arbitrary cut-off most frequently used is 25% of marrow replacement by leukemic lymphoid blasts. If malignant lymphoblasts account for less than 25%, the disease is staged and treated as lymphoblastic lymphoma involving the bone marrow. For AML, the French-AmericanBritish (FAB) classification12,13 requires at least 30% bone marrow myeloblasts, while in the World Health Organization (WHO) classification,2,14,15 a diagnosis of AML can be established with 20% myeloblasts or more. When there are sufficient myeloblasts to establish a diagnosis of AML, further classification of the disease requires assessment of the degree of differentiation and lineage (detailed below). If myeloblasts represent less than 20% or 30% of the marrow cells, myelodysplastic syndromes and the accelerated phase of a myeloproliferative disorder (such as CML) will have to be considered in the differential diagnosis. If there is no increase in marrow blasts, and depending on the presence of hypercellularity, myeloid and megakaryocyte hyperplasia or dysplasia, a low-grade myelodysplastic syndrome or the chronic (stable) phase of a myeloproliferative disorder are more likely. As discussed above, if there is an increase in bone marrow blasts, the lineage of these cells should be established. A combination of morphologic examination and cytochemical staining is usually relatively accurate in distinguishing between lymphoid and myeloid blasts (Table 2.1). However, in a certain proportion of cases, the morphologic

23

Table 2.1 Morphologic and cytochemical characteristics of blasts in the major subtypes of acute leukemia Stain/Feature Romanowsky stain Cell size Nucleus Shape/outline

Chromatin Nucleoli Cytoplasm Color Amount Granules

Auer rods Nuclear:cytoplasm (N/C) ratio Periodic acid Schiff (PAS) stain Myeloperoxidase stain Sudan Black B stain Esterase stains Alpha naphthyl butyrate (ANBE) Alpha naphthyl acetate (ANAE) Naphthol ASD chloroacetate (CAE) Acid phosphatase stain

Lymphoblasts

Myeloblasts

Monoblasts

Erythroblasts

Megakaryoblasts

Variable

Large

Large

Large

Variable

Round or indented, deeply cleaved

Round or indented

Round, indented, lobulated

Round, binucleated or multinucleated

Condensed or finely dispersed 0–2, small and inconspicuous (L1) or prominent (L2) Basophilic (deeply basophilic in L3 subtype) Scant to moderate

Finely dispersed

Finely dispersed

Round, multilobulated, binucleated or multinucleated Finely dispersed

1–4, prominent

1–3, prominent

0–5, prominent

Condensed or finely dispersed 0–3, variable size

Lightly basophilic

Lightly basophilic to blue-gray Moderate to abundant

Deeply basophilic

Basophilic

Scant

Usually absent; amphophilic granules in some cases (“granular ALL”); abundant azurophilic granules in rare cases Absent Variable, high (L1) or low (L2)

Usually present

Usually present, fine azurophilic or amphophilic

Absent

Variable, often with surface budding Usually absent

Usually present Typically low

Absent Low

Absent Low

Absent Variable

Coarse granules or blocks

Negative

Negative Usually negative; weakly positive in granular ALL

Positive Positive

Usually negative; sometimes fine or coarse granulation Positive or negative Positive or negative

Strongly positive, coarsely granular pattern Negative Negative

Negative or fine granular positivity Negative Negative

Negative; rarely weakly positive in granular ALL Negative or weakly positive

Negative

Positive

Negative

Negative

Negative or positive (not inhibited by fluoride) Positive

Diffuse positivity (inhibited by fluoride) Positive or negative

Positive (not inhibited by fluoride) Negative

Localized positivity (partially inhibited by fluoride) Negative

Negative

Negative

Negative

Localized positivity

Negative or weakly positive Positive in T-cell and in some pre-B ALL

Moderate

Diagnosis and classification

Fig. 2.2 ALL, L1 (FAB). Small blasts with indistinct nucleoli, with an admixture of some larger blasts. This spectrum of small and larger blasts is common in ALL. (Wright-Giemsa, ×1000; see color plate 2.2 for full-color reproduction.)

Fig. 2.3 ALL, L2 (FAB). Blasts with prominent nucleoli and moderate amounts of cytoplasm, with an admixture of smaller blasts. Such cases overlap morphologically with AML and emphasize the importance of ancillary studies to assign the correct lineage in acute leukemia (Wright-Giemsa, ×1000; see color plate 2.3 for full-color reproduction.)

and cytochemical findings may be ambiguous, such that immunophenotypic analysis may be required to make this distinction. The morphologic features seen on bone marrow smear examination may suggest either lymphoid or myeloid differentiation of leukemic cells, but with the exception of Auer rods in myeloblasts, none of these findings are lineagespecific. Lymphoblasts tend to be relatively small (identical to or twice the size of small lymphocytes) with scant, often light-blue cytoplasm; a round, clefted or slightly indented nucleus; fine to slightly coarse and clumped chromatin; and inconspicuous nucleoli (Figs. 2.2 to 2.4).12,16 Cytoplas-

Fig. 2.4 B-ALL (FAB ALL, L3) with the t(8;14). Blasts are characterized by intensely basophilic cytoplasm, regular nuclear features, prominent nucleoli, and cytoplasmic vacuolization. (Wright-Giemsa, ×1000; see color plate 2.4 for full-color reproduction.)

Fig. 2.5 ALL, L1 with prominent cytoplasmic vacuoles. Note the scant, lightly basophilic cytoplasm, and inconspicuous nucleoli (by comparison with Fig. 2.4). Such cases may be mistaken for B-ALL. Vacuolation is not unique to ALL, L3 (Burkitt) leukemia and other cytologic features have to be considered when making this diagnosis. (Wright-Giemsa, ×600; see color plate 2.5 for full-color reproduction.)

mic vacuoles (Fig. 2.5)17–19 and granules (Fig. 2.6)20–23 are found in lymphoblasts in some cases of ALL; such granules are usually amphophilic (fuchsia) rather than the deep purple of primary myeloid granules or the color of any of the secondary myeloid granules. In some cases, lymphoblasts have a “hand-mirror” shape,24–26 but this feature is not lineage specific. Myeloblasts have a more heterogeneous morphology, which depends on their lineage and differentiation. Generally, they tend to be larger than

25

26

Mihaela Onciu and Ching-Hon Pui

Fig. 2.6 ALL with cytoplasmic granules. Fuchsia-colored granules are present in the cytoplasm of numerous blasts. Such granules may lead to a mistaken diagnosis of AML, but the granules are negative for MPO and myeloid-pattern SBB. Immunophenotyping will confirm a diagnosis of ALL, usually of precursor B-cell lineage. Granular ALL may display granular positivity for esterase stains. (Wright-Giemsa, ×1000; see color plate 2.6 for full-color reproduction.)

Fig. 2.7 AML with minimal granulocytic differentiation (FAB M1). Blasts are large and somewhat irregular, with moderate amounts of cytoplasm but little cytoplasmic differentiation. (WrightGiemsa, ×1000; see color plate 2.7 for full-color reproduction.)

lymphoblasts, with round or indented nuclei, fine chromatin, one to several distinct nucleoli, blue to blue-gray cytoplasm, and variable numbers of cytoplasmic granules (Figs. 2.7 to 2.9). Auer rods (elongated red crystalline rods consisting of coalesced lysosomal granules) are pathognomonic of malignant myeloblasts (see Fig. 2.9).27 In the hypergranular variant of acute promyelocytic leukemia (APL), the leukemic cells are promyelocytes with abundant cytoplasmic granulation and prominent, often numerous Auer rods (Fig. 2.10). In the microgranular

Fig. 2.8 Myeloperoxidase positivity in AML, demonstrated by yellow staining against a Romanowsky-stained background. (o-Toluidine stain with dilute Giemsa counterstain, ×1000; see color plate 2.8 for full-color reproduction.)

Fig. 2.9 AML with granulocytic differentiation (FAB M2). Differentiating granulocyte precursors are admixed with myeloblasts. Several blasts contain Auer rods (arrows). (Wright-Giemsa, ×1000; see color plate 2.9 for full-color reproduction.)

variant of APL the cytoplasmic granulation is minimal or absent (Fig. 2.11). Monoblasts are large, often with a folded or indented nucleus, fine chromatin, one to three large nucleoli, and abundant blue or blue-gray, frequently vacuolated cytoplasm that may contain fine amphophilic granules (Figs. 2.12 to 2.14). Erythroblasts are large, with centrally located nuclei, sometimes binucleated or multinucleated, deeply basophilic (blue) cytoplasm, fine nuclear chromatin, and prominent nucleoli (Figs. 2.15 and 2.16).28 Megakaryoblasts are highly polymorphic, ranging from small cells with scant cytoplasm and fine or dense chromatin to large cells with abundant cytoplasm, fine chromatin, and one to several nucleoli (Fig. 2.17).29 They

Diagnosis and classification

Fig. 2.10 Hypergranular acute promyelocytic leukemia (FAB M3, sometimes designated M3h). The neoplastic cells are abnormal hypergranular promyelocytes with reddish granules and occasional clefted nuclei. Several promyelocytes contain multiple Auer rods (so-called “faggot cells”). (Wright-Giemsa, ×1000; see color plate 2.10 for full-color reproduction.)

Fig. 2.11 Microgranular acute promyelocytic leukemia (FAB M3v). The leukemic process is characterized by cells with bilobed and grooved nuclei, and sparse cytoplasmic granulation. (Wright-Giemsa, ×1000; see color plate 2.11 for full-color reproduction.)

may be bi- or multinucleated, may have cytoplasmic blebs or platelets on their surface,29,30 and may form blast cell clumps, mimicking metastatic small cell tumors.31,32 Cytochemical staining improves the accuracy and reproducibility of lineage assessment and is required for traditional AML subclassification according to the FAB and WHO criteria.2,12,14,15,33 The advent of flow cytometric analysis has rendered many of the traditional cytochemical stains obsolete. However, myeloperoxidase (MPO), Sudan Black B (SBB), and nonspecific esterase (NSE) stains,

Fig. 2.12 Acute myelomonocytic leukemia (FAB M4). The leukemic cell population includes large blasts, with irregular and reniform nuclei, promonocytes and monocytes. Esterase staining is often positive in such cases. (Wright-Giemsa, ×1000; see color plate 2.12 for full-color reproduction.)

Fig. 2.13 Acute monoblastic leukemia (FAB M5). Blasts are large and uniform, with abundant blue-gray cytoplasm, and may have cytoplasmic vacuolation and amphophilic granules. (Wright-Giemsa, ×1000; see color plate 2.13 for full-color reproduction.)

including alpha naphthyl butyrate esterase (ANB) and alpha naphthyl acetate esterase (ANA), have remained useful in this regard (Table 2.1). MPO staining detects myeloperoxidase in the peroxisomes of neutrophilic, eosinophilic, and monocytic precursors (see Fig. 2.8). Myeloblasts and some monoblasts stain positively with MPO, while lymphoblasts are uniformly negative. Auer rods are usually strongly positive and can often be identified more readily with MPO than with Romanowsky staining.34 By standard criteria of the FAB Cooperative Group, a leukemic process is considered MPO positive if reactivity is present in 3% or more of the blast

27

28

Mihaela Onciu and Ching-Hon Pui

Fig. 2.14 Alpha naphthyl butyrate esterase reactivity in acute monoblastic leukemia. ANB positivity is characterized by intense, diffuse reddish-brown cytoplasmic staining, typical of monoblastic leukemia. (ANB stain with hematoxylin counterstain, ×1000; see color plate 2.14 for full-color reproduction.)

Fig. 2.15 AML with predominant erythroid differentiation (FAB M6, also termed M6a). An infiltrate of myeloblasts is present admixed with dysplastic erythroid precursors. (Wright-Giemsa, ×1000; see color plate 2.15 for full-color reproduction.)

cells.12 This 3% threshold for MPO positivity requires careful interpretation by the reviewer, as it may be exceeded (in some cases of ALL) by staining of normal residual myeloid precursors. Several subtypes of AML (including minimally differentiated AML, megakaryoblastic leukemia and some erythroleukemias, as well as a subset of the acute monoblastic leukemia) lack MPO reactivity.35,36 MPO staining of unfixed, unstained smears must be performed without undue delay, as myeloperoxidase is unstable and may be undetectable after 7 to 10 days of slide storage. This constraint is of less importance to institutional laboratories than to reference centers. SBB is a direct nonenzymatic stain of phospholipids in the membranes of granules, principally those in myeloid

Fig. 2.16 Acute erythroblastic leukemia (also termed FAB M6b). The blasts are large with basophilic cytoplasm, resembling normal erythroblasts, and may show vacuolization. Immunophenotypic analysis confirms erythroid differentiation of the blasts. (Wright-Giemsa, ×1000; see color plate 2.16 for full-color reproduction.)

Fig. 2.17 Acute megakaryoblastic leukemia (FAB M7). The blasts have prominent surface blebs, bi- or multinucleation, and may occasionally form cohesive clusters, mimicking metastatic tumor. (Wright-Giemsa, ×1000; see color plate 2.17 for full-color reproduction.)

precursors; its reactivity parallels that of MPO but is usually more intense. By standard FAB criteria, a leukemic process is considered SBB positive if 3% or more of the blasts are positive. Rare ALL specimens are weakly positive for SBB, although this reactivity is restricted to intensive SBB staining procedures and produces gray granular staining rather than the dense black staining characteristic of myeloid granules. The SBB staining pattern in ALL lacks the Golgi zone staining seen in myeloid precursors. It is most often seen in granular ALL, where it appears to be associated to the membrane of lysosomal granules.37–39 Reactivity to SBB is stable for months in unfixed, unstained air-dried smears.

Diagnosis and classification

Esterase enzymes in monocytic precursors can be stained with either ANB or ANA as the substrate (Fig 2.14).40 Typically, the cytoplasm of monoblasts stains strongly and diffusely. Although the reaction can be completely inhibited with sodium fluoride, this step is unnecessary in most cases. Occasionally, myeloblasts stain weakly with ANA, with no inhibition by sodium fluoride. Megakaryoblasts stain negatively with ANB and typically show multifocal punctate reactivity with ANA, which is incompletely inhibited by sodium fluoride.41–45 Reactivity to these enzymes is stable for months in unfixed, unstained smears. In our experience, other stains are of limited value for the diagnosis and subclassification of leukemia. The naphtholASD-chloroacetate (CAE) or specific esterase stain identifies secondary lysosomal granules in maturing granulocytic precursors.12,46 The cytoplasmic staining pattern is diffuse granular; however, CAE is not as sensitive as MPO or SBB for identifying myeloblasts, as it stains only secondary lysosomes. Importantly, CAE remains stable in paraffin-embedded tissue and may be useful for diagnosis of granulocytic sarcoma (chloroma) in tissue sections (Leder stain).47 Smears from 1% to 2% of ALL cases may show weak granular cytoplasmic staining with CAE.48 These positive results are largely restricted to cases of granular ALL and are not apparent in tissue sections. Normal eosinophils are CAE negative, but abnormal eosinophils are positive in some subtypes of AML. The periodic acid Schiff (PAS) reaction stains glycogen in immature cells and is positive in most ALL cases, producing a fine-to-coarse granular staining pattern.49–51 It is not useful for routine classification of acute leukemia, as it produces a weak and diffuse staining pattern in one-third of AML cases49–51 and may produce intense block positivity in monoblasts and in erythroid precursors of MDS and erythroleukemia.52,53 PAS staining may also be positive in some small round cell tumors. Although normal eosinophils are PAS-negative, their leukemic counterparts may show granular PAS reactivity in some subtypes of AML. Acid phosphatase is present in all blood cell types, eliciting generally strong reactivity within the Golgi region in leukemic T lymphoblasts.54 However, some cases of B-cell precursor ALL may show an acid phosphatase staining pattern similar to that in T-cell cases, rendering the stain useless for ALL subclassification.55

Classification of acute leukemia The modern classification of leukemias requires the integration of morphologic examination, immunophenotypic (flow cytometric) analysis, cytogenetics and molecular findings. The morphologic examination has remained the

“gold standard” in establishing the diagnosis and guiding the selection of further studies. For that purpose, morphologic criteria, combined with cytochemical and immunophenotypic findings as established by the FAB Cooperative Group are still applied.12,13,29,35 However, identification of biologically significant leukemia subtypes and further risk stratification require the knowledge of immunophenotype and of cytogenetic and molecular lesions, as reflected in other leukemia classifications.14,56,57

Acute lymphoblastic leukemia Morphologic classification (the FAB system) The FAB classification system originally defined three subtypes of ALL (L1, L2, and L3),12,16 based solely on morphologic features (Figs. 2.2 to 2.5; Table 2.2). However, subsequent insights into the immunophenotype and biology of Burkitt lymphoma revealed that the L3 subtype of ALL represents the leukemic phase of this high-grade non-Hodgkin lymphoma with a mature B-cell immunophenotype. Furthermore, extensive studies have documented the importance of immunophenotypic, cytogenetic and molecular features of ALL in risk stratification and the lack of correlation between these latter findings and the L1 and L2 morphologic subtypes.58–68 Hence, the FAB classification of ALL has been largely abandoned in practice. Since the features outlined by the FAB classification are important in making the distinction between ALL and the leukemic phase of Burkitt lymphoma, which has major therapeutic implications, we have included a description of these morphologic categories in this chapter. L1-type blast cells are predominantly small (up to twice the diameter of a small lymphocyte) with homogeneous, finely dispersed-to-clumped nuclear chromatin, inconspicuous or absent nucleoli and scant deeply basophilic cytoplasm. ALL classified as L1 consists predominantly of morphologically homogeneous L1 blasts (Figs. 2.2 and 2.5). L2-type blasts are larger than twice the size of small lymphocytes, with prominent heterogeneity in blast size. The nuclear chromatin may be finely dispersed to coarsely condensed, with nuclear outline irregularities, and prominent nucleoli. The cytoplasm is usually abundant, with variable degrees of basophilia. ALL classified as L2 consists predominantly of blasts with considerable morphologic heterogeneity (Fig. 2.3).12 In practice, most ALLs show a morphologic spectrum between the L1 and L2 subtypes and therefore, even with the introduction of a scoring system,16 the distinction between these categories has remained somewhat arbitrary. It is important to note that both L1 and L2 blasts may contain cytoplasmic vacuoles in up to 30% of ALL cases, and in some cases

29

30

Mihaela Onciu and Ching-Hon Pui

Table 2.2 Classification of acute leukemias according to the revised French-American-British (FAB) criteria

FAB category

Percentage within the category

ALL L1

82

L2

15

L3

3

AML M0

2

M1

10–18

M2

27–29

M3

5–10

M4

16–25

M5

13–22

M6

1–3

M7

4–8

Diagnostic criteria

Small blasts with scanty cytoplasm, smooth-to-variably indented nuclear outline, fine-to-condensed nuclear chromatin and inconspicuous nucleoli in most blasts; often a variable percentage of larger blasts present Large and more heterogeneous blasts with moderately abundant cytoplasm, irregularly shaped nuclei, variable chromatin pattern, and prominent nucleoli; often a variable percentage of smaller blasts present Large, generally homogeneous blasts with finely stippled chromatin, round nuclei, prominent nucleoli, moderately abundant deeply basophilic cytoplasm with prominent vacuolization Large, usually agranular blasts, lacking Auer rods; negative for myeloperoxidase and Sudan Black B (SBB) by cytochemistry; expression of at least one myeloid antigen (e.g. CD13, CD33) by flow cytometry Myeloblasts with occasional Auer rods and/or cytochemical positivity for myeloperoxidase or SBB that take up ≥90% of nonerythroid cells; maturing myeloid cells 5% dysplastic eosinophilic precursors in the bone marrow Predominantly monocytic differentiation (≥80% of the marrow cells are of monocytic lineage, including monoblasts, promonocytes, and more mature monocytes); M5a has predominance of monoblasts (≥80% of leukemic cells); M5b shows maturation of the monocytic precursors (50 chromosomes, with almost 100% accuracy.220 A later study using higher density oligonucleotide arrays containing probes for most of the identified genes in the human genome has confirmed these data and the feasibility of classifying ALL cases using this approach (Fig. 2.22).221 In a study of T-cell ALL, Ferrando et al.129 identified distinct gene expression signatures of leukemic cells that corresponded to differentiation arrest at specific stages of normal thymocyte development: LYL1 + signature (pro-T), HOX11 + (early cortical thymocyte), and TAL1 + (late cortical thymocyte). Moreover, H0X11 and MLL-ENL subgroups were associated with a favorable prognosis. More recently, using gene expression profiling, we found that different genetic or lineage subtypes of ALL share common pathways of genomic response to the same treatment.222 The changes in gene expression were treatment specific, and the expression profiles could be used to illuminate differences in cellular responses to drug combinations versus single agents. The genes identified to date have also provided insights into the underlying biology of different leukemic subtypes and may become targets for novel therapy.129,218–220,223–225 Similar results have been obtained for pediatric and adult cases of acute myeloid leukemia.226–228 In addition to identifying gene expression profiles that correlate with the known molecular subtypes of AML, such studies have identified new prognostic groups among cases with normal cytogenetics and have shown that some of the AML subtypes (e.g. AML with MLL gene rearrangement and AML with FLT3 internal tandem duplications) are heterogeneous with respect to their gene expression profiles, with potential biologic and prognostic implications.227,229 Thus, analyses of the type described above may become an integral part of the diagnostic work-up of acute leukemia. Expression profiling of micro RNAs, which reflect the lineage and differentiation state of leukemic cells, could further enhance our ability to classify leukemia.230 It appears likely that, as our knowledge of leukemia pathobiology evolves, genetic or pathogenetic classification systems will supplant more traditional strategies, eventually becoming the gold standard for patient management.

3

4 5

6

7

8

9

10

11

12

13

14

15

16

REFERENCES 17 1 Miller, R. W., Young, J. L., Jr., & Novakovic, B. Childhood cancer. Cancer, 1995; 75: 395–405. 2 Hasle, H., Niemeyer, C. M., Chessells, J. M., et al. A pediatric approach to the WHO classification of myelodysplas-

18

tic and myeloproliferative diseases. Leukemia, 2003; 17: 277– 82. Casey, T. P. Chronic lymphocytic leukaemia in a child presenting at the age of two years and eight months. Australas Ann Med, 1968; 17: 70–4. Sardemann, H. Chronic lymphocytic leukemia in an infant. Acta Paediatr Scand, 1972; 61, 213–6. Sonnier, J. A., Buchanan, G. R., Howard-Peebles, P. N., Rutledge, J., & Smith, R. G. Chromosomal translocation involving the immunoglobulin kappa-chain and heavy-chain loci in a child with chronic lymphocytic leukemia. N Engl J Med, 1983; 309: 590–4. Yoffe, G., Howard-Peebles, P. N., Smith, R. G., Tucker, P. W., & Buchanan, G. R. Childhood chronic lymphocytic leukemia with (2;14) translocation. J Pediatr, 1990; 116: 114–7. Gajjar, A., Ribeiro, R., Hancock, M. L., et al. Persistence of circulating blasts after 1 week of multiagent chemotherapy confers a poor prognosis in childhood acute lymphoblastic leukemia. Blood, 1995; 86: 1292–5. Ellis, L. D., Johnson, B. J., & Westerman, M. P. Needle biopsy of bone and marrow: an experience with 1,445 biopsies. Arch Intern Med, 1964; 114: 213–21. Jamshidi, K. & Swaim, W. R. Bone marrow biopsy with unaltered architecture: a new biopsy device. J Lab Clin Med, 1971; 77: 335–42. Goldenberg, A. S. & Tiesinga, J. J. Clinical experience with a new specimen capturing bone marrow biopsy needle. Am J Hematol, 2001; 68: 189–93. Aboul-Nasr, R., Estey, E. H., Kantarjian, H. M., et al. Comparison of touch imprints with aspirate smears for evaluating bone marrow specimens. Am J Clin Pathol, 1999; 111: 753–8. Bennett, J. M., Catovsky, D., Daniel, M. T., et al. Proposals for the classification of the acute leukaemias. French-AmericanBritish (FAB) Co-operative Group. Br J Haematol, 1976; 33: 451–8. Bennett, J. M., Catovsky, D., Daniel, M. T., et al. Proposed revised criteria for the classification of acute myeloid leukemia. A report of the French-American-British Cooperative Group. Ann Intern Med, 1985; 103: 620–5. Jaffe, E. S., Harris, N. L., Stein, H., & Vardiman, J. W., eds. World Health Organization Classification of Tumours. Tumours of Haematopoietic and Lymphoid Tissues (Lyon, France: IARC Press, 2001). Vardiman, J. W., Harris, N. L., & Brunning, R. D. The World Health Organization (WHO) classification of the myeloid neoplasms. Blood, 2002; 100: 2292–302. Bennett, J. M., Catovsky, D., Daniel, M. T., et al. The morphological classification of acute lymphoblastic leukaemia: concordance among observers and clinical correlations. Br J Haematol, 1981; 47: 553–61. Lilleyman, J. S., Hann, I. M., Stevens, R. F., Richards, S. M., & Eden, O. B. Blast cell vacuoles in childhood lymphoblastic leukaemia. Br J Haematol, 1988; 70: 183–6. Lilleyman, J. S., Hann, I. M., Stevens, R. F., et al. Cytomorphology of childhood lymphoblastic leukaemia: a prospective

Diagnosis and classification

19

20

21

22

23 24

25

26

27

28

29

30

31

32

study of 2000 patients. United Kingdom Medical Research Council’s Working Party on Childhood Leukaemia. Br J Haematol, 1992; 81: 52–7. Potter, V. P., Sorell, M., Baglivo, J. A., Sather, H., & Miller, D. R. Prognostic significance of vacuoles in L1 lymphoblasts in childhood acute lymphoblastic leukaemia: a report from Children’s Cancer Study Group. Br J Haematol, 1984; 56: 215– 22. Cantu-Rajnoldi, A., Invernizzi, R., Biondi, A., et al. Biological and clinical features of acute lymphoblastic leukaemia with cytoplasmic granules or inclusions: description of eight cases. Br J Haematol, 1989; 73: 309–14. Cerezo, L., Shuster, J. J., Pullen, D. J., et al. Laboratory correlates and prognostic significance of granular acute lymphoblastic leukemia in children. A Pediatric Oncology Group study. Am J Clin Pathol, 1991; 95: 526–31. Darbyshire, P. J. & Lilleyman, J. S. Granular acute lymphoblastic leukaemia of childhood: a morphological phenomenon. J Clin Pathol, 1987; 40: 251–3. Stein, P., Peiper, S., Butler, D., et al. Granular acute lymphoblastic leukemia. Am J Clin Pathol, 1983; 79: 426–30. Miller, D. R., Steinherz, P. G., Feuer, D., Sather, H., & Hammond, D. Unfavorable prognostic significance of hand mirror cells in childhood acute lymphoblastic leukemia. A report from the childrens Cancer Study Group. Am J Dis Child, 1983; 137: 346– 50. Schumacher, H. R., Champion, J. E., Thomas, W. J., Pitts, L. L., & Stass, S. A. Acute lymphoblastic leukemia – hand mirror variant. An analysis of a large group of patients. Am J Hematol, 1979; 7: 11–17. Sjogren, U. & Garwicz, S. Prognostic significance of amoeboid movement configuration in lymphoid cells from children with acute lymphoblastic leukaemia. Scand J Haematol, 1980; 24: 335–9. Hayhoe, F. G., Quaglino, D., & Flemans, R. J. Consecutive use of Romanowsky and periodic-acid-Schiff techniques in the study of blood and bone marrow cells. Br J Haematol, 1960; 6: 23–5. Kowal-Vern, A., Cotelingam, J., & Schumacher, H. R. The prognostic significance of proerythroblasts in acute erythroleukemia. Am J Clin Pathol, 1992; 98: 34–40. Bennett, J. M., Catovsky, D., Daniel, M. T., et al. Criteria for the diagnosis of acute leukemia of megakaryocyte lineage (M7). A report of the French-American-British Cooperative Group. Ann Intern Med, 1985; 103: 460–2. Huang, M. J., Li, C. Y., Nichols, W. L., Young, J. H., & Katzmann, J. A. Acute leukemia with megakaryocytic differentiation: a study of 12 cases identified immunocytochemically. Blood, 1984; 64: 427–39. Penchansky, L., Taylor, S. R., & Krause, J. R. Three infants with acute megakaryoblastic leukemia simulating metastatic tumor. Cancer, 1989; 64: 1366–71. Pui, C. H., Rivera, G., Mirro, J., et al. Acute megakaryoblastic leukemia. Blast cell aggregates simulating metastatic tumor. Arch Pathol Lab Med, 1985; 109: 1033–5.

33 Head, D. R., Cerezo, L., Savage, R. A., et al. Institutional performance in application of the FAB classification of acute leukemia. The Southwest Oncology Group experience. Cancer, 1985; 55: 1979–86. 34 Hanker, J. S., Ambrose, W. W., James, C. J., et al. Facilitated light microscopic cytochemical diagnosis of acute myelogenous leukemia. Cancer Res, 1979; 39: 1635–9. 35 Bennett, J. M., Catovsky, D., Daniel, M. T., et al. Proposal for the recognition of minimally differentiated acute myeloid leukaemia (AML-MO). Br J Haematol, 1991; 78: 325–9. 36 Venditti, A., Del Poeta, G., Stasi, R., et al. Minimally differentiated acute myeloid leukaemia (AML-M0): cytochemical, immunophenotypic and cytogenetic analysis of 19 cases. Br J Haematol, 1994; 88: 784–93. 37 Charak, B. S., Advani, S. H., Karandikar, S. M., et al. Sudan black B positivity in acute lymphoblastic leukemia. Acta Haematol, 1988; 80: 199–202. 38 Stass, S. A., Pui, C. H., Melvin, S., et al. Sudan black B positive acute lymphoblastic leukaemia. Br J Haematol, 1984; 57: 413– 21. 39 Tricot, G., Broeckaert-Van Orshoven, A., Hoof, A. Van, & Verwilghen, R. L. Sudan Black B positivity in acute lymphoblastic leukaemia. Br J Haematol, 1982; 51: 615–21. 40 Li, C. Y., Lam, K. W., & Yam, L. T. Esterases in human leukocytes. J Histochem Cytochem, 1973; 21: 1–12. 41 Koike, T. Megakaryoblastic leukemia: the characterization and identification of megakaryoblasts. Blood, 1984; 64: 683–92. 42 Peterson, B. A. & Levine, E. G. Uncommon subtypes of acute nonlymphocytic leukemia: clinical features and management of FAB M5, M6 and M7. Semin Oncol, 1987; 14: 425–34. 43 Pombo De Oliveira, M. S., Gregory, C., Matutes, E., Parreira, A., & Catovsky, D. Cytochemical profile of megakaryoblastic leukaemia: a study with cytochemical methods, monoclonal antibodies, and ultrastructural cytochemistry. J Clin Pathol, 1987; 40: 663–9. 44 Pui, C. H., Williams, D. L., Scarborough, V., et al. Acute megakaryoblastic leukaemia associated with intrinsic platelet dysfunction and constitution ring 21 chromosome in a young boy. Br J Haematol, 1982; 50: 191–200. 45 Ribeiro, R. C., Oliveira, M. S., Fairclough, D., et al. Acute megakaryoblastic leukemia in children and adolescents: a retrospective analysis of 24 cases. Leuk Lymphoma, 1993; 10: 299–306. 46 Yam, L. T., Li, C. Y., & Crosby, W. H. Cytochemical identification of monocytes and granulocytes. Am J Clin Pathol, 1971; 55: 283–90. 47 Neiman, R. S., Barcos, M., Berard, C., et al. Granulocytic sarcoma: a clinicopathologic study of 61 biopsied cases. Cancer, 1981; 48: 1426–37. 48 Keifer, J., Abromowitch, M., & Stass, S. A. Chloroacetate esterase positivity in acute lymphoblastic leukemia. Am J Clin Pathol, 1985; 83: 647–9. 49 Bennett, J. M. & Dutcher, T. F. The cytochemistry of acute leukemia: observations on glycogen and neutral fat in bone marrow aspirates. Blood, 1969; 33: 341–7.

41

42

Mihaela Onciu and Ching-Hon Pui

50 Hayhoe, F. G. J. & Cawley, J. C. Acute leukemia:cellular morphology, cytochemistry and fine structure. Clin Haematol, 1972; 1: 49–94. 51 Snower, D. P., Smith, B. R., Munz, U. J., & McPhedran, P. Reevaluation of the periodic acid-Schiff stain in acute leukemia with immunophenotypic analyses. Arch Pathol Lab Med, 1991; 115: 346–50. 52 Quaglino, D. & Hayhoe, F. G. Periodic-acid-Schiff positivity in erythroblasts with special reference to Di Guglielmo’s disease. Br J Haematol, 1960; 6: 26–33. 53 Skinnider, L. F. & Ghadially, F. N. Glycogen in erythroid cells. Arch Pathol, 1973; 95: 139–41. 54 Catovsky, D., Cherchi, M., Greaves, M. F., et al. Acidphosphatase reaction in acute lymphoblastic leukaemia. Lancet, 1978; 1: 749–51. 55 Head, D. R., Borowitz, M., Cerezo, L., et al. Acid phosphatase positivity in childhood acute lymphocytic leukemia. Am J Clin Pathol, 1986; 86: 650–3. 56 Morphologic, immunologic, and cytogenetic (MIC) working classification of acute lymphoblastic leukemias. Report of the workshop held in Leuven, Belgium, April 22–23, 1985. First MIC Cooperative Study Group. Cancer Genet Cytogenet, 1986; 23: 189–97. 57 Morphologic, immunologic and cytogenetic (MIC) working classification of the acute myeloid leukaemias. Second MIC Cooperative Study Group. Br J Haematol, 1988; 68: 487–94. 58 Cline, M. J. The molecular basis of leukemia. N Engl J Med, 1994; 330: 328–36. 59 Kersey, J., Nesbit, M., Hallgren, H., et al. Evidence for origin of certain childhood acute lymphoblastic leukemias and lymphomas in thymus-derived lymphocytes. Cancer, 1975; 36: 1348–52. 60 Look, A. T. Oncogenic transcription factors in the human acute leukemias. Science, 1997; 278: 1059–64. 61 Pui, C. H., Raimondi, S. C., Dodge, R. K., et al. Prognostic importance of structural chromosomal abnormalities in children with hyperdiploid (greater than 50 chromosomes) acute lymphoblastic leukemia. Blood, 1989; 73: 1963–7. 62 Pui, C. H., Behm, F. G., Singh, B., et al. Heterogeneity of presenting features and their relation to treatment outcome in 120 children with T-cell acute lymphoblastic leukemia. Blood, 1990; 75: 174–9. 63 Pui, C. H., Behm, F. G., & Crist, W. M. Clinical and biologic relevance of immunologic marker studies in childhood acute lymphoblastic leukemia. Blood, 1993; 82: 343–62. 64 Rubin, C. M., Le Beau, M. M., Mick, R., et al. Impact of chromosomal translocations on prognosis in childhood acute lymphoblastic leukemia. J Clin Oncol, 1991; 9: 2183–92. 65 Secker-Walker, L. M., Lawler, S. D., & Hardisty, R. M. Prognostic implications of chromosomal findings in acute lymphoblastic leukaemia at diagnosis. Br Med J, 1978; 2: 1529–30. 66 Sen, L. & Borella, L. Clinical importance of lymphoblasts with T markers in childhood acute leukemia. N Engl J Med, 1975; 292: 828–32. 67 Trueworthy, R., Shuster, J., Look, T., et al. Ploidy of lymphoblasts is the strongest predictor of treatment outcome in

68

69

70 71

72 73

74

75

76

77

78

79

80

81

82 83

B-progenitor cell acute lymphoblastic leukemia of childhood: a Pediatric Oncology Group study. J Clin Oncol, 1992; 10: 606– 13. Williams, D. L., Tsiatis, A., Brodeur, G. M., et al. Prognostic importance of chromosome number in 136 untreated children with acute lymphoblastic leukemia. Blood, 1982; 60: 864–71. Dyment, P. G., Savage, R. A., & McMahon, J. T. Anomalous azurophilic granules in acute lymphoblastic leukemia. Am J Pediatr Hematol Oncol, 1982; 4: 207–11. Hecht, J. L. & Aster, J. C. Molecular biology of Burkitt’s lymphoma. J Clin Oncol, 2000; 18: 3707–21. Bene, M. C., Castoldi, G., Knapp, W., et al. Proposals for the immunological classification of acute leukemias. European Group for the Immunological Characterization of Leukemias (EGIL). Leukemia, 1995; 9: 1783–6. Brouet, J. C. & Seligmann, M. The immunological classification of acute lymphoblastic leukemias. Cancer, 1978; 42: 817–27. Campana, D., Dongen, J. J. van, Mehta, A., et al. Stages of Tcell receptor protein expression in T-cell acute lymphoblastic leukemia. Blood, 1991; 77: 1546–54. Rothe, G. & Schmitz, G. Consensus protocol for the flow cytometric immunophenotyping of hematopoietic malignancies. Working Group on Flow Cytometry and Image Analysis. Leukemia, 1996; 10: 877–95. Drexler, H. G., Thiel, E., & Ludwig, W. D. Review of the incidence and clinical relevance of myeloid antigen-positive acute lymphoblastic leukemia. Leukemia, 1991; 5: 637–45. Fink, F. M., Koller, U., Mayer, H., et al. Prognostic significance of myeloid-associated antigen expression on blast cells in children with acute lymphoblastic leukemia. The Austrian Pediatric Oncology Group. Med Pediatr Oncol, 1993; 21: 340–6. Kurec, A. S., Belair, P., Stefanu, C., et al. Significance of aberrant immunophenotypes in childhood acute lymphoid leukemia. Cancer, 1991; 67: 3081–6. Pui, C. H., Behm, F. G., Singh, B., et al. Myeloid-associated antigen expression lacks prognostic value in childhood acute lymphoblastic leukemia treated with intensive multiagent chemotherapy. Blood, 1990; 75: 198–202. Uckun, F. M., Nachman, J. B., Sather, H. N., et al. Clinical significance of Philadelphia chromosome positive pediatric acute lymphoblastic leukemia in the context of contemporary intensive therapies: a report from the Children’s Cancer Group. Cancer, 1998; 83: 2030–9. Pui, C. H., Rubnitz, J. E., Hancock, M. L., et al. Reappraisal of the clinical and biologic significance of myeloid-associated antigen expression in childhood acute lymphoblastic leukemia. J Clin Oncol, 1998; 16: 3768–73. Putti, M. C., Rondelli, R., Cocito, M. G., et al. Expression of myeloid markers lacks prognostic impact in children treated for acute lymphoblastic leukemia: Italian experience in AIEOPALL 88–91 studies. Blood, 1998; 92: 795–801. Pui, C. H. Acute lymphoblastic leukemia in children. Curr Opin Oncol, 2000; 12: 3–12. Pui, C. H., Campana, D., & Evans, W. E. Childhood acute lymphoblastic leukaemia – current status and future perspectives. Lancet Oncol, 2001; 2: 597–607.

Diagnosis and classification

84 Ferrando, A. A. & Look, A. T. Clinical implications of recurring chromosomal and associated molecular abnormalities in acute lymphoblastic leukemia. Semin Hematol, 2000; 37: 381–95. 85 Harrison, C. J., Martineau, M., & Secker-Walker, L. M. The Leukaemia Research Fund/United Kingdom Cancer Cytogenetics Group Karyotype Database in acute lymphoblastic leukaemia: a valuable resource for patient management. Br J Haematol, 2001; 113: 3–10. 86 Recommendations for a morphologic, immunologic, and cytogenetic (MIC) working classification of the primary and therapy-related myelodysplastic disorders. Report of the workshop held in Scottsdale, Arizona, USA, on February 23–25, 1987. Third MIC Cooperative Study Group. Cancer Genet Cytogenet, 1988; 32: 1–10. 87 Harrison, C. J. The detection and significance of chromosomal abnormalities in childhood acute lymphoblastic leukaemia. Blood Rev, 2001; 15: 49–59. 88 Pui, C. H., Williams, D. L., Raimondi, S. C., et al. Hypodiploidy is associated with a poor prognosis in childhood acute lymphoblastic leukemia. Blood, 1987; 70: 247–53. 89 Pui, C. H., Williams, D. L., Roberson, P. K., et al. Correlation of karyotype and immunophenotype in childhood acute lymphoblastic leukemia. J Clin Oncol, 1988; 6: 56–61. 90 Pui, C. H., Carroll, A. J., Head, D., et al. Near-triploid and neartetraploid acute lymphoblastic leukemia of childhood. Blood, 1990; 76: 590–6. 91 Pui, C. H., Carroll, A. J., Raimondi, S. C., et al. Clinical presentation, karyotypic characterization, and treatment outcome of childhood acute lymphoblastic leukemia with a near-haploid or hypodiploid less than 45 line. Blood, 1990; 75: 1170–7. 92 Raimondi, S. C., Pui, C. H., Hancock, M. L., et al. Heterogeneity of hyperdiploid (51–67) childhood acute lymphoblastic leukemia. Leukemia, 1996; 10: 213–24. 93 Secker-Walker, L. M., Chessells, J. M., Stewart, E. L., et al. Chromosomes and other prognostic factors in acute lymphoblastic leukaemia: a long-term follow-up. Br J Haematol, 1989; 72: 336– 42. 94 Heerema, N. A., Nachman, J. B., Sather, H. N., et al. Hypodiploidy with less than 45 chromosomes confers adverse risk in childhood acute lymphoblastic leukemia: a report from the Children’s Cancer Group. Blood, 1999; 94: 4036–45. 95 Look, A. T., Melvin, S. L., Williams, D. L., et al. Aneuploidy and percentage of S-phase cells determined by flow cytometry correlate with cell phenotype in childhood acute leukemia. Blood, 1982; 60: 959–67. 96 Look, A. T., Roberson, P. K., Williams, D. L., et al. Prognostic importance of blast cell DNA content in childhood acute lymphoblastic leukemia. Blood, 1985; 65: 1079–86. 97 Ito, C., Kumagai, M., Manabe, A., et al. Hyperdiploid acute lymphoblastic leukemia with 51 to 65 chromosomes: a distinct biological entity with a marked propensity to undergo apoptosis. Blood, 1999; 93: 315–20. 98 Pui, C. H., Crist, W. M., & Look, A. T. Biology and clinical significance of cytogenetic abnormalities in childhood acute lymphoblastic leukemia. Blood, 1990; 76: 1449–63.

99 Rubnitz, J. E., Downing, J. R., Pui, C. H., et al. TEL gene rearrangement in acute lymphoblastic leukemia: a new genetic marker with prognostic significance. J Clin Oncol, 1997; 15: 1150–7. 100 Shurtleff, S. A., Buijs, A., Behm, F. G., et al. TEL/AML1 fusion resulting from a cryptic t(12;21) is the most common genetic lesion in pediatric ALL and defines a subgroup of patients with an excellent prognosis. Leukemia, 1995; 9: 1985–9. 101 Filatov, L. F., Behm, F. G., & Pui, C. H. Childhood acute lymphoblastic leukemias with equivocal chromosome markers of the t(1;19) translocation. Genes Chromosom Cancer, 1995; 13: 99–103. 102 Privitera, E., Kamps, M. P., Hayashi, Y., et al. Different molecular consequences of the 1;19 chromosomal translocation in childhood B-cell precursor acute lymphoblastic leukemia. Blood, 1992; 79: 1781–8. 103 Pui, C. H., Raimondi, S. C., Hancock, M. L., et al. Immunologic, cytogenetic, and clinical characterization of childhood acute lymphoblastic leukemia with the t(1;19) (q23; p13) or its derivative. J Clin Oncol, 1994; 12: 2601–6. 104 Rieder, H., Kolbus, U., Koop, U., et al. Translocation t(1;22) mimicking t(1;19) in a child with acute lymphoblastic leukemia as revealed by chromosome painting. Leukemia, 1993; 7: 1663–6. 105 Behm, F. G., Raimondi, S. C., Frestedt, J. L., et al. Rearrangement of the MLL gene confers a poor prognosis in childhood acute lymphoblastic leukemia, regardless of presenting age. Blood, 1996; 87: 2870–7. 106 Raimondi, S. C., Frestedt, J. L., Pui, C. H., et al. Acute lymphoblastic leukemias with deletion of 11q23 or a novel inversion (11)(p13q23) lack MLL gene rearrangements and have favorable clinical features. Blood, 1995; 86: 1881–6. 107 Loh, M. L. & Rubnitz, J. E. TEL/AML1-positive pediatric leukemia: prognostic significance and therapeutic approaches. Curr Opin Hematol, 2002; 9: 345–52. 108 Romana, S. P., Poirel, H., Leconiat, M., et al. High frequency of t(12;21) in childhood B-lineage acute lymphoblastic leukemia. Blood, 1995; 86: 4263–9. 109 Romana, S. P., Le Coniat, M., & Berger, R. t(12;21): a new recurrent translocation in acute lymphoblastic leukemia. Genes Chromosomes Cancer, 1994; 9: 186–91. 110 Baruchel, A., Cayuela, J. M., Ballerini, P., et al. The majority of myeloid-antigen-positive (My+) childhood B-cell precursor acute lymphoblastic leukaemias express TEL-AML1 fusion transcripts. Br J Haematol, 1997; 99: 101–6. 111 Weir, E. G. & Borowitz, M. J. Flow cytometry in the diagnosis of acute leukemia. Semin Hematol, 2001; 38: 124–38. 112 Ramakers-van Woerden, N. L., Pieters, R., Loonen, A. H., et al. TEL/AML1 gene fusion is related to in vitro drug sensitivity for L-asparaginase in childhood acute lymphoblastic leukemia. Blood, 2000; 96: 1094–9. 113 Hunger, S. P., Galili, N., Carroll, A. J., et al. The t(1;19)(q23;p13) results in consistent fusion of E2A and PBX1 coding sequences in acute lymphoblastic leukemias. Blood, 1991; 77: 687–93. 114 Izraeli, S., Henn, T., Strobl, H., et al. Expression of identical E2A/PBX1 fusion transcripts occurs in both pre-B and

43

44

Mihaela Onciu and Ching-Hon Pui

115

116

117

118

119

120

121

122

123

124

125

126

127

early pre-B immunological subtypes of childhood acute lymphoblastic leukemia. Leukemia, 1993; 7: 2054–6. Arico, M., Valsecchi, M. G., Camitta, B., et al. Outcome of treatment in children with Philadelphia chromosome-positive acute lymphoblastic leukemia. N Engl J Med, 2000; 342: 998– 1006. Silva, M. L., Fernandez, T. S., de Souza, M. H., et al. M-BCR rearrangement in a case of T-cell childhood acute lymphoblastic leukemia. Med Pediatr Oncol, 1999; 32: 455–6 Ribeiro, R. C., Abromowitch, M., Raimondi, S. C., et al. Clinical and biologic hallmarks of the Philadelphia chromosome in childhood acute lymphoblastic leukemia. Blood, 1987; 70: 948–53. Borkhardt, A., Wuchter, C., Viehmann, S., et al. Infant acute lymphoblastic leukemia – combined cytogenetic, immunophenotypical and molecular analysis of 77 cases. Leukemia, 2002; 16: 1685–90. Chessells, J. M., Harrison, C. J., Watson, S. L., Vora, A. J., & Richards, S. M. Treatment of infants with lymphoblastic leukaemia: results of the UK Infant Protocols 1987–1999. Br J Haematol, 2002; 117: 306–14. Chessells, J. M., Harrison, C. J., Kempski, H., et al. Clinical features, cytogenetics and outcome in acute lymphoblastic and myeloid leukaemia of infancy: report from the MRC Childhood Leukaemia working party. Leukemia, 2002; 16: 776–84. Heerema, N. A., Sather, H. N., Ge, J., et al. Cytogenetic studies of infant acute lymphoblastic leukemia: poor prognosis of infants with t(4;11) – a report of the Children’s Cancer Group. Leukemia, 1999; 13: 679–86. Pui, C. H., Gaynon, P. S., Boyett, J. M., et al. Outcome of treatment in childhood acute lymphoblastic leukaemia with rearrangements of the 11q23 chromosomal region. Lancet, 2002; 359: 1909–15. Hilden, J. M., Smith, F. O., Frestedt, J. L., et al. MLL gene rearrangement, cytogenetic 11q23 abnormalities, and expression of the NG2 molecule in infant acute myeloid leukemia. Blood, 1997; 89: 3801–5. Smith, F. O., Rauch, C., Williams, D. E., et al. The human homologue of rat NG2, a chondroitin sulfate proteoglycan, is not expressed on the cell surface of normal hematopoietic cells but is expressed by acute myeloid leukemia blasts from poorprognosis patients with abnormalities of chromosome band 11q23. Blood, 1996; 87:1123–33. Pieters, R., den Boer, M. L., Durian, M., et al. Relation between age, immunophenotype and in vitro drug resistance in 395 children with acute lymphoblastic leukemia – implications for treatment of infants. Leukemia, 1998; 12: 1344–8. Stam, R. W., Boer, M. L.den, Meijerink, J. P., et al. Differential mRNA expression of Ara-C-metabolizing enzymes explains Ara-C sensitivity in MLL gene-rearranged infant acute lymphoblastic leukemia. Blood, 2003; 101: 1270–6. Schneider, N. R., Carroll, A. J., Shuster, J. J., et al. New recurring cytogenetic abnormalities and association of blast cell karyotypes with prognosis in childhood T-cell acute lymphoblas-

128

129

130

131

132

133

134

135 136

137

138

139

140

141

tic leukemia: a pediatric oncology group report of 343 cases. Blood, 2000; 96: 2543–9. Ballerini, P., Blaise, A., Busson-Le Coniat, M., et al. HOX11L2 expression defines a clinical subtype of pediatric T-ALL associated with poor prognosis. Blood, 2002; 100: 991–7. Ferrando, A. A., Neuberg, D. S., Staunton, J., et al. Gene expression signatures define novel oncogenic pathways in T cell acute lymphoblastic leukemia. Cancer Cell, 2002; 1: 75–87. Ulich, T. R., del Castillo, J., Yin, S. M., & Egrie, J. C. The erythropoietic effects of interleukin 6 and erythropoietin in vivo. Exp Hematol, 1991; 19: 29–34. Kotylo, P. K., Seo, I. S., Smith, F. O., et al. Flow cytometric immunophenotypic characterization of pediatric and adult minimally differentiated acute myeloid leukemia (AML-M0). Am J Clin Pathol, 2000; 113: 193–200. Roumier, C., Eclache, V., Imbert, M., et al. M0 AML, clinical and biologic features of the disease, including AML1 gene mutations: a report of 59 cases by the Groupe Francais d’Hematologie Cellulaire (GFHC) and the Groupe Francais de Cytogenetique Hematologique (GFCH). Blood, 2003; 101: 1277–83. Arber, D. A., Stein, A. S., Carter, N. H., et al. Prognostic impact of acute myeloid leukemia classification. Importance of detection of recurring cytogenetic abnormalities and multilineage dysplasia on survival. Am J Clin Pathol 2003; 119: 672–80. Cantu-Rajnoldi, A., Biondi, A., Jankovic, M., et al. Diagnosis and incidence of acute promyelocytic leukemia (FAB M3 and M3 variant) in childhood. Blood, 1993; 81: 2209–10. Chan, K. W., Steinherz, P. G., & Miller, D. R. Acute promyelocytic leukemia in children. Med Pediatr Oncol, 1981; 9: 5–15. Rovelli, A., Biondi, A., Cantu, R. A., et al. Microgranular variant of acute promyelocytic leukemia in children. J Clin Oncol, 1992; 10: 1413–18. Biondi, A., Rovelli, A., Cantu-Rajnoldi, A., et al. Acute promyelocytic leukemia in children: experience of the Italian Pediatric Hematology and Oncology Group (AIEOP). Leukemia, 1994; 8: 1264–8. Biondi, A., Luciano, A., Bassan, R., et al. CD2 expression in acute promyelocytic leukemia is associated with microgranular morphology (FAB M3v) but not with any PML gene breakpoint. Leukemia, 1995; 9: 1461–6. Guglielmi, C., Martelli, M. P., Diverio, D., et al. Immunophenotype of adult and childhood acute promyelocytic leukaemia: correlation with morphology, type of PML gene breakpoint and clinical outcome. A cooperative Italian study on 196 cases. Br J Haematol, 1998; 102: 1035–41. Falini, B., Flenghi, L., Fagioli, M., et al. Immunocytochemical diagnosis of acute promyelocytic leukemia (M3) with the monoclonal antibody PG-M3 (anti-PML). Blood, 1997; 90: 4046–53. Chen, Z., Brand, N. J., Chen, A., et al. Fusion between a novel Kruppel-like zinc finger gene and the retinoic acid receptoralpha locus due to a variant t(11;17) translocation associated with acute promyelocytic leukaemia. EMBO J, 1993; 12: 1161–7.

Diagnosis and classification

142 Bennett, J. M., Catovsky, D., Daniel, M. T., et al. Hypergranular promyelocytic leukemia: correlation between morphology and chromosomal translocations including t(15;17) and t(11;17). Leukemia, 2000; 14: 1197–200. 143 Sainty, D., Liso, V., Cantu-Rajnoldi, A., et al. on behalf of Group Franc¸ais d’Haematologie Cellulaire, Group Francais de Cytogenetique Hematologique, UK Cancer Cytogenetics Group and BIOMED 1 European Community-Concerted Action “Molecular Cytogenetic Diagnosis in Haematological Malignancies”. A new morphologic classification system for acute promyelocytic leukemia distinguishes cases with underlying PLZF/RARA gene rearrangements. Blood, 2000; 96: 1287– 96. 144 Baer, M. R., Stewart, C. C., Lawrence, D., et al. Acute myeloid leukemia with 11q23 translocations: myelomonocytic immunophenotype by multiparameter flow cytometry. Leukemia, 1998; 12: 317–25. 145 Malkin, D. & Freedman, M. H. Childhood erythroleukemia: review of clinical and biological features. Am J Pediatr Hematol Oncol, 1989; 11: 348–59. 146 Mirchandani, I. & Palutke, M. Acute megakaryoblastic leukemia. Cancer, 1982; 50: 2866–72. 147 Dastugue, N., Lafage-Pochitaloff, M., Pages, M. P., et al. Cytogenetic profile of childhood and adult megakaryoblastic leukemia (M7): a study of the Groupe Francais de Cytogenetique Hematologique (GFCH). Blood, 2002; 100: 618–26. 148 Athale, U. H., Razzouk, B. I., Raimondi, S. C., et al. Biology and outcome of childhood acute megakaryoblastic leukemia: a single institution’s experience. Blood, 2001; 97: 3727–32. 149 Byrd, J. C., Dodge, R. K., Carroll, A., et al. Patients with t(8;21)(q22;q22) and acute myeloid leukemia have superior failure-free and overall survival when repetitive cycles of highdose cytarabine are administered. J Clin Oncol, 1999; 17: 3767– 75. 150 Grimwade, D., Walker, H., Oliver, F., et al. The importance of diagnostic cytogenetics on outcome in AML: analysis of 1,612 patients entered into the MRC AML 10 trial. The Medical Research Council Adult and Children’s Leukaemia Working Parties. Blood, 1998; 92: 2322–33. 151 Raimondi, S. C., Chang, M. N., Ravindranath, Y., et al. Chromosomal abnormalities in 478 children with acute myeloid leukemia: clinical characteristics and treatment outcome in a cooperative pediatric oncology group study-POG 8821. Blood, 1999; 94: 3707–16. 152 Rubnitz, J. E., Raimondi, S. C., Hall, S., et al. Characteristics and outcome of t(8;21)-positive childhood acute myeloid leukemia: a single institution’s experience. Leukemia, 2002; 16: 2072–7. 153 Felice, M. S., Zubizarreta, P. A., Alfaro, E. M., et al. Good outcome of children with acute myeloid leukemia and t(8;21)(q22;q22), even when associated with granulocytic sarcoma: a report from a single institution in Argentina. Cancer, 2000; 88: 1939–44. 154 Andrieu, V., Radford-Weiss, I., Troussard, X., et al. Molecular detection of t(8;21)/AML1-ETO in AML M1/M2: correlation

155

156

157

158

159

160

161

162

163

164

165

166

167

with cytogenetics, morphology and immunophenotype. Br J Haematol, 1996; 92: 855–65. Haferlach, T., Bennett, J. M., Loffler, H., et al. Acute myeloid leukemia with translocation (8;21). Cytomorphology, dysplasia and prognostic factors in 41 cases. AML Cooperative Group and ECOG. Leuk Lymphoma, 1996; 23: 227–34. Nucifora, G., Dickstein, J. I., Torbenson, V., et al. Correlation between cell morphology and expression of the AML1/ETO chimeric transcript in patients with acute myeloid leukemia without the t(8;21). Leukemia, 1994; 8: 1533–8. Swirsky, D. M., Li, Y. S., Matthews, J. G., et al. 8;21 translocation in acute granulocytic leukaemia: cytological, cytochemical and clinical features. Br J Haematol, 1984; 56: 199–213. Hurwitz, C. A., Raimondi, S. C., Head, D., et al. Distinctive immunophenotypic features of t(8;21)(q22;q22) acute myeloblastic leukemia in children. Blood, 1992; 80: 3182–8. Ferrara, F., Di Noto, R., Annunziata, M., et al. Immunophenotypic analysis enables the correct prediction of t(8;21) in acute myeloid leukaemia. Br J Haematol, 1998; 102: 444–8. Arthur, D. C. & Bloomfield, C. D. Partial deletion of the long arm of chromosome 16 and bone marrow eosinophilia in acute nonlymphocytic leukemia: a new association. Blood, 1983; 61: 994–8. Bitter, M. A., Le Beau, M. M., Larson, R. A., et al. A morphologic and cytochemical study of acute myelomonocytic leukemia with abnormal marrow eosinophils associated with inv(16)(p13q22). Am J Clin Pathol, 1984; 81: 733–41. Le Beau, M. M., Larson, R. A., Bitter, M. A., et al. Association of an inversion of chromosome 16 with abnormal marrow eosinophils in acute myelomonocytic leukemia. A unique cytogenetic-clinicopathological association. N Engl J Med 1983; 309: 630–6. Haferlach, T., Winkemann, M., Loffler, H., et al. The abnormal eosinophils are part of the leukemic cell population in acute myelomonocytic leukemia with abnormal eosinophils (AML M4Eo) and carry the pericentric inversion 16: a combination of May-Grunwald-Giemsa staining and fluorescence in situ hybridization. Blood, 1996; 87: 2459–63. Adriaansen, H. J., te Boekhorst, P. A., Hagemeijer, A. M., et al. Acute myeloid leukemia M4 with bone marrow eosinophilia (M4Eo) and inv(16)(p13q22) exhibits a specific immunophenotype with CD2 expression. Blood, 1993; 81: 3043–51. Byrd, J. C., Mrozek, K., Dodge, R. K., et al. Pretreatment cytogenetic abnormalities are predictive of induction success, cumulative incidence of relapse, and overall survival in adult patients with de novo acute myeloid leukemia: results from Cancer and Leukemia Group B (CALGB 8461). Blood, 2002; 100: 4325–36. Haferlach, T., Schoch, C., Loffler, H., et al. Morphologic dysplasia in de novo acute myeloid leukemia (AML) is related to unfavorable cytogenetics but has no independent prognostic relevance under the conditions of intensive induction therapy: results of a multiparameter analysis from the German AML Cooperative Group studies. J Clin Oncol, 2003; 21: 256–65. Razzouk, B. I., Raimondi, S. C., Srivastava, D. K., et al. Impact of treatment on the outcome of acute myeloid leukemia with

45

46

Mihaela Onciu and Ching-Hon Pui

168

169

170

171 172

173

174

175

176

177

178

179

180

181

182

inversion 16: a single institution’s experience. Leukemia, 2001; 15: 1326–30. Tosi, P., Visani, G., Ottaviani, E., et al. Inv(16) acute myeloid leukemia cells show an increased sensitivity to cytosine arabinoside in vitro. Eur J Haematol, 1998; 60: 161–5. Harbott, J., Mancini, M., Verellen-Dumoulin, C., Moorman, A. V., & Secker-Walker, L. M. Hematological malignancies with a deletion of 11q23: cytogenetic and clinical aspects. EU Concerted Action 11q23 Workshop participants. Leukemia, 1998; 12: 823–7. Rubnitz, J. E., Raimondi, S. C., Tong, X., et al. Favorable impact of the t(9;11) in childhood acute myeloid leukemia. J Clin Oncol, 2002; 20: 2302–9. Pui, C. H. & Relling, M. V. Topoisomerase II inhibitor-related acute myeloid leukaemia. Br J Haematol, 2000; 109: 13–23. Dann, E. J. & Rowe, J. M. Biology and therapy of secondary leukaemias. Best Pract Res Clin Haematol, 2001; 14: 119– 37. Foucar, K., McKenna, R. W., Bloomfield, C. D., Bowers, T. K., & Brunning, R. D. Therapy-related leukemia: a panmyelosis. Cancer, 1979; 43: 1285–96. Leone, G., Voso, M. T., Sica, S., Morosetti, R., & Pagano, L. Therapy related leukemias: susceptibility, prevention and treatment. Leuk Lymphoma, 2001; 41: 255–76. Michels, S. D., McKenna, R. W., Arthur, D. C., & Brunning, R. D. Therapy-related acute myeloid leukemia and myelodysplastic syndrome: a clinical and morphologic study of 65 cases. Blood, 1985; 65: 1364–72. Armitage, J. O., Carbone, P. P., Connors, J. M., et al. Treatmentrelated myelodysplasia and acute leukemia in non-Hodgkin’s lymphoma patients. J Clin Oncol, 2003; 21: 897–906. Le Deley, M. C., Leblanc, T., Shamsaldin, A., et al. Risk of secondary leukemia after a solid tumor in childhood according to the dose of epipodophyllotoxins and anthracyclines: a case-control study by the Societe Francaise d’Oncologie Pediatrique. J Clin Oncol, 2003; 21: 1074–81. Pui, C. H., Ribeiro, R. C., Hancock, M. L., et al. Acute myeloid leukemia in children treated with epipodophyllotoxins for acute lymphoblastic leukemia. N Engl J Med, 1991; 325: 1682–7. Block, A. W., Carroll, A. J., Hagemeijer, A., et al. Rare recurring balanced chromosome abnormalities in therapy-related myelodysplastic syndromes and acute leukemia: report from an international workshop. Genes Chromosomes Cancer, 2002; 33: 401–12. Beaumont, M., Sanz, M., Carli, P. M., et al. Therapy-related acute promyelocytic leukemia. J Clin Oncol, 2003; 21: 2123– 37. Detourmignies, L., Castaigne, S., Stoppa, A. M., et al. Therapyrelated acute promyelocytic leukemia: a report on 16 cases. J Clin Oncol, 1992; 10: 1430–5. Quesnel, B., Kantarjian, H., Bjergaard, J. P., et al. Therapyrelated acute myeloid leukemia with t(8;21), inv(16), and t(8;16): a report on 25 cases and review of the literature. J Clin Oncol, 1993; 11: 2370–9.

183 Andersen, M. K., Larson, R. A., Mauritzson, N., et al. Balanced chromosome abnormalities inv(16) and t(15;17) in therapyrelated myelodysplastic syndromes and acute leukemia: report from an international workshop. Genes Chromosomes Cancer, 2002; 33: 395–400. 184 Matutes, E., Morilla, R., Farahat, N., et al. Definition of acute biphenotypic leukemia. Haematologica, 1997; 82: 64–6. 185 Reinhardt, D., Zimmermann, M., Langebrake, C., et al. Acute mixed lineage leukemia in childhood. Blood, 2002; 100: 69a. 186 Carbonell, F., Swansbury, J., Min, T., et al. Cytogenetic findings in acute biphenotypic leukaemia. Leukemia, 1996; 10: 1283–7. 187 Killick, S., Matutes, E., Powles, R. L., et al. Outcome of biphenotypic acute leukemia. Haematologica, 1999; 84: 699–706. 188 Pui, C. H., Raimondi, S. C., Head, D. R., et al. Characterization of childhood acute leukemia with multiple myeloid and lymphoid markers at diagnosis and at relapse. Blood, 1991; 78: 1327–37. 189 Pane, F., Frigeri, F., Camera, A., et al. Complete phenotypic and genotypic lineage switch in a Philadelphia chromosomepositive acute lymphoblastic leukemia. Leukemia, 1996; 10: 741–5. 190 Bierings, M., Szczepanski, T., van Wering, E. R., et al. Two consecutive immunophenotypic switches in a child with immunogenotypically stable acute leukaemia. Br J Haematol, 2001; 113: 757–62. 191 Tsimberidou, A. M., Kantarjian, H. M., Estey, E., et al. Outcome in patients with nonleukemic granulocytic sarcoma treated with chemotherapy with or without radiotherapy. Leukemia, 2003; 17: 1100–3. 192 Tallman, M. S., Hakimian, D., Shaw, J. M., et al. Granulocytic sarcoma is associated with the 8;21 translocation in acute myeloid leukemia. J Clin Oncol, 1993; 11: 690–7. 193 Jenkin, R. D., Al Shabanah, M., Al Nasser, A., et al. Extramedullary myeloid tumors in children: the limited value of local treatment. J Pediatr Hematol Oncol, 2000; 22: 34–40. 194 Johansson, B., Gray, A., Kullendorff, C. M., et al. Granulocytic sarcoma in body cavities in childhood acute myeloid leukemias with 11q23/MLL rearrangements. Genes Chromosomes Cancer, 2000; 27: 136–42. 195 Peterson, L., Dehner, L. P., & Brunning, R. D. Extramedullary masses as presenting features of acute monoblastic leukemia. Am J Clin Pathol, 1981; 75: 140–8. 196 Bown, N. P., Rowe, D., & Reid, M. M. Granulocytic sarcoma with translocation (9;11)(p22;q23): two cases. Cancer Genet Cytogenet, 1997; 96: 115–7. 197 Menasce, L. P., Banerjee, S. S., Beckett, E., & Harris, M. Extramedullary myeloid tumour (granulocytic sarcoma) is often misdiagnosed: a study of 26 cases. Histopathology, 1999; 34: 391–8. 198 Oliva, E., Ferry, J. A., Young, R. H., et al. Granulocytic sarcoma of the female genital tract: a clinicopathologic study of 11 cases. Am J Surg Pathol, 1997; 21: 1156–65. 199 Ritter, J. H., Goldstein, N. S., Argenyi, Z., & Wick, M. R. Granulocytic sarcoma: an immunohistologic comparison with

Diagnosis and classification

200

201

202 203

204

205

206

207 208

209

210

211

212

213

214

peripheral T-cell lymphoma in paraffin sections. J Cutan Pathol, 1994; 21: 207–16. Bennett, J. M., Catovsky, D., Daniel, M. T., et al. Proposals for the classification of the myelodysplastic syndromes. Br J Haematol, 1982; 51: 189–99. Estey, E. H., Keating, M. J., Smith, T. L., et al. Prediction of complete remission in patients with refractory acute leukemia treated with AMSA. J Clin Oncol, 1984; 2: 102–6. Seymour, J. F., & Estey, E. H. The prognostic significance of auer rods in myelodysplasia. Br J Haematol, 1993; 85: 67–76. Forty-four cases of childhood myelodysplasia with cytogenetics, documented by the Groupe Francais de Cytogenetique Hematologique. Leukemia, 1997; 11: 1478–85. Hasle, H., Jacobsen, B. B., & Pedersen, N. T. Myelodysplastic syndromes in childhood: a population based study of nine cases. Br J Haematol, 1992; 81: 495–8. Hasle, H., Wadsworth, L. D., Massing, B. G., McBride, M., & Schultz, K. R. A population-based study of childhood myelodysplastic syndrome in British Columbia, Canada. Br J Haematol, 1999; 106: 1027–32. Luna-Fineman, S., Shannon, K. M., Atwater, S. K., et al. Myelodysplastic and myeloproliferative disorders of childhood: a study of 167 patients. Blood, 1999; 93: 459–66. Mielot, F. Childhood myelodysplastic syndromes. Pediatr Hematol Oncol, 1999; 16: 283–4. Mandel, K., Dror, Y., Poon, A., & Freedman, M. H. A practical, comprehensive classification for pediatric myelodysplastic syndromes: the CCC system. J Pediatr Hematol Oncol, 2002; 24: 596–605. Sasaki, H., Manabe, A., Kojima, S., et al. Myelodysplastic syndrome in childhood: a retrospective study of 189 patients in Japan. Leukemia, 2001; 15: 1713–20. Passmore, S. J., Hann, I. M., Stiller, C. A., et al. Pediatric myelodysplasia: a study of 68 children and a new prognostic scoring system. Blood, 1995; 85: 1742–50. Passmore, S. J., Chessells, J., Kempski, H., et al. Pediatric myelodysplastic syndromes and juvenile myelomonocytic leukaemia in the UK: a population-based study of incidence and survival. Br J Haematol, 2003; 121: 758–67. Arceci, R. J., Longley, B. J., & Emanuel, P. D. Atypical cellular disorders. In V. C. Broudy, J. L. Abkowitz, & J. M. Vose, eds. Hematology, American Society of Hematology Education Program Book, 2002, pp. 297–314. http://www.asheducationbook.org/ cgi/content/full/2002/11297. Emanuel, P. D., Shannon, K. M., & Castleberry, R. P. Juvenile myelomonocytic leukemia: molecular understanding and prospects for therapy. Mol Med Today, 1996; 2: 468–75. Niemeyer, C. M., Arico, M., Basso, G., et al. Chronic myelomonocytic leukemia in childhood: a retrospective analysis of 110 cases. European Working Group on Myelodysplastic Syndromes in Childhood (EWOG-MDS). Blood, 1997; 89: 3534–43.

215 Tartaglia, M., Niemeyer, C. M., Song, X., et al. Somatic PTPN11 mutations in juvenile myelomonocytic leukemia, myelodysplastic syndromes and acute myeloid leukemia. Blood, 2002; 100: 141a. 216 Arico, M., Biondi, A., & Pui, C. H. Juvenile myelomonocytic leukemia. Blood, 1997; 90: 479–88. 217 Castro-Malaspina, H., Schaison, G., Briere, J., et al. Philadelphia chromosome-positive chronic myelocytic leukemia in children. Survival and prognostic factors. Cancer, 1983; 52: 721–7. 218 Golub, T. R., Slonim, D. K., Tamayo, P., et al. Molecular classification of cancer: class discovery and class prediction by gene expression monitoring. Science, 1999; 286: 531–7. 219 Armstrong, S. A., Staunton, J. E., Silverman, L. B., et al. MLL translocations specify a distinct gene expression profile that distinguishes a unique leukemia. Nat Genet, 2002; 30: 41–7. 220 Yeoh, E. J., Ross, M. E., Shurtleff, S. A., et al. Classification, subtype discovery, and prediction of outcome in pediatric acute lymphoblastic leukemia by gene expression profiling. Cancer Cell, 2002; 1: 133–43. 221 Ross, M. E., Zhou, X., Song, G., et al. Classification of pediatric acute lymphoblastic leukemia by gene expression profiling. Blood, 2003; 102: 2951–9. 222 Cheok, M. H., Yang, W., Pui, C. H., et al. Treatment-specific changes in gene expression discriminate in vivo drug response in human leukemia cells. Nat Genet, 2003; 34: 85–90. 223 Pui, C.-H., Relling, M. V., & Downing, J. R. Acute lymphoblastic leukemia. N Engl J Med, 2004; 350: 1535–48. 224 Cario, G., Stanulla, M., Fine, B. M., et al. Distinct gene expression profiles determine molecular treatment response in childhood acute lymphoblastic leukemia. Blood, 2005; 105: 821–6. 225 Zaza, G., Cheok, M., Yang, M., et al. Gene expression and thioguanine nucleotide disposition in acute lymphoblastic leukemia after in vivo mercaptopurine treatment. Blood, 2005; May 19 [Epub ahead of print] PMID: 15905191. 226 Bullinger, L., Dohner, K., Bair, E., et al. Use of gene-expression profiling to identify prognostic subclasses in adult acute myeloid leukemia. N Engl J Med, 2004; 350, 1605–16. 227 Ross, M. E., Mahfouz, R., Onciu, M., et al. Gene expression profiling of pediatric acute myelogenous leukemia. Blood 2004; 104: 3679–87. 228 Valk, P. J., Verhaak, R. G., Beijen, M. A., et al. Prognostically useful gene-expression profiles in acute myeloid leukemia. N Engl J Med, 2004; 350: 1617–28. 229 Lacayo, N. J., Meshinchi, S., Kinnunen, P., et al. Gene expression profiles at diagnosis in de novo childhood AML patients identify FLT3 mutations with good clinical outcomes. Blood 2004; 104: 2646–54. 230 Lu, J., Getz, G., Miska, E. A., et al. MicroRNA expression profiles classify human cancers. Nature, 2005; 439: 834–8.

47

3 Epidemiology and etiology Logan G. Spector, Julie A. Ross, Leslie L. Robison, and Smita Bhatia

Introduction

Incidence and trends

The acute leukemias of childhood are a heterogeneous group of diseases. In reviewing the descriptive and analytic epidemiology of these malignancies, we have emphasized specific subgroups, as defined by morphology [the French-American-British (FAB) classification], cytogenetic features, or molecular markers. There is evidence that specific subtypes of leukemia may have distinct etiologies, and that molecular abnormalities associated with particular subtypes may be linked with specific causal mechanisms. Moreover, the mutations produced at the successive stages of leukemogenesis, from initiation through induction to promotion, may all involve separate etiologic processes. It is also important to note that changes over time in diagnostic practice and precision may account in part for some reported epidemiologic trends. Moreover, changes in terminology and classification schemes for leukemia make it difficult to perform direct comparisons among studies, especially if risk factors differ for different subgroups. However, in assessing risk factors, studies of the childhood leukemias present several methodologic advantages. The interval between exposure to putative risk factors and the onset of leukemia may be shorter, recall of exposures is likely to be better, and intervening factors may be fewer than those associated with adult leukemias. These characteristics of childhood leukemia may facilitate identification of the most likely risk factors for each leukemia subtype. Furthermore, they lend themselves to an approach that includes both population studies and molecular epidemiologic techniques, permitting the design of research to assess genetic-environmental causal interactions.

In the United States, the acute leukemias represent 31% of malignancies occurring among children under the age of 15 years.1 Acute lymphoblastic leukemia (ALL) comprises 85% of childhood acute leukemias. Annual incidence rates (per million population) for ALL and acute myeloid leukemia (AML) are 30.9 and 5.6, respectively. The two diagnostic categories – ALL and AML – are further subdivided based on leukemic cell features. Childhood ALL is classified by FAB morphology (L1, L2, and L3) and by immunophenotype (B cell, early pre-B, pre-B, and T cell). Childhood AML is classified morphologically into eight distinct morphologic subgroups: M0 (myeloid leukemia with minimal differentiation), M1 (acute myeloblastic without maturation), M2 (acute myeloblastic with maturation), M3 (acute promyelocytic), M4 (acute myelomonocytic), M5 (acute monocytic), M6 (erythroleukemia), and M7 (acute megakaryocytic). Age-specific incidence patterns demonstrate a characteristic peak between the ages of 2 and 5 years for childhood ALL (Fig. 3.1). The incidence rates for AML are highest in infancy and are fairly uniform in older children (Fig. 3.1).2 In the United States, during 1986 and 1995, the incidence in the 0 to 4 year age group was 10.3 per million, and 5.0 per million and 6.2 per million in the 5 to 9 and 10 to 14 year age groups, respectively. Comparable rates have been reported elsewhere in Europe and in Britain. In childhood ALL, males are more often affected than females, with the notable exception of a female predominance in infancy. In contrast, there is no clear pattern in the male-to-female ratio for childhood AML. A striking difference in the incidence of ALL exists between white and black children. The excess incidence

C Cambridge University Press 2006. Childhood Leukemias, ed. Ching-Hon Pui. Published by Cambridge University Press.


48

Epidemiology and etiology

Table 3.1 Annual standardized rates (per million population) of acute lymphoblastic leukemia, from birth to 14 years of age, in selected regionsa

Population Costa Rica Finland Canada Hong Kong Sweden Australia, New South Wales Germany Norway US, SEER (whites) Italy Hungary United Kingdom, England and Wales Czechoslovakia, Slovakia Cuba Kuwait (Kuwaiti) Japan Brazil, Goiania New Zealand (Maori) US, SEER (blacks) Brazil, Belem Israel (Jews) China, Tianjin India, Bombay Cancer Registry

Both sexes

Males

Females

46.3 41.9 41.0 40.6 40.1 39.9 39.0 38.3 38.0 37.9 33.5 32.8

51.7 41.8 44.8 50.6 40.9 56.2 43.6 39.3 41.3 38.9 37.0 35.7

40.7 42.0 36.9 29.9 39.3 43.2 34.1 37.3 34.5 36.8 29.9 29.7

28.4 25.4 24.3 22.6 21.9 21.9 20.8 18.8 18.6 17.4 16.0

31.8 27.6 27.1 25.6 21.8 30.3 22.0 20.5 18.7 19.0 19.8

24.8 23.1 21.4 19.5 22.1 13.1 19.6 17.4 18.4 15.7 12.0

Abbreviations: SEER, Surveillance, Epidemiology, and End Results Program. a Data are from Ferlay et al.5

Fig. 3.1 Age-specific incidence and gender and race ratios for childhood acute leukemia in the United States. (Data derived from Gurney et al.1 )

among white children is apparent in most age groups. The white-to-black ratio for AML in the United States is 1.2. The higher rate of AML among Hispanics1,3 is contributed by acute promyelocytic leukemia (APL), raising the question of genetic predisposition to APL and/or exposure to distinct environmental factors.4 Substantial geographic variation exists in childhood leukemia incidence rates.5 Internationally, annual incidence rates of childhood ALL range from 9 to 47 per million for males and from 7 to 43 per million for females

(Table 3.1). Incidence rates for ALL are highest in the United States (among white children), Australia, Costa Rica, and Germany. Rates are intermediate in most European countries and lowest in India and among black children in the United States. By contrast, the incidence of AML is highest in China, Japan, and among the Maori of New Zealand, with intermediate rates in Australia, the United States, and the United Kingdom. India, Kuwait and the Canadian Atlantic Provinces have the lowest reported rates of childhood AML (Table 3.2). Several investigators have examined temporal trends for leukemia, but the findings are difficult to interpret. This is primarily because of the various time periods covered and the different methods of analysis used. Some investigators analyzed the 0- to 14-year age group as a whole, while others have focused on specific age groups for each sex and by

49

50

Logan G. Spector et al.

Table 3.2 Annual standardized rates (per million population) of acute myeloid leukemia, from birth to 14 years of age, in selected regions worldwidea Population

Both sexes

Males

Females

New Zealand (Maori) Costa Rica Australia, New South Wales Norway Italy Japan China, Tianjin Germany Sweden Canada United Kingdom, England and Wales US, SEER (blacks) US, SEER (whites) Cuba Hungary Finland Brazil, Goiania Israel (Jews) Czechoslovakia, Slovakia India, Bombay Cancer Registry Hong Kong Brazil, Belem Kuwait (Kuwaiti)

14.4 8.9 8.0

15.5 8.5 8.3

13.3 9.2 7.7

8.0 7.9 7.2 6.7 6.7 6.7 6.3 6.3

9.4 7.4 7.8 6.8 7.1 6.0 6.9 6.2

6.5 8.3 6.5 6.6 6.3 7.5 5.6 6.5

6.2 6.0 5.7 5.5 5.4 5.3 5.3 5.1 4.8

5.6 5.9 5.1 6.0 4.4 5.1 4.9 4.4 5.2

6.8 6.0 6.4 5.0 6.4 5.5 5.7 5.9 4.4

3.9 3.5 2.0

5.2 4.5 0.6

2.6 2.7 3.4

Abbreviations: SEER, Surveillance, Epidemiology, and End Results Program. a Data are from Ferlay et al.5

leukemia subtype. Some studies on secular trends in the incidence of childhood leukemia have identified increasing rates over time.6 From 1974 through 1991, the average annual percentage increase for all leukemias was 0.9%, with average annual increases of 1.6% and 0.6% for ALL and AML, respectively. This increase may be attributable in part to changes in diagnostic classification during the time period studied. Nevertheless, it is noteworthy that the increase is modest and the largest average annual increases occurred among children diagnosed with ALL and AML during the first 2 years of life (2.4% and 2.5%, respectively). A similar increase in the incidence of childhood acute leukemia was reported from regions of the United Kingdom,7 and the incidence of childhood AML, but not all of ALL, was reported to have increased in Australia.8 Stable childhood leukemia rates have been reported from other

registries, including the Greater Delaware Valley Pediatric Tumor Registry,9 the nationwide German Registry,10 and the SEER Registry, where Linet et al.11 reported no substantial change in the incidence of childhood leukemia diagnosed in the United States, between 1975 and 1995.

Genetic factors Acute leukemia is a clonal disorder of the hematopoietic system, arising from mutations in a single cell that are passed on to all of its descendants. In most cases, the genetic abnormalities that give rise to acute leukemia are acquired rather than inherited. As many as 5% of acute leukemias, however, are associated with inherited genetic syndromes.12 In addition, a variety of normal inherited polymorphisms in genes may contribute indirectly to the risk of leukemia: these include genes that encode enzymes involved in carcinogen metabolism and detoxification and those that are involved in the immune response to infections.13

Cytogenetic abnormalities Many acquired chromosomal abnormalities have been found in childhood leukemia,14 as described in more detail in Chapter 9. Abnormalities are generally more prevalent in, but not restricted to, specific morphologic subtypes. Some of the more common abnormalities include translocations involving the MLL gene at chromosome band 11q23 in infants (both ALL and AML), the t(8;21) in M2 AML, the t(15;17) in M3 AML, trisomy 8 in AML, and the t(9;22) and t(1;19) in ALL.15–21 Recent studies indicate that the most common reciprocal translocation in ALL is the t(12;21)(p13;q22), which occurs in about 25% of cases and fuses the TEL and AML1 genes.22 This translocation is invisible karyotypically and therefore went undetected in early studies. A significant proportion of ALL cases (about 25%) exhibit leukemic cell hyperdiploidy. With the exception of infant leukemias,14,17 almost 80% of which have an abnormality involving the MLL gene at 11q23, there have been no epidemiologic studies exploring associations between environmental exposures and specific chromosomal abnormalities, partly because of the heterogeneous clinical presentation of these abnormalities and their relatively low frequency.

Genetic syndromes Several genetic syndromes have been associated with an increased risk of childhood leukemia.14 Studies suggest a 10- to 20-fold increased risk of leukemia (both ALL and

Epidemiology and etiology

AML) in children with Down syndrome,23,24 and some reports suggest up to a 600-fold increased risk for one subtype of AML (M7).25 The reasons for this increased risk are unclear, although a gene (AML1) associated with certain cases of AML has been identified in a chromosomal site (band 21q22) believed responsible for the Down syndrome phenotype.26 Other genetic syndromes associated with both childhood ALL and AML include Bloom syndrome, neurofibromatosis type 1, Schwachman syndrome, and ataxia telangiectasia.27–32 Kostmann granulocytic leukemia and Fanconi anemia are associated with AML.33,34 In addition, there is a familial form of AML (familial monosomy 7) in which two or more siblings develop leukemia before the age of 20.35 Although specific associations have been described mainly as case reports, data on the proportion of cases of leukemia with a known genetic etiology or an association with specific genetic syndromes are limited. Several studies report that ∼2.5% of the children with leukemia have a recognized genetic condition, which is almost entirely accounted for by Down syndrome.36–38 To evaluate the risk of leukemia associated with congenital anomalies, a series of matched case-control studies have been conducted. In one such study by the Children’s Cancer Group, children with ALL and AML were compared with matched regional population controls. More congenital anomalies were found in the index child with ALL or AML than in control subjects. The congenital anomalies included Down syndrome, congenital heart defect, and multiple birth marks.39 Another study from the United Kingdom reported a lower frequency of congenital anomalies among children with leukemia or lymphoma, when compared with children with solid tumors. They hypothesized that mutations resulting in the development of leukemias and lymphomas occur at a much later stage in development, in the cells committed to hematopoiesis.40

Familial patterns There have been several reports of familial aggregation of childhood leukemia.29,41–44 Although this finding may represent an inherited predisposition, the available studies do not rule out shared environmental factors. One of the most informative associations is the higher degree of concordance of leukemia among twins (particularly monozygotes), which is highly age-dependent, occurring mostly in infants.45–50 However, in a study of leukemia in a pooled series from the United States, Canada and the United Kingdom, only three concordant pairs (1.5%) were found among 197 pairs in which at least one twin had leukemia, with the concordance rate for monozygotic twins reported as

3.9%.51 Thus, although the concordance rate in twins of the same sex (likely monozygotes) is higher than the zero concordance rate reported for twins of unlike sex (dizygotes), the concordance rate in twins of like sex is quite variable.51 There are molecular data to suggest that in twin pairs with leukemia, the leukemic clone develops in one fetus and disseminates to the other via a shared placental circulation rather than resulting from an inherited mutation.52,53 Recent molecular studies also indicate that concordant ALLs in older twin children (3 to 11 years) have an in utero origin, coupled in these cases with a protracted latency.54,55

Cancer in offspring of patients treated for childhood leukemia There are relatively few reports of the incidence of cancer in the offspring of individuals treated for childhood cancer. Hawkins et al.56 estimated that the proportion of heritable cases among the cancer survivors is unlikely to exceed 5%. Furthermore, recent reports indicate that offspring of subjects previously receiving chemotherapy and/or radiotherapy for childhood malignancies do not exhibit latent chromosomal instability.57

Other conditions in relatives of cases with leukemia A recent case-control study provides support for a positive association between childhood acute leukemia and a family history of hematologic neoplasms and solid tumors (particularly gastrointestinal tumors and melanoma), with a stronger association for patients with AML.58 Further studies need to be conducted to corroborate this report and to identify the biologic basis of this association.

Demographic and environmental risk factors Although the specific causes of most pediatric and adult leukemias are not known, several environmental and demographic features (summarized together with predisposing genetic syndromes in Table 3.3) have been associated with increased risk. These include ionizing radiation and exposure to certain chemicals, particularly organic solvents such as benzene.59 Probably the most extensively studied risk factor is in utero exposure to ionizing radiation from diagnostic x-rays, a well-established risk factor for both childhood ALL and AML (although accounting for only a small proportion of cases).60–63 Other factors associated with increased risk include parental occupational exposures to hydrocarbons and pesticides, maternal alcohol use and cigarette smoking during pregnancy, and a

51

52

Logan G. Spector et al.

Table 3.3 Risk factors for childhood acute leukemia, by degree of certainty Degree of certainty

Acute lymphoblastic leukemia

Acute myeloid leukemia

Generally accepted risk modifiers

Males Age (2–5 years) High socioeconomic status Race (whites > blacks) In utero x-ray exposure Postnatal radiation (therapeutic) Down syndrome Neurofibromatosis type I Bloom syndrome Shwachman syndrome Ataxia telangiectasia Increased birth weight Breast feeding Maternal history of fetal loss

Race (Hispanic) Chemotherapeutic agents (alkylating agents, topoisomerase II inhibitors) Down syndrome Fanconi anemia Neurofibromatosis type 1 Bloom syndrome Shwachman syndrome Familial monosomy 7 Kostmann granulocytopenia

Suggestive of increased risk

Limited evidence

Probably not associated

Parental smoking prior to or during pregnancy Parental occupational exposures Postnatal infections Diet Maternal alcohol consumption during pregnancy Electric and magnetic fields Postnatal use of chloramphenicol Vitamin K prophylaxis in newborns Ultrasound Indoor radon

maternal history of prior fetal loss. Both exposure to electromagnetic fields and paternal preconception exposure to radiation have also been proposed as possible risk factors. These associations remain controversial because of inconsistencies in the results of epidemiologic investigations,64–66 and the lack of biologically plausible etiologic theories.

Ionizing radiation The leukemogenic potential of ionizing radiation has been well documented in studies of survivors of the atomic bombing of Japan in 194567 and of occupational exposure incurred by early radiation scientists.68 In children, exposure in utero to diagnostic x-rays is associated with an increased risk of both ALL and AML, with relative risks of 1.5 to 1.7 for ALL.32 It is reasonable to expect that in utero xray exposure levels would have declined substantially over time, given the knowledge of health-related risks associated with radiation exposure and the reduced level of radiation required for diagnostic procedures. Meeting these expecta-

Maternal alcohol consumption during pregnancy Prenatal and child exposure to pesticides Parental solvent exposure Maternal marijuana use during pregnancy Indoor radon Postnatal use of chloramphenicol

tions are reports from studies conducted in Sweden, United Kingdom and the United States that have demonstrated higher risk estimates among children born in earlier eras (e.g. 1930s to 1950s) compared to more recent times.69–71 The causal nature of this association has been called into question,72,73 however, because studies of Japanese survivors of the atomic bomb revealed no increased incidence of leukemia associated with in utero exposure. Moreover, there are concerns regarding the lack of good exposure assessment and the potential for recall bias in case-control studies related to this subject. Nevertheless, a strong argument in favor of causality is provided by the dose–response pattern that is evident for in utero exposure to diagnostic x-rays (i.e. increased risk with increasing numbers of exposures).74 Studies of children who have received radiation therapy for the treatment of Hodgkin disease, Langerhans cell histiocytosis, thymic enlargement, and tinea capitis show a slightly elevated risk of leukemia, particularly AML.75–78 Postnatal diagnostic radiation, in contrast, does not appear to increase risk, although a recent study suggested risk

Epidemiology and etiology

may be increased in the presence of polymorphisms that decrease the effectiveness of DNA repair genes.79–81 Even if leukemia risk is increased, the potential number of cases that currently might be attributed to in utero and/or postnatal x-ray exposure would likely be very small when one considers the modest magnitude of risk and the limited level of exposure. Exposures to ionizing radiation from fallout from atomic bombs or accidentally released by nuclear power plants, from background radiation and radon, and from parental employment in the nuclear power industry have also been proposed to increase the risk of leukemia. Evidence in support of an association between childhood acute leukemia and radiation from nuclear fallout is quite weak. Because of the widespread fallout from Chernobyl, a series of investigations were undertaken in affected countries, including Sweden,82 Finland,83 United Kingdom,84 Scotland,85 Germany,86 and Greece.87 Except for a recent report of a transient increase in infant acute leukemia in northern Greece,88 no increase in the incidence of leukemia has been identified in the areas contaminated by the Chernobyl reactor accident.89 Similarly, reports relating to nuclear fallout follow-up accidents at Three Mile Island have not provided strong evidence of an increased risk of childhood leukemia.90 However, a reactor accident in 1957 in Chelyabinsk in the former Soviet Union resulted in exposure to radiation levels that may have reached 4 Gy. A recent analysis found that this exposure was associated with a subsequent regional increase in the incidence of leukemia.91 Nuclear fallout affecting regions of Nevada and Utah, as a result of nuclear weapon testing in Nevada, led to studies to determine the potential impact on childhood leukemia, utilizing estimates of geographic site-specific radiation doses to the bone marrow. No significant trend between the estimated dose and risk of leukemia mortality was found.92 However, the risk of death from leukemia was found to be higher in persons receiving the highest exposure level (6–30 mGy) compared to the lowest-dose group (0–2.9 mGy). All of these associations are controversial because of the difficulty of extrapolating from the high acute doses of radiation experienced by the Japanese survivors of the atomic bomb to these much smaller or chronic exposures. Also questionable is the evidence for an association with nuclear energy production and nuclear fuel reprocessing. Studies addressing childhood leukemia have been conducted in the United Kingdom,93,94 France,95 United States,96,97 Germany,98 and Canada.99 Overall, these reports do not lend support for an increased incidence of childhood leukemia, although some do report an increased incidence or mortality. In fact, the childhood acute leukemia clusters

around the nuclear reprocessing plants of Sellafield and Dounreay in the United Kingdom are now thought to result from unique sociodemographic features of the local communities, with population mixing and infection probably playing a role, rather than to any direct effect of proximity to the plants themselves.100 A possible role of paternal preconception radiation exposure has been the subject of recent interest. At least two studies have reported an increased risk of childhood leukemia in offspring of fathers exposed to radiation either occupationally101 or diagnostically.102 However, other studies have found no such associations.103,104 Again, the lack of a plausible biological mechanism and the absence of an increased risk of leukemia in the offspring of survivors of the atomic bomb argue against this proposed association. Natural background radiation from terrestrial sources and cosmic radiation is thought to account for about 5% of all leukemia cases in adults and children.73 Although there is no clear association between the risk of leukemia and cumulative radon exposure in uranium miners,73 it has been proposed that some naturally occurring circumstances might result in the accumulation of a leukemogenic dose of radon in fat-containing bone marrow.105 The leukemogenic effects of radon have recently come under investigation. Although ecologic investigations have provided some evidence for a correlation,105,106 case-control studies incorporating measurement of indoor radon levels have not found higher levels within homes of children with ALL107 and AML.108–111

Nonionizing radiation Public concern has been raised regarding the leukemogenic potential of low-energy electromagnetic fields produced by residential power supply and appliances. The concern is prompted in part by the ubiquitous nature of the exposure – a feature that poses methodologic problems because of the difficulty of finding an unexposed comparison population. In addition, the biologic plausibility of the association is questionable.112 Laboratory studies indicate that electromagnetic fields may produce adverse biological effects; however, they do not release sufficient energy to damage DNA. Results from studies to date are inconsistent, and the evaluation of exposure has often relied on surrogate measures, such as the physical configuration of power lines and their distance from homes. Three of the more recently reported investigations from the United States,66 Canada,113 and the United Kingdom,114 provide rather convincing evidence that electric and magnetic field exposure is not associated with a significantly increased risk of childhood ALL. A recent study failed to show any association

53

54

Logan G. Spector et al.

between exposure to magnetic fields inside infant incubators and the risk of childhood leukemia.115 Nonetheless, results of meta-analysis have been interpreted to suggest that risk may be increased at the highest exposure levels (i.e. > 0.4 muT).116,117 However, it is important to note that even if there is an increased risk at this highest level of exposure, the proportion of the population exposed to such high levels is extremely small and thus the attributable risk would be negligible.

Chemical exposures Pesticide exposure (occupational or home use) has been reported to be associated with childhood leukemias in several studies.118–121 In a multicenter case-control study of 204 cases of AML and matched community controls, Buckley et al.120 reported a positive association with maternal occupational exposure to pesticides before, during and after pregnancy and with paternal occupational exposure during the same time period.120 This association was particularly strong for younger children with the M4 (monocytic) or M5 (myelomonocytic) subtype of AML. Furthermore, there was an independent association with maternal exposure to household fly-sprays, pesticides, garden and agricultural sprays, and home treatment by insect exterminators, in the month before the last menstrual cycle and during the index pregnancy, and with direct exposure of the index child to household and garden insecticides. These associations were independent of parental occupational exposure. Although there is a large body of literature on the association between pesticide exposure and childhood leukemia,122,123 the studies are limited by the nonspecific nature of the exposure, by the reliance of these studies on parental self-reporting and by insufficient evidence for a causal relationship. Moreover, use of pesticides may be an indicator of rural isolation, and hence a possible confounding variable because the patterns of exposure to infection may be associated with population mixing. Parental exposure to solvents has also been associated with an increased risk of childhood leukemia in several studies, although other studies report no such association.118,120,124–127 An increased risk of childhood leukemia associated with paternal exposure to chlorinated solvents has been reported by several investigators,118,120,128 but consistent association with maternal exposure to solvents is apparent.129 However, there is some evidence to indicate that maternal occupational exposure to hydrocarbons may be associated with childhood leukemia, which is compatible with the observation that maternal benzene is associated with AML in adults, but these findings need further substantiation.120 No clear association with residential

proximity to industrial sources of hydrocarbons has been observed.130–132 A recent study revealed a significant association between childhood leukemia and substantial participation by household members in some common household activities involving organic solvents.133 Studies of proximity to motor vehicle traffic and, by proxy, of benzenecontaining exhaust have produced mixed results.134,135 A recent study in which the distance measure was confirmed to be correlated with measures of vehicle exhaust found no relationship with leukemia.136 A positive association between total reported duration of paternal occupational exposure to lead and AML has been reported by Buckley et al.,120 as has an association between maternal occupational exposure to metal dusts and fumes and lead.120,124 However, ecological studies have failed to show an association between leukemia and proximity to industrial facilities with an increased exposure to metal or metal fumes.137,138 An association between maternal occupational exposure to wood dust before conception of the index child and childhood leukemia has been reported, although few women were exposed during or after pregnancy.128 Significantly elevated relative risks have been reported for paternal occupational exposure to wood dust before and near conception and during the gestational and postnatal periods.139

Lifestyle Diet and vitamin supplement use In pediatric malignancies, research has focused primarily upon the use of vitamin supplements during pregnancy,140,141 with fewer studies focusing on maternal diet. A recent case-control study in Australia suggested that maternal folate supplementation during pregnancy may be associated with a decreased risk of childhood ALL.142 This study, however, was very small (83 cases) and lacked sufficient detail regarding exposure. Details of studies reporting the association of maternal diet during pregnancy and infant leukemia are described in detail elsewhere in this chapter. The role of diet of the index child in the development of childhood leukemia has not been investigated extensively. Results of investigations exploring the association between the intake of certain food items thought to be precursors or inhibitors of N-nitroso compounds have been controversial. An increased risk has been associated with consumption of processed meats, such as hot dogs, while others have failed to show such an association.118,141,143 No associations have been reported with postnatal use of vitamins

Epidemiology and etiology

by the index child,144 consumption of fish, dried milk, fruit juice or canned foods,60 although a decreased risk with long-term use of cod liver oil has also been reported.119

Alcohol consumption by parents The majority of studies relate solely to alcohol consumption of the mother and only to consumption during the pregnancy leading to the birth of the index child. These studies have reported an increased risk of AML (particularly in very young children) associated with maternal alcohol consumption during pregnancy.145–147 A recent study of infant leukemia147 found a dose–response relationship with AML; risk was most pronounced for the M1 and M2 subtypes. There was only a modest increase in risk of ALL in this study, while another study reported no association between maternal alcohol consumption and ALL.125 This series of reports has implications for discussions of the etiology of infant leukemia later in this chapter. No association between paternal alcohol consumption and childhood leukemia has been found.147,148

independent of associations with occupational and household pesticide exposure, paternal occupational exposure to paints, and pigments, metal dusts and saw dust. There was no significant association with paternal use of marijuana in the year before conception of the index child.

Maternal reproductive history Fetal loss Several studies have reported an increased risk of childhood leukemia (both AML and ALL) in association with a maternal history of fetal loss,125,160–162 while one study reported an inverse relationship.119 Although some data suggest that this association is confined to cases diagnosed at a very young age,161 a recent study was unable to confirm this finding.163 A history of fetal loss may indicate some common environmental exposure, an inherited genetic defect with variable effects on the fetus, or both.

Maternal age and birth order Tobacco smoking by parents Several studies have examined the role of parental smoking in the development of childhood leukemia. However, issues such as reporting bias (greater likelihood of positive associations being published) limit the quality of the reports. The association with maternal cigarette smoking has been inconsistent. Some studies have found an increased risk of leukemia in the children of women who smoked during pregnancy,149–151 whereas others have reported no increased risk of either ALL119,125,152–156 or AML.145,147 Two recent studies have suggested an association between paternal preconception cigarette smoking and the risk of childhood leukemia in offspring.156,157 A recent large case-control study concluded that parental smoking during pregnancy or exposure to cigarette smoke shortly after birth is unlikely to contribute substantially to the risk of childhood leukemia in North America.158 In summary, the literature shows no consistent association between leukemia and parental exposure to tobacco. Sandler et al.159 evaluated the cancer risk from cumulative household exposures to cigarette smoke in a casecontrol study. Cancer risk was greater for individuals with exposure during both childhood and adulthood than for individuals with exposure during one period only.159 An association between AML (M4 or M5) and the reported use of mind-altering drugs (primarily marijuana) by the mother in the year before or during the index pregnancy has been described.140 The association with marijuana exposure was

Accumulation of chromosomal aberrations and mutations during the maturation of germ cells is a mechanism hypothesized for the association between increasing maternal age and cancer in the offspring. Most studies have failed to show an association between leukemia and maternal age.160,163 However, two recent analyses of the Swedish Family-Cancer Database revealed a maternal age effect for childhood leukemia, as did a very large British case-control study.164–166 Importantly, these results were obtained after exclusion of children with Down syndrome, which is associated with both leukemia and advanced maternal age at birth. Although most studies fail to show a positive association between birth order and childhood leukemia, there are reports of a decreasing trend in the incidence of childhood ALL and AML with increasing birth order, adjusted for age, sex, calendar period and maternal age at birth of child.167 A large case-control study, similarly adjusted, found a significant trend in the incidence of ALL, but not AML, with increasing birth order.166

Birth weight and length at birth High birth weight has been found to increase the risk for both ALL and AML before the age of 5 years with a fair consistency in larger studies.160,164,167,168 Birth weight is likely a marker for endogenous risk factors. Accordingly, the level of insulin-like growth factor 1 has been found to

55

56

Logan G. Spector et al.

be highly correlated with birth weight,169 and high levels of thyroid-stimulating hormones, which is inversely correlated with birth weight, has been found to be protective against ALL.170

Unique populations of interest Because the etiology of childhood leukemia is largely unknown, investigations focusing on unique patient or disease subsets may prove particularly fruitful. In this section, we briefly summarize theories that have been proposed to explain the etiology of infant leukemia and the 2- to 5-year age peak in the incidence of childhood ALL.

Infant leukemias Nearly 80% of infant leukemias present with a specific cytogenetic abnormality involving the MLL gene on chromosome band 11q23.17 Molecular analysis of MLL gene rearrangements in acute leukemias in identical twin infants and studies of neonatal blood samples have shown that this abnormality occurs in utero.52,171 The identification of MLL gene abnormalities in treatment-related AMLs that develop following therapy for a histologically distinct primary cancer suggests relationships that are explored in the following sections.

Treatment-related leukemias AML arising after exposure to genotoxic chemotherapy is a heterogeneous collection of diseases characterized by distinct chromosomal abnormalities. One subset of therapyrelated AMLs is associated with exposure to alkylating agents. The chromosomal abnormalities 5q- and monosomy 7 are commonly observed in leukemic cells in this group (as they are in de novo adult AML). The other major subset of therapy-induced leukemia is associated with exposure to the epipodophyllotoxin drugs teniposide (VM-26) and etoposide (VP-16).172 A high proportion of epipodophyllotoxin-associated AMLs are of the M4 (monocytic) or M5 (myelomonocytic) subtype and have abnormalities at chromosome band 11q23 involving rearrangements of the MLL gene.173 The same genetic abnormality is also found in some secondary AMLs associated with exposure to anthracyclines (e.g. daunorubicin and doxorubicin). These two classes of chemotherapeutic agents share a common mechanism of action that involves binding to and inhibition of DNA topoisomerase II. Topoisomerase II is an enzyme that catalyzes breakage and resealing of DNA, a function that prevents tangling of helical,

duplexed DNA strands during replication.174 Binding of an epipodophyllotoxin or other inhibitor to this enzyme introduces the potential for faulty recombination between chromosomal regions undergoing simultaneous breakage. This faulty recombination can result in the 11q23/MLL rearrangement that characterizes some of these therapyinduced AMLs.

The topoisomerase II inhibitor hypothesis Because significant proportions of epipodophyllotoxinassociated AMLs and infant leukemias exhibit the same chromosomal abnormalities, it is reasonable to ask whether these two groups of leukemias share a common causal mechanism. As noted previously, infant leukemia has been causally linked to an abnormality acquired in utero, which likely results from maternal/fetal exposure to carcinogens during pregnancy. In light of the association between the epipodophyllotoxins and MLL gene rearrangements in secondary leukemias, it has been proposed that in utero exposure to agents that inhibit DNA topoisomerase II function (including epipodophyllotoxins found in both medicinal and dietary sources) might also play a causal role in infant leukemia.14 A subset of mothers of infants with leukemia diagnosed at 1 year or less of age (controls, selected by random digit dialing, were from three multicenter case-control studies of childhood leukemia in the United States) was reapproached and supplemental information on maternal diet during index pregnancy was sought. This study attempted to test the hypothesis that infant leukemia characterized by 11q23 abnormalities may result from exposure to naturally occurring topoisomerase inhibitors, such as caffeine, and a variety of fruits and vegetables.14 Although no association emerged between diet and leukemia in general or ALL, there was a statistically significant association for AML with increasing consumption of dietary topoisomerase II inhibitors.175,176 A recent study has shown that bioflavonoids, natural substances in food as well as in dietary supplements, can cause site-specific DNA cleavage in the MLL breakpoint cluster region (BCR) in vivo.177 These results suggest that maternal ingestion of bioflavonoids may induce MLL breaks and potential translocations in utero leading to infant leukemia. In utero exposures and their association with infant leukemia were also assessed in a recent study. Use of cigarettes and alcohol, the ingestion of certain herbal medicines and drugs classified as “DNA damaging” and exposure to pesticides were associated with an increased risk of MLL fusion-positive leukemias.178 Despite their common chromosomal rearrangement, it is likely that infant leukemias with 11q23 rearrangements

Epidemiology and etiology

are a heterogeneous group of diseases rather than a single entity. In the study referred to previously,175 the majority of significant associations with dietary topoisomerase II inhibitors were with AML rather than ALL. AML and ALL exhibit distinct chromosomal translocations involving the MLL gene and could have different etiologies, as breaks in the MLL gene occur at different distances from the topoisomerase II binding site, depending on the subtype of acute leukemia. Moreover, inherited susceptibility may play a role in leukemogenesis that is mediated by the actions of topoisomerase II inhibitors. Thus, it is clear that the topoisomerase II inhibitor hypothesis requires further investigation.

The childhood ALL age peak A peak in the incidence of CD10+ B-cell precursor ALL (also known as common ALL, or cALL) occurs in developed countries among children between the ages of 2 and 5 years. This pattern was first noted in the United Kingdom and among white children in the United States, appearing later among U.S. black children.179 Epidemiologic evidence supports the view that many childhood leukemias, especially those that occur during this age peak are the consequence of a rare, abnormal response (brought on by unusual timing, perhaps in combination with individual genetic susceptibility) to a common infection.180

Population mixing The evidence, indirect but compelling, for an etiologic role for infection is provided by studies of leukemia in the context of population mixing. To test the idea that population mixing might provide the conditions under which infection could play a role in childhood leukemia, Kinlen181 performed a series of observational studies in the United Kingdom involving relatively isolated populations affected by significant population mixing or movement. In each case, an increased relative risk of childhood acute leukemia was noted subsequent to the population mixing. Moreover, the leukemia risk decreased after the initial epidemiclike increase, suggesting that immunization of the population had occurred. This effect can be interpreted in terms of the epidemiology and population dynamics of common infections.182 Susceptible individuals who live in areas where a specific infection is not endemic are placed at risk when brought into contact with infected carriers. In the present context, leukemia would be the rare end result of infection for some individuals. Whereas Kinlen’s original studies examined extreme instances of population mixing, such as military

encampments, more recent studies have applied various quantitative measures to wide geographic areas. Several such studies have found an increased risk of childhood leukemia (and, in some instances, trends) with greater population growth and with greater diversity of immigrants.183–188 However, the literature is not entirely in agreement, since associations of leukemia with population mixing have been found for rural but not urban areas and vice versa. Also, one study found an inverse association of ALL with the diversity of the migrants’ origins.189 Interpretation is hampered by the aggregation of diverse types of leukemia; to date only one study has analyzed separately the cALL subtype.186 Other lines of evidence also support an infection-related etiology for childhood leukemia. The 2- to 5-year age peak of cALL is found among affluent populations in developed societies. Many lines of evidence (e.g. international incidence rate comparisons, time trends, geographic distributions, and community characteristics) combine to suggest that cALL is a disease of affluent societies involving a rare response to common infection(s). Delayed first exposure is thought to contribute to the pathogenesis of several diseases associated with affluence in which infection is implicated, the pathologic precedent being paralytic poliomyelitis.190 The timing and pattern of exposure to common infections early in life is considered critical to the production of an appropriate immune response. Factors associated with affluence, including relative social isolation and higher levels of hygiene, may lead to a delayed exposure to common infections. Moreover, the practice of prolonged breast feeding (which, in addition to its nutritional and immunologic benefits, may also provide early exposure to common viruses and bacteria191 ) has declined among affluent populations. Breast feeding has been consistently shown to protect against leukemia, the protection increasing with the length of feeding.192 Lack of early exposure to infections may leave the immune system unprepared for infection at a later time and could lead, in some cases, to an abnormal immunologic response that increases the risk of leukemia. Evidence in support of this contention includes the observation of an inverse association of leukemia with time in attendance at day-care centers,193–195 although this observation is not consistent.196,197 Also relevant is the observation of an inverse association of leukemia with birth order.166,167 Studies that have inquired directly about infection history or have searched for serologic evidence of infection, both of which are problematic in case-control studies, have not identified particular candidate microbial agents either serologically198–203 or through surveys.119,144,194,197,204–209 There is no direct evidence as to the nature of the infection

57

58

Logan G. Spector et al.

or of the abnormal response. It could be (as with feline leukemia virus in cats and HTLV-1 in human beings) a single but common transforming virus to which the exposed individual has an inadequate protective response. Alternatively, the immune system might mount an overexuberant or unbalanced response to some microbial antigens or superantigens.210 In the absence of direct biologic evidence for an infectious etiology in childhood leukemia, further epidemiologic studies are required. Future studies should be designed to test whether the risk of cALL increases in association with certain major histocompatibility types,211 with reduced infections in infancy, with the absence of prolonged breast feeding, and with a lack of social contact with other children in infancy. Some of these associations have been reported in smaller case-control studies and caseseries surveys.182,212

Interactions between genetic and environmental factors Biologic markers (biomarkers) related to genetic mutational events can provide a means of evaluating interactions between inherited genetic susceptibility and environmental health hazards in the etiology of childhood acute leukemias. Biomarkers of both exposure and effect not only provide bases for assessing interactions but may be indicative of future disease risk. At present, however, there is little information on the predictive value of these assays for populations or individuals. The following sections discuss some biomarkers currently under investigation that may generate new insights into leukemogenesis.

Microsatellite instability Microstatellite DNA is noncoding DNA consisting of short tandem repeat sequences that are unique to each individual. Microsatellite instability (MSI) is characterized by mutations in these short tandem repeat sequences, which appear to reflect multiple replication errors brought on by defective mismatch repair genes and which contribute to a “mutator phenotype.”213 MSI has been implicated in several human malignancies, including hereditary nonpolyposis colon cancer, gliomas, and lung cancer. Few studies have investigated MSI in childhood leukemia. The largest such study to date examined 48 primary samples from ALL patients and found that five (10%) exhibited MSI.214 Of interest, the authors found that several of the sites of instability were located in chromosomal regions associated with childhood ALL (including regions containing the TEL gene

in chromosome 12p, the p16 gene on chromosome 9, and the long arm of chromosome 6). Although the overall frequency of instability was low, these data suggest that localized MSI may identify a fragile chromosomal region that could result in an alteration of surrounding target genes and thus lead to leukemia in some children. In another study, Baccichet et al.215 demonstrated alterations in microsatellite patterns in five of six patients with childhood T-cell ALL. Further evidence of genomic instability was provided by a loss of heterozygosity in chromosomes 6p, 9p, and 12p. In fact, two-thirds of these patients had deletion of chromosome 9p21, the location of the tumor suppressor gene p16. Larger studies are needed to further characterize the role of MSI in childhood leukemia.

Mutation frequencies Somatic mutations in the hypoxanthine-guanine phosphoribosyltransferase (HPRT) gene are rare occurrences in the T lymphocytes of normal individuals. Lacking pathogenic significance, these events can serve as biomarkers for assessing environmental genotoxicity. Finette et al.216 demonstrated an early childhood HPRT mutational spectrum that was quite distinct from the adult background spectrum. This age-frequency distribution of HPRT mutations correlates with the age-frequency distribution of childhood ALL and merits further exploration.216

HLA-DR and susceptibility to childhood ALL Since the demonstration of the influence of the major histocompatibility complex (MHC) on mouse leukemia, an HLA association has been considered as a possible genetic risk factor. Dorak et al.217 have demonstrated a moderate association with the most common allele in the HLA-DR53 group, HLA-DRB1*04, that was stronger in males.217 In addition, homozygosity for HLA-DRB4*01, encoding the HLA-DR53 specificity was increased among patients. These associations could possibly suggest a malespecific increase in homozygosity for HLA-DRB4*01. The cross-reactivity between HLA-DR53 and H-2Ek, extensive mimicry of the immunodominant epitope of HLA-DR53 by several carcinogenic viruses, and the extra amount of DNA in the vicinity of the HLA-DRB4 gene argue for the case that HLA-DRB4*01 may be one of the genetic risk factors for childhood ALL. Comparison of DQA1 and DQB1 alleles in a case-control study revealed that male but not female patients had a higher frequency of DQA1*0101/*0104 and DQB1*0501, thus suggesting a male-associated susceptibility haplotype in ALL, supporting an infectious etiology.218

Epidemiology and etiology

Susceptibility to childhood leukemia: influence of CYP1A1, CYP2D6, GSTM1 and GSTT1 genetic polymorphisms Several investigators have examined the role of polymorphisms in genes encoding drug-metabolizing enzymes such as glutathione S-transferases and cytochrome P-450 in the development of pediatric cancers. Both genes are involved in carcinogen metabolism and have been shown to influence the risk of a variety of adult cancers.219–222 Davies et al.223 demonstrated that the GSTM1 null genotype is significantly more frequent among childhood AML, particularly the M3 and M4 FAB groups.223 Chen et al.224 compared the frequency of the null phenotype for GSTTI or GSTM1, or both, in children with ALL with that in healthy controls. Their results showed that the double-null genotype of GSTT1 and GSTM1 is more common among black than white children with ALL. However, Davies et al.225 failed to show any association between GST polymorphisms and the risk of developing childhood ALL.225 Possible links between the risk of ALL and inducibility of the drug-metabolizing enzyme CYP1A1 have been hypothesized.226 The results of these studies indicate a possible role of gene-environment interaction in the etiology of childhood leukemia that needs to be explored in greater detail.

Issues and future directions Impressive advances in the understanding of leukemia cell biology and the treatment of childhood acute leukemia stand in striking contrast to the relatively limited progress toward understanding the etiology of this heterogeneous group of diseases. Epidemiologic research addressing the etiology of childhood cancer, while considerable in volume, has been limited because of several issues. These include the retrospective nature of study designs, restricting the incorporation of biologic and clinical parameters, and the difficulties in identifying a sufficiently large study population to allow subgroup analyses. In the context of casecontrol studies, the design of choice for all childhood cancer epidemiologic studies, recall bias is a major concern. With regard to biologic samples, the main tissue studied has been blood. Studies that seek to use blood for biomarkers are likely to face difficulties with poor participation rates, especially in relation to control subjects. These barriers may be overcome by the development of techniques using very small samples, or the use of saliva. Case-control studies that examine evidence of infection face problematic temporality. There is mounting evidence that the cellular and molecular characteristics of biologically distinct

subgroups may define the most promising approaches for future epidemiologic/etiologic investigations. Specifically, it may be most productive to evaluate environmental and/or genetic hypotheses within well-defined, wellcharacterized, homogeneous groups of leukemias. Standard analytic epidemiologic investigations of more broadly defined populations of childhood leukemia have provided a relatively limited amount of etiologically relevant data over the past three decades. Several large epidemiologic studies – most of which incorporate detailed biological characterization of cases – are currently nearing completion in North America and Europe. If these investigations fail to identify biologically relevant associations, it is possible that the etiology of the majority of childhood acute leukemias may remain elusive for decades to come.

REFERENCES 1 Gurney, J. G., Severson, R. K., Davis, S., et al. Incidence of cancer in children in the United States. Sex-, race-, and 1-year agespecific rates by histologic type. Cancer, 1995; 75: 2186–95. 2 Linet, M. S. & Devesa, S. S. Descriptive epidemiology of childhood leukaemia. Br J Cancer, 1991; 63: 424–9. 3 Glazer, E. R., Perkins, C. I., Young, J. L., Jr., et al. Cancer among Hispanic children in California, 1988–1994: comparison with non-Hispanic white children. Cancer, 1999; 86: 1070–9. 4 Douer, D., Preston-Martin, S., Chang, E., et al. High frequency of acute promyelocytic leukemia among Latinos with acute myeloid leukemia. Blood, 1996; 87: 308–13. 5 Ferlay, J., Bray, F., Pisani, P., & Parkin, D. M. GLOBOCAN 2000: Cancer Incidence, Mortality and Prevalence Worldwide. (Lyon, France: IARC Press, 2001). 6 Gurney, J. G., Davis, S., Severson, R. K., et al. Trends in cancer incidence among children in the U. S. Cancer, 1996; 78: 532–41. 7 Blair, V. & Birch, J. M. Patterns and temporal trends in the incidence of malignant disease in children: I. Leukaemia and lymphoma. Eur J Cancer, 1994; 30A: 1490–8. 8 McWhirter, W. R. & Petroeschevsky, A. L. Incidence trends in childhood cancer in Queensland, 1973–1988. Med J Aust, 1991; 154: 453–5. 9 Bunin, G. R., Feuer, E. J., Witman, P. A., & Meadows, A. T. Increasing incidence of childhood cancer: report of 20 years experience from the greater Delaware Valley Pediatric Tumor Registry. Paediatr Perinat Epidemiol, 1996; 10: 319–38. 10 Kaatsch, P., Haaf, G., & Michaelis, J. Childhood malignancies in Germany – methods and results of a nationwide registry. Eur J Cancer, 1995; 31A: 993–9. 11 Linet, M. S., Ries, L. A., Smith, M. A., Tarone, R. E., & Devesa, S. S. Cancer surveillance series: recent trends in childhood cancer incidence and mortality in the United States. J Natl Cancer Inst, 1999; 91: 1051–8.

59

60

Logan G. Spector et al.

12 Taylor, G. M. & Birch, J. M. The hereditary basis of human leukemia. In Henderson, E. S., Lister, T. A., & Greaves, M. F., eds., Leukemia (Philadelphia, PA: W. B. Saunders, 1996), pp. 210–45. 13 Norppa, H. Cytogenetic markers of susceptibility: influence of polymorphic carcinogen-metabolizing enzymes. Environ Health Perspect, 1997; 105(Suppl. 4): 829–35. 14 Spector, L. G., Xie, Y., Robison, L. L., et al. Maternal diet and infant leukemia: the DNA topoisomerase II inhibitor hypothesis: a report from the children’s oncology group. Cancer Epidemiol Biomarkers Prev, 2005; 14: 651–5. 15 Puchkova, G. P., Prigogina, E. L., Fleischmann, E. W., Drosdova, T. S., & Mayakwa, S. A. Chromosome abnormalities in chronic myeloid leukemia in children. Hum Genet, 1983; 64: 257–62. 16 Woods, W. G., Nesbit, M. E., Buckley, J., et al. Correlation of chromosome abnormalities with patient characteristics, histologic subtype, and induction success in children with acute nonlymphocytic leukemia. J Clin Oncol, 1985; 3: 3–11. 17 Cimino, G., Lo Coco, F., Biondi, A., et al. ALL-1 gene at chromosome 11q23 is consistently altered in acute leukemia of early infancy. Blood, 1993; 82: 544–6. 18 Pui, C. H., Raimondi, S. C., Murphy, S. B., et al. An analysis of leukemic cell chromosomal features in infants. Blood, 1987; 69: 1289–93. 19 Stone, R. M. & Mayer, R. J., The unique aspects of acute promyelocytic leukemia. J Clin Oncol, 1990; 8: 1913–21. 20 Secker-Walker, L. M., Berger, R., Fenaux, P., et al. Prognostic significance of the balanced t(1;19) and unbalanced der(19)t(1;19) translocations in acute lymphoblastic leukemia. Leukemia, 1992; 6: 363–9. 21 Crist, W., Carroll, A., Shuster, J., et al. Philadelphia chromosome positive childhood acute lymphoblastic leukemia: clinical and cytogenetic characteristics and treatment outcome. A Pediatric Oncology Group study. Blood, 1990; 76: 489–94. 22 Shurtleff, S. A., Buijs, A., Behm, F. G., et al. TEL/AML1 fusion resulting from a cryptic t(12;21) is the most common genetic lesion in pediatric ALL and defines a subgroup of patients with an excellent prognosis. Leukemia, 1995; 9: 1985–9. 23 Robison, L. L. & Neglia, J. P. Epidemiology of Down syndrome and childhood acute leukemia. Prog Clin Biol Res, 1987; 246: 19–32. 24 Fong, C. T. & Brodeur, G. M. Down’s syndrome and leukemia: epidemiology, genetics, cytogenetics, and the mechanisms of leukemogenesis. Cancer Genet Cytogenet, 1987; 28: 55–76. 25 Zipursky, A., Thorner, P., De Harven, E., Christensen, H., & Doyle, J. Myelodysplasia and acute megakaryoblastic leukemia in Down’s syndrome. Leuk Res, 1994; 18: 163–71. 26 Sacchi, N. Down syndrome and chromosome 21 abnormalities in leukaemia. Baillieres Clin Haematol, 1992; 5: 815–31. 27 Mullvihill, J. J. Congenital and genetic diseases. In J. F. Fraumeni, Jr., ed., Persons at High Risk of Cancer (San Diego, CA: Academic Press, 1975), pp. 3–31. 28 Bader, J. L. & Miller, R. W. Neurofibromatosis and childhood leukemia. J Pediatr, 1978; 92: 925–9. 29 German, J., Bloom, D., & Passarge, E. Bloom’s syndrome. VII. Progress report for 1978. Clin Genet, 1979; 15: 361–7.

30 Woods, W. G., Roloff, J. S., Lukens, J. N., & Krivit, W. The occurrence of leukemia in patients with the Shwachman syndrome. J Pediatr, 1981; 99: 425–8. 31 Hecht, F. & Hecht, B. K. Cancer in ataxia-telangiectasia patients. Cancer Genet Cytogenet, 1990; 46: 9–19. 32 Linet, M. The Leukemias: Epidemiological Aspects (New York: Oxford University Press, 1985). 33 Bloomfield, C. D. & Brunning, R. D. Acute leukemia as a terminal event in nonleukemic hematopoietic disorders. Semin Oncol, 1976; 3: 297–317. 34 Zeidler, C., Reiter, A., Yakisan, E., et al. Long-term treatment with recombinant human granulocyte colony stimulating factor in patients with severe congenital neutropenia. Klin Padiatr, 1993; 205: 264–71. 35 Shannon, K. M., Turhan, A. G., Rogers, P. C., & Kan, Y. W. Evidence implicating heterozygous deletion of chromosome 7 in the pathogenesis of ftreatial leukemia associated with monosomy 7. Genomics, 1992; 14: 121–5. 36 Narod, S. A., Stiller, C., & Lenoir, G. M. An estimate of the heritable fraction of childhood cancer. Br J Cancer, 1991; 63: 993–9. 37 Robison, L. L., Nesbit, M. E., Jr., Sather, H. N., et al. Down syndrome and acute leukemia in children: a 10-year retrospective survey from Children’s Cancer Study Group. J Pediatr, 1984; 105: 235–42. 38 Watson, M. S., Carroll, A. J., Shuster, J. J., et al. Trisomy 21 in childhood acute lymphoblastic leukemia: a Pediatric Oncology Group study (8602). Blood, 1993; 82: 3098–102. 39 Mertens, A. C., Wen, W., Davies, S. M., et al. Congenital abnormalities in children with acute leukemia: a report from the Children’s Cancer Group. J Pediatr, 1998; 133: 617–23. 40 Narod, S. A., Hawkins, M. M., Robertson, C. M., & Stiller, C. A. Congenital anomalies and childhood cancer in Great Britain. Am J Hum Genet, 1997; 60: 474–85. 41 Anderson, R. Familial leukemia. Am J Dis Child, 1951; 81: 313–322. 42 Maklin, M. Inheritance of cancer of the stomach and large intestine in man. J Natl Cancer Inst, 1960; 24: 551–571. 43 Gunz, F. W., Gunz, J. P., Vincent, P. C., et al. Thirteen cases of leukemia in a family. J Natl Cancer Inst, 1978; 60: 1243–50. 44 Farwell, J. & Flannery, J. T. Cancer in relatives of children with central-nervous-system neoplasms. N Engl J Med, 1984; 311: 749–53. 45 MacMahon, B. & Levy, M. A. Prenatal origin of leukemia: evidence from twins. N Engl J Med, 1964; 270: 1082–1085. 46 Keith, L., Brown, E. R., Ames, B., Stotsky, M., & Keith, D. M. Leukemia in twins: antenatal and postnatal factors. Acta Genet Med Gemellol, 1976; 25: 336–41. 47 Chaganti, R. S., Miller, D. R., Meyers, P. A., & German, J. Cytogenetic evidence of the intrauterine origin of acute leukemia in monozygotic twins. N Engl J Med, 1979; 300: 1032–4. 48 Hartley, S. E. & Sainsbury, C. Acute leukaemia and the same chromosome abnormality in monozygotic twins. Hum Genet, 1981; 58: 408–10. 49 Strong, L. C. Genetics, etiology, and epidemiology of childhood cancer. In W. W. Sutow, T. J. Vietti & D. J. Ferrbach, eds.,

Epidemiology and etiology

50 51

52

53

54

55

56

57

58

59 60 61

62

63

64

65

66

67

68

Clinical Pediatric Oncology (St. Louis, MO: Mosby, 1984), pp. 14–41. Elwood, P. C. Possible explanation of the high concordance for acute leukemia in monozygotic twins. Lancet, 1971; 1: 699. Buckley, J. D., Buckley, C. M., Breslow, N. E., et al. Concordance for childhood cancer in twins. Med Pediatr Oncol, 1996; 26: 223–9. Ford, A. M., Ridge, S. A., Cabrera, M. E., et al. In utero rearrangements in the trithorax-related oncogene in infant leukaemias. Nature, 1993; 363: 358–60. Maia, A. T., Ford, A. M., Jalali, G. R., et al. Molecular tracking of leukemogenesis in a triplet pregnancy. Blood, 2001; 98: 478–82. Ford, A. M., Pombo-de-Oliveira, M. S., McCarthy, K. P., et al. Monoclonal origin of concordant T-cell malignancy in identical twins. Blood, 1997; 89: 281–5. Ford, A. M., Bennett, C. A., Price, C. M., et al. Fetal origins of the TEL-AML1 fusion gene in identical twins with leukemia. Proc Natl Acad Sci U S A, 1998; 95: 4584–8. Hawkins, M. M., Draper, G. J., & Winter, D. L. Cancer in the offspring of survivors of childhood leukaemia and non-Hodgkin lymphomas. Br J Cancer, 1995; 71: 1335–9. Bajnoczky, K., Khezri, S., Kajtar, P., et al. No chromosomal instability in offspring of survivors of childhood malignancy. Cancer Genet Cytogenet, 1999; 109: 79–80. Perrillat, F., Clavel, J., Jaussent, I., et al. Family cancer history and risk of childhood acute leukemia (France). Cancer Causes Control, 2001; 12: 935–41. Grenwald. Benzene and leukemia. Crit Rev Toxicol, 2002; 32: 155–210. Stewart, A. A survey of childhood malignancies. Br Med J, 1958; 2: 1495–507. MacMahon, B. & Newill, V. A. Birth characteristics of children dying of malignant neoplasms. J Natl Cancer Inst, 1962; 28: 231–44. Gibson, R. W., Bross, I. D. J., Graham, S., et al. Leukemia in children exposed to multiple risk factors. N Engl J Med, 1968; 279: 906–9. Stewart, A. & Kneale, G. W. Radiation dose effects in relation to obstetric x-rays and childhood cancers. Lancet, 1970; 1: 1185–8. Savitz, D. A., Wachtel, H., Barnes, F. A., John, E. M., & Tvrdik, J. G. Case-control study of childhood cancer and exposure to 60-Hz magnetic fields. Am J Epidemiol, 1988; 128: 21–38. London, S. J., Thomas, D. C., Bowman, J. D., et al. Exposure to residential electric and magnetic fields and risk of childhood leukemia. Am J Epidemiol, 1991; 134: 923–37. Linet, M. S., Hatch, E. E., Kleinerman, R. A., et al. Residential exposure to magnetic fields and acute lymphoblastic leukemia in children. N Engl J Med, 1997; 337: 1–7. Preston, D. L., Kusumi, S., Tomonaga, M., et al. Cancer incidence in atomic bomb survivors. Part III. Leukemia, lymphoma and multiple myeloma, 1950–1987. Radiat Res, 1994; 137(2 Suppl.): S68–97. Doll, R. Hazards of ionising radiation: 100 years of observations on man. Br J Cancer, 1995; 72: 1339–49.

69 Harvey, E. B., Boice, J. D., Jr., Honeyman, M., & Flannery, J. T. Prenatal x-ray exposure and childhood cancer in twins. N Engl J Med, 1985; 312: 541–5. 70 Rodvall, Y., Pershagen, G., Hrubec, Z., et al. Prenatal X-ray exposure and childhood cancer in Swedish twins. Int J Cancer, 1990; 46: 362–5. 71 Mole, R. H. Childhood cancer after prenatal exposure to diagnostic X-ray examinations in Britain. Br J Cancer, 1990; 62: 152–68. 72 Boice, J. D. & Inskip, P. D. Radiation-induced leukemia. In E. S. Henderson, T. A. Lister, & M. F. Greaves, eds., Leukemia (Philadelphia, PA: W. B. Saunders, 1996), pp. 195–209. 73 Darby, S. C. & Weiss, H. A. Human studies in radiation leukemogenesis. In J. H. Hendry & B. I. Lord, eds., Radiation Toxicology: Bone Marrow and Leukemia (London: Taylor & Francis, 1995), pp. 337–53. 74 Doll, R. & Wakeford, R. Risk of childhood cancer from fetal irradiation. Br J Radiol, 1997; 70: 130–9. 75 Simpson, C. I., Hempelmann, L. H., & Fuller, L. M. Neoplasms in children treated with x-rays in infancy for thymic enlargement. Radiology, 1955; 64: 840–55. 76 Ron, E. & Modon, B. Thyroid and other neoplasms following childhood scalp irradiation. In J. D. Boice & J. F. Fraumeni, eds., Radiation Carcinogenesis, Epidemiology and Biological Significance (New York: Raven Press, 1984), pp. 139–51. 77 Tucker, M. A., Coleman, C. N., Cox, R. S., Varghese, A. & Rosenberg, S. A. Risk of second cancers after treatment for Hodgkin’s disease. N Engl J Med, 1988; 318: 76–81. 78 Egeler, R. M., Neglia, J. P., Arico, M., et al. Acute leukemia in association with Langerhans cell histiocytosis. Med Pediatr Oncol, 1994; 23: 81–5. 79 Infante-Rivard, C., Mathonnet, G., & Sinnett, D. Risk of childhood leukemia associated with diagnostic irradiation and polymorphisms in DNA repair genes. Environ Health Perspect, 2000; 108: 495–8. 80 Linos, A., Kyle, R. A., Elveback, L. R., & Kurland, L. T. Leukemia in Olmsted County, Minnesota, 1965–1974. Mayo Clin Proc, 1978; 53: 714–18. 81 Boice, J. D., Jr. The danger of X-rays – real or apparent? N Engl J Med, 1986; 315: 828–30. 82 Hjalmars, U., Kulldorff, M., & Gustafsson, G. Risk of acute childhood leukaemia in Sweden after the Chernobyl reactor accident. Swedish Child Leukaemia Group. BMJ, 1994; 309: 154–7. 83 Auvinen, A., Hakama, M., Arvela, H., et al. Fallout from Chernobyl and incidence of childhood leukaemia in Finland, 1976– 92. BMJ, 1994; 309: 151–4. 84 Cartwright, R. A., McKinney, P. A., Alexander, F. E., & Ricketts, J. Leukemia in young children. Lancet, 1988; 2: 960. 85 Gibson, B. E., Eden, O. B., Barrett, A., Stiller, C. A., & Draper, G. J. Leukaemia in young children in Scotland. Lancet, 1988; 2: 630. 86 Michaelis, J., Kaletsch, U., Burkart, W., & Grosche, B. Infant leukaemia after the Chernobyl accident. Nature, 1997; 387: 246. 87 Petridou, E., Proukakis, C., Tong, D., et al. Trends and geographical distribution of childhood leukemia in Greece in

61

62

Logan G. Spector et al.

88

89

90 91

92

93

94

95 96

97

98

99

100

101

102

103

relation to the Chernobyl accident. Scand J Soc Med, 1994; 22: 127–31. Petridou, E., Trichopoulos, D., Dessypris, N., et al. Infant leukaemia after in utero exposure to radiation from Chernobyl. Nature, 1996; 382: 352–3. Parkin, D. M., Cardis, E., Masuyer, E., et al. Childhood leukemia following the Chernobyl accident: the European Childhood Leukemia-Lymphoma Incidence Study (ECLIS). Eur J Cancer, 1992; 29A: 87–95. Ivanov, E. P., Tolochoko, G., Lazarev, V. S. & Shuvaeva, L. Child leukaemia after Chernobyl. Nature, 1993; 365: 702. United Nations Scientific Committee on the Effects of Atomic Radiation (UNSCEAR). Sources and Effects of Ionizing Radiation. New York: United Nations, 1994. Stevens, W., Thomas, D. C., Lyon, J. L., et al. Leukemia in Utah and radioactive fallout from the Nevada test site. A case-control study. JAMA, 1990; 264: 585–91. Cook-Mozaffari, P. J., Darby, S. C., Doll, R., et al. Geographical variation in mortality from leukaemia and other cancers in England and Wales in relation to proximity to nuclear installations, 1969–78. Br J Cancer, 1989; 59: 476–85. Bithell, J. F., Dutton, S. J., Draper, G. J., & Neary, N. M. Distribution of childhood leukaemias and non-Hodgkin’s lymphomas near nuclear installations in England and Wales. BMJ, 1994; 309: 501–5. Hill, C. & Laplanche, A. Overall mortality and cancer mortality around French nuclear sites. Nature, 1990; 347: 755–7. Jablon, S., Hrubec, Z. & Boice, J. D., Jr. Cancer in populations living near nuclear facilities. A survey of mortality nationwide and incidence in two states. JAMA, 1991; 265: 1403–8. Jablon, S., Hrubec, Z. & Boice, J. D., Jr. Cancer in Populations Living Near Nuclear Facilities. NIH Publication 90–874). (Bethesda, MD: Public Health Service, Department of Health and Human Services, 1990). Michaelis, J., Keller, B., Haaf, G., & Kaatsch, P. Incidence of childhood malignancies in the vicinity of west German nuclear power plants. Cancer Causes Control, 1992; 3: 255–63. Clarke, E. A., McLaughlin, J., & Anderson, T. W. Childhood Leukaemia Around Canadian Nuclear Facilities: Phase II. Final Report. (Ottawa, Canada: Atomic Energy Control Board, 1991). Committee of Medical Aspects of Radiation in the Environment (COMARE). The Incidence of Cancer and Leukaemia in Young People in the Vicinity of the Sellafield Site, West Cumbria: Further Studies and an Update of the Situation Since the Publication of the Report of the Black Advisory Group in 1984. (London: Her Majesty’s Stationery Office, 1996). Gardner, M. J. Father’s occupational exposure to radiation and the raised level of childhood leukemia near the Sellafield nuclear plant. Environ Health Perspect, 1991; 94: 5–7. Shu, X. O., Reaman, G. H., Lampkin, B., et al. Association of paternal diagnostic X-ray exposure with risk of infant leukemia. Investigators of the Children’s Cancer Group. Cancer Epidemiol Biomarkers Prev, 1994; 3: 645–53. McLaughlin, J. R., Clarke, E. A., Nishri, E. D., & Anderson, T. W. Childhood leukemia in the vicinity of Canadian nuclear facilities. Cancer Causes Control, 1993; 4: 51–8.

104 Parker, L., Craft, A. W., Smith, J., et al. Geographical distribution of preconceptional radiation doses to fathers employed at the Sellafield nuclear installation, West Cumbria. BMJ, 1993; 307: 966–71. 105 Henshaw, D. L., Eatough, J. P., & Richardson, R. B. Radon as a causative factor in induction of myeloid leukaemia and other cancers. Lancet, 1990; 335: 1008–12. 106 Alexander, F. E., McKinney, P. A., & Cartwright, R. A. Radon and leukaemia. Lancet, 1990; 335: 1336–7. 107 Lubin, J. H., Linet, M. S., Boice, J. D. Jr., et al. Case-control study of childhood acute lymphoblastic leukemia and residential radon exposure. J Natl Cancer Inst, 1998; 90: 294– 300. 108 Steinbuch, M., Weinberg, C. R., Buckley, J. D., Robison, L. L., & Sandler, D. P. Indoor residential radon exposure and risk of childhood acute myeloid leukaemia. Br J Cancer, 1999; 81: 900–6. 109 Kaletsch, U., Kaatsch, P., Meinert, R., et al. Childhood cancer and residential radon exposure – results of a population-based case-control study in Lower Saxony (Germany). Radiat Environ Biophys, 1999; 38: 211–5. 110 UK Childhood Cancer Study Investigators. The United Kingdom Childhood Cancer Study of exposure to domestic sources of ionising radiation: 1: radon gas. Br J Cancer, 2002; 86: 1721–6. 111 Axelson, O., Fredrikson, M., Akerblom, G., & Hardell, L. Leukemia in childhood and adolescence and exposure to ionizing radiation in homes built from uranium-containing alum shale concrete. Epidemiology, 2002; 13: 146–50. 112 Watts, G. Power to confuse. BMJ, 2005; 330: 1293. 113 McBride, M. L., Gallagher, R. P., Theirault, G., et al. Powerfrequency electric and magnetic fields and risk of childhood leukemia in Canada. Am J Epidemiol, 1999; 149: 831–42. 114 Anonymous. Exposure to power-frequency magnetic fields and the risk of childhood cancer. UK Childhood Cancer Study Investigators. Lancet, 1999; 354: 1925–31. 115 Soderberg, K. C., Naumburg, E., Anger, G., et al. Childhood leukemia and magnetic fields in infant incubators. Epidemiology, 2002; 13: 45–9. 116 Ahlbom, A., Day, N., Feychting, M., et al. A pooled analysis of magnetic fields and childhood leukaemia. Br J Cancer, 2000; 83: 692–8. 117 Greenland, S., Sheppard, A. R., Kaune, W. T., Poole, C., & Kelsh, M. A. A pooled analysis of magnetic fields, wire codes, and childhood leukemia. Childhood Leukemia-EMF Study Group. Epidemiology, 2000; 11: 624–34. 118 Lowengart, R. A., Peters, J. M., Cicioni, C., et al. Childhood leukemia and parents’ occupational and home exposures. J Natl Cancer Inst, 1987; 79: 39–46. 119 Shu, X. O., Gao, Y. T., Brinton, L. A., et al. A population-based case-control study of childhood leukemia in Shanghai. Cancer, 1988; 62: 635–44. 120 Buckley, J. D., Robison, L. L., Swotinsky, R., et al. Occupational exposures of parents of children with acute nonlymphocytic leukemia: a report from the Children’s Cancer Study Group. Cancer Res, 1989; 49: 4030–7.

Epidemiology and etiology

121 Leiss, J. K. & Savitz, D. A. Home pesticide use and childhood cancer: a case-control study. Am J Public Health, 1995; 85: 249–52. 122 Zahm, S. H. Childhood leukemia and pesticides. Epidemiology, 1999; 10: 473–5. 123 Daniels, J. L., Olshan, A. F., & Savitz, D. A. Pesticides and childhood cancers. Environ Health Perspect, 1997; 105: 1068–77. 124 Shu, X. O., Stewart, P., Wen, W. Q., et al. Parental occupational exposure to hydrocarbons and risk of acute lymphocytic leukemia in offspring. Cancer Epidemiol Biomarkers Prev, 1999; 8: 783–91. 125 Steensel-Moll, H. A. van, Valkenburg, H. A., Vandenbroucke, J. P., & Zanen, G. E. van. Are maternal fertility problems related to childhood leukaemia? Int J Epidemiol, 1985; 14: 555–9. 126 Gold, E. B., Diener, M. D., & Szklo, M. Parental occupations and cancer in children – a case-control study and review of the methodologic issues. J Occup Med, 1982; 24: 578–84. 127 McKinney, P. A., Roberts, B. E., O’Brien, C., et al. Chronic myeloid leukaemia in Yorkshire: a case control study. Acta Haematol, 1990; 83: 35–8. 128 Shaw, G., Lavey, R., Jackson, R., & Austin, D. Association of childhood leukemia with maternal age, birth order, and paternal occupation. A case-control study. Am J Epidemiol, 1984; 119: 788–95. 129 Infante-Rivard, C., Mur, P., Armstrong, B., Alvarez-Dardet, C., & Boulmar, F. Acute lymphoblastic leukaemia among Spanish children and mothers’ occupation: a case-control study. J Epidemiol Community Health, 1991; 45: 11–5. 130 Schuz, J., Kaletsch, U., Meinert, R., Kaatsch, P., & Michaelis, J. Risk of childhood leukemia and parental self-reported occupational exposure to chemicals, dusts, and fumes: results from pooled analyses of German population-based case-control studies. Cancer Epidemiol Biomarkers Prev, 2000; 9: 835–8. 131 Lyons, R. A., Monaghen, S. P., Heaven, M., et al. Incidence of leukaemia and lymphoma in young people in the vicinity of the petrochemical plant at Baglan Bay, South Wales, 1974 to 1991. Occup Environ Med, 1995; 52: 225–8. 132 Sans, S., Elliott, P., Kleinschmidt, I., et al. Cancer incidence and mortality near the Baglan Bay petrochemical works, South Wales. Occup Environ Med, 1995; 52: 217–24. 133 Freedman, D. M., Stewart, P., Kleinerman, R. A., et al.. Household solvent exposures and childhood acute lymphoblastic leukemia. Am J Public Health, 2001; 91: 564–7. 134 Pearson, R. L., Wachtel, H., & Ebi, K. L. Distance-weighted traffic density in proximity to a home is a risk factor for leukemia and other childhood cancers. J Air Waste Manag Assoc, 2000; 50: 175–80. 135 Reynolds, P., Elkin, E., Scalf, R., Von Behren, J., & Neutra, R. R. A case-control pilot study of traffic exposures and early childhood leukemia using a geographic information system. Bioelectromagnetics, 2001; Suppl. 5: S58–68. 136 Reynolds, P., Von Behren, J., Gunier, R. B., et al. Traffic patterns and childhood cancer incidence rates in California, United States. Cancer Causes Control, 2002; 13: 665–73. 137 Alexander, F., Cartwright, R., McKinney, P. A., & Ricketts, T. J. Investigation of spacial clustering of rare diseases: childhood

138

139

140

141

142

143

144

145

146

147

148

149

150

151

malignancies in North Humberside. J Epidemiol Community Health, 1990; 44: 39–46. Wulff, M., Hogberg, U., & Sandstrom, A. Cancer incidence for children born in a smelting community. Acta Oncol, 1996; 35: 179–83. Feychting, M., Plato, N., Nise, G., & Ahlbom, A. Paternal occupational exposures and childhood cancer. Environ Health Perspect, 2001; 109: 193–6. Robison, L. L., Buckley, J. D., Daigle, A. E., et al. Maternal drug use and risk of childhood nonlymphoblastic leukemia among offspring. An epidemiologic investigation implicating marijuana (a report from the Children’s Cancer Study Group). Cancer, 1989; 63: 1904–11. Sarasua, S. & Savitz, D. A. Cured and broiled meat consumption in relation to childhood cancer: Denver, Colorado (United States). Cancer Causes Control, 1994; 5: 141–8. Thompson, J. R., Gerald, P. F., Willoughby, M. L., & Armstrong, B. K. Maternal folate supplementation in pregnancy and protection against acute lymphoblastic leukaemia in childhood: a case-control study. Lancet, 2001; 358: 1935–40. Peters, J. M., Preston-Martin, S., London, S. J., et al. Processed meats and risk of childhood leukemia (California, USA). Cancer Causes Control, 1994; 5: 195–202. Buckley, J. D., Buckley, C. M., Ruccione, K., et al. Epidemiological characteristics of childhood acute lymphocytic leukemia. Analysis by immunophenotype. The Children’s Cancer Group. Leukemia, 1994; 8: 856–64. Severson, R. K., Buckley, J. D., Woods, W. G., Benjamin, D., & Robison, L. L. Cigarette smoking and alcohol consumption by parents of children with acute myeloid leukemia: an analysis within morphological subgroups – a report from the Children’s Cancer Group. Cancer Epidemiol Biomarkers Prev, 1993; 2: 433–9. Duijn, C. M. van, Steensel-Moll, H. A. van, Coebergh, J. W., & Zanen, G. E. van. Risk factors for childhood acute nonlymphocytic leukemia: an association with maternal alcohol consumption during pregnancy? Cancer Epidemiol Biomarkers Prev, 1994; 3: 457–60. Shu, X. O., Ross, J. A., Pendergrass, T. W., et al. Parental alcohol consumption, cigarette smoking, and risk of infant leukemia: a Children’s Cancer Group study. J Natl Cancer Inst, 1996; 88: 24–31. Sorahan, T., Lancashire, R., Prior, P., Peck, I., & Stewart, A. Childhood cancer and parental use of alcohol and tobacco. Ann Epidemiol, 1995; 5: 354–9. Stjernfeldt, M., Berglund, K., Lindsten, J., & Ludvigsson, J. Maternal smoking and irradiation during pregnancy as risk factors for child leukemia. Cancer Detect Prev, 1992; 16: 129–35. Mucci, L. A., Granath, F., & Cnattingius, S. Maternal smoking and childhood leukemia and lymphoma risk among 1,440,542 Swedish children. Cancer Epidemiol Biomarkers Prev, 2004; 13: 1528–33. John, E. M., Savitz, D. A., & Sandler, D. P. Prenatal exposure to parents’ smoking and childhood cancer. Am J Epidemiol, 1991; 133: 123–32.

63

64

Logan G. Spector et al.

152 Buckley, J. D., Hobbie, W. L., Ruccione, K., et al. Maternal smoking during pregnancy and the risk of childhood cancer [letter]. Lancet, 1986; 2: 519–20. 153 Pershagen, G., Ericson, A., & Otterblad-Olausson, P. Maternal smoking in pregnancy: does it increase the risk of childhood cancer? Int J Epidemiol, 1992; 21: 1–5. 154 McKinney, P. A. & Stiller, C. Maternal smoking during pregnancy and the risk of childhood leukaemia. Lancet, 1986; 2: 519. 155 Li, F. P. Maternal smoking during pregnancy and the risk of childhood cancer. Lancet, 1986; 2: 520. 156 Ji, B. T., Shu, X. O., Linet, M. S., et al. Paternal cigarette smoking and the risk of childhood cancer among offspring of nonsmoking mothers. J Natl Cancer Inst, 1997; 89: 238–44. 157 Sorahan, T., Lancashire, R. J., Hulten, M. A., Peck, I., & Stewart, A. M. Childhood cancer and parental use of tobacco: deaths from 1953 to 1955. Br J Cancer, 1997; 75: 134–8. 158 Brondum, J., Shu, X. O., Steinbuch, M., et al. Parental cigarette smoking and the risk of acute leukemia in children. Cancer, 1999; 85: 1380–8. 159 Sandler, D. P., Wilcox, A. J., & Everson, R. B. Cumulative effects of lifetime passive smoking on cancer risk. Lancet, 1985; 1: 312–5. 160 Kaye, S. A., Robison, L. L., Smithson, W. A., et al. Maternal reproductive history and birth characteristics in childhood acute lymphoblastic leukemia. Cancer, 1991; 68: 1351–5. 161 Yeazel, M. W., Buckley, J. D., Woods, W. G., Ruccione, K., & Robison, L. L. History of maternal fetal loss and increased risk of childhood acute leukemia at an early age. A report from the Children’s Cancer Group. Cancer, 1995; 75: 1718–27. 162 Gibson, R. W., Bross, I. D. J., & Graham, S. Leukemia in children exposed to multiple risk factors. N Engl J Med, 1968; 279: 906–9. 163 Ross, J. A., Potter, J. D., Shu, X. O., et al. Evaluating the relationships among maternal reproductive history, birth characteristics, and infant leukemia: a report from the Children’s Cancer Group. Ann Epidemiol, 1997; 7: 172–9. 164 Westergaard, T., Andersen, P. K., Pedersen, J. B., et al. Birth characteristics, sibling patterns, and acute leukemia risk in childhood: a population-based cohort study. J Natl Cancer Inst, 1997; 89: 939–47. 165 Hemminki, K., Kyyronen, P., & Vaittinen, P. Parental age as a risk factor of childhood leukemia and brain cancer in offspring. Epidemiology, 1999; 10: 271–5. 166 Dockerty, J. D., Draper, G., Vincent, T., Rowan, S. D., & Bunch, K. J. Case-control study of parental age, parity and socioeconomic level in relation to childhood cancers. Int J Epidemiol, 2001; 30: 1428–37. 167 Zack, M., Adami, H. O., & Ericson, A. Maternal and perinatal risk factors for childhood leukemia. Cancer Res, 1991; 51: 3696–701. 168 Hjalgrim, L. L., Rostgaard, K., Hjalgrim, H., et al. Birth weight and risk for childhood leukemia in Denmark, Sweden, Norway, and Iceland. J Natl Cancer Inst, 2004; 96: 1549–56 169 Petridou, E., Skalkidou, A., Dessypris, N., et al. Endogenous risk factors for childhood leukemia in relation to the IGF system

170

171

172

173

174 175

176

177

178

179

180 181 182 183

184

185

(Greece). The Childhood Haematologists-Oncologists Group. Cancer Causes Control, 2000; 11: 765–71. Lei, U., Wohlfahrt, J., Hjalgrim, H., et al. Neonatal level of thyroid-stimulating hormone and acute childhood leukemia. Int J Cancer, 2000; 88: 486–8. Gale, K. B., Ford, A. M., Repp, R., et al. Backtracking leukemia to birth: identification of clonotypic gene fusion sequences in neonatal blood spots. Proc Natl Acad Sci U S A, 1997; 94: 13 950–4. Pui, C. H., Ribeiro, R. C., Hancock, M. L., et al. Acute myeloid leukemia in children treated with epipodophyllotoxins for acute lymphoblastic leukemia. N Engl J Med, 1991; 325: 1682–7. Broeker, P. L., Super, H. G., Thirman, M. J., et al. Distribution of 11q23 breakpoints within the MLL breakpoint cluster region in de novo acute leukemia and in treatment-related acute myeloid leukemia: correlation with scaffold attachment regions and topoisomerase II consensus binding sites. Blood, 1996; 87: 1912–22. Wang, J. C. DNA topoisomerases. Annu Rev Biochem, 1996; 65: 635–92. Ross, J. A., Potter, J. D., Reaman, G. H., Pendergrass, T. W., & Robison, L. L. Maternal exposure to potential inhibitors of DNA topoisomerase II and infant leukemia (United States): a report from the Children’s Cancer Group. Cancer Causes Control, 1996; 7: 581–90. Ross, J. A. Maternal diet and infant leukemia: a role for DNA topoisomerase II inhibitors? Int J Cancer Suppl, 1998; 11: 26–8. Strick, R., Strissel, P. L., Borgers, S., Smith, S. L., & Rowley, J. D. Dietary bioflavonoids induce cleavage in the MLL gene and may contribute to infant leukemia. Proc Natl Acad Sci U S A, 2000; 97: 4790–5. Alexander, F. E., Patheal, S. L., Biondi, A., et al. Transplacental chemical exposure and risk of infant leukemia with MLL gene fusion. Cancer Res, 2001; 61: 2542–6. Ross, J. A., Potter, J. D., & Robison, L. L. Infant leukemia, topoisomerase II inhibitors, and the MLL gene. J Natl Cancer Inst, 1994; 86: 1678–80. Greave, M. F. Infant leukaemia biology, aetiology and treatment. Leukemia, 1996; 10: 372–7. Kinlen, L. J. Epidemiological evidence for an infective basis in childhood leukaemia. Br J Cancer, 1995; 71: 1–5. Anderson, R. M. & May, R. M. Immunisation and herd immunity. Lancet, 1990; 335: 641–5. Boutou, O., Guizard, A. V., Slama, R., Pottier, D., & Spira, A. Population mixing and leukaemia in young people around the La Hague nuclear waste reprocessing plant. Br J Cancer, 2002; 87: 740–5. Langford, I. Childhood leukaemia mortality and population change in England and Wales 1969–73. Soc Sci Med, 1991; 33: 435–40. Stiller, C. A. & Boyle, P. J. Effect of population mixing and socioeconomic status in England and Wales, 1979–85, on lymphoblastic leukaemia in children. BMJ, 1996; 313: 1297– 300.

Epidemiology and etiology

186 Dickinson, H. O. & Parker, L. Quantifying the effect of population mixing on childhood leukaemia risk: the Seascale cluster. Br J Cancer, 1999; 81: 144–51. 187 Koushik, A., King, W. D., & McLaughlin, J. R. An ecologic study of childhood leukemia and population mixing in Ontario, Canada. Cancer Causes Control, 2001; 12: 483–90. 188 Dickinson, H. O., Hammal, D. M., Bithell, J. F., & Parker, L. Population mixing and childhood leukaemia and non-Hodgkin’s lymphoma in census wards in England and Wales, 1966–87. Br J Cancer, 2002; 86: 1411–3. 189 Parslow, R. C., Law, G. R., Feltbower, R., Kinsey, S. E., & McKinney, P. A. Population mixing, childhood leukaemia, CNS tumours and other childhood cancers in Yorkshire. Eur J Cancer, 2002; 38: 2033–40. 190 Baccate, E. M. Social patterns of antibody to poliovirus. Lancet, 1983; 1: 778–83. 191 Dworsky, M., Yow, M., Stagno, S., Pass, R. F., & Alford, C. Cytomegalovirus infection of breast milk and transmission in infancy. Pediatrics, 1983; 72: 295–9. 192 Parker, L. Breast-feeding and cancer prevention. Eur J Cancer, 2001; 37: 155–8. 193 Ma, X., Buffler, P. A., Selvin, S., et al. Daycare attendance and risk of childhood acute lymphoblastic leukaemia. Br J Cancer, 2002; 86: 1419–24. 194 Perrillat, F., Clavel, J., Auclerc, M. F., et al. Day-care, early common infections and childhood acute leukaemia: a multicentre French case-control study. Br J Cancer, 2002; 86: 1064–9. 195 Gilham, C., Peto, J., Simpson, J. et at. Day care in infancy and risk of childhood acute lymphoblastic leukaemia: findings from UK case-contol study. BMJ, 2005; 330: 1294. 196 Rosenbaum, P. F., Buck, G. M., & Brecher, M. L. Early child-care and preschool experiences and the risk of childhood acute lymphoblastic leukemia. Am J Epidemiol, 2000; 152: 1136–44. 197 Neglia, J. P., Linet, M. S., Shu, X. O., et al. Patterns of infection and day care utilization and risk of childhood acute lymphoblastic leukaemia. Br J Cancer, 2000; 82: 234–40. 198 Gahrton, G., Wahren, B., Killander, D., & Foley, G. E. Epstein– Barr and other herpes virus antibodies in children with acute leukemia. Int J Cancer, 1971; 8: 242–9. 199 Groves, F. D., Sinha, D., Kayhty, H., Goedert, J. J., & Levine, P. H. Haemophilus influenzae type b serology in childhood leukaemia: a case-control study. Br J Cancer, 2001; 85: 337–40. 200 Heegaard, E. D., Jensen, L., Hornsleth, A., & Schmiegelow, K. The role of parvovirus B19 infection in childhood acute lymphoblastic leukemia. Pediatr Hematol Oncol, 1999; 16: 329–34. 201 MacKenzie, J., Gallagher, A., Clayton, R. A., et al. Screening for herpesvirus genomes in common acute lymphoblastic leukemia. Leukemia, 2001; 15: 415–21. 202 MacKenzie, J., Perry, J., Ford, A. M., Jarrett, R. F., & Greaves, M. JC and BK virus sequences are not detectable in leukaemic samples from children with common acute lymphoblastic leukaemia. Br J Cancer, 1999; 81: 898–9. 203 Salonen, M. J., Siimes, M. A., Salonen, E. M., Vaheri, A., & Koskiniemi, M. Antibody status to HHV-6 in children with leukaemia. Leukemia, 2002; 16: 716–9.

204 Dockerty, J. D., Skegg, D. C., Elwood, J. M., et al. Infections, vaccinations, and the risk of childhood leukaemia. Br J Cancer, 1999; 80: 1483–9. 205 McKinney, P. A., Juszczak, E., Findlay, E., Smith, K., & Thomson, C. S. Pre- and perinatal risk factors for childhood leukaemia and other malignancies: a Scottish case control study. Br J Cancer, 1999; 80: 1844–51. 206 Schuz, J., Kaatsch, P., Kaletsch, U., Meinert, R., & Michaelis, J. Association of childhood cancer with factors related to pregnancy and birth. Int J Epidemiol, 1999; 28: 631–9. 207 McKinney, P. A., Cartwright, R. A., Saiu, J. M., et al. The interregional epidemiological study of childhood cancer (IRESCC): a case control study of aetiological factors in leukaemia and lymphoma. Arch Dis Child, 1987; 62: 279–87. 208 Naumburg, E., Bellocco, R., Cnattingius, S., Jonzon, A., & Ekbom, A. Perinatal exposure to infection and risk of childhood leukemia. Med Pediatr Oncol, 2002; 38: 391–7. 209 Chan, L. C., Lam, T. H., Li, C. K., et al. Is the timing of exposure to infection a major determinant of acute lymphoblastic leukaemia in Hong Kong? Paediatr Perinat Epidemiol, 2002; 16: 154–65. 210 Greaves, M. F. & Alexander, F. E. An infectious etiology for common acute lymphoblastic leukemia in childhood? Leukemia, 1993; 7: 349–60. 211 Taylor, G. M. & Birch, J. M. The hereditary basis of human leukemia. In E. S. Henderson, T. A. Lister, & M. F. Greaves, eds., Leukemia (Philadelphia, PA: W. B. Saunders, 1996), pp. 210–45. 212 Petridou, E., Kassimos, D., Kalmanti, M., et al. Age of exposure to infections and risk of childhood leukaemia. BMJ, 1993; 307: 774. 213 Tasaka, T., Lee, S., Spira, S., et al. Microsatellite instability during the progression of acute myelocytic leukaemia. Br J Haematol, 1997; 98: 219–21. 214 Takeuchi, S., Seriu, T., Tasaka, T., et al. Microsatellite instability and other molecular abnormalities in childhood acute lymphoblastic leukaemia. Br J Haematol, 1997; 98: 134–9. 215 Baccichet, A., Benachenhou, N., Couture, F., Leclerc, J. M., & Sinnett, D. Microsatellite instability in childhood T cell acute lymphoblastic leukemia. Leukemia, 1997; 11: 797–802. 216 Finette, B. A., Poseno, T., & Albertini, R. J. V(D)J recombinasemediated HPRT mutations in peripheral blood lymphocytes of normal children. Cancer Res, 1996; 56: 1405–12. 217 Dorak, M. T., Lawson, T., Machulla, H. K., et al. Unravelling an HLA-DR association in childhood acute lymphoblastic leukemia. Blood, 1999; 94: 694–700. 218 Taylor, G. M., Robinson, M. D., Binchy, A., et al. Preliminary evidence of an association between HLA-DPB1*0201 and childhood common acute lymphoblastic leukaemia supports an infectious aetiology. Leukemia, 1995; 9: 440–3. 219 Rothman, N., Wacholder, S., Caporaso, N. E., et al. The use of common genetic polymorphisms to enhance the epidemiologic study of environmental carcinogens. Biochim Biophys Acta, 2001; 1471: C1–10. 220 Nakachi, K., Imai, K., Hayashi, S., & Kawajiri, K. Polymorphisms of the CYP1A1 and glutathione S-transferase genes associated

65

66

Logan G. Spector et al.

with susceptibility to lung cancer in relation to cigarette dose in a Japanese population. Cancer Res, 1993; 53: 2994–9. 221 Hirvonen, A., Nylund, L., Kociba, P., Husgafvel-Pursiainen, K., & Vainio, H. Modulation of urinary mutagenicity by genetically determined carcinogen metabolism in smokers. Carcinogenesis, 1994; 15: 813–5. 222 Chen, H., Sandler, D. P., Taylor, J. A., et al. Increased risk for myelodysplastic syndromes in individuals with glutathione transferase theta 1 (GSTT1) gene defect. Lancet, 1996; 347: 295–7. 223 Davies, S. M., Robison, L. L., Buckley, J. D., et al. Glutathione S-transferase polymorphisms in children with

myeloid leukemia: a Children’s Cancer Group study. Cancer Epidemiol Biomarkers Prev, 2000; 9: 563–6. 224 Chen, C. L., Liu, O., Pui, C. H., et al. Higher frequency of glutathione S-transferase deletions in black children with acute lymphoblastic leukemia. Blood, 1997; 89: 1701–7. 225 Davies, S. M., Bhatia, S., Ross, J. A., et al. Glutathione Stransferase genotypes, genetic susceptibility, and outcome of therapy in childhood acute lymphoblastic leukemia. Blood, 2002; 100: 67–71. 226 Blumer, J. L., Dunn, R., Esterhay, M. D., Yamashita, T. S., & Gross, S. Lymphocyte aromatic hydrocarbon responsiveness in acute leukemia of childhood. Blood, 1981; 58: 1081–8.

Part II Cell biology and pathobiology

4 Anatomy and physiology of hematopoiesis Connie J. Eaves and Allen C. Eaves

Introduction Hematopoiesis refers to all aspects of the process of blood cell production. Understanding this process requires a comprehensive knowledge of both the anatomy and the physiology of the blood-forming system. Here, the anatomy of the hematopoietic system is viewed as the distinguishable stages of differentiation that together make up the complete hierarchy of hematopoietic cells. These stages reflect the changes that initially endow cells in the embryo with hematopoietic differentiation potential (a step referred to as specification), in addition to those that subsequently constitute the processes of lineage restriction and terminal differentiation. The physiology of hematopoiesis refers to the dynamic aspects of these events and covers issues such as the determination of alternate outcomes, at both the cellular and molecular level, as well as their modulation during development and in response to injury or disease. Leukemias arise from clonal accumulations of mutations that impact the production and differentiation of blood cells. Because of the low probability of such events, a large number of divisions is thought to be required for their successive acquisition. It is therefore not surprising that, in many leukemias, the first leukemogenic mutation appears to take place in a hematopoietic stem cell. Moreover, in spite of the acute picture of many leukemias, there is growing evidence that they may result from relatively subtle perturbations of the mechanisms that regulate normal hematopoiesis. A framework for understanding normal hematopoiesis is therefore essential to obtaining new insights into the nature and better management of these diseases.

The concept of a hierarchical model of hematopoiesis Normal adult blood contains large numbers of highly specialized cells that perform critical physiological functions. Because the life-span of most of these cells is relatively short, their replacement and the mechanisms that control new blood cell output are key to survival. Red blood cells (RBCs, also called erythrocytes) are the most numerous cell type in the blood. Their primary role is to transport oxygen from the lung to peripheral tissues and, for this, the concentration of RBCs in the blood needs to be tightly maintained at 5 × 1012 per liter. In the average adult with a 5-liter blood volume, this requires a controlled output of 2 × 1011 new RBCs each day to replace those that have reached the end of their normal 120-day life-span. Each liter of adult blood also contains approximately 7 × 109 white blood cells (WBCs, also called leukocytes). These are responsible for the elimination of bacteria and viruses. Most of the WBCs are either neutrophilic granulocytes (4 × 109 per liter) or different kinds of lymphocytes [B cells, T cells and natural killer (NK) cells, 2 × 109 per liter] with smaller numbers of monocytes, and eosinophilic and basophilic granulocytes making up the remainder. The life-span of human neutrophils in the blood is very short (approximately 1 day), and the number of neutrophils generated each day has been estimated to be approximately 6 × 1010 . Platelets constitute another important blood cell type because of their role in controlling the clotting process. Platelets are produced by fragmentation of megakaryocytes in the bone marrow and are present in the blood of normal adults at a concentration of approximately 300 × 109 per liter. The average life-span of the human platelet is

C Cambridge University Press 2006. Childhood Leukemias, ed. Ching-Hon Pui. Published by Cambridge University Press.


69

70

Connie J. Eaves and Allen C. Eaves

9 days, which dictates a production rate of approximately 25 × 1010 per day (albeit from a much smaller number of megakaryocytes).1 Thus, even with respect to just the myeloid lineages, hematopoiesis in the normal adult provides for a daily output of approximately half a trillion new blood cells of six different lineages. The first insights into how new blood cells are formed came from light microscopic studies of bone marrow preparations. This led to the description of a hierarchy of precursors for each myeloid lineage. The relationship of these morphologically defined cells to successive cell divisions was then deduced from pulse-labeling experiments.2–4 These studies established that the terminal differentiation of erythroid and granulopoietic cells in adults is completed over the course of three to five cell generations that cover a period of 5 to 7 days. At the same time, the number of cells in each of these lineages is progressively amplified, although cell death is used to finetune cell numbers in successive generations. These observations laid the foundation of a model of hematopoiesis in which differentiation is unidirectional and irreversible and leads to a predominance of terminally differentiating cells. The concept of a persisting population of more primitive cells able to produce multiple types of blood cells was first inferred from the observation in 1951 that the bone marrow of patients with myeloproliferative disorders (MPDs) contains elevated numbers of precursors of all lineages, even though only one mature blood cell type is increased in the circulation.5 More definitive evidence of such pluripotent hematopoietic cells was provided a few years later from experiments showing the production in sublethally irradiated mice of lymphoid and myeloid cells carrying the same clonal cytogenetic markers.6 This was accompanied by the related discovery in humans of a consistent unique cytogenetic abnormality, the Philadelphia (Ph) chromosome, in erythroid, granulopoietic and megakaryocytic cells in patients with chronic myeloid leukemia (CML).7,8 Formal evidence that a population of transplantable cells with long-term lympho-myeloid-reconstituting potential is maintained in normal adult human bone marrow was first reported in 1989.9 Together, these findings established the functional identity of pluripotent hematopoietic stem cells; i.e. cells able to produce some undifferentiated progeny that remain competent to produce all blood cell lineages, as well as progeny that have begun to differentiate irreversibly. However, as discussed in more detail below, only in the last 15 years have quantitative assays specific for murine hematopoietic stem cells defined in this way been devised. Analogous assays for human hematopoietic stem cells are still being

refined. In the interim, much confusion has arisen from the common use of the term stem cell to refer to primitive cells that display features that are shared by, but are not exclusive to hematopoietic stem cells (e.g. expression of CD34). Hematopoietic stem cells, in addition to being able to give rise to progenitors of the various blood cell lineages, must have mechanisms for blocking the activation of their latent differentiation potential. These mechanisms allow some of their progeny to remain in the same undifferentiated, but competent state, thereby allowing the production of new cohorts of blood cells to be sustained throughout adult life. Stem cell divisions that produce at least one daughter stem cell are called self-renewal divisions. If both progeny remain as stem cells, or if both begin to differentiate, the division is described as “symmetric.” If only one of the progeny retains its stem cell properties, the division is described as “asymmetric.” The introduction of functional assays that detect the unique developmental profiles of both murine and human hematopoietic cells with more limited proliferative and differentiation abilities also started in the 1960s. These indicated that execution of the full hematopoietic differentiation process can span many cell generations, even before the acquisition of changes that are overt at a morphological level. Accordingly, a much more extensive hierarchy of intermediate, functionally distinguished cell types is now envisaged (Fig. 4.1). The terms precursor and progenitor have been widely adopted to distinguish between intermediate cell types that either already have, or have not yet, reached the point when morphologic features of a particular lineage first appear. It can be seen that the progenitor pool comprises a large and heterogeneous spectrum of cell types. Some of these are transplantable and may play a significant role in the immediate supply of mature blood cells, particularly in the first month of hematologic recovery in myeloablated recipients of hematopoietic transplants. The development of lymphoid cells differs from the development of myeloid cells in several respects. For example, the terminal differentiation of lymphoid cells is often not accompanied by gross morphologic changes. In addition, reproducible procedures for supporting their growth and differentiation in vitro have been more difficult to establish, particularly for cells of human origin. On the other hand, extensive progress has been made in the identification of sequential stages of lymphoid differentiation based on changes in the molecules they display on the cell surface, and in the transcription factors they express. However, this aspect of hematopoiesis is reviewed in detail elsewhere and therefore is not covered further in this chapter.

Anatomy and physiology of hematopoiesis

Myeloblasts

and

Myelocytes

Fig. 4.1 The process of hematopoietic cell differentiation in normal adults viewed as a developmental hierarchy of functionally distinguished cell types. See text for discussion of developmental stages and for abbreviations of cell types. (Reprinted, with permission, from Eaves.10 )

The developmental origin of hematopoiesis During embryogenesis, the first recognizeable blood cells are nucleated RBCs and macrophages that appear in the extraembryonic yolk sac blood islands. In mice, these are derived from primitive erythroid and macrophage progenitors that develop in the proximal regions of the egg cylinder at the primitive streak stage.11 This is followed rapidly by the appearance in the yolk sac of hematopoietic cells with multilineage myeloid potential and an ability to produce definitive RBCs.11–15 Such cells also arise concurrently, but independently, in the developing aorticgonad-mesonephros (AGM) region inside the embryo from primitive mesenchymal cells with both endothelial and hematopoietic potential (hema-angioblasts). Current evidence suggests that the AGM region is the primary source of hematopoietic stem cells with lympho-myeloid differentiation potential and an ability to reconstitute irradiated mice.16 More limited studies of human embryos indicate a similar intraembryonic origin of human hematopoietic stem cells.17 After the circulation has developed and the heart begins to function, the fetal liver becomes the primary hematopoietic organ with subsequent colonization of the developing thymus, spleen and

bone marrow.13,18,19 Throughout adult life, hematopoiesis continues predominantly in the bone marrow. However, reactivation of extramedullary hematopoiesis in the liver and spleen can occur after hyperstimulation with growth factors, during marrow regeneration, or in association with the development of an MPD (e.g. CML). Investigation of the potentialities of the cells produced during the genesis of the hematopoietic system have indicated that the first cells to display the key properties of hematopoietic stem cells (multilineage differentiation potential and an ability to execute many divisions without loss of this potential) are not generated until after the first mature blood cells have already been produced. Thus, the mechanism(s) regulating when the first lineage restriction events begin in a multipotent hematopoietic cell (the selfrenewal decision) and the mechanism(s) regulating how lineage restriction will be effected (the determination decision) can be activated independently in time. This suggests that these mechanisms are likely to be molecularly distinct, as illustrated schematically in Fig. 4.2,20 and serves to introduce the idea of additional complexities in the process by which blood cells are produced in different situations. Many other aspects of hematopoiesis also change during development. These include the particular growth

71

72

Connie J. Eaves and Allen C. Eaves

Embryo MD

perturbations in the phenotype or behavior of leukemic populations,47 since they may be secondary to an increased turnover rate, or to the reactivation of pathways that are more characteristic of fetal rather than adult hematopoietic cells.

Adult MD

1

3

PC

HSC

5

2

HSC Wnt

4

PC 2

Fig. 4.2 Schematic model of the progressive development of primitive hematopoietic cells. During embryogenesis, cells with pluripotent hematopoietic potential arise (step 1) from mesodermal precursors (MD). These pluripotent cells (PC) generate macrophages and/or erythroid cells (step 2) directly but do not have long-term self-maintaining ability. Later, mesodermal precursors generate a different type of pluripotent cell (step 3) that meets the definition of a hematopoietic stem cell (HSC) because it can divide without immediately activating its latent competence for multilineage differentiation. According to the signals these HSCs receive, they thus either give rise to progeny pluripotent cells (step 4) that then rapidly differentiate, or they can self-renew (step 5). Recent evidence suggests that the Wnt--catenin pathway may be involved in regulating the choice between steps 4 and 5.21 The separately timed appearance of cells with hematopoietic and self-renewal ability suggests that the regulation of lineage selection and self-renewal involve distinct molecular mechanisms. (Reprinted, with permission, from Eaves20 )

factors to which cells at comparable stages of differentiation respond, 22–28 the time taken for differentiating hematopoietic cells to transit the cell cycle,29,30 and the choice of certain lineage-specific genes expressed during terminal blood cell differentiation (e.g. globin genes in erythroid cells31,32 and T-cell receptor genes in developing T-cells33 ). Changes also occur in the proportion of primitive cells that are cycling (i.e. that are not in G0 ) and in the phenotype of these cells, as a result of their altered cycling/activation status (as discussed further below).34–41 Hematopoietic stem cells present at different times during development have been shown to differ in their average clonal output of daughter stem cells42–44 and in the ordering of the lineage restriction process they undergo.45,46 All of these variations are relevant to understanding apparent

Functional assays for different hematopoietic cell types Early steps in hematopoietic cell differentiation that precede the onset of terminal maturation events take place in cells that share a similar “undifferentiated” (blast) morphology. In normal adult bone marrow, such cells constitute less than 5% of all the cells present. Identification of the various steps that hematopoietic stem cells undergo before the appearance of overt morphologic changes characteristic of particular lineages has been made possible by the development of a series of functional assays that detect the growth and differentiation properties of hematopoietic cells when they are stimulated to proliferate and differentiate either in vivo or in vitro. Combining the use of retrospective functional end points with a strategy that allows these to be assigned to individual cells in the original test cell suspension, makes it possible to measure progenitor frequencies, often when these are less than 10−5 . This gives such functional assays the extraordinary sensitivity required to detect very rare progenitors and to monitor their numbers and responses to different manipulations in a quantitative fashion. Much of what we now know of the process of normal hematopoiesis has thus been obtained using functional assays that allow cells at different stages of hematopoietic cell differentiation to be discriminated by virtue of the numbers and types of mature progeny they generate, and the speed and conditions under which this is achieved. Critical to the utility of functional assays for detecting primitive hematopoietic cells is the use of reproducible conditions for stimulating the cells being evaluated, both in terms of assessing their differentiation potential (number of lineages represented among the daughter cells produced) and in assessing their proliferative potential (total number of mature cells produced). The use of well-defined, objective end points for these parameters is also an important issue. In vivo assays generally depend on transplanting various numbers of test cells into hosts given partially or severely myeloablative treatments to create conditions that will strongly promote stem/progenitor cell stimulation while decreasing the competition from endogenous stem/progenitor cells. However, it is difficult to completely

Anatomy and physiology of hematopoiesis

eliminate the host stem cells by such treatments without having lethal effects on other tissues. Therefore, deriving definitive conclusions about the regenerative potential of transplanted cells requires the use of a donor that is compatible, but genetically distinguishable from the recipient so that the donor (versus host) origin of the blood cells regenerated can be unequivocally established. In vitro assays typically make use of defined (or at least standardized) conditions to stimulate particular primitive cell types to divide and differentiate in a reproducible manner. They must also support the terminal differentiation of all lineage(s) whose representation is required to establish the potentiality(ies) of the original cell being tested. This may require the adoption of sequential culture conditions that differ significantly. A key feature of both in vivo and in vitro functional assays for various primitive hematopoietic cell types is their ability to measure the frequency in a given test population of individual cells with the properties of interest. One of the attractions of the first in vivo assay developed (the spleen colony assay, see below) was the fact that progenitor frequency data could be obtained simply by counting the number of macroscopic colonies visible on the surface of the spleen.48 However, for measurements of the frequency of most other types of cells with in vivo reconstituting ability, limiting dilution methods have to be used.49,50 The frequency of the initial cell of interest is then calculated using Poisson statistics and the method of maximum likelihood.51 Both direct colony-scoring and limiting-dilution analysis make use of a minimal cell output criterion to detect the input cells of interest. Accordingly, these methods circumvent the problems of inferring input cell numbers from “activities” measured in bulk assays, which can be skewed by changes in the average output potential of the cells being quantified. Both of these methods for quantifying the frequency of a given cell type rely on the validity of two important assumptions: (i) that the readout is linearly related to the number of cells tested over the range used, and (ii) that a single cell is responsible for a positive score. These assumptions have been validated for some populations48,52–58 but remain a potential concern when the test cells have been manipulated or altered genetically. A careful appreciation of the specificity of a functional assay for a particular type of hematopoietic cell is another important issue in interpreting data derived from the use of such procedures. In fact, many of the functional assays described for hematopoietic cells detect populations of cells with different biologic properties. Thus, a change in the total number of cells detected with a particular assay will not necessarily mean that all subsets in the original

test population identified by that assay will have changed in parallel. For example, rapid production of mature blood cells in vivo is a property of many types of primitive hematopoietic cell types, ranging from those with a very short-term lineage-restricted growth potential to those with life-long multilineage repopulating ability. Thus, early blood cell recovery data can provide misleading information about the numbers of stem cells transplanted. Similarly, biologically distinct types of primitive hematopoietic cells can be stimulated to form colonies in vitro under the same growth factor conditions. In both cases, the most reliable parameters for discriminating between subsets of cells at different stages of hematopoietic differentiation have been the minimal number of divisions that occur prior to the appearance of the first mature progeny and the durability of their continued output. Another limitation of functional assays for primitive hematopoietic cells is their dependence on entire programs of differentiation being faithfully completed. For some assays, this may require waiting for a period of many weeks or even months. In addition, cells that are more primitive than those that may be adversely affected can masquerade as being defective or may fail to be detected even though they, themselves, are functionally normal or present in normal numbers. Such a situation could arise in cells with altered genomes, or in cells subjected to certain nonphysiological treatments. Similarly, the loss of a feature required for a given cell type to be detected, but not key to its intrinsic differentiation status or proliferative potential, will preclude its identification. Either of these situations can lead to confusion about the biologic processes targeted by various treatments or mutations. Functional assays have also been usefully applied to investigations of many types of transformed hematopoietic cell populations, particularly where the process of differentiation is minimally altered. Examples of diseases where morphologically normal blood cells are produced include both the MPDs and the myelodysplastic diseases (MDS). In contrast, the block in differentiation characteristic of acute myeloid leukemia (AML), has made the development of analogous functional assays for these cells more difficult. Nevertheless, in AML, other features of the early stages of differentiation appear frequently to be retained. This has allowed the use of both cell surface phenotyping and functional assays to construct a hierarchical organization of cells within the neoplastic clone of many types of human AML.59,60 Such a model is also consistent with the occurrence of the first leukemogenic mutation(s) in a hematopoietic stem cell.61,62 Because of their central importance to the investigation of both normal and leukemic hematopoiesis, the most

73

74

Connie J. Eaves and Allen C. Eaves

commonly used functional assays are reviewed in greater detail below.

5% and 20%67,81,82 and may be influenced by a number of cellular and host parameters (reviewed by Lord83 ).

Competitive repopulating units (CRUs)

Cells with in vivo repopulating activity Colony-forming unit-spleen (CFU-S) The spleen colony assay detects a rare subpopulation of primitive murine hematopoietic cells that are stimulated to generate localized colonies of terminally differentiating cells in the spleens of mice given a myeloablative treatment, or in whom endogenous hematopoiesis is genetically compromised.48,63 If cells with this ability are present in the spleen in sufficiently low numbers, their progeny may, within the first week and a half, be visualized on the surface of the spleen as discrete macroscopic colonies containing a few hundred thousand cells. Most such colonies contain maturing cells of a single lineage64–66 and do not contain cells that can generate new spleen colonies in secondary hosts67 or multilineage colonies in vitro.68 After longer periods, larger spleen colonies containing several million cells may be seen. These typically contain both terminally differentiating granulopoietic and erythroid cells,52 as well as cells that can generate secondary multilineage colonies both in vivo67 and in vitro.68,69 The cells from which all of these spleen colonies arise are referred to operationally as “colony-forming units-spleen,” or CFU-S. However, CFU-S are not a biologically homogeneous population. Careful time-course analyses,68,70 and cell-separation71–74 and biologic characterization studies,75 have established that the CFU-S assay detects a spectrum of early hematopoietic cell types. Some CFU-S are lineage-restricted and have quite limited proliferative potential. These properties account for the small size and rapid, but transient, appearance of the spleen colonies they generate. Other CFU-S have multilineage myeloid differentiation potential, extensive proliferative potential and an independently variable self-renewal potential.52,67,69 A subset of CFU-S can also generate cells able to permanently reconstitute lympho-myelopoiesis in irradiated hosts.76,77 However, the converse is not true; i.e. most cells able to permanently reconstitute lymphomyelopoiesis in irradiated hosts do not have CFU-S activity.78,79 CFU-S appear early in ontogeny and are subsequently maintained throughout adulthood, primarily in the marrow.80 When injected intravenously, only a proportion of these cells end up in the spleen and form macroscopically sized colonies within 2 weeks. This proportion (the seeding fraction) has been estimated to range between

Studies of the in vivo-generated progeny of hematopoietic cells that had acquired unique radiation-induced cytogenetic markers provided the first evidence in mice that large, persisting, multilineage clonal populations of hematopoietic cells could be produced.6,84 Subsequently, similar observations were made using transplants of hematopoietic cells that were genetically marked by unique retroviral inserts.85–87 These and additional serial transplantation experiments88 established the remarkable proliferative potential of normal hematopoietic stem cells but were not useful for quantifying them or for investigating their heterogeneity. The competitive repopulating unit (CRU) assay provides the specificity required for the exclusive quantification of hematopoietic stem cells with life-long blood cellproducing activity.49,89 As shown schematically in Fig. 4.3, this assay simply involves injecting serial dilutions of test cells into congenic hosts and then measuring the fraction of mice in each group with detectable levels of circulating B-cell, T-cell and myeloid cells derived from the transplanted cells after a period of at least 4 months. The recipients are also given a supplemental transplant (or a less myelotoxic treatment) to ensure that they all survive independent of whether they receive any CRUs in the test cells injected. The 4-month interval is used to ensure that the multilineage progeny of cells with limited self-renewal ability will have disappeared.74,90,91 The frequency of CRUs in the test cell suspension can then be calculated with Poisson statistics. Although the contribution of individual CRUs to the mature B-, T- and myeloid cells in the peripheral blood varies greatly at any given time and over time,91,92 it is possible to derive average clone sizes for CRUs from different sources. These average output values can then be used for extensive comparative analyses in studies that would not be practical to undertake with limiting dilution methodologies.93 The CRU assay should not be confused with the longterm competitive transplantation methodology developed by Harrison.94,95 This procedure compares the relative stem cell activity of two cell populations (again using genetic markers to distinguish their progeny) but does not measure their frequency. Additional statistical methods can be applied to the variance data obtained from such experiments to calculate the frequencies of the input stem cell populations. However, the validity of the values thus derived is contingent on the assumption that the average

Anatomy and physiology of hematopoiesis

Test cells

Irradiation

> 16 wks

37% negative mice*

No. of test cells/mouse 2 × 105 compromised cells (2 × serial BMT) or 105 normal BM cells or No cells using sublethally irradiated W41 recipients

*Positive = >0.5% Iymphoid and >0.5% myeloid repopulation by test cells

Fig. 4.3 The mouse CRU assay. This procedure uses the principles of limiting-dilution analysis to measure the frequency of cells in a given suspension that have transplantable long-term repopulating ability and can individually generate both lymphoid and myeloid progeny.89 The treatment of the host is chosen to maximize the sensitivity of the assay by reducing the competing endogenous stem cell population to a minimum and creating an environment in which the engrafting stem cells will be optimally stimulated. This can be achieved by pretreating normal mice with a myeloablative treatment (e.g. a lethal dose of radiation) or by giving c-kit mutant mice (whose stem cells are defective27,63 ) a sublethal treatment. In order for a limiting dilution analysis of the stem cell content of the test cell suspension to be performed, the recipients must be able to survive regardless of whether they receive any stem cells in the test cells injected. Survival of normal recipients is assured by cotransplanting them with hematopoietic cells of the same genotype that contain sufficient numbers of short-term repopulating cells but minimal numbers of long-term repopulating cells. Survival of c-kit mutant hosts is similarly assured by pretreating them with a dose of radiation that allows significant numbers of endogenous cells to survive. The differentiated blood cell progeny of the test cells and the recipients must be genetically distinguishable and assessed at a time when they can safely be assumed to represent the exclusive output of cells with life-long stem cell potential. Strains of mice congenic with the C57B1/6 mouse are typically used to allow the blood cell progeny of the test cells to be uniquely identified by CD45 (Ly5) allotype markers or glucose phosphate isomerase isoform differences. Poisson statistics are then used to calculate the frequency of CRUs in the original test cell suspension from the proportions of mice whose blood does not contain both lymphoid and myeloid cells of donor origin at 4 months post-transplant (or longer). Blood cells present after 4 months are assumed to be derived from stem cells with life-long hematopoietic activity. BM, bone marrow; BMT, bone marrow transplantation.

cell outputs are not different, which is not an invariant feature of different stem cell populations. Cell populations in which at least 40% of the cells can be detected as CRUs in single cell transplant experiments have been isolated by a variety of strategies demonstrating the robustness of this assay.92,96,97 These experiments also indicate that the ultimate marrow seeding efficiency of long-term repopulating stem cells must be almost 100%. Nevertheless, only 10% of injected CRUs can be detected in the bone marrow 24 hours after they have been injected intravenously.98 The full explanation for this apparent discrepancy is not yet known. One possibility is that many CRUs initially locate in other tissues, as has been shown to occur in mice assessed at later times post-transplantation.99–101 Alternatively, CRUs may seed the marrow efficiently, but then rapidly lose this ability while still retaining their full developmental properties. Support for this latter concept is provided by the observation that CRUs acquire a reversible engraftment defect when they are in the S/G2 /M phases of the cell cycle,102,103

in contrast to cells with shorter-term repopulating activities, such as those detectable as CFU-S.104,105 Primitive human hematopoietic cells can also home efficiently to the marrow of various xenogeneic hosts, allowing the characterization of human hematopoietic cells with in vivo repopulating ability. Intravenous injection of human hematopoietic cells into neonatal or sublethally irradiated adult mice with various genetically determined immunodeficiency states (and hence containing greatly reduced numbers of B, T and NK cells) can lead to repopulation of the marrow of these mice with large numbers of human lymphoid and myeloid cells.106–111 Intraperitoneal injections of preimmune fetal sheep produce a similar result.112,113 Both types of experimental models take advantage, not only of the immune tolerance of the hosts, but also of the growth-promoting conditions in the fetal and early postnatal marrow, or induced by a myelosuppressive pretreatment. Prior removal of immunocompetent human cells from the cells to be injected is important to prevent the development of graft-versus-host disease, which can be

75

76

Connie J. Eaves and Allen C. Eaves

fatal.112,114 Nonobese diabetic-scid/scid (NOD/SCID) mice and NOD/SCID-β2microglobulin−/− mice have been the most widely used hosts to date and have the advantage that they breed easily and are efficiently engrafted by specific subsets of human hematopoietic cells,108,115 if these are injected intravenously before the mice reach 10 weeks of age.116 A significant limitation of these particular mice is their rapid development of fatal endogenous thymomas between 6 and 12 months of age, which precludes their use for long-term follow-up studies.114 Hopefully, the alternative use of long-lived Rag2−/− γc−/− mice110,111 or NOD/SCID-nu/nu mice117 may circumvent this problem in the future. In all of these hosts, human cells belonging to multiple hematopoietic lineages are generated, even in the absence of exogenously supplied human growth factors. In addition, human cells capable of repopulating secondary fetal sheep118–120 or secondary immunodeficient mice108,121–123 are produced in the primary hosts. This finding implies that these hosts can be engrafted by human hematopoietic stem cells and that the xenogeneic environment can support the self-renewal of these cells. In sublethally irradiated immunodeficient murine hosts, an early wave of predominantly erythroid cells and some megakaryocytes is seen in the first 2 to 3 weeks. These cells are then rapidly replaced by a larger population of predominantly pre-B cells and cells of the neutrophil and macrophage lineages.108,122,124,125 Species-specific factors are thus also not essential to support some multilineage human hematopoietic cell differentiation. However, growth factor administration can enhance the production of both primitive and mature human cell types, indicating that those produced by the host are likely suboptimal in type and/or amount.121,126–128 In NOD/SCID and NOD/SCID-β2microglobulin−/− mice, the production of mature human erythroid and megakaryocytic cells,129 and mature B cells and T/NK cells122,123,130 is particularly compromised. This may be related in part to the poor survival of human cells in the bloodstream of such mice.126,130 The human cell population regenerated in NOD/SCID mice thus typically contains a disproportionately high number of more primitive (CD34+ ) cells in comparison to primary human hematopoietic tissues. Adult cells appear more sensitive to the mechanisms responsible for the poor terminal differentiation of their progeny than are human cord blood or fetal liver transplants, and the compromised output of differentiated progeny from adult cells can be partially reversed by the administration of pharmacologic doses of human-specific growth factors.126,127

Interestingly, transgenic NOD/SCID mice engineered to produce human interleukin-3 (IL-3), human granulocyte colony-stimulating factor (G-CSF) and human Steel factor have increased numbers of human cells in the blood and enhanced human granulopoiesis in the marrow, with reduced output of primitive cells and more mature human cells of other lineages.131 This suggests that this method of achieving sustained exposure to high levels of these growth factors may lead to premature exhaustion of the graft through a continuous mobilization of primitive cells into the circulation. The time allowed to elapse before assessing the human progeny produced in the mice can be a critical variable when using this end point to infer the differentiated state of the injected cells responsible for their generation.108,123,132,133 This issue is particularly relevant in interpreting data obtained from experiments with immunodeficient murine hosts that cannot be followed for more than 5 months, since studies in the sheep model have suggested that the output of cells from short-term repopulating cells may be significant for 6 to 9 months.119 In addition, the immune status of the host can differentially affect the ability of different subsets of human hematopoietic cells to engraft and/or survive.108 The most immunocompromised mice are engrafted with the greatest spectrum of human repopulating cells, and there is a progressive selection in favor of the most primitive cells as the NK activity of the murine host increases. Thus, only the most primitive cells efficiently engraft NOD/SCID mice and their progeny constitute the majority of the cells present after a few weeks. NOD/SCID mice thus serve as a relatively selective host for the early quantification of the most primitive type(s) of hematopoietic cells in human tissues. In contrast, the NOD/SCID-β2microglobulin−/− mouse is highly permissive for multiple types of human repopulating cells. End points for detecting three distinct subsets of human cells with different short- and long-term repopulating activities in these hosts have been devised (Fig. 4.4). The features used to distinguish these three cell types include differences in their immunophenotypes, differences in their engrafting abilities when passing through S/G2 /M, and differences in their abilities to generate progeny that will repopulate secondary NOD/SCID mice.108,123,132,134 Limiting-dilution approaches, similar to those developed for measuring murine CRUs, are used to measure the frequency of each of these cell types in different sources of human cells.50,108,129,135 Frequency values for normal adult human marrow, mobilized peripheral blood, cord blood and 3- to 4-month gestational fetal liver are given in Table 4.1.

Anatomy and physiology of hematopoiesis

STRC-ML (34 + 38 −)

Human cells/mouse BM

A

Cell cycle-unrestricted engraftment, do not make NOD/SCID repopulating cells

B

NOD/SCID-b2m −/− 107

STRC-M (34 + 38 +)

LTRC-ML (34 + 38 −)

NOD/SCID

105

Cell cycle-restricted engraftment, self-renew 3

6

13

3

Table 4.1 Frequencies of human CRUs (NOD/SCID lympho-myeloid repopulating cells) in various sources of hematopoietic cells

Tissue

106

6

13

77

Fetal liver (12–20 weeks) Cord blood Adult bone marrow G-CSF-mobilized blood

CRU frequency per 105 CD34+ cells (95% CI)

Reference

9 (5–15)

Holyoake et al.129

6 (3–11) 0.8 (0.5–1.3) 0.06 (0.03–0.1)

Holyoake et al.129 Holyoake et al.129 Van der Loo et al.516

Time post-transplant (weeks)

Fig. 4.4 Different types of human cells with repopulating activity in sublethally irradiated immunodeficient mice (reprinted, with permission, from Glimm et al.108 ). (A) Consistently higher levels of human hematopoietic cells present in the marrow of NOD/SCID-β2microglobulin−/− (NOD/SCID-β2m−/− ) mice as compared with NOD/SCID mice after their transplantation with the same number of normal adult human bone marrow cells (depleted of mature cells expressing erythroid, granulopoietic, megakaryopoietic, or lymphoid lineage markers). (B) Explanation for these differences in terms of the differential ability of human cells with short- and long-term repopulating ability (STRC and LTRC) to engraft the two mouse strains. Most of the first human cells produced in the marrow of the more immunodeficient NOD/SCID-β2microglobulin−/− mice are derived from a type of human repopulating cell that is myeloid -restricted and has very short-lived repopulating activity. The progeny of these cells are then superceded by a second cohort of human lymphoid and myeloid cells derived from a cell that is more primitive but still not self-sustaining. Human cells with long-term repopulating ability and self-renewal activity also engraft the NOD/SCIDβ2microglobulin−/− mice but, because of their relatively low numbers, their progeny in the NOD/SCID-β2microglobulin−/− mice are greatly outnumbered by the progeny of the short-term repopulating cells. Neither of the two types of human short-term repopulating cell engrafts NOD/SCID mice efficiently. Therefore, repopulation of NOD/SCID mice can provide a relatively selective early measure of human cells with long-term repopulating activity.

Abbreviations: CRV, competitive repopulating unit; G-CSF, granulocyte colony-stimulating factor; CI, confidence interval.

specific factors required for the growth of the T-ALL cells.148 The pattern of malignant disease that develops in the engrafted mice usually mimics that seen in the patient from whom the malignant cells were taken. In addition, the AML-repopulating cells share many features with normal human CRUs, including a low frequency and a quiescent CD34+ CD38− phenotype.60,149,150 On the other hand, differences between transplantable normal and AML stem cells have also been demonstrated; for example, leukemic stem cells usually do not express detectable Thy-1 in contrast to their normal counterparts.151 Engraftment of mice with cells from patients with MDS152,153 or chronic phase CML154,155 has also been reported. However, the typically low numbers of very primitive neoplastic cells and/or the predominance of residual normal stem cells in diagnostic marrow or blood samples from patients with these diseases has made it very difficult to develop useful xenotransplant models or to characterize the most primitive subsets of malignant cells.

Cells with in vitro hematopoietic activity Long-term culture-initiating cells (LTC-ICs)

Cells from patients with a variety of hematologic disorders have also been used to successfully engraft immunodeficient mice, and in many instances evidence of the production of neoplastic or leukemic human cells has been documented. These xenotransplant experiments include AML,60,136–138 B-lineage acute lymphoid leukemia (ALL),139–141 CML in blast crisis,142,143 myeloma144–146 and chronic lymphocytic leukemia (CLL).147 Successful transplantation of immunodeficient mice with human T-lineage ALL (T-ALL) has required the prior transplantation of other human cells, presumably to produce unknown species-

Another assay for quantifying a very primitive hematopoietic cell population is one that detects a cell that can initiate sustained hematopoiesis when cocultured on stromal feeder layers.54,156,157 The assay for these so-called long-term culture-initiating cells (LTC-ICs) was developed from the observation that mature granulocytes and macrophages can be produced for several months when bone marrow cells are cultured in media containing horse serum and corticosteroids.158–161 Subsequent studies showed that the primary need for serum was to generate a competent feeder layer of stromal cells that

78

Connie J. Eaves and Allen C. Eaves

stimulate the proliferation and differentiation of very primitive hematopoietic cells in the absence of exogenously supplied growth factors.162 To ensure that the latter end point can be quantified independently of the ability of cells in the test suspension to form a competent feeder layer, irradiated pre-established marrow feeders54,156 or irradiated monolayers of a number of human163,164 or murine28,165–168 fibroblast cell lines are used. Many studies have provided evidence of heterogeneity in the phenotypes of different sources of fibroblasts and in their supportive activity,163,164,169–171 but definitive correlations with particular growth factor-producing profiles have not been established. A role of notch-ligands has been suggested,172–175 and the production of specific types of heparan sulfates176 that may help to colocalize primitive hematopoietic cells and growth factors177 may be another feature of some supportive stromal cells. This concept is certainly consistent with the finding that larger numbers of human LTC-ICs, including a more primitive subset, are detected when test cells are cocultured on competent murine fibroblasts that have been engineered to produce ng/ml quantities of human G-CSF, human IL-3, and human Steel factor constitutively.178 Such feeders are, therefore, now routinely used to optimize human LTC-IC detection.179 However, definitive characterization of the mechanisms responsible for the underlying functionality of stromal cells in LTC-IC assays has remained elusive. The principles underlying the LTC-IC assay are similar to those used to develop the in vivo CRU assay. The LTC system, like the irradiated host, supports the proliferation and differentiation of mature cells (granulocytes and macrophages) from progenitors at multiple stages of differentiation. Therefore, sufficient time must be allowed to elapse for all intermediate types of progenitors in the original test suspension to have exhausted their proliferative potential so that the cells ultimately detected can be safely assumed to have been exclusively derived from a very primitive subpopulation. Early studies indicated that this condition was met after an interval of 4 to 5 weeks when the standard in vitro colony-forming cell (CFC) assay (see below) was used to detect the progeny of the input cells.156,157 However, it was subsequently found that simply prolonging this interval to 6 weeks provides significantly greater specificity for very primitive human cells.178,180 LTC-IC frequencies are best determined by limiting-dilution analysis. If, however, the average output of progeny CFCs per LTCIC is known, the total output of CFCs in a bulk culture can simply be divided by this value, assuming first that the number of LTC-ICs used to initiate the assay is linearly related to the number of CFCs they will produce (at the test cell

Table 4.2 Frequencies of human LTC-ICs in various sources of hematopoietic cells

Tissue Fetal liver (12–20 weeks) Cord blood Adult bone marrow G-CSF-mobilized blood Normal adult blood

LTC-IC frequency (± SEM) 760 ± 170 per 106 low-density cells 310 ± 410 per 106 low-density cells 430 ± 280 per 106 low-density cells 270 ± 40 per 106 low-density cells 0.40 ± 0.14 per ml

Reference Pawliuk et al.43 Hogge et al.178 Hogge et al.178 Hogge et al.178 Eaves et al.517

Abbreviations: LTC-IC, long-term culture-initiating cells; G-CSF, granulocyte colony-stimulating factor.

dose tested) and, second, that the total number of LTC-ICs actually assayed is sufficient to accommodate the highly variable CFC output exhibited by individual LTC-ICs.54,55 Both the murine55 and human181–184 LTC-IC assays have been further modified to allow detection of input cells with B and NK lymphoid potential as well as myeloid differentiation potential. Characterization of murine CRUs and LTC-ICs (defined by a 4-week end point) has shown that the cells detected by both of these assays have similar frequencies and properties throughout ontogeny.28,55,157 In addition, some LTCICs generate progeny that are detectable as CRUs.157,185–187 Taken together, these findings indicate a close relationship between the populations detected by both assays. Nevertheless they are not identical, or at least they are not detected at the same efficiency in vivo and in vitro, since a suspension of 40% pure murine CRUs was found to contain only 25% LTC-ICs.92 In addition, the maintenance of mouse bone marrow cells under LTC conditions results in a more rapid loss of cells detectable as CRUs than of cells detectable as LTC-ICs.188 The human LTC-IC assay also detects very primitive hematopoietic cells that appear to overlap with, but are not identical to, CRUs as shown also by immunophenotyping studies and investigations of changes in their numbers under different conditions in vitro.40,50,189–191 Values for LTC-IC frequencies in different normal human tissues are given in Table 4.2. The LTC-IC assay can also detect a rare subpopulation of primitive leukemic progenitor cells present at highly variable frequencies in the marrow and blood of newly diagnosed CML192 and AML patients.193,194 Interestingly, the detection of either CML or AML LTC-ICs is not enhanced

Anatomy and physiology of hematopoiesis

Table 4.3 Frequencies of human CFCs in various sources of hematopoietic cellsa Tissue

CFU-E

BFU-E

CFU-GM

CFU-Mk

BFU-Mk

CFU-GEMM

Fetal liver (12–20 weeks) Cord blood Adult bone marrow Normal adult blood

130 ± 50

2900 ± 800

2500 ± 1400

120 ± 90

21 ± 10

490 ± 320

4±1 4±1

58 ± 10 27 ± 5

120 ± 30 35 ± 6

26 ± 11 35 ± 11

15 ± 8 14 ± 7

8±4 2±1

9±2

340 ± 50

98 ± 12

ND

ND

19 ± 3

Abbreviations: CFU-E/GM/Mk/GEMM, colony-forming unit-erythrocyte/granulocyte-macrophage/ megakaryocyte/granulocyte-erythroid-megakaryocyte-macrophage; BFUMk, burst-forming unitmegakaryocyte. a All values shown are mean numbers ± SEM. For fetal liver, cord blood and adult marrow, these are expressed per 105 low-density cells44 and for normal blood per ml.517

by the use of feeders engineered to produce G-CSF, IL-3, and Steel factor,195,196 presumably because of the autocrine mechanisms that are active in these cells.197,198 Residual normal LTC-ICs are also often prevalent in these samples and the progeny of the leukemic LTC-ICs may not be distinguishable from those produced by the normal LTC-ICs except by genotyping or clonality analyses.195,196,199,200

Colony-forming cells (CFCs) Most, if not all, types of hematopoietic progenitor cells can proliferate and differentiate in liquid suspension culture when provided with appropriate nutrients and growth factors. However, the ability of the most primitive hematopoietic cells (both normal and leukemic) to generate such responses may be compromised if they are immediately suspended in a semisolid matrix.201–204 The acquisition of an ability to proliferate in three-dimensional cultures by later hematopoietic progenitor cell types has thus formed the basis of a variety of clonal assay cultures for their separate detection and quantitation. In most cases, optimization of the sensitivity and specificity of in vitro colony assays has required the identification of culture conditions and endpoints that are unique for colonies from each progenitor type to be assessed. Media supplements to stimulate the production of colonies from all stages of lineage-restricted progenitors on each of the major myeloid pathways as well as some multi-potent progenitor cells have been identified and are now commercially available as standardized reagents. For progenitor quantitation, all components are usually added at concentrations that will saturate the growth requirements of the cells produced throughout the duration of the culture.

These requirements may vary according to the types of colonies being generated, independent of any growth factors that might be endogenously produced. Colony assays have also been used extensively for investigations of the types (or concentrations) of factors that can support, stimulate, enhance or block the proliferation and differentiation of various types of normal and leukemic progenitor cells. The most commonly monitored types of CFCs are described below and their frequencies in different human hematopoietic tissues are given in Table 4.3.

Granulopoietic colony-forming cells (CFU-GMs) Colonies of granulocytes and macrophages were the first to be generated in culture205–207 and remain the easiest to obtain and recognize. The progenitors of such colonies were originally called CFU-Cs (colony-forming units-culture) prior to their characterization in mice as a population distinct from CFU-S.208 When conditions were defined that allowed the proliferation and maturation of other lineages in semisolid culture media, evidence of CFCs with restricted granulopoietic activity was obtained and the term CFU-GM (colony-forming unit-granulocytemacrophage) was widely adopted to refer to these. CFUGMs are stimulated to proliferate and differentiate into colonies of mature granulocytes and/or macrophages by any single or combination of growth factors, including GMCSF, IL-3, G-CSF, IL-6, and Steel factor.209,210 IL-5 selectively stimulates eosinophil progenitors.211,212 Different stages of granulopoietic progenitors can be recognized by their ability to make colonies of different sizes, the ultimate size being fixed by the number of divisions the original progenitor undergoes before

79

80

Connie J. Eaves and Allen C. Eaves

all progeny have reached a stage of maturity at which they can no longer divide. However, on the granulopoietic pathways, the number and rate of late divisions can be influenced by the growth factors to which the cells are exposed.213 Growth factors may also salvage some early granulopoietic cells from apoptosis.214 Nevertheless, progenitors of large and small colonies of granulocytes and/or macrophages generated under standardized culture conditions, have distinct characteristics suggesting that they represent a sequence of cells with decreasing proliferative potential.215,216 In addition, they appear to be regulated differently in vivo. For example, in the marrow of normal adult humans, the CFC-GMs that produce very large colonies of neutrophils and macrophages belong to a very slowly cycling population, whereas those that make smaller colonies appear to be maintained in a state of rapid turnover.217

Erythroid colony-forming cells (CFU-Es, BFU-Es) Erythroid progenitors are detected by their ability to produce colonies of maturing erythroblasts. When these are of murine origin, the erythroblasts produced are identified by their small size and tendency to adhere to one another to form tight clusters of up to approximately 50 tiny cells.218,219 The hemoglobin they contain usually cannot be visualized directly, although it can be readily demonstrated histochemically.219 Terminally differentiating human erythroid cells are somewhat larger and more robust in vitro and the hemoglobin they produce is sufficient to give them a distinctive reddish color in the living state. This allows colonies containing human erythroblasts to be uniquely recognized in cultures that contain colonies of other types of cells, or colonies that contain both erythroid cells and other types of mature cells. Different stages of erythroid progenitor cell development are readily distinguished by the different-sized colonies of erythroblasts that they can generate. The most differentiated erythroid colony-forming cells are called CFU-Es (colony-forming units-erythroid). They display a limited proliferative potential of three to seven divisions resulting in the rapid formation of tight colonies of mature hemoglobinized erythroblasts and are thought to be the immediate precursors of cells recognized morphologically as proerythroblasts. Erythroid progenitors that produce larger colonies are called BFU-Es (burst-forming unitserythroid). This term was coined to reflect the sudden final “burst-like” appearance of terminally differentiating erythroid cells in larger colonies due to the semisynchronous proliferation and differentiation behavior of the cells from which they arise.218,220–222 Primitive and mature

BFU-Es are readily distinguished by the time they take to produce terminally differentiating erythroid cells and display properties that allow their physical and biologic separation as distinct stages of differentiation, including differences in their dependence on stimulation by factors other than erythropoietin (EPO); e.g. Steel factor, IL-3, IL-6, or GM-CSF.69,217,223,224 In the adult, CFU-Es and terminally differentiating erythroid cells require continued stimulation by EPO to stay alive,225 even though they make some erythropoietin themselves.226 The addition of EPO to erythroid colony assays is thus essential to the detection of CFU-Es and BFUEs with a few notable exceptions. These include certain disorders in which a stem cell has acquired a mutation(s) that allow(s) its differentiating erythroid progeny to bypass the normal requirement for EPO stimulation. The best example of this is polycythemia vera (PV),222,227–231 although an abnormal ability of erythropoietic cells to survive and mature in the absence of exogenously supplied EPO is also seen in cells from some patients with CML232,233 and essential thrombocytosis (ET).234–237 Interestingly, the inability of adult erythroid precursors to survive in the absence of stimulation by EPO is also a developmentally acquired property. Thus, the generation of the first primitive RBCs is completely EPO-independent,24 while the definitive erythroid precursors subsequently produced in the fetus are more responsive to low concentrations of EPO than are their adult counterparts.22,23,231

Megakaryopoietic colony-forming cells (CFU-Mks, BFU-Mks) Megakaryocytopoiesis is unique in its termination in a cell that undergoes successive endomitoses to produce megakaryocytes of increasing ploidy. A mature 32n megakaryocyte would thus correspond to an erythroid colony containing 16 erythroblasts. This is why a criterion of only two megakaryocytes is commonly used as the minimum required to identify a megakaryocyte colony. The progenitors of megakaryocyte colonies are referred to as CFU-Mks (colony-forming cells-megakaryocyte) and BFUMks (burst-forming units-megakaryocyte) for the progenitors of larger megakaryocyte colonies (which often resemble erythroid BFU-E-derived colonies and may overlap with them). Murine megakaryocytes can be specifically identified by their content of acetylcholinesterase, which can be stained histochemically.238 This is not the case for human megakaryocytes; however, these cells can be specifically identified by immunohistochemical methods that detect their expression of platelet-specific antigens, such as CD41

Anatomy and physiology of hematopoiesis

(GpIIb/III). Quantitation of human CFC/BFU-Mks thus requires generating the colonies in a matrix such as agarose or collagen that not only supports their optimal growth but also allows the entire culture to be fixed and stained. Human megakaryopoietic progenitors are exquisitely sensitive to the inhibitory effects of TGF- and hence are best cultured under serum-free conditions.239–241 Their growth and differentiation can be stimulated effectively by a variety of soluble factors, including IL-3, IL-6, GM-CSF, Steel factor, and thrombopoietin (TPO).241–243 Different stages of megakaryocyte progenitor development, like those of their granulopoietic and erythroid counterparts, can be defined on the basis of the number of megakaryocytes the progenitor will produce and, hence, the time required for this process to be complete in a given colony.241,244,245

Pluripotent colony-forming cells (CFU-GEMMs) Cells with multilineage hematopoietic differentiation potentialities can also generate colonies of mature progeny in semisolid cultures.246–249 Those that generate colonies containing erythroid cells and megakaryocytes as well as granulocytes and macrophages are referred to as CFUGEMMs (colony-forming cells-granulocyte, erythroid, megakaryocytic, macrophage). CFU-GEMM-derived colonies are often large, indicative of the high proliferative potential that cells with unrestricted myeloid differentiation potential would be expected to possess. For this reason, a very large colony size has been used by some as the sole criterion for identifying a subset of very primitive progenitors referred to as high proliferative potential-CFCs (HPP-CFCs).250,251 HPP-CFCs likely overlap extensively with those identified as CFU-GEMMs, and a shared feature of the assays for both is the addition of multiple growth factors to the assay cultures.249–252 Because of their uncommitted differentiated status, these primitive progenitors will, during their initial divisions, generate progeny that can be detected as CFCs when replated into secondary semisolid culture assays. The types of daughter cells thus detected may include CFU-GEMMs as well as lineage-restricted CFCs.253–255 A subset of CFUGEMMs also show delayed entry into the first cell cycle in vitro,256,257 allowing their separate transient recognition as small colonies of “blasts” that can represent almost pure populations of CFCs.254,258 In mice, the progenitors of a large proportion of these “blast colonies” can, when appropriately stimulated, generate CFU-S,248 NK cells, B- and T-lymphoid progenitors,259,260 and even occasional CRUs,202 as well as progenitors of all of the myeloid lineages, as long as IL-1 and IL-3 are not present in excessive amounts.261

CFCs in acute leukemia Conditions for obtaining colonies of leukemic blasts in semisolid assays of primary blood and marrow samples have also been available for many years.262–265 Leukemic CFCs are typically cultured in semisolid media containing the same growth factors that are used to stimulate their normal counterparts (i.e. Steel factor, GM-CSF, G-CSF, IL-3, IL-6, and more recently, flk2/flt3-ligand and TPO).196,266–269 Although abnormal autocrine mechanisms are clearly activated in the leukemic CFCs of some patients,197,198 this does not abrogate their dependence on added growth factors to achieve optimal growth in low density cultures.53,196 Many of the colonies generated by sources of highly enriched AML blast populations fail to complete normal differentiation programs, providing early confidence of their origin from leukemic progenitors. However, in AML, the initial transforming event leading to the generation of a dominant (but mutant) clone does not necessarily perturb the subsequent ability of the progeny to differentiate normally.61,62,270 Even in some cases of ALL, where the initial transforming event appears to occur in a lympho-myeloid stem cell, colonies of genetically abnormal, but morphologically normal terminally differentiating cells may be detected.271,272 Thus, not all blast cell colonies are leukemic, and some leukemic progenitors can make colonies of normally differentiated cells. Both of these facts have thus made it difficult to use colony assays for the routine assessment and characterization of leukemic progenitor populations. Nevertheless, blast colony assays have been important in establishing the concept of a progenitor hierarchy within human leukemic populations and have begun to allow their properties and abnormal regulation to be characterized.59,194

Use of phenotypic markers to describe different stages of hematopoietic cell differentiation The power of phenotype analysis The retrospective and often lengthy nature of functional assays for detecting primitive hematopoietic cells has given great impetus to the search for more immediate methods to quantify hematopoietic cell populations, particularly in cases where the cells of interest cannot be distinguished by standard morphologic criteria or are too rare for this approach to be useful. Multiparameter flow cytometry offers great power in this regard because it allows cells to be distinguished on the basis of molecularly determined features in an objective and quantitative

81

82

Connie J. Eaves and Allen C. Eaves

fashion, with a high degree of specificity and reproducibility, and at high resolution. In addition, this technology can be applied in a sterile manner to live cells and used not only for their analysis, but also for their separation into viable subsets defined by the analysis. These isolated cells can then be assayed for their functional attributes. In this way, the phenotype of different functionally defined cell populations can be identified. In the last few years, as reagents for detecting an increasing number of markers have become available, significant progress has been made in the description of a hierarchy of primitive hematopoietic cell types in terms of changes in their phenotype. The most commonly analyzed markers are cell surface antigens against which specifically reactive monoclonal antibodies have been made. These antibodies can then be labeled either directly or indirectly (via a secondary antibody) with a unique fluorochrome and used to distinguish cells as positive or negative on the basis of their acquired fluorescence. In general, marker antigens have been grouped into two categories. The socalled lineage markers refer to cell surface antigens whose expression on hematopoietic cells is generally confined to particular lineages of terminally differentiating cells. Common examples are: CD3, CD4 and CD8 for T-lineage cells, B220/CD45RA, CD19 and CD20 for B-lineage cells, CD56 for NK-lineage cells, CD13, CD14, CD15, CD66b and CD11b for GM-lineage cells, glycophorin A (Ter119) for erythroid cells and CD41 and CD61 for MK-lineage cells. The use of flow cytometry to monitor cells positive for these markers has greatly facilitated studies of their differentiation under a variety of conditions both in vitro and in vivo. In addition, this approach has provided evidence of some programmatic differentiation in leukemic blasts that do not undergo the additional morphologic changes characteristic of normal terminally differentiating cells. The concept of lineage-specific markers is also widely exploited for obtaining populations of hematopoietic cells that are enriched in the more primitive elements that have not yet begun to express any of these markers.50,167,273 Forward and side (90◦ ) light-scattering characteristics (related to cell size and granularity, respectively) are additional parameters that have been useful in the characterization and separation of subpopulations of hematopoietic cells. Staining with various fluorescent dyes allows primitive hematopoietic cells to be distinguished on the basis of their unique ABC transporter activities.41,77,274–276 Similarly, their differential expression of aldehyde dehydrogenase can be detected by staining with fluorescent substrates for this enzyme.277,278 Analysis of their replication kinetics has been detected by quantitative losses in fluorescence after staining with stable membrane dyes such as

PKH 26279 or stable cytoplasmic dyes such as carboxyfluorescein diacetate succinimidyl ester (CFSE).280,281 Figure 4.5 depicts a hierarchy of phenotypes from adult mouse bone marrow that have been cross-matched with functional (and transcriptional) attributes.282,283 A picture of sequentially changing phenotypes with progressive stages of primitive human hematopoietic cell differentiation is also emerging,108,167,178,284,285 but equivalent functional homogeneity of the phenotypes defined thus far remains to be achieved. Although these studies do suggest that some changes in immunophenotype are shared during hematopoietic cell differentiation in the two species, important differences have also been revealed. Three notable examples discussed in detail below are CD34, CD38 and Flk2/Flt3. Another example is the “side population” (SP) phenotype revealed after staining cells with the lipophilic dye Hoechst 33342.286

Phenotype instability and the limitations of phenotype analysis CD34, an L-selectin-binding sialomucin, is expressed on the surface of a small subpopulation ( neutrophil maturation, monocytic cells, megakaryoblasts, early erythroblasts

80% of AML, 30% of My+ALL

FC

40 and 116 kDa transmembrane GP,

Lympho/hematopoietic stem cells

ALL, AML, LBL, vascular tumors

FC, IHC b

sialomucin, two forms with one having a truncated cytoplasmic domain, stromal cell adhesion

and progenitors, small-vessel

Platelet gpIV, collagen,

Platelets, megakaryocytes, erythroblasts and RBCs, monocytes, macrophages

M4, M5, M6, and M7 AML

FC

Megakaryocytes, platelets

M7AML, Glanzmann's

FC

P.falciparum

receptor recognition and phagocytosis of apoptotic cells, platelet adhesion and aggregation CD41

gpIIb(asubunit)ofCD41/CD61

endothelium

(gpIIb/gpIIIa) complex, composed of gpllb-a and gpllb-g subunits, platelet aggregation, receptor for fibronectin, fibrinogen, and von Willebrand factor

thromboasthenia

CD42b

Platelet gplba, forms heterodimer with [3 chain (CD42c) and complexes with gpIX/ (CD42a), von Willebrand factor-ristocetin receptor

Megakaryocytes, platelets

M7 AML, Bernard-Soulier syndrome

FC

CD45

180 to 220 kDa GP, leukocyte common antigen (LCA), tyrosine phosphatase, different isoforms characteristic of different subsets of hematopoietic cells, critical for lymphocyte activation

All leukocytes, weakly expressed by

ALL, AML, lymphoma

FC, IHC

very early erythroblasts

CD45RO 180 kDa glycoprotein, CD45 isoform

Thymic and mature T cells, monocytes, neutrophils

T-cell leukemia and lymphoma, M5 AML

FC, IHC

CD56

140 kDa isoform on NK cells, cytotoxic T cells, subset of CD4+ T cells, neural tissues

T- and NK-cell

FC, IHC

175 to 220 kDa transmembrane GP, neural adhesion molecule (N-CAM), many isoforms

leukemia/lymphoma, some t(8;21) + and t(15;17)+ AMLs, M4 and M5 AML, neuroendocrine tumors

Immunophenotyping

153

Table 7.1 (cont.) Marker

Description/function

Normal cell expression

Associated disease states

Method of detection

CD61

gpIIIa([3subunit)ofCD41/CD61 (gpIIb/gpIIIa) and CD51/CD61 complexes, adhesion to diverse matrix proteins

Megakaryocytes and platelets with

M7 AML, Glanzmann's

FC, IHC

CD41/CD61, monocytes c

thromboasthenia

CD64

72 kDa GP, Fc-y RI, endocytosis of IgG antigen complexes, phagocytosis

Monocytes, histiocytes, neutrophils, early myeloid and monocytic precursors, dendritic cells

Myeloid and monocytic AML, neutrophil expression increases in infection

FC

CD65

Carbohydrate carried by lipid and maybe protein, poly-N-acetyl-lactosamine, unknown function

Myeloid and monocytic cells

Myeloid and monocytic AML,

FC

90 kDa GPI-linked GP, member of CEA antigen family, regulator of adhesion activity?

Granulocytes, epithelial cells

Subset of Precursor-BALL

FC

110 kDa transmembrane GP, primarily in

Monocytes, macrophages, mast cells, neutrophils, basophils, dendritic cells, subset of myeloid progenitors, activated T cells, subset of B cells, osteoclasts

M4 and M5 AML, subset of

FC, IHC

CD66c

CD68

cytoplasmic granules, endocytosis, lysosomal trafficking

some My+ALL

precursor-TALL/LBL

CD79a

mb-1 gene product Iga, associates with CD79b, signal transduction for surface immunoglobulin

Precursor and mature B cells

Precursor and mature B-cell leukemia and lymphomas some T-ALLandAML

FC, IHC

CD79b

B29 gene product Igg, associates with CD79a, signal transduction for surface immunoglobulin

Precursor and mature B cells

Mature B-cell leukemia and

FC

43 kDa transmembrane GP

Nonfollicular dendritic cells,

Dendritic leukemia, Langerhans

Langerhans cells

histiocytosis AML, rare T-precursor ALL, mastocytosis

FC, IHC

Precursor-B ALL, rare precursor-T

FC

CD83

lymphoma

CD117

143 kDa transmembrane GP, tyrosine kinase c-kit, stem cell factor receptor, a growth factor receptor; cell adhesion?

Hematopoietic stem and progenitor cells, mast cells

CD179a

16 to 18 kDa polypeptide, VpreB surrogate light-chain component disulfide-linked to Ig(x to form pre-BCR, transduces signals for B-cell differentiation and proliferation

Early pre-B and pre-B cells

lambda 5/14.1 surrogate light-chain component disulfide-linked to Ig(x to form pre-BCR, transduces signals for B-cell differentiation and proliferation

Early pre-B and pre-B cells

Sialoglycoprotein glycophorin A,

Erythroid cells

CD179b

CD235a

FC

ALL and AML

Precursor-B ALL, rare precursor-T

FC

ALL and AML

M6 AML, some M7 AML

FC, IHC

proposed functions include minimizing RBC aggregation and inhibition of lysis Abbreviations: GP, glycoprotein; FC, flow cytometry; IHC, immunohistochemistry; NK, natural killer; TCR, T-cell receptor; MHC, major histocompatibility complex; Th, helper T cells; Tc, cytotoxic T cells; CLL, chronic lymphocytic leukemia; LBL, lymphoblastic lymphoma; FCC, follicular center cell; BCR, B-cell receptor; MDS, myelodysplastic syndrome; LPS; lipopolysaccharide, My+ALL, myeloid antigen-positive AML; R-S, Reed-Sternberg cell; LGLL; large granular lymphocyte leukemia; CFU, colony-forming unit; GEMM, granulocyte/erythoid/myeloid/monocytic; GM, granulocyte/monocytic; G, granulocytic; BFU-E, burst forming unit-erythroid; EBV, Epstein-Barr virus; HIV, h u m a n immunodeficiency virus; HTLV-1, h u m a n T-cell lymphotrophic virus; ATLL, adult T-cell lymphoma/leukemia; Ig, immunoglobulin; GPI, glycosylphosphatidylinisotol. a b c

CD20 antibodies differ for FC and IHC. Different antibodies optimal for blast cells and endothelium. Nonspecific binding to monocytes.

154

Fred G. Behm

Early pre-B (pro-B)

Pre-B

Late pre-B (transitional)

Mature-B (naïve-B)

Fig. 7.1 Schematic of normal stages of B-cell maturation. Numbers on cell surface refer to CD (cluster of differentiation) antigens. The bars below the cell diagram represent the expression of proteins associated with the pre-B-cell receptor (pre-BCR) and the B-cell receptor (BCR). A lymphoid progenitor (not shown) expressing CD34, CD10, and weak CD19 precedes the first identifiable cell committed to B-lineage differentiation. Pseudo-light chains (\J/ LC) CD179a and CD197b first appear on the surface of early pre-B cells (pro-B). The early pre-B cell has rearranged immunoglobulin heavy-chain (HR) and germline immunoglobulin light-chain (LG) genes. The early pre-B cell expresses terminal deoxynucleotidyl transferase (TDT), \\i LC, CD 10, CD 19, CD24, and CD34. The pre-B stage is heralded by the appearance of cytoplasmic |x immunoglobulin (Ig) heavy chains and the movement of CD22 from the cytoplasm to the cell surface. The late pre-B (transitional) cell displays surface pre-BCR that comprises Ig|x complexed to t|< LC and noncovalently bound to CD79a and CD79b (|x/\J/LC/CD79). Rearrangement of Ig light-chain (LR) genes precedes the synthesis of light chains \ and K. The |x heavy chain is bound by disulfide bonds to either \ or K light chains to form IgM. The mature or nai've-B-cell stage is initiated with the movement of IgM to the cell surface. The mature B cell loses TDT and fully expresses the BCR, a complex of IgM/CD79a/CD79b.

T-cell differentiation and maturation and have provided a framework for immunologic classifications of leukemias and lymphomas. The newest classifications of hematopoietic and lymphoid malignancies are based on immunophenotyping, cytogenetic, and molecular genetic studies.2 Immunophenotyping serves to establish or confirm the diagnosis of hematopoietic and lymphoid malignancies. However, its usefulness in predicting treatment response has largely been replaced by the prognostic importance of cytogenetics and molecular genetic studies. B-lineage cells The maturation of bone marrow progenitor cells to mature B lymphocytes proceeds through stages that can be iden-

Pre-BCR complex (lguA|/LC CD79)

BCR complex (IgM CD79)

Fig. 7.2 Schematics of the pre-B-cell receptor (pre-BCR) and B-cell receptor (BCR) complexes. The pre-BCR consists of pseudo-light chains (t|< LC) VpreB and \ 5 (CD179a and CD179b, respectively) bound by disulfide bonds to immunoglobulin (Ig) |x heavy chains. This Ig|x/\J/ LC receptor is noncovalently bound to CD79a and CD79b. The pre-BCR first appears on the surface of late pre-B (transitional) cells and persists into the early stage of mature B-cell maturation. Mature or naive B cells express IgM, the BCR, which is composed of immunoglobulin |x heavy and light chains (either A. or K). CD79a and CD79b serve as the signaling molecules for the BCR. Binding of antigen by IgM results in tyrosyl phosphorylation of the cytoplasmic tails of the CD79 molecules, leading to activation of cellular signaling pathways.

tified by the pattern of cellular Ig protein expression (Fig. 7.1). The earliest bone marrow cells committed to B-lineage development have rearrangements of the diversity and joining (DJ) regions of the Igheavy-chain gene but do not synthesize immunoglobulin proteins. These proB or early-pre-B cells express surface CD 10, CD 19, CD24, CD34, CD45 (commonleukocyte antigen), HLA-DR, cytoplasmic CD22, CD79a, CD79b, and nuclear TDT.3~6 The expression of CD45 is weak initially, but increases with cell maturation.7 The pseudo-light-chain components of the pre-B-cellreceptor,CD179a(preV5) andCD179b(X5),make their debut in the cytoplasm of pro-B cells.8 Functional rearrangement of the Ig heavy-chain gene heralds the appearance of (j, heavy chains in the cytoplasm and promotes the differentiation of pro-B cells to the preB stage (Fig. 7.2).8 Young pre-B lymphocytes are the first cells to express surface CD22, followed by the appearance of CD20. During the late stages of pre-B-cell maturation (transitionalpre-B phase) lymphocytes weakly express surface (j, heavy chains but not K or X Ig light chains. Instead, (j, chains are transported to the cell surface in the company of the noncovalently linked pseudo-light chains CD179a and CD179b (Fig. 7.1).8-12 The CD179a/CD179b/Ig|j, complex

Immunophenotyping

Extra-follicular

Lymphoid follicle and germinal center

na

Progenitor Bcell

Centrocyte CD19 CD20 CD23 CD10 Centroblast IgG

Fig. 7.3 Diagram of B-cell maturation in the bone marrow with further differentiation after the cells have transited to peripheral lymphoid tissues via the circulatory system. Naive B cells leave the bone marrow and upon binding antigen undergo blastogenesis with the production of plasma cells and memory B cells in either extra-follicular or follicular lymphoid tissues. In the lymphoid follicle, B cells pass through several stages of differentiation that result in immunoglobulin isotype switches and a selection process that yields cells capable of producing high-affinity immunoglobulins. A partial list of B-cell-associated antigens by their CD (cluster of differentiation) names is given for the B cells portrayed.

is referred to as the pre-B-cell receptor (pre-BCR) (Fig. 7.2). CD 179a and CD 179b are encoded by the genes on chromosome 22 that also carries the X lglight-chain gene. Although these genes share partial homology with the variable and constant regions, respectively, of the X light chain, they do not undergo rearrangement. Disulfide-linked Iga (CD79a) and Ig(J (CD79b) molecules are noncovalently associated with the pre-BCR in a manner analogous to their association with IgM B-cell receptor (BCR). Expression of the pre-BCR results in termination of further rearrangement of the heavy-chain gene and leads to the induction of active pre-B-cell proliferation. Cells that fail to generate a preBCR undergo apoptosis.8, 13, 14 As thepre-B stage ends, TDT, CD34, and CD 10 disappear. Rearrangement and transcription of Ig light-chain genes leads to the formation of complete Ig molecules and to the maturation of transitional pre-B cells to mature B lymphocytes. The mature B cell is defined by the expression of surface IgM BCR (Fig. 7.2). For a brief time, the young mature B cells may coexpress the pre-BCRs and IgM BCRs.15 The BCR is a multiprotein structure comprising an antigenbinding membrane Ig molecule, that consists of two heavy and two light Ig chains linked by disulfide bonds, noncovalently associated with signal transducing heterodimers

CD79a and CD79b. The short three-amino-acid cytoplasmic tail of the JL heavy chain has no intrinsic signaling capacity. Instead, the antigen-stimulated IgM molecules induces conformational changes in CD79a/CD79b heterodimers that serve as signal transduction molecules via their cytoplasmic domains. The cytoplasmic domains subsequently induce signal transduction by binding to the Src family kinases Lyn and Fyn, protein tyrosine kinase SYK, and other phosphoproteins.10,13,16 B cells unable to produce functional IgM BCRs undergo apoptosis in the bone marrow. The expression of surface IgM is accompanied by an increased CD20 density and appearance of CD21, a receptor for C3d and Epstein-Barr virus. The product of bone marrow lymphopoiesis is an immune-competent B cell commonly referred to as a 'naive' or 'virgin' B cell to distinguish it from mature B cells stimulated by antigen. Naive B cells migrate from the bone marrow to colonize the follicular regions or B-cellzones of lymphnodes, spleen, Peyer's patches, tonsilar and other secondary lymphoid tissues. Germinal centers develop in the lymphoid follicles in response to proliferation of antigen-primed B cells. In the germinal centers, BCR diversification results from somatichypermutationoftheIggenes (Fig.7.3).17,18 Insecondary lymphoid tissues, further rearrangements of the Ig

155

156

Fred G. Behm

Cortex

Medulla TCRαβ

CD34 CD7 HLA-DR CD33 (CD2)

Pro-T

CD34 CD7 HLA-DR cyCD3 CD1a CD2

Pre-T1

CD7 cyCD3 CD1a CD2 CD5 CD4

Pre-T2

Pre-T3

DP

SP

Fig. 7.4 Schematic of T-cell maturation in the cortical and medullary regions of the thymus. Expression of T-cell-associated antigens are listed as CD groups. The lower panel shows expected sequential expressions of T-cell receptor gene (TCR) transcripts: pre-Tαa (pTαa), pre-TCRαb (pTαb), TCRδ, TCRγ, TCRβ, TCRα. Nuclear expression of terminal deoxynucleotidyl transferase (TDT) is indicated in the cell nuclei. Bone marrow derived pro-T cells migrate to the subcapsular region of thymic lobules. Under the influence of thymic chemokines and thymic stromal and epithelial cells, pro-T cells are induced to begin T-lineage differentiation and maturation. The first recognizable T cell, pre-T1, expresses cytoplasmic CD3 and TCR8 transcripts. Pre-T cells transition first through the cortex and then into medullary regions of the lobule before leaving the thymus as mature or naive T cells. Five maturational stages are represented: pre-T1, pre-T2, pre-T3, CD4/CD8 double-positive (DP), and CD4 or CD8 single-positive (SP). Some studies support two additional stages of maturation: late pre-T3 and pre-DP, during which surface CD3 is temporarily lost or internalized.

T-lineage cells

thymus.19"22 The T-cellprecursor is abone marrow-derived cell that expresses CD34, CD7, CD33, CD45RA, HLA-DR, possibly CD2, and little or no CD117 and CD90.20-26 The blood-borne progenitor cells arrive in the outer cortical layer of thymic lobules, move through the cortex to the cortico-medullary junction, and take up residence in the medulla (Fig. 7.4). During this journey a variety of chemokines and stromal cell chemokine receptors are encountered that induce differentiation and proliferation.27"29 The subcapsular progenitor cells or proT cells express CD34, CD7, and HLA-DR, but not CD1a or CD3. Pro-T cells still have their T-cell receptor 8,7, a and (3 genes (TCRG, TCRD TCRA, and TCRB) in germline configuration and retain the capacity to generate T cells, NK cells, dendritic cells, and myeloid elements. 24,25,28

The commitment of bone marrow and fetal liver-derived progenitor cells to T-cell lineage takes place in the

The earliest cells committed to T-cell development, pre-T1 cells, are found in the outer cortical areas of the thymic lobules. They are identified by their expression

heavy-chain gene can result in production of IgD, IgG, or IgA. Germinal B cells (centroblasts and centrocytes) share some similarities with pre-B cells, in that they express CD19, CD22, and CD10 and have a high propensity for apoptosis. However, unlike pre-B cells, germinal B cells do not express TDT or CD34. Centrocytes not induced to die by apoptosis leave the germinal center to become memory B cells or form plasmablasts that home to the bone marrow or medullary region of lymphoid tissues with subsequent maturation to plasma cells. Plasma cells express CD38, CD79a, CD 138, and can secrete large amounts of Ig, but do not display surface Ig, CD20, CD22, or HLA-DR. A subset of plasma cells may weakly express CD 19 and CD45.

Immunophenotyping

Pre-TCR complex pTαβ

TCR complex a p

Fig. 7.5 Pre-T-cell receptor (pre-TCR) and T-cell receptor (TCR) with associated CD3 transmembrane proteins. The pre-TCR complex a a b b consists of TCR(3 and two other proteins, pre-Ta (pTa ) and pre-Ta (pTa ). The pre-TCR and TCR molecules require CD3 for cell signaling. This antigen is a complex of 8, -y, and two ε chains. Two £ chains are also part of, but not unique to, the CD3 complex. Signal transduction by pre-TCR and TCR depends on the cytoplasmic immunoreceptor tyrosine-based activation motif (ITAM) of the CD3 proteins. After binding of the ligand by the TCR, phosphorylation of tyrosines of ITAMs creates binding sites for downstream signaling molecules and initiation of signaling events. of CD34, CD7, CD1a, CD2, cytoplasmic CD3, and TDT, but not CD4 or CD8 (double-negative cells because of the absence of CD4 and CD8 expression) (Fig. 7.4). TCRG and TCRD genes are rearranged at this stage and partial D-/p rearrangement of TCRB also takes place. PreT1 cells lose the potential to produce NK, dendritic, and myeloid cells. The pre-T1 stage is followed by sequential expression of CD4 and CD8aa, identifying pre-T2 and early pre-T3 stages, respectively. During the pre-T3 stage TCR(3 proteins appear in the cytoplasm with productive rearrangement of TCRB genes. The (3 proteins bind to surrogate pre-Ta proteins to form pre-TCRB receptors that are transported to the cell surface in the company of CD3 (Fig. 7.5). Expression of the pre-TCRp results in further pre-T-cell expansion and differentiation to a common thymocyte stage characterized by CD7/CD5/CD2/CD4/ CD8a(3-positive cells (double-positive cells because of the coexpressionof CD4 and CD8). These double-positive cells display low levels of CD3 (or CD3l° cells) and may also express CD10 and CD21. 2 7 , 3 0 , 3 1 The expression of pre-TCRp receptors induces rearrangements at the TCRA locus. 2 7 , 2 8 In the pre-T4 stage, productive rearrangement of TCRA results in the pro-

duction of TCRa molecules that in turn combine with TCR(3 proteins to form TCRa(3 (Fig. 7.5). The TCRa(3/CD3 complex moves to the surface of double-positive T cells. With the appearance of TCRapJ, cell surface CD3 intensity increases, resulting in CD3 hi cells, while expressions of CD1, CD10, CD21, and TDT rapidly diminish. Most of the TCRa(3 double-positive cells die by apoptosis as a result of the TCR receptors not being able to recognize self-peptideMCH complexes on thymic epithelial cells (death by failing positive selection). Surviving medullary CD3+, CD4+ and CD3+, CD8+ cells are releasedfrom the thymus into circulatory system to home to T-cell zones of peripheral lymphoid tissues.

Myeloid and monocytic lineages Myeloid and monocytic lineages exhibit many antigens in common (Fig. 7.6). The earliest myeloblasts weakly express CD34 and HLA-DR, but these are lost before the promyelocyte stage.32 CD 13 first appears on the surface of committed granulocyte-monocyteCD34-positive progenitors, increases with granulocyte maturation, and decreases

157

158

Fred G. Behm

Myeloblast

locyte

Myelocyte

Neutrophil G^J

HLA-DR CD33 CD13± CD4 CD15 CD14

Fig. 7.6 Schematic of myelocytic and monocytic maturation. Early progenitor cells committed to myeloid and monocytic development (CFU-GM) express CD33, CD34, CD117, HLA-DR and weak (±) CD13. At the myeloblast and monoblast stages, CD34 and CD117 disappear and CD13 intensity increases. Cells in the mid-stages of myelocytic and monocytic maturation show the expression of CD15. CD65 and CD66 also appear in the late myelocyte stage. HLA-DR is lost early in myelocytic differentiation but persists with monocytic maturation. Strong expression of CD 14 distinguishes the mature monocyte from myelocytic lineage cells. CD4 is weakly expressed throughout monocytic development.

Fig. 7.7 Comparison of myeloperoxidase (MPO), elastase, and lactoferrin expression with myelocytic lineage maturation. The MPO messenger RNA (mRNA) and the proenzyme form of MPO (proMPO) are present in the earliest stages of myelocytic development. Small quantities of enzymatically active MPO are present in early myeloblasts and increase with myelocytic maturation. Primary granules contain MPO and elastase. Lactoferrin appears in early myelocytes and is packaged in specific or secondary granules.

The CD33 antigen is expressed by stem cells that give rise to mixed hematopoietic colonies with in vitro assays. The expression of CD33 decreases slightly with myeloid mat5 33 uration, but is evident at all stages of monocytic develwith monocytic development.. , CD 13 is also present opment. CD64 or Fc-/RI appears at the CFU-GM stage in secretory granules. Granulocyte activation or apoptoand persists through granulocyte and monocyte maturasis results in release of storage CD13 and increased cell tion, but its intensity is greater on monocytic cells. CD65, surface expression. Density-gradient separations and RBC which is structurally similar to CD 15 and shares its cellular lysis techniques used in immunophenotype testing can distribution, is first detected on granulocytic and monoinduce release of granule-based CD13 and enhance surcytic precursor cells that have lost CD34. CD65 density face CD13 density. increases with maturation to neutrophils and monocytes. Glycoprotein CD14 is expressed strongly on monocytes CD36 (glycoprotein IV) is found on immature and mature and macrophages but only weakly, if at all, on immature monocytes but not the granulocytic series. Two intracellumonocytic cells. Neutrophils also produce small amounts lar molecules, myeloperoxidase and lactoferrin, are useful of CD14. Glycosyl phosphatidylinositol (GPI) links CD14 markers for assessing myeloid differentiation. Myeloperoxto the cell surface of monocytes. Monocytes of patients idase is a component of primary or azurophilic granules; with paroxysmal nocturnal hemaglobulinuria lack funclactoferrin is an enzyme contained in secondary or spetional GPI and hence CD14. cific granules. Late myeloblasts and promyelocytes contain Molecules of CD 15 are saccharide antigens having a myeloperoxidase but not lactoferrin, whereas myelocytes, common terminal pentasaccharide, lacto-N-fucopentose metamyelocytes, and neutrophils produce myeloperoxiIII or the Lewis antigen. The CD 15 family also includes a siadase plus lactoferrin (Fig. 7.7).34 lylated (CD15s) form. Some bone marrow CD34+, CD117+ progenitor cells express CD15 or CD15s. CD15 expression increases with monocyte and granulocyte maturation. Similar to CD 13, CD 15 is present inbothprimary and secondary Megakaryocytic and erythroid lineages granules, and with granulocyte or monocyte stimulation increases in surface expression. Neutrophil CD 15 expres- Mature megakaryocytes express platelet-associated antision is reduced in patients or stem cell donors receiving gens CD9, CD36, CD41, CD42, CD61, and Factor VIII.35,36 G-CSF. Less is known about the surface antigen pattern of

Immunophenotyping

Megakaryoblast

MicromegaK3ryocyt6

Megakaryocyte ^ __^^^

Platelets •

- v

•.••;

CD41/61 CD36 CD42 Factor VIM

CD41/61,CD36 CD42, Factor VIII

RBC

CD34 CD117 CD33 HLA-DR CD41±

CD33± CD71 CD36 CD235a±

CD71 CD36 CD235 Hgb

CD71 CD36 CD235

Hgb

CD71 CD36 CD235

Hgb

CD36 CD235

Hgb

Fig. 7.8 Schematic of megakaryocytic and erythroid maturation. A common progenitor cell (CFU-EM) expresses CD33, CD34, CD117, HLA-DR, and weak (±) CD41. CFU-EM progenitors give rise to megakaryocytic and erythroid lineages. Megakaryocytic maturation is accompanied by increased CD41a expression and sequential appearance of CD36 and CD42b. Factor VIII antigen appears relatively late in megakaryocytic development. The earliest identifiable erythroid precursor, the proerythroblast, expresses CD36 and low amounts of CD235a (glycophorin A). Hemoglobin (Hgb) appears relatively late in erythroid maturation.

megakaryoblasts and their immediate parent cells. Studies of megakaryoblastic leukemia cells and normal marrow cell cultures support an orderly appearance and increasing density of platelet antigens with megakaryocytic maturation (Fig. 7.8) .36,37 Young megakaryoblasts have surface CD4, CD33, CD34, and CD117.38-40 The CD41a (glycoprotein IIb/IIIa complex) and CD61 (glycoprotein IIIa) molecules appear early in megakaryocytic development, followed by the expression of CD42b (glycoprotein lb) and CD36 (glycoprotein IV). Factor VIII-related proteins appear relatively late, being fully developed on recognizable megakaryocytes. Relatively few cell surface antigens are unique to early erythroid precursor cells. Immunophenotypic descriptions of erythropoiesis are derived from studies of cultured normal erythroid precursors and leukemic cell lines and, thus, may not accurately reflect in vivo states.41"44 CD34 is expressed only by the earliest erythroid progenitor cells. The transferrin receptor (CD71) is present on all stages of erythroblastic development and on reticulocytes, but is lost with formation of the mature red blood cells (RBCs) (Fig. 7.8). Proerythroblasts, and basophilic, polychromatic, and orthochromatic normoblasts express

CD36. Glycophorin A (CD235a) appears during the late proerythroblast stage and shows its greatest intensity on the mature RBC. Hemoglobin is a specific marker for erythroid lineage but is detectable only after the basophilic normoblast stage.

Methods used in immunophenotyping Immunohistochemistry

Immunohisto chemical techniques use combined immunologic and chemical reactions to reveal surface and intracellular antigens. A primary antibody to the antigen in question is recognized by a secondary antibody conjugated to horseradish peroxidase or alkaline phosphatase (Fig. 7.9). Alternatively, advantage is taken of the strong binding of avidin substrates to biotin. Secondary antibodies are biotinylated and subsequently detected by labeling with streptavidin conjugated to peroxidase of phosphatase. The enzymes react with diaminobenzidine/H2O2 or an alkaline phosphatase substrate to produce a color reaction. The major advantages of immunohisto chemical

159

160

Fred G. Behm

Goat anti-mouse conjugated with HRP

Goat anti-mouse conjugated with biotin

Avidin

nated this problem. The clinical laboratory now has a wide selection of commercially available antibodies to facilitate the diagnosis of hematopoietic and lymphoid neoplasms. Antigens easily detected by immunohistochemistry and useful in the differential diagnosis of pediatric hematologic and lymphoid malignancies are noted in Table 7.1.

Biotin HRP

Flow cytometry

Fig. 7.9 Diagrams of two different immunohistochemical staining procedures. (A) Indirect immunohistochemical method. A primary antibody is used to detect a cellular antigen (Ag). The primary antibody is identified by a secondary antibody conjugated to horseradish peroxidase (HRP). In the presence of a benzidine substrate and H2O2, HRP catalyzes the production of a colored reaction product that can be observed by routine light microscopy. (B) Avidin-biotin-HRP complex immunohistochemical method. The secondary antibody is conjugated with biotin. Biotin molecules have high affinity for avidin that can be bonded to HRP or alkaline phosphatase. As compared with the indirect immunohistochemical method, the avidin-biotin-HRP complexes provide more HRP molecules per detected antigen and hence a more sensitive immunohistochemical technique.

assays are their excellent sensitivity, retention of cell morphology by light microscopy, minimal sample requirement, and applicability to blood and marrow smears and paraffin-embedded tissue including bone marrow biopsies. Immunohistochemical assays are not commonly used for the initial diagnosis and classification of most leukemias, but are particularly valuable when "dry" bone marrow aspirates and peripheral blood specimens yield insufficient numbers of leukemic cells for flow cytometric analysis. Immunohistochemistry studies of processed bone marrow biopsies enable reliable detection of B- and T-precursor ALLs and AMLs.45"55 This technique can aid in the diagnosis of megakaryoblastic leukemia (M7 AML) from marrow core biopsies when cells cannot be obtain by a needle aspirate.56 Also, these assays are sometimes the only way to establish a diagnosis of granulocytic or monocytic sarcoma in afixedtissue biopsy.57"59 In the past, the deleterious effect of formalin and mercuric tissuefixativeson cellular antigens was the major disadvantage of immunohistochemistry. However, the development of new antibodies and of improved immunologic methods for detecting antigens masked by these tissuefixativeshas largely elimi-

The lineage and maturational stage of neoplastic cells can be examined by immunofluorescent microscopy or flow cytometry with antibodies conjugated to fluorochromes. Immunofluorescent microscopic methods are generally successful, but multiparameter flow cytometry offers the advantages of quantitative measurements of antigen expression and rapid analysis of a large number of cells. A past disadvantage offlowcytometry, especially with samples containing small numbers of malignant cells, was the difficulty of discriminating between neoplastic and normal hematopoietic cells. Current analysis based on CD45 antigen intensity expression and light side scatter largely overcome this limitation.60"63 This analysis takes advantage of the different intensities of CD45 and light side scatter properties of the major hematopoietic lineages. For example, normal lymphocytes strongly express CD45 but produce little light side scatter. By contrast, late myeloid cells express CD45 more weakly than lymphocytes but emit strong light side scatter signals. The simultaneous display of CD45 and light side scatter on a two-parameter histogram permits discrimination between the different cellular components of normal bone marrow (Fig. 7.10). Lymphoid and myeloid leukemias show characteristic histogram patterns of CD45 and light side scatter expressions that largely resemble their normalhematopoietic or lymphoid counterparts and thus, facilitate their identification (Figs. 7.11 and 7.12).

Immunophenotyping panel

With few exceptions, most leukocyte antigens fail to retain their lineage specificity in malignant processes. However, the lineage of over 98% of acute leukemias is discernible with appropriately designed panels of monoclonal antibodies directed toward relatively lineage-restricted antigens.64,65 The antibody screening panel for acute leukemias used at St. Jude Children's Research Hospital (SJCRH) was designed to include at least one very sensitive and one relatively specific marker for each major hematopoietic and lymphoid lineage (Fig. 7.13). Leukemic processes can be analyzed with this panel by use of flow cytometry, immunofluoromicroscopy, or immunohistochemistry. Flow cytometry and immunofluoromicroscopy

Immunophenotyping

B L" IL-

RBC

Light side scatter (SS) intensity

•

Fig. 7.10 Flow cytometric dot plot histograms of CD45 intensity versus light side scatter (SS) for two normal pediatric bone marrow aspirates. Lymphocytic, monocytic, myeloid precursor, and blast populations have distinct intensities of CD45 and light SS, facilitating their identification in a two-dimensional dot plot histogram. In both histograms, discrete populations of bone marrow cells are identified: mature lymphocytes (L), immature B lymphocytes (IL), monocytes (Mo), myeloid precursors (My), and erythroid elements (RBC). (A) Infant marrow containing a large number of immature B lymphocytes. (B) Adolescent marrow with fewer immature B lymphocytes and more myeloid cells.

?'..*-'B

Light side scatter (SS) intensity

Fig. 7.11 Characteristic dot plot histograms of acute leukemias graphed as light side scatter (SS) versus CD45 expression based on Beckman-Coulter flow cytometric analysis. Non-neoplastic lymphocytic, monocytic, and myeloid cell populations are indicated by L, Mo, and My, respectively. The leukemic blast populations are labeled B. The CD45 and light SS signal intensities for leukemic processes are comparable with their normal marrow hematopoietic counterparts (compare with Fig. 7.10). (A) Early pre-B ALL. (B) Precursor-T ALL. (C) Acute myeloblastic leukemia, M1 AML. (D) Acute promyelocytic leukemia, M3 AML. (E) Acute monoblastic leukemia, M5 AML. (F) Acute megakaryoblastic leukemia, M7 AML. All studies were performed on bone marrow samples enriched for mononuclear cells and leukemic blasts by a density-gradient separation technique.

161

Fred G. Behm

My

..&-.& 101

102

103

104

10

101

CD45

'

""IF"

40

My

•V:

o _

'

101

102

103

CD45

My+>'-

If-.

0 " •RBC. *

"Ms

RBC

.i -

s=i •

10

_

:

2l

z> _ o

104

: # ;

5

5

O "

o

40

D

10 3

LLI

o _

2

CD45

10

• Ly

:

)

162

10 1

10 2

10 3

10 4

10 2

CD45

10 3

10 2

CD45

103

CD45

Fig. 7.12 Characteristic dot plot histograms of acute leukemias graphed as CD45 versus light side scatter (SS) based on Becton-Dickinson flow cytometric analysis. The axes for CD45 and SS are reversed compared with Fig. 7.11. Differences in instrumentation result in slightly different histogram representations of normal and leukemic cell populations. Lymphocytic, myelocytic, monocytic, and erythroid populations are indicated by Ly, My, Mo, and RBC, respectively. The leukemic blast populations are identified by the large arrows and normal marrow cells by thin-line arrows. (A) Early pre-B ALL. Leukemic blasts show a spectrum of negative-to-weak CD45 expression and low intensity light SS. (B) Mature B-cell ALL (Burkitt leukemia). Leukemic cells strongly express CD45 and their light SS properties extend into the monocyte region. (C) Precursor-TALL. In general, blasts of precursor-TALL strongly express CD45 while having relatively low intensity light SS. (D) Acute myeloblastic leukemia (M1 AML). (E) Acute myelomonocytic leukemia (M4 AML). Large, open arrow points to the myeloblast component. Large, shaded arrow identifies the monocytic component. (F) Acute monoblastic leukemia (M5 AML). All studies were performed on bone marrow samples that incorporated a red blood cell lysis step.

Acute leukemia

CD45 CD19 CD22 CD79a* CD7 CD3* MPO* CD13 CD33 CD117 CD61 CD235

± + + + ± ± -

B-lineage

CD45 CD19 CD22 CD79a* CD7 CD3* MPO* CD13 CD33 CD117 CD61 CD235

+ ± + + ± ± -

T-lineage

CD45 CD19 CD22 CD79a* CD7 CD3* MPO* CD13 CD33 CD117 CD61 CD235

+ ± ± ± ± ± ± ± -

+

CD45+ CD19 CD22 CD79a* CD7 CD3* MPO* CD13 CD33 CD117 CD61 CD235

+ _ _ ± _ _ ± ± ± + _

Myeloid/ Megakaryocytic monocytic

CD45 CD19 CD22 CD79a* CD7 CD3* MPO* CD13 CD33 CD117 CD61 CD235

± ± ± ± ± -

Erythroid

Fig. 7.13 Example of a basic acute leukemia immunophenotype screening panel. Immunologic studies are necessary to confirm or establish the lineage of myeloperoxidase (MPO)-negative leukemias. Additional antigenic studies are required to distinguish among subtypes of lymphoid and myeloid leukemias. Antigens are identified by their CD (clusters of differentiation) groups. CD235 represents glycophorin A. Asterisks indicate cytoplasmic antigen expression.

require a prior cell permeabilization step to expose myeloperoxidase (MPO) and for optimal detection of CD3 and CD79a.66-68 The SJCRH panel identifies the lineage of over 98% of childhood acute leukemias with samples rich in neoplastic cells. Examples of frequently encountered immunophenotype profiles of ALL and AML are presented in Table 7.2. If specimens have a small leukemic component, it may be necessary to include antibodies to several other antigen groups to differentiate between normal and neoplastic cells. The subclassification of Band T-precursor ALL and AML requires the study of additional leukocyte antigens. Although CD34 and TDT are not lineage-restricted antigens, screening for their presence is useful in distinguishing blastic from mature cell malignancies. Important aspects of TDT, CD34, and the antigens includedinthe SJCRH immunophenotype screening panel are discussed below.

CD45

The common leukocyte antigen (CD45) is a tyrosine phosphatase expressed by all leukocytes and their progenitors.

Immunophenotyping

Table 7.2 Examples of immunophenotype profiles of acute leukemias and their lineage assignments using the SJCRH screening panel

Immunophenotype profile Example

CD19

CD22

CD79aa

CD7

CD3a

CD13

1 2 3 4 5 6 7 8 9 10 11 12 13

CD33

CD117

MPOa

CD61

+ +

14 15 16 17 18

CD235a

Lineage interpretation B B (My+ALL) B (My+ALL) T T T (My+ALL) Myeloid Myeloid Myeloid (Ly+AML) Myeloid (Ly+AML) Myeloid (Ly+AML) Megakaryocytic Megakaryocytic (Ly+AML) Myeloid/erythroidb Erythroidc Biphenotypic d Biphenotypic d Biphenotypic d

Abbreviations and symbols: My+ALL, myeloid antigen-positive acute lymphoblastic leukemia; Ly+AML, lymphoid antigen-positive acute myeloid leukemia (see text and Table 7.12 for descriptions of these leukemia profiles); +, positive; - , negative; + / - , positive or negative. a Cytoplasmic antigen expression. b Characteristic of acute erythroleukemia, M6 AML. c Characteristic of erythroblastic leukemia with no significant myeloblast component. d Mixed lymphoid-myeloid lineage (see Table 7.12 for definition of biphenotypic leukemia).

It functions in antigen receptor signaling by dephosphorylation of Src kinase.69 CD45 also acts as a Janus kinase (JAK) tyrosine phospatase to regulate cytokine receptor signaling in differentiation and proliferation of hematopoietic cells. CD45 appears very early in B-cell development and persists throughout maturation up to but not through the plasma cell stage.1 Similarly, T and myeloid cells at all stages of maturation express CD45. The cell surface intensity of CD45 increases with B- and T-cell maturation but stays relatively constant with myeloid maturation. Very early erythroblasts express CD45 but at levels that are not easily discernible from cellular autofluoresence by routine flow cytometry. Megakaryocytes and platelets do not display CD45. All acute leukemias, except for approximately 10% of B-precursor ALLs and some M7 AMLs have easily detectable levels of CD45.70 Low-level CD45 expression in ALL correlates with favorable clinical and laboratory features, including lower leukocyte counts, B-cell lineage, and hyperdiploidy (>51 chromosomes).70,71 CD45 is very important in differentiating leukemia and lymphoma from small cell tumors. Small cell tumors (e.g. neuroblastoma, Ewing sarcoma, and rhabdomyosarcoma) can involve the

bone marrow and morphologically mimic leukemia but do not synthesize CD45. CD19, CD22, and CD79a

The synthesis of immunoglobulins is the hallmark of Blineage commitment. However, Ig molecules are insensitive markers for diagnosing precursor-B ALL in that only about 25% of such cases will express these proteins. By contrast, CD 19, CD22, and cytoplasmic CD79a are expressed in almost every case of B-lineage ALL. The sensitivity of these latter markers is compromised by their atypical expression in other types of acute leukemia. Cases of AML and T-ALL can also weakly express CD 19, CD22, or CD79a, but coexpression of CD 19 with CD22 or CD79a is indicative of a B-lineage commitment with the exception of rare examples of biphenotypic leukemia (Table 7.2). Thus, it is recommended that immunophenotyping studies of acute leukemia include CD 19 plus CD22 or CD79a. CD19 alone is expressed by the earliest of precursor B cells and remains detectable through B-cell maturation up to the plasma cell (Figs. 7.1 and 7.3).1 Follicular dendritic cells in lymphoid tissues also express CD 19, which is not

163

164

Fred G. Behm

detectable in epithelial and soft tissue tumors. This antigen is a signal transduction molecule that participates in Blymphocyte development, activation, and differentiation. The intensity of CD19 expressionis useful in distinguishing between precursor-B ALL and other leukemias that aberrantly display this antigen. The aberrant expression of CD 19 by some cases of AML and T-ALL is very weak in comparison to strong expression in precursor-B ALL. CD22 is a B cell restricted glycosylated protein that acts as an adhesion receptor and signaling molecule.1 It is expressed at all stages of B cell differentiation, initially in the cytoplasm of early precursor B cells and on the surface of more mature B cells (Fig. 7.2). Similar to their paucity of CD19, plasma cells do not express CD22, while basophils express CD22 but not CD19. With the appropriate monoclonal antibodies, over 98% of precursor-B ALLs show relatively strong surface expression of CD22. Although normal myeloid- and T-cells and the blasts of T-ALL do not express this antigen, certain anti-CD22 antibodies can react with a non-CD22 cytoplasmic protein in some AMLs.72 A small number of myeloperoxidase-positiveAMLs weakly express surface CD22. CD79a forms a heterodimer with CD79b that is noncovalently bound to immunoglobulin to form the BCR (Fig. 7.2). 1 The heterodimer transmits signals into the cytoplasm upon antigen binding by cell surface immunoglobulin. CD79a appears prior to CD 19 in B-cell ontogeny.73 Current commercially available monoclonal antibodies only detect the cytoplasmic domain of CD79a. With cell permeabilization techniques and flow cytometry over 98% of precursor B ALLs will have detectable CD79a. The leukemic blasts of some cases of precursor-B ALL with the t(4;11) translocation involving the MLL gene can express little or no CD79a. Initial studies found CD79a only in normal and neoplastic B cells,74,75 but more recent investigations show that CD79a is also weakly expressed by a minority of T-ALLs and AMLs.51 , 76 - 80 The CD79b antigen is restricted to B cells, but is an insensitive marker for most cases of precursor-B ALL.75 CD7 and CD3

CD7 is a glycoprotein found on the surface of pluripotent hematopoietic stem cells, thymic and mature T cells, and NK cells.1 CD7 appears to be a coactivator molecule involved in cytokine secretion and cellular adhesion.1 It is a very sensitive marker of precursorT-ALL but lacks lineage specificity. Virtually every case of precursor-T ALL expresses CD7; thus, its absence on leukemic blasts mitigates against that diagnosis. Expression by precursor-T ALL blasts is strong as compared to normal mature T cells. Unfortunately for lineage assignment purposes, CD7 is also weakly expressed by approx-

imately 50% of acute megakaryoblastic leukemias, many acute myeloid leukemias, and a small percentage of 81 84 B-lineage ALLs (Table 7.2). ~ The expression of CD7 by precursor-B ALLappears to lack clinical significance, butits expression by AMLs maybe associated with a lower overall survival rate. 82 " 84 CD3 consists of six polypeptides divided into three dimers (7 /ε, 8/e, and t, It,) that associate with either the TCR 1 22 25 28 ct(J or 78 proteins to form the CD3-TCR complex. , , , CD3 is also associated with a pre-TCR molecule on thymic T cells. Although the ε chain of CD3 has been described in fetal cells and pro thymic T cells with the potential of differentiating along myeloid or natural killer (NK) lineages, the complete CD3 complex is expressed only by T cells in nor85 mal postnatal tissues. The youngest thymic cell committed to T-lineage development contains cytoplasmic but not 26 85 86 surface CD3. , , Because many precursor-T ALL cases express only cytoplasmic CD3, test methodologies must include a procedure that permeabilizes the leukemic cell membrane to expose CD3. With these methods, every case of T-ALL has detectable cytoplasmic CD3, providing an excellent T-lineage-restricted antigen for immunophenotyping studies.86"88 Although many monoclonal CD3 antibodies only detect the ε chain when combined with the 7 or 8 chain of the CD3 complex, some polyclonal antibodies to CD3ε detect CD3ε proteins independently.89 This raises the possibility that rare cases of leukemia reacting with the latter polyclonal antibodies may correspond to a pro-T or early NK-cell stage of development rather than a committed T-lineage process. CD33 and CD13

CD33 is a transmembrane glycoprotein that is expressed by myeloid progenitors (GFU-GEMM, CFU-GM, CFU-G, BFUE), granulocyte precursors, mature granulocytes, monocytes, and macrophages but not normal lymphocytes.1,90 CD33 has no known expression outside of hematopoiesis, and its biologic function is not understood. CD 13 is similar to CD33 in its expression by granulocytic and monocytic precursors and their progeny.1,91 This metallopeptidase is identical to aminopeptidase N, which degrades regulatory pep tides produced by a wide variety of cell types. CD 13 is also expressed on a subpopulation of large granular lymphocytes and cells of vascular endothelium, renal proximal tubules, intestinal brush border, bone marrow stroma, and osteoclasts.1 Small numbers of normal precursor-B cells in the marrow (< 5 x 1CT3) and reactivated mature B cells express CD13.92,93 CD13 expression can be induced by in vitro stimulation with B cell growth factor.94 Antibodies to many myeloid- andmonocytic-associated antigens, including CD13, CD14, CD15, CD33, CD64, and CD65, are available for identifying the blast cells of AML.

Immunophenotyping

However, only CD13 and CD33 have proved to be sensitive markers of AML. Over 95% of AMLs express CD13 and/or CD33. In general, expression of CD33 is more intense and CD 13 less intense in monocytic than in myelogenous acute leukemias. Unfortunately, CD 13 andCD33 are notvery specific, in that 15% to 35% of ALLs also weakly express one or both of these antigens, although the weak specificity of CD13 and CD33 is offset by their sensitivity for AML when interpretedin the context of the expression (or lack thereof) of other antigens listed in the SJCRH screening panel (Table 7.2) .95 Myeloperoxidase The cytoplasmic granules of myeloid, monocytic, and eosinophilic cells contain peroxidases that defend against invading microorganisms. One of these, myeloperoxidase (MPO), a specific marker for myelocytic and monocytic lineages, is indispensable in the diagnosis and classification of acute leukemias.2,96 The MPO gene isfirsttranscribed in CD34-positive progenitor cells that give rise to myeloblasts (Fig. 7.7) ,97~101 while myeloperoxidase is packaged together with other leukocyte enzymes in primary granules. The enzymatic form of MPO is detected by cytochemical reactions with benzidine compounds. Newer approaches use immunofluorescence or immunohistochemical assays with monoclonal antibodies to detect MPO protein.102,103 The cytochemical test for peroxidase requires the presence of functional enzyme, whereas the antibody test needs only the intact protein. The enzymatic activity of MPO decays rapidly with time, but old cytologic preparations retain sufficient antigenic sites for antibody binding. The proenzyme form of MPO that is cleaved to produce the dimeric enzymatic form of MPO is also detected by anti-MPO staining, providing a test for very immature myeloid cells. The anti-MPO test may be more sensitive than the cytochemical assay for MPO. Indeed, cases of acute leukemia cytochemically negative for MPO may test positive with anti-MPO This may be due to reactions with nonfuncassays. tional, degraded, or proenzymatic forms of MPO. Whether the anti-MPO method is more sensitive than the enzymatic reaction is controversial.109 Several investigations detected a surprisingly, if not suspiciously, high percentage of ALLs expressing MPO by immunologic methods.110,111 Significant differences between enzymatic and antigenic MPO expression are not observed at SJCRH, where these tests are performed within hours of each other on aliquots of the same leukemic specimen. CD117

The CD 117 antigen is the product of the proto-oncogene cKITand belongs to a family of growth factor receptors with tyrosine kinase activity.1,112,113 This receptor is found on

over half of all CD34+ cells and megakaryocytic, erythroid, granulocytic, and monocytic lineage-restricted progenitor cells in normal bone marrow. Bone marrow mast cells, a subset of NK cells, and some early prothymic T cells also have detectable CD 117. With its ligand (mast cell factor, stem cell factor, or Steel factor), CD 117 is thought to play a crucial role in early hematopoiesis. Only a rare case of T-precursor ALL and almost no cases of precursor-B ALL express CD117.114"118 Blasts of up to 90% of cases of AML express CD117, and it appears to be more highly associated with AML than either CD13 or CD33.119 All subtypes of AML can express CD117.103'117"121 The close association with AML makes CD 117 a valuable addition to a marker panel designed for lineage determination.65,119 In childhood and adult AML, CD117 is strongly associated with the expression of CD34 and CD7, but not with other clinical features or with prognosis. 108,117,119,122

TDT This nuclear DNA polymerase participates in the addition of nucleotides to the N regions of Ig and TCR genes undergoing rearrangement.123 Thus, TDT is normally detected in the nuclei of immature B and T lymphocytes and disappears with lymphocyte maturation (Figs. 7.1 and 7.4).124,125 In healthy persons, cells bearing TDT are present in the bone marrow and thymus, but in very small numbers in the peripheral blood and lymphoid tissues.126,127 TDT is readily detected with immunofluorescence and immunohistochemical techniques using cell smears, touch preparations, cytospin smears, or frozen tissue sections. All of these tissue preparations rapidly lose TDT if left at room temperature for more than 24 hours. TDT is masked from anti-TDT antibodies in marrow and other tissue biopsies fixed in formalin, but immunologic reactivity can be restored by microwaving or by other antigen retrieval techniques.49 Flow cytometry with methods that partially permeabilize cells to expose nuclear proteins to antibodies to TDT is in common use. The blasts of more than 90% of B- and T-lineage ALLs and lymphoblastic lymphomas harbor easily detectable TDT.128,129 Sensitive flow cytometric and immunohistochemical assays will reveal TDT in 10% to 45% of AMLs.130,131 Although TDT assays are not useful for determining lineage, they are helpful in several diagnostic situations. For example, assays are positive in lymphoblastic but not myeloblastic crisis of chronic myelogenous leukemias.2 The presence of TDT-positive blasts in spinal, pleural, or peritoneal fluid, testicular biopsy specimens, or other nonlymphoid tissues is indicative of a lymphoblastic malignancy. TDT is also helpful in distinguishing lymphoblastic lymphoma from Burkitt and other lymphomas.2 Dual immunofluorescence techniques use anti-TDT plus anti-CD3 to detect minimal residual

165

166

Fred G. Behm

Table 7.3 Immunologic classification of B-lineage ALL

Immunologic marker (% cases positive for marker) Subtype

CD19

CD20

CD22

CD79aa

CD10

cylg|x

slg|x

slgKOrA.

Frequency

Early pre-B (pro-B) Pre-B Transitional pre-B (late pre-B) Mature B

100 100 100 100

35 45 55 99

99 100 100 100

99 100 100 99

95 100 100 50

0 100 100 100

0 1 100 ~95 b

0 0 0 ~95 b

60-65% 20-25% 10-12% 3-5%

Abbreviations: cylg|x, cytoplasmic immunoglobulin mu heavy chain; slg|x, surface immunoglobulin mu heavy chain; slgK or \, surface immunoglobulin kappa or lambda light chain. a Cytoplasmic expression. b From 2% to 5% of cases with t(8;14), t(2;8), or t(8;22) may lack surface immunoglobulin or express only cytoplasmic immunoglobulin. disease in the b o n e m a r r o w of patients with T-lineage ALL. 132

CD34

The transmembrane sialoglycoprotein CD34 is expressed by early hematopoietic progenitors of all lineages, endothelial cells of high endothelial venules, bone marrow stromal cells, peripheral nerve sheath cells, and osteoclasts.133"137 This glycoprotein may play a role in progenitor cell localization and stromal adhesion in the b one marrow. Less than 10% of cells inpostnatalbone marrows express CD34. Primitive multipotent hematopoietic stem cells express CD34 but not CD38 or other leukocyte-associated antigens. However, the majority of normal CD34-positive marrow cells correspond to later stages of stem cell commitment and coexpress CD38, HLA-DR, CD33, and/or CD19.138 Smaller populations of CD34 cells express CD4, CD10, CD7, and/or CD41. In normal bone marrows most CD34+ cells appear committed to the myeloid lineage (CD13+, CD33+) or to lymphoid development (CD19+'", CD10+). CD34 expression progressively decreases as hematopoietic progenitors differentiate. The majority of acute leukemias are CD34-positive, whereas the chronic leukemias are negative.2 About 70% of B-lineage and30% of T-lineage ALLs are CD34-positive. In adult AML, CD34 expression correlates with a M1 or M2 morphology by French-American-British (FAB) criteria, as well as leukemias evolving from myelodysplastic syndrome, karyotypic abnormalities of chromosome 5 or 7, and a lower remission induction rate.139"143 In pediatric AML, CD34 expression also correlates withM1 andM2morphologies but not with chromosome 5 and 7 abnormalities, and appears to have no prognostic significance.144,145 In childhood precursor-B ALL, CD34 expression by leukemic blasts is associated with an age of 1 to 10 years, hyperdiploidy (>50 chromosomes), the absence of central ner-

vous system (CNS) leukemia, a n d a favorable response to therapy. 1 4 6 , 1 4 7

Classification of specific leukemias and lymphomas B-lineage ALL

Several classifications of B-lineage ALL have been proposed. The SJCRH classification recognizes four subtypes discernible by their pattern of immunoglobulin expression (Table 7.3).148 Other classifications include additional subgroups based on schemes of immunoglobulin, CD10, CD179a, or CD179b expression.149"152 Although initial studies demonstrated an association of treatment response with different subtypes of precursorB ALL, improved treatment approaches and recognition of the overriding significance of genotypic abnormalities has marginalized the importance of immunologic subgrouping in these leukemias. The WHO classification divides acute B-lymphoid neoplasms into precursor-B lymphoblastic leukemia/lymphoma and Burkitt leukemia/lymphoma with further subgrouping based on cytogenetic abnormalities.2 Early pre-B ALL

The leukemic blasts of early pre-B ALL resemble a normal marrow B-precursor cell that lacks immunoglobulins (Fig. 7.1). Although Ig heavy-chain genes are usually rearranged in these leukemias, immunoglobulins are not detectable. The leukemic cells of all early pre-B ALL cases expressCD19andHLA-DR(Table7.3).Allbutrarecases display surface CD22 and/or cytoplasmic CD22, and almost all have cytoplasmic CD79a.63,69,81-83 CD10 and TDT are detectable in over 90% of cases, andmore than 75% express CD34.147,148 The CD20 antigen that normally appears with

Immunophenotyping

Table 7.4 Antigen expression profiles typical of B-lineage ALL with various cytogenetic features

Karyotype t(4:11)(q21;q23) t(11;19)(q23;q13.3) t(12;21)(p12;q22) t(1;19)(q23;p13) Hyperdiploidat(1;19) t(9;22)(q34;q11) Hyperdiploida t(8;14)(q24;q32) Normal

Genes involved

ALL subtype

AF4-MLL MLL-ENL TEL-AML1 PBX1-E2A

EPB EPB PB, EPB PB, TBP EPB PB, EPB PB, EPB Mature B EPB, PB

ABL-BCR MYC-IGH

Leukocyte antigen expression profile CD45

CD34

CD22

CD10

CD13

CD15

CD24

/-

-10 + -to +

CD33

CD66c

NG2

-to +

Abbreviations and symbols: NG2, nonhematopoietic chondroitin proteoglycan sulfate; EPB, early pre-B; PB, pre-B; TPB, transitional preB;+, weakly positive; ++, moderately positive, +++, strongly positive; -, negative; - / + , negative more often than positive; + / - , positive more often than negative. a More than 51 chromosomes. the production of (j, heavy chains is present in varying proportions of blasts in many cases. Up to 10% of early pre-B ALLs do not have detectable CD45.70 Leukemias harboring rearrangements of the MLL gene, resulting from the t(4;11), t(11;19), and t(9;11) chromosomal translocations, are usually classified as early pre-B ALL, although examples have been described with a preB and precursor-T immunophenotype. 153,154 The blasts of t(4;11)+ B-lineage ALL usually present with a characteristic antigenic profile: CD19+, CD22+, CD24 lo/ -, and CD10- (Table 7.4).154,155 This differs from other B-lineage cases, which almost always show strong expression of CD10 and CD24. Further, most t(4;11)+ cases express myeloid-associated CD 15 or CD65 antigens. 154 , 155 ThepanT-cell CD7 antigen and myeloid-associated antigens CD 13 and CD33 can also be present. Precursor-B ALLs with a t(9;11) or t(11;19) often have immunophenotypes similar to those associated with the t(4;11) (Table 7.4).156 Cellsurface chondroitin protoglycan sulfate, a nonhematopoietic cellular molecule detected by the monoclonal antibody 7.1, is present in almost all precursor-B ALLs and over one-half ofAMLs having a rearrangedMLL gene. 157~161 T-cell-associated CD2 is rarely expressed by early preB ALL and may be associated with poorer treatment outcome. 162 " 164 Pre-B ALL

About 25% of newly diagnosed ALLs have a preB immunophenotype. 148 Like early pre-B ALL, this subtype expresses CD 19, CD22, CD79a, and HLA-DR (Table 7.3). By definition, the lymphoblasts of pre-B ALL exhibit cytoplasmic Ig (cIg) (j, chains without detectable

surface (sIg) (j, chains. 148,165 Rearrangement of Ig lightchain genes is evident in some of these leukemias, but K and X proteins are not detectable. Over 95% of these leukemias express CD10 and TDT, but only two-thirds express CD34.147,148,166 In contrast to the normalbone marrow pre-B lymphocyte, blasts of pre-B ALL may lack or only weakly express surface CD20, which is thought to function in B-cell activation and proliferation by regulating transmembrane Ca2+ conductance and cell cycle progression. 1 Studies of precursor-B ALL, including early pre-B and preB ALL, suggest that expression of CD20 maybe associated with a poorer treatment response. 167 In comparison to earlypre-B ALL, pre-B ALL is more often associated with higher leukocyte counts, elevated serum lactic acid levels, fewer than 51 chromosomes or a DNA indexless than 1.16, and recurrent chromosome translocations. Between 20% and 25% of pre-B ALLs harbor either the t(1;19)(q23;p13) or der(19)t(1;19)(q23;p13).148,168 Patients with these translocations require more intensive therapy to obtain a satisfactory treatment response. The antigen expression profile of CD19+, CD22+, CD20 ±, CD34~, CD45hi, clgjju+ is characteristic of ALLs with the t(1;19) but lacks specificity (Table 7.4).167,169 A few well-studied cases of early pre-B ALL also carry a t(1;19).167,169,170 However, in those cases the blasts are CD34+, hyperdiploid (chromosome number >51), and lack evidence of the chimeric E2A-PBX1, which is transcript that is uniformly detected in t(1;19)+ pre-BALL.170 Polyclonal and monoclonal antibodies are now available for the flow cytometric detection of the PBX1 portion of the chimeric E2A-PBX1 protein in the nucleus of t(1;19)+leukemic blasts.171 Immunologic testing for E2A-PBX1 provides rapid screening for leukemias with

167

168

Fred G. Behm

Table 7.5 Features of B-lineage ALL expressing surface immunoglobulin

Transitional pre-B Mature B (naive B-cell type) Mature B (GC B-cell type)

Extramedullary mass

FAB subtype

No No Yesb

L1orL2 L1orL2 L3 (L2)

Immunologic marker expression TDT CD34

CD20

CD10

cylg|x

slg|x

slgK or \

Karyotype Variablea Variablea t(8;14)(q24;q32), t(2;8)(p12;q24), or t(8;22)(q24;q11)

Abbreviations and symbols: GC, germinal center; cylg|x, cytoplasmic immunoglobulin mu heavy chain; slg|x, surface immunoglobulin mu heavy chain; slgK or \, surface immunoglobulin kappa or lambda light chain; +, positive; —.negative; +/—, positive or negative; +(—), rarely negative. a t(1;19)(p13;q32) frequent, no t(8;14)(q24;q32), t(2;8)(p12;q24), or t(8;22)(q24;q11), and no MYCrearrangement. b Majority of patients present with lymphomatous masses of the ileo-cecum, gonads, or head and neck. c Less than 5% of cases express IgA or IgG without IgM;