Myelodysplastic Syndromes (2009) PDF

Hematology Second
about the book… Edition

Written by a team of leading authorities in pathogenesis, diagnostic techniques, and
Myelodysplastic
clinical management strategies in myelodysplastic syndromes (MDS), this text provides a
concise, easy-to-follow review of the advances in the science, classification, diagnosis, and
Pathobiology and Clinical Management

Myelodysplastic Syndromes
management of these challenging conditions.
An ideal source for hematologists, oncologists, and researchers, this Second Edition
Syndromes
features:
• a new eight-page color insert
• 200 color and black-and-white illustrations
• reworked content, organized into three sections: MDS epidemiology and biology;
MDS diagnosis, classification, and prognosis; and MDS therapy
Pathobiology and
• 21 thoroughly updated chapters, reflecting a shift in topical focus as dictated by
the evolution of the field
New topics in Myelodysplastic Syndromes include:
• del(5q) and 5q-syndrome, including new biological insights such as RPS14
down-regulation
• MDS-MPN overlap syndromes including CMML
Clinical Management
Second Edition
• the “reborn” DNA methyltransferase inhibiting nucleoside analogs, azacitidine
and decitabine
• the potential importance of iron overload in MDS
• global genomics approaches, such as gene expression arrays, array-based comparative
genomic hybridization, and single nucleotide polymorphism arrays
• the latest classification revisions, including the 2008 changes to the WHO classification
• the role of mitochondrial ferritin accumulation and other mitochondrial metabolic Edited by
anomalies in MDS
• other newly defined disease-associated alterations, such as abnormalities of
B lymphocyte populations
David P. Steensma
• three recently FDA-approved drugs for MDS: azacitidine, lenalidomide, and decitabine
about the editor...
DAVID P. STEENSMA is a Consultant and Associate Professor of Medicine and Oncology at
Mayo Clinic, Rochester, Minnesota, USA. Originally from the New York City suburbs, he
received his M.D. from the Pritzker School of Medicine at the University of Chicago,
Chicago, Illinois, USA. Dr. Steensma completed his clinical training in internal medicine,
hematology and medical oncology at Mayo Clinic, and research training at the Weatherall
Institute of Molecular Medicine in Oxford, England. His laboratory efforts focus on the
molecular genetics of MDS, and he is also currently conducting several clinical trials aimed
at improving outcomes and quality of life for patients with myelodysplastic syndromes
(MDS) and various forms of anemia. He is a member of the North Central Cancer Treatment
Group, Mayo Clinic Cancer Research Consortium, and the Eastern Cooperative Oncology
Group Leukemia Committee. Steensma
Printed in the United States of America H7439
Myelodysplastic
Syndromes
BASIC AND CLINICAL ONCOLOGY
Series Editor
Bruce D. Cheson
Professor of Medicine and Oncology
Head of Hematology
Georgetown University
Lombardi Comprehensive Cancer Center
Washington, D.C.
1. Chronic Lymphocytic Leukemia: Scientific Advances and Clinical Develop-

ments, edited by Bruce D. Cheson
2. Therapeutic Applications of Interleukin-2, edited by Michael B. Atkins and
James W. Mier
3. Cancer of the Prostate, edited by Sakti Das and E. David Crawford
4. Retinoids in Oncology, edited by Waun Ki Hong and Reuben Lotan
5. Filgrastim (r-metHuG-CSF) in Clinical Practice, edited by George Morstyn
and T. Michael Dexter
6. Cancer Prevention and Control, edited by Peter Greenwald, Barnett S.
Kramer, and Douglas L. Weed
7. Handbook of Supportive Care in Cancer, edited by Jean Klastersky, Stephen
C. Schimpff, and Hans-Jörg Senn
8. Paclitaxel in Cancer Treatment, edited by William P. McGuire and Eric K.
Rowinsky
9. Principles of Antineoplastic Drug Development and Pharmacology, edited
by Richard L. Schilsky, Gérard A. Milano, and Mark J. Ratain
10. Gene Therapy in Cancer, edited by Malcolm K. Brenner and Robert C. Moen
11. Expert Consultations in Gynecological Cancers, edited by Maurie Markman
and Jerome L. Belinson
12. Nucleoside Analogs in Cancer Therapy, edited by Bruce D. Cheson, Michael
J. Keating, and William Plunkett
13. Drug Resistance in Oncology, edited by Samuel D. Bernal
14. Medical Management of Hematological Malignant Diseases, edited by Emil
J Freireich and Hagop M. Kantarjian
15. Monoclonal Antibody-Based Therapy of Cancer, edited by Michael L.
Grossbard
16. Medical Management of Chronic Myelogenous Leukemia, edited by Moshe
Talpaz and Hagop M. Kantarjian
17. Expert Consultations in Breast Cancer: Critical Pathways and Clinical Deci-
sion Making, edited by William N. Hait, David A. August, and Bruce G.
Haffty
18. Cancer Screening: Theory and Practice, edited by Barnett S. Kramer, John
K. Gohagan, and Philip C. Prorok
19. Supportive Care in Cancer: A Handbook for Oncologists: Second Edition,
Revised and Expanded, edited by Jean Klastersky, Stephen C. Schimpff, and
Hans-Jörg Senn
20. Integrated Cancer Management: Surgery, Medical Oncology, and Radiation
Oncology, edited by Michael H. Torosian
21. AIDS-Related Cancers and Their Treatment, edited by Ellen G. Feigal,
Alexandra M. Levine, and Robert J. Biggar
22. Allogeneic Immunotherapy for Malignant Diseases, edited by John Barrett
and Yin-Zheng Jiang
23. Cancer in the Elderly, edited by Carrie P. Hunter, Karen A. Johnson, and
Hyman B. Muss
24. Tumor Angiogenesis and Microcirculation, edited by Emile E. Voest and
Patricia A. D’Amore
25. Controversies in Lung Cancer: A Multidisciplinary Approach, edited by
Benjamin Movsas, Corey J. Langer, and Melvyn Goldberg
26. Chronic Lymphoid Leukemias: Second Edition, Revised and Expanded,
edited by Bruce D. Cheson
27. The Myelodysplastic Syndromes: Pathology and Clinical Management,
edited by John M. Bennett
28. Chemotherapy for Gynecological Neoplasms: Current Therapy and Novel
Approaches, edited by Roberto Angioli, Pierluigi Benedetti Panici, John J.
Kavanagh, Sergio Pecorelli, and Manuel Penalver
29. Infections in Cancer Patients, edited by John N. Greene
30. Endocrine Therapy for Breast Cancer, edited by James N. Ingle and Mitchell
Dowsett
31. Anemia of Chronic Disease, edited by Guenter Weiss, Victor R. Gordeuk,
and Chaim Hershko
32. Cancer Risk Assessment, edited by Peter G. Shields
33. Thrombocytopenia, edited by Keith R. McCrae
34. Treatment and Management of Cancer in the Elderly, edited by Hyman B.
Muss, Carrie P. Hunter, and Karen A. Johnson
35. Innovative Leukemia and Lymphoma Therapy, edited by Gertjan J. L.
Kaspers, Bertrand Coiffier, Michael C. Heinrich, and Elihu Estey
36. Myelodysplastic Syndromes, Second Edition: Pathobiology and Clinical
Management, edited by David P. Steensma
Myelodysplastic
Syndromes
Pathobiology and
Clinical Management
Second Edition
Edited by
David P. Steensma
Mayo Clinic
Rochester, Minnesota, USA
Informa Healthcare USA, Inc.
52 Vanderbilt Avenue
New York, NY 10017

C 2009 by Informa Healthcare USA, Inc.
Informa Healthcare is an Informa business
No claim to original U.S. Government works
Printed in the United States of America on acid-free paper
10 9 8 7 6 5 4 3 2 1
International Standard Book Number-10: 1-4200-7439-3 (Hardcover)
International Standard Book Number-13: 978-1-4200-7439-0 (Hardcover)
This book contains information obtained from authentic and highly regarded sources. Reprinted
material is quoted with permission, and sources are indicated. A wide variety of references are
listed. Reasonable efforts have been made to publish reliable data and information, but the author
and the publisher cannot assume responsibility for the validity of all materials or for the conse-
quence of their use.
No part of this book may be reprinted, reproduced, transmitted, or utilized in any form by any
electronic, mechanical, or other means, now known or hereafter invented, including photocopying,
microfilming, and recording, or in any information storage or retrieval system, without written
permission from the publishers.
For permission to photocopy or use material electronically from this work, please access
www.copyright.com (http://www.copyright.com/) or contact the Copyright Clearance Center, Inc.
(CCC) 222 Rosewood Drive, Danvers, MA 01923, 978-750-8400. CCC is a not-for-profit organiza-
tion that provides licenses and registration for a variety of users. For organizations that have been
granted a photocopy license by the CCC, a separate system of payment has been arranged.
Trademark Notice: Product or corporate names may be trademarks or registered trademarks, and
are used only for identification and explanation without intent to infringe.
Library of Congress Cataloging-in-Publication Data

Myelodysplastic syndromes : pathobiology and clinical management. – 2nd ed. /
edited by David P. Steensma.
p. ; cm. – (Basic and clinical oncology ; 36)
Includes bibliographical references and index.
ISBN-13: 978-1-4200-7439-0 (hardcover : alk. paper)
ISBN-10: 1-4200-7439-3 (hardcover : alk. paper)
1. Myelodysplastic syndromes. I. Steensma, David P. II. Series.
[DNLM: 1. Myelodysplastic Syndroms–physiopathology. 2. Myelodysplastic
Syndroms–therapy. W1 BA813W v.36 2008 / WH 380 M99593 2008]
RC645.73.M95 2008
616.4’1–dc22 2008035268
For Corporate Sales and Reprint Permissions call 212-520-2700 or write to: Sales Department,
52 Vanderbilt Avenue, 7th floor, New York, NY 10017.
Visit the Informa Web site at
www.informa.com
and the Informa Healthcare Web site at
www.informahealthcare.com
Preface to the Second Edition
The landscape of the myelodysplastic syndromes (MDS) has changed consid-

erably since the first edition of Myelodysplastic Syndromes: Pathobiology and
Clinical Management, edited by venerable MDS morphologist Dr. John Bennett,
was published in 2002. In these 6 years since that text appeared, the Food and Drug
Administration in the United States has approved three new medications for MDS
indications (azacitidine in 2004, lenalidomide in 2005, and decitabine in 2006),
and clinicians and morphologists have grown comfortable enough with the 2001
disease re-classification by the World Health Organization to demand a revision.
MDS is now captured by the Surveillance, Epidemiology, and End Results (SEER)
program of the National Cancer Institute, promising more accurate epidemiolog-
ical numbers. Supportive care strategies were altered by increased realization of
the potential harm of chronic transfusional iron overload in MDS, by approval of
an oral iron chelator, deferasirox, and by the introduction of a long-acting erythro-
poiesis stimulating agent, darbepoetin alfa. The range of patients suffering from
MDS who are eligible for hematopoietic stem cell transplantation has expanded,
as nonmyeloablative approaches allow older and sicker patients to be considered
for this procedure.
Our understanding of the pathobiology of MDS has perhaps not kept up with
the myeloproliferative disorders, where the landmark discovery of the common
JAK2 V617F mutation in 2005 led to major changes in diagnosis and disease
classification. Still, there have been some exciting advances, with much hope
for the near future as biotechnology techniques continue to get better, cheaper,
and faster. Global genomic approaches (e.g., cDNA expression arrays, array-
based comparative genomic hybridization, and single nucleotide polymorphism
arrays) have yielded new insights into gene expression patterns in MDS; several
provocative findings from these global approaches promise to stimulate new lines
of research. The role of abnormal mitochondrial ferritin accumulation and other
mitochondrial metabolic anomalies in MDS is now better appreciated, and other
disease-associated alterations, such as abnormalities of B lymphocyte populations
in MDS, are also receiving increased attention.
This edition of Myelodysplastic Syndromes: Pathobiology and Clinical Man-
agement again assembles a team of knowledgeable contributors chosen from
among the world’s leading MDS authorities for their understanding of pathogen-
esis, diagnostic techniques, and clinical management strategies in MDS. After an
introductory chapter on the history of the nosologic concept of MDS, the first
major section of the book focuses on MDS epidemiology and biology, followed
iii
iv Preface to the Second Edition
by a section on diagnosis, classification, and prognosis, and concluding with a sec-

tion on MDS therapy. New areas of concentration in this book include a chapter
on del(5q) and 5q syndrome—a group of conditions that has assumed increasing
importance because of the striking clinical success with lenalidomide therapy—as
well as two chapters on lenalidomide and on the reborn DNA methyltransferase
inhibiting nucleoside analogs. We are also fortunate to be able to include new chap-
ters devoted to the challenging diagnosis of chronic myelomonocytic leukemia and
other MDS cases with myeloproliferative features, global genomic approaches to
MDS, the role of mitochondria in MDS, and the potential importance of iron
overload in MDS. In addition, all the other chapters from the first edition have
been thoroughly updated; in some cases, merging chapters or shifting the topical
focus as dictated by the evolution of the field has also been done.
The growing interest in MDS in the medical and pharmaceutical communi-
ties is attested to by the increased numbers of attendees at the biennial international
symposia organized by the MDS Foundation, as well as the packed crowds that
now routinely fill MDS sessions at large general hematology meetings such as
the annual meeting of the American Society of Hematology and the congresses
of the European Hematology Association. Therefore, it seems timely to bring out
a new edition of this text, which the contributors and I believe will be of value
for investigators engaged in MDS research, clinicians diagnosing and caring for
patients with MDS, and interested students of this challenging and puzzling group
of hematopoietic failure states.
David P. Steensma
Contents
Preface . . . . iii
Contributors . . . . vii
1. The Myelodysplastic Syndromes: History and Classification 1

David P. Steensma and John M. Bennett
2. Epidemiology of the Myelodysplastic Syndromes 27

Terry J. Hamblin
3. The Cytogenetics and Molecular Biology of the Myelodysplastic

Syndromes 49
Harold J. Olney and Michelle M. Le Beau
4. Genomic Approaches in MDS 87

Andrea Pellagatti
5. The Role of Apoptosis in MDS 107

Yatato Yoshida
6. The Role of Mitochondria in MDS 127

Norbert Gattermann
7. Defects in Iron Metabolism and Iron Overload in MDS 153

Luca Malcovati
8. Therapy-Related Myelodysplastic Syndrome and Myeloid Leukemia 173

Lucy A. Godley and Richard A. Larson
9. Diagnosis of Myelodysplastic Syndromes: Criteria and Challenges 195

Attilio Orazi and James W. Vardiman
10. Hypocellular Myelodysplastic Syndrome: Relationship to Aplastic Anemia

and Hypocellular Acute Myeloid Leukemia 227
Tomoko Hata and Masao Tomonaga
11. Diagnostic and Prognostic Utility of Flow Cytometry in MDS 247

Denise A. Wells and Michael R. Loken
v
vi Contents
12. Molecular Pathogenesis of the 5q− Syndrome 267

Jacqueline Boultwood and James S. Wainscoat
13. Chronic Myelomonocytic Leukemia and Myelodysplastic Syndrome/

Myeloproliferative Overlap Syndromes 285
Phuong L. Nguyen and Curtis A. Hanson
14. MDS in Children 311

Henrik Hasle and Charlotte M. Niemeyer
15. Prognostic Factors in the Assessment of Patients with Myelodysplastic

Syndromes 347
Aristoteles Giagounidis, Carlo Aul and Ulrich Germing
16. Therapeutic Strategies: The Approach to Care of Patients with MDS, and
Criteria for Treatment Response 393
Pierre Fenaux, Lionel Ades and Claude Gardin
17. Management of Cytopenias in MDS 413

Luca Malcovati, David T. Bowen and Eva Hellström-Lindberg
18. Immune Dysregulation and the Role for Immunotherapy in

Myelodysplastic Syndrome (MDS) 437
Kebede Hussein and David P. Steensma
19. Lenalidomide Therapy in MDS 457

Rami Komrokji, Aristoteles Giagounidis and Alan F. List
20. DNA Methyltransferase Inhibitor Therapy in the Treatment of

Myelodysplastic Syndromes 485
Steven D. Gore
21. Intensive Chemotherapy and Stem Cell Transplantation in Myelodysplastic

Syndromes 497
Guillermo F. Sanz and Theo de Witte
Index . . . . 527
Contributors
Lionel Ades Hôpital Avicenne (Assistance Publique-Hôpitaux de Paris),

Université Paris 13, France
Carlo Aul Medizinische Klinik II (Hämatologie, Onkologie und
Immunologie), St. Johannes-Hospital Duisburg, Duisburg, Germany
John M. Bennett James P. Wilmot Cancer Center and University of Rochester,
Rochester, New York, U.S.A.
Jacqueline Boultwood Leukaemia Research Fund Molecular Haematology
Unit, Nuffield Department of Clinical Laboratory Sciences, John Radcliffe
Hospital, Oxford, U.K.
David T. Bowen Department of Hematology, St James’s Institute of
Oncology, Leeds, U.K.
Pierre Fenaux Hôpital Avicenne (Assistance Publique-Hôpitaux de Paris),
Claude Gardin Hôpital Avicenne (Assistance Publique-Hôpitaux de Paris),
Norbert Gattermann Department of Hematology, Oncology, and Clinical
Immunology, Heinrich-Heine-University, Düsseldorf, Germany
Ulrich Germing Klinik für Hämatologie, Onkologie und klinische
Immunologie, Heinrich-Heine Universität Düsseldorf, Düsseldorf, Germany
Aristoteles Giagounidis St. Johannes Hospital, Medizinische Klinik II, An der
Abtei 7–11, Duisburg, Germany
Lucy A. Godley Department of Medicine and Cancer Research Center,
University of Chicago, Chicago, Illinois, U.S.A.
Steven D. Gore Sidney Kimmel Comprehensive Cancer Center at Johns
Hopkins, Baltimore, Maryland, U.S.A.
Terry J. Hamblin University of Southampton, Consultant Hematologist,
Royal Bournemouth Hospital, Southampton, U.K.
Curtis A. Hanson Department of Laboratory Medicine and Pathology, Mayo
Clinic, Rochester, Minnesota, U.S.A.
vii
viii Contributors
Henrik Hasle Department of Pediatrics, Aarhus University Hospital Skejby,

Aarhus, Denmark
Tomoko Hata Department of Hematology, Atomic Bomb Disease Institute,
Nagasaki University Graduate School of Biomedical Sciences, Sakamoto,
Nagasaki City, Japan
Eva Hellström-Lindberg Karolinska Institutet and Karolinska University
Hospital Huddinge, Stockholm, Sweden
Kebede Hussein Division of Hematology, Department of Medicine, Mayo
Rami Komrokji H. Lee Moffitt Cancer and Research Institute, Tampa,
Florida, U.S.A.
Richard A. Larson Department of Medicine and Cancer Research Center,
Michelle M. Le Beau Department of Medicine and Cancer Research Center,
Alan F. List H. Lee Moffitt Cancer and Research Institute, Tampa, Florida,
U.S.A.
Michael R. Loken Hematologics, Inc., Seattle, Washington, U.S.A.
Luca Malcovati Department of Hematology, University of Pavia Medical
School and Fondazione IRCCS Policlinico San Matteo, Pavia, Italy
Phuong L. Nguyen Department of Laboratory Medicine and Pathology, Mayo
Charlotte M. Niemeyer Division of Paediatric Hematology and Oncology,
Department of Pediatrics and Adolescent Medicine, Albert-Ludwigs-University,
Mathildenstrasse 1, Freiburg, Germany
Harold J. Olney Université de Montréal, CHUM Hôpital Notre-Dame,
Montreal, Quebec, Canada
Attilio Orazi Department of Pathology and Laboratory Medicine, Weill
Medical College of Cornell University, New York, New York, U.S.A.
Andrea Pellagatti Leukaemia Research Fund Molecular Haematology Unit,
Nuffield Department of Clinical Laboratory Sciences, John Radcliffe Hospital,
Oxford, U.K.
Guillermo F. Sanz Department of Hematology, Hospital Universitario La Fe,
Valencia, Spain
David P. Steensma Division of Hematology, Department of Medicine, Mayo
Contributors ix
Masao Tomonaga Department of Hematology, Atomic Bomb Disease

Institute, Nagasaki University Graduate School of Biomedical Sciences,
Sakamoto, Nagasaki City, Japan
James W. Vardiman Department of Pathology, University of Chicago Medical
Center, Chicago, Illinois, U.S.A.
James S. Wainscoat Leukaemia Research Fund Molecular Haematology
Unit, Nuffield Department of Clinical Laboratory Sciences, John Radcliffe
Hospital, Oxford, U.K.
Denise A. Wells Hematologics, Inc., Seattle, Washington, U.S.A.
Theo de Witte Department of Hematology, Radboud University Hospital,
Nijmegen, The Netherlands
Yatato Yoshida The Center for Hematological Diseases, Takeda General
Hospital, Kyoto, Japan
1
The Myelodysplastic Syndromes:
History and Classification
David P. Steensma
Division of Hematology, Department of Medicine, Mayo Clinic, Rochester,
Minnesota, U.S.A.
John M. Bennett
James P. Wilmot Cancer Center and University of Rochester,
Rochester, New York, U.S.A.
INTRODUCTION: WHAT ARE “MYELODYSPLASTIC SYNDROMES”?

The myelodysplastic syndromes (MDS) have probably existed as long as the
human lifespan was stretched beyond the reproductive age. But the tools required
for detection of MDS—accurate hemocytometers, biological stains capable
of highlighting intracellular detail, and microscopes with reduced spherical
aberration—and chromatic distortion—came into widespread use only at the end
of the 19th century. In addition, clinicians rarely attempted bone marrow exami-
nation of living patients for diagnostic purposes until 1929, when a technique for
sternal marrow aspiration was reported by Mikhail Arinkin in Leningrad (1,2).
So the definition of MDS as a discrete clinical syndrome had to wait until the
20th century—just as the molecular solutions to MDS will surely belong to the
21st century, for similar reasons related to evolving technology.
The medical community currently applies the term MDS to a diverse collec-
tion of disorders of blood cell formation. Collectively, these conditions are defined
by three cardinal features. The first common attribute of MDS is bone marrow
failure—i.e., ineffective hematopoiesis (3). Disturbed erythropoiesis is especially
frequent, but neutropenia and thrombocytopenia may also be present. Marrow
1
2 Steensma and Bennett
failure accounts for the nearly universal finding of peripheral blood cytopenias
in MDS, often despite a hypercellular marrow. Second, the abnormal “dysplastic”
appearance of blood and bone marrow cells in MDS represents the morphological
correlate of disordered cell maturation. Observation of characteristic morpho-
logical abnormalities is necessary, but not sufficient, for a diagnosis of MDS to
be made with confidence (4). Finally, acquired, clonally restricted chromosomal
abnormalities can be detected in myeloid cells from most patients with MDS (5),
and clonal evolution due to genomic instability may result in disease progression
to a difficult-to-treat form of acute myeloid leukemia (AML) (6). While half of the
patients with MDS have a normal G-banded karyotype at the time of diagnosis,
newer tools allow the human genome to be scanned at a higher resolution than is
possible with conventional karyotyping, and preliminary experiments with these
technologies suggest that clonally restricted genetic abnormalities will eventually
be found in all MDS cases (7,8).
But these statements are only generalities. The rest of this book explores
the full breadth and depth of MDS, including other common MDS-associated
pathobiological traits. These characteristics include dysregulation of apoptosis,
disordered iron metabolism, dysfunctional heme biosynthesis, and deranged epi-
genetic patterning, just to name a few. With so many inter-case distinctions and
only a few unifying themes, the MDS still commonly confound clinicians, perturb
pathologists, and cause consternation for categorizers seeking a robust and orderly
classification of hematological neoplasms.
EARLY DEVELOPMENTS
As is the case with many other disorders of relatively recent definition, it is
possible to search through old case reports and single-institution series to find
disease descriptions that are consistent with a diagnosis of MDS. MDS was first
recognized as a form of anemia refractory to treatment with hematinics—iron, and,
after 1926, liver extract (9) (Table 1). The concept of MDS as a stubborn yet benign
form of anemia, rather than a dangerous smoldering malignancy, is still widespread
among physicians, even though J. L. Hamilton-Paterson in London and several
French investigators connected refractory anemia and subsequent development of
AML as early as the 1940s (10–14).
Because the authors of early reports often chose to emphasize a single aspect
of MDS (e.g., characteristics of the anemia, cell morphology, or the relationship
to acute leukemia and other diseases) in their published papers, they used many
different terms to describe what they were seeing (Table 1). Obsolete terms to
describe MDS-like conditions include “pseudoaplastic anemia” (15), “refractory
normoblastic anemia” (16), “DiGuglielmo syndrome” (17), “preleukemic anemia”
(10,18), “refractory megaloblastic anemia” (19), “hyperplastic panmyelosis” (20),
“smoldering acute leukemia” (21), and many others. Even in the early 21st century,
issues related to MDS terminology continue to be center of discussions for its
classification and minimal diagnostic criteria (22).
The Myelodysplastic Syndromes: History and Classification 3
Table 1 Selected Important Events in the History of Classification and Treatment of

Myelodysplastic Syndromes (MDS)
Year Event Source
1926 Giovanni Di Guglielmo in Naples describes “acute erythremia,” a (79)

heterogeneous group of disorders of erythropoiesis termed “Di
Guglielmo syndrome” by William Dameshek in 1951; some
cases of Di Guglielmo syndrome would now be called MDS.
1926 George Minot and William Murphy of Boston describe the cure of (9)
pernicious anemia with raw liver, often considered the beginning
of the modern age of hematology; patients who do not respond to
liver therapy are labeled as having “refractory anemia.”
1938 C.P. Rhoads and W. Halsey Barker in New York report 100 cases of (80)
refractory anemia, highlighting the frequency of macrocytosis
and leukopenia, and distinguishing primary cases from patients
with previous exposure to hematological toxins.
1949 Pathologist J. L. Hamilton-Paterson in Edgware (North London) (10)
describes “preleukaemic anemia,” connecting refractory anemia
with the subsequent development of leukemia for the first time in
the English medical literature. Several French investigators come
to similar conclusions.
1953 In the Journal of the American Medical Association, Matthew (18)
Block, Leon Jacobsen, and William Bethard in Chicago report 12
patients with “preleukemic acute human leukemia,” substantially
increasing awareness of the condition in the United States.
1956 Sven-Erik Björkman, a young clinical investigator in Malmö, (81)
describes 4 cases of acquired idiopathic sideroblastic anemia. By
the 1970s, this syndrome would be considered a form of MDS.
1963 Three “smoldering acute leukemia” cases are reported by Jack (21)
Rheingold and colleagues in Washington, D. C., who highlight
the possibility of death from infection prior to “burst[ing] forth
into a more typical picture of acute leukemia.”
1968 A team in Minneapolis reports the first successful human allogeneic (82–84)
bone marrow transplantation (BMT) done for a child with
congenital immunodeficiency. In 1978, a French group reports an
unsuccessful attempt at syngeneic BMT without conditioning for
a patient with “preleukemia,” and in 1979, a German group
reports a more successful result. Allogeneic BMT for MDS is
widespread by the mid-1980s, and some investigators (especially
in Europe) also use autologous BMT.
1976 John Bennett and colleagues of the French-American-British (FAB) (25)
Cooperative Group propose a classification of acute leukemia,
including 2 dysmyelopoietic syndromes: refractory anemia with
excess blasts and chronic myelomonocytic leukemia.
(Continued)
Table 1 Selected Important Events in the History of Classification and Treatment of

Myelodysplastic Syndromes (MDS) (Continued)
Year Event Source
1982 The FAB Cooperative Group publishes a revised and expanded (26)
version of their classification system for MDS (Table 2).
1985 Ghulam Mufti and his colleagues in the United Kingdom devise (37,39,40)
the Bournemouth Score, an influential early system for MDS
risk assessment. This is followed by the Sanz score (1989), the
Düsseldorf score (1992), Lille score (1996), and others,
eventually culminating in the International Prognostic Scoring
System (IPSS) (1997) (Tables 4–6).
1988 The First International Symposium on MDS is held in Innsbruck.
Subsequent symposia took place in Bournemouth (1991),
Chicago (1994), Barcelona (1997), Prague (1999), Stockholm
(2001), Paris (2003), Nagasaki (2005), and Florence (2007);
future meetings are scheduled in Patras (2009), Edinburgh
(2011), and Berlin (2013).
1994 Incorporation of the Myelodysplastic Syndromes Foundation
(http://www.mds-foundation.org/), an advocacy group for
patients with MDS and their families, which awards qualifying
health care facilities an “MDS Center of Excellence”
distinction. The Foundation is based in Crosswicks, New Jersey;
a European office opened in London in 2007.
1997 Peter Greenberg of Stanford University and a multinational group (85)
of investigators publish the IPSS, which for more than a decade
is the most widely used prognostic tool for de novo MDS.
2000 An International Working Group (IWG) publishes a proposal for (86)
standardized response criteria for clinical trials enrolling MDS
patients.
2001 The World Health Organization (WHO) publishes the final version (54)
of a revised “Classification of Tumours of the Haematopoietic
and Lymphoid Tissues,” which reclassified MDS using the
backbone of the 1982 FAB scheme.
2003 An international group proposes a classification of pediatric MDS. (63)
2004 Approval of azacitidine, the first therapy indicated for MDS (all (87)
FAB subtypes), by the Food and Drug Administration (FDA) in
the United States.
2005 Approval of lenalidomide by the FDA for lower-risk MDS patients (88)
with del(5q).
2006 Approval of decitabine by the FDA for all FAB subtypes of MDS. (89)
2006 Revision of the IWG standardized response criteria. (90)
2008 Revision of WHO classification of MDS.
One of the first thorough literature reviews of “preleukemia” was published

in 1973 by Matti Saarni (Tampere, Finland) and James Linman (Portland,
Oregon), who had been together at Mayo Clinic in Rochester, Minnesota, in
1970 (23). The authors struggled to gather 143 published cases, because the
classification of many cases were difficult due to the varying case definitions
employed and the incomplete nature of early reports. Although Saarni and Linman
emphasized the uniformity of many clinical features of what would later be
known as MDS, including a consistently high risk of progression to leukemia, the
designation of “preleukemia” that was in vogue at that time soon fell out of favor.
Once investigators realized that only a minority (∼25%) of patients with MDS
develops AML, and that the most common cause of MDS-associated death is
actually infection due to neutropenia and neutrophil dysfunction, “preleukemia”
seemed an incomplete descriptor. As early as 1976, the question of whether all
patients with MDS would eventually develop leukemia if they were to live long
enough was vigorously debated (24)—a question as esoteric and difficult as the
old kōan about whether a tree makes a sound when it falls unwitnessed deep in a
forest.
EVOLVING CRITERIA FOR MDS DIAGNOSIS AND RISK ASSESSMENT

The French-American-British (FAB) Classification System
The first formal diagnostic criteria for MDS were developed by the French-
American-British (FAB) Cooperative Group during a series of meetings in 1974–
1975, where investigators reviewed representative cases and attempted to refine the
nomenclature and classification of the acute leukemias. The FAB group included
seven members: John M. Bennett (Rochester, New York), Daniel Catovsky
(London), Marie-Thérèse Daniel (Paris), Georges Flandrin (Paris), Harvey R.
Gralnick (Bethesda, Maryland), and the late David A.G. Galton (London; died
2006), and Claude S. Sultan (Paris; died 1992).
Ultimately, the FAB group published 14 papers between 1976 and 2002,
all of them concerning the classification of various hematological diseases. In
the first of these publications (25), the authors identified two broad categories
of “dysmyelopoietic syndrome” that could be confused with acute leukemia—
conditions they termed “refractory anemia with excess of blasts (RAEB)” and
“chronic myelomonocytic leukemia (CMML).” RAEB was defined in the FAB’s
1976 report as dyserythropoietic changes in a hypercellular marrow containing
10% to 30% blasts and promyelocytes. Other key features of RAEB included
erythropoietic hyperplasia (with or without the presence of ringed sideroblasts)
and neutropenia, often with hypogranular granulocytes. In contrast, CMML was
defined by peripheral blood monocytosis (a threshold of ⬎1000/␮L was chosen to
define monocytosis), again in a hypercellular marrow with 10% and 30% blasts and
promyelocytes. Additional key features of CMML included increased numbers
of promonocytes and monocytes in the marrow and elevated serum lysozyme
concentrations (lysozyme is a leukocyte enzyme that was formerly used as a
surrogate marker for monocyte/macrophage activity). Finally, the FAB group’s

initial publication mentioned, “a range of conditions for which an immediate
recommendation to start [antileukemic] therapy cannot be made or may not be
indicated. These are disorders associated with bone-marrow hypercellularity in
which confusion with acute myeloid leukemia is possible” (10).
In a seminal 1982 follow-up report (26), the FAB group chose to use the
term “MDS” rather than “dysmyelopoietic syndrome.” The FAB investigators also
acknowledged that in the 6 years since their first publication, it had “become clear
that the range of morphologic appearances in the PB [peripheral blood] and BM
[bone marrow] consistent with a diagnosis of MDS is very wide, that there is a
great variation in the risk of transformation to a blastic phase, and that the risk
appears to be correlated with the morphologic features.” Another limitation of the
1976 report subsequently addressed in 1982 was that the two broad categories of
dysmyelopoietic syndrome defined earlier—that is, RAEB and CMML—did not
incorporate well-defined features for predicting progression to leukemia. This was
an overt recognition that the ideal MDS classification system is not merely a roster
of categories useful for epidemiological purposes, but also has the power to predict
median survival and the risk of AML progression, thus allowing stratification of
patients for therapy.
In order to try to achieve these goals, between 1980 and 1982, the FAB coop-
erative group examined many more cases of MDS, trying to determine whether
morphological features could be used to better differentiate subtypes according
to the likelihood of AML progression. The outcome of this review was the FAB
group’s 1982 revised classification of MDS (Table 2) (26). The 1982 revision
added formal definitions for three new MDS subtypes: refractory anemia (RA),
refractory anemia with ringed sideroblasts (RARS), and refractory anemia with
excess blasts in transformation (RAEB-t). In addition, the FAB refined their pre-
vious definitions of RAEB and CMML.
The revised RAEB and CMML categories excluded cases with 21% to
30% marrow blasts, and the RAEB category now included cases with 5% to 9%
marrow blasts (as well as retaining those with 10–20% blasts). The 1982 revision
also clearly defined AML as ⬎30% marrow blasts regardless of the presence
of dysmyelopoiesis. The addition of RARS to the revised FAB classification
was based on evolving understanding of the points of overlap between acquired
idiopathic sideroblastic anemia and the various forms of MDS (27), while the
15% ringed sideroblast diagnostic threshold resulted from a report by Sultan and
colleagues indicated that regardless of the proportion of blasts, patients with more
than 20% ringed sideroblasts have a natural history distinct from those with fewer
ringed sideroblasts (28).
One of the most useful parts of the FAB group’s 1982 report was the inclusion
of a roster of specific cytological and histochemical features that morphologists
could assess when analyzing possible MDS cases. These features, which made
MDS classification more reproducible amongst morphologists globally, are out-
lined in Table 3. Contemporary MDS diagnosis is discussed more thoroughly in
Chapter 9.
Table 2 The 1982 French-American-British (FAB) Cooperative Group Revised

Classification of the Myelodysplastic Syndromes (26)
Absolute
Myeloblasts Myeloblasts Ringed monocytes Auer rods
in periph- in bone siderob- in present in
eral blood marrow lasts peripheral bone
MDS subtype (%) (%) (%) blood marrow?
Refractory ⬍1 ⬍5 ≤15 - No
anemia
(RA)
Refractory ⬍1 ⬍5 ⬎15 - No
anemia with
ringed
sideroblasts
(RARS)
Refractory ⬍5 5–20 - - No
anemia with
excess blasts
(RAEB)
Refractory ⬍5 21–30 - - Yes or no
anemia with
excess blasts
in transfor-
mation
(RAEB-t)
Chronic ⬍5 ≤20 - ⬎1 × 109 /L No
myelomono-
cytic
leukemia
(CMML)
Limitations of the 1982 FAB Classification

The FAB group’s revised classification system proved to be a milestone in the
study of MDS. These definitions laid the conceptual groundwork for further study
of the biology of MDS, and the system’s utility and prognostic value were soon
validated by multiple groups (29–31). For more than 15 years, the 1982 FAB
classification was the primary system used around the world for MDS diagnosis
and risk assessment.
Despite these successes, growing experience with the revised FAB classi-
fication highlighted several limitations. Hypoplastic MDS (32), childhood MDS
(33,34), treatment-related MDS, and difficult-to-categorize transitional forms were
not readily classified using the FAB system (35). Marrow fibrosis was also not
mentioned, although fibrosis was subsequently associated with poorer survival in
Table 3 Important Cell Morphological Features to Assess in Suspected MDS Cases

Proportion of bone marrow and peripheral blood blasts on manual aspirate differential
Dyserythropoietic features
Proportion of ringed sideroblasts on marrow aspirate prussian blue reaction
Circulating nucleated red cells
Multi-nuclearity in erythroid precursors
Nuclear fragments
Nuclear and cytoplasmic morphological abnormalities
Proportion of erythroblasts
Dysgranulopoietic features
Nuclear abnormalities
Hypogranular cytoplasm
Auer rods
Dysmegakaryocytopoietic features
Micromegakaryocytes
Large mononuclear forms
Multiple small nuclei
Reduced numbers
Monocyte count
Marrow cellularity (age-adjusted)
patients who would otherwise fit into lower risk groups (36). In addition, within
each FAB class, variability in patient survival still proved to be quite large, with
some patients living many years with minimal treatment while others developed
problems quickly. By the early 1990s, it had become clear that additional objective
prognostic criteria were needed in order to risk-stratify patients with MDS more
accurately.
Assessment of Prognosis: Steps Towards the International Prognostic

Scoring System (IPSS)
In order to address the need for a better prognostic tool, between 1985 and 1996, a
number of investigators retrospectively examined their databases of MDS patients,
attempting to assess the strengths and weaknesses of the FAB classification and to
derive new scoring systems that might better predict patient outcomes (31,37–42).
Some of the important prognostic features defined by these analyses are found in
Table 4. Collectively, these retrospective studies demonstrated the following:
1. In multivariate regression analyses, a poor prognosis was apparent for older

patients (⬎60 years) diagnosed with FAB subtypes RAEB and RAEB-t,
though the outcome of those with RAEB (5–20% marrow blasts) was more
variable. Other high-risk features included increasing proportion of marrow
blasts (≥5%), and increasing number and degree of peripheral blood cytope-
nias, particularly a platelet count of ⬍100 × 109 /L.
Table 4 Comparison of Several Prognostic Systems in MDS Proposed After the 1982 FAB Classification
Prognostic system (lead author)
Dutch Japanese German Lille IPSS

Bournemouth (Kerkhofs) Spanish Düsseldorf (Toyama) (Maschek) (Morel) (Greenberg)
(Mufti) (37) (30) (Sanz) (39) (Aul) (40) (91) (42) (41) (85)
Year of 1985 1987 1989 1992 1993 1994 1996 1997

publication
Number of 141 237 370 235 401 569 203 816
patients
assessed
Key adverse Marrow blast FAB RAEB or Increasing Marrow FAEB FAEB RAEB Increasing Increasing
prognostic proportion RAEB-t marrow blasts RAEB or or RAEB-t marrow marrow blast
markers ≥5% (RAEB-t Increased blast ≥5% RAEB-t Increasing blast proportion
especially bad) marrow or proportion LDH ⬎ 200 Complex marrow proportion Complex
Platelets blood blast Low platelet U/L karyotype blast Complex karyotype or
⬍100 × 109 /L proportion count Hemoglobin Low blood proportion karyotype chromosome
Neutrophils All Low ≤9 g/dL counts Presence of Low platelet 7 abnormality
⬍2.5 × 109 /L metaphases hemoglobin Platelets ALIP count Number of
The Myelodysplastic Syndromes: History and Classification
Hemoglobin abnormal Low white ≤100 × Presence of Hemoglobin cytopenias

⬍10 g/dL on count 109 /L myelofi- ⬍10 g/dL Age ⬎ 60
karyotyping Older age brosis Low or high
white count
Age ⬎60
Only series with >100 MDS patients are included.

Abbreviations: IPSS, International Prognostic Scoring System; FAB, French-American-British; RAEB, refractory anemia with excess blasts; RAEB-t, refractory
anemia with excess blasts in transformation; ALIP, abnormal localization of immature myeloid precursors.
9
2. The presence of a specific morphological pattern in the bone marrow, abnormal

localization of immature myeloid precursors (ALIP), correlated with signif-
icantly shorter survival, higher risk FAB subtypes, increased marrow blast
proportion, and increased risk of AML evolution (42–44).
3. Patients with RA were found to have a modestly increased frequency of
evolution to AML relative to those diagnosed with RARS, especially if the
dysplasia in RARS was restricted to the erythroid lineage (45).
Cytogenetic analysis grew increasingly common in clinical practice through-
out the 1980s and 1990s, and during this time the repertoire of MDS-associated
karyotypic abnormalities expanded markedly. Investigators also began to appre-
ciate the importance of cytogenetic results in predicting MDS outcome (46–49).
However, before the late 1990s, few multi-institutional studies had addressed
cytogenetic parameters as predictors of clinical outcomes for MDS patients.
The IPSS
In 1996, more than 15 prominent physicians from Europe, Japan, and the United
States convened as part of an International MDS Risk Assessment Workshop
(IMRAW) in order to address the continuing need for a reliable MDS prog-
nostic tool. The investigators’ task was to analyze multiple clinical, demographic,
hematological, and cytogenetic variables in 816 primary MDS patients from seven
large previously reported risk-based studies that had generated prognostic systems
(Table 4). Patients with MDS as a consequence of prior exposure to radiotherapy
or cytotoxic chemotherapy were excluded, and most patients (⬎85%) included in
the IMRAW cohort had been treated with supportive care alone, without disease-
modifying therapy. Data on survival were available from all 816 patients, and
AML transformation data were available from 759 patients (93%).
Univariate analysis showed that the major variables predicting poorer out-
come in MDS were a higher risk FAB classification, increased percentage of bone
marrow blasts, increased number of cytopenias (2 or 3 cytopenias were found to be
worse than 0 or 1), high-risk cytogenetic pattern (as defined below), age ⬎60 years,
and male gender. Variables predicting AML evolution included FAB classification,
percentage of marrow blasts, cytogenetic pattern, and number of cytopenias.
With respect to the karyotype, the IPSS investigators found that complex
cytogenetic abnormalities (≥3 chromosomal anomalies) or chromosome 7 anoma-
lies were associated with the worst outcomes. Better survival was observed in
patients with normal cytogenetics, with isolated interstitial deletions of the long
arm of chromosomes 5 or 20 [del(5q) or del(20q)], or isolated loss of the Y chro-
mosome (−Y). All other cytogenetic patterns were designated “intermediate,” but
since many less common karyotypes were present in too few patients (⬍5) in the
IMRAW cohort to be able to make a valid assessment of risk, a better term might
have been “indeterminate.”
On multivariate analysis, five variables maintained independent predictive
ability for both overall survival and AML transformation: percentage of marrow
Table 5 The 1997 International Prognostic Scoring System (IPSS) for Myelodysplastic
Syndromes (85)
Category score (sum all 3 for overall IPSS score)

Prognostic
factor 0 (best) 0.5 1 1.5 2 (worst)
Marrow blasts ⬍5 5–10 – 11–20 21–30b

(%)
Karyotype Good: normal, Intermediate: Poor: abnormal – –
isolated −Y, all chromosome 7,
isolated karyotypes or a complex
del(5q), or not defined karyotype (3 or
isolated as good or more
del(20q) poor anomalies)
Peripheral 0 or 1 2 or 3 – – –
blood
cytopeniasa
a IPSS definition of peripheral blood cytopenias: Hemoglobin, <10 g/dL; absolute neutrophil count
<1800/mm3 ; platelet count <100,000/mm3 .
b No longer considered MDS (redefined as AML by WHO in 2001).
blasts, number of cytopenias, cytogenetic subgroup (good, intermediate, and

poor, as defined above), age, and gender (all p ⬍ 0.0001) (Table 5). Marrow blast
proportion, the number of cytopenias, and the cytogenetic risk group were then
chosen as the final set of three variables to construct a prognostic model (Table 6).
These variables were selected on both clinical and statistical grounds, as they
were statistically significant for both survival and AML evolution.
The IMRAW investigators assigned a risk score for each of these three
prognostic variables, with the most weight given to marrow blast proportion. These
numbers were used to develop a scoring system, the IPSS (Table 6). Using the IPSS,
patients could be stratified into four well-defined risk groups (low, intermediate-1,
intermediate-2, and high risk) with respect to survival and AML evolution.
The IPSS showed greater discrimination between good- and poor-prognosis
patients than three previously popular published prognostic systems [i.e., the FAB
classification (26), the Spanish system (39), and the Lille score (41)], despite
narrower entry criteria (i.e., patients whose MDS had been treated were largely
excluded in the IMRAW analysis). The IPSS also better clarified the role of age as
a prognostic factor in MDS—younger patients in the lower 2 IPSS risk groups had
a greater probability of survival than older patients in those risk groups, but there
was little difference between young and old in the higher risk categories. The lack
of importance of age in higher risk MDS may be due to the greater propensity of
the higher IPSS risk group patients to develop AML, which is an adverse clinical
outcome regardless of age.
12
Table 6 Risk Stratification of International Prognostic Scoring System (IPSS)

Time until 25% of
Median survival (yr) for Median survival (yr) for surviving patients in
Median survival patients 60-years old patients over 60-years category developed
Risk category Total score (yr) or younger (n = 205) old (n = 611) leukemia (yr)
Low risk 0 5.7 11.8 4.8 9.4

Intermediate-1 0.5 or 1.0 3.5 5.2 2.7 3.3
(int-1)
Intermediate-2 1.5 or 2.0 1.2 1.8 1.1 1.1
(int-2)
High 2.5 or greater 0.4 0.3 0.5 0.2
Scoring system: a point value from 0 to 2.0 is determined for each of the 3 prognostic factors in Table 4, and the 3 values are summed to obtain the total IPSS
score. Interpretation is given in the Table.
Steensma and Bennett
The sole exception to greater discrimination by the IPSS was with respect
to CMML, as patients diagnosed with CMML were present in all the 4 IPSS risk
groups, and their outcomes assessment by IPSS proved to be quite inaccurate
(50). Even today, a highly accurate prognostic system for CMML remains elusive,
though the 2002 MD Anderson system is a useful starting point (51). The MD
Anderson CMML classification system includes four adverse prognostic features:
a hemoglobin level less than 12.0 g/dL, the presence of circulating immature
myeloid cells, an absolute lymphocyte count above 2.5 × 109 /L, and marrow blast
proportion greater than 10%.
More than 10 years after its publication, the IPSS remains the most compre-
hensive and widely used prognostic scoring system developed for MDS, due to the
broad geographic distribution of the patients evaluated by the IMRAW, stringent
entry criteria, and central review of cytogenetics. The IPSS has been used to clas-
sify patients for many clinical trials of novel therapies, and it is helpful in day-to-
day clinical practice for risk assessment and choice of therapy. However, the IPSS
does have a number of limitations, which will be discussed in the later sections.
The World Health Organization (WHO) Classifications of MDS

In the late 1990s, there was a major effort to reclassify all hematopoietic and
lymphoid malignancies under the auspices of the WHO, as the WHO had not
updated its neoplasia classification system since 1981 (52). The goal of this
classification modernization effort was to better incorporate newly defined clinical
entities (e.g., the various subtypes of lymphoma identified by flow cytometry and
molecular techniques in the late 1980s and 1990s) and to produce more homoge-
neous and clinically relevant disease categories. To this end, beginning with a 1997
meeting at Airlie House in northern Virginia, more than 100 hematologists and
hematopathologists from around the world debated the most useful classification
of hematopoietic and lymphoid tumors (53). The final result of this effort was
published in 2001, and included a revised MDS classification system built on the
backbone of the FAB system (Table 7) (54).
The salient features of the 2001 WHO system were elimination of the
RAEB-t category, redefinition of AML as ≥20% marrow or peripheral blood
myeloblasts instead of ⬎30%, recognition that multilineage dysplasia in indi-
viduals previously classified as RA or RARS worsens prognosis (55), splitting
of RAEB into two categories (type 1, 5–9% blasts and type 2, 10–19% blasts),
and movement of CMML out of MDS and into a separate myeloproliferative
disorder (MPD)/MDS overlap category. MDS associated with isolated del(5q)
was established as a separate category within the MDS cluster, and the provisional
entity of RARS with thrombocytosis (RARS-T) was included as a MPD/MDS
overlap syndrome. Although discrete categories for hypoplastic MDS, MDS with
fibrosis, and treatment-related MDS were not created, a separate designation was
added for “unclassifiable MDS” not clearly fitting other categories, so it is likely
Table 7 2001 World Health Organization (WHO) Classification of Neoplastic Diseases

of the Hematopoietic and Lymphoid Tissues (54)
Categories relevant to the myelodysplastic Approximate proportion of

syndrome (MDS) MDS cases (%)(92)
Myelodysplastic syndromes
Refractory anemia
With ringed sideroblasts (pure sideroblastic anemia, 5–10
RARS)
Without ringed sideroblasts (RA) 10–15
Refractory cytopenia with multilineage dysplasia
With ringed sideroblasts (RCMD-RS) 10–15
Without ringed sideroblasts (RCMD) 25–30
Refractory anemia with excess blasts
With 5–9% myeloblasts (RAEB-1) 20
With 10–19% myeloblasts (RAEB-2) 20
5q− syndromea ⬍5
Myelodysplastic syndrome, unclassifiable variable
Myelodysplastic/myeloproliferative syndromes
Chronic myelomonocytic leukemia (CMML)b
Atypical chronic myelogenous leukemia
Juvenile myelomonocytic leukemia
Relevant acute myeloid leukemia (AML) categoriesc
AML with multilineage dysplasia
With prior myelodysplastic syndrome
Without prior myelodysplastic syndrome
AML and myelodysplastic syndromes, therapy-related
Alkylating-agent related
Epipodophyllotoxin-related
Other types
a Like RAEB, CMML can also be subdivided based on myeloblast count.
b The 5q− syndrome is narrowly defined by the WHO to include only cases with de novo isolated
del(5q) and the characteristic morphologic findings of hypolobated megakaryocytes and less than 5%
marrow myeloblasts (93).
c AML is defined by the WHO as a myeloid disorder with ≥20% myeloblasts, or with any blast count
if clonal karyotypic findings include inv(16), t(8;21) or t(15;17).
that these entities are more uniformly categorized in the WHO system than in the
1982 FAB system.
The WHO system was quickly validated by several groups of investigators
(56), including those who control the Düsseldorf MDS registry, which includes
data on 1600 patients diagnosed with primary MDS seen between 1970 and 1999
(57). These retrospective data demonstrated the capacity of the WHO classification
system to identify relatively poorer prognosis patients, such as those having multi-
lineage dysplasia or 10% to 19% blasts, who would have been grouped with better
prognosis patients (RA/RARS and RAEB with 5% and 9% blasts, respectively) in
the revised FAB system. Also, the refractory cytopenia with multilineage dysplasia
(RCMD) categories allowed for inclusion of a large group of MDS patients who
were formally unclassifiable using the 1982 FAB classification, since the FAB sys-
tem stressed the predominance of anemia in RA and RARS cases and minimized
the importance of other cytopenias. In addition, at least in the Düsseldorf registry,
the troublesome CMML group was almost completely eliminated from MDS by
reclassification. Overall, the WHO MDS criteria reduced the number of unclas-
sifiable cases, provided more homogeneity within MDS categories, and yielded
more uniform and accurate prognostic information, at least when applied retro-
spectively. Prospective studies validating the 2001 WHO criteria have been slow in
coming, but are still necessary to provide data for future revision of these criteria.
The 2008 WHO Revision

As with all classification systems, real-world testing also highlighted some limita-
tions of the 2001 WHO MDS classification (58). For instance, some investigators
questioned the subjective nature of the determination of dysplasia in specific cell
lineages. There is considerable inter-observer variability in the threshold required
to label a lineage as dysplastic, and the WHO did not mandate the use of spe-
cial histochemical stains (e.g., dual esterase) or flow cytometry—techniques which
sometimes highlight dysplasia in myeloid cells that is difficult to discern on routine
staining (59). Occasional patients present with MDS-associated morphological
changes and a karyotypic abnormality, yet have neutropenia or thrombocytopenia
with little anemia (60), and the 2001 WHO system does not classify these patients
well. Issues regarding pediatric MDS classification and classification of atypical
cases (e.g., hypoplastic MDS, MDS with fibrosis, a normal appearing marrow
with a cytogenetic abnormality) continue to be troublesome.
As a consequence of these and other concerns, the WHO classification
system is currently being revised. An updated classification will be published
in the fall of 2008, but, as mentioned by Dr. James Vardiman (one of the
classification’s lead authors) in chapter 9 of this book, the 2008 revision will not
offer major changes from the current nosology. RARS with thrombocytosis will
move from a provisional entity to a recognized MPD/MDS overlap subtype, based
on the finding of a high frequency of JAK2 V617F in patients with a high platelet
count and ⬎15% ringed sideroblasts (JAK2 V617F is a mutation more commonly
associated with MPD than MDS) (61,62). New categories of “refractory
thrombocytopenia” and “refractory neutropenia” (for patients with cytopenias
and dysplasia, but without anemia) will be added under a category of “Refractory
Unilineage Cytopenia(s)” that will also include RA. The distinction between
RCMD and RCMD with ringed sideroblasts has proven questionable, and these
two entities will be placed into a single subtype in the revised classification. On
historical and etymological grounds, the name “ring sideroblast” will be preferred
to “ringed sideroblasts.” Finally, the presence of cytopenia(s) associated with a
characteristic MDS-associated cytogenetic abnormality, such as abnormalities of
chromosome 5, 7, 11q, or 20q—even in the absence of significant dysplasia (i.e.,

⬎10% abnormal-appearing cells in any lineage) or an increase in marrow blast
proportion (≥5%)—will be considered MDS, unclassifiable.
The pediatric MDS community has taken matters into their own hands
and has generated an independent classification system for childhood cases
(63). In future revisions of both the standard WHO system and the pediatric
MDS classification, molecular markers and flow cytometry will increasingly be
incorporated.
Currently, there are a small number of patients who present with cytopenias
that could be caused by MDS, but who have no karyotypic or definitive marrow
morphologic evidence of MDS (64). These patients are often labeled as having
“Idiopathic cytopenia(s) of undetermined (or unknown) significance” (ICUS) (4),
and are sometimes informally called “not quite MDS” or “not yet MDS.” Careful
follow-up of these patients is necessary, as some will eventually go on to develop
overt MDS.
Beyond the IPSS

Since the publication of the IPSS in 1997, a number of investigators have proposed
modifications of the system, or have suggested inclusion of additional prognostic
markers in MDS risk assessment. Some of the proposed high-risk markers are
molecular abnormalities, such as TP53 mutations (65) and NRAS/KRAS mutations
(66); however, testing for these has not yet been incorporated into routine clinical
practice. Other easily obtainable indicators, such as the mean platelet volume, were
advocated as prognostically useful by some investigators (67) but then could not
be confirmed by others (68,69), so their true importance is unclear. Furthermore,
patients with secondary, exposure-related MDS were not included in the IMRAW
cohort and are not classified well by the IPSS, and there is also uncertainty about
how well the IPSS assesses risk for patients poorly classified by WHO and FAB
classification systems.
One area where there is clearly a room for improvement in the IPSS is with
respect to the roster of cytogenetic abnormalities included in the system. Since the
original publication of the IPSS, it has become clear that some of the unspecified
karyotypes that by default were assigned to the IPSS “intermediate” (or “indeter-
minant”) cytogenetic risk group are actually associated with a relatively benign
clinical course (e.g., trisomy 15, deletions of chromosome 9q), whereas others
have a uniformly poor outlook [e.g., t(3;21); abnormalities of chromosome 17p]
(5,70). Recently, a German-Austrian consortium led by Detlef Haase in Göttingen
has combined forces with the Leukemia Group at the MD Anderson Cancer Cen-
ter in Houston, Texas, to examine the importance of various karyotypes in a data
set including 3860 MDS patients, 55% of whom had cytogenetic abnormalities
(71). These efforts defined a much larger set of good-risk, intermediate-risk, and
poor-risk karyotypes to assist clinicians in risk assessment (Table 8) (5).
Table 8 Newer Cytogenetic Risk Groups Defined by the German-Austrian Consortium
Good risk Intermediate risk Poor risk
Anomaly Median survival (mo) Anomaly Median survival (mo) Anomaly Median survival (mo)
del(9q) NR del(11q) 26.1 4–6 abnormalities 8.7

Abnormal 15 NR +8 23.0 ⬎6 abnormalities 5.0
del(12p) 108 t(11q23) 20.0 t(5q) 4.4
+21 101 Abnormal 3q 19.9
del(5q) 77.2 +19 19.8
del(20q) 71.0 del(7q) 19.0
−X 56.4 3 abnormalities 17.1
Normal 53.4 −5 14.6
−Y 39.0 −7 14.0
Abnormal 1 34.7
t(7q) 34.7
The Myelodysplastic Syndromes: History and Classification
t(11q) 32.1
−21 32.0
The data are based on 2072 patients, 1084 with cytogenetic abnormalities (5). Boldface type indicates the most common cytogenetic findings. NR, median
survival not reached during the duration of the study. An updated version, merged with MD Anderson dataset, is pending.
17
Table 9 World Health Organization Classification–Based Prognostic Scoring System

(WPSS) (76)
Score (points)
Prognostic variable 0 1 2 3
WHO MDS category RA, RARS, 5q− RCMD, RCMD-RS RAEB-1 RAEB-2
IPSS karyotype class Good Intermediate Poor -
Transfusion No Yes - -
requiring?a
Score (sum) WPSS risk category Median survival (mo)b
0 Very low 141

1 Low 66
2 Intermediate 48
3–4 High 26
5–6 Very High 9
a RBC transfusion dependency was defined as having at least one RBC transfusion every 8 wk over a
period of 4 mo.
b Validation cohort. Median survival in the learning cohort was slightly shorter for all but the very high
risk category patients, ranging from 12 to 103 mo.

Abbreviations: WHO, 2001 World Health Organization classification of hematopoietic tumors; MDS,
myelodysplastic syndrome; RA, refractory anemia; RARS, refractory anemia with ringed sideroblasts;
5q−, myelodysplastic syndrome with isolated del(5q) and marrow blasts less than 5%; RCMD,
refractory cytopenia with multilineage dysplasia; RCMD-RS, refractory cytopenia with multilineage
dysplasia and ringed sideroblasts; RAEB-1, refractory anemia with excess of blasts-1; RAEB-2,
refractory anemia with excess of blasts-2.
Patients’ IPSS risk scores increase based on their number of peripheral blood
cytopenias (0, 1, 2, or 3 lineages affected). However, the severity of cytopenias
may be just as important in determining the natural history in MDS as the number
of cytopenias that are present. Several groups have recently suggested that a low
platelet count is of particular importance in MDS prognosis, and the degree of
thrombocytopenia may help refine prognosis further within IPSS risk categories
(68,72,73). More work is needed in this area in order to define the prognostic
importance of both quantitative and qualitative defects of peripheral blood cells.
The IPSS is heavily weighted to the marrow blast proportion, with the num-
ber of cytopenias and cytogenetic risk category contributing to a lesser degree. A
number of investigators have questioned this characteristic, and data are emerging
that high-risk cytogenetic findings may actually be as important, or even more
important, than the other two components of the IPSS (74). For example, com-
pared to patients with a normal karyotype and ⬍5% blasts, patients with a complex
karyotype that includes abnormalities of chromosome 5 or 7 have a hazard ratio
of 3.88 for death, whereas those with 11% to 20% blasts have a hazard ratio of
only 1.64 (75).
The IPSS includes patients with 21% to 30% marrow blasts (currently
considered AML) who would no longer be classified as MDS by the WHO.
Recognizing this limitation, a European group led by Luca Malcovati and Mario
Cazzola in Pavia, Italy, has proposed an IPSS update (76) that incorporates the
2001 WHO classification as well as an additional independent poor prognostic
factor in MDS, transfusion dependence (77). This so-called “WHO classification–
based Prognostic Scoring System” (WPSS) divides patients into five groups with
median survival time ranging from less than 1 year to more than 8 years (Table 9).
Although there is still considerable intra-category variability in life expectancy
using the WPSS, this refinement in prognostication power may be helpful in
discussions with some patients.
A few other individual prognostic markers in MDS have been proposed
since the IPSS, but are awaiting validation. One of the most prominent is the
proposed independent prognostic importance of a high ferritin level (see chap. 7).
A retrospective analysis of 467 patients diagnosed with MDS at Pavia between
1992 and 2002 indicated that a ferritin over 1000 ng/mL was an independent risk
factor for death in patients with lower risk disease (RA and RARS), but that
ferritin’s prognostic value was not signficant in higher risk patients, presumably
because of the more potent impact of other adverse disease features (74). These
results await confirmation by other groups.
IS IT APPROPRIATE TO CLASSIFY MDS AS “CANCER”?

Because MDS were first described in the early 20th century as difficult form
of anemia, the perception has lingered that MDS are a benign group of bone
marrow disorders rather than a set of smoldering malignancies that markedly
reduce patients’ expected survival. Given this history, it is not surprising that the
National Cancer Institute’s (NCI) Survey, Epidemiology, and End Results (SEER)
database and other cancer registries did not track MDS cases until very recently. It
is still possible to find reassuring statements on many Internet sites to the effect that
MDS are not forms of cancer, and that it has a relatively good outlook compared
to other hematological disorders such as leukemia.
Although there are conditions currently captured under the MDS umbrella
that are best described as nonneoplastic [e.g., pure sideroblastic anemia (45), which
sometimes results from germline mutations in genes such as ALAS2], the parallels
between MDS and other hematological malignancies are striking (78). MDS are
clonal disorders that are characterized by cytogenetic and molecular abnormalities
that overlap substantially with AML. The median life expectancy overall is only
2–3 years, and many patients require intensive therapy with cytotoxic agents. In
the United States, most patients with MDS are treated by oncologists, the NCI
funds MDS research, and oncology journals routinely publish papers with an MDS
focus. Therefore, for the time being, it seems best to continue to classify MDS
with other malignancies.
CONCLUSION
There is still much work to be done toward defining better classification and
prognostic systems for MDS—systems that are useful both for individual patient
risk assessment and to help clinicians choose therapy of appropriate intensity.
Careful assessment of cell morphology in blood and bone marrow will remain
the starting point of MDS diagnosis for at least the next 10 years, but it seems
likely that in the longer term, the largest gains will come from the new molecular
technologies now widely available to investigators. The authors’ hope that
prognostic scoring systems like the IPSS will become obsolete not just because
of refinements in molecular classification of MDS, but also because the better
understanding of disease biology and definition of new therapeutic targets leads
to dramatic improvement in MDS treatments. If the new molecular tools live up
to their promise, one day the 1997 IPSS will be viewed as hopelessly pessimistic.
REFERENCES
1. Arinkin MI. Die intravitale untersuchungsmethodik des knochenmarks. Folia Haema-
tol 1929; 38:233–240.
2. Parapia LA. Trepanning or trephines: A history of bone marrow biopsy. Br J Haematol
2007; 139(1):14–19.
3. Bagby GC, Lipton JM, Sloand EM, et al. Marrow failure. Hematology (Am Soc
Hematol Educ Program) 2004; 318–336.
4. Valent P, Horny HP, Bennett JM, et al. Definitions and standards in the diagnosis and
treatment of the myelodysplastic syndromes: Consensus statements and report from
a working conference. Leuk Res 2007; 31(6):727–736.
5. Haase D, Germing U, Schanz J, et al. New insights into the prognostic impact of the
karyotype in MDS and correlation with subtypes: Evidence from a core dataset of
2124 patients. Blood 2007; 110(13):4385–4395.
6. Bagby GC, Meyers G. Bone marrow failure as a risk factor for clonal evolution:
Prospects for leukemia prevention. Hematology (Am Soc Hematol Educ Program)
2007:40–46.
7. Gondek LP, Haddad AS, O’Keefe CL, et al. Detection of cryptic chromosomal lesions
including acquired segmental uniparental disomy in advanced and low-risk myelodys-
plastic syndromes. Exp Hematol 2007; 35(11):1728–1738.
8. Mohamedali A, Gaken J, Twine NA, et al. Prevalence and prognostic significance of
allelic imbalance by single-nucleotide polymorphism analysis in low-risk myelodys-
plastic syndromes. Blood 2007; 110(9):3365–3373.
9. Minot GR, Murphy WP. Treatment of pernicious anemia by a special diet. J Am Med
Assoc 1926; 87:470–476.
10. Hamilton-Paterson JL. Pre-leukaemic anaemia. Acta Haematol (Basel) 1949; 2:309–
316.
11. Chevallier P. Sur la terminologie des leucoses et les affections-frontieres: les odoleu-
coses. Sangre (Barc) 1942–1943; 15:587–593.
12. Dustin MP. Myelose aplasique récidivante chez un enfant. Crise leucémoide aigue
terminale. Sangre 1944; 16:394.
13. Marchal G, Deprez V, Blanc G. Syndrome agranulocytaire par chimiothérapie

transformé crypto-leucémie aigue a l’évolution lente. Sangre 1944; 16(133).
14. Paraf A. Anémie pernicieuse ave mégaloblastose méedullaire rsistante a’
l’hépatothérapie terminaison en leucose aigue. Rev Hemat 1949; 4:655.
15. Luzzatto AM. Sull’ anemia grave megaloblastica senza reporto ematologico cor-
rispondente (anemia pseudoaplastica). Riv veneta di sc med Venezia 1907; 47:193–
212.
16. Dacie JV, Smith MD, White JC, et al. Refractory normoblastic anaemia: A clinical
and haematological study of seven cases. Br J Haematol 1959; 5:56–82.
17. Dameshek W. The DiGuglielmo syndrome revisited. Blood 1969; 34(5):567–572.
18. Block M, Jacobson LO, Bethard WF. Preleukemic acute human leukemia. JAMA
1953; 152:1018–1028.
19. Lehrer RI, Goldberg LS, Apple MA,et al. Refractory megaloblastic anemia with
myeloperoxidase-deficient neutrophils. Ann Intern Med 1972; 76(3):447–453.
20. Di Guglielmo G. Chronic hyperplastic panmyelosis. Sangre 1950; 21(5):405–
408.
21. Rheingold JJ, Kaufman R, Adelson E, et al. Smoldering acute leukemia. N Engl J
Med 1963; 268:812–815.
22. Steensma DP, Tefferi A. The myelodysplastic syndrome(s): A perspective and review
highlighting current controversies. Leuk Res 2003; 27(2):95–120.
23. Saarni MI, Linman JW. Preleukemia. The hematologic syndrome preceding acute
leukemia. Am J Med 1973; 55(1):38–48.
24. Anonymous. Is “preleukemic states” an adequate designation? Blood Cells 1976;
2:347–351.
25. Bennett JM, Catovsky D, Daniel MT, et al. Proposals for the classification of the acute
leukaemias. French-American-British (FAB) Co-operative Group. Br J Haematol
1976; 33(4):451–458.
26. Bennett JM, Catovsky D, Daniel MT, et al. Proposals for the classification of the
myelodysplastic syndromes. Br J Haematol 1982; 51(2):189–199.
27. Dameshek W. Sideroblastic anaemia: Is this a malignancy? Br J Haematol 1965;
11:52–58.
28. Juneja SK, Imbert M, Sigaux F, et al. Prevalence and distribution of ringed sideroblasts
in primary myelodysplastic syndromes. J Clin Pathol 1983; 36(5):566–569.
29. Delacretaz F, Schmidt PM, Piguet D, et al. Histopathology of myelodysplastic syn-
dromes. The FAB classification (proposals) applied to bone marrow biopsy. Am J
Clin Pathol 1987; 87(2):180–186.
30. Kerkhofs H, Hermans J, Haak HL, et al. Utility of the FAB classification for myelodys-
plastic syndromes: Investigation of prognostic factors in 237 cases. Br J Haematol
1987; 65(1):73–81.
31. Foucar K, Langdon RM, II, Armitage JO, et al. Myelodysplastic syndromes. A clinical
and pathologic analysis of 109 cases. Cancer 1985; 56(3):553–561.
32. Tuzuner N, Cox C, Rowe JM, et al. Hypocellular myelodysplastic syndromes (MDS):
New proposals. Br J Haematol 1995; 91(3):612–617.
33. Passmore SJ, Hann IM, Stiller CA, et al. Pediatric myelodysplasia: A study of 68
children and a new prognostic scoring system. Blood 1995; 85(7):1742–1750.
34. Locatelli F, Zecca M, Pession A, et al. Myelodysplastic syndromes: The pediatric
point of view. Haematologica 1995; 80(3):268–279.
35. Rosati S, Anastasi J, Vardiman J. Recurring diagnostic problems in the pathology of

the myelodysplastic syndromes. Semin Hematol 1996; 33(2):111–126.
36. Buesche G, Teoman H, Wilczak W, et al. Marrow fibrosis predicts early fatal marrow
failure in patients with myelodysplastic syndromes. Leukemia 2008; 22(2):313–322.
37. Mufti GJ, Stevens JR, Oscier DG, et al. Myelodysplastic syndromes: A scoring system
with prognostic significance. Br J Haematol 1985; 59(3):425–433.
38. Coiffier B, Adeleine P, Gentilhomme O, et al. Myelodysplastic syndromes. A mul-
tiparametric study of prognostic factors in 336 patients. Cancer 1987; 60(12):3029–
3032.
39. Sanz GF, Sanz MA, Vallespi T, et al. Two regression models and a scoring sys-
tem for predicting survival and planning treatment in myelodysplastic syndromes: A
multivariate analysis of prognostic factors in 370 patients. Blood 1989; 74(1):395–
408.
40. Aul C, Gattermann N, Heyll A, et al. Primary myelodysplastic syndromes: Analysis
of prognostic factors in 235 patients and proposals for an improved scoring system.
Leukemia 1992; 6(1):52–59.
41. Morel P, Declercq C, Hebbar M, et al. Prognostic factors in myelodysplastic syn-
dromes: Critical analysis of the impact of age and gender and failure to identify a
very-low-risk group using standard mortality ratio techniques. Br J Haematol 1996;
94(1):116–119.
42. Maschek H, Gutzmer R, Choritz H, et al. Life expectancy in primary myelodysplastic
syndromes: A prognostic score based upon histopathology from bone marrow biopsies
of 569 patients. Eur J Haematol 1994; 53(5):280–287.
43. De Wolf-Peeters C, Stessens R, Desmet V, et al. The histological characterization
of ALIP in the myelodysplastic syndromes. Pathol Res Pract 1986; 181(4):402–
407.
44. Verhoef G, De Wolf-Peeters C, Kerim S, et al. Update on the prognostic implication
of morphology, histology, and karyotype in primary myelodysplastic syndromes.
Hematol Pathol 1991; 5(4):163–175.
45. Gattermann N, Aul C, Schneider W. Two types of acquired idiopathic sideroblastic
anaemia (AISA). Br J Haematol 1990; 74(1):45–52.
46. Jacobs RH, Cornbleet MA, Vardiman JW, et al. Prognostic implications of morphol-
ogy and karyotype in primary myelodysplastic syndromes. Blood 1986; 67(6):1765–
1772.
47. Pierre RV, Catovsky D, Mufti GJ, et al. Clinical-cytogenetic correlations in myelodys-
plasia (preleukemia). Cancer Genet Cytogenet 1989; 40(2):149–161.
48. Sole F, Prieto F, Badia L, et al. Cytogenetic studies in 112 cases of untreated myelodys-
plastic syndromes. Cancer Genet Cytogenet 1992; 64(1):12–20.
49. Seo IS, Li CY, Yam LT. Myelodysplastic syndrome: Diagnostic implications of cyto-
chemical and immunocytochemical studies. Mayo Clin Proc 1993; 68(1):47–53.
50. Germing U, Kundgen A, Gattermann N. Risk assessment in chronic myelomonocytic
leukemia (CMML). Leuk Lymphoma 2004; 45(7):1311–1318.
51. Onida F, Kantarjian HM, Smith TL, et al. Prognostic factors and scoring systems in
chronic myelomonocytic leukemia: A retrospective analysis of 213 patients. Blood
2002; 99(3):840–849.
52. Sobin LH. The international histological classification of tumours. Bull World Health
Organ 1981; 59(6):813–819.
53. Harris NL, Jaffe ES, Diebold J, et al. World Health Organization classification of
neoplastic diseases of the hematopoietic and lymphoid tissues: Report of the Clinical
Advisory Committee meeting—Airlie House, Virginia, November 1997. J Clin Oncol
1999; 17(12):3835–3849.
54. Jaffe ES, Harris NL, Stein H, eds. WHO Classification: Tumours of Haematopoietic
and Lymphoid Tissues. Lyon: International Agency For Research on Cancer (IARC)
Press, 2001.
55. Rosati S, Mick R, Xu F, et al. Refractory cytopenia with multilineage dysplasia:
Further characterization of an ‘unclassifiable’ myelodysplastic syndrome. Leukemia
1996; 10(1):20–26.
56. Howe RB, Porwit-MacDonald A, Wanat R, et al. The WHO classification of MDS
does make a difference. Blood 2004; 103(9):3265–3270.
57. Germing U, Gattermann N, Strupp C, et al. Validation of the WHO proposals for a
new classification of primary myelodysplastic syndromes: A retrospective analysis of
1600 patients. Leuk Res 2000; 24(12):983–992.
58. Greenberg P, Anderson J, de Witte T, et al. Problematic WHO reclassification of
myelodysplastic syndromes. Members of the International MDS Study Group. J Clin
Oncol 2000; 18(19):3447–3452.
59. Nosslinger T, Reisner R, Koller E, et al. Myelodysplastic syndromes, from French-
American-British to World Health Organization: Comparison of classifications on
431 unselected patients from a single institution. Blood 2001; 98(10):2935–2941.
60. Qian J, Xue Y, Pan J,et al. Refractory thrombocytopenia, an unusual myelodysplas-
tic syndrome with an initial presentation mimicking idiopathic thrombocytopenic
purpura. Int J Hematol 2005; 81(2):142–147.
61. Szpurka H, Tiu R, Murugesan G, et al. Refractory anemia with ringed sideroblasts
associated with marked thrombocytosis (RARS-T), another myeloproliferative con-
dition characterized by JAK2 V617 F mutation. Blood 2006; 108(7):2173–2181.
62. Steensma DP, Dewald GW, Lasho TL, et al. The JAK2 V617F activating tyrosine
kinase mutation is an infrequent event in both “atypical” myeloproliferative disorders
and myelodysplastic syndromes. Blood 2005; 106(4):1207–1209.
63. Hasle H, Niemeyer CM, Chessells JM, et al. A pediatric approach to the WHO
classification of myelodysplastic and myeloproliferative diseases. Leukemia. 2003;
17(2):277–282.
64. Wimazal F, Fonatsch C, Thalhammer R, et al. Idiopathic cytopenia of undetermined
significance (ICUS) versus low risk MDS: The diagnostic interface. Leuk Res 2007;
31(11):1461–1468.
65. Horiike S, Kita-Sasai Y, Nakao M,et al. Configuration of the TP53 gene as an inde-
pendent prognostic parameter of myelodysplastic syndrome. Leuk Lymphoma 2003;
44(6):915–922.
66. Side LE, Curtiss NP, Teel K, et al. RAS, FLT3, and TP53 mutations in therapy-
related myeloid malignancies with abnormalities of chromosomes 5 and 7. Genes
Chromosomes Cancer 2004; 39(3):217–223.
67. Bowles KM, Warner BA, Baglin TP. Platelet mass has prognostic value in patients
with myelodysplastic syndromes. Br J Haematol 2006; 135(2):198–200.
68. Palmer SR, Tefferi A, Hanson CA,et al. Platelet count is an IPSS-independent risk fac-
tor predicting survival in refractory anaemia with ringed sideroblasts. Br J Haematol
2008; 140(6):722–725.
69. Germing U, Platzbecker U, Giagounidis A,et al. Platelet morphology, platelet mass,
platelet count and prognosis in patients with myelodysplastic syndromes. Br J Haema-
tol 2007; 138(3):399–400.
70. Sole F, Espinet B, Sanz GF, et al. Incidence, characterization and prognostic signif-
icance of chromosomal abnormalities in 640 patients with primary myelodysplastic
syndromes. Grupo Cooperativo Espanol de Citogenetica Hematologica. Br J Haema-
tol 2000; 108(2):346–356.
71. Haase D, Estey EH, Steidl C, et al. Abstract #247. Multivariate evaluation of the
prognostic and therapeutic relevance of cytogenetics in a merged European-American
cohort of 3860 patients with MDS. Blood 2007; 110(11).
72. Garcia-Manero G, Shan J, Faderl S, et al. A prognostic score for patients with lower
risk myelodysplastic syndrome. Leukemia 2008; 22(3):538–543.
73. Kao JM, McMillan A, Greenberg PL. Abstract #2457. International MDS Risk Anal-
ysis Workshop (IMRAW)/IPSS re-analyzed: Impact of depth of cytopenias on clinical
outcomes in MDS. Am J Hematol 2008; in press.
74. Malcovati L, Porta MG, Pascutto C, et al. Prognostic factors and life expectancy in
myelodysplastic syndromes classified according to WHO criteria: A basis for clinical
decision making. J Clin Oncol 2005; 23(30):7594–7603.
75. Schanz J, Estey EH, Steidl C, et al. Abstract #248. Multivariate analysis suggests that
the prognostic impact of poor cytogenetics is potentially underestimated in the IPSS.
Blood 2007; 110(11).
76. Malcovati L, Germing U, Kuendgen A, et al. Time-dependent prognostic scoring
system for predicting survival and leukemic evolution in myelodysplastic syndromes.
J Clin Oncol 2007; 25(23):3503–3510.
77. Cazzola M, Malcovati L. Myelodysplastic syndromes—Coping with ineffective
hematopoiesis. N Engl J Med 2005; 352(6):536–538.
78. Steensma DP. Are myelodysplastic syndromes “cancer”? Unexpected adverse conse-
quences of linguistic ambiguity. Leuk Res 2006; 30(10):1227–1233.
79. Di Guglielmo G. Eritremie acute. Boll Soc Med Chir Cremona 1926; 1:665–673.
80. Rhoads CP, Barker WH. Refractory anemia: Analysis of 100 cases. JAMA 1938;
110:794–796.
81. Bjorkman SE. Chronic refractory anemia with sideroblastic bone marrow: A study of
four cases. Blood 1956; 11:250–259.
82. Gatti RA, Meuwissen HJ, Allen HD,et al. Immunological reconstitution of sex-linked
lymphopenic immunological deficiency. Lancet 1968; 2(7583):1366–1369.
83. Rozman C, Granena A, Nomdedeu B, et al. Attempt of isogeneic bone marrow
transplantation without conditioning in a preleukaemic syndrome, negative result
[proceedings]. Pathol Biol (Paris) 1978; 26(1):47.
84. Bhaduri S, Kubanek B, Heit W, et al. A case of preleukemia–Reconstitution of
normal marrow function after bone marrow transplantation (BMT) from identical
twin. Blutalkohol 1979; 38(2):145–149.
85. Greenberg P, Cox C, LeBeau MM, et al. International scoring system for evaluating
prognosis in myelodysplastic syndromes. Blood 1997; 89(6):2079–2088.
86. Cheson BD, Bennett JM, Kantarjian H, et al. Report of an international working
group to standardize response criteria for myelodysplastic syndromes. Blood 2000;
96(12):3671–3674.
87. Silverman LR, Demakos EP, Peterson BL, et al. Randomized controlled trial of
azacitidine in patients with the myelodysplastic syndrome: A study of the cancer and
leukemia group B. J Clin Oncol 2002; 20(10):2429–2440.
88. List A, Kurtin S, Roe DJ, et al. Efficacy of lenalidomide in myelodysplastic syndromes.
N Engl J Med 2005; 352(6):549–557.
89. Kantarjian H, Issa JP, Rosenfeld CS, et al. Decitabine improves patient outcomes in
myelodysplastic syndromes: Results of a phase III randomized study. Cancer 2006;
106(8):1794–1803.
90. Cheson BD, Greenberg PL, Bennett JM, et al. Clinical application and proposal for
modification of the International Working Group (IWG) response criteria in myelodys-
plasia. Blood 2006; 108(2):419–425.
91. Toyama K, Ohyashiki K, Yoshida Y, et al. Clinical implications of chromosomal
abnormalities in 401 patients with myelodysplastic syndromes: A multicentric study
in Japan. Leukemia 1993; 7(4):499–508.
92. Komrokji R, Bennett JM. The myelodysplastic syndromes: Classification and prog-
nosis. Curr Hematol Rep 2003; 2(3):179–185.
93. Greenberg P, Anderson J, de Witte T, et al. Problematic WHO reclassification of
myelodysplastic syndromes. Members of the International MDS Study Group. J Clin
Oncol 2000; 18(19):3447–3452.
2
Epidemiology of the Myelodysplastic
Syndromes
Terry J. Hamblin
University of Southampton, Consultant Hematologist, Royal Bournemouth
Hospital, Southampton, U.K.
INCIDENCE
Difficulties in Establishing the True Incidence of MDS
The true incidence of the myelodysplastic syndromes (MDS) is not known because
these disorders were defined only relatively recently. The 9th edition of the Interna-
tional Classification of Diseases (ICD-9) (published in 1977) (1) did not recognize
them as distinct nosological entities, and therefore they tended to get lost under
a variety of headings. Chronic myelomonocytic leukemia (CMML) was included
in ICD-9, but only as a variety of chronic myeloid leukemia (ICD-9 code 205.1).
In addition, in the ICD-9 manual, the term “myelodysplasia” refers to congenital
defects of the spinal canal, further confusing the matter. The more specialized diag-
nostic manual, the International Classification of Diseases for Oncology (ICD-O
series), suffers from similar defects. Therefore, cancer registries, death certificates,
and hospital discharge data have been of little value in establishing the prevalence
of MDS. Indeed, the nonuniform classification schemes for MDS are exempli-
fied by a Swedish study that found MDS recorded in cancer registries under five
different diagnoses (2).
When the ICD-10 became available in 1992 (second edition, 2002) (3),
MDS were included under the heading “Neoplasms of Uncertain or Unknown
Behaviour” and coded under block D46. The French-American-British (FAB)
classification of MDS (4) was used in ICD-10, although CMML was included
27
28 Hamblin
separately among the “Monocytic leukaemia” subgroup of “Malignant neoplasms,

stated or presumed to be primary, of lymphoid, haematopoietic and related tissue”
at block C93.1.
An early estimate (1978) of the incidence of MDS in the United States (U.
S.) by James Linman and Grover Bagby (5) suggested approximately 1500 new
cases of “preleukemia” per year, but their definition was limited to cases with fewer
than 5% blasts in the marrow, and this figure is clearly a marked underestimate.
The National Cancer Institute (NCI) Surveillance, Epidemiology and End Results
(SEER) Program has only recently published American data on the incidence of
MDS (6). MDS only became a reportable disease for the United States cancer
registries after 2001, when the new third edition of the specialized International
Classification of Disease for Oncology (ICD-O-3) became applicable (7). Initial
results from SEER suggest that more than 10,268 new US MDS cases occur every
year, which is more than six times the original estimate from the late 1970s.
One of the central problems in establishing the true incidence of MDS is
the difficulty in making the diagnosis with confidence. Especially with respect to
refractory anemia (RA), the diagnostic features may be extremely subtle, and for
the new category of single lineage RA introduced by the World Health Organi-
zation (WHO) in 2001, this is even more so (8). Such diagnostic subtleties are
highlighted by a study of 54 “healthy” people between 18 and 85 years by Silvia
Fernández-Ferrero and Fernando Ramos (9) in León, Spain. These investigators
demonstrated that subtle dysplastic features occur not uncommonly in the bone
marrow of healthy subjects, whereas they occur most frequently in the elderly
and among smokers. Dyserythropoiesis was observed in 0.4% to 7.6% of cells in
subjects under 50 years of age and 4.0% to 28% of cells in subjects older than 50,
while dysgranulopoiesis was found in 3% to 12.5% and 6% to 29% of cells in the
two groups, respectively. In subjects with a history of smoking, the percentage of
cells affected is raised by an average of 3.7%.
The problem of establishing the true incidence of MDS is further compli-
cated by the suggestion that some patients might have “biological MDS” without
showing morphological features of one of these syndromes. In 1990, Allan Jacobs
in Cardiff coined the terms “not quite MDS” and “not yet MDS” to describe
this phenomenon (10). He initiated a study to look for clonal hematopoiesis in
patients who survived for a long time after having received chemotherapy for lym-
phoma. Although the assays used have proven insensitive, since skewed Lyoniza-
tion occurs relatively frequently in normal women as they age (11), individual case
reports suggest that “not yet MDS” (more recently termed “idiopathic cytopenia of
undeterminated/uncertain significance”, ICUS) is a legitimate entity that presents
serious diagnostic challenges (12).
Results from the SEER Database in the United States

The SEER Program of the NCI is an authoritative source of information on
cancer incidence and survival in the United States. SEER currently collects and
Epidemiology of the Myelodysplastic Syndromes 29
publishes cancer incidence and survival data from 17 population-based cancer

registries covering approximately 26% of the United States population. The SEER
population groups closely mirror the United States population mix, while the rural
populations are slightly underrepresented, the nonwhite populations are slightly
overrepresented.
Based on the first available SEER figures for MDS, the incidence of MDS
is 3.6 cases per 100,000 persons per year (males, 4.9; females, 2.9) with a further
0.4 per 100,000 presenting with CMML. Diagnoses are categorized in the SEER
database according to a mixture of the FAB and WHO classifications; the database
retains the FAB classification’s refractory anemia with excess of blasts in transfor-
mation (RAEB-t) as well as the various WHO categories. However, more than half
of the cases were registered as “MDS, not otherwise specified (NOS)”, suggesting
that diagnosticians reporting to SEER often used neither the FAB scheme nor the
WHO classification. Where distinctions were made, 32% of cases were RA, 22%
RA with ringed sideroblasts (RARS), 27% RA with excess of blasts (RAEB),
and 8% refractory cytopenia with multilineage dysplasia (RCMD). Only 3.5%
of cases were therapy related, and a similar proportion had the 5q deletion—a
much smaller proportion than in most databases of MDS patients that include
cytogenetic findings.
More than 70% of patients with MDS in the SEER database presented over
the age of 70, and two-thirds of those were over the age 80. The incidence was
slightly higher in whites than in other races. The 3-year survival rate was only 35%,
with older males having the worst survival. Despite the very large patient base
captured by SEER, it is likely that, because of ascertainment and classification
difficulties, these figures seriously underestimate the true incidence of MDS in the
United States.
European Population Studies

United Kingdom
In the mid-1980s, the Leukaemia Research Fund (LRF) in the United Kingdom
conducted a study in 26 counties in England and Wales surveying about 16 million
people (13), just over one-quarter of the combined population of those nations.
Hospital-based hematologists retrospectively reported cases diagnosed in their
laboratories to a central registry. This method of ascertainment has several flaws,
because many patients with mild MDS are diagnosed on incidental blood tests that
may not result in further investigation or hospital admission. Permanent records
are only likely to be kept if a bone marrow examination has been performed, and
older patients especially may decline this investigation. In addition, patients aged
80 and over (who represented 37% of MDS cases in the SEER study) were not
reported on in this study.
Nevertheless, the LRF found a much higher incidence of MDS than had
been expected. Over a 5-year period, 1806 cases were reported, for an overall
incidence of 3.6 per 100,000 patients. This figure disguises a four-fold variation
30 Hamblin
between counties, with centers that had a hematologist interested in the syndrome
recording higher incidences of MDS.
More than 90% of cases consisted of individuals ⬎55 years of age at diag-
nosis. The male to female ratio was 1.4:1, but owing to the greater number of
women alive at older ages, the standardized, gender-specific, incidence rates were
even more biased towards males, being 4.69 and 2.51 per 100,000, respectively.
In the age group 75 to 79, the incidence rates were 34 per 100,000 for males and
17 per 100,000 for females.
A subsequent LRF publication (14) explored the incidence in individuals
over 80 years of age, and extended the data collection period from 5 to 10 years.
During the 10-year period, 1919 cases were reported among those aged ⬎80
compared with 3641 among those aged ⬍80. In other words, 34.5% of cases
occurred in an age group that comprises only 3.9% of the British population (i.e.,
persons older than 80 years of age). The standardized incidence rates for the ⬎80
age group were 61.72 per 100,000 for males, 28.37 per 100,000 for females, and
38.85 per 100,000 overall. Even these figures are probably underestimates, as the
reported incidence of MDS peaks in the 85 to 89 age group at 41.7 per 100,000,
and decreases in the ⬎90 age group. There is no logical reason for the reported
decline in the ⬎90 age group other than a reluctance to investigate the very old
with a bone marrow examination to confirm a suspected diagnosis of MDS.
Germany
A second major investigation to determine the incidence of MDS was undertaken
by the Düsseldorf group (15). The study, a retrospective reanalysis of bone marrow
aspirates, covered a 16-year period from 1975 to 1990. The catchment population
studied was 1.2 million, and there were 18,416 bone marrow evaluations during the
study period. Of these, 584 cases of MDS could be recognized from the diagnostic
reports. Prior to 1983—that is, before the FAB classification was in use—the fol-
lowing diagnoses were surveyed to find possible MDS cases: panmyelopathy with
hypercellular marrow, sideroblastic anemia, CMML, and smoldering leukemia.
In this investigation, the most reliable incidence data came from a sub-
study of the 547,000 people of the town of Düsseldorf itself, where more accurate
estimates of the denominator could be obtained. For the period 1986 to 1990, the
overall incidence was estimated to be 4.1 cases per 100,000 persons, very similar
to the incidence found by the LRF and SEER studies (i.e., 3.6 per 100,000 for
both). An increasing incidence with the age was also confirmed: in the ⬍50 age
group, MDS incidence was 0.22 per 100,000; in the 50 to 69 age group, 4.88 per
100,000; and in the ⬎70 age group, 22.8 per 100,000.
Other Studies
Following the publication of these observations, several groups went on to estimate
the incidence of MDS in their own communities. A Swedish study found an
incidence of 3.6 cases per 100,000 persons (16), and a French study observed
an incidence of 3.2 per 100,000 (17). A small study from the Basque region in
Table 1 Annual Incidence Rates of MDS as Estimated by Individual Studies

Country or community Annual incidence rate
studied Years studied (per 100,000) Reference
England and Wales 1984–1988 3.6 overall (13)

(initial LRF study)
England and Wales (LRF 1984–1993 38.85 for persons ⬎80 yr (14)
follow-up study)
Düsseldorf, Germany 1975–1990 4.1 (15)
Sweden 1978–1992 3.6 (16)
France 1980–1990 3.2 (17)
French Basque region 1993–1996 7.7 (18)
Japan 1991 1 (19)
Bournemouth, U.K. 1981–1990 12.6 (20)
Somerset County, U.K. 1985–1993 9.3 (21)
SEER, U.S.A. 2001–2005 4.0 (8)
southern France (18), found an incidence of 7.7 per 100,000; in contrast to a

Japanese study (19) that found an incidence of only 1 per 100,000. Additionally,
two of the U.K. counties with the highest incidence rates in the LRF study have
since reported their data separately: Bournemouth found a rate of 12.6 per 100,000
(20) while Somerset observed an incidence of 9.3 per 100,000 (21). All studies
bar the German data have shown the same male predominance, and all have shown
the same increasing incidence with age. Table 1 summarizes the annual incidence
rates of MDS for various regions, as estimated by the studies mentioned above.
Is the Incidence of MDS Increasing?

In 1991, an opinion poll was conducted among a group of hematologists recognized
internationally for their interest in MDS about whether or not the incidence of
MDS was changing over time (2). Of the 41 hematologists surveyed, 91% felt
that the incidence of these disorders had increased by at least 100% in the past
10 to 20 years. More than 80% felt that the rise was real and not simply due to
improved diagnosis or wider recognition. Most felt that the most likely cause was
increased exposure to leukemogenic agents in the workplace or during medical
treatment. Atomic bombs, nuclear power station accidents, automobile exhausts,
and pesticides were named as the most likely culprits.
In the LRF study (13), there was evidence for an increasing rate of diagnosis
of MDS in the United Kingdom over a relatively short time period. In 1984, the
standardized incidence rate was 75% of the mean defined in that study, while
in 1988, it was 125%. This trend was confirmed by the subsequent publication
from the same group covering the next 5 years (14). Over a 10-year period, the
standardized incidence rate had gone from 2.38 per 100,000 (in 1984) to 4.59 per
100,000 (in 1993). The Düsseldorf group reported a similar increase. An annual
32 Hamblin
incidence of 1.3 per 100,000 for the years 1976 to 1980 increased to 4.1 per
100,000 for the years 1986 to 1990 (15). Interestingly, in those UK counties with
the highest incidence in the LRF study, there has been no increase during the
period between 1980 and 2000 (20,21).
The Düsseldorf group (22) considered whether there was evidence for
increased exposure to medical or environmental leukemogens during the obser-
vation period, but concluded that this explanation was not convincing. Only 31
cases were felt to be secondary to treatment with ionizing radiation or cytotoxic
drugs, and in all but a further 12, exposure to high levels of organic solvents
could be ruled out. The investigators felt that the most likely explanation for the
apparent increase in incidence in MDS was a greater willingness to perform bone
marrow investigations on older subjects. In 1975, patients older than 60 years of
age comprised 42% of those having bone marrow aspirates, compared with 54%
in 1990. During the same period the percentage of studied patients who were ⬎80
years of age climbed from 2.5% to 9%.
A later publication from the Düsseldorf group (23) confirmed that the appar-
ent increase in incidence of MDS is due to better case finding. Although there
was an increase from 1 per 100,000 to 4 per 100,000 between 1976 and 1986,
there was no corresponding increase between 1986 and 2002. The initial rise was
attributed to a greater awareness of the syndrome among hematologists since the
publication of the FAB group’s MDS classifications in 1976 and 1982 (4).
To What Extent is MDS Incidence Underestimated?

The possibility that there is a large group of asymptomatic and undiagnosed
patients with MDS was investigated by the Bournemouth group (20). This group
had already claimed the highest annual incidence of MDS (12.6 per 100,000) of any
group in the world, and had found an incidence of 89 per 100,000 in the ⬎80 age
group. From among 10,200 individuals registered with a single family practitioner
group practice of six partners located close to a hospital, the Bournemouth group
identified a cohort of 2926 older than age 55. These individuals were invited to
give a blood sample, either by attending the out-patient phlebotomy service or by
a home visit. A total of 1388 (47%) people agreed, and blood samples were drawn
and analyzed. Five years later, 785 of the 1388 (57%) were still alive and living in
the area; these individuals were invited to give a second sample. On this occasion,
657 samples were analyzed. In the first survey, three previously unknown and
asymptomatic cases of sideroblastic anemia were discovered, while in the second
survey, two new cases of CMML were discovered. Both of the latter patients had
normal blood counts 5 years back.
Over the 10-year period that encompassed these surveys, an additional 10
patients from the same family practice had been diagnosed as having MDS follow-
ing conventional investigation of symptoms or discovery of blood abnormalities
during evaluation of unrelated disease. Out of these 10, three had been too young
to be included in the initial cohort study (one with 5q− syndrome, one with RAEB,
and one with RAEB-t), four (two with CMML, and one each with RA and RAEB)
had declined the initial invitation to come for a blood test, while two (with RA and
RARS) had a normal blood count on the first occasion. Two other patients with
MDS joined the practice from other parts of the country during the 10-year period.
Although the small numbers make it difficult to have great confidence in the pre-
cise annual incidence of MDS estimated for this practice (14.7 per 100,000), this
incidence is of the same order of magnitude as that reported for the Bournemouth
group overall (i.e., 12.6 per 100,000). This finding suggests that for every case of
MDS known to hematologists, there are perhaps two other asymptomatic cases
that could be discovered by intensive screening.
The high prevalence of asymptomatic MDS in the elderly was confirmed
by geriatrician Yichayaou Beloosesky and colleagues in Israel (24), who studied
3275 patients admitted to a geriatric department for the “cognitively different”
at Rabin Medical Center. Over a 4-year period, 245 (7.5%) patients were found
to have unexplained cytopenia, macrocytosis or monocytosis. Out of these 245,
37 (15%; 1.1% of the total patient cohort) patients were ultimately diagnosed as
having MDS.
Incidence of MDS in Relation to Other Hematological Malignancies

According to the LRF study (13,14), MDS is by far the most common hematolog-
ical malignancy in the very old (⬎age 80) subjects, being three times as common
as acute myelogenous leukemia (AML), twice as common as chronic lymphocytic
leukemia (CLL) and myeloma, and more common than all non-Hodgkin lym-
phomas (NHL) put together. MDS becomes more common than AML when age
increases above 65 years; MDS and AML are equally common among those in
the 60 to 64 years of age group; and AML is more common among those younger
than age 60.
The Düsseldorf group found MDS to be twice as common as AML for the
general population and three times as common for the ⬎70 age group (15). In
Sweden (16), the two conditions were thought to be equally common; whereas in
Bournemouth, MDS was believed to be six times as common as AML (20).
In the WHO classification (8), some of the AML in the elderly is included as
AML with trilineage dysplasia, which may be part of the same disease process as
MDS. Thus, the distinction between these two may to some extent simply reflect
how assiduously asymptomatic MDS cases are sought prior to AML progression.
Differences Between MDS Epidemiological Patterns in Various Countries

Attention has been drawn to the fact that MDS appears to occur at a younger age
in countries other than those in Europe and North America (25). Reports from
Zimbabwe, Turkey, India, Korea, China, Thailand, and Japan (26–32) suggested
that in these countries, MDS is not mostly confined to the elderly. In Thailand, for
example, 32% of patients were younger than 40 years of age (29), compared with
34 Hamblin
only 2.6% in the US SEER statistics. In part, this is because the population as a
whole is younger in African and many Asian countries than in the West, but this
is not the case for Japan, where also the age of MDS patients is younger than in
the West.
In common with most Asian countries, in South Korea, physicians found
that patients with MDS had different characteristics from those in the Western
countries (30). Patients here were not only younger, but also their survival was
shorter, and the influence of prognostic factors was different. The percentage of
blasts in the bone marrow and the amount of dysgranulopoiesis were important
prognostic indicators, but the number and degree of cytopenias were not. Fur-
thermore, compared to patients in the West, Korean patients were more likely to
die from transformation to acute leukemia, and less likely to die from cytopenias.
Abnormalities of chromosome 5 were less likely to occur than in the Western
countries. Similarly, a Chinese study (32) found that RARS and CMML were less
common than what was found in Western studies, and that the incidence of MDS
was higher in rural than in urban areas. The Chinese investigators also found a
higher incidence of complex karyotypes than that in Western series.
Faced with these differences between European and Asian populations,
Akira Matsuda in Saitama, Ulrich Germing in Düsseldorf, and their colleagues
(33) compared the clinical features of Japanese patients with RA to a similar
cohort in Germany. Japanese patients were significantly younger and had more
severe cytopenias, yet they had a significantly more favorable prognosis. Japanese
patients also had a significantly lower risk of transformation to AML. German
patients designated RA by the FAB classification of MDS were more likely to be
designated RCMD by the WHO classification than the Japanese patients.
Since the distribution of population by age is similar in Germany and Japan,
these differences in age at presentation of MDS are likely to reflect real differences
between populations. Despite a higher incidence of thrombocytopenia among the
Japanese patients, these patients had a better overall survival than the German
ones. Because the International Prognostic Scoring System (IPSS) (34) is heavily
dependent on the number of cytopenias, it was found to be useful in predicting
prognosis for the German patients but not for the Japanese patients. Given the
collaborative nature of this study, we can be confident that the differences between
Asian and European patients are not simply due to differences in interpretation
of MDS diagnosis. Whether these differences in MDS epidemiological patterns
between countries are due to genetic or environmental factors (or both) remains
to be determined.
MDS in Younger Adults

There have been four published series that focused on adult MDS patients younger
than 50 years of age (35–38), all from Europe or North America. In three of these
studies, the overall proportion of MDS patients who fell in this age group was
6.7%, 9.5%, and 8.5%.
Because of the option of treatment by allogeneic stem cell transplantation

for this younger group, there is a danger that series assembled by tertiary referral
centers will be biased in favor of more severe cases, and this seems to have been
the case in the French and American series (35,36). But in a large unselected group
of patients drawn from a defined geographic area, Andrea Kuendgen in Düsseldorf
and her German colleagues found no difference between young and old patients
with regard to the distribution of IPSS risk groups (38). While the RAEB-t (FAB)
and RAEB 2 (WHO) subtypes were overrepresented in this series, so was the
RA (WHO) morphological subtype, and both RARS and CMML (FAB) were
underrepresented. Both this study and an Italian study (37) found more females
than expected among younger patients with MDS.
The most important finding of the German MDS study was related to overall
survival of younger patients (38). It has become common practice to offer allo-
geneic stem cell transplantation to younger patients with MDS, and this seemed
fully justified in patients with higher risk IPSS scores, since the actuarial median
survival was less than 12 months. However, for those with lower risk disease, the
actuarial median survival was not yet reached at 30 years. Indeed, the 20-year
overall survival for the IPSS low-risk group was 86%, and the median survival for
the intermediate-1 group was more than 14 years. In contrast, following allogeneic
stem cell transplantation, even in experienced centers, the long-term disease-free
survival is no higher than 60% for patients with IPSS low- or intermediate-1–
risk disease (39). Therefore, for younger patients with lower risk MDS, a rush to
transplantation may not be justified.
Pierre Fenaux in Paris and his colleagues in France (35) have suggested
that among younger MDS patients, there is a higher proportion of familial cases
and cases secondary to occupational exposure than is seen in older groups. Both
this French series and the German study (38) also found higher than expected
numbers of younger patients with abnormalities of chromosome 7, a karyotypic
abnormality frequently associated with secondary MDS.
MDS in Childhood
Childhood MDS will be considered in detail in chapter 14, but a brief reference
to its epidemiology is appropriate here. In the LRF study (13,14), childhood
MDS was found to be very rare, with only 21 cases reported in patients less
than 20-year old over the whole 10-year surveillance period (there are more than
12 million children in the United Kingdom, so the survey population potentially
included at least 2 million children having age less than 20 years). However, this
incidence is probably an underestimate owing to issues of accurate diagnosis and
classification, since the FAB system for MDS classification does not fit childhood
cases well (e.g., pediatric sideroblastic anemia is almost always due to a congenital
mutation in heme biosynthetic enzymes, and adult-like CMML is almost unknown
in children).
36 Hamblin
Table 2 Annual Incidence of Childhood MDS (Incidence Per 100,000)

Disorder Denmark British Columbia United Kingdom
DS MDS/AML 0.8 1.0 0.61

JMML 1.3 1.0 0.63
MDS (other) 2.1 1.4 0.81
Abbreviations: DS MDS/AML, myelodysplastic syndromes (MDS) or acute myeloblastic leukemia

(AML) occurring in patients with Down syndrome; JMML, juvenile myelomonocytic leukemia.
A more suitable classification of childhood MDS has emerged over the past
decade (40). Refractory cytopenia is the most common diagnosis in children, and
includes more than half of reported cases. RAEB is also seen, but the distinction
between MDS and AML is best defined by clinical behavior and is not simply a
product of blast count. Two other special MDS-like syndromes are recognized in
pediatric populations: juvenile myelomonocytic leukemia (JMML) and monosomy
7 syndrome. The clinical and epidemiological features of JMML are quite unlike
CMML (41). JMML is almost always seen in patients less than 5 years of age, and
there is a strong male predominance.
In children with Down syndrome, there are two distinct myeloid abnor-
malities. Transient myeloproliferative syndromes, most of which have dysplastic
features, are seen in perhaps 10% of newborns with Down syndrome, but almost
always undergo spontaneous remission with 3 months (42). A true acute leukemia
also occurs in these patients, and is usually classified as acute megakaryoblastic
leukemia (M7). Down syndrome megakaryoblastic leukemia cases, too, typically
have dysplastic features, and the distinction between MDS and AML is made on
the bone marrow blast cell count (43).
There have been several attempts to ascertain the incidence of childhood
MDS. All have limitations, but the most authoritative studies to date have come
from Denmark, British Columbia, and the United Kingdom (44–46). Table 2
shows the different incidence rates estimated by these studies for children aged
16 and under. The lower incidence in the British study compared to the other two
exemplifies the difficulty in classifying MDS in children, as this analysis excluded
patients who developed more than 30% blasts in the bone marrow within 3 months
of diagnosis of MDS, and also excluded patients with granulocytic sarcoma despite
having fewer than 30% blasts in the marrow.
Monosomy 7 is the commonest chromosomal abnormality in childhood
MDS, while other karyotypic abnormalities, especially those involving chromo-
some 5, are rare. While monosomy 7 carries a poor prognosis in childhood MDS
generally, in JMML, it is associated with a better outcome (46). The IPSS is not
useful in childhood MDS (47), as only a platelet count ⬍ 100 × 109 /L and marrow
blasts cells of ⬎5% are associated with poorer outcomes. For JMML, an “FPC”
(Fetal hemoglobin, Platelets, Cytogenetics) scoring system is useful, in which one
point is awarded each for HbF level ⬎10%, platelet count ⬍ 40 × 109 /L, or a
complex karyotype, and patients with higher scores have worse outcomes (48).
PREDISPOSING CAUSES
Most cases of MDS are idiopathic. In a small subset, there is a preexisting hemato-
logical disease such as aplastic anemia (49) or paroxysmal nocturnal hemoglobin-
uria (50), and therapy-related MDS, following treatment for other cancers is also
well recognized. Exploration of the genetic and environmental predisposing fac-
tors for primary MDS is important, as it may give insights into MDS etiology.
Genetic Factors
Congenital
Childhood MDS is often associated with predisposing germline genetic disorders,
including Down syndrome. Several other unusual syndromes have also been asso-
ciated with MDS-like disorders. For instance, Shwachman–Diamond syndrome
and Noonan syndrome may predispose to MDS. Shwachman–Diamond syndrome
(51) is an autosomal recessive disorder characterized by pancreatic insufficiency,
metaphyseal dysostosis, and bone marrow dysfunction. Noonan syndrome (52) is
a dysmorphic syndrome with multiple craniofacial abnormalities, cardiac abnor-
malities, hypertelorism, webbed neck, and prominent pads on fingers and toes.
Some confusion may arise with a disorder that goes by the same acronym as
MDS (Miller–Dieker syndrome) and which, like myelodysplastic syndromes, is
associated with a cytogenetic abnormality involving the short arm of chromosome
17. The Miller–Dieker syndrome (53) consists of lissencephaly, abnormal facies,
and growth retardation, but it is not a bone marrow disorder.
Neurofibromatosis is another genetic disorder that may predispose to MDS.
Type 1 neurofibromatosis is a relatively common autosomal dominant disorder
with an estimated incidence of 1 in 3500 (54). About 15% of patients with JMML
have the clinical features of neurofibromatosis, and a further 15% have mutations
of the neurofibromatosis gene NF1 (54) but without other clinical features of that
syndrome. Among patients with neurofibromatosis generally, however, JMML
develops only very rarely.
Fanconi syndrome was first described in 1927 as form of pancytopenia asso-
ciated with various physical abnormalities including altered skin pigmentation,
hypoplastic thumbs and radii, undeveloped genitalia, and abnormalities of head
and neck (55). Fanconi syndrome, sometimes called Fanconi anemia, is defined
by characteristic chromosomal breaks after clastogenic stress. Aplastic anemia
occurs in two-thirds of all individuals with Fanconi syndrome, and MDS occurs in
at least 2% to 3% by age 20, often heralded by the development of monosomy 7.
Dyskeratosis congenita is an inherited disease characterized by the triad of
abnormal skin pigmentation, nail dystrophy, and mucosal leucoplakia (56). Most
cases have an X-linked recessive form of inheritance, with the causative gene,
38 Hamblin
DKC1, mapped to Xq28. Around two-thirds of deaths among individuals with

dyskeratosis congenita are due to bone marrow failure and in approximately 1%
to 2% of these, features of MDS are recognized.
Kostmann syndrome (57) describes severe congenital neutropenia that
responds to treatment with granulocyte colony–stimulating factor (G-CSF). Muta-
tions in a gene (ELA2) encoding neutrophil elastase are common. MDS and acute
leukemia are rare complications.
In general, about one-third of cases of childhood MDS have an associated
predisposing condition. In addition to those mentioned, disorders associated with
MDS have included platelet storage pool disorder, Pierre Robin syndrome, Xan-
thogranulomata, Seckel syndrome, chronic infantile neurological cutaneous arthri-
tis, Milroy disease, various constitutional chromosomal abnormalities including
trisomy 8 mosaicism and abnormalities of chromosomes 5 and 7, diverse immun-
odeficiencies, and congenital autoimmune syndromes (45,46,48,58–60).
Familial Clustering
Onset of MDS in adulthood may also have a familial or genetic basis, as demon-
strated by one analysis (61) that identified 5/193 individuals with first-degree rel-
atives who also had MDS, higher than would be expected based on the estimated
population incidence of MDS. Sometimes the same constitutional abnormalities
on chromosomes 5 and 7 are implicated in each case in a family (62–64), but
often no linkage is found (65–67). Occasionaly patients with familial Pelger-Huet
anomaly may develop full blown MDS (68).
The Influence of Polymorphisms
An underlying genetic alteration may also be present in MDS cases that seem
to have been caused by environmental toxins or by exposure to cytotoxic drugs.
The molecular epidemiology of hematologic neoplasms has been well reviewed
by Ofer Shpilberg in Tel Aviv and his colleagues (69). The term “molecular
epidemiology” was coined in 1982 (70). It is defined as “the science that deals
with the contribution of potential genetic and environmental risk factors identified
at the molecular and biochemical level, to the etiology, distribution and control
of disease in groups of relatives and in populations” (69). Polymorphisms of
genes coding for proteins involved in DNA repair, DNA synthesis, and drug and
toxin metabolism have all been implicated in the causation of primary MDS and
treatment-related MDS/AML.
Although decline in the efficiency of DNA repair mechanisms (includ-
ing nucleotide excision repair, telomere maintenance and nonhomologous end-
joining) is a natural consequence of aging, leading to diminished stem cell
functional capacity (71,72), some patients with MDS also have an inactivat-
ing Ser327Cys polymorphism in the gene coding for 8-oxoG DNA glycosy-
lase (hOGG1), a component of the base excision repair pathway (73). Other
genetic polymorphisms that have been reported to increase AML risk and may
also increase susceptibility to MDS include the G/C variant at position -135 in the
5 untranslated region of the RAD51 gene and the Thr241Met variant of the XRCC3
gene—both genes encode factors involved in DNA repair (74). This complex area
of research was recently reviewed by Xi et al. (75). Over 520 different amino acid
substitution variants have been identified in the systematic screening of 91 human
DNA repair genes for sequence variation. At least one-third of these are classed
as possibly, or probably, contributing to an increase in DNA damage.
The methylene tetrahydrofolate reductase (MTHFR) enzyme directs tetrahy-
drofolate toward methionine synthesis and away from deoxyuridine monophos-
phate (dUMP) synthesis; methionine is required for DNA synthesis and repair.
Two MTHFR polymorphisms, C677 T and A1298 C, are associated with reduced
enzyme activity. In a recent publication, both have been associated with an
increased risk of development of treatment-related MDS/AML (76).
Polymorphisms in Drug and Carcinogen-Metabolizing Enzymes

Metabolism of cytotoxic drugs and of potentially carcinogenic compounds is a two-
stage process. In phase I, the substance is converted to its maximally mutagenic
metabolite by cytochrome P450 enzymes. These enzymes are encoded by genes
with numerous polymorphisms (77). For example, the most abundant component
of the cytochrome P450 system in human liver is CYP3 A, which metabolizes
etoposide, teniposide, cyclophosphamide, ifosfamide, and the vinca alkyloids. A
variant in the 5 promoter region of this gene (CYP3 A-V) diminishes production
of an epipodophyllotoxin catechol metabolite that is a precursor of the DNA-
damaging quinine. At least one study has indicated that this enzyme variant is
much less likely to be present in patients with treatment-related MDS/AML (78).
The phase I hepatic enzyme CYP2E1 metabolizes benzene to benzene oxide,
which spontaneously forms phenol, and the same enzyme further metabolizes phe-
nol to hydroquinone, a potent genotoxin. Detoxification of hydroquinone requires a
phase II enzyme, nicotinamide adenine dinucleotide (P)H:quinine oxidoreductase
(NQO1). Defects in the NQO1 pathway lead to accelerated telomere shortening
and clonally restricted hematopoiesis (79). An NQO1 polymorphism at codon
187 was studied by Tamoki Naoe and colleagues in Japan (80). The homozygous
Ser/Ser genotype was significantly more common in treatment-related MDS/AML
than in de novo AML or healthy individuals.
The most important phase II detoxifying enzyme is glutathione S-transferase
(GST), which is involved in the detoxification of adriamycin, BCNU, bleomycin,
chlorambucil, cisplatin, cyclophosphamide, etoposide, melphalan, mitomycin C,
mitoxantrone, and vincristine by catalyzing their conjugation to glutathione. It
has been postulated that deficiency of GST might be associated with a greater
risk of MDS. Functional polymorphisms exist for at least three genes that encode
GSTs, including GSTM1, GSTT1 and GSTP1 (81). Both GSTM1 and GSTT1
genes also have a “null” variant allele in which the entire gene is absent. At least
one group (82) found an odds ratio of 4.3 (95%CI, 2.5–7.4) for the GSTT1 null
genotype and MDS/AML, and this was confirmed by another group (83), who
40 Hamblin
found that in individuals with the GSTT1 null genotype, the odds ratio for disease
risk were raised to 2.65 (95%CI, 1.27–5.52) for de novo MDS, 4.62 (95%CI,
1.48–14.4) in therapy related AML and 2.94 (95%CI, 1.07–8.07) in AML with
trilineage dysplasia.
However, another group (84) found the GSTT1 null genotype similarly
distributed in MDS patients as in a healthy population, but an odds ratio of
2.0 (95%CI, 1.3–3.1) for the GSTM1 null genotype in patients with AML/MDS
compared with controls. Subsequently, three groups have demonstrated the double
null GSTT1/GSTM1 genotype to be overrepresented in Caucasian, Asian, and
Hispanic patients with primary and secondary MDS, AML, and aplastic anemia
(85–87).
Finally, one group found no increase in either GSTT1 or GSTM1 null genes
in patients with treatment-related MDS/AML, but individuals with at least one
GSTP1 codon 105 Val allele were significantly overrepresented in this group
compared with de novo AML cases (odds ratio, 1.81; 95%CI, 1.39–5.09) (81).
Environmental Factors
Anemic episodes preceding the development of acute leukemia after exposure
to ionizing radiation were reported anecdotally as early as the 1930s (88–90).
Survivors of the atom bomb attacks in Japan exhibited features in their blood that
would be today termed as MDS (91). It is now well established that secondary
MDS occurs following bone marrow injury by alkylating agents and ionizing
radiation. This topic will be covered in detail in chapter 8.
In addition to the effects of radiation and alkylators, the effect of the topoiso-
merase II inhibitors such as the epipodophyllotoxins (92) and the anthracyclines
and mitoxantrone (93) in generating chromosomal translocations, particularly
affecting the MLL gene at chromosomal band 11q23, are also well known. Less
familiar are the effects of antimetabolites such as 5-fluorouracil (94), azathioprine
(95), and fludarabine (96). Furthermore, the use of G-CSF for chronic therapy of
neutropenia or to generate autologous stem cells for peripheral blood harvest has
been implicated in the causation of secondary MDS (97), but this area remains
controversial.
The chromosomal abnormalities most commonly found in therapy-related
MDS, mainly involving chromosomes 5 and 7, are also most frequently found
in primary MDS. This implies that similar, though unknown, chemicals might be
responsible for marrow injury in apparently de novo cases. The earliest reports
of an environmental toxin causing “preleukemia” related to benzene (98–100).
Raw benzene is rarely encountered these days, but a number of case control
studies have explored the possibility that exposure to other common chemicals or
environmental factors might have a role in the pathogenesis of MDS.
A pilot study, evaluating a methodology previously used to link environmen-
tal exposure and solid tumors, implicated exposure to petrol (gasoline) and diesel
liquid and vapor, and ammonia (101) as risk factors for MDS. A larger, definitive
study by the same group identified exposure to ionizing radiation, metals, halo-
genated organics and petroleum products as significant risk factors (102). A United
States study found an increased exposure to pesticides and solvents among MDS
cases compared to controls (103) and this was confirmed by an Italian study (104).
Next, a Japanese pilot study found drinking alcohol to be the only significant risk
factor for MDS (105), but a larger study by the same group then found the risk
to be confined to former rather than current drinkers, and the larger study also
implicated the use of hair dyes (106).
Most recently, a French study has found that agricultural workers, textile
operators, health professionals, and machine operators all have a significantly
greater risk of developing MDS than people employed in other occupations. The
high risk also applies to those living near industrial plants, smokers, and those
exposed to mineral oil (107). However, long-term surveillance studies of workers
in the petrochemical industry have thus far found no increased incidence of any
hematological disease (108).
CONCLUDING REMARKS
As our understanding of MDS has increased, it has become clear why it is so
difficult to establish its true incidence. While the distinction between MDS and
AML has become clearer, separating MDS from some other hematological dis-
orders (such as myeloproliferative neoplasms) continues to be problematic, and
the distinction between MDS and normality has become blurred. It is clear that
there is a “hinterland” of undiagnosed disease that will only be recognized if it is
diligently sought. However, there may be little point in discovering such cases if
they will never become clinically relevant.
Consideration of predisposing genetic and environmental factors contributes
to our understanding of the epidemiology of MDS. The molecular epidemiology
of MDS is proving to be extremely complex, and we do not yet understand how
silent polymorphisms interact with unknown genotoxins to produce the disease.
In the next decade, we can expect that unraveling of this conundrum will lead us
to both prevention and improved therapy.
REFERENCES
1. International Classification of Diseases. Manual of the International Statistical Clas-
sification of Diseases, Injuries and Causes of Death, 9th ed. Geneva: World Health
Organization, 1977.
2. Reizenstein P, Dabrowski L. Increasing prevalence of the myelodysplastic syndrome.
An international Delphi study. Anticancer Res 1991; 11:1069–1070.
3. International Statistical Classification of Diseases and Health Related Problems,
(The) ICD-10 Second Edition. Geneva: World Health Organisation, 2005.
myelodysplastic syndromes. Br J Haematol 1982; 51:189–199.
5. Linman EG, Bagby GC. The preleukemic syndrome (hemopoietic dysplasia). Cancer
1978; 42(Suppl 2):854–864.
42 Hamblin
6. Ries LAG, Melbert D, Krapcho M, et al., eds. SEER Cancer Statistics Review,
1975–2004, National Cancer Institute. Bethesda, MD, http://seer.cancer.gov/csr/
1975 2004/, based on November 2006 SEER data submission, posted to the SEER
web site, 2007 (accessed December 10, 2007.)
7. International Classification of Diseases for Oncology, 3rd ed. (ICD-O-3). Geneva:
World Health Organisation, 2000.
8. Harris NL, Jaffe ES, Diebold J, et al. World Health Organization classification
of neoplastic diseases of the hematopoietic and lymphoid tissues: Report of the
Clinical Advisory Committee meeting—Airline House, Virginia, November 1997. J
Clin Oncol 2000; 18:2788–2789.
9. Fernandez-Ferrero S, Ramos F. Dyshaemopoietic bone marrow features in healthy
subjects are related to age. Leuk Res 2001; 25:187–189.
10. Jacobs A. Leukaemia Research Fund annual guest lecture 1990. Genetics lesions in
preleukaemia. Leukemia 1991; 5:277–282.
11. Busque L, Mio R, Mattioli J, et al. Non-random X-inactivation patterns in normal
females: Lyonization ratios vary with age. Blood 1996; 88:59–65.
12. Cachia PG, Taylor C, Thompson PW, et al. Non-dysplastic myelodysplasia.
Leukemia 1994; 8:677–681.
13. Cartwright RA, Alexander FE, McKinney PA, et al. Leukaemias and Lymphoma:
An Atlas of distribution Within Areas of England and Wales 1984–1988. London:
Leukaemia Research Fund, 1990.
14. Cartwright RA, McNally RJQ, Rowland DJ, et al. The Descriptive Epidemiology
of Leukaemia and Related Conditions in Parts of the United Kingdom 1984–1993.
London: Leukaemia Research Fund, 1997.
15. Aul C, Gattermann N, Schneider W. Age-related incidence and other epidemiological
aspects of myelodysplastic syndromes. Br J Haematol 1992; 82:358–367.
16. Radlund A, Thiede T, Hansen S, et al. Incidence of myelodysplastic syndromes in a
Swedish population. Eur J Haematol 1995; 54;153–156.
17. Mayanadie M, Verret C, Moskovtchenko P, et al. Epidemiological characteristics of
myelodysplastic syndrome in a well-defined French population. Br J Cancer 1996;
74:288–290.
18. Bauduer F, Ducout L, Dastugue N, et al. Epidemiology of myelodysplastic syn-
dromes in a French general hospital of the Basque country. Leuk Res 1998; 22:205–
208.
19. Shimizu H, Matsushita Y, Aoki K, et al. Prevalence of myelodysplastic syndromes
in Japan. Int J Hematol 1995; 61:17–22.
20. Williamson PJ, Kruger AR, Reynolds PJ, et al. Establishing the incidence of
myelodysplastic syndrome. Br J Haematol 1994; 87:743–745.
21. Phillips MJ, Cull GM, Ewings M. Establishing the incidence of myelodysplastic
syndrome. Br J Haematol 1994; 88:896–897.
22. Aul C, Germing U, Gattermann N, et al. Increasing incidence of myelodysplastic
syndromes: Real or fictitious? Leuk Res 1998; 22:93–100.
23. Germing U, Strupp C, Kündgen A, et al. No increase in age-specific incidence of
myelodysplastic syndromes. Haematologica 2004; 89:905–910.
24. Beloosesky Y, Cohen AM, Grosman B, et al. Prevalence and survival in myelodys-
plastic syndrome of the refractory anemia type in hospitalized cognitively different
geriatric patients. Gerontology 2000; 46:323–327.
25. Kuendgen A, Matsuda A, Germing U. Differences in epidemiology between Western

and Eastern countries: Ethnic differences or environmental influences. Leuk Res
2007; 31:103–104.
26. Mukiibi JM, Paul B. Myelodysplastic syndromes (MDS) in Central Africans. Trop
Geogr Med 1994; 46:17–19.
27. Paydas S, Kocak R. Myelodysplastic syndrome in young age. Ann Intern Med 1994;
3:29–32.
28. Oguma S, Yoshida Y, Uchino H, et al. Clinical characteristics of Japanese patients
with primary myelodysplastic syndromes: A co-operative study based on 838 cases.
Anemia Study Group of the Ministry of Health and Welfare. Leuk Res 1995; 19:219–
225.
29. Intragumtornchai T, Prayoonwiwat W, Swasdikul D, et al. Myelodysplastic syn-
dromes in Thailand: A retrospective pathologic and clinical analysis of 117 cases.
Leuk Res 1998; 22:453–460.
30. Lee JH, Lee JH, Shin YR, et al. Application of different prognostic scoring sys-
tems and comparison of the FAB and WHO classifications in Korean patients with
myelodysplastic syndrome. Leukemia 2003; 17:305–313.
31. Chatterjee T, Dixit A, Mohapatra M, et al. Clinical, haematological and histomor-
phological profile of adult myelodysplastic syndrome. Study of 96 cases in a single
institute. Eur J Haematol 2004; 73:93–97.
32. Chen B, Zhao WL, Jin J, et al. Clinical and cytogenetic features of 508 Chinese
patients with myelodysplastic syndrome and comparison with those in Western
countries. Leukemia 2005; 19:767–775.
33. Matsuda A, Germing U, Jinnai I, et al. Difference in clinical features between
Japanese and German patients with refractory anemia in myelodysplastic syndromes.
Blood 2005; 106:2633–2640.
prognosis in myelodysplastic syndromes. Blood 1997; 89:2079–2088.
35. Fenaux P, Preudhomme C, Estienne M, et al. De novo myelodysplastic syndromes
in adult aged 50 or less. A report on 37 cases. Leuk Res 1990; 14:1053–1059.
36. Chang KL, O’Donnell MR, Slovak ML, et al. Primary myelodysplasia occurring in
adults under 50 years old: A clinicopathologic study of 52 patients. Leukemia 2002;
16:623–631.
37. Breccia M, Mengarelli A, Mancini M, et al. Myelodysplastic syndromes in patients
under 50 years old: A single institution experience. Leuk Res 2005; 29:749–754.
38. Kuendgen A, Strupp C, Aivado M, et al. Myelodysplastic syndromes in patients
younger than age 50. J Clin Oncol 2006; 24:5358–5365.
39. Appelbaum FR, Anderson J. Allogeneic bone marrow transplantation for myelodys-
plastic syndrome: Outcomes analysis according to IPSS score. Leukemia 1998;
Suppl. 1:S25–S29.
classification of myelodysplastic and myeloproliferative diseases. Leukemia 2003;
17:277–282.
41. Emanuel PD. Myelodysplasia and myeloproliferative disorders in childhood: An
update. Br J Haematol 1999; 105:852–863.
42. Zipursky A, Brown E, Christensen H, et al. Leukemia and/or myeloproliferative
syndrome in neonates with Down syndrome. Semin Perinatol 1997; 21:97–101.
44 Hamblin
43. Lange BJ, Kobrinsky N, Barnard DR, et al. Distinctive demography, biology and
outcome of acute myeloid leukemia and myelodysplastic syndrome in children with
Down syndrome: Children’s Cancer Group studies 2681 and 2891. Blood 1998;
91:608–615.
44. Hasle H, Kerndrup G, Jacobsen BB. Childhood myelodysplastic syndrome in Den-
mark: incidence and predisposing conditions. Leukemia 1995; 9:1569–1572.
45. Hasle H, Wadsworth LD, Massing BG, et al. A population-based study of child-
hood myelodysplastic syndrome in British Columbia, Canada. Br J Haematol 1999;
106:1027–1032.
46. Passmore SJ, Chessells JM, Kempski H, et al. Paediatric myelodysplastic syndromes
and juvenile myelomonocytic leukaemia in the UK: A population-based study of
incidence and survival. Br J Haematol 2003; 121:758–767.
47. Hasle H, Baumann I, Bergsträsser E, et al. The International Prognostic Scor-
ing System (IPSS) for childhood myelodysplastic syndrome (MDS) and juvenile
myelomonocytic leukemia (JMML). Leukemia 2004; 18:2008–2014.
children and a new prognostic scoring system. Blood 1995; 85:1742–1750.
49. de Planque MM, Kluin-Nelemans HC, van Krieken HJ, et al. Evolution of acquired
severe aplastic anaemia to myelodysplasia and subsequent leukaemia in adults. Br J
Haematol 1988; 70:55–62.
50. Ishihara S, Nakakuma H, Kawaguchi T, et al. Two cases showing clonal progres-
sion with full evolution from aplastic anemia-paroxysmal nocturnal hemoglobinuria
syndrome to myelodysplastic syndromes and leukemia. Int J Hematol 2000; 72:206–
209.
51. Faber J, Lauener R, Wick F, et al. Shwachman–Diamond syndrome: Early bone
marrow transplantation in a high risk patient and new clues to pathogenesis. Eur J
Pediatr 1999; 158:995–1000.
52. Bertola DR, Sugayama SM, Albano LM, et al. Noonan syndrome: A clinical and
genetic study of 31 patients. Rev Hosp Clin Fac Med Sao Paulo 1999; 54:147–150.
53. Stratton RF, Dobyns WB, Airhart SD, et al. New chromosomal syndrome: Miller–
Dieker syndrome and monosomy 17p13. Hum Genet 1984; 67:193–200.
54. Side LE, Emanuel PD, Taylor B, et al. Mutations of the NF1 gene in children with
juvenile myelomonocytic leukemia without clinical evidence of neurofibromatosis,
type 1. Blood 1998; 92:267–272.
55. Alter BP. Inherited bone marrow failure syndromes. In: Handin RI, Lux SE, Stossel
TP, ed. Blood: Principles and Practice of Hematology. Philadelphia: JB Lippincott
Company, 1995:227–291.
56. Dokal I. Dyskeratosis Congenita in all its forms. Br J Haematol 2000; 110:768–779.
57. Kostmann R. Infantile genetic agranulocytosis: A new recessive lethal disease in
man. Acta Paediatr Scand 1956; 45:1–12.
58. Grimwade DJ, Stephenson J, De Silva C, et al. Familial MDS with 5q- abnormality.
Br J Haematol 1993; 84:536–538.
59. Kardos G, Veerman AJ, de Waal FC, et al. familial sideroblastic anemia with emer-
gence of monosomy 5 and myelodysplastic syndrome. Med Pediatr Oncol 1996;
26:54–56.
60. Lune-Fineman S, Shannon KM, Atwater SK, et al. Myelodysplastic and myelopro-
liferative disorders of childhood: A study of 167 patients. Blood 1999; 93:459–466.
61. Lucas GS, West RR, Jacobs A. Familial myelodysplasia. BMJ 1989; 299:551.
62. Olopade OI, Roulston D, Baker T, et al. Familial myeloid leukemia associated with
the loss of the long arm of chromosome 5. Leukemia 1996; 10:669–674.
63. Wakita A, Komatsu H, Banno S, et al. Myelodysplastic syndrome developed in a
mother and her son whose bone marrow karyotype showed monosomy 7. Rinsho
Ketueki 1996; 37:311–316.
64. Kardos G, Veerman AJ, de Waal FC,et al. Familial sideroblastic anemia with emer-
gence of monosomy 5 and myelodysplastic syndrome. Med Pediatr Oncol 1996;
26:54–56.
65. Gao Q, Horwitz M, Roulsdon D, et al. Susceptibility gene for familial acute myeloid
leukaemia associated with loss of 5q and/or 7q is not localized on the commonly
deleted portion of 5q. Genes Chromosomes Cancer 2000; 28:264–272.
66. Kumar T, Mandla SG, Greer WL. Familial myelodysplastic syndrome with early
age of onset. Am J Hematol 2000; 64:53–58.
67. Mandala SG, Goobie S, Kumar RT, et al. Genetic analysis of familial myelodysplas-
tic syndrome: absence of linkage to chromosomes 5q31 and 7q22. Cancer Genet
Cytogenet 1998; 106:113–118.
68. Hiraga H, Yabe H, Nagai K, et al. Myelodysplastic syndrome in a patient with
familial Pelger-Huet anomaly. Rinsho Ketuski 1991; 32:1453–1457.
69. Shpilberg O, Dorman JS, Shahar A, et al. Molecular epidemiology of hematological
neoplasms present status and future directions. Leuk Res 1997; 21:265–284.
70. Perera FP, Weinstein IB. Molecular epidemiology and carcinogen-DNA adduct
detection: New approaches to human cancer causation. J Chronic Dis 1982; 35:581–
600.
71. Rossi DJ, Bryder D, Seita J, et al. Deficiencies in DNA damage repair limit the
function of haematopoietic stem cells with age. Nature 2007; 447:725–729.
72. Nijnik A, Woodbine L, Marchetti C, et al. DNA repair is limiting for haematopoietic
stem cells during ageing. Nature 2007; 447:686–690.
73. Jankowska AM, Gondek LP, Szpurka H, et al. Base excision repair dysfunction in a
subgroup of patients with myelodysplastic syndrome. Leukemia 2007 Dec 6 (Epub
ahead of print).
74. Bhatla D, Gerbing RB, Alonzo TA, et al. DNA repair polymorphisms and outcome
of chemotherapy for acute myelogenous leukemia: A report from the children’s
oncology group. Leukemia 2007 Nov 22 (Epub ahead of print).
75. Xi T, Jones IM, Mohrenweiser HW. Many amino acid substitution variants identified
in DNA repair genes during human population screenings are predicted to impact
protein function. Genomics 2004; 83:970–979.
76. Guillem VM, Collado M, Terol MJ, et al. Role of MTHFR (677, 1298) haplotype
in the risk of developing secondary leukemia after treatment of breast cancer and
hematological malignancies. Leukemia 2007; 21:1413–1422.
77. Leone G, Pagano L, Ben-Yehuda D, et al. Therapy-related leukemia and myelodys-
plasia: Susceptibility and incidence. Haematologica 2007; 92:1389–1398.
78. Rund D, Krichevsky S, Bar-Cohen S, et al. Therapy-related leukemia: Clinical char-
acteristics and analysis of new molecular risk factors in 96 adult patients. Leukemia
2005; 19:1919–1928.
79. Fern L, Pallis M, Carter G, et al. Clonal haemopoiesis may occur after conventional
chemotherapy and is associated with accelerated telomere shortening and defects
46 Hamblin
in the NQO1 pathway; possible mechanisms leading to an increased risk of t-

AML/MDS. Br J Haematol 2004; 126:63–71.
80. Naoe T, Takeyama K, Yokozawa T, et al. Analysis of genetic polymorphism in
NQO1, GST-M1, GST-T1, and CYP3A4 in 469 Japanese patients with therapy-
related leukemia/myelodysplastic syndrome and de novo acute myeloid leukemia.
Clin Cancer Res 2000; 6:4091–4095.
81. Allan JM, Wild CP, Rollinson S, et al. Polymorphism in glutathione S-transferase
P1 is associated with susceptibility to chemotherapy-induced leukemia. Proc Natl
Acad Sci U S A 2001; 98:11592–11597.
82. Chen H, Sandler DP, Taylor JA, et al. Increased risk for myelodysplastic syndrome
in individuals with glutathione transferase theta (GSTT1) gene defect. Lancet 1996;
347:295–297.
83. Sasai Y, Horiike S, Misawa S, et al. Genotype of glutathione S-transferase and other
genetic configurations in myelodysplasia. Leuk Res 1999; 23:975–981.
84. Davies SM, Robison LL, Buckley JD, et al. Glutathione S-transferase polymorphisms
in children with myeloid leukemia: A Children’s Cancer Group Study. Cancer Epi-
demiol Biomarkers Prev 2000; 9:563–566.
85. Arruda VR, Lima CS, Grignoli CR, et al. Increased risk for acute myeloid leukaemia
in individuals with glutathione S-transferase mu 1 (GSTM1) and theta 1 (GSTT1)
gene defects. Eur J Haematol 2001; 66:383–388.
86. Haase D, Binder C, Bünger J, et al. Increased risk for therapy-associated hematologic
malignancies in patients with carcinoma of the breast and combined homozygous
gene deletions of glutathione transferases M1 and T1. Leuk Res 2002; 26:249–254.
87. Sutton JF, Stacey M, Kearns WG, et al. Increased risk for aplastic anemia and
myelodysplastic syndrome in individuals lacking glutathione S-transferase genes.
Pediatr Blood Cancer 2004; 42:122–126.
88. Moreland HS. The occurrence of malignancy in radioactive persons. Am J Cancer
1931; 15:2435–2516.
89. Hamilton-Paterson JL. Pre-leukaemic anaemia. Acta Haematol 1949; 2:307–316.
90. Evans RD. The effects of skeletally deposited alpha-ray emitters in man. Br J Radiol
1966; 39:881–895.
91. Finch SC, Hoshino T, Lamphere JP, et al. Peripheral blood changes preceding the
development of leukemia in atom bomb survivors in Hiroshima and Nagasaki. In:
Seno S, Takaku F, Irino S, eds. Topics in Haematology, 16th International Congress
of Haematology. Amsterdam: Exerpta Medica, 1977:97–98.
92. Felix CA, Walker AH, Lange BJ, et al. Association of CYP3A4 genotype with
treatment-related leukemia. Proc Natl Acad Sci U S A 1998; 95:13176–13181.
93. Le Deley MC, Suzan F, Cutuli B, et al. Anthracyclines, mitoxantrone, radiotherapy,
and granulocyte colony-stimulating factor: Risk factors for leukemia and myelodys-
plastic syndrome after breast cancer. J Clin Oncol 2007; 25:292–300.
94. Tucker A, Guler N. Therapy related acute myeloid leukemia after exposure to 5-
fluorouracil: A case report. Hematol Cell Ther 1999; 41:195–196.
95. Offman J, Opelz G, Doehler B, et al. Defective DNA mismatch repair in acute
myeloid leukemia/myelodysplastic syndrome after organ transplantation. Blood
2004; 104:822–828.
96. Tam CS, Seymour JF, Prince HM, et al. Treatment-related myelodysplasia following
fludarabine combination chemotherapy. Haematologica 2006; 91:1546–1550.
97. Socie G, Mary JY, Schrezenmeier H, et al. Granulocyte-stimulating factor and severe
aplastic anemia: A survey by the European Group for Blood and Marrow Transplan-
tation (EBMT). Blood 2007; 109:2794–2796.
98. Browning E. Toxicity and Metabolism of Industrial Solvents. Amsterdam: Elsevier,
1965:3–65.
99. Aksoy M, Erdem S, Dincol G. Types of leukemia in chronic benzene poisoning.
Acta Haematol 1976; 55:65–72.
100. Aksoy M. Malignancies due to occupational exposure to benzene. Am J Indust Med
1985; 7:395–402.
101. Farrow A, Jacobs A, West RR. Myelodysplasia, chemical exposure and other envi-
ronmental factors. Leukemia 1989; 3:33–35.
102. West RR, Stafford DA, Farrow A, et al. Occupational and environmental exposures
and myelodysplasia: A case control exposure. Leuk Res 1995; 19:127–139.
103. Goldberg H, Lusk E, Moore J, et al. Survey of exposure to genotoxic agents in
primary myelodysplastic syndrome: Correlation with chromosomal patterns and
data on patients without hematological disease. Cancer Res 1990; 50:6876–6881.
104. Rigolin GM, Cuneo A, Roberti MG, et al. Exposure to myelotoxic agents and
myelodysplasia: Case-control study and correlation with clinicobiological findings.
Br J Haematol 1998; 103:189–197.
105. Ido M, Nagata C, Kawakami N, et al. A case-control study of myelodysplastic
syndromes among Japanese men and women. Leuk Res 1996; 20:727–731.
106. Nagata C, Shimizu H, Hirashima K, et al. Hair dye use and occupational exposure
to organic solvents as risk factors for myelodysplastic syndrome. Leuk Res 1999;
23:57–62.
107. Nisse C, Haguenoer JM, Grandbastien B, et al. Occupational and environmental risk
factors of the myelodysplastic syndromes in the North of France. Br J Haematol
2001; 112:927–935.
108. Tsai SP, Fox EE, Ransdell JD, et al. A hematology surveillance study of petrochem-
ical workers exposed to benzene. Regul Toxicol Pharmacol 2004; 40:67–73.
3
The Cytogenetics and Molecular Biology
of the Myelodysplastic Syndromes
Harold J. Olney
Université de Montréal, CHUM Hôpital Notre-Dame,
Montreal, Quebec, Canada
Michelle M. Le Beau
Department of Medicine and Cancer Research Center, University of Chicago,
Chicago, Illinois, U.S.A.
INTRODUCTION
The myelodysplastic syndromes (MDS) include a heterogenous group of clonal
bone marrow disorders characterized by the presence of dysplastic maturation
of hematopoietic cells, coupled with one or more peripheral cytopenias and a
propensity to progress to an acute leukemia (1,2). While cytologic dysplasia is the
cardinal feature of MDS, there are a number of other conditions that can present
a similar histopathologic picture: nutritional deficiencies (e.g., vitamin B12 and
folate), toxins, infections, and congenital disorders must all be excluded. There-
fore, documentation of a cytogenetically abnormal clone can provide important
information in support of a diagnosis of MDS (3,4).
The current MDS-related diagnostic entities are established based on World
Health Organization (WHO) criteria as discussed in chapters 1 and 9. This classifi-
cation is based on bone marrow histological findings, blast count, and cytogenetic
findings (Table 1). The natural history of the disease, including the risk of leukemic
transformation, is significantly worsened with an increasing marrow blast count
and the presence of a high-risk cytogenetic abnormality.
49
50 Olney and Le Beau
Table 1 World Health Organization MDS Classification System

Marrow blasts Cytogenetic
Disease (%) Clinical presentation abnormalities (%)
Refractory anemia ⬍5 Anemia 25

(RA)
RA with ringed ⬍5 Anemia, ≥15% ringed 10
sideroblasts sideroblasts in
(RARS) erythroid precursors
5q− Syndrome ⬍5 Anemia, normal 100
platelets
Refractory ⬍5 Bicytopenia or 50
cytopenia with pancytopenia
multilineage
dysplasia
(RCMD)
RCMD with ringed ⬍5 Bicytopenia or 50
sideroblasts pancytopenia, ≥15%
ringed sideroblasts
Refractory anemia 5–9 Cytopenias ± blasts 50–70
with excess (⬍5%)
blasts-1
(RAEB-1)
Refractory anemia 10–20 Cytopenias, blasts 50–70
with excess present
blasts-2
(RAEB-2)
Myelodysplastic ⬍5 Neutropenia or 50
syndrome— thrombocytopenia
unclassified
Chronic ⬍20 Monocytosis 25–50
myelomonocytic (⬎1000/␮l), total
leukemia leukocytes
(CMML)— ≤ 13000/␮l
nonproliferative
type
The cytogenetic evaluation of a bone marrow sample from patients with

an MDS has become an integral part of clinical care. Not only does this analysis
confirm the diagnosis, but also it is invaluable in defining the prognosis and median
survival, as well as estimating the risk for progression to acute myeloid leukemia
(AML). On a more fundamental level, cytogenetic analysis has been instrumental
in establishing the clonality of these syndromes as well as providing hints into
the pathobiology of these entities. This chapter will review the most frequently
encountered abnormalities and explore their clinical and genetic features.
The Cytogenetics and Molecular Biology of the Myelodysplastic Syndromes 51
DIAGNOSIS
Establishing Clonality
The diagnosis of all hematological malignancies, including MDS, begins with
the appropriate clinical evaluation combined with expert analysis of pathological
and genetic features. For the clinician, cytogenetic analysis plays a vital role in
the management of patients with MDS, including confirmation of the diagnosis
(especially important when morphological findings are subtle), prognostication,
and selection of appropriate therapy. Cytogenetic analysis is also the most widely
available and standardized technique for identifying clonality in MDS. In fact,
the WHO has included recurring cytogenetic abnormalities in the classification of
several subtypes of MDS with distinct clinical presentations and natural histories,
as discussed later (4).
At the time of diagnosis, recurring chromosomal abnormalities are found
in 40% to 70% of patients with primary MDS and in 95% of patients with
therapy-related MDS (t-MDS) (5) (see chap. 8); identification of these confirms
the presence of a neoplastic process. The frequency of cytogenetic abnormalities
increases with the severity of disease (Table 1). Beyond karyotyping, the analysis
of mutated oncogenes or tumor suppressor genes has also been used to confirm
the clonal nature of MDS and to provide prognostic information. Other diag-
nostic techniques are often not definitive. Aberrant in vitro growth patterns of
hematopoietic progenitor cells are a characteristic of MDS (6), yet techniques to
assess these patterns are neither widely available, nor are these findings specific
for MDS. Immunophenotyping protocols (7–9) (see chap. 11) and microarray
techniques (10,11), including array comparative genomic hybridization and single
nucleotide polymorphism (SNP) arrays to detect copy-neutral loss of heterozy-
gosity (see chap. 4), hold potential for future clinical diagnostic and prognostic
applications.
Cytogenetic Analysis
A variety of recurring cytogenetic abnormalities are seen in MDS (Table 2). These
findings are not exclusive to MDS and may also be seen in AML or other clonal
myeloid neoplasms. Most recurring cytogenetic abnormalities found in MDS are
unbalanced, most commonly the result of the loss of a whole chromosome or a
deletion of part of a chromosome, but unbalanced translocations and more complex
derivative (rearranged) chromosomes can also be found (Figs. 1 and 2). The most
common cytogenetic abnormalities encountered in MDS are del(5q), −7, and +8,
which have been incorporated into several prognostic scoring systems for MDS,
including the International Prognostic Scoring System (IPSS) (see chaps. 1 and
15). Clones with unrelated abnormalities, one of which typically has a gain of
chromosome 8, are seen at a greater frequency (∼5% vs. ∼1%) in patients with
MDS than in patients with AML.
52
Table 2 Recurring Chromosomal Abnormalities in Myelodysplastic Syndromes

Diseasea Chromosome abnormality Frequency (%) Involved genesb Consequence
+8 10
MDS unbalanced −7/del(7q) 10
−5/del(5q) 10
del(20q) 5–8
−Y 5
i(17q)/t(17p)/−17 3–5 TP53 Loss of function
−13/del(13q) 3
del(11q) 3
del(12p)/t(12p) 3
del(9q) 1–2
idic(X)(q13) 1–2
Balanced t(1;3)(p36.3;q21) 1 MMEL1 RPN1 Deregulation of
MMEL1—
transcriptional
activation?
t(2;1)(p21;q23)/t(11q23) 1 MLL MLL fusion
protein—altered
transcriptional
regulation
inv(3)(q21q26.2) 1 RPN1 MDS1/EVI1 Fusion protein
t(6;9)(p23;q34) 1 DEK NUP214 Fusion protein—nuclear
pore protein
Olney and Le Beau
−7/del(7q) 50
Therapy-related MDS −5/del(5q) 40–45
dic(5;17)(q11.1–13;p11.1–13) 5 TP53 Loss of function, DNA
damage response
der(1;7)(q10;p10) 3
t(3;21)(q26.2;q22.1) 3 RPL22L1 RUNX1 RUNX1 fusion
protein—altered
transcriptional
regulation
t(11;16)(q23;p13.3)/t(11q23) 2 MLL CREBBP MLL fusion
protein—altered
transcriptional
regulation
CMML t(5;12)(q32;p13) 2–5 PDGFRB ETV6/TEL Fusion protein—altered
signaling pathway
a MDS, myelodysplastic syndrome; CMML, chronic myelomonocytic leukemia.
b Genes are listed in order of citation in the karyotype, e.g., for the t(11;16), MLL is at 11q23 and CREBBP is at 16p13.3.
The Cytogenetics and Molecular Biology of the Myelodysplastic Syndromes
53
Primary MDS t-MDS
Normal karyotype Other balanced abnormalities
Balanced abnormalities Abnormal chromosomes 5 and/or 7
Figure 1 Type of karyotypic abnormalities in MDS.
A handful of specific cytogenetic abnormalities, including MDS with iso-

lated del(5q) (the 5q− syndrome) (12), −17/loss of 17p (the 17p− syndrome)
(13), and the isodicentric X chromosome (associated with RARS with a high like-
lihood of transformation to AML) (14), are associated with morphologically and
Primary MDS t-MDS/t-AML
Normal
Both 5/7 Both 5/7 8% t(11q23)
14% Normal 23% 4%
Balanced
40%
4%
Abnl 7
9% Other
12%
Abnl 5
9%
Abnl 17p
5% t(11q23) Abnl 5
+8 2% 21%
9% Abnl 7
del(20q)
28%
5%
Figure 2 Recurring chromosomal abnormalities in MDS.

clinically distinct subsets of MDS. In rare cases, recurring balanced translocations

have been reported. Abnormalities characteristic of acute leukemia without a prior
myelodysplastic phase, such as the t(15;17), inv(16) and t(8;21), can be identified
rarely in MDS (15). The t(9;22) rearrangement, diagnostic of chronic myeloge-
nous leukemia (CML) and a subtype of acute lymphoblastic leukemia (ALL), has
only rarely been reported in MDS, and when seen may indicate the presence of
two myeloid neoplasms in a single patient (i.e., CML + MDS).
Fluorescence In Situ Hybridization (FISH)

In contrast to classical cytogenetic analysis, FISH can evaluate interphase as well as
metaphase cells in a rapid and efficient manner (16,17). The primary advantage of
FISH is its simplified analysis permitting the evaluation of an increased number of
cells compared to classical karyotyping, thereby greatly increasing the sensitivity.
FISH can also be applied to histologically stained cells allowing a direct correlation
of the status of the genetic target within morphologically characterized cells.
However, the technique evaluates specific alterations based on probe selection,
rather than the entire chromosomal complement. Probes suitable for clinical use
are not available for all recurring abnormalities of interest, and variation in the
cytogenetic abnormality (with either complex rearrangements or differences in
breakpoints) may not be detected with conventional probes. In MDS, commercially
available probes have been developed for the detection of 11q23 translocations
involving MLL, −Y, −5/del(5q), −7/del(7q), +8, del(20q), del(13q), del(11q),
and −17/loss of 17p.
PROGNOSIS AND CLINICAL CORRELATIONS

The value of cytogenetic analysis in predicting survival and risk of leukemic
transformation during a patient’s clinical course has been well established (18–22).
Among the few independent variables identified that predict clinical outcomes in
MDS, cytogenetic findings form the cornerstone of successful prognostic scoring
systems (23,24). The best-validated scoring system, the IPSS (23), identified
marrow blast count, number of peripheral cytopenias, and cytogenetic findings
as the three variables most useful in prognostication (Table 2). Its application is
limited to the time of diagnosis.
The WHO diagnostic classification integrates the characteristics of the first
two IPSS elements into the diagnosis. In the WHO prognostic scoring system
(WPSS), WHO diagnostic classification can be combined with the IPSS cytoge-
netic risk groups, as well as the clinical need for blood transfusion, in order to
determine a more dynamic prognostic score that may have advantages compared
to the IPSS. The WPSS has the advantage of being a time-dependent system that
can be used throughout the course of the disease (24). In lower risk patients,
additional laboratory findings (such as ferritin and ␤2-microglobulin levels) may
help identify those patients with a worse prognosis who may benefit from early
Table 2 Cytogenetic Abnormalities in the International Prognosis Scoring System

25% AML Median survival
Cytogenetic abnormalities progression (yr) (yr)
Favorable risk Normal karyotype 5.6 3.8

Isolated del(5q)
Isolated del(20q)
Isolated −Y
Intermediate Other abnormalities 1.6 2.4
risk
Poor risk −7/del(7q) 0.9 0.8
Complex karyotypes
therapeutic interventions (25). With larger datasets, rarer recurring cytogenetic

abnormalities are being examined, which allow a refining of the cytogenetic risk
groups and provide the clinician with more information to predict the expected
outcome for patients, albeit with the caveats associated with retrospective studies
(22).
Therapeutic options are increasing in MDS patients, as several drugs are
now approved for their care by various regulatory agencies (see chap. 16). While
cytogenetic analysis has always been used to establish prognosis, which dic-
tated therapeutic decisions in the most global sense (supportive care vs. a trial of
remission-inducing cytotoxic chemotherapy regimens), a new era of targeted ther-
apy was initiated with the recognition of the sensitivity of specific genetic subtypes
of MDS to particular therapies. Success stories include PDGFRB translocations,
which are seen in rare cases of chronic myelomonocytic leukemia (CMML) and
are sensitive to the tyrosine kinase inhibitor imatinib mesylate (26), and the del(5q)
cases that respond to the immunomodulating drug lenalidomide (27) (see chap.
19). With further molecular understanding of the underlying abnormalities in
MDS, treatment of patients will be individualized to the specific chromosomal
abnormalities underlying the disease process in each patient.
EVOLUTION OF THE KARYOTYPE

Over time, the identification of new abnormalities in the karyotype often coincides
with a change in the behavior of the disease, usually toward a more aggressive
course, and may herald incipient leukemia. Cytogenetic evolution includes the
appearance of an abnormal clone where only normal cells have been seen pre-
viously, or progression from the presence of a single clone (often with a simple
karyotype) to multiple related, or occasionally unrelated, abnormal clones. The
abnormal clones may evolve acquiring additional abnormalities with disease pro-
gression, and typically resolve with remission of disease following treatment. In
published series, most MDS patients die of bone marrow failure, close to half
progress to acute leukemia, and a few die of unrelated intercurrent illness.
The natural history of MDS is generally characterized by one of three
clinical scenarios: (1) a gradual worsening of pancytopenia where the marrow
blast count is found to be increasing (the karyotype typically remains stable);
(2) a relatively stable clinical course followed by an abrupt change with a clear
leukemic transformation, typically with a change in the karyotype with the gain
of secondary clones, and complex karyotypes; or (3) a stable course over many
years without significant change in the marrow blast counts when reevaluated,
and a stable karyotype (28). Few series with sequential cytogenetic studies have
been published, and most series are small with short follow-up periods (29–31).
Nonetheless, karyotypic evolution in MDS is associated with transformation to
acute leukemia in about 60% of cases and reduced survival, particularly for those
patients who evolve within a short period of time (⬍100 days) (30).
CYTOGENETIC FINDINGS IN MDS

Normal Karyotype
A normal karyotype is found in 30% to 60% of patients with MDS. This group of
patients is almost certainly genetically heterogeneous. Technical factors may pre-
clude the detection of chromosomally abnormal cells, or leukemogenic alterations
can occur at the molecular level and be undetectable with standard cytogenetic
methods. Nonetheless, despite this heterogeneity, these cases are a standard ref-
erence for comparison of outcomes. The International MDS Risk Assessment
Workshop found that patients with a normal karyotype fall within the favorable
risk group. The median survival for these patients is 3.8 years, and the time to
progression to AML of 25% of this cohort was 5.6 years (23).
Loss of the Y Chromosome: −Y

The clinical and biological significance of the loss of the Y chromosome, −Y, is
unknown. Loss of the Y chromosome has not only been observed in a number of
malignant diseases, but has also been reported to be a phenomenon associated with
aging. The United Kingdom Cancer Cytogenetics Group undertook a comprehen-
sive analysis of this abnormality in both normal and neoplastic bone marrows
(32). A −Y could be identified in 7.7% of patients without evidence of a malig-
nant hematologic disease, and in 10.7% of patients with MDS; thus, the presence
of a −Y alone was not reliable in documenting a malignant process. Analysis of a
large series of 215 male patients found that patients with a hematological disease
had a significantly higher percentage of cells with a −Y (52% vs. 37%; p = 0.036)
(33). In this series, the presence of −Y in ⬎75% of metaphase cells accurately
predicted a malignant hematological disease. While loss of a Y chromosome may
not be diagnostic of MDS, once the disease is identified by clinical and pathologic
means, the International MDS Risk Analysis Workshop found that −Y as the sole
cytogenetic abnormality conferred a favorable outcome (23).
Deletions of Chromosome 20: del(20q)

A deletion of the long arm of chromosome 20, del(20q), is seen in ∼5% of
MDS cases and 7% of t-MDS cases (5). Clinical features characterizing MDS
patients with isolated del(20q) include low-risk disease (usually RA), low rate of
progression to AML, and prolonged survival (median of 45 months vs. 28 months
for other MDS patients) (34). Morphologically, the presence of a del(20q) is
associated with prominent dysplasia in the erythroid and megakaryocytic lineages
(35). The International MDS Risk Analysis Workshop noted that patients with a
del(20q) observed in association with a complex karyotype identified a poor-risk
group with a median survival for the entire poor-risk group of 9.6 months, whereas
the prognosis for patients with an isolated del(20q) was favorable (23).
Loss of Chromosome 5 or del(5q): −5/del(5q)

In MDS or AML arising de novo, loss of a whole chromosome 5, or a deletion
of its long arm, −5/del(5q), are observed in 10% to 20% of patients, whereas it
is identified in 40% of patients with t-MDS/t-AML (Fig. 3) (5,36). A significant
occupational exposure to potential carcinogens is present in many patients with
AML or MDS de novo and either −5/del(5q) or a −7/del(7q) (discussed below),
suggesting that abnormalities of chromosome 5 or 7 may be a marker of mutagen-
induced hematological malignant diseases (37).
In primary MDS, abnormalities of chromosome 5 are characteristically
observed in the 5q− syndrome (described below; see also chap. 12). More
del(5) (q14q33) del(7) (q11.2q36)
q14 q11.2
q33
q36
5 del(5q) 7 del(7q)
Figure 3 Deletions of 5q and 7q in myeloid neoplasms. In this del(5q), breakpoints occur

in q14 and q33 resulting in interstitial loss of the intervening chromosomal material. In this
del(7q), breakpoints occur in q11.2 and q36. In both cases, the critical commonly deleted
segments are lost. Normal chromosome 5 and 7 homologs are shown for comparison.
commonly, however, they are observed in RAEB as part of a complex karyotype.

Clinically, the patients with del(5q) coupled with other cytogenetic abnormalities
have a poor prognosis, with early progression to leukemia, resistance to treatment,
and short survival. Abnormalities of 5q are associated with previous exposure to
standard and high-dose alkylating agent therapy, including use in immunosuppres-
sive regimens (38–41). A role for exposure to benzene (42), as well as therapeutic
ionizing radiation (43,44), as risk factors for MDS is emerging.
The 5q− Syndrome

MDS with an isolated del(5q) (5q− syndrome) represents a distinct clinical
syndrome characterized by a del(5q) as the sole karyotypic abnormality (12,45).
Unlike the male predominance in MDS in general, the 5q− syndrome has an
overrepresentation of females (2:1). The initial laboratory findings are usually
a macrocytic anemia with a normal or elevated platelet count. The diagnosis is
usually RA (in two-thirds), or RAEB (in one-third); some investigators exclude
patients with excess blasts from the definition of 5q− syndrome. On bone
marrow examination, abnormalities in the megakaryocytic lineage (particularly
micromegakaryocytes) are prominent. These patients have a favorable outcome,
in fact the best of any MDS subgroup, with low rates of leukemic transformation,
and a relatively long survival of several years duration (23,45). The loss of a single
copy of the RPS14 gene may be involved in the pathogenesis of this syndrome,
as described below (46).
Gain of Chromosome 8: +8
The incidence of a gain of chromosome 8 in MDS is ∼10%. This abnormality is
observed in all MDS subgroups, varying in frequency with age, gender, and prior
treatment with cytotoxic agents or radiation (5,19,23,47). The prognostic signif-
icance of the gain of chromosome 8 in MDS patients is not fully characterized.
The situation is complicated in that +8 is often associated with other recurring
abnormalities known to have prognostic significance, for example, −5/del(5q) or
−7/del(7q), and may also be seen in isolation as a separate clone unrelated to
the primary clone in up to 5% of cases. The presence of cryptic abnormalities
at other sites within the genome in association with +8 has also been described
in some cases using molecular methods (48), which may explain the variability
in the clinical course reported in patients with trisomy 8. The International MDS
Risk Analysis Workshop ranked this abnormality in the intermediate-risk group
(23), and this ranking remains unchanged with the newly proposed time-dependent
score of the WPSS (24). In univariate analysis, one large confirmatory study found
that +8 as a sole abnormality had a worse behavior than expected for an inter-
mediate IPSS risk group, which was also the case in a large retrospective study
(22,49). This latter study found that the prognosis improved with one additional
abnormality, but worsened with more than one additional abnormality.
Loss of Chromosome 7 or del(7q): −7/del(7q)

A −7/del(7q) is observed as the sole abnormality in ∼5% of adult patients with
de novo MDS (21,49), but in ∼50% of children with de novo MDS (50) and in
∼55% of patients with t-MDS (Fig. 3) (5,36). It can occur in three clinical settings
[reviewed in (51)]: (1) de novo MDS and AML, (2) myeloid leukemia associated
with constitutional predisposition, and (3) t-MDS/t-AML. The similar clinical and
biological features of the myeloid disorders associated with −7/del(7q) suggest
that the same gene(s) is altered in each of these contexts. The IPSS considers the
−7/del(7q) to be a poor prognostic cytogenetic finding (23).
“Monosomy 7 Syndrome” has been described in young children (see
chap. 14). It is characterized by a preponderance of males (∼4:1), hepato-
splenomegaly, leukocytosis, thrombocytopenia, and a poor prognosis (52,53).
Juvenile myelomonocytic leukemia (JMML, previously known as juvenile chronic
myelogenous leukemia; see chap. 14), is a myelodysplastic/myeloproliferative
(MDS/MPD) disease in the WHO classification, and shares many features with
this entity; −7 is observed either at diagnosis or as a new cytogenetic finding asso-
ciated with disease acceleration (51). An emerging paradigm is that −7 cooperates
with deregulated signaling via the RAS pathway in the pathogenesis of JMML.
Activation of the RAS pathway occurs as a result of mutations in the NRAS or KRAS
gene, inactivating mutations in the gene encoding NF1, a negative regulator of RAS
proteins, or activating mutations in the gene encoding the PTPN11/SHP2 phos-
phatase, a positive regulator of RAS proteins. In constitutional disorders associated
with a predisposition to myeloid neoplasms, including Fanconi anemia, neurofi-
bromatosis type 1, and severe congenital neutropenia, a −7/del(7q) is the most fre-
quent cytogenetic abnormality detected. As with −5/del(5q), occupational or envi-
ronmental exposure to mutagens including chemotherapy, radiotherapy, benzene
exposure, and smoking (54), as well as severe aplastic anemia (regularly treated
with immunosuppressive agents alone) have been associated with −7/del(7q).
The 17p− Syndrome

Loss of the short arm of chromosome 17 (17p−) has been reported in up to 10%
of patients with MDS. This loss can result from various abnormalities, including
simple deletions, unbalanced translocations, dicentric rearrangements (particu-
larly with chromosome 5), or less often, −17 or isochromosome formation (55).
The dic(5;17)(q11.1–13;p11.1–13) is a frequent recurring rearrangement (56,57).
Approximately one-third of these patients have t-MDS (58) and most have com-
plex karyotypes. The most common additional changes are −7/del(7q) and +8.
Morphologically, the 17p− syndrome is associated with a characteristic
form of dysgranulopoiesis combining pseudo-Pelger-Hüet hypolobulation and the
presence of small granules in granulocytes. Clinically, the disease is aggressive
with resistance to treatment and short survival. The TP53 (p53) gene, an important
tumor suppressor gene that functions in the cellular response to DNA damage, is
located at 17p13.1. In these cases, one allele of TP53 is typically lost as a result of
the abnormality of 17p; an inactivating mutation in the second allele on the remain-
ing, morphologically normal chromosome 17 occurs in ∼70% of cases (56,57).
Complex Karyotypes
Complex karyotypes are variably defined, but generally involve the presence of
≥3 chromosomal abnormalities. The majority of cases with complex karyotypes
involve unbalanced chromosomal abnormalities leading to the loss of genetic
material, on chromosomes 5, 7, or both. Complex karyotypes are observed in
∼20% of patients with primary MDS, and in as many as 90% of patients with
t-MDS, and carry a poor prognosis (22–24,28,59,60).
RARE RECURRING TRANSLOCATIONS

Translocations of 11q23
The MLL (Mixed Lineage Leukemia) gene (also known as ALL1, HTRX, HRX) is
involved in over 50 reciprocal translocations in acute leukemia (61). In a European
workshop of 550 patients with 11q23 abnormalities, 28 cases (5.1%) presented
with an MDS, and five others with such an abnormality had evolved from t-MDS
to t-AML prior to cytogenetic analysis, for a total of 6% of all cases examined.
One-fourth of these cases had t-MDS (62). Other abnormalities, including complex
karyotypes and a −7/del(7q), frequently accompany the 11q23 abnormalities in
both primary MDS and t-MDS. No association with FAB subgroup was identified,
although RA was overrepresented, and RARS was underrepresented as compared
to most series of MDS patients. The median survival was short (19 months)
with leukemic transformation in ∼20% of cases. The classic association of prior
exposure to topoisomerase II inhibitors with the development of t-MDS/t-AML
with translocations of 11q23 was not confirmed in this workshop, but this may
simply reflect the relatively small number (n = 23) of cases with full treatment
details (63).
Slightly less than 12% of the 162 patients with 11q23 involvement included
in an International Workshop on MDS and Leukemia following cytotoxic treat-
ment presented with a t-MDS (44,64). One third (6/19) of these patients had
progression to an acute leukemia (5 AML, 1 ALL). This study also did not find
a clear association of 11q23 abnormalities with FAB subtype. The most common
translocations were t(9;11)(p22;q23) in six cases, t(11;19)(q23;p13.1) in three
cases, and t(11;16)(q23;p13.3) in three cases.
The t(11;16) Translocation

The t(11;16)(q23;p13.3) occurs primarily in t-MDS, but rare cases have presented
as t-AML (Fig. 4) (65). The t(11;16) is unique among MLL translocations in that
most patients have t-MDS. The MLL gene on chromosome 11 is fused with the
t(11;16) (q23;p13.3)
p13.2
q23
11 der(11) 16 der(16)
Figure 4 t(11;16)(q23;p13.3): in the t(11;16), breakpoints occur in 11q23 and 16p13.3,

followed by a reciprocal exchange of chromosomal material. The 5 end of the MLL gene at
11q23 is fused to the 3 end of the CREBBP gene from 16p13.3 to form the MLL/CREBBP
fusion gene on the der(11). Arrowheads indicate the breakpoints. Normal chromosome 11
and 16 homologs are shown for comparison.
CREBBP (CREB binding protein) gene on chromosome 16. The MLL protein
is a histone methytransferase that assembles in protein complexes that regulate
gene transcription, for example, HOX genes during embryonic development, via
chromatin remodeling. CREBBP is a histone acetyltransferase involved in tran-
scriptional control via histone acetylation, which mediates chromosome decon-
densation, thereby facilitating transcription. Both genes have multiple transloca-
tion partners in various hematological disorders; thus elucidating their function is
providing new insights in leukemia research.
The Platelet-Derived Growth Factor Receptor Beta Translocations

The t(5;12)(q32;p13) is observed in ∼1% of patients with CMML. The transloca-
tion creates a fusion gene, and the encoded fusion protein contains the 5 portion of
TEL gene on chromosome 12 (also known as ETV6) and the 3 portion of PDGFRB,
encoding the beta chain of the platelet-derived growth factor receptor (PDGFRB)
on chromosome 5 (66). The PDGFRB kinase activity is perturbed contributing
to the transformed phenotype. TEL encodes a transcriptional repressor, and is
promiscuously involved in translocations with some 40 genes in hematologic
malignancies (61). Interest has increased in identifying this translocation, which
predicts for a response to imatinib mesylate, a selective inhibitor of the tyrosine
kinase activity of the PDGFRB protein (26). Similarly, PDGFRB participates in
other rare translocations involving genes encoding the membrane-associated pro-
tein HIP1 (Huntington interacting protein 1) in the t(5;7)(q32;q11.2) (67), the small
GTPase RABEP1 (Rabaptin 1) in the t(5;17)(q32;p13) (68), CCDC6 (a ubiqui-
tous coiled-coil domain protein of unknown function) in the t(5;10)(q32;q21)
(69) to produce CMML, and CEV14 (clonal evolution–related gene on
chromosome 14, also known as TRIP11, thyroid hormone receptor interactor

11) in the t(5;14)(q32;q32) in a case of AML (70). A unifying feature of these
various translocations is the presence of eosinophilia.
Translocations of 3q
The t(3;21)(q26.2;q22.1) has been linked to AML arising after cytotoxic therapy.
This abnormality was first recognized in CML in blast crisis (71) and later in
t-MDS/t-AML (72). The RPL22L1 (EAP) gene at 3q26.2 encodes a highly
expressed small nuclear protein associated with EBV small RNA (EBER1),
and is fused with the RUNX1 (Runt-related transcription factor, also known as
AML1) gene at 21q22.1, producing a fusion protein that retains the DNA-binding
sequences of RPL22L1, with loss of function of RUNX1.
Further work has identified two additional genes 400–750 kb centromeric
to RPL22L1, also at 3q26.2, namely MDS1 (MDS associated sequences) and
EVI1 (Ecotropic Virus Insertion site −1) encoding nuclear transcription factors
containing DNA-binding zinc finger domains. The proteins encoded by these two
genes are identical other than an N-terminal extension of 12 amino acids in the
MDS1 protein, representing a splicing variant. As for EAP, RUNX1 transcripts
fused with MDS1/EVI1 sequences are also produced in some cases, and result
in the loss of the first 12 amino acids, producing a novel EVI1 protein; the
phenotype is one of arrested differentiation, which leads to apoptosis in vitro
(73). MDS1/EVI1 also serves as a translocation partner with the ribosome binding
protein RPN1 (ribophorin 1) (74) and/or the C3ORF27 gene; the latter encodes a
poorly characterized protein in fetal development (75) in the inv(3)(q21q26.2) or
the t(3;3)(q21;q26.2), which are karyotypes associated with normal or increased
platelet counts.
Common features of myeloid diseases associated with abnormalities of 3q
are a previous history of cytotoxic exposure, prominent bone marrow dysplasia,
and a poor prognosis. Abnormalities of chromosome 7 [−7/del(7q)] are observed
in most cases with rearrangements of 3q. In an International Workshop on Therapy-
Related Hematologic Disease, inv(3)/t(3;3) abnormalities were the most frequent
of the 3q abnormalities (76).
THE GENETICS OF THE MYELODYSPLASTIC SYNDROMES

Molecular Models for Chromosome Abnormalities in MDS
As described earlier, many of the recurring chromosomal abnormalities in MDS
lead to the loss of genetic material. The genetic consequences of a deletion may be
a reduction in the level of one or more critical gene products (haploinsufficiency),
or complete loss of function. The latter model, widely known as the “Knudson
two-hit model”, predicts that loss of function of both alleles of the target gene
would occur: in one instance, through a detectable chromosomal loss or deletion,
and in the other, as a result of a subtle inactivating mutation or another mechanism,

such as transcriptional silencing (77).
A clinical example that may illustrate this principle is t-MDS/t-AML. The
relatively long latency period between the time of exposure and t-MDS/t-AML
with abnormalities of chromosomes 5 or 7 (∼5 years) is compatible with a two-step
mechanism, in which two mutations of a target gene must occur in a hematopoietic
stem/progenitor cell. These patients may have two normal alleles initially, one of
which is mutated as a result of therapy. Subsequent stochastic loss of the other allele
in a bone marrow stem cell would then contribute to leukemogenesis. Alternatively,
because t-AML develops in only 5% to 10% of patients who are treated for a
primary tumor, these individuals may have inherited a predisposing mutant allele;
subsequent mutagenic exposure may then induce the second mutation, eventually
giving rise to leukemia when additional cooperating mutations arise. In these cases,
characterization of the predisposing mutations will be important in identifying
individuals who are at risk of developing t-AML, which might aid in the selection
of the appropriate therapy for the primary malignant disease.
In an alternative model, loss of only a single copy of a gene may result
in a reduction in the level of one or more critical gene products (haploinsuffi-
ciency). There is growing evidence that a number of leukemia-related genes are
haploinsufficient, for example, TP53, SPI1/PU.1, and RUNX1 (78–82). In humans,
haploinsufficiency of the RUNX1 gene results in a familial platelet disorder with
a predisposition to AML (80,83,84). Support for this mechanism was provided by
the subsequent identification of point mutations in the RUNX1 gene in sporadic
cases of MDS and AML (84). Despite intensive efforts, neither homozygous dele-
tions nor inactivating mutations in the remaining allele of candidate genes located
within the commonly deleted segments (CDS) have been detected in myeloid
leukemia cells characterized by deletions of 5q, 7q, or 20q in MDS and AML.
This observation is compatible with a haploinsufficiency model in which loss of
one allele of the relevant gene (or genes) alters cell fate. Finally, it remains theo-
retically possible that loss of function of different genes on 5q or 7q (outside of
the CDS) contributes to the pathogenesis of AML, or that loss of more than one
gene, that is, a contiguous gene syndrome, is necessary for leukemogenesis.
Molecular Analysis of the −5/del(5q)

Several groups of investigators have defined a CDS on the long arm of
chromosome 5 predicted to contain a myeloid tumor suppressor gene that is
involved in the pathogenesis of MDS and AML (Fig. 5) (85–87). By cytogenetic
and FISH analyses, Michelle Le Beau and her colleagues defined a 970 kb
CDS within 5q31 flanked by D5S479 and D5S500 (85,88). Molecular analysis
of 20 candidate genes within the CDS of 5q31 neither revealed inactivating
mutations in the remaining alleles, nor was there any evidence of transcriptional
silencing (85,88) (Lucy Godley and Michelle Le Beau, unpublished data).
Commonly deleted segment of 5q
CSF2/IL3
TCF1B
IL9 Fairman et al..
q11
SMAD5
q12 D5S393
q13 D5S89 MDS/AML &
q14 D5S399 t-MDS/t-AML
Zhao et al.
q15 SPOCK
q21 D5S479
q22 D5S414 970 kb
q23 EGR1 Horrigan et al.
D5S500
q31
D5S476
q32
CTNNA1
q33
CSF1R Jaju et al.
q34
PDGFRB
q35 5q-syndrome
RPS14
5q SPARC
GLRA1
Figure 5 Ideogram of the long arm of chromosome 5 showing chromosome markers and
candidate genes within the commonly deleted segments (CDSs) as reported by various
investigators. The proximal CDS in 5q31 was identified in MDS, AML, and t-MDS/AML,
whereas the distal CDS in 5q32–33 was identified in the 5q− syndrome.
These observations are compatible with a haploinsufficiency model, and several

candidate haploinsufficient genes have been identified on 5q.
One such candidate is the early growth response 1 gene (EGR1). EGR1, a
member of the WT1 family of transcription factors, is downstream of cytokine
signaling pathways, and is a direct transcriptional activator of TP53 and CDKN1
A/p21. Joslin and colleagues at the University of Chicago demonstrated that loss
of a single allele of Egr1 cooperates with mutations induced by an alkylating
agent in the development of myeloid diseases in mice (89). Liu and colleagues
demonstrated that the gene encoding alpha-catenin (CTNNA1, located 285 kb distal
to the CDS defined in 5q31) is expressed at lower levels in AML or MDS with a
del(5q), than in other AMLs or normal hematopoietic stem cells (90). Moreover,
restoration of CTNNA1 expression in HL-60 cells (used as a model for AML with
loss of 5q) resulted in reduced proliferation and apoptotic cell death. These studies
raise the possibility that loss of expression of EGR1 and CTNNA1 in HSCs may
contribute to the pathogenesis of MDS or AML with a del(5q).
Molecular analyses of the 5q− syndrome suggest that a different region
is involved. Boultwood and her colleagues in Oxford identified a 1.5-Mb CDS
within 5q32 between D5S413 and GLRA1 which contains 44 genes (91). This
region is distal to the CDS in 5q31 found in the patients with RAEB-1, RAEB-2,
and AML with del(5q). Ebert and colleagues used an RNA interference screen to
reduce expression levels of each gene within the 5q CDS, and identified the gene
encoding RPS14, a ribosomal protein, as a haploinsufficient tumor suppressor
gene that has striking effects on erythropoiesis and minimal effects on megakary-
opoiesis (46). Moreover, they demonstrated that expressing RPS14 in CD34+
cells from patients with the 5q− syndrome–enhanced erythroid differentiation
and normalized the activation level of genes specifically expressed in red blood
cell precursors. RPS14 is an essential component of the 40 S subunit of ribo-
somes (sites of protein synthesis), and ribosome synthesis is impaired in CD34+
cells from 5q− syndrome patients. It is notable that two other ribosomal genes—
RPS19 and RPS25—are mutated in individuals with Diamond Blackfan anemia,
a congenital form of anemia that shares some features of the 5q− syndrome (46).
The role of RPS14 in 5q− syndrome, and whether this gene plays a role in the
pathogenesis of other subtypes of MDS or AML, remain to be determined.
In summary, the existing data suggest that there are two nonoverlapping
CDSs in 5q31 and 5q32 in clonal myeloid disorders. The proximal segment in
5q31 is likely to contain a tumor suppressor gene involved in the pathogenesis
of both de novo and therapy-related MDS/AML. Band 5q32 contains a second
myeloid tumor suppressor gene involved in the pathogenesis of the 5q− syndrome.
Molecular Analysis of the −7/del(7q)

As with the −5/del(5q), the breakpoints and extent of the deletions of 7q in patients
have been investigated to identify a CDS (92–94). Data from cytogenetic, FISH,
and loss of heterozygosity (LOH) studies, performed in a number of laboratories,
paint a complex picture of 7q deletions in myeloid malignancies; however, there
is general agreement that 7q22 is involved in a majority of cases. Defining a
consistent CDS has been hampered by the relatively low frequency of del(7q)
compared to the complete loss of chromosome 7, as well as the existence of
multiple and sometimes complex cytogenetic abnormalities in most cases.
By using cytogenetic and FISH analysis of 81 patients with de novo and
therapy-related MDS/AML with a del(7q), Le Beau and colleagues identified
two distinct CDSs, a 2.52-Mb CDS within 7q22 spanning the interval containing
LRCC17 and SRPK2, and a second, less frequent, region in q32–33 (92). Each
of the candidate genes within the CDS at 7q22 has been evaluated for mutations
(95); however, no inactivating mutations have been identified in the remaining
allele. Tosi and her colleagues in Oxford evaluated patients with 7q abnormalities
and identified an interesting patient with a complex karyotype and a t(7;7), who
had a deletion associated with the translocation breakpoint of 150 kb spanning
the CUTL1 locus in 7q22, slightly centromeric to the CDS defined by Le Beau
and colleagues (94). Recently, Döhner and colleagues in Heidelberg reported the
analysis of a large series of patients with abnormalities of 7q using FISH. Whereas
most patients had large deletions, they identified an ∼2-Mb deleted segment in
proximal q22, that overlaps with the proximal portion of the CDS defined by
Le Beau and colleagues, but extends more proximally, and includes the CUTL1,
RASA4, EPO, and FBXL13 genes in 7q22.1 (96).
Molecular Analysis of the del(20q)

By using FISH analysis, combined with LOH studies, investigators have identified
an interstitial CDS of 4 Mb within 20q12 that is flanked by D20S206 proximally
and D20S424 distally, containing a number of genes. Despite the availability of
detailed physical maps and transcripts maps, the identity of a myeloid tumor sup-
pressor gene on 20q is unknown (97,98). Recent studies have implicated the genes
encoding topoisomerase 1 and lethal(3) malignant brain tumor (L3MBTL), which
is related to the Polycomb group family of transcriptional repressors. Although
L3MBTL is not mutated in MDS, reduced or absent L3MBTL expression may be
relevant in some cases of myeloid leukemia (99).
Alterations in Gene Function

A growing body of evidence suggests that mutations of multiple genes mediate
the pathogenesis and progression of MDS. The involved genes fall into two main
classes, namely, genes encoding hematopoietic transcription factors or proteins
that regulate cytokine signaling pathways. There is an increase in the frequency
of molecular mutations from low-risk to high-risk MDS, or AML evolving from
MDS, emphasizing the role of these mutations in disease progression. Moreover,
the progression from the early stages of MDS to AML is often accompanied
by the acquisition of molecular mutations that are known to play an important
role in AML, e.g., NRAS point mutations. A detailed review of these genes
is beyond the scope of this chapter. Tables 4 and 5 provide a partial list and
overview of some of the salient features of genes implicated in the pathogenesis
of MDS.
The RAS signaling cascade is downstream of a number of activated cytokine
receptors, including the FLT3, IL3, and GM-CSF receptors; thus, this signaling
pathway plays a pivotal role in hematopoiesis. In MDS, constitutively activating
point mutations of NRAS, typically involving codons 12, 13, or 61, have been
detected in 10% to 15% of cases. These mutations have been associated with a poor
prognosis with a higher incidence of transformation to AML and shorter survival
(100–105). Targeted therapies, including the farnesyl transferase inhibitors and
imatinib, interrupt various steps in the RAS signaling pathways (26,106).
Mutations of the FMS-like tyrosine kinase 3 (FLT3) gene, including both
point mutations within the tyrosine kinase domain (TKD) and internal tandem
duplications (ITDs), are among the most common genetic changes seen in AML,
occurring in 15% to 35% of cases. FLT3-ITD mutations are associated with a
poor prognosis, particularly in those cases with loss of the remaining wild-type
FLT3 allele. In MDS, FLT3-ITD mutations are rare in RA/RARS, and increase
to ∼3% and 12% in RAEB and AML following MDS, respectively (105). Thus,
68
Table 4 Frequency of Molecular Mutations in MDS

MDS subtypea
MDS total t-MDS/t-AML

Mutated gene RA (%) RARS (%) RAEB (%) RAEB-t (%) CMML (%) (%) (%)
FLT3 (ITD) 0 0 2.5 8 4.5 2.4 0

FLT3 (TKD) NA NA NA NA NA 1 ⬍1
NRAS 10 13 9% 13 35 10–15 10
KIT D816 0 NA ⬍1 ∼10 0 ∼1 NA
MLL (ITD) 2 NA 3 NA 3 3 2–3
RUNX1 4 NA 10 15 20 10–15 15–30
TP53 NA NA NA NA NA 5–10 25–30
PTPN11 NA NA NA NA NA ∼1 3
NPM1 NA NA NA NA NA Rare 4–5
CEBPA NA NA 4 NA 10–20 1–8 Rare
JAK2V617F Rare Rare Rare Rare 3 2–5 2–5
a Note that the FAB nomenclature for MDS is used in this table, reflecting the available literature. RAEB-t, refractory anemia with excess blasts in transformation
(reclassified as acute myeloid leukemia in the WHO classification); NA, not available.
Olney and Le Beau
Table 5 Genes Altered in the Myelodysplastic Syndromes

Gene Alteration Associated features Reference
BCL2 Overexpressed in all Encodes a protein product that (42,128,129)

FAB subtypes suppresses apoptosis
No correlation with survival
Highest in higher risk entities
where apoptosis is reduced
CEBPA Mutated in 1–8% of Encodes a transcription factor (113)
MDS, higher in essential for granulopoiesis
CMML Mutations appear to have no
(10–20%) effect on time to overall
progression or overall survival
CSF1R/ Mutated in 12–20%, Encodes the macrophage (102,130)
FMS increased with colony-stimulating factor
higher risk MDS receptor with TK activity
Karyotype predominantly
normal
Increased transformation to
AML and poor survival
FLT3 Internal tandem Encodes a class III receptor TK (105,131,132)
duplication (ITD) involved in cytokine signaling
in ∼10% of MDS and stem cell differentiation
and AML with ITD results in constitutive
trilineage activation of protein
dysplasia associated with progression to
AML and poor prognosis
Frequently observed with a
normal karyotype in AML
GCSFRG Point mutations Encodes the G-CSF receptor (133)
identified Severe congenital neutropenia
(SCN) patients with G-CSF
receptor defects can progress
to MDS and/or AML, with
loss of chromosome 7 and
NRAS/KRAS mutations
Mutation alone is not sufficient
for transformation
JAK2V617F Mutated in 2–5% of Encodes a TK component of (117)
MDS cytokine signaling pathways
Activating mutations result in
constitutive signaling
Mutation in 60% of patients with
RARS with thrombocytosis,
an unclassified MDS/MPD
(Continued)
Table 5 Genes Altered in the Myelodysplastic Syndromes (Continued)

KIT Overexpressed; rare Encodes the stem cell factor receptor (134,135)
mutations in May provide an autocrine growth
MDS, KITD816 pathway
MDR1 Expressed in ∼60% Encodes a transmembrane drug efflux (136)
pump
May be involved in resistance of MDS to
drug therapy associated with
monosomy 7
MDM2 Overexpressed in Encodes a protein (murine double (137,138)
∼70% minute-2) which abrogates the function
of p53 via ubiquitination/degradation
Gene amplification not detected
Associated with unfavorable cytogenetic
abnormalities shorter remission
duration
MLL Internal tandem Encodes a histone methyltransferase that (105)
duplication in 3% regulates gene transcription via
of MDS chromatin remodeling
Increased mutation frequency in AML
following MDS
MPL Overexpressed in Encodes the thrombopoietin receptor (139,140)
∼45% of CMML, Higher expression in RAEB and RAEB-t
and ∼40% of associated with poor prognosis,
RAEB, RAEB-t; increased progression to AML
underexpressed Correlated with
(∼50% of normal dysmegakaryocytopoiesis
levels) in most
RA patients
NF1 Loss and mutations Encodes neurofibromin, a tumor (141,142)
identified, suppressor gene product, that functions
particularly in as a GTPase activating (GAP) protein
pediatric to downregulate RAS function
MDS/MPS High incidence of MDS and AML in
children with NF1
NPM1 Mutations rare in Encodes a protein with diverse functions (110)
MDS, 5% in in the cell, including chromatin
t-MDS remodeling, genome stability,
ribosome biogenesis, DNA duplication
and transcriptional regulation
Mutations result in C-terminus
alterations, and aberrant protein
localization to the cytoplasm

NRAS Mutated in 10–15%; GTPase component of cytokine (102,105)

overexpressed in signaling pathways
RA, RARS Activating mutations result in
constitutive signaling
associated with monocytic
component
Increased risk of progression to
AML overexpression may be
an early event in
transformation
Associated with −7/del(7q)
CDKN2B/ Decreased expression Closely associated with deletion (118)
p15INK4B via DNA or loss of 7q
methylation in 68% Independently associated with
of t-MDS/t-AML poor survival
PTPN11 Somatic missense A non-receptor tyrosine (121)
mutations in 33% phosphatase that relays
of JMML patients signals from activated growth
factor receptors to RAS
proteins
Mutations of NRAS/KRAS, NF1,
and PTPN11 are mutually
exclusive
RUNX1/AML1 Mutated in 10–15% Encodes the DNA-binding (107,108)
MDS, 15–30% subunit of the heterodimeric
t-MDS core-binding factor (CBF)
complex, which is essential
for definitive hematopoiesis
Point mutations in the Runt
(DNA-binding) domain result
in loss of function and a
dominant negative effect
Associated with activating
mutations of the RAS
pathway, −7/del(7q), and a
shorter overall survival
Telomerase Increased activity late Enzyme complex responsible (143,144)
(including in disease, for chromosome telomere
TERT, TR, particularly TERT maintenance and replication
and TP1) Abnormal telomere maintenance
may be an early indication of
genetic instability
Telomeres shortened with
disease progression
(Continued)

TP53 Mutated in 5–25%; Checkpoint protein, which monitors (114,115,145)

higher frequency integrity of genome, and arrests
in t-MDS cell cycle in response to DNA
damage
Loss of wild type allele
Observed as both early and late
genetic event in MDS
Associated with rapid progression and
poor outcome seen with loss of 17p,
−5/del(5q), suggesting pathogenic
exposure to carcinogens
Identifies poorer prognosis within each
IPSS subgroup
WT1 Associated with Correlated with blast counts and (146)
overexpression cytogenetic abnormalities
(∼70% of MDS, Significantly correlated with IPSS
and 100% of score
t-AML)
FLT3-ITD may represent a secondary event associated with MDS progression,

rather than the initiating event in the pathogenesis of MDS. Point mutations of the
FLT3 TKD (codons 835 or 836 of the second tyrosine kinase domain) are noted
in 5% to 8% of AML, but are rare in MDS (105).
Recurring translocations involving the MLL gene at 11q23 are uncommon in
MDS; however, a partial tandem duplication of exons 3–9, 3–10, or 3–11 has been
reported in MDS (3%), and in a higher percentage of cases of AML arising from
MDS (7%) (105). The MLL protein is a histone methyltransferase that assembles
in protein complexes that regulate gene transcription via chromatin remodeling.
Target genes of the MLL transcriptional regulatory complexes include the HOX
genes, which have a critical role in development as well as hematopoiesis.
The Runt-related transcription factor 1 (RUNX1), also known as AML1,
encodes the DNA-binding subunit of the heterodimeric core-binding factor (CBF)
complex, which is essential for definitive hematopoiesis. Point mutations in the
RUNX1 Runt (DNA-binding) domain have been reported in AML and MDS (12%),
particularly in MDS secondary to atomic bomb radiation exposure or treatment
with cytotoxic therapy, and increase with the severity of the disease. Moreover,
RUNX1 mutations are associated with activating mutations of the RAS pathway,
−7/del(7q), and a shorter overall survival (107,108).
Mutations of NPM1 also occur frequently in AML (35% of adult cases),
but are less frequent in patients with recurring cytogenetic abnormalities. In the
absence of FLT3 mutations, NPM1 mutations are associated with a favorable

prognosis (109). NPM1 mutations, most commonly involve exon 12, resulting
in alterations at the C-terminus, that is, replacement of tryptophan(s) at position
288 and 290 and creation of a nuclear export signal (NES) motif, which mediates
aberrant localization of the protein to the cytoplasm. Few studies have examined
NPM1 mutations in MDS; however, mutations have been reported in rare cases of
MDS, and in ∼5% of patients with t-MDS (110). It is notable that the NPM1 gene
located at 5q35 is not mutated in MDS with a del(5q) (111).
Mutations of the CCAAT/enhancer-binding protein-alpha transcription fac-
tor gene (CEPBA) occur in 6% to 15% of AML de novo, and are generally asso-
ciated with a favorable prognosis (112). Out-of-frame insertions and deletions
in the N-terminal region, which result in a truncated dominant-negative 30 kDa
protein that lacks transactivation activity, and in-frame insertions and deletions
in the C-terminal region are common. Similar mutations have been identified in
MDS, ranging in frequency from 0% to 8% of cases. In MDS, CEBPA mutations
are seen at the highest frequency in CMML (15–20%), and mutations may occur
either as an early or late event in the course of the disease (113). In this small
series, CEBPA mutations had no effect on time to AML progression or overall
survival.
The TP53 tumor suppressor gene encodes an essential checkpoint protein
that monitors the integrity of the genome, and arrests cell cycle progression in
response to DNA damage. Mutations of TP53 are observed in primary MDS
(5–10%) and, more commonly, in t-MDS (25–30%) (114,115). The spectrum
of mutations includes missense mutations in exons 4–8, as well as loss of the
wild-type allele, typically as a result of a cytogenetic abnormality of 17p. TP53
mutations may occur as either an early or late event in the course of the disease,
and are associated with rapid progression, and a poor outcome. In t-MDS, TP53
mutations are associated with −5/del(5q) and a complex karyotype.
Point mutations of JAK2 (JAK2V617F ), resulting in a constitutively active
tyrosine kinase (TK) that is able to activate JAK-STAT signaling via the erythro-
poietin receptor, thrombopoietin receptor (MPL), or the G-CSF receptor, have
been identified in rare cases of MDS (2–5%) and CMML (3%) (116). An excep-
tion is refractory anemia with ringed sideroblasts with thrombocytosis (RARS-T),
a “myelodysplastic/myeloproliferative syndrome, unclassified” by the WHO clas-
sification (see chap. 9), in which ∼60% of patients have the JAK2V617F mutation,
and present with higher WBCs and platelet counts (117).
The role of epigenetic changes in the pathogenesis and treatment of MDS is
becoming increasingly important. Transcriptional silencing via DNA methylation
of the CDKN2B (p15INK4B ) gene is observed in a high percentage of patients with
t-MDS, and is associated with −7/del(7q), and a poor prognosis (118). Moreover,
the frequency of CDKN2B methylation increases with progression from RA to
RAEB-t. Other genes that may be affected by DNA methylation include the
CTNNA1 gene on 5q.
Genetic Pathways Leading to MDS

An important aspect of leukemia biology is the elucidation of the spectrum of chro-
mosomal abnormalities and molecular mutations that cooperate in the pathways
leading to leukemogenesis, which may lead to the identification of therapeutic tar-
gets and facilitate the development of targeted therapies (119). Kelly and Gilliland
have described an emerging paradigm in AML, namely, the cooperation between
constitutively activated tyrosine kinase molecules, such as FLT3, and aberrant tran-
scription factor fusion proteins (120). In this model, the activated tyrosine kinase
(Class I mutation) confers a proliferative and/or anti-apoptotic activity, whereas
the fusion protein (Class II mutation) impairs normal differentiation pathways,
but has a limited effect on cellular proliferation. Although our understanding of
the association of chromosomal abnormalities with gene mutations in MDS is
incomplete, several patterns of cooperating mutations have emerged, suggesting
that there are multiple genetic pathways leading to MDS.
In general, mutation of multiple Class I genes (or Class II genes) is mutually
exclusive; however, there is cooperativity between Class I and Class II mutations.
For example, −7/del(7q) has been associated with activating mutations of the
RAS pathway (activating KRAS, NRAS, or PTPN11 mutations, or inactivating
mutations of NF1), and RUNX1 mutations, as well as methylation silencing of the
CDKN2B (p15INK4B ) gene (118,121,122). TP53 (p53) mutations are uncommon
in this subgroup. In contrast, MDS associated with −5/del(5q) is associated with
TP53 mutations and a complex karyotype (114,122).
Gene expression profiling of MDS and t-MDS has also provided support for
the concept of distinct molecular and genetic subsets of MDS. Using expression
profiling of CD34+ cells in t-MDS and t-AML, Qian and colleagues (123) found
that patients with a −5/del(5q) have a higher expression of genes involved in cell
cycle control (CCNA2, CCNE2, CDC2), checkpoints (BUB1), or growth (MYC),
and loss of expression of the gene encoding interferon consensus sequence binding
protein (ICSBP/IRF8). A second subgroup of t-AML, including patients with
−7/del(7q) is characterized by downregulation of transcription factors involved in
early hematopoiesis (TAL1, GATA1, and EKLF), and overexpression of proteins
involved in signaling pathways in myeloid cells (FLT3), and cell survival (BCL2).
Similarly, expression profiling of CD34+ cells from patients with primary
MDS revealed that the expression profile of 11 selected genes could accurately
distinguish low-risk from high-risk MDS cases (124,125). Pellagatti and col-
leagues in Oxford performed expression profiling analysis of CD34+ cells from
55 patients with primary MDS; the most highly represented gene category in
both up and downregulated genes were genes involved in signal transduction. The
expression profile in MDS revealed many similarities to the pattern observed in
interferon-gamma stimulated normal CD34+ cells, for example, the two most
upregulated genes were IFIT1 and IFITM1. With respect to specific subsets, cells
from patients with RARS had a unique expression profile characterized by upreg-
ulation of mitochrondrial-related genes, and genes involved in heme synthesis,
for example, ALAS2. Interestingly, hierarchical clustering separated MDS patients

with a del(5q) from those with a normal karyotype or other abnormalities. Genes
that were differentially overexpressed in cells with a del(5q) included a num-
ber of the histone genes within the HIST1 cluster, genes encoding actin-binding
or myosin-related proteins, including ARPC2, CORO1C, MYL6, CAPZA2, and
WASPIP, as well as megakaryocyte/platelet-associated genes, including PF4V1,
PPBP, THBS1, GP1BA, and CD61. Notably, numerous genes mapping to 5q were
underexpressed.
In other studies, these investigators demonstrated that CD34+ cells from
patients with the 5q− syndrome also have a distinct expression profile (126). A
reduction in the expression level of several genes mapping within the CDS was con-
sistent with haploinsufficiency, such as SPARC, RPS14, RMB22, and CSNK1A1.
Pathway analysis revealed deregulation of the WNT/␤-catenin pathway, a critical
regulator of hematopoietic stem cells, as well as the protein ubiquitination path-
way. In a similar study, global expression profiling of the stem cell population
in the 5q− syndrome, suggested that the disease arises in CD34+CD38−Thy1+
stem cells (127).
At present, there are a number of unanswered questions regarding the molec-
ular pathogenesis of MDS. For example, we neither know the full spectrum of
genetic mutations in MDS within each pathway, nor do we know the order in which
these mutations occur, and the prognostic significance associated with many coop-
erating mutations. Several possible models are outlined in Figure 6. A variety of
experimental evidence suggests that the recurring chromosomal abnormalities in
MDS and AML are likely to be the initiating event. With respect to the recurring
translocations, the rearrangement is likely to occur in a hematopoietic progenitor
cell or, in some cases, in a committed myeloid progenitor cell. Leukemogenesis
may entail a linear process in which the initiating mutation leads to a specific
pattern of stepwise, additional mutations that complete malignant transformation.
In MDS, the process may vary somewhat in that the initiating mutations may occur
in a hematopoietic stem cell. In the setting of a normal bone marrow microen-
vironment, the initiating mutation may result in clonal expansion coupled with
emerging genetic instability (or the selection of cooperating mutations that lead to
instability), and the development of a clonal population. Selective pressures cre-
ated by both the microenvironment as well as the initiating events would lead to
the acquisition of additional mutations necessary to complete malignant transfor-
mation. Alternatively, MDS may arise in the setting of an abnormal bone marrow
microenvironment, resulting in the generation of multiple populations with vary-
ing initiating events. Some clonal populations may persist, whereas others may
undergo cell death, and yet others may go on to acquire additional mutations
necessary to complete malignant transformation. The latter model would account
for the observation of unrelated cytogenetic clones in the bone marrow of MDS
patients, as well as the observation of persistent dysplasia in MDS or AML patients
following therapy. Emerging technologies, such as the ability to culture stromal
Pathways To MDS and AML
AML,
fusion
t(8;21) N/KRAS, FLT3-ITD, KIT, or NPM1
proteins
CMP AML
q)
l(5 persists
/ de
–5 her
ot TP53 complex
persists
Normal BM
environment HSCs –7/del(7q) NRAS CDKN2B
Abnormal MDS/AML
environment
Figure 6 Models for the genetic pathways leading to MDS. See text for a discussion of
the alternative models. In the lower panel, the examples of the −5/del(5q) and −7/del(7q)
are used to illustrate the models of MDS arising in the setting of a normal bone marrow
environment versus an abnormal bone marrow environment, respectively. It is possible that
either abnormality can arise in both settings, and that each model may occur.
cell populations, and proteomics and genomics technologies may facilitate the
evaluation of these various models.
Emerging Technologies
Recent advances in microarray technology have enabled high-resolution genome-
wide genotyping using SNPs. This technology facilitates genome-wide association
studies for the identification of disease susceptibility loci, as well as the identifi-
cation of acquired abnormalities, such as genetic imbalances, for example, cryptic
deletions, duplications, and LOH without concurrent changes in the gene copy
number (referred to as copy-neutral LOH). Several recent studies have validated the
diagnostic utility of this technology in MDS. In the largest study, Gondek and col-
leagues used 250 K (250000 SNPs) arrays to examine 174 patients (94, MDS; 33,
AML following MDS; 47, MDS/MPD); the results were validated by the microar-
ray analysis of germline DNA, as well as quantitative PCR analysis (11). Acquired,
copy-neutral LOH was identified in 20% of MDS, 23% of MDS→AML, and 35%
of MDS/MPD (particularly CMML). Collectively, abnormalities were detected
in a higher proportion of cases as compared to conventional cytogenetic analysis
(78% vs. 59% for MDS). New lesions detected by microarray analysis included
copy-neutral LOH of 6p21.2-pter, 11q13.5-qter, 4q23-qter, 7q11.23-qter, and
7q22.1. With respect to the prognostic value of SNP arrays, patients with a normal
karyotype in whom new lesions were detected by SNP array analysis had a reduced
overall survival (16 months vs. 39 months) than those without new lesions. When
the presence of newly identified SNP array lesions were factored into the IPSS
classification, the survival curves diverged for patients originally classified as IPSS
intermediate-1, suggesting that SNP arrays provide additional information allow-
ing for better prognostic resolution (median survival 28 months vs. 9 months;
p = 0.03). Thus, the results of these studies suggest that SNP array analysis may
have future diagnostic application, and may complement conventional cytogenetic
analysis in risk stratification and the selection of therapy.
CONCLUDING REMARKS
The role of cytogenetic analysis in the MDSs remains a pivotal element for estab-
lishing the diagnosis and prognosis, and informs therapeutic decisions, including
the initiation of specific treatments. Cytogenetic studies are also important in the
follow-up of altered clinical behavior of the disease. MDS-associated recurring
abnormalities, while rarely specific for these disease entities, have not only pro-
vided insight into prognosis but also are beginning to shed light on the molecular
pathogenesis of these heterogeneous disorders. Coupling careful clinical observa-
tion with both classical cytogenetic techniques and newer genomics technologies
will refine our understanding of these often unpredictable myeloid diseases.
REFERENCES
1. Vardiman JW. The new World Health Organization classification of myeloid neo-
plasms: Q&A with James W. Vardiman, MD. Clin Adv Hematol Oncol 2003; 1:18,21.
hematopoiesis. N Engl J Med 2005; 352:536–538.
3. Bowen D, Culligan D, Jowitt S, et al. Guidelines for the diagnosis and therapy of
adult myelodysplastic syndromes. Br J Haematol 2003; 120:187–200.
4. Jaffe ES HN, Stein H, Vardiman JW, eds. World Health Organization Classification
of Tumours: Pathology and Genetics of Tumours of Haematopoietic and Lymphoid
Tissues. Lyon: IARC Press, 2001.
5. Vallespi T, Imbert M, Mecucci C, et al. Diagnosis, classification, and cytogenetics
of myelodysplastic syndromes. Haematologica 1998; 83:258–275.
6. Spitzer G, Verma DS, Dicke KA, et al. Subgroups of oligoleukemia as identified by
in vitro agar culture. Leuk Res 1979; 3:29–39.
7. Wells DA, Benesch M, Loken MR, et al. Myeloid and monocytic dyspoiesis as
determined by flow cytometric scoring in myelodysplastic syndrome correlates with
the IPSS and with outcome after hematopoietic stem cell transplantation. Blood
2003; 102:394–403.
8. van de Loosdrecht AA, Westers TM, Westra AH, et al. Identification of distinct
prognostic subgroups in low and intermediate-1-risk myelodysplastic syndromes by
flow cytometry. Blood 2008; 111:1067–1077.
9. Kussick SJ, Fromm JR, Rossini A, et al. Four-color flow cytometry shows strong
concordance with bone marrow morphology and cytogenetics in the evaluation for
myelodysplasia. Am J Clin Pathol 2005; 124:170–181.
10. Walker J, Flower D, Rigley K. Microarrays in hematology. Curr Opin Hematol 2002;
9:23–29.
11. Gondek LP, Tiu R, O’Keefe CL, et al. Chromosomal lesions and uniparental disomy
detected by SNP arrays in MDS, MDS/MPD, and MDS-derived AML. Blood 2008;
111:1534–1542.
12. Van den Berghe H, Michaux L. 5q-, twenty-five years later: A synopsis. Cancer
Genet Cytogenet 1997; 94:1–7.
13. Jary L, Mossafa H, Fourcade C, et al. The 17p− syndrome: A distinct myelodys-
plastic syndrome entity? Leuk Lymphoma 1997; 25:163–168.
14. Dewald GW, Pierre RV, Phyliky RL. Three patients with structurally abnormal X
chromosomes, each with Xq13 breakpoints and a history of idiopathic acquired
sideroblastic anemia. Blood 1982; 59:100–105.
15. Rowley JD. The role of chromosome translocations in leukemogenesis. Semin Hema-
tol 1999; 36:59–72.
16. Gozzetti A, Le Beau MM. Fluorescence in situ hybridization: Uses and limitations.
Semin Hematol 2000; 37:320–333.
17. Kearney L. The impact of the new fish technologies on the cytogenetics of haema-
tological malignancies. Br J Haematol 1999; 104:648–658.
18. Jotterand M, Parlier V. Diagnostic and prognostic significance of cytogenetics in
adult primary myelodysplastic syndromes. Leuk Lymphoma 1996; 23:253–266.
19. Morel P, Hebbar M, Lai JL, et al. Cytogenetic analysis has strong independent
prognostic value in de novo myelodysplastic syndromes and can be incorporated in
a new scoring system: A report on 408 cases. Leukemia 1993; 7:1315–1323.
20. Sole F, Luno E, Sanzo C, et al. Identification of novel cytogenetic markers with
prognostic significance in a series of 968 patients with primary myelodysplastic
syndromes. Haematologica 2005; 90:1168–1178.
21. Toyama K, Ohyashiki K, Yoshida Y, et al. Clinical and cytogenetic findings of
myelodysplastic syndromes showing hypocellular bone marrow or minimal dyspla-
sia, in comparison with typical myelodysplastic syndromes. Int J Hematol 1993;
58:53–61.
karyotype in MDS and correlation with subtypes: evidence from a core dataset of
2124 patients. Blood 2007; 110:4385–4395.
J Clin Oncol 2007; 25:3503–3510.
25. Garcia-Manero G, Shan J, Faderl S, et al. A prognostic score for patients with lower
risk myelodysplastic syndrome. Leukemia 2008; 22:538–543.
26. Apperley JF, Gardembas M, Melo JV, et al. Response to imatinib mesylate in patients
with chronic myeloproliferative diseases with rearrangements of the platelet-derived
growth factor receptor beta. N Engl J Med 2002; 347:481–487.
27. List A, Dewald G, Bennett J, et al. Lenalidomide in the myelodysplastic syndrome
with chromosome 5q deletion. N Engl J Med 2006; 355:1456–1465.
28. Hamblin TJ, Oscier DG. The myelodysplastic syndrome—A practical guide. Hema-
tol Oncol 1987; 5:19–34.
29. de Souza Fernandez T, Ornellas MH, Otero de Carvalho L, et al. Chromosomal alter-
ations associated with evolution from myelodysplastic syndrome to acute myeloid
leukemia. Leuk Res 2000; 24:839–848.
30. Geddes AA, Bowen DT, Jacobs A. Clonal karyotype abnormalities and clinical
progress in the myelodysplastic syndrome. Br J Haematol 1990; 76:194–202.
31. Horiike S, Taniwaki M, Misawa S, et al. Chromosome abnormalities and karyotypic
evolution in 83 patients with myelodysplastic syndrome and predictive value for
prognosis. Cancer 1988; 62:1129–1138.
32. Loss of the Y chromosome from normal and neoplastic bone marrows. United King-
dom Cancer Cytogenetics Group (UKCCG). Genes Chromosomes Cancer 1992;
5:83–88.
33. Wiktor A, Rybicki BA, Piao ZS, et al. Clinical significance of Y chromosome loss
in hematologic disease. Genes Chromosomes Cancer 2000; 27:11–16.
34. Wattel E, Lai JL, Hebbar M, et al. De novo myelodysplastic syndrome (MDS)
with deletion of the long arm of chromosome 20: A subtype of MDS with distinct
hematological and prognostic features? Leuk Res 1993; 17:921–926.
35. Kurtin PJ, Dewald GW, Shields DJ, et al. Hematologic disorders associated with
deletions of chromosome 20q: A clinicopathologic study of 107 patients. Am J Clin
Pathol 1996; 106:680–688.
36. Godley LA, Larson RA. The syndrome of therapy-related myelodysplasia and
myeloid leukemia. In: Bennett JM, ed. The Myelodysplastic Syndromes: Patho-
biology and Clinical Management. New York: Marcel Dekker Inc., 2002:136–
176.
37. West RR, Stafford DA, White AD, et al. Cytogenetic abnormalities in the myelodys-
plastic syndromes and occupational or environmental exposure. Blood 2000;
95:2093–2097.
38. Aul C, Bowen DT, Yoshida Y. Pathogenesis, etiology and epidemiology of myelodys-
plastic syndromes. Haematologica 1998; 83:71–86.
39. Larson RA, LeBeau MM, Vardiman JW, et al. Myeloid leukemia after hematotoxins.
Environ Health Perspect 1996; 104(Suppl. 6):1303–1307.
40. McCarthy CJ, Sheldon S, Ross CW, et al. Cytogenetic abnormalities and therapy-
related myelodysplastic syndromes in rheumatic disease. Arthritis Rheum 1998;
41:1493–1496.
41. Pedersen-Bjergaard J, Andersen MK, Christiansen DH. Therapy-related acute
myeloid leukemia and myelodysplasia after high-dose chemotherapy and autolo-
gous stem cell transplantation. Blood 2000; 95:3273–3279.
42. Hayes RB, Yin SN, Dosemeci M, et al. Benzene and the dose-related incidence
of hematologic neoplasms in China. Chinese Academy of Preventive Medicine—
National Cancer Institute Benzene Study Group. J Natl Cancer Inst 1997; 89:1065–
1071.
43. Fenaux P, Lucidarme D, Lai JL, et al. Favorable cytogenetic abnormalities in sec-
ondary leukemia. Cancer 1989; 63:2505–2508.
44. Rowley JD, Olney HJ. International workshop on the relationship of prior therapy
to balanced chromosome aberrations in therapy-related myelodysplastic syndromes
and acute leukemia: Overview report. Genes Chromosomes Cancer 2002; 33:331–
345.
45. Boultwood J, Lewis S, Wainscoat JS. The 5q– syndrome. Blood 1994; 84:3253–
3260.
46. Ebert BL, Pretz J, Bosco J, et al. Identification of RPS14 as a 5q– syndrome gene
by RNA interference screen. Nature 2008; 451:335–339.
47. Paulsson K, Sall T, Fioretos T, et al. The incidence of trisomy 8 as a sole chromoso-
mal aberration in myeloid malignancies varies in relation to gender, age, prior iatro-
genic genotoxic exposure, and morphology. Cancer Genet Cytogenet 2001; 130:160–
165.
48. Paulsson K, Heidenblad M, Strombeck B, et al. High-resolution genome-wide array-
based comparative genome hybridization reveals cryptic chromosome changes in
AML and MDS cases with trisomy 8 as the sole cytogenetic aberration. Leukemia
2006; 20:840–846.
tol 2000; 108:346–356.
50. Kardos G, Baumann I, Passmore SJ, et al. Refractory anemia in childhood: A
retrospective analysis of 67 patients with particular reference to monosomy 7. Blood
2003; 102:1997–2003.
51. Luna-Fineman S, Shannon KM, Lange BJ. Childhood monosomy 7: Epidemiology,
biology, and mechanistic implications. Blood 1995; 85:1985–1999.
52. Emanuel PD. Myelodysplasia and myeloproliferative disorders in childhood: An
update. Br J Haematol 1999; 105:852–863.
53. Martinez-Climent JA, Garcia-Conde J. Chromosomal rearrangements in childhood
acute myeloid leukemia and myelodysplastic syndromes. J Pediatr Hematol Oncol
1999; 21:91–102.
54. Bjork J, Albin M, Mauritzson N, et al. Smoking and myelodysplastic syndromes.
Epidemiology 2000; 11:285–291.
55. Johansson B, Mertens F, Mitelman F. Cytogenetic deletion maps of hematologic
neoplasms: Circumstantial evidence for tumor suppressor loci. Genes Chromosomes
Cancer 1993; 8:205–218.
56. Lai JL, Preudhomme C, Zandecki M, et al. Myelodysplastic syndromes and acute
myeloid leukemia with 17p deletion. An entity characterized by specific dysgranu-
lopoiesis and a high incidence of P53 mutations. Leukemia 1995; 9:370–381.
57. Wang P, Spielberger RT, Thangavelu M, et al. Dic(5; 17): A recurring abnormality
in malignant myeloid disorders associated with mutations of TP53. Genes Chromo-
somes Cancer 1997; 20:282–291.
58. Merlat A, Lai JL, Sterkers Y, et al. Therapy-related myelodysplastic syndrome and
acute myeloid leukemia with 17p deletion. A report on 25 cases. Leukemia 1999;
13:250–257.
59. Thirman MJ, Larson RA. Therapy-related myeloid leukemia. Hematol Oncol Clin
North Am 1996; 10:293–320.
60. Le Beau MM, Albain KS, Larson RA, et al. Clinical and cytogenetic correlations in 63
patients with therapy-related myelodysplastic syndromes and acute nonlymphocytic
leukemia: Further evidence for characteristic abnormalities of chromosomes no. 5
and 7. J Clin Oncol 1986; 4:325–345.
61. Zhang Y, Rowley JD. Chromatin structural elements and chromosomal translocations
in leukemia. DNA Repair (Amst) 2006; 5:1282–1297.
62. Bain BJ, Moorman AV, Johansson B, et al. Myelodysplastic syndromes associated
with 11q23 abnormalities. European 11q23 Workshop participants. Leukemia 1998;
12:834–839.
63. Secker-Walker LM. General report on the european union concerted action workshop
on 11q23, London, U.K.: May 1997. Leukemia 1998; 12:776–778.
64. Bloomfield CD, Archer KJ, Mrozek K, et al. 11q23 balanced chromosome aberra-
tions in treatment-related myelodysplastic syndromes and acute leukemia: Report
from an international workshop. Genes Chromosomes Cancer 2002; 33:362–
378.
65. Rowley JD, Reshmi S, Sobulo O, et al. All patients with the T(11;16)(q23;p13.3)
that involves MLL and CBP have treatment-related hematologic disorders. Blood
1997; 90:535–541.
66. Golub TR, Barker GF, Lovett M, et al. Fusion of PDGF receptor beta to a novel
ets-like gene, tel, in chronic myelomonocytic leukemia with t(5;12) chromosomal
translocation. Cell 1994; 77:307–316.
67. Ross TS, Bernard OA, Berger R, et al. Fusion of Huntingtin interacting protein 1 to
platelet-derived growth factor beta receptor (PDGFbetaR) in chronic myelomono-
cytic leukemia with t(5;7)(q33;q11.2). Blood 1998; 91:4419–4426.
68. Magnusson MK, Meade KE, Brown KE, et al. Rabaptin-5 is a novel fusion partner
to platelet-derived growth factor beta receptor in chronic myelomonocytic leukemia.
Blood 2001; 98:2518–2525.
69. Kulkarni S, Heath C, Parker S, et al. Fusion of H4/D10S170 to the platelet-derived
growth factor receptor beta in BCR-ABL-negative myeloproliferative disorders with
a t(5;10)(q33;q21). Cancer Res 2000; 60:3592–3598.
70. Abe A, Emi N, Tanimoto M, et al. Fusion of the platelet-derived growth factor
receptor beta to a novel gene CEV14 in acute myelogenous leukemia after clonal
evolution. Blood 1997; 90:4271–4277.
71. Rubin CM, Larson RA, Bitter MA, et al. Association of a chromosomal 3;21 translo-
cation with the blast phase of chronic myelogenous leukemia. Blood 1987; 70:1338–
1342.
72. Rubin CM, Larson RA, Anastasi J, et al. t(3;21)(q26;q22): A recurring chromoso-
mal abnormality in therapy-related myelodysplastic syndrome and acute myeloid
leukemia. Blood 1990; 76:2594–2598.
73. Sood R, Talwar-Trikha A, Chakrabarti SR, et al. MDS1/EVI1 enhances TGF-
beta1 signaling and strengthens its growth-inhibitory effect but the leukemia-
associated fusion protein AML1/MDS1/EVI1, product of the t(3;21), abrogates
growth-inhibition in response to TGF-beta1. Leukemia 1999; 13:348–357.
74. Martinelli G, Ottaviani E, Buonamici S, et al. Association of 3q21q26 syndrome
with different RPN1/EVI1 fusion transcripts. Haematologica 2003; 88:1221–1228.
75. Pekarsky Y, Rynditch A, Wieser R, et al. Activation of a novel gene in 3q21 and
identification of intergenic fusion transcripts with ecotropic viral insertion site I in
leukemia. Cancer Res 1997; 57:3914–3919.
76. Block AW, Carroll AJ, 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–
412.
77. Knudson AG Jr. Mutation and cancer: Statistical study of retinoblastoma. Proc Natl
Acad Sci USA 1971; 68:820–823.
78. Venkatachalam S, Shi YP, Jones SN, et al. Retention of wild-type p53 in tumors from
p53 heterozygous mice: reduction of p53 dosage can promote cancer formation.
EMBO J 1998; 17:4657–4667.
79. Rosenbauer F, Wagner K, Kutok JL, et al. Acute myeloid leukemia induced by
graded reduction of a lineage-specific transcription factor, PU.1. Nat Genet 2004;
36:624–630.
80. Song WJ, Sullivan MG, Legare RD, et al. Haploinsufficiency of CBFA2 causes
familial thrombocytopenia with propensity to develop acute myelogenous leukaemia.
Nat Genet 1999; 23:166–175.
81. French JE, Lacks GD, Trempus C, et al. Loss of heterozygosity frequency at the Trp53
locus in p53-deficient (+/−) mouse tumors is carcinogen- and tissue-dependent.
Carcinogenesis 2001; 22:99–106.
82. Fero ML, Rivkin M, Tasch M, et al. A syndrome of multiorgan hyperplasia with
features of gigantism, tumorigenesis, and female sterility in p27(Kip1)-deficient
mice. Cell 1996; 85:733–744.
83. Michaud J, Wu F, Osato M, et al. In vitro analyses of known and novel RUNX1/AML1
mutations in dominant familial platelet disorder with predisposition to acute myel-
ogenous leukemia: Implications for mechanisms of pathogenesis. Blood 2002;
99:1364–1372.
84. Nakao M, Horiike S, Fukushima-Nakase Y, et al. Novel loss-of-function muta-
tions of the haematopoiesis-related transcription factor, acute myeloid leukaemia
1/runt-related transcription factor 1, detected in acute myeloblastic leukaemia and
85. Lai F, Godley LA, Joslin J, et al. Transcript map and comparative analysis of the
1.5-Mb commonly deleted segment of human 5q31 in malignant myeloid diseases
with a del(5q). Genomics 2001; 71:235–245.
86. Fairman J, Chumakov I, Chinault AC, et al. Physical mapping of the minimal region
of loss in 5q− chromosome. Proc Natl Acad Sci U S A 1995; 92:7406–7410.
87. Horrigan SK, Arbieva ZH, Xie HY, et al. Delineation of a minimal interval and
identification of 9 candidates for a tumor suppressor gene in malignant myeloid
disorders on 5q31. Blood 2000; 95:2372–2377.
88. Zhao N, Stoffel A, Wang PW, et al. Molecular delineation of the smallest commonly
deleted region of chromosome 5 in malignant myeloid diseases to 1–1.5 Mb and
preparation of a PAC-based physical map. Proc Natl Acad Sci U S A 1997; 94:6948–
6953.
89. Joslin JM, Fernald AA, Tennant TR, et al. Haploinsufficiency of EGR1, a candidate
gene in the del(5q), leads to the development of myeloid disorders. Blood 2007;
110:719–726.
90. Liu TX, Becker MW, Jelinek J, et al. Chromosome 5q deletion and epigenetic
suppression of the gene encoding alpha-catenin (CTNNA1) in myeloid cell trans-
formation. Nat Med 2007; 13:78–83.
91. Boultwood J, Fidler C, Strickson AJ, et al. Narrowing and genomic annotation
of the commonly deleted region of the 5q− syndrome. Blood 2002; 99:4638–
4641.
92. Le Beau MM, Espinosa R III, Davis EM, et al. Cytogenetic and molecular delineation
of a region of chromosome 7 commonly deleted in malignant myeloid diseases.
Blood 1996; 88:1930–1935.
93. Fischer K, Frohling S, Scherer SW, et al. Molecular cytogenetic delineation of

deletions and translocations involving chromosome band 7q22 in myeloid leukemias.
Blood 1997; 89:2036–2041.
94. Tosi S, Scherer SW, Giudici G, et al. Delineation of multiple deleted regions in 7q
in myeloid disorders. Genes Chromosomes Cancer 1999; 25:384–392.
95. Curtiss NP, Bonifas JM, Lauchle JO, et al. Isolation and analysis of candidate myeloid
tumor suppressor genes from a commonly deleted segment of 7q22. Genomics 2005;
85:600–607.
96. Döhner K, Habdank M, Rucker FG, et al. Molecular characterization of distinct hot
spot regins on chromosome 7q in myeloid leukemias. Blood 2006; 108; Abstract
2349.
97. Bench AJ, Nacheva EP, Hood TL, et al. Chromosome 20 deletions in myeloid malig-
nancies: Reduction of the common deleted region, generation of a PAC/BAC contig
and identification of candidate genes. UK Cancer Cytogenetics Group (UKCCG).
Oncogene 2000; 19:3902–3913.
98. Wang PW, Eisenbart JD, Espinosa R III, et al. Refinement of the smallest commonly
deleted segment of chromosome 20 in malignant myeloid diseases and development
of a PAC-based physical and transcription map. Genomics 2000; 67:28–39.
99. MacGrogan D, Kalakonda N, Alvarez S, et al. Structural integrity and expression
of the L3MBTL gene in normal and malignant hematopoietic cells. Genes Chromo-
somes Cancer 2004; 41:203–213.
100. Neubauer A, Greenberg P, Negrin R, et al. Mutations in the ras proto-oncogenes in
patients with myelodysplastic syndromes. Leukemia 1994; 8:638–641.
101. Beaupre DM, Kurzrock R. RAS and leukemia: From basic mechanisms to gene-
directed therapy. J Clin Oncol 1999; 17:1071–1079.
102. Padua RA, Guinn BA, Al-Sabah AI, et al. RAS, FMS and p53 mutations and poor
clinical outcome in myelodysplasias: A 10-year follow-up. Leukemia 1998; 12:887–
892.
103. de Souza Fernandez T, Menezes de Souza J, Macedo Silva ML, et al. Correla-
tion of N-ras point mutations with specific chromosomal abnormalities in primary
myelodysplastic syndrome. Leuk Res 1998; 22:125–134.
104. Tien HF, Wang CH, Chuang SM, et al. Cytogenetic studies, ras mutation, and clinical
characteristics in primary myelodysplastic syndrome. A study on 68 Chinese patients
in Taiwan. Cancer Genet Cytogenet 1994; 74:40–49.
105. Bacher U, Haferlach T, Kern W, et al. A comparative study of molecular mutations in
381 patients with myelodysplastic syndrome and in 4130 patients with acute myeloid
leukemia. Haematologica 2007; 92:744–752.
106. Kurzrock R, Kantarjian HM, Cortes JE, et al. Farnesyltransferase inhibitor R115777
in myelodysplastic syndrome: Clinical and biologic activities in the phase 1 setting.
Blood 2003; 102:4527–4534.
107. Chen CY, Lin LI, Tang JL, et al. RUNX1 gene mutation in primary myelodysplastic
syndrome—The mutation can be detected early at diagnosis or acquired during
disease progression and is associated with poor outcome. Br J Haematol 2007;
139:405–414.
108. Niimi H, Harada H, Harada Y, et al. Hyperactivation of the RAS signaling pathway
in myelodysplastic syndrome with AML1/RUNX1 point mutations. Leukemia 2006;
20:635–644.
109. Falini B, Mecucci C, Tiacci E, et al. Cytoplasmic nucleophosmin in acute myeloge-

nous leukemia with a normal karyotype. N Engl J Med 2005; 352:254–266.
110. Pedersen-Bjergaard J, Andersen MT, Andersen MK. Genetic pathways in the patho-
genesis of therapy-related myelodysplasia and acute myeloid leukemia. Hematology
Am Soc Hematol Educ Program 2007; 2007:392–397.
111. Shiseki M, Kitagawa Y, Wang YH, et al. Lack of nucleophosmin mutation in patients
with myelodysplastic syndrome and acute myeloid leukemia with chromosome 5
abnormalities. Leuk Lymphoma 2007; 48:2141–2144.
112. Leroy H, Roumier C, Huyghe P, et al. CEBPA point mutations in hematological
malignancies. Leukemia 2005; 19:329–334.
113. Shih LY, Huang CF, Lin TL, et al. Heterogeneous patterns of CEBPalpha mutation
status in the progression of myelodysplastic syndrome and chronic myelomonocytic
leukemia to acute myelogenous leukemia. Clin Cancer Res 2005; 11:1821–1826.
114. Christiansen DH, Andersen MK, Pedersen-Bjergaard J. Mutations with loss of het-
erozygosity of p53 are common in therapy-related myelodysplasia and acute myeloid
leukemia after exposure to alkylating agents and significantly associated with dele-
tion or loss of 5q, a complex karyotype, and a poor prognosis. J Clin Oncol 2001;
19:1405–1413.
115. Kita-Sasai Y, Horiike S, Misawa S, et al. International prognostic scoring system and
TP53 mutations are independent prognostic indicators for patients with myelodys-
plastic syndrome. Br J Haematol 2001; 115:309–312.
116. Steensma DP, Dewald GW, Lasho TL, et al. The JAK2 V617 F activating tyrosine
and myelodysplastic syndromes. Blood 2005; 106:1207–1209.
117. Zipperer E, Wulfert M, Germing U, et al. MPL 515 and JAK2 mutation analysis in
MDS presenting with a platelet count of more than 500 × 10(9)/l. Ann Hematol
2008; 87:413–415.
118. Christiansen DH, Andersen MK, Pedersen-Bjergaard J. Methylation of p15INK4B
is common, is associated with deletion of genes on chromosome arm 7q and predicts
a poor prognosis in therapy-related myelodysplasia and acute myeloid leukemia.
Leukemia 2003; 17:1813–1819.
119. Pedersen-Bjergaard J, Andersen MK, Christiansen DH, et al. Genetic pathways in
therapy-related myelodysplasia and acute myeloid leukemia. Blood 2002; 99:1909–
1912.
120. Kelly LM, Gilliland DG. Genetics of myeloid leukemias. Annu Rev Genomics Hum
Genet 2002; 3:179–198.
121. Loh ML, Vattikuti S, Schubbert S, et al. Mutations in PTPN11 implicate the SHP-2
phosphatase in leukemogenesis. Blood 2004; 103:2325–2331.
122. Side LE, Curtiss NP, Teel K, et al. RAS, FLT3, and TP53 mutations in therapy-
related myeloid malignancies with abnormalities of chromosomes 5 and 7. Genes
Chromosomes Cancer 2004; 39:217–223.
123. Qian Z, Fernald AA, Godley LA, et al. Expression profiling of CD34+ hematopoi-
etic stem/progenitor cells reveals distinct subtypes of therapy-related acute myeloid
leukemia. Proc Natl Acad Sci U S A 2002; 99:14925–14930.
124. Pellagatti A, Cazzola M, Giagounidis AA, et al. Gene expression profiles of CD34+
cells in myelodysplastic syndromes: Involvement of interferon-stimulated genes and
correlation to FAB subtype and karyotype. Blood 2006; 108:337–345.
125. Hofmann WK, de Vos S, Komor M, et al. Characterization of gene expression

of CD34+ cells from normal and myelodysplastic bone marrow. Blood 2002;
100:3553–3560.
126. Boultwood J, Pellagatti A, Cattan H, et al. Gene expression profiling of CD34+ cells
in patients with the 5q− syndrome. Br J Haematol 2007; 139:578–589.
127. Nilsson L, Eden P, Olsson E, et al. The molecular signature of MDS stem cells
supports a stem-cell origin of 5q myelodysplastic syndromes. Blood 2007; 110:3005–
3014.
128. Parker JE, Mufti GJ, Rasool F, et al. The role of apoptosis, proliferation, and the Bcl-
2-related proteins in the myelodysplastic syndromes and acute myeloid leukemia
secondary to MDS. Blood 2000; 96:3932–3938.
129. Lepelley P, Soenen V, Preudhomme C, et al. Bcl-2 expression in myelodysplas-
tic syndromes and its correlation with hematological features, p53 mutations and
prognosis. Leukemia 1995; 9:726–730.
130. Ridge SA, Worwood M, Oscier D, et al. FMS mutations in myelodysplas-
tic, leukemic, and normal subjects. Proc Natl Acad Sci USA 1990; 87:1377–
1380.
131. Kiyoi H, Towatari M, Yokota S, et al. Internal tandem duplication of the FLT3 gene
is a novel modality of elongation mutation which causes constitutive activation of
the product. Leukemia 1998; 12:1333–1337.
132. Horiike S, Yokota S, Nakao M, et al. Tandem duplications of the FLT3 receptor gene
are associated with leukemic transformation of myelodysplasia. Leukemia 1997;
11:1442–1446.
133. Tidow N, Kasper B, Welte K. Clinical implications of G-CSF receptor mutations.
Crit Rev Oncol Hematol 1998; 28:1–6.
134. Arland M, Fiedler W, Samalecos A, et al. Absence of point mutations in a functionally
important part of the extracellular domain of the c-kit proto-oncogene in a series of
patients with acute myeloid leukemia (AML). Leukemia 1994; 8:498–501.
135. Siitonen T, Savolainen ER, Koistinen P. Expression of the c-kit proto-oncogene
in myeloproliferative disorders and myelodysplastic syndromes. Leukemia 1994;
8:631–637.
136. Zochbauer S, Gsur A, Gotzl M, et al. MDR1 gene expression in myelodysplastic
syndrome and in acute myeloid leukemia evolving from myelodysplastic syndrome.
Anticancer Res 1994; 14:1293–1295.
137. Faderl S, Kantarjian HM, Estey E, et al. The prognostic significance of p16(INK4
a)/p14(ARF) locus deletion and MDM-2 protein expression in adult acute myeloge-
nous leukemia. Cancer 2000; 89:1976–1982.
138. Bueso-Ramos CE, Manshouri T, Haidar MA, et al. Multiple patterns of MDM-2
deregulation in human leukemias: Implications in leukemogenesis and prognosis.
Leuk Lymphoma 1995; 17:13–18.
139. Ogata K, Tamura H. Thrombopoietin and myelodysplastic syndromes. Int J Hematol
2000; 72:173–177.
140. Bouscary D, Preudhomme C, Ribrag V, et al. Prognostic value of c-mpl expression
in myelodysplastic syndromes. Leukemia 1995; 9:783–788.
141. Shannon KM, O’Connell P, Martin GA, et al. Loss of the normal NF1 allele from
the bone marrow of children with type 1 neurofibromatosis and malignant myeloid
disorders. N Engl J Med 1994; 330:597–601.
142. Gallagher A, Darley RL, Padua R. The molecular basis of myelodysplastic syn-
dromes. Haematologica 1997; 82:191–204.
143. Xu D, Gruber A, Peterson C, et al. Telomerase activity and the expression of telom-
erase components in acute myelogenous leukaemia. Br J Haematol 1998; 102:1367–
1375.
144. Li B, Yang J, Andrews C, et al. Telomerase activity in preleukemia and acute
myelogenous leukemia. Leuk Lymphoma 2000; 36:579–587.
145. Misawa S, Horiike S. TP53 mutations in myelodysplastic syndrome. Leuk Lym-
phoma 1996; 23:417–422.
146. Cilloni D, Saglio G. WT1 as a universal marker for minimal residual disease detection
and quantification in myeloid leukemias and in myelodysplastic syndrome. Acta
Haematol 2004; 112:79–84.
4
Genomic Approaches in MDS
Andrea Pellagatti
Leukaemia Research Fund Molecular Haematology Unit, Nuffield Department of
Clinical Laboratory Sciences, John Radcliffe Hospital, Oxford, U.K.
INTRODUCTION
The advent of microarray technology in the last decade has revolutionized the
investigation of human disease. While previously available technologies allowed
to study only one or a few genes at the same time, microarrays are capable of
interrogating thousands of genes simultaneously. There are several different types
of microarray platforms that vary in the type of probes immobilized on the array
and the kind of information that can be derived. For example, gene expression
microarrays are designed to measure changes in gene expression at the RNA
level, while comparative genomic hybridization (CGH) and single nucleotide
polymorphism (SNP) microarrays are used to look for genomic gains and losses
and changes in gene copy number at the DNA level.
GENE EXPRESSION MICROARRAYS

In gene expression microarrays, the detection of mRNA transcript abundance
relies on the specific hybridization of labeled nucleic acids derived from the
starting RNA to known gene sequences immobilized on glass slides or other
suitable surfaces. There are two prevalent platforms for gene expression microar-
ray analysis: spotted arrays, in which specific cDNA species or oligonucleotides
are spotted onto glass microscope slides, and synthetic oligonucleotide arrays
(e.g., Affymetrix GeneChipTM technology), in which short oligonucleotides are
synthesized in situ on the glass surface using a process known as photolithography.
The most basic data analysis is usually aimed at identifying genes differentially
87
88 Pellagatti
expressed between diseased and normal samples, or between disease subgroups.

Hierarchical clustering is often used to visualize differences in gene expression
patterns—the resulting tree (or “dendrogram”) groups together those samples
having similar expression levels across a set of selected genes.
Early Microarray Studies in MDS

The first landmark study in leukemia using gene expression microarrays dates back
to 1999, when Todd Golub in Boston and his colleagues investigated the differences
in the expression profiles of samples from 11 patients with acute myeloid leukemia
(AML) and those of samples from 27 patients with acute lymphoblastic leukemia
(ALL) (1). This study was able to identify a class predictor of 50 genes that could
distinguish AML cases from ALL cases, and the accuracy of this predictor was
validated using an independent sample set.
It was not until 2 years later that the first study on the use of microarray
technology in myelodysplastic syndromes (MDS) was published. Akira Miyazato
and his colleagues in Japan investigated the differences in the transcriptome of
AC133+ hematopoietic stem cells (HSC) in five MDS patients (three patients with
MDS-associated leukemia and two patients with RAEB) and five patients with
de novo AML (2), using a microarray platform containing 2304 genes. Several
MDS-specific and AML-specific genes were identified, with particular emphasis
on DLK (official gene symbol: DLK1) as a newly discovered candidate MDS-
specific marker. DLK belongs to the superfamily of epidermal growth factor–like
proteins and plays a direct role in stromal cell–supported HSC proliferation (3).
DLK upregulation in HSC might play a role in the increased proliferation of
these cells in MDS. DLK overexpression in MDS was further investigated using
real-time quantitative PCR in 22 MDS, 31 AML, and 8 CML samples. Expression
levels of DLK were increased by ⬎2-fold in 12 of 22 (55%) MDS patients and 3 of
31 (10%) AML patients, while no DLK expression was detected in CML patients.
DLK represents therefore the first candidate MDS-specific gene identified using
expression microarray technology that could be important for the diagnosis and
molecular pathogenesis of MDS (2).
A second paper using gene expression microarrays for the study of MDS
appeared 1 year later (2002). Wolf-Karsten Hofmann in Berlin and his colleagues
studied the expression profiles of the CD34+ cells isolated from 11 MDS patients
(seven IPSS lower risk and four higher risk) and four healthy controls, using a
microarray platform containing 9670 human genes (Affymetrix HG-U95Av2) (4).
With a cut-off of 5-fold for changes in gene expression, it was shown that 278
genes were differentially expressed between lower risk MDS and healthy controls,
125 genes between higher risk MDS and healthy controls, and 344 between higher
risk and lower risk MDS. Genes downregulated in lower risk MDS included RAI3,
IEX1, and STIP1—all encoding proteins that play a role in the protection of cells
from stress caused by mutagens, radiation, or other factors. Moreover, using class
membership prediction, a set of 11 genes was identified that could discriminate
Genomic Approaches in MDS 89
with high accuracy between lower risk MDS patients, higher risk MDS patients,
and healthy controls (4). Interestingly, this study also reported upregulation of the
DLK1 gene, which was shown to be overexpressed in lower risk MDS patients
compared with healthy controls.
Another study, by Masuzu Ueda and colleagues in Japan, compared the
gene expression profiles of the AC133+ HSC populations isolated from 11 MDS
patients with RA (refractory anemia), 5 with RAEB (refractory anemia with
excess of blasts), 14 with MDS-associated leukemia, and 2 healthy volunteers
(5). For this purpose, the same platform (containing 2304 genes) was used as
in the first microarray study in MDS by Miyazato et al. Differences in gene
expression were identified between samples from individuals with good prognosis
(healthy controls and RA) and those of individuals with poor prognosis (RAEB
and MDS-associated leukemia). Eleven genes, including NDUFV1, LH2, and
PNMA2, were expressed at higher levels in the poor prognosis group, while genes
with higher expression in the good prognosis group included PIASy and PIASx-␤.
The expression levels of PIASy were further investigated by real-time quantitative
PCR in the AC133+ cells of 13 MDS patients with RA, 9 with RAEB, 13 with
MDS-associated leukemia, and 2 healthy controls. In concordance with the array
results, PIASy expression levels were higher in healthy controls and patients with
RA than in patients with RAEB or MDS-associated leukemia (5). None of the
prognostically important genes identified in this study appeared in the previously
described study by Hofmann et al. (4). However, it has to be pointed out that
these two studies differed in the array platform used, the HSC population chosen
(AC133+ vs. CD34+), and the patient classification adopted (FAB–WHO vs.
IPSS); all these factors, together with the small number of patients analyzed, are
likely to have contributed to the lack of concordance between these datasets.
A Larger Microarray Study in CD34+ Cells

To reduce such variability, studies involving a larger number of patients and more
complete microarray platforms are clearly needed. The first step in this direction
was represented by a study by Andrea Pellagatti in Oxford and his colleagues
on the CD34+ cells of a group of 55 MDS patients and 11 healthy controls (6).
A comprehensive array (Affymetrix HG U133 Plus 2.0) was used, covering over
47,000 transcripts representing 39,000 human genes. The expression profiles of
MDS CD34+ cells shared many similarities to reported interferon-␥ –induced
gene expression in normal CD34+ cells (7). Several interferon-stimulated genes
(ISG) were upregulated in MDS CD34+ cells; indeed the two most upregulated
genes, IFIT1 and IFITM1, are both ISG. It is possible that this upregulation of ISG
may be responsible for some of the hematological characteristics of MDS, such
as peripheral blood cytopenias.
In agreement with the previous studies by Miyazato et al. and Hofmann
et al. (2,4), this study also found DLK1 to be upregulated in a high proportion of
MDS patients (60%). Several genes were downregulated by more than 2-fold in
90 Pellagatti
the majority of the MDS patients studied, including Gravin/AKAP12, ARPP21,

ELA2, CD24, MME and EBF (6). Several of the downregulated genes found in
this study were in common with an earlier microarray study by Alex Sternberg
and colleagues, which had reported reduced expression of genes involved in B-
lymphocyte development in patients with RA and a normal karyotype (8). Both
studies agree on the lower expression levels of nine B-lymphocyte development
genes (CD24, POU2AF1, VPREB1, VPREB3, CD79B, DNTT, PAX5, LEF1 and
BACH1) in MDS patients with RA.
Another goal of this study was to determine whether distinct gene expres-
sion profiles were associated with specific FAB subtypes. Hierarchical clustering
was performed using the 457 significantly differentially expressed genes between
patients with RA, patients with RARS (refractory anemia with ringed sideroblasts),
and patients with RAEB. Most of the MDS patients with RARS were grouped
into a distinct cluster, while patients with RA and patients with RAEB showed
considerable overlap (Fig. 1 see colour insert). The gene expression profiles of the
CD34+ cells from patients with RARS were characterized by the upregulation of
mitochondrial-related genes, in particular of those involved in heme synthesis (e.g.,
ALAS2). The fact that RARS has a distinct gene expression profile suggests the
presence of a common underlying pathophysiological basis to this disease (6).
This study by Pellagatti et al. also identified several genes differentially
expressed between early MDS (subtype RA) and advanced MDS (subtype RAEB-
2). The most significant genes included CASP3, PRKAR2B, LEF1 and EMS1 (all
with higher expression levels in patients with RA than RAEB-2), and FLT3 and
DHRS8 (with lower expression levels in patients with RA than RAEB-2). These
genes differentially expressed between early and advanced MDS may have value
as prognostic markers or markers of disease progression (6).
Effect of Chromosomal Abnormalities on Gene Expression

In the study by Pellagatti et al., the relationship between karyotype and gene
expression was also investigated. In particular, the investigators analyzed dif-
ferences in gene expression between MDS patients with a deletion on the long
arm of chromosome 5 [del(5q)], patients with RA and a normal karyotype, and
patients with other karyotypes not including del(5q). A total of 889 differentially
expressed genes were identified that resulted in patients with a del(5q) clustering
as a distinct family, while no separation was observed between cases with normal
karoytype and other karyotypes (Fig. 2 see colour insert). Approximately 40% of
the probe sets showing lower expression levels in patients with a del(5q) mapped
to chromosome 5q, demonstrating that the presence of a del(5q) results in a gene
dosage effect. Among upregulated genes in patients with a del(5q), several were
representatives of the histone HIST1 gene cluster at chromosome 6p21, and others
were genes related to the actin cytoskeleton (6).
Another karyotype-related gene expression microarray study investigated
MDS patients with monosomy 7 and trisomy 8, which are among the most
Down Up
RA RARS RAEB
Figure 1 Hierarchical clustering (or dendrogram) of 457 significantly differentially

expressed genes in the CD34+ cells of patients with RA, patients with RARS, and patients
with RAEB. In this type of graph, the MDS samples are clustered according to their similar-
ity in expression levels of the selected genes. Similarly, the genes are clustered according to
their profile across the patients studied. Each row represents a single gene and each column
a separate MDS patient sample. Source: Adapted from Ref. 6. (see color insert)
frequently observed cytogenetic abnormalities in MDS (9,10). MDS patients with

monosomy 7 and trisomy 8 have distinct clinical courses, responses to therapy,
and survival probabilities (11–14). In an attempt to determine if the presence
of these karyotypic abnormalities results in specific changes in gene expression,
Guibin Chen at the National Institutes of Health in Bethesda and his colleagues
compared the expression profiles of CD34+ cells obtained from 2 MDS patients
with monosomy 7, 11 MDS patients with trisomy 8, and 8 healthy controls, using
the Affymetrix U95Av2 platform (9760 genes) (15). The results showed that 177
92 Pellagatti
Down Up
Del(5q) Normal karyotype Other karyotype
Figure 2 Hierarchical clustering of 889 significantly differentially expressed genes

between MDS patients with a del(5q), patients with RA and a normal karyotype, and
patients with other karyotypes. Each row represents a single gene and each column a
separate MDS patient sample. Source: Adapted from Ref. 6. (see color insert)
genes were deregulated in patients with monosomy 7, more than in patients with
trisomy 8 (115 deregulated genes), as compared with healthy controls. In partic-
ular, the CD34+ cells of both groups of patients with monosomy 7 and trisomy
8 showed a deregulation of genes involved in HSC proliferation (ZNF261, B94)
and blood cell function (PF4, CD41b, DEFA4, CCR2, CD14). In patients with
monosomy 7, there was a peculiar upregulation of genes inducing leukemia trans-
formation and tumorigenesis (HOX9 A, BRCA2, PRAME, BMI-1 and PLAB), and
downregulation of genes involved in the control of cell growth and differentiation
(PDGF, inhibin alpha and beta). The CD34+ cells of patients with trisomy 8
were instead characterized by upregulation of genes implicated in immune and
inflammatory response (TGF-␤, TGF-␤ receptor, IL-10, IL-7R, VCAM-1, and

c-Maf ), and downregulation of anti-apoptotic genes (BIRC5, NAIP) (15).
These karyotype-related studies on del(5q), monosomy 7, and trisomy 8
in MDS clearly demonstrate that stratification of MDS patients by karyotype is
required for accurate interpretation of gene expression data. Indeed, the strong
influence of karyotype on gene expression was also demonstrated in a study of
AML blast cells, where it was shown that in patients with trisomy 8, the median
expression of genes located on chromosome 8 was higher, whereas in those with
AML and monosomy 7 or del(5q), the median expression of genes located in the
deleted regions was lower (16).
Studies in 5q− Syndrome

When del(5q) occurs as the sole karyotypic abnormality in de novo MDS patients
who have less than 5% blasts in the bone marrow, patients are classified as having
“5q− syndrome” (17,18). The 5q− syndrome is one of the most distinct forms
of MDS, and microarray studies investigating this particular MDS subtype have
shown interesting results in relation to the deregulated genes identified and to the
gene dosage effect caused by the presence of the del(5q).
A study by Lars Nilsson and colleagues investigated the gene expression
profiles of highly purified CD34+CD38−Thy1+ cells from four patients with
5q− syndrome and from five healthy controls, using an oligonucleotide microar-
ray platform containing ∼27,000 genes (19). Firstly, as a confirmation of the
gene-dosage effect observed in patients with a del(5q), the authors found downreg-
ulation of 34 of 37 genes located at 5q31–q32. Also, in agreement with previous
microarray studies in MDS, upregulation of both IFITM1 and DLK1 was observed.
Other deregulated genes highlighted in this study were BMI1, upregulated in
5q− syndrome cases, and CEBPA, downregulated in 5q− syndrome cases (19).
Another microarray study on 5q− syndrome investigated the transcriptome
of the CD34+ cells of 10 patients with 5q− syndrome, 14 patients with RA and
a normal karyotype, and 16 healthy controls, using the Affymetrix HG U133 Plus
2.0 array platform (39,000 genes) (20). In agreement with the gene dosage effect
reported in the studies mentioned before, the majority of the genes assigned to the
commonly deleted region (CDR) of the 5q− syndrome at 5q31–q32 (21) showed
reduced expression levels in patients with 5q− syndrome, consistent with the loss
of one allele. Two interesting candidate genes showing haploinsufficiency in the
5q− syndrome are SPARC and RPS14 (20). SPARC is a tumor suppressor gene
and RPS14 is a component of the 40 S ribosomal subunit. Haploinsufficiency for
RPS14 has been recently suggested to be the mechanism behind the hematologic
phenotype of the 5q− syndrome (22). Two other genes mapping to the CDR,
RBM22 and CSNK1A1, showed a ⬎50% reduction in expression levels, consistent
with the downregulation of the remaining allele. Differentially expressed genes
mapping to other chromosomes include SPAG6, WIG-1, BMI1 (upregulated in
5q− syndrome) and DPH5 (downregulated in 5q− syndrome) (20). The finding
94 Pellagatti
of BMI1 upregulation is in agreement with the above-mentioned study by Nils-

son et al. (19). This paper addressed the question of deregulated pathways in the
5q− syndrome. Several significantly deregulated gene pathways were identified
in patients with the 5q− syndrome including the Wnt/␤-catenin signaling path-
way and the protein ubiquitination pathway. This study concluded that several
of the genes mapping to the CDR of the 5q− syndrome may play a role in the
pathogenesis of this disorder (20).
Lenalidomide in MDS
Until recently, no effective treatment was available for MDS patients, with most
patients receiving blood transfusion and/or supportive care as the only therapy. This
has changed with the introduction of drugs such as lenalidomide. Lenalidomide
has shown dramatic therapeutic effects in patients with low-risk MDS and del(5q)
(23,24). The molecular targets and mode of action of lenalidomide in MDS are
still unknown, and so far there has been only one published microarray study
addressing this question.
In that microarray study, erythroblast cultures from nine MDS patients with a
del(5q) and from eight healthy controls were used to evaluate the in vitro effects of
lenalidomide on their global gene expression profile (25). Several genes were sig-
nificantly downregulated in response to lenalidomide treatment, including genes
involved in erythropoiesis, such as HBA2, HBB, SPTA1, GYPA, GYPB, ALAS2
and KLF1. Many genes were also significantly upregulated by lenalidomide: in
particular, the four genes SPARC, VSIG4, PPIC and TPBG were upregulated by
more than 2-fold in all MDS del(5q) and all healthy control samples analyzed. The
observed upregulation and increased protein expression of the tumor suppressor
gene SPARC [Figs. 3(A) and 3(B see colour insert)] is of particular interest since its
anti-proliferative, anti-adhesive, and anti-angiogenic properties mirror the known
effects of lenalidomide. Moreover, SPARC is located at 5q31–q32, within the CDR
in MDS 5q− syndrome, and was the only representative of the 44 genes located
within this region to be deregulated by lenalidomide [Fig. 3(C)]. These data show
that the upregulation of SPARC may contribute to the potent effects of lenalido-
mide in MDS with del(5q) (25). A recent report found that SPARC may have
biomarker potential in non-Hodgkin’s lymphoma (NHL), as it was upregulated in
NHL-derived cell lines sensitive to lenalidomide, but not in lenalidomide-resistant
cell lines (26). These data show that microarray analysis of gene expression can
help understand the molecular mechanisms of drug action (27–30).
Studies in Mature Cells

Most of the published expression microarray–based studies in MDS have focused
on hematopoietic progenitor cells, such as AC133+ or CD34+ cells. However, a
few studies have also addressed the gene expression profiles of more differentiated
cell types, such as neutrophils (31,32). Peripheral blood neutrophils originate from
A Untreated B
Lenaliomide Untreated Lenalidomide
2500
2000
1500
1000
500
0
MDS Healthy control
Expression level of SPARC

C
4.5
Genomic Approaches in MDS
4
3.5
3
2.5
2
1.5
1
0.5
Expresssion level
0
B
L
00
U6
E
S
XA
O 0
M
O
A
N
L
16
N
F3
AR 6
P
RP22
C A1
DC A
CD69
CD 74
NS 4
G3 03
G A
M 1
GMBP
T D6
C 4
TIG F1
AT SI
C 2
C 2
L
NK R
RP L7
AN IM3
C X1
ND Z3
J4 9
PD T1
TC PO
SP 6A7
DC 1A1
RBC1B
M T15
A
RGFRB
SY RC
AB
A
ZN IP1
S 36
MS 326
SH S14
Q8 FAT
GR PX3
CC K2
Q9 TN
C OX
A
P E
SL 3TC
SL 26A
FL TED2
KI IL17 2
CS SF1
MG A019
PP DG 6A
Figure 3 Lenalidomide treatment increases SPARC expression in cultured erythroblasts. (A) SPARC gene
expression levels in erythroblast cultures from nine MDS patients with a del(5q) and from eight healthy controls.
The expression profile was evaluated in lenalidomide-treated and untreated cells from each patient/control after
7 days of culture. (B) SPARC immunofluorescent staining in lenalidomide-treated and untreated cells from one
MDS del(5q) patient. Staining was performed on cytocentrifuged cells from day 7 of culture, corresponding to
the cells analyzed with gene expression profiling. Nuclei are stained with DAPI. (C) Effects of lenalidomide
treatment on the expression levels of the genes mapping to the CDR of the 5q− syndrome (41 of 44 are
represented on the Affymetrix arrays used) in the erythroblasts from MDS del(5q) patients. The graph shows
95
the average fold change for each gene after lenalidomide treatment, with the SPARC gene highlighted in red.
Source: Adapted from Ref. 25. (see color insert)
96 Pellagatti
the clonal expansion of the malignant MDS clone, and are readily accessible, as
they can be easily separated from blood samples instead of the more invasive bone
marrow aspirations.
In one such study, cDNA microarrays containing 6000 genes were used
to determine the expression profiles of the neutrophils populations of 22 MDS
patients and 2 AML patients, as compared with those of the neutrophils of 7
healthy controls (31). The disease heterogeneity commonly observed in MDS was
reflected in a high level of heterogeneity in gene expression patterns between MDS
patients. Nevertheless, several genes were found to be commonly up or downreg-
ulated. The most upregulated genes include RAB20, ZNF183, ARG1 and ACPL,
while the most downregulated genes include COX2, CD18, KIAA0001, FOS, IL-7
receptor, ACT2 and IFI56. Statistical analysis showed that 71 genes were sig-
nificantly differentially expressed between patients with the 5q− syndrome and
patients with RA and a normal karyotype, with the resulting hierarchical cluster-
ing separating the two groups into distinct families; this is in agreement with the
previous studies in CD34+ cells showing a distinct expression profile in patients
with 5q− syndrome. The expression levels of 27 genes were significantly different
between MDS subtypes RA, RARS and RAEB, and AML. Hierarchical clustering
performed using these 27 genes grouped together the MDS patients with RAEB
and the patients with AML. Clear differences in the expression pattern of the
neutrophils were observed in paired samples from one patient obtained before and
after leukemic transformation. The results of this study support the possibility that
disease progression in MDS may be studied without resorting to bone marrow
examination (31).
Another gene expression microarray–based study of neutrophils investigated
the rare condition of ␣-thalassemia arising in the context of MDS (ATMDS). In
ATMDS, ␣-thalassemia occurs as an acquired abnormality in individuals with
myelodysplasia (33–35). In patients with ATMDS, there is a downregulation of
the ␣-globin genes, but prior to this microarray study no structural abnormalities
in cis to the ␣-globin genes had been detected, suggesting an association with
a trans-acting mutation. The expression profiles of purified neutrophils from the
peripheral blood of a newly diagnosed ATMDS individual were compared with
those of pooled neutrophils from seven healthy controls (32). Of all the 6000
genes present on the cDNA microarray used, ATRX was the gene with the lowest
expression ratio [Figs. 4(A see colour insert) to 4(C)]. Germline mutations in the
ATRX gene are present in the ATR-X syndrome, a rare and severe form of X-linked
mental retardation, which is also characterized by a mild form of ␣-thalassemia
(36). This observation made ATRX a very good candidate gene for ATMDS.
The ATMDS microarray result was confirmed using real-time quantitative
PCR, and indeed ATRX gene expression levels in the ATMDS patient were only
3% to 4% of those in healthy controls, while no reduction was found in 13 MDS
patients without ␣-thalassemia [Fig 4(D)]. Sequence analysis of the ATRX gene
in the ATMDS patient tested highlighted a mutation in a canonical splice donor
site, possibly responsible for nonsense-mediated decay of the mRNA. Since the
B D
160%
140%
120%
100%
3.00
2.80
2.60
2.40
2.20
2.00
1.80
1.60
1.40
1.20
1.00
0.80
0.60
0.40
0.20
0.00
Gene expression ratio

C 80%
60%
40%
20%
0%
Control MDS ATMDS
0.50
0.40
0.30
0.20
0.10
0.00
Figure 4 Gene expression profiling in an individual with ATMDS. (A) Portion of the
microarray image containing one spot corresponding to the ATRX gene. Yellow spots rep-
resent genes with normal expression levels, while red and green spots represent genes
downregulated and upregulated, respectively, in the individual with ATMDS. (see color
insert) (B) Distribution of the gene expression ratios between neutrophils from the individ-
ual with ATMDS and neutrophils from a mixed pool of healthy controls. All genes on the
cDNA microarray are shown and values in the range 0–0.5 are boxed. (C) Magnification of
the distribution plot in the range 0–0.5, with the three probes representing the ATRX gene
highlighted in red. (D) Real-time quantitative PCR data on ATRX gene expression in neu-
trophils from 7 normal controls, 13 individuals with MDS, and the individual with ATMDS.
Source: Adapted from Ref. 32.
98 Pellagatti
publication of this study, DNA sequencing of the ATRX gene in archival material
from several ATMDS patients identified other point mutations and/or splicing
abnormalities in the ATRX gene (37). This is the first example of the discovery of
a gene mutation in human cancer from the application of microarray technology
(32). This methodology therefore not only can identify changes in gene expression,
but may also pin-point genes that harbor mutations affecting the mRNA integrity.
CGH AND SNP MICROARRAYS

CGH and SNP microarrays are technologies that are increasingly widely used to
detect genomic gains and losses or gene copy number changes in human disease.
In microarray CGH, large fragments of genomic DNA or oligonucleotides are
spotted on the glass surface of the array, and each spot on the array corresponds to
a known chromosomal location. Genomic DNA obtained from a diseased sample
and from a cytogenetically normal reference sample are labeled with different
fluorophores and then hybridized simultaneously to the CGH array. The ratio of
the fluorescent signals from both labeled samples for each spot then indicates if a
loss or gain for that particular gene sequence has occurred.
SNP microarrays are designed to target thousands of single nucleotide
polymorphisms, small genetic changes or variations that occur in the DNA.
With this type of platform, only one genomic DNA sample is required for each
hybridization, and the array provides combined information on the presence
of chromosomal copy number changes and regions of loss of heterozygosity
(LOH) (Fig. 5). It is also possible to determine whether the LOH is associated
with chromosomal loss/deletion or with uniparental disomy (UPD) (Fig. 5).
Uniparental disomy indicates the presence of two copies of a chromosome pair
from one parent and no copy from the other parent, and can result from incomplete
chromosome segregation or from mitotic recombination. The detection of DNA
regions affected by UPD is a prerogative of SNP microarrays, as both conventional
metaphase cytogenetics and CGH microarrays are unable to measure UPD.
Early Studies with CGH Arrays in MDS

Compared with expression microarray–based studies, the use of CGH and SNP
microarrays in MDS is more recent. The first report using CGH microarrays
appeared in 2003; however, that study was very limited as it included only three
MDS cases, and the platform used contained only 59 DNA regions spanning
selected oncogenes (38). A second study was published 2 years later in which 37
MDS patients with complex karyotypes were investigated using nonmicroarray
high-resolution CGH. An array-based CGH platform was subsequently used in
this study, but it was limited to three samples and represented subtelomeric regions
only (39).
Clearly, these pioneering studies, while they demonstrated the feasibility of
the new technology, were limited by the small number of samples analyzed and
A B
Chr 5 Chr 5
CN plots
Heterozygous
SNPs
LOH region
C D
Chr 11 Chr 13
UPD region
Figure 5 Examples of graphs generated from SNP microarray analysis. In each graph,
the red dots indicate the signal intensity for each SNP and the blue lines represent an
averaged value. The green bars represent the position of heterozygous SNPs, and blue bars
represent the areas affected by LOH. If the LOH region is not associated with copy number
changes, then a pink bar shows that the LOH is the result of UPD. (A) Example of a normal
chromosome 5. No alterations are seen in the copy number (CN) plots, and heterozygous
SNPs are evenly distributed. (B) Example of a deletion on chromosome 5q. There is a SNP
signal intensity reduction in the CN plots and a lack of heterozygous SNPs that indicates
LOH. (C) Example of UPD on chromosome 11. The presence of an LOH region which
is not associated with copy number changes indicates an area of UPD. (D) Example of
a gain on chromosome 13. There is an increase of SNP signal intensity in the CN plots.
Source: Courtesy of L. Wang, LRF Molecular Haematology Unit, Oxford, U.K.
100 Pellagatti
of restricted DNA regions contained on the array platforms that were used. The
first publication including MDS samples that used a more comprehensive array
CGH platform appeared in 2006—microarray slides containing over 32,000 BAC
clones and covering at least 98% of the human genome were used to test four MDS
and six AML samples, all with trisomy 8 as the sole karyotypic abnormality (40).
Microarray CGH was able to confirm the presence of trisomy 8 in all cases and,
in addition, found potential leukemia-associated cryptic chromosome changes in
4 of the 10 cases. Only 1 of these 4 cases was an MDS sample, which had a
karyotypically occult 0.8 Mb loss at 17q24.3. Despite the use of a comprehensive
platform, this study was limited by the small sample size.
More Recent SNP Microarray Studies in MDS

Since these earlier reports, several papers using mostly SNP microarrays have
been published, which clearly show that this technology is being increasingly
successfully adopted for the study of MDS.
In the first part of 2007, Lukasz Gondek and his colleagues at Cleveland
Clinic published a study of 66 patients with MDS and 29 healthy control samples,
using a commercial array platform containing 50,000 SNPs (Affymetrix 50 K
SNP arrays) (41). Importantly, this study included healthy controls and addressed
the pivotal issue of copy number polymorphisms which are present in the normal
population; defects found in MDS patients in DNA regions affected by copy num-
ber polymorphisms in healthy controls were not considered significant. In general,
the SNP microarrays not only confirmed all abnormalities found by metaphase
cytogenetics, but also identified a number of new cryptic copy number changes
and several regions of UPD. While metaphase cytogenetics reported a normal
karyotype for 39% of the patients included in the study, using SNP microarrays
this proportion decreased to 12%, meaning that new lesions were identified in
∼70% of patients with normal karyotype. In addition, extra abnormalities were
found in 83% of patients with defects identified by routine karyotyping. Also, in
80% of the cases with unsuccessful metaphase cytogenetics, SNP technology was
able to identify chromosomal lesions.
Among chromosomes known to be affected in MDS, loss of material within
chromosomes 1, 5, 7, and 11, and gain of chromosome 8 were observed. Copy
number changes previously not reported in MDS included changes on chromosome
1p31.1, add(2)(q37), and del(14)(q11). Uniparental disomy was observed in 33%
of the patients, and mostly affecting regions on chromosomes 7q and 11q. This
study was later updated by increasing the number of MDS patients and healthy
controls to 72 and 33, respectively, using the same platform containing 50,000
SNPs (42). In addition to the previously described findings, which remained
virtually unchanged, the authors reported the gain of regions on chromosome 4p
in 4 of 18 patients with ringed sideroblasts. Also, isolated areas of UPD were found
in 4 of the 33 healthy controls analyzed, affecting small areas on chromosomes
6q, 10q, 11q, and 13q, and ranging in size from 2.35 to 4.69 Mb.
The same group subsequently published another paper containing data on

174 patients (94 MDS, 33 secondary AML, and 47 MDS/MPD) and 76 healthy con-
trols. In that follow-up analysis, a higher resolution platform containing 250,000
SNPs was used (Affymetrix 250 K SNP arrays) (43). Chromosomal defects were
identified by SNP microarrays in 78% of MDS patients, 75% of MDS/MPD
patients, and in 77% of secondary AML patients. These proportions were higher
than those obtained with metaphase cytogenetics: 59%, 37%, and 53%, respec-
tively. Uniparental disomy was found in 20% of MDS patients, 35% of MDS/MPD
patients, and in 23% of secondary AML patients. This study also addressed the
question of the prognostic value of these findings: patients with normal karyotype
according to metaphase cytogenetics were divided into two groups, depending on
whether or not new lesions had been identified by SNP microarrays. Patients with
new lesions had a reduced overall survival.
Another paper on SNP microarray analysis in lower risk MDS was published
by Azim Mohamedali in London and his colleagues (44). These authors studied
119 low-risk MDS patients and 33 control subjects with no history of malignancy,
using the Affymetrix 250 K SNP arrays. Some of the samples were also analyzed
using a lower resolution platform (Affymetrix 50 K SNP arrays) and a higher
resolution platform (Affymetrix 500 K SNP arrays). While additional copy number
changes or regions of LOH ⬎2 Mb were identified by using 250 K SNP arrays
as compared with 50 K SNP arrays, no additional regions were found by using
the more expensive 500 K SNP arrays as compared with the 250 K SNP arrays.
This highlights the important point that 250 K SNP microarrays could represent
the ideal platform for an optimal balance between sensitivity and cost. In this
study, UPD regions ⬎2 Mb were found in 55 of 119 MDS patients; these regions
were distributed across the genome, with the most notable chromosomes with
frequent regions of UPD being chromosomes 1, 2, 3, 4, and 6. In particular, UPD
regions on chromosome 4q were found in 25% of patients with RARS, in 12% of
patients with RCMD and normal karyotype, in 17% of patients with RAEB, and
in 6% of patients with 5q− syndrome. Novel copy number changes not identified
by cytogenetics analysis were found in 20 MDS patients; chromosomes most
frequently associated with copy number changes were chromosomes 1, 6, 12,
and 19. No significant difference in the overall survival was observed between
patients with or without UPD regions ⬎2 Mb. Univariate analysis showed that
both copy number changes (in particular deletions) and IPSS score significantly
affected survival, although with multivariate analysis the IPSS score was the only
predictive variable for overall survival.
Studies on MDS with Del(5q)

The above-mentioned study by Mohamedali et al. included data on 16 patients
with 5q− syndrome (44). SNP microarrays confirmed the presence of the del(5q)
in all cases, and only one patient showed an additional copy number change, a
deletion on chromosome 21. Uniparental disomy regions ⬎2 Mb were found in
102 Pellagatti
six of the 16 cases with 5q− syndrome; chromosomes with more than one UPD
region were chromosomes 1, 4, 6, 16, 17, 18, and 22.
A study by Christina Evers and her colleagues in Düsseldorf (45) used a
44 K oligonucleotide array CGH platform to analyze 13 cases with an isolated
del(5q). Of these 13 cases, 12 were patients with MDS (9 of which with 5q−
syndrome) and 1 patient had AML. Array CGH allowed a more precise mapping
of the del(5q) end points and identified additional aberrations, with four of those
aberrations occurring in two or more samples. However, three of these additional
aberrations mapped to known copy number polymorphisms also present in healthy
populations.
Our group in Oxford has also used SNP microarrays to investigate copy
number changes and LOH in MDS del(5q) patients (46). Using Affymetrix 50 K
SNP microarrays, 42 cases of MDS with del(5q) (including 21 patients with 5q−
syndrome) and 45 healthy controls were studied. SNP microarrays confirmed
the presence of a del(5q) in all cases, and the boundaries of the CDR in 5q−
syndrome patients confirmed the previous mapping data (21). While no additional
copy number changes were identified in the 21 patients with the 5q− syndrome,
copy number losses or gains were detected in 10 of the other 21 del(5q) patients.
Uniparental disomy regions ⬎2 Mb were found in 19 of 21 patients with 5q−
syndrome and in all other 21 del(5q) patients. The chromosomes with the highest
frequency of UPD regions were chromosome 4 in the 5q− syndrome patient group,
and chromosomes 6 and 13 in the other del(5q) patient group. Recurrent regions
of UPD found in two or more patients were on chromosomes 4q, 6q, 8q, and 9p.
The finding of UPD regions on chromosome 4 is in agreement with the previously
mentioned study by Mohamedali et al. (44), and this interesting finding is being
investigated further with support from the MDS Foundation.
CONCLUSIONS
Several groups of investigators have now published studies harnessing the extreme
power of gene expression microarrays for the study of MDS. The results observed
are still quite heterogeneous, and this reflects diverse platforms, patient subsets,
and types of cells studied, as well as the heterogeneity of pathogenetic processes
in MDS. However, there are already several common findings, and surely these
will increase with future studies on a larger number of patients using the most
comprehensive platforms available. Also, pathway analysis is predicted to improve
with the increasingly better annotation of gene functions and gene pathways. The
identification by expression microarrays of deregulated genes and the knowledge
of their role in cell pathways will inevitably form the basis of a better understanding
of the pathophysiology of MDS.
For CGH/SNP microarray studies, one of the biggest challenges is the
determination of copy number polymorphisms in the healthy population in order
to distinguish these from changes that may be disease associated. Ideally, these
types of investigations should include constitutive DNA samples obtained from
all patients studied. However, this is often logistically problematic, and therefore
it is necessary to study large cohorts of controls to establish the incidence of
polymorphisms at each locus in the healthy population.
CGH/SNP microarrays have a higher resolution than conventional cytoge-
netic methods such as G-banded karyotyping. The arrays can also provide infor-
mation on samples where metaphase cytogenetics was not done or failed, and on
archival material, as these technologies require DNA samples rather than a fresh
sample containing dividing cell populations. SNP microarrays also offer the addi-
tional feature of UPD determination. However, metaphase cytogenetics maintains
the advantage of detecting chromosomal translocations and abnormalities present
in small clones. It is therefore most likely that in future both conventional cytoge-
netic methods and SNP/CGH microarrays will coexist for diagnostic purposes.
In conclusion, CGH/SNP microarrays can provide accurate information
on chromosomal gain/loss and gene copy number changes, as well as the loca-
tion of regions affected by UPD. Gene expression microarrays can then tell if
such changes also result in differences in the expression of genes mapping to
these regions. The ultimate target is the integration of data obtained with both
gene expression microarrays and with CGH/SNP arrays on samples obtained
from the same patient, and this will provide a more complete and extremely useful
picture of both genome and transcriptome status of MDS patients.
REFERENCES
1. Golub TR, Slonim DK, Tamayo P, et al. Molecular classification of cancer: Class dis-
covery and class prediction by gene expression monitoring. Science 1999; 286:531–
537.
2. Miyazato A, Ueno S, Ohmine K, et al. Identification of myelodysplastic syndrome-
specific genes by DNA microarray analysis with purified hematopoietic stem cell
fraction. Blood 2001; 98:422–427.
3. Moore KA, Pytowski B, Witte L, et al. Hematopoietic activity of a stromal cell
transmembrane protein containing epidermal growth factor-like repeat motifs. Proc
Natl Acad Sci U S A 1997; 94:4011–4016.
4. Hofmann WK, de Vos S, Komor M, et al. Characterization of gene expression of
CD34+ cells from normal and myelodysplastic bone marrow. Blood 2002; 100:3553–
3560.
5. Ueda M, Ota J, Yamashita Y, et al. DNA microarray analysis of stage progression
mechanism in myelodysplastic syndrome. Br J Haematol 2003; 123:288–296.
7. Zeng W, Miyazato A, Chen G, et al. Interferon-gamma-induced gene expression in
CD34 cells: Identification of pathologic cytokine-specific signature profiles. Blood
2006; 107:167–175.
8. Sternberg A, Killick S, Littlewood T, et al. Evidence for reduced B-cell progenitors
in early (low-risk) myelodysplastic syndrome. Blood 2005; 106:2982–2991.
104 Pellagatti
9. Maserati E, Aprili F, Vinante F, et al. Trisomy 8 in myelodysplasia and acute leukemia

is constitutional in 15–20% of cases. Genes Chromosomes Cancer 2002; 33:93–97.
10. Fenaux P. Chromosome and molecular abnormalities in myelodysplastic syndromes.
Int J Hematol 2001; 73:429–437.
11. Maciejewski JP, Risitano A, Sloand EM, et al. Distinct clinical outcomes for cytoge-
netic abnormalities evolving from aplastic anemia. Blood 2002; 99:3129–3135.
12. Paulsson K, Sall T, Fioretos T, et al. The incidence of trisomy 8 as a sole chromosomal
aberration in myeloid malignancies varies in relation to gender, age, prior iatrogenic
genotoxic exposure, and morphology. Cancer Genet Cytogenet 2001; 130:160–165.
tol 2000; 108:346–356.
highlighting current controversies. Leuk Res 2003; 27:95–120.
15. Chen G, Zeng W, Miyazato A, et al. Distinctive gene expression profiles of CD34 cells
from patients with myelodysplastic syndrome characterized by specific chromosomal
abnormalities. Blood 2004; 104:4210–4218.
16. Schoch C, Kohlmann A, Dugas M, et al. Genomic gains and losses influence expres-
sion levels of genes located within the affected regions: A study on acute myeloid
leukemias with trisomy 8, 11, or 13, monosomy 7, or deletion 5q. Leukemia 2005;
19:1224–1228.
17. Giagounidis AA, Germing U, Wainscoat JS, et al. The 5q- syndrome. Hematology
2004; 9:271–277.
18. Boultwood J, Lewis S, Wainscoat JS. The 5q- syndrome. Blood 1994; 84:3253–3260.
19. Nilsson L, Eden P, Olsson E, et al. The molecular signature of MDS stem cells supports
a stem-cell origin of 5q myelodysplastic syndromes. Blood 2007; 110:3005–3014.
in patients with the 5q- syndrome. Br J Haematol 2007; 139:578–589.
21. Boultwood J, Fidler C, Strickson AJ, et al. Narrowing and genomic annotation of the
commonly deleted region of the 5q- syndrome. Blood 2002; 99:4638–4641.
22. Ebert BL, Pretz J, Bosco J, et al. Identification of RPS14 as a 5q- syndrome gene by
RNA interference screen. Nature 2008; 451:335–339.
N Engl J Med 2005; 352:549–557.
25. Pellagatti A, Jadersten M, Forsblom AM, et al. Lenalidomide inhibits the malignant
clone and up-regulates the SPARC gene mapping to the commonly deleted region in
5q- syndrome patients. Proc Natl Acad Sci U S A 2007; 104:11406–11411.
26. Zhang LH, Schafer P, Muller G, et al. Lenalidomide displays direct anti-non-
Hodgkin’s lymphoma (NHL) cell activity in association with enhanced SPARC
expression but independent of its ability to strongly inhibit NHL cell VEGF pro-
duction in vitro. Blood 2007; 3473.
27. Wang XD, Reeves K, Luo FR, et al. Identification of candidate predictive and surrogate
molecular markers for dasatinib in prostate cancer: Rationale for patient selection and
efficacy monitoring. Genome Biol 2007; 8:R255.
28. Rosenwald A, Chuang EY, Davis RE, et al. Fludarabine treatment of patients with
chronic lymphocytic leukemia induces a p53-dependent gene expression response.
Blood 2004; 104:1428–1434.
29. Huang CY, Beer TM, Higano CS, et al. Molecular alterations in prostate carcinomas
that associate with in vivo exposure to chemotherapy: Identification of a cytoprotec-
tive mechanism involving growth differentiation factor 15. Clin Cancer Res 2007;
13:5825–5833.
30. Miller WR, Larionov AA, Renshaw L, et al. Changes in breast cancer transcriptional
profiles after treatment with the aromatase inhibitor, letrozole. Pharmacogenet Genom
2007; 17:813–826.
31. Pellagatti A, Esoof N, Watkins F, et al. Gene expression profiling in the myelodysplas-
tic syndromes using cDNA microarray technology. Br J Haematol 2004; 125:576–583.
32. Gibbons RJ, Pellagatti A, Garrick D, et al. Identification of acquired somatic mutations
in the gene encoding chromatin-remodeling factor ATRX in the alpha-thalassemia
myelodysplasia syndrome (ATMDS). Nat Genet 2003; 34:446–449.
33. Weatherall DJ, Old J, Longley J, et al. Acquired haemoglobin H disease in leukemia:
Pathophysiology and molecular basis. Br J Haematol 1978; 38:305–322.
34. Higgs DR, Bowden DK. Clinical and laboratory features of ␣-thalassemia syndromes.
In: Nagel RL, ed. Disorders of Hemoglobin. Cambridge:Cambridge University Press,
2001:431–469.
35. Higgs DR, Wood WG, Barton C, et al. Clinical features and molecular analysis of
acquired hemoglobin H disease. Am J Med 1983; 75:181–191.
36. Gibbons RJ, Higgs DR. Molecular-clinical spectrum of the ATR-X syndrome. Am J
Med Genet 2000; 97:204–212.
37. Steensma DP, Higgs DR, Fisher CA, et al. Acquired somatic ATRX mutations in
myelodysplastic syndrome associated with alpha thalassemia (ATMDS) convey a
more severe hematologic phenotype than germline ATRX mutations. Blood 2004;
103:2019–2026.
38. Martinez-Ramirez A, Urioste M, Calasanz MJ, et al. Array comparative genomic
hybridization analysis of myelodysplastic syndromes with complex karyotypes. A
technical evaluation. Cancer Genet Cytogenet 2003; 144:87–89.
39. Martinez-Ramirez A, Urioste M, Melchor L, et al. Analysis of myelodysplastic syn-
dromes with complex karyotypes by high-resolution comparative genomic hybridiza-
tion and subtelomeric CGH array. Genes Chromosomes Cancer 2005; 42:287–298.
40. Paulsson K, Heidenblad M, Strombeck B, et al. High-resolution genome-wide array-
based comparative genome hybridization reveals cryptic chromosome changes in
AML and MDS cases with trisomy 8 as the sole cytogenetic aberration. Leukemia
2006; 20:840–846.
41. Gondek LP, Tiu R, Haddad AS, et al. Single nucleotide polymorphism arrays com-
plement metaphase cytogenetics in detection of new chromosomal lesions in MDS.
Leukemia 2007; 21:2058–2061.
plastic syndromes. Exp Hematol 2007; 35:1728–1738.
43. Gondek LP, Tiu R, O’Keefe CL, et al. Chromosomal lesions and uniparental dis-
omy detected by SNP arrays in MDS, MDS/MPD and MDS-derived AML. Blood.
Prepublished Oct 22, 2007; DOI 10.1182/blood-2007-05-092304.
106 Pellagatti
44. Mohamedali A, Gaken J, Twine NA, et al. Prevalence and prognostic significance of
allelic imbalance by single-nucleotide polymorphism analysis in low-risk myelodys-
plastic syndromes. Blood 2007; 110:3365–3373.
45. Evers C, Beier M, Poelitz A, et al. Molecular definition of chromosome arm 5q dele-
tion end points and detection of hidden aberrations in patients with myelodysplastic
syndromes and isolated del(5q) using oligonucleotide array CGH. Genes Chromo-
somes Cancer 2007; 46:1119–1128.
46. Wang L, Fidler C, Giagounidis A, et al. Genome-wide analysis of copy number
changes and loss of heterozygosity in myelodysplastic syndrome with del(5q) using
high-density SNP arrays. Haematologica 2008; 93:994–1000.
5
The Role of Apoptosis in MDS
Yatato Yoshida
The Center for Hematological Diseases, Takeda General Hospital, Kyoto, Japan
INTRODUCTION
The myelodysplastic syndromes (MDS) encompass a group of clonal stem cell
disorders characterized by both bone marrow failure and a predilection to develop
acute myeloid leukemia (AML). While the majority of patients with MDS are
initially diagnosed with lower-grade disease, nearly two-thirds will ultimately suc-
cumb to complications of peripheral cytopenias or of progression to AML. The
World Health Organization (WHO) classification (1) and the International Prog-
nostic Scoring System (IPSS) (2) provide a useful framework for classification and
risk stratification of individual patients, but these systems do not take into account
the considerable clinical heterogeneity of MDS and their diverse biology. MDS per
se represent many different conditions, probably not just a single disease. A marked
variability exists among patients even within the same subtype or IPSS risk group.
One biologically striking feature present in most MDS cases is deregula-
tion of apoptosis, or abnormal “programmed cell death”. Apoptosis deregulation
may partially explain the classical MDS-associated feature of bone marrow fail-
ure coexisting with a cellular marrow. Abnormalities of apoptosis appear to play
an important part in both early stage of bone marrow failure and, later on, in
progression to leukemia. The ineffective hematopoiesis in early MDS may be
accounted for by an excessive apoptosis of hematopoietic progenitor and differ-
entiating cells. In contrast, later stage MDS is marked by a progressive increase
of immature cells ultimately culminating in leukemic transformation, presumably
due to a progressive suppression of apoptosis and increased clonal expansion
and/or cell proliferation (3–5).
107
108 Yoshida
This chapter discusses the role of apoptosis in MDS with specific focus on
the mechanisms of apoptosis and its deregulation. For the sake of convenience,
most references included in this chapter are those published since the previous
edition of this book. Apoptosis literature is often confusing because of the varied
terminology used to describe the same factor, and reviews may also be challenging
to understand because some studies have looked at expression of apoptosis-factors
at the protein level while the others only at the RNA level. For the most part, the
studies described below referred to changes detected at the protein level, except
where specified.
PROGRAMMED CELL DEATH AND APOPTOSIS

Programmed Cell Death
Apoptosis occurs by mechanisms that have been at least partially conserved
throughout evolution. In the nematode Caenorhabditis elegans, 131 of the 1090
somatic cells generated during development undergo apoptosis at a precise time
and anatomical location, determined by each cell’s genetic endowment. Hence,
developmental biologists prefer the term “programmed cell death” to “apoptosis”.
In C. elegans, this form of cell death is dependent on the actions of at least three
key genes: ced-3, ced-4, and ced-9. The cloning of these three genes ultimately led
to the discovery of homologues in other species, including humans (Fig. 1). The
human homologue of the protein encoded by ced-9 is Bcl-2 (B cell lymphoma);
that of ced-4, Apaf-1 (apoptotic protease activating factor 1), and that of ced-3,
the caspase (cysteine-requiring aspartate-directed proteases) family. These factors
are discussed in more detail below.
The Bcl-2 family of proteins is an important group of apoptosis regulators.
They function as homodimers or heterodimers, and it is the ratio of the anti-
apoptotic versus pro-apoptotic proteins that determines the cellular susceptibility
to death signals. The survival-promoting, anti-apoptotic members of the Bcl-
2 family include Bcl-2 itself, Bcl-XL , and Mcl-1, while the death-promoting,
pro-apoptotic members of the family are Bax, Bcl-XS , Bad, Bak, Bik, and Bid.
Collectively, these proteins act as “gatekeepers” for a variety of apoptotic signals.
The final “execution” step of apoptosis is caused by the activation of a family
of cysteine proteases known as caspases, which represent the effector molecules of
apoptosis. Caspases are present as inactive proenzyems and, when activated, cleave
and activate each other in a cascade-like fashion (the so-called caspase cascade).
Through caspase-mediated cleavage of a number of intracellular structures such
as the cytoskelton and a wide range of nuclear and cytoplasmic proteins, most of
the morphological and biochemical events associated with apoptosis ensue.
Apoptosis in Hematopoietic System

Apoptosis is essential to the maintenance of homeostasis in multicellular organ-
isms, as are control of cell proliferation, differentiation, activation, and cell
The Role of Apoptosis in MDS 109
Caenorhabditis elegans Death signal Humans
Growth factor Death receptors

ces-1 receptors
ces-2
AKT Caspase-8
PKA
Bad, Bik
Egl-1 ced-9 Bcl-2 Bid, Bax
ced-4 Apaf-1
ced-3 caspase-9
Apoptosis Apoptosis
Figure 1 A cell death program evolutionary conserved from nematodes to humans. Three
genes encode factors that play a key role in the control of programmed cell death in the
nematode Caenorhabditis elegans. Ced-9, a Bcl-2 homologue, is important for cell survival,
while Ced-3 and Ced-4 are required for the induction of cell death. Ced-3 is homologous to
the caspase family and Ced-4 serves as an activator of Ced-3. Death signals initiate the cell
death pathway by inhibiting the functions of anti-apoptotic proteins (Ced-9 in C. elegans
and Bcl-2 in humans) or by activating factors which can suppress the functions of anti-
apoptotic proteins (such as Egl-1 and Bad). Inhibition of Ced-9 or Bcl-2 leads to triggering
of the next step in the suicide program, namely, activation of Ced-4 in C. elegans or Apaf-1
(apoptotic protease activating factor 1) in humans, which sets off the final executors of
apoptosis, Ced-3, or the caspase cascade.
removal. Cell cycle progression, with sequences of DNA replication, mitosis,

and cell division, is a tightly controlled and complicated process that, when dereg-
ulated, may become dangerous not only to a single cell, but also to the whole
organism. In steady-state hematopoiesis, blood cells undergo constant renewal
from hematopoietic stem cells. A delicate balance between cell proliferation and
cell death determines the control of cell number of each blood cell lineage.
In terms of tissue kinetics, apoptosis may be a mechanism that counter-
balances the effect of cell proliferation by mitotic division. In fact, deregulated
apoptosis has been implicated as a fundamental pathogenetic mechanism in a
variety of human diseases. Excessive apoptotic cell death may cause cytopenias
and marrow failure, such as the one that has been demonstrated in aplastic anemia
110 Yoshida
or large granular lymphocyte leukemia. On the other hand, inefficient elimination

of abnormal cells may lead to the development of neoplasia such as AML.
THE INTRINSIC PATHWAY OF APOPTOSIS

Apoptosis can be triggered in a cell through either an extrinsic pathway or an
intrinsic pathway (Fig. 2). The extrinsic pathway is initiated through the stimu-
lation of the transmembrane death receptors, such as the Fas and tumor necrosis
factor-alpha (TNF-␣) receptors, which span the cell outer membrane. In contrast,
Extrinsic Intrinsic
Death-including DNA damage

signaling complex
P53
Fas-L
Death Promoting
Bcl-2 protein (Bid)
Fas
Mitochondria
FADD
FLIP tBid
Cytochrome-c
Caspase-8 Bid
Apaf-1 Smac/DIABLO
Caspase-9
Granzyme B
Effector caspase
(caspase 3, 6, 7) IAPs
Apoptosis
Figure 2 The intrinsic and extrinsic pathways of apoptosis. (Left): The extrinsic (cyto-
plasmic) pathway is triggered by activation of specific “death ligands” such as CD95 (Fas)
and tumor necrosis factor-alpha receptor (TNFR) 1. Activation of Fas and TNFR by Fas-L
and TNF, respectively, leads to the activation of caspase-8 through the trimerization of death
receptors and recruitment of the adaptor molecule FADD. The TNFR-mediated pathway
utilizes TRADD protein in recruitment of FADD. FLIP, a mutated form of caspase-8, binds
to FADD, thus inhibiting the binding and activation of caspase-8. Activated caspase-8 is
able to induce apoptosis through a mitochondrial pathway by cleaving Bid. (Right): The
internal (mitochondrial) signal pathway is driven by the inactivation of Bcl-2 leading to
the release of cytochrome c from the mitochondria and activation of the death signal. The
inhibitor of apoptosis (IAP) family of proteins prevents cell death by binding to and inhibit-
ing active caspase-9 and are negatively regulated by IAP-binding mitochondrial proteins,
such as the mammalian protein Smac [second mitochondrial–derived activator of caspase
/DIABLO (direct IAP binding protein with low PI)]. Both pathways converge to a final
common pathway involving the activation of caspase cascade, culminating in the death of
the cell.
the intrinsic pathway of apoptosis is initiated through the release of signal factors
by mitochondria within the cell.
The Extrinsic Pathway of Apoptosis

In the extrinsic pathway, signal molecules known as ligands are released by other
cells and bind to transmembrane death receptors on the target cell to induce
apoptosis. For example, natural killer cells possess the Fas ligand (Fas-L) on their
surface. The binding of the Fas-L to Fas receptor (one of the death receptors) on
a target cell will trigger multiple receptors to aggregate together on the surface
of the target cell. The aggregation of these receptors recruits an adaptor protein
known as Fas-associated death domain protein (FADD) on the cytoplasmic side
of the receptors. FADD, in turn, recruits caspase-8, an initiator caspase, to form
the death-inducing signal complex (DISC). During recruitment of caspase-8 to
the DISC, caspase-8 is activated and is then able to directly activate caspase-3, an
effector caspase, thereby initiating degradation of the cell. Active caspase-8 can
also cleave Bid protein (a Bcl-2 protein family member) to tBid, which acts as a
signal on the membrane of mitochondria to facilitate the release of cytochrome c
and initiate the intrinsic pathway of apoptosis, described below.
Other death receptors include TNF-␣ receptor and receptors for TRAIL
(TNF-related apoptosis inducing ligand). Together Fas, TNF-␣, and TRAIL recep-
tors form the TNF receptor superfamily, and have significant sequence and struc-
tural homology. Several nonfunctional decoy death receptors also exist, and
relative expression of functional receptor versus decoy receptor may alter cell
susceptibility to apoptosis.
The Intrisic Pathway of Apoptosis

The intrinsic pathway of apoptosis is triggered by cellular stress, specifically mito-
chondrial stress caused by factors such as DNA damage (e.g., induced by ionizing
radiation or chemotherapy). Upon receiving a stress signal, the mitochondrial
membrane polarization is lost and pro-apoptotic Bcl-2 family proteins in the cyto-
plasm, Bax, and Bid, bind to the outer membrane of the mitochondria to signal
the release of the internal contents. However, the signal of Bax and Bid is not by
itself enough to trigger a full release of these contents. Bak, another pro-apoptotic
protein that resides within the mitochondria, is also needed to fully promote the
release of cytochrome c and the other pro-apoptotic intramembrane content from
the mitochondria.
Following its escape from mitochondria, cytochrome c forms a complex
in the cytoplasm with adenosine triphosphate (ATP), an energy molecule, and
Apaf-1, an enzyme. This complex then activates caspase-9, an initiator protein.
In return, the activated caspase-9 works together with the complex of cytochrome
c, ATP, and Apaf-1 to form an “apoptosome”, which activates caspase-3, the
effector protein that initiates degradation. Besides the release of cytochrome c
from the intramembrane space, the intramembrane contents released also contains
112 Yoshida
apoptosis inducing factor (AIF) to facilitate DNA fragmentation, and Smac/

DIABLO proteins to inhibit the inhibitor of apoptosis protein family (IAPs).
APOPTOSIS-RELATED PARAMETERS IN MDS

TNF-␣, Fas-Ligand, and TRAIL in MDS
Overexpression of negative regulators of hematopoiesis, such as TNF-␣, Fas-L,
and TRAIL, may contribute to apoptosis in MDS (3–5). Upregulation of TOLL-
like receptor (TLR), a member of conserved transmembrane receptors that initiates
a complex cascade of events including altered regulation of thousands of genes,
has also been reported in MDS (6). Table 1 lists representative reports on the role
of TNF-␣, Fas-L, and TRAIL in MDS.
Several groups have reported upregulation of Fas or Fas-L in a subset of MDS
cases (7,8). Likewise, upregulation of TNF-␣ levels in MDS has been reported by
Joachim Deeg in Seattle and colleagues (9) and by Gerald Stifter and colleagues
in Innsbruck (10), as well as by Azra Raza and her coworkers and others. Deeg
and colleagues found no relationship between TNF-␣ levels and FAB subtypes or
IPSS risk group, whereas the latter suggested a correlation between TNF-␣ level
and the degrees of anemia.
TRAIL and its receptors have been shown to be constitutively expressed
in some marrow samples taken from patients with MDS. In culture, TRAIL can
induce apoptosis of MDS marrow cells (11), and suppress MDS progenitor growth
(12). The existence of pre-mRNA splice variants for TRAIL receptors and other
factors in the death receptor pathway hints at the complexity of apoptosis regulation
and the challenges in defining specific apoptosis defects associated with MDS.
Caspase Activity in MDS

Elevated caspase-3 and caspase-1 levels have been reported in MDS samples by
studies using enzymatic assays and Western blots (Table 2) (13,14). Increased
caspase-3 activity has been correlated with enhanced apoptosis in early MDS, as
assessed by techniques such as TUNEL (terminal deoxynucleotidyl transferase–
mediated dUTP-biotin nick end labeling) and Annexin V positivity (a measure
of phosphatidylserine translocation at the cell membrane—an early apoptosis-
associated event) (13,15). Refractory anemia with ringed sideroblasts (RARS), in
particular, may be associated with an elevated caspase-3 activity and a high rate
of apoptosis. Altered mitochondrial membrane potential has been implicated in
erythroid apoptosis in RARS (16). Expression of caspases 8, 5, 3, 2, and 1 has also
been reported to have a positive correlation with apoptosis in de novo MDS (17).
NF-Kappa B Activation in MDS

NF-␬B (nuclear factor kappa B) is a transcription factor that regulates a wide
variety of genes, including many involved in cell proliferation and survival
Table 1 TNF-␣, Fas-Ligand and TRAIL Expression in MDS
Mundle et al. (7) Gupta et al. (8) Deeg et al. (9) Stifter et al. (10)
Method mRNA ELISA Immunohisto chemistry

The Role of Apoptosis in MDS
Parameters Fas, Fas-L, caspases 1 and 3 Fas-L TNF-␣, Fas, Fas-L TNF-␣
examined
N 17 MDS, 4 N, 3 post-MDS 50 MDS, 6 N, 10 de novo 80 MDS 89 MDS, 12 N
AML AML, 6 post-MDS AML
Results All mRNA detectable except Fas-L high in MDS, highest TNF-␣ high in MDS TNF-␣ overexpressed in
caspase 3 (undetectable in in AML MDS
N)
Notes ISEL high in MDS (2.6%), FAS-L correlates with FAB TNF-␣ no relation to FAB, TNF-␣ level highest in RA,
low in N (0.7%), and and degree of anemia IPSS, Fas-L inversely correlates with BM
AML (0.2%) related to cytogenetic risk cellularity and degree of
anemia
Abbreviations: ISEL, in situ end labeling; ELISA, enzyme-linked immunosorbent assay.

113
114
Table 2 Caspase Activity in MDS

Economopoulou et al.
Boudard et al. (13) Boudard et al. (14) Bouscary et al. (15) Mattes et al. (16) (17)
Method Fluorogenic assay Fluorogenic assay Fluorogenic assay mRNA

Western Western
Parameters Caspases 1 and 3 Caspases 3 and 8, Bid Caspase-3 Caspase-3, ROS Caspase, granzyme B,
examined production ⌬ ψm Bcl-2 family protein
N 83 MDS 9 RA, 3 RARS 54 MDS 9 RARS 46
Results Both caspase-3 and Caspase-3 activity High apoptosis in High caspase-3 and Correlation in
apoptosis (TUNEL) elevated after 24 hr RA/RARS (44.8%) low ⌬ ψm caspases 8, 5, 3, 2, 1
higher in RA/RARS culture than in advanced correlated with and apoptosis, Bfl1
than in advanced MDS (21.4%), Annexin V, ROS no and mcl1 high in
MDS caspase-3 higher in difference from N RAEB-1
RA/RARS than in
N
Notes Specific caspase Fas blocking no effect Broad spectrum Mitochondrial
inhibitors increase on caspase 3, 8, or caspase inhibitor membrane potential
progenitor growth Bid Z-VAD-FMK implicated in
decreases Annexin erythroid apoptosis
V+ cells but no in RARS
effect on progenitor
growth
Abbreviations: ψm, mitochondrial transmembrane potential; TUNEL, TdT-mediated dUTP-biotin nick end labeling;
ROS, reactive oxygen species.
Yoshida
Table 3 NF-␬B in MDS

Carvalho et al. (18) Braun et al. (19)
Method Nuclear membrane extraction and Nuclear membrane extraction and

electrophoretic mobility shift electrophoretic mobility shift
⌬␺ m ⌬ ⌿m
Annexin V Annexin V
NF-␬B binding I␬-B measurement
Immunoblot Immunoblot
Parameters NF-␬B NF-␬B
examined Caspase-3, Endo G, AIF, Bcl-xL, Caspase-3, I␬-B-P, C-IAP2, Endo
C-IAP2 G, AIF
N 10 MDS, 5 AML 50 MDS, 7 controls
Results Blasts from high-risk MDS & CD34+ cells from high-risk MDS
AML NF-␬B activation and nuclear
Constitutive activation of NF-␬B translocation
Dependent on activation of I-␬B NF-␬B inhibitor triggers
kinase complex mitochondrial death pathway
IKK␥ /NEMO antagonists–induce
caspase independent loss of
⌬␺ m
Notes Survival of high-risk MDS and NF-␬B activation detected
AML cells dependent on in high-risk but not in low-risk
IKK␥ /NEMO MDS,
restricted to cytogenetically
abnormal cells
Abbreviations: ψm, mitochondrial transmembrane potential; AIF, apoptosis-inducing factor; EndoG,

endonuclease G; c-IAP2, cellular inhibitor of apoptosis protein 2; Iκ-B, inhibitor of NF-κB; IKK, Iκ-B
kinase; NEMO, NF-κB essential modulator.
(including apoptosis regulation). NF-␬B activation correlates with MDS

progression—CD34+ cells from higher risk MDS patients show increased NF-␬B
activity compared to lower risk patient samples (Table 3) (18,19). TNF-␣ is one
of the factors that can induce NF-␬B translocation from cytoplasm to the nucleus
with its concomitant activation.
Another activating factor is FLICE-like inhibitory protein (FLIP), an anti-
apoptotic protein that is involved, along with TNF-␣, in an autoamplification loop
with NF-␬B (20,21). (FLICE is an acronym for FADD-like IL-1b–converting
enzyme.) Two splice variants of FLIP, FLIPSHORT and FLIPlong , have been studied
in MDS marrow cells. In lower risk MDS, FLIPSHORT mRNA expression was
reported to be high by one group (22), but low by another (23). These conflicting
results may be due, in part, to technical differences.
P38 MAP Kinase Activation in MDS

P38 MAPK (mitogen-activated protein kinase) activation is important in regulating
the growth inhibitory signals of TNF-␣, transforming growth factor beta (TGF-␤),
116 Yoshida
Table 4 Expression of Pro-Apoptotic and Anti-Apoptotic Proteins in MDS

Boudard et al. (27) Gianelli et al. (28) Invernizzi et al. (29)
Method Immunohistochemistry mRNA immunohis- Immunohistochemistry,

tochemistry TUNEL
Parameters Anti-apoptotic: Bcl-2, Survivin Survivin
examined Bcl-Xl
Pro-apoptotic: Bax,
Bad, Bak, Bcl-Xs
N 168 MDS 49 BM aspirates, 17 34 CMML, 10 MDS,
BM biopsies 41 AML
Results Anti-apoptotic: Bcl-2, Survivin mRNA and Apoptosis higher in
Bcl-Xl high in protein high in MDS and CMML
advanced MDS MDS than in N or, AML
Pro-apoptotic: Bax, High mRNA in int-1 Survivin mRNA high
Bad, Bak, Bcl-Xs than int-2 or in CMML than MDS
high in low-risk high-risk MDS or AML
MDS
Notes Expression of Increased survivin in
pro-apoptotic MDS especially
protein associated in low-risk MDS
with longer survival
and high
anti-apoptotic
protein associated
with shorter survival
and interferon. P38 MAPK is constitutively activated or phosphorylated in marrow

cells of lower risk MDS and this finding correlates with enhanced apoptosis (24). In
one experiment, P38 MAPK inhibition decreases apoptosis of CD34+ progenitor
cells from MDS patients and restores hematopoiesis (24,25). Thus, P38 MAPK
may be implicated in ineffective hematopoiesis in MDS, and its inhibition may
have therapeutic role (24–26).
Pro-Apoptotic and Anti-Apoptotic Proteins

As described above, among the most important regulators of apoptotic pathway
are the Bcl-2 family proteins. These family members accelerate or inhibit apop-
tosis with the ratio of pro-apoptotic versus anti-apoptotic molecules determining
whether a cell survives or undergoes programmed cell death. An immunohisto-
chemical study of anti-apoptotic (Bcl-2, Bcl-Xl ) and pro-apoptotic (Bax, Bad,
Bak, Bcl-XS ) proteins suggested that levels of anti-apoptotic proteins were ele-
vated in higher risk MDS and those of pro-apoptotic proteins decreased in lower
risk MDS (Table 4) (27).
Another protein, survivin, an IAP family protein that has a role both in
counteracting apoptosis and in regulating cell division, was studied in MDS by
one group of investigators. Increased survivin expression and elevated protein
levels were found in low-risk MDS (28,29). All these results are in line with
the general notion that apoptosis is excessive in early, lower risk MDS but is
diminished in advanced, higher risk MDS.
Mitochondrial Changes in MDS

Abnormalities of mitochondrial functions have long been postulated in siderob-
lastc anemias. Nearly half of erythroid progenitors from MDS exhibit spontaneous
release of cytochrome c from mitochondria with ensuing activation of caspase-9.
G-CSF inhibits cytochrome c release and suppresses apoptosis, most noticeably
in cells from patients with RARS (30).
Role of the Marrow Microenvironment in MDS in Potentiating Apoptosis

Abnormal stromal cell function may contribute to apoptosis in MDS (31,32).
Flores-Figueroa (33) reported high production of TNF-␣ and IL-6 by marrow
stromal cells, contributing to high apoptosis induction. However, in co-culture
system with MDS stromal cells, a direct cell-to-cell contact rather than the release
of soluble factors seemed to play a major role (31,32).
Cell Proliferation Versus Apoptosis in MDS

High rates of apoptosis and simultaneously increased proliferation of marrow cells
is a cell kinetic hallmark of MDS. Excess cell proliferation has been suggested
by increased expression of proliferation markers and incorporation of BRdU (34–
36). Suneel Mundle and colleagues (37) proposed a “signal antonomy” where
a significant proportion of S-phase cells are simultaneously apoptotic in MDS;
in their study, this signal antonomy was linked to an altered expression of the
transcription factor E2F1. Increased proliferation is especially seen in advanced
MDS, and inhibition of apoptosis rather than excessive apoptosis is closely related
to AML transformation (34).
APOPTOSIS AND MDS DISEASE PROGRESSION

MDS progression arises through multiple mechanisms that alter the levels of
CD34+ cell apoptosis and proliferation. For instance, disease progression may
be accompanied by a fall in pro-apoptotic versus anti-apoptotic protein ratios,
as discussed earlier (34). Likewise, at least one group has argued that refractory
anemia with excess blasts in transformation (RAEB-t) differs from de novo AML
in a way that apoptosis is still higher in RAEB-t than in de novo AML (38).
Other signaling pathways, such as phosphatidylinositol 3-kinase (PI3 K)/
AKT/mammalian target of rapamycin (mTOR), have been reported to be activated
118 Yoshida
in high-risk MDS in immunohistochemical and flow cytometric studies (39,40).

Rapamycin was capable of inducing apoptotic death of CD33+ cells from higher
risk MDS, indicating a possible therapeutic importance to this observation (40).
In addition to the deregulation of apoptosis, accumulation of genetic mutations
may be involved in progression to AML. A recent comprehensive survey of the
molecular mutations in MDS versus AML (41) has compared the frequency of
FLT3-length mutations (FLT3-LM), FLT3-TKD, MLL-partial tandem duplica-
tions (MLL-PTD), NRAS codon 12 mutations, and KIT D816 mutations among
381 MDS and 4130 AML patients. The prevalence of each of these mutations
except for FLT3-TKD increased during progression along the spectrum from
lower risk MDS to higher risk MDS and to AML.
EPIGENETIC MODIFICATIONS
Epigenetic changes of DNA, which are acquired alterations in nucleotide struc-
ture (e.g., cytosine methylation) without a change in nucleotide sequences, may
be involved in the pathobiology of many malignancies (see chap. 21 for details).
In MDS, a number of epigenetic changes have been demonstrated, such as hyper-
methylation of cytosine residues and abnormal histone acetylation patterns, both
of which are induced by the DNA methyltransferase enzymes. Genes often hyper-
methylated in MDS include P15 (a cyclin-dependent kinase and a putative tumor
suppressor), DAPK (a serine/threonine kinase and regulator of apoptosis), SOCS1
(a negative regulator of cytokine signaling), and RASSF1A (a negative regulator of
Ras signaling) (42,43). These epigenetic changes are believed to result in silencing
of tumor suppressor genes, thus impairing normal cell differentiation and aug-
menting AML transformation. They also induce perturbation of cell cycle control,
apoptosis, differentiation, and DNA repair (43,44). Reversal of DNA methylation
by hypomethylating agents such as azacytidine and decitabine and histone deacety-
lase inhibitors can restore function and allow for normal cell differentiation, as
well as upregulation of pro-apoptotic genes and downregulation of anti-apoptotic
genes (45). The details of these changes have not yet been worked out.
CONTROVERSIAL ISSUES RELATED TO APOPTOSIS IN MDS

Apoptosis in Relation to the Stages of Cell Differentiation
The nature and the stage of cell differentiation of apoptotic cells in MDS remains
the subject of considerable debate. Earlier reports based on apoptosis-related
parameters indicated that apoptosis in MDS is mainly seen in CD34+ progen-
itors (3). However, subsequent studies have been inconsistent. A TUNEL and
immunohistochemical study by Alessandro Pecci and his colleagues in Pavia,
Italy, reported a significantly higher apoptotic rate in CD34− cells than in CD34+
cells (46). An ultrastructural study (47) showed apoptosis in all stages of differen-
tiation ranging from blasts to terminally differentiated cells in all three lineages.
Another analysis by a French group indicated that erythroid apoptosis specifically

begins early in erythropoiesis but increases with erythroid differentiation (48).
Differences in CD34+ and CD34− cell apoptosis between FAB subtypes
have also been examined. Tsoplou and colleagues in Patras (49) detected excessive
apoptosis in both CD34+ and CD34− cells in good prognosis MDS, while in
poorer prognosis MDS, apoptosis was restricted to CD34+ cells, for unclear
reasons. Gaëtan Berger and colleagues in France (50) found that samples from
patients with early stage MDS had higher levels of progenitor cells apoptosis and
lower levels of apoptosis were present in blast cells in advanced MDS. Yet another
report found both CD34+ and CD34− fractions in MDS showed comparable
degrees of apoptosis. However, when cell groups were isolated and tested for
apoptosis in short-term cultures, the CD34+ fraction had lower propensity to
undergo apoptosis; the authors concluded that the CD34− fraction is a possible
source for pro-apoptotic signaling (51).
Clinical Implications
Most investigators currently favor the view that apoptosis is more pronounced in
early stages of MDS than in advanced stages, and that apoptosis diminishes as
the disease progresses. But little is known of the clinical relevance of apoptosis
in MDS. If apoptosis is truly responsible for ineffective hematopoiesis in MDS,
the degree of cytopenias might be expected to correlate with apoptosis. But only
some reports are consistent with this, as discussed above (3,4). Factors impairing
our ability to make clear conclusions include a variety of methods used to assess
apoptosis, methodological difficulties, and limitations inherent to each method,
as well as a marked heterogeneity of MDS. A simple and reliable technique to
evaluate apoptosis in the most relevant cells in MDS would be important before
measurement of apoptosis rates could be practiced in the clinical setting, but such
a technique has thus far proven elusive.
Is Apoptosis in MDS Restricted to Clonal Cells?

Elaine Sloand and her colleagues at the National Institutes of Health in Bethesda
(52) have assessed apoptosis in samples from MDS patients with trisomy 8, and
they showed that apoptotic markers, Fas expression and caspase-3 activity, were
high in both trisomy 8 clones as well as cytogenetically normal cells, although
the levels were generally higher in trisomy 8 clones. Sloand and colleagues fur-
ther suggested that trisomy 8 cells undergo incomplete apoptosis and nonetheless
are capable of colony formation and growth (53). Another study also reported
preserved growth potential despite enhanced apoptosis of trisomy 8 clone (54).
Apoptosis was reported in both clonal and nonclonal cells in one study that used
FISH as a marker of clonality and in situ end labeling (ISEL) or Annexin V as mark-
ers of apoptosis (55). However, apoptosis was found more frequently in normal,
nonclonal cells in MDS. If excessive apoptosis involves normal cells to a greater
extent than clonal cells, even if proliferation rates were similar, this would give a
survival advantage to clonal cells and allow slow expansion of the MDS clone.
120 Yoshida
Different cytogenetic abnormalities may also involve discrete molecular

changes. As described in chapter 4, a recent gene expression profiling study of
CD34+ cells from patients with MDS suggested downregulation of genes impli-
cated in apoptosis inhibition in trisomy 8, in contrast to upregulation of genes
related to leukemic transformation, tumorigenesis, and apoptosis; and downreg-
ulation of genes controlling cell growth and differentiation in monosomy 7 (56).
Trisomy 8 clones may be highly susceptible to apoptosis. “Clonality” in all these
studies refers to known cytogenetic abnormalities, and one limitation of this work
is that molecular clonality is still possible even if it is not associated with demon-
strable chromosomal changes.
THERAPEUTIC MANIPULATIONS OF DEREGULATED APOPTOSIS

IN MDS
Lower Risk MDS
Theoretically, therapies targeting increased apoptosis in low-risk MDS might act
through enhancing survival signals or by blocking death signals. Combined use of
hematopoietic growth factors (i.e., G-CSF and Epo) improves anemia, especially
in patients with RARS, reportedly through suppression of apoptosis in erythroid
progenitors. This growth factor–induced suppression of apoptosis is mediated by
inhibition of cytochrome c release from mitochondria (57).
Several efforts have been made to inhibit the activity of TNF-␣, as described
in more detail in chapter 19. Clinical trials of TNF-␣ inhibitors in MDS have
included the use of soluble TNF-␣ receptor fusion protein (etanercept) and a
chimeric anti–TNF-␣ monoclonal antibody (infliximab). The former has low-to-
moderate clinical efficacy in low-risk MDS (58,59), while the latter showed some
activity in pilot studies (60).
The dramatic effects of immunomodulatory drug lenalidomide in MDS
associated with an isolated 5q chromosome deletion include not only the improve-
ment in hemoglobin and achievement of transfusion independence, but also the
disappearance of 5q− chromosome abnormality (61). These results are poten-
tially consistent with a view that the 5q− clone undergoes selective apoptotic
death during lenalidomide therapy, sparing the normal clone. This same drug has
been shown to have some role in amelioration of anemia in lower risk MDS not
associated with del(5q) (62), but is of much more limited efficacy in that setting.
It is unknown whether the mechanisms of lenalidomide action actually involve
apoptosis or whether the effects are the same or not in del(5q) and nondel(5q)
MDS.
Higher Risk MDS

Therapies attempting at facilitating apoptotic death of neoplastic cells of high-risk
MDS have included DNA methylation inhibitors, histone deacetylase (HDAC)
inhibitors, farnesyl-transferase inhibitors, and angiogenesis inhibitors. The extent
to which these agents induce cytoreduction via apoptosis versus more conventional
cytotoxicity is unclear. They are discussed in more detail in chapter 20 and 21.
CONCLUSION AND FUTURE DIRECTIONS

Since the early 1990s, the role of apoptosis in MDS pathobiology has been a
topic of active investigation. Apoptosis biology is complex, and this makes the
task of translating knowledge of apoptosis into the clinic challenging. Still, sev-
eral currently available agents may, either directly or indirectly, alter the rate of
apoptosis in MDS. Future studies should be directed towards (1) the development
of a simple and reliable marker for the evaluation of apoptosis in clinical samples;
(2) further elucidation of the apoptotic machinery and its regulators; (3) improved
understanding of the role of apoptosis in the pathophysiology and progression of
MDS, with careful controlling for stage of disease and cell populations studied;
and (4) establishment of effective combinations of various compounds to improve
clinical and hematological conditions of needy patients.
REFERENCES
1. Vardiman JW, Harris NL, Brunning RD. The World Health Organization (WHO)
classification of the myeloid neoplasms.Blood 2002; 100(7):2292–2302.
2. Greenberg P, Cox C, Le Beau MM, et al. International scoring system for evaluating
prognosis in myelodysplastic syndromes.Blood 1997; 89(6):2079–2088.
3. Yoshida Y. The role of apoptosis in the myelodysplastic syndromes. In: Bennett JB,
ed. The Myelodysplastic Syndromes. Pathobiology and Clinical Management. 1st ed.
New York, NY: Marcel Dekker, 2002:177–201.
4. Parker JE, Mufti GJ. The myelodysplastic syndromes. A matter of life and death. Acta
Haematol 2004; 111(1–2):78–99.
5. Kerbauy DB, Deeg HJ. Apoptosis and antiapoptotic mechanisms in the progression
of myelodysplastic syndrome. Exp Hematol 2007; 35(11):1739–1746.
6. Maratheftis CI, Andreakos E, Moutsopoulos HM, et al. Toll-like receptor-4 is up-
regulated in hematopoietic progenitor cells and contribute to increased apoptosis in
myelodysplastic syndromes. Clin Cancer Res 2007; 13(4):1154–1160.
7. Mundle SD, Mativi BY, Bagai K, et al. Spontaneous down-regulation of Fas-associated
phosphatase-1 may contribute to excessive apoptosis in myelodysplastic marrows. Int
J Hematol 1999; 708(2):83–90.
8. Gupta P, Niehans GA, LeRoy SC, et al. Fas ligand expression in the bone marrow
in myelodysplastic syndtomes correlates with FAB subtype and anemia, and predicts
survival. Leukemia 1999; 13(1):44–53.
9. Deeg HJ, Beckham C, Loken MR, et al. Negative regulators of hemopoiesis and
stroma function in patients with myelodysplastic syndrome. Leuk Lymphoma 2000;
37(3–4):405–414.
10. Stifter G, Heiss S, Gastl G, et al. Over-expression of tumor necrosis factor-alfa in
bone marrow biopsies from patients with myelodysplastic syndromes. Eur J Hematol
2005; 75(6):485–491.
122 Yoshida
11. Zang DY, Goodwin RG, Loken MR, et al. Expression of tumor necrosis factor-related
apoptosis-inducing ligand, Apo-2L and its receptors in myelodysplastic syndrome:
Effects on in vitro hemopoieis. Blood 2001; 98(10):3058–3065.
12. Plasilova M, Zivny J, Jelinek J, et al. TRAIL (Apo2L) suppresses growth of primary
human leukemia and myelodysplasia progenitors. Leukemia 2002; 16(1):67–73.
13. Boudard D, Sordet O, Vasselon C, et al. Expression and activity of caspases 1 and 3
in myelodysplastic sydndromes. Leukemia 2000; 14(12):2045–2051.
14. Boudard D, Sordet O, Piselli S, et al. Increased caspase-3 activity in refractory ane-
mias: Lack of evidence for Fas pathway implication. Leukemia 2002; 16(11):2343–
2345.
15. Bouscary D, Chen YL, Guesnu M, et al. Activity of caspase-3/CPP32 enzyme is
increased in “early stage” myelodysplastic syndromes with excessive apoptosis, but
caspase inhibition does not enhance colony formation in vitro. Exp Hematol 2000;
28(7):784–791.
16. Mattes TW, Meyer G, Samii K, et al. Increased apoptosis in acquired sideroblastic
anaemia. Br J Haematol 2000; 111(3):843–852.
17. Economopoulou C, Pappa V, Kontsioti F, et al. Analysis of apoptosis regulatory genes
expression in the bone marrow (BM) of adult de novo myelodysplastic syndromes
(MDS). Leuk Res 2008; 32(1):61–69.
18. Carvalho G, Fabre C, Braun T, et al. Inhibition of NEMO. The regulatory subunit
of the IKK complex, induces apoptosis in high-risk myelodysplastic syndrome and
acute myeloid leukemia. Oncogene 2007; 26(16):2299–2307.
19. Braun T, Carvalho G, Coquelle A, et al. NF-κB constitutes a potential therapeutic
targets in high-risk myelodysplastic syndrome. Blood 2006; 107(3):1156–1165.
20. Kerbauy DM, Lesnikov V, Abbasi N, et al. NF-κB and FLIP in arsenic trioxide (ATO)-
induced apoptosis in myelodysplastic syndromes (MDS). Blood 2005; 106(12):3917–
3925.
21. Fabre C, et al. NF-κB inhibition sensitizes to starvation-induced cell death in
high-risk myelodysplastic syndrome and acute myeloid leukemia. Oncogene 2007;
26(28):4071–4083.
22. Benesch M, Platzbecker U, Ward J, et al. Expression of FLIPlong and FLIPshort in bone
marrow mononuclear and CD34+ cells in patients with myelodysplastic syndrome:
Correlation with apoptosis. Leukemia 2003; 17(12):2460–2466.
23. de Melo Campos P, Traina F, Duarte da SSA, et al. Reduced expression of FLIPshort
in bone marrow of low risk myelodysplastic syndrome. Leuk Res 2007; 31(6):853–
857.
24. Navas TA, Mohindru M, Estes M, et al. Inhibition of overactivated p38 MAPK
can restore hematopoiesis in myelodysplastic syndrome progenitors. Blood 2006;
108(13):4170–4177.
25. Katsoulidis E, Li Y, Yoon P, et al. Role of the p38 mitogen-activated protein kinase
pathway in cytokine-mediated hematopoietic suppression in myelodysplastic syn-
dromes. Cancer Res 2005; 65(19):9029–9037
26. Zhou L, Opalinska J, Verma A. p38 MAP kinase regulates stem cell apoptosis in
human hematopoietic failure. Cell Cycle 2007; 6(5):534–537.
27. Boudard D, Vasselon C, Berthéas MF, et al. Expression and prognostic significance of
Bcl-2 family proteins in myelodysplastic syndromes. Am J Hematol 2002; 70(2):115–
125.
28. Gianelli U, Fracchiolla NS, Cortelezzi A, et al. Survivin expression in “low-risk” and
“high-risk” myelodysplastic symdromes. Ann Hematol 2007; 86(3):185–189.
29. Invernizzi R, Travaglino E, Benatti C, et al. Survivin expression, apoptosis and prolif-
eration in chronic myelomonocytic leukemia. Eur J Haematol 2006; 76(6):494–501.
30. Tehranchi R, Fadeel B, Frosblom AM, et al. Granulocyte colony-stimulating factor
inhibits spontaneous cytochrome c release and mitochondria-dependent apoptosis
of myelodysplastic syndrome hematopoietic progenitors. Blood 2003; 101(3):1080–
1086.
31. Aizawa S, Hiramoto M, Hoshi H, et al. Establishment of stromal cell line from an
MDS patient which induced an apoptotic change in hematolopoietic and leukemic
cells in vitro. Exp Hematol 2000; 28(2):148–155.
32. Tauro S, Hepburn MD, Peddie CM, et al. Functional disturbance of marrow stromal
microenvironment in the myelodysplastic syndromes. Leukemia 2002; 16(5):785–
790.
33. Flores-Figueroa E, Gutiérrez-Espı́ndola G, Montesinos JJ, et al. In vitro characteri-
zation of hematopoietic microenvironment cells from patients with myelodysplastic
syndromes. Leuk Res 2002; 26(7):677–686.
34. Parker JE, Mufti GJ, Rasoo F, et al. The role of apoptosis, proliferation, and the
Bcl-2-related proteins in the myelodysplastic syndromes and acute myeloid leukemia
secondary to MDS. Blood 2000; 96(12):3932–3938.
35. Lin CW, Manshouri T, Jilani I, et al. Proliferation and apoptosis in acute and chronic
leukemias and myelodysplastic syndrome. Leuk Res 2002; 26(6):551–559.
36. Shimazaki K, Ohshima K, Suzumiya J, et al. Evaluation of apoptosis as a prognostic
factor in myelodysplastic syndromes. Br J Haematol 2000; 110(3):584–590.
37. Mundle SD, Mativi BY, Cartlidge JD, et al. Signal antonomy unique to myelodys-
plastic marrows correlates with altered expression of E2F1. Br J Haematol 2000;
109(2):376–381.
38. Huh YO, Jilani I, Estey E, et al. More cell death in refractory anemia with excess blasts
in transformation than in acute myeloid leukemia. Leukemia 2002; 16(11):2249–2252.
39. Nyåkern M, Tazzari PL, Finelli C, et al. Frequency elevation of Akt kinase phspho-
rylation in blood marrow and peripheral blood mononuclear cells from high-risk
myelodysplastic syndrome patients. Leukemia 2006; 20(2):230–238.
40. Follo MY, Mongiorgi S, Bosi C, et al. The Akt/mammalian target of rapamycin
signal transduction pathway is activated in high-risk myelodysplastic syndromes and
influences cell survival and proliferation. Cancer Res 2007; 67(9):4287–4294.
41. Bacher U, Haferlach T, Kern W, et al. A comparative study of molecular mutations
in 381 patients with myelodysplastic syndrome and 4130 patients with acute myeloid
leukemia. Haematologica 2007; 92(6):744–752.
42. Hofmann WK, Takeuchi S, Takeuchi N, et al. Comparative analysis of hypermethyla-
tion of cell cycle control and DNA-mismatch repair genes in low-density and CD34+
bone marrow cells from patients with myelodysplastic syndrome. Leuk Res 2006;
30(11):1347–1353.
43. Oki Y, Issa JJP. Treatment options in advanced myelodysplastic syndrome, with
emphasis on epigenetic therapy. Int J Hematol 2007; 86(4):306–314.
44. Hopfer O, Komoro M, Koehler IS, et al. DNA methylation profiling of myelodys-
plastic syndrome progenitor cells during in vitro lineage-specific differentiation. Exp
Hematol 2007; 35(5):712–723.
124 Yoshida
45. Bishton M, Kinealy M, Johnstone R, et al. Epigenetic targets in hematological malig-

nancies: Combination therapies with HDACis and demethylating agents. Expert Rev
Anticancer Ther 2007; 7(10):1439–1449.
46. Pecci A, Travaglino E, Klersy C, et al. Apoptosis in relation to CD34 antigen expres-
sion in normal and myelodysplastic marrow. Acta Haematol 2003; 109(1):29–34.
47. Shetty V, Hussaini S, Broady-Robinson L, et al. Intramedullary apoptosis of
hematopoietic cells in myelodysplastic syndrome patients can be massive: Apop-
totic cells recovered from high-density fraction of bone marrow aspirates. Blood
2000; 96(4):1388–1392.
48. Claessens YE, Bouscary D, Dupont JM, et al. In vitro proliferation and differentiation
of erythroid progenitors from patients with myelodysplastic syndromes: Evidence for
Fas-dependent apoptosis. Blood 2002; 99(5):1594–1601.
49. Tsoplou P, Kouraklis-Symeonidas A, Thanopoulou E, et al. Apoptosis in patients with
myelodysplastic syndromes: Differential involvement of marrow cells in ‘good’ versus
‘poor’ prognosis patients and correlation with apoptosis-related genes. Leukemia
1999; 13(10):1554–1563.
50. Berger G, Hunault-Berger M, Rachieru P, et al. Increased apoptosis in mononucleated
but not in CD34+ cells in blastic forms of myelodysplastic syndromes. Haematol J
2001; 2(2):87–96.
51. Mundle S, Venugopal P, Shetty V, et al. The relative extent and propensity of CD34+
vs. CD34- cells to undergo apoptosis in myelodysplastic marrows. Int J Hematol
1999; 69(3):152–159.
52. Sloand EM, Kim S, Fuhrer M, et al. Fas-mediated apoptosis is important in regulating
cell replication and death in trisomy 8 hematopoietic cells but not in cells with other
cytogenetic abnormalities. Blood 2002; 100(13):4427–4432.
53. Sloand EM, Pfannes L, Chen G, et al. CD34 cells from patients with trisomy
8 myelodysplastic syndromes (MDS) express early apoptotic markers but avoid
programmed cell death by up-regulation of antiapoptotic proteins. Blood 2007;
109(6):2399–2405.
54. Span LFR, Ruuten E, Gemmink A, et al. Bone marrow mononuclear cells of MDS
patients are characterized by in vitro proliferation and increased apoptosis indepen-
dently of stromal interactions. Leuk Res 2007; 31(12):1659–1667.
55. Li X, Bryant CE, Deeg HJ. Simultaneous demonstration of clonal chromosome abnor-
malities and apoptosis in individual marrow cells in myelodysplastic syndrome. Int J
Hematol 2004; 80(2):140–145.
abnormalities. Blood 2004; 104(13):4210–4218.
57. Tehranchi R, Fadeel B, Schmidt-Menge J, et al. Antiapoptotic role of growth fac-
tors in the myelodysplastic syndrome: Concordance between in vitro and in vivo
observations. Clin Cancer Res 2005; 11(17):6291–6299.
58. Deeg HJ, Gotlib J, Beckham C, et al. Soluble TNF receptor fusion protein (ethaner-
cept) for the treatment of myelodysplastic syndrome: a pilot study. Leukemia 2002;
16(2):162–164.
59. Maciejewski JP, Resitano AM, Sloand EM, et al. A pilot study of the recombinant
soluble human tumour necrosis factor receptor (p75)-Fc fusion protein in patients
with myelodysplastic syndrome. Br J Haematol 2002; 117(1):119–126.
60. Stasi R, Amadori S, Newland AC, et al. Infliximab chimeric antitumor necrosis factor-
a monoclonal antibody a potential treatment for myelodysplastic syndromes. Leuk
Lymphoma 2005; 46(5):509–516.
61. List A, Dewald G, Bennett JM, et al. Lenalidomide in the myelodysplastic syndrome
with chromosome 5q deletion. N Engl J Med 2006; 355(14):1456–1465.
62. Raza A, Reeves JA, Feldman EJ, et al. Phase II study of lenalidomide in transfusion-
dependent, low- and intermediate-1-risk myelodysplastic syndromes with karyotypes
other than deletion 5q. Blood 2008; 111(1):86–93.
6
The Role of Mitochondria in MDS
Norbert Gattermann
Department of Hematology, Oncology, and Clinical Immunology,
Heinrich-Heine-University, Düsseldorf, Germany
INTRODUCTION
A role for mitochondria in the pathophysiology of myelodysplastic syndrome
(MDS) has been suspected for several years. However, it has been difficult to tell
whether any of the observed mitochondrial changes in hematopoietic tissues from
MDS patients are closely related to the primary cause of these syndromes, or
merely represent participation of mitochondria in the apoptotic activity of MDS
bone marrow cells (see chap. 5).
Recent findings support the hypothesis that primary mitochondrial defects
can contribute to the phenotype of myelodysplasia. This chapter summarizes the
relevant findings and indicates how mitochondrial dysfunction may at least partly
explain several features of MDS pathology.
FINDINGS ON LIGHT AND ELECTRON MICROSCOPY

The presence of ringed sideroblasts is a common morphological change in MDS
and may thus be closely connected with a basic pathomechanism of myelodys-
plasia. Elucidating the cause of mitochondrial iron overload demonstrated as the
ringed sideroblast phenotype should therefore shed some light on other riddles of
MDS, too.
Ringed sideroblasts are not only confined to RARS (according to the FAB
classification) but are also found in refractory anemia (RA), refractory anemia
with excess of blasts (RAEB), and even in some cases of RAEB-t. Allan Jacobs
and David Bowen in the United Kingdom (1) stated that in RA “the number of
127
128 Gattermann
erythroblasts with ringed siderotic granules may vary from 1% to 90% and there
is no clear demarcation between ‘sideroblastic’ and ‘nonsideroblastic’ cases”.
Claude Sultan’s group in Paris (2), examined the prevalence and distribution of
ringed sideroblasts in 133 patients with primary MDS by using light microscopy.
Ringed sideroblasts ranging from 1% to 86% of cells were found in 57% of cases.
Among 46 patients with ⬍5% blasts in the bone marrow (corresponding to FAB
categories RA and RARS), a great majority (87%) had more than 20% ringed
sideroblasts. Among 65 patients with RAEB, ringed sideroblasts were present in
40%. One-fifth of the patients with RAEB had more than 20% ringed sideroblasts.
Among 22 patients with RAEB-t, seven had ringed sideroblasts. Only two patients
with RAEB-t had more than 20% ringed sideroblasts.
Electron microscopy is more sensitive than light microscopy at detecting
mitochondrial iron overload as well as other mitochondrial changes. In the 1970s,
Jorge Maldonado at Mayo Clinic and his colleagues (3) studied the erythrocytic
line in refractory anemia and myelomonocytic leukemia and found iron overload,
including the presence of large numbers of definitely pathologic sideroblasts, in all
their patients with preleukemia. There were iron deposits in a significant number of
the mitochondria and in a large number of the normoblasts. Often, but not always,
the presence of iron in the mitochondria was accompanied by degenerative changes
consisting of swelling, vacuolization, and other signs of internal disarray such as
rupture or separation of the cristae and formation of myelin figures.
Sakura and colleagues in Japan (4) studied the ultrastructural abnormalities
of erythroblasts in 30 patients with RA. Iron-laden mitochondria were found in
erythroblasts of intermediate stage maturation in 30% of the patients, and the
incidence of erythroblasts with this abnormality in each patient ranged from 0%
to 28.6% (mean ± SD = 4.3 ± 8%). Cohen et al. (5) used transmission elec-
tron microscopy to examine bone marrow aspirates from 26 patients with MDS,
including three patients with RARS. Iron deposits in the mitochondria were often
seen, accompanied by marked alterations of the mitochondrial structure. Arjan van
de Loosdrecht in Amsterdam and his colleagues (6) also investigated the ultra-
structural characteristics of erythroblasts in MDS. Among 22 patients, only two
had been diagnosed as sideroblastic anemia (RARS). Nevertheless, 16 out of 22
patients (73%) showed iron-laden mitochondria. In 55% of the cases, the mitochon-
dria were enlarged with or without disruption of internal cristae and/or mitochon-
drial membranes, which was significantly associated with accumulation of iron. In
all patients, erythroblasts showed extensive cytoplasmic vacuolization. The latter
finding is also characteristic of erythroblasts in patients with Pearson syndrome,
a congenital disorder giving rise to sideroblastic anemia (among other features).
Pearson syndrome is caused by large deletions of mitochondrial DNA (7).
Recently, E.J. Houwerzijl and colleagues in Groningen (8) found that ery-
throid precursors from patients with low-risk MDS display ultrastructural features
of autophagy. Autophagy is a potent degradation mechanism that can turn over
cytosolic proteins and eliminate entire organelles. In MDS, autophagy may be
increased in an attempt to remove defective and iron-laden mitochondria. At the
The Role of Mitochondria in MDS 129
same time, autophagy may support erythroblast survival because, under condi-
tions of reduced nutrient availability, this evolutionary process of catabolism of
intracellular organelles generates energy. However, prolonged autophagy can lead
to nonapoptotic, type II programmed cell death. In myelodysplastic syndromes,
autophagy may be an indicator of disrupted nutrient and energy metabolism, pos-
sibly caused by mitochondrial dysfunction.
MITOCHONDRIAL DYSFUNCTION IN MDS?

Mitochondria are often called the “powerhouse” of the cell because ATP regener-
ation by the mitochondrial respiratory chain (RS) is the most important function
of these organelles. First evidence of RS dysfunction in the bone marrow of MDS
patients was provided by Aoki (9), who examined several mitochondrial enzymes
in 61 patients with primary acquired sideroblastic anemia. Cytochrome c oxi-
dase (COX) and oligomycin-sensitive ATPase, both components of the respiratory
chain, had reduced activity compared with controls, whereas citrate synthase, an
enzyme of the mitochondrial matrix, was not impaired. The reduced activity of
COX and oligomycin-sensitive ATPase was not secondary to excess mitochon-
drial iron because enzyme measurements were performed using the patients’ gran-
ulocytes, which do not show mitochondrial iron loading. Every case of primary
acquired sideroblastic anemia had decreased COX activity in mature granulocytes.
Matthes et al. observed decreased mitochondrial membrane potential in erythrob-
lasts of patients with RARS (10), but later found no biochemical evidence of COX
deficiency in bone marrow homogenates from five patients with RARS (11).
David Bowen and Clare Peddie (12) measured oxygen consumption by
peripheral blood mononuclear cells (MNCs) from patients with MDS and healthy
controls. Oxygen consumption by MNCs from MDS patients was significantly
decreased, pointing to a defect in mitochondrial respiration. James Thompson in
Philadelphia and his colleagues (13) recently found greater sensitivity of MDS
progenitors or their progeny to the nonphysiologic oxygen tensions routinely used
in vitro. Taking into consideration the accumulating evidence for a low oxygen
tension in the hematopoietic stem cell (HSC) niche and a rising gradient of oxygen
tension during subsequent differentiation, the authors hypothesized that MDS
results from a lesion that causes apoptosis only at oxygen tensions higher than
that of the hematopoietic stem cell niche. A primary defect of the mitochondrial
respiratory chain, leading to increased ROS production, would nicely explain the
findings of Thompson and coworkers. Indicators of increased oxidative stress have
been found in the plasma (14), peripheral blood cells (15), and bone marrow cells
(16,17) of patients with MDS (18).
POSSIBLE CAUSES OF MITOCHONDRIAL DEFECTS

IN MYELODYSPLASIA
Mitochondrial dysfunction can be caused by mutations of mitochondrial DNA
(mtDNA) as several subunits of the respiratory chain are encoded in the
130 Gattermann
mitochondrial genome. Accumulation of mtDNA mutations is pivotal to the mito-

chondrial theory of ageing, which could therefore offer an explanation for the
well-known age distribution of MDS patients. The possible role of mtDNA muta-
tions is considered in more detail in subsequent sections of this chapter. However,
if mitochondrial defects are postulated to contribute to MDS pathology, it must
be emphasized that they are not necessarily related to the mitochondrial genome.
Nuclear encoded mitochondrial defects can also produce an MDS-like picture.
A group working with zebrafish found that loss of HSPA9B recapitulates
the ineffective hematopoiesis of the MDS (19). HSPA9B, which is encoded in
the critical deleted region of chromosome 5, is a highly conserved, ubiquitously
expressed mitochondrial chaperone. It has been implicated in a number of cellular
functions, but most prominently as a chaperone for the transport, folding, and
assembly of mitochondrial matrix proteins. The zebrafish mutants of HSPA9B
were anemic at the onset of circulation and showed significantly delayed devel-
opment of red blood cells. There appeared to be an arrest at the erythroblast
stage, followed by apoptosis. There was a reduction in erythrocytes and also in
myeloperoxidase-positive granulocytes, as well as hematopoietic progenitors in
the kidneys. Interestingly, haploid insufficieny increased apoptosis specifically
in adult hematopoietic progenitors. Craven and colleagues concluded that their
analysis of the zebrafish mutant indicated that a primary defect in a mitochondrial
chaperone is sufficient to produce characteristic symptoms of the MDS blood
disorders and implicates mitochondrial dysfunction in the pathogenesis of these
syndromes.
The importance of HSPA9B was confirmed by Chen and colleagues (20)
in primary human CD34+ hematopoietic cells, which were purified from cord
blood samples and infected with a shRNA expressing lentivirus, thereby reducing
HSPA9B expression. HSPA9B knockdown cultures showed reduced cell numbers,
an increase in apoptosis, and a decrease in the number of cells entering S-phase.
Furthermore, there was a decreased number of glycophorin A–positive cells, con-
comitant with retention of CD34+ expression on cells. A delay in erythroid and
myeloid differentiation was confirmed by cell morphology. Lentiviral knockdown
of the murine orthologue of HSPA9B (Hspa9a) in transduction/transplantation
experiments resulted in a loss of transduced cells over 2 months, implicating a cell
intrinsic defect when Hspa9a levels are reduced in vivo. Collectively, the results
implied that reduced HSPA9B expression in human CD34+ progenitor cells leads
to abnormal proliferation, increased apoptosis, and altered differentiation, key
features of ineffective hematopoiesis, in a dose-dependent manner.
Another interesting link between impaired mitochondrial function and dis-
turbed erythropoiesis was recently established by Stuart Orkin’s group in Boston
(21). These workers examined the role that the retinoblastoma gene product (Rb),
a central regulator of the cell cycle, plays in erythropoiesis. In a mouse model,
deletion of Rb in the erythroid compartment resulted in an anemia due to inef-
fective erythropoiesis. A partial block in differentiation occurred at the transition
from early-to-late erythroblasts. Unexpectedly, in addition to a failure to exit the
cell cycle, mitochondrial biogenesis failed to be upregulated at this transition

point. The disturbance in mitochondrial activities was established through gene
expression analysis, measures of respiratory function, as well as measurements
of mtDNA content in sorted cell populations. The failure to upregulate mito-
chondrial biogenesis appeared to contribute to the block in differentiation. The
links between erythroid differentiation and mitochondrial function were validated
by chemical inhibition of mitochondrial biogenesis and knockdown of the mito-
chondrial transcriptional coactivator PGC-1-beta in cultured erythroid cells. The
ineffective erythropoiesis seen in erythroid cells lacking Rb closely resembled the
features of the RA seen in the context of MDS.
Even exogenous causes of mitochondrial dysfunction may lead to an MDS-
like picture. This is suggested by chloramphenicol toxicity. The antibiotic not
only inhibits bacterial but also mitochondrial protein synthesis, via its direct
action on the large ribosomal subunit of the organelle, thereby impairing the
synthesis of respiratory-chain subunits encoded by the mitochondrial genome.
High-dose chloramphenicol also acts as an inhibitor of the complex I segment (i.e.,
NADH dehydrogenase) of the respiratory chain (22). Prolonged administration
of chloramphenicol causes ineffective hematopoiesis with severe bone marrow
dysplasia, sometimes including the formation of ringed sideroblasts (23).
Copper deficiency is another exogenous factor that can mimick MDS
(24,25). Copper is an essential component of COX, that is, complex IV of the mito-
chondrial respiratory chain. Therefore, copper deficiency can cause impairment
of RS function (25). In patients who developed copper deficiency after prolonged
parenteral nutrition lacking trace elements, a moderate number of ringed siderob-
lasts has been observed. The anemia of copper deficiency can be accompanied by
neutropenia with an absence of late myeloid forms in the bone marrow. Typically,
erythroid and myeloid precursors in the marrow are vacuolated. Surprisingly,
in very severe copper deficiency in the experimental animal, mitochondria
did not show iron overload (26). Iron remained in the cytoplasm, presenting
as scattered ferritin molecules as well as ferritin molecules accumulating in
siderosomes. Apparently, mitochondria were unable to import much iron through
the inner mitochondrial membrane, probably due to insufficient membrane
potential.
The above-mentioned findings emphasize that mitochondrial defects can
result from a variety of different causes. Since mitochondria are the site of heme
synthesis and thus play a pivotal role in the metabolism of erythroid precursors,
any major disturbance of mitochondrial function has the potential to produce
ineffective erythropoiesis.
MITOCHONDRIAL THEORY OF AGEING, APPLIED

TO HEMATOPOIETIC STEM CELLS
Mitochondrial dysfunction in elderly MDS patients is compatible with the mito-
chondrial or free-radical theory of ageing (27,28). The basis of this theory is
132 Gattermann
age-dependent accumulation of oxidative damage. The most important argument

in favor of the oxygen free radical theory of ageing is the elongation of life span by
a low-calory diet, which is thought to be mediated through diminished production
of reactive oxygen species (ROS) (29,30). Mitochondria are the main source of
ROS production and can also be expected to suffer most of the oxidative dam-
age, since short-lived radical molecules will not diffuse very far from their site
of origin. The free-radical theory of ageing must take into account that damaged
mitochondrial lipids, proteins, and carbohydrates will be rejuvenated by normal
mitochondrial turnover (31,32). However, the recycling machinery may not be
perfect.
A component of mitochondria which is turned over but is still prone to
damage accumulation is mtDNA. Its proximity to the cell’s major source of free
radicals renders it more susceptible to damage than nuclear DNA. Drastic oxidation
of mtDNA molecules probably leads to their complete degradation, which may
trigger cell death. Damage to mtDNA that results in mtDNA mutations may be
even more relevant to the ageing process. Such mtDNA damage is detectable
by repair systems only during the short period of nucleotide mismatch. Once
damaged mtDNA has been replicated, the error will be propagated. The number
of mtDNA molecules carrying somatic mutations may steadily accumulate with
age, with deleterious effects on respiratory chain function, and mitochondrial
energy production.
Evidence is accumulating that, as we grow older, the mitochondrial genome
becomes mutated in the majority of our cells (33–40). This is not necessarily
related to occupational hazards or other environmental factors. Rather, it appears
to be a natural consequence of mitochondrial energy production. Reactive oxygen
species damage mtDNA, which is not protected by histones (41,42). In addition,
DNA repair capacity in the mitochondria is less efficient than in the nucleus.
Together, this results in a mutation rate 10–20-fold higher than in nuclear DNA
(43–46). Since all 13 protein genes on the small (16,569 kb) mitochondrial genome
encode subunits of the mitochondrial respiratory chain, pathogenic mtDNA muta-
tions in these protein genes lead to impaired cellular respiration if the proportion
of mutant mtDNA molecules in the cell exceeds a certain threshold (47). Muta-
tions in mitochondrial tRNAs and ribosomal RNAs, which are also part of the
mitochondrial genome, can have similar effects by interfering with the expression
of the mitochondrial-encoded protein genes.
Which cells acquire mutations of mtDNA? Apparently, no cell type or tissue
is spared. However, in highly proliferative tissues such as intestinal mucosa or bone
marrow, mtDNA mutations occurring in differentiated cells are swiftly eliminated,
because these cells have a short life span. In order to persist in a proliferative tissue,
mtDNA mutations must occur in stem cells. Neal Young and his group at the NIH
looked at CD34+ cell-derived colonies from the bone marrow and found marked
mtDNA sequence heterogeneity in adult bone marrows. Twenty-five percent of
adult probands, with an average age in the low fourties, had developed such
sequence heterogeneity through acquired mutations of mtDNA (48). Single-cell
analysis of CD34+ cells yielded an even higher mtDNA heterogeneity, with around
38% of cells showing deviation from corresponding aggregate sequences (49). In
colonies derived from cord blood CD34+ cells, no mtDNA mutations were found
(48).
A recent study showed that age-dependent accumulation of mtDNA muta-
tions in murine hematopoietic stem cells was modulated by the nuclear genetic
background and was not associated with increase in ROS or senescence (50). This
suggests that the accumulation of mtDNA mutations may not be as straightfor-
ward as first thought to be and that an individual’s nuclear background may affect
their susceptibility to the accumulation of mtDNA mutations and the resulting
phenotype (51). At present, the question, whether mitochondrial mutations cause
mammalian aging, or are merely correlated with it, is an area of intense debate
(52–54).
AGE-RELATED ACCUMULATION OF MTDNA MUTATIONS:

IMPACT ON BONE MARROW FUNCTION?
The ability of somatic mtDNA mutations to expand intracellularly, as well as their
apparently high incidence, reinforces the possibility that these mutations may be
actively involved in various physiologicial processes such as ageing and degen-
erative disease. Age-related accumulation of somatic mtDNA rearrangements has
been demonstrated in many tissues (33), including the bone marrow (48,55). If
somatic mtDNA mutations in cardiac muscle contribute to age-related heart fail-
ure, and hematopoietic stem cells also accumulate mtDNA mutations, why is
age-related bone marrow failure not a common cause of death?
Evidence has been presented that hematopoietic stem cells (HSC) are subject
to ageing (56–58). Although they have the capacity to maintain hematopoiesis at
steady state, HSC populations in aged individuals may not always have the reserves
necessary to respond to replicative stresses. The number of HSC and progenitor
cells does not significantly decline in old age, and in some mouse strains, numbers
actually increase. The important effects of ageing therefore may be on HSC quality
rather than quantity. It is tempting to speculate that the quality of hematopoietic
stem cells deteriorates as they acquire and amplify mtDNA mutations.
During mitosis, mtDNA mutations are subject to random segregation and
will therefore be concentrated in certain stem cells. The fate of mtDNA mutations
is then probably determined by their severity. Drastic mutations such as large
deletions may not be compatible with stem cell survival and will thus be eliminated
by death of the affected stem cell(s) (55). Less severe mtDNA mutations may be
compatible with stem cell survival and will thus persist. However, stem cells
harboring such mutations may be of inferior quality. Their differentiated progeny,
being more dependent on mitochondrial activity, may suffer maturational defects
and functional impairment and may die from apoptosis, thereby contributing to
ineffective hematopoiesis.
134 Gattermann
In the hematopoietic system, an age-related deterioration in the quality of

stem cells is probably compensated by increasing the recruitment of quiescent
stem cells into the proliferating pool, thereby preventing age-related bone marrow
failure. However, continuously increased proliferation in the HSC compartment
may, in the long term, increase the risk of malignant transformation.
The effects that accumulating mtDNA mutations may have in the
hematopoietic system are now being investigated in a mouse model of premature
ageing, using mice expressing defective mtDNA polymerase gamma (Polg)
(59,60). A defective Polg allele eliminates the proofreading activity of Polg,
and homozygous mutant mice rapidly accumulate mtDNA mutations. Starting at
9 months of age, mutant mice display multiple features of accelerated aging and
develop a progressive and ultimately lethal anemia. The anemia is macrocytic.
Despite the profound anemia, the marrow is hypercellular, displaying megakary-
ocytic dysplasia (paucity of cytoplasm and abnormal nuclear lobulation), and an
increase in immature MPO+ myeloid precursors (61). These features fulfil the
diagnostic criteria for a myeloid dysplasia without evidence of myeloproliferation.
In competitive marrow transplants, the mutant bone marrow cells showed a
competitive disadvantage, suggesting cell-intrinsic defects at the level of the
hematopoietic stem cell (61). It is not yet clear, how the massive accumulation
of mtDNA mutations in the Polg mouse models exert their unfavorable effects.
No increase in DNA, RNA, protein, or lipid markers of oxidative stress were
observed (60,62). Instead, induction of apoptosis, particularly in tissues with
rapid cellular turnover, may be the mechanism of aging in Polg mutator
mice.
As mentioned above, pathogenic mtDNA mutations usually create malfunc-
tion of mitochondrial respiration. Recognizing the lack of convincing evidence to
confirm the pathophysiological relation between respiration defects in hematopoi-
etic cells and expression of anemia, Inoue et al. (63) recently transplanted bone
marrow cells carrying pathogenic mtDNA with a large-scale deletion into normal
mice. This resulted in macrocytic anemia. The mice showed abnormalities of ery-
throid differentiation and a weak erythropoietic response to hematopoietic stress.
Age-related accumulation of the large 5-Kb “common deletion” of mtDNA has
already been demonstrated in the human bone marrow (55).
MTDNA MUTATIONS IN THE CONTEXT OF CLONAL BONE

MARROW DISORDERS
Despite their possible contribution to MDS pathology, mtDNA mutations cannot
be expected to be the sole cause of a MDS, or any other clonal bone marrow
disease, because they are unlikely to provide the cellular growth advantage that is
necessary for clonal expansion. Initially, any somatic mutation of mtDNA in the
bone marrow is confined to the small clone consisting of the affected stem cell
and its progeny. Therefore, in a polyclonal marrow, a particular mtDNA mutation
cannot do much harm to the hematopoietic system as a whole. However, a special
situation arises if a stem cell harboring mutant mtDNA incurs additional genomic
damage in the cell nucleus. Nuclear DNA mutations may confer an abnormal
cellular growth advantage, leading to clonal expansion and ultimately to a clonal
bone marrow disease. The phenotype of this disorder may be influenced by the
mutant mtDNA that probably existed in the stem cell prior to transformation. This
situation can be encountered in patients with MDS (64–68).
We determined the frequency and spectrum of somatic mtDNA mutations
in the bone marrow of patients with MDS (69). The analysis included 104 patients
with MDS (24 RA, 32 RARS, 34 RAEB, 7 RAEB-t, 7 CMML), 3 patients with
AML from MDS, and 36 patients with myeloproliferative disease (23 CML,
9 PV, 4 IMF). Mutation scanning was performed using heteroduplex analysis with
denaturing HPLC (dHPLC). The entire mitochondrial genome was amplified in 67
overlapping PCR fragments, carefully optimized regarding DNA melting profiles
(70). Abnormal dHPLC findings were confirmed by DNA sequencing.
Heteroplasmic mtDNA mutations, mostly transitions, were identified in 56%
of MDS and 44% of MPD patients. In MDS, mutation frequency increased with
age and more advanced disease. Mutational spectra showed no hotspots and were
similar in different types of MDS (Fig. 1). Heteroplasmic mutations generally did
not represent known polymorphisms, and about half of them affected conserved
amino acids or nucleotides. Mutations were less frequent in protein encoding
genes (50 per 106 base pairs) than other mitochondrial genes (transfer RNAs,
1
12S CyB
ND6
16S
FAB-type ND5
RA
ND1
RARS
RAEB
RAEB-t
AML
ND2 CMML ND4
ND4L
ND3
COX1
COX3
COX2 AP6
Figure 1 Distribution of acquired heteroplasmic mtDNA mutations in MDS patients. 1,

Origin of base numbering in control region; 12S, 12S ribosomal RNA; 16S, 16S ribosomal
RNA; ND1–6, NADH dehydrogenase subunits; COX1–3, cytochrome-c-oxidase subunits;
ATP6, ATP synthase F0 subunit 6 (the overlapping ATP8 gene is not labeled); CyB,
cytochrome b subunit (transfer RNA genes are not labeled). (see color insert)
136 Gattermann
ribosomal RNAs, and control region; about 80 per 106 base pairs). The study
thus yielded a high frequency of acquired, clonally expanded mtDNA mutations
in MDS. However, the functional importance of these mutations remains unclear,
considering that genotype correlates poorly with phenotype in mitochondrial
diseases. It is not surprising that no mutational hotspots were identified. All protein
genes in the mitochondrial genome code for respiratory-chain subunits, and the
mitochondrial tRNAs and rRNAs are also required for the synthesis of these
subunits.
Therefore, all pathogenic mtDNA mutations (in protein genes, tRNAs, or
rRNAs) have a final common pathway, namely, respiratory chain dysfunction,
irrespective of their location in the mtDNA. Accordingly, the absence of muta-
tional hotspots is not at variance with a possible role of mtDNA mutations in the
pathogenesis of MDS (or other diseases).
A search for mtDNA mutations in 10 cases of MDS, performed by Myung
Geun Shin and his colleagues in Neal Young’s Group at the National Institutes
of Health in Bethesda (71), revealed some homoplasmic sequence alterations
but, surprisingly, no heteroplasmic mutations. This may be due to the fact that
mutation scanning was based on direct DNA sequencing, which is less well suited
for detection of heteroplasmy.
As shown by our patients with MPD, clonally expanded mtDNA mutations
are not specific to MDS. The study therefore supports the idea that hematopoietic
stem cells in general suffer age-related damage to mtDNA. This concept is also
supported by findings obtained by several groups (72–74), who demonstrated
that among patients with different types of leukemia, about 40% to 60% show a
clonally expanded somatic mtDNA mutation. It seems likely that in most instances
such mtDNA mutations pre-exist in the leukemic stem cell and thus become, upon
clonal transformation, a clonal marker.
However, mtDNA mutations may also occur during growth of the leukemic
clone, at least those found in the noncoding D-loop of mtDNA (74,75). Further-
more, the occurrence of mtDNA mutations may be enhanced by cytotoxic drugs.
An analysis of mtDNA from 20 patients with chronic lymphocytic leukemia (CLL)
revealed that primary CLL cells from patients with prior chemotherapy had a sig-
nificantly higher frequency of heteroplasmic mtDNA mutations than did those
from untreated patients (76).
Since many mtDNA mutations identified in solid tumors or clonal bone
marrow disorders are silent mutations or mutations changing poorly conserved
nucleotides/amino acids (72,77), or simply represent sequencing errors (78), it is
inappropriate to claim that mtDNA mutations are universally important in carcino-
genesis or leukemogenesis. On the other hand, it is equally unjustified to claim that
all mtDNA mutations are irrelevant. Why should mutations at well-conserved sites
of mtDNA be innocent in tumor cells, while causing all sorts of neurological and
other disease phenotypes in nonmalignant cells (34,47)? Mitochondrial defects
may contribute to the phenotype of the MDS clone in various ways, as explained
in the following section.
PARTIAL EXPLANATION OF MDS FEATURES ON THE BASIS

OF MITOCHONDRIAL DYSFUNCTION
Sideroblastic Phenotype
Several examples (see above) illustrate that erythroid precursors can develop mito-
chondrial iron accumulation as a consequence of mitochondrial dysfunction. Of
particular interest, the rare congenital disorder called Pearson syndrome provides
a strong hint that mtDNA defects can be involved in the pathogenesis of the
sideroblastic phenotype. Pearson’s syndrome is characterized by permanent lactic
acidosis, pancreatic insufficiency, other metabolic derangements, and severe RA,
with moderate numbers of ringed sideroblasts in the bone marrow. The anemia
is usually accompanied by neutropenia and thrombocytopenia. The bone marrow
shows dysplastic changes, including prominent vacuolization of precursor cells.
Pearson’s syndrome is caused by large deletions of mtDNA (7).
Until now, the link between mtDNA damage and the sideroblastic pheno-
type in Pearson syndrome has remained elusive. We offer the following hypothesis,
explaining the sideroblastic phenotype as a consequence of mitochondrial respi-
ratory chain dysfunction (Fig. 2). During the final step of heme synthesis in the
Mitochondrion Cytoplasm Mitochondrion Cytoplasm
Fe3+
Ferro- Ferro-
Heme chelatase chelatase
Fe2+
O2
+ IV IV
2H + 1/2 Ο2
H+ H+
H2O c H O c
2
III H+ III
H+
e– Q e– Q
II II
e– e –
Complex I H+ Complex I H+
ADP +P ADP+P
ATP ATP
Normal RC defect
Figure 2 Pathogenesis of mitochondrial iron overload as a consequence of respiratory

chain dysfunction. A respiratory chain defect will slow down oxygen consumption and
will thus leave more oxygen than usual in the mitochondrial matrix. Therefore, some of
the imported Fe2+ will become oxidized. Fe3+ is rejected by ferrochelatase and will thus
accumulate in the mitochondrial matrix.
138 Gattermann
mitochondria, the enzyme ferrochelatase inserts iron into protoporphyrin IX. Fer-
rochelatase processes ferrous iron (Fe2+ ) but cannot utilize ferric iron (Fe3+ ) (79).
In ringed sideroblasts, mitochondrial iron accumulates as ferric iron (Fe3+ ) (80).
We believe that this is attributable to a failure of the respiratory chain (RC) to
effectively remove oxygen from the mitochondrial matrix. Oxygen consumption
in erythropoietic cells appears to be stimulated by uncoupling protein-2, which
is expressed during erythroid cell maturation (81,82). Oxygen consumption, and
the resulting low oxygen concentration in the mitochondrial matrix, would help
to keep iron in the reduced form. A respiratory chain defect will decrease O2
consumption and will thus increase O2 in the mitochondrial matrix. If iron, after
crossing the inner mitochondrial membrane as Fe2+ (83–85), becomes oxidized
(→Fe3+ ), it will be rejected by ferrochelase and will thus accumulate in the
mitochondrial matrix.
Previous explanations proposed for ringed sideroblast formation are inad-
equate. A defect in the heme synthetic pathway (protoporphyrin synthesis), still
suggested in some textbooks, is very unlikely because the end product of this path-
way, protoporphyrin IX, is elevated rather than reduced in idiopathic sideroblastic
anemia (86). Ferrochelatase is also unlikely to be the culprit. First, it has been
demonstrated that increased red cell protoporphyrin concentrations are not cor-
related with low ferrochelatase activities (87). Second, Steensma and colleagues
at Mayo Clinic (88) recently performed a candidate gene mutation analysis and
found no somatic missense mutations of the ferrochelatase gene and its promoter in
patients with acquired idiopathic sideroblastic anemia (AISA). These workers also
analyzed ABCB7 and PUS1, genes implicated in rare congential SA syndromes,
but again found no coding mutations in patients with acquired SA.
It has been shown that increased expression of mitochondrial ferritin (MtF)
is an early event in the development of mitochondrial iron accumulation (89,90).
This feature can be utilized for diagnosing sideroblastic anemia by flow cytom-
etry (91). However, the increase in MtF is probably a secondary phenomenon.
This is suggested by the association of MtF with cell types characterized by high
metabolic activity and oxygen consumption, which indicates a role in protecting
mitochondria from iron-dependent oxidative damage (92). Furthermore, overex-
pression of MtF caused cytosolic iron depletion (93), which is not observed in
sideroblastic anemia.
Mitochondrial dysfunction may inhibit iron utilization in erythroblasts even
in the absence of a sideroblastic phenotype. If respiratory chain activity is heavily
compromised, as in severe experimental copper deficiency, the membrane potential
of erythroblast mitochondria may be insufficient to support the extensive iron
import that is needed for heme synthesis. This may preclude the expression of the
sideroblastic phenotype.
Impaired Erythroid Maturation

Since 4 of the 8 steps of the heme synthetic pathway are located in the mitochon-
dria, heme synthesis may suffer if mitochondria are not working properly. Heme
synthesis is critically dependant on mitochondrial functions such as proper iron

handling. Impaired heme synthesis, in turn, has untoward effects on erythroid mat-
uration. Diverse inhibitors of heme synthesis induce apoptosis of human erythroid
progenitor cells (94), and the presence of apoptosis in ineffective erythropoiesis
may be secondary to impaired heme synthesis. Similarly, Osamu Nakajima in Yam-
agata and his colleagues (95) found that heme deficiency, caused by a disrupted
first step in the heme biosynthetic pathway, resulted in impaired differentiation in
the erythroid lineage. Another example relates to mitoferrin (Mfrn), which appears
to be the principal mitochondrial iron importer essential for heme biosynthesis in
vertebrate erythroblasts. George Shaw in Boston and his colleagues demonstrated
that Mfrn−/− hematopoietic cells were unable to undergo terminal erythroid
maturation (85).
As heme synthesis, together with globin chain synthesis, is the main biosyn-
thetic activity of erythroblasts, the latter are likely to be particularly vulnerable
to mitochondrial dysfunction. Mitochondrial defects may therefore provide an
answer to the question, why anemia is an early and usually the predominant
feature of MDS.
Granulocyte Dysfunction
The consequences of impaired heme synthesis may not be confined to erythroid
cells. Heme is an important prosthetic group of several enzymes, for example,
myeloperoxidase (MPO). Therefore, the unexplained finding of partial MPO defi-
ciency in MDS may be attributable to impaired mitochondrial production of heme.
Flavocytochrome b558, which shows reduced expression in neutrophils from
patients with MDS (96), is a heme derivative, too, and thus depends on mitochon-
drial heme synthesis. Like MPO deficiency, a deficiency of flavocytochrome b558
can compromise antimicrobial activity, because the molecule is a component of
the NADPH oxidase complex, which contributes to the respiratory burst.
Premature Death of Megakaryocytes

Apparently, mitochondria are involved in the premature death of megakaryocytes
(MK) in the bone marrow of MDS patients. Edo Vellenga in Groningen and
coworkers demonstrated, by ultrastructural analysis, that cell death in mature
and immature MK in MDS is not typical of apoptosis because it is a caspase-
independent process, described by these authors as “necrosis-like programmed
cell death (PCD)” (97,98). Guido Kroemer’s group in Paris recently confirmed
that immature MK in MDS die without caspase-3 activation (99). However, the
premature death of MDS MK was accompanied by the mitochondrial release of
cytochrome c, Smac/DIABLO and endonuclease G, a caspase-independent death
effector, as well as loss of the mitochondrial membrane potential and plasma
membrane phosphatidylserine exposure before definitive loss of viability. Thus,
a stereotyped pattern of mitochondrial alterations accompanied differentiation-
associated MK death in MDS. Emphasizing these mitochondrial processes in the
lethal event, Thorsten Braun and colleagues (Kroemer’s group), in contrast to
140 Gattermann
Vellenga’s group, classified the premature death of MK in MDS in the category

of “apoptosis-like cell death” (99).
Increased Apoptosis
An increased level of apoptosis in bone marrow cells is widely regarded as a char-
acteristic feature of low-risk MDS. Whether it is mainly driven by intrinsic defects
such as genetic and epigenetic changes, or by extrinsic signals from an altered bone
marrow microenvironment or an autoimmune response, remains controversial. At
least in part, increased apoptosis may be connected with mitochondrial dysfunc-
tion. Mitochondria are the switchboard of the apoptosis machinery (100) and
contribute to caspase activation, thereby triggering apoptosis. In low-risk MDS,
exacerbation of this pathway, even in the presence of elevated Epo concentrations,
has been reported to cause a high rate of apoptosis, ineffective erythropoiesis, and
increased phagocytosis (101–106). Several observations argue for an important
role of the Fas-L/Fas pathway in this situation (107,108). However, it is hard to
determine whether mitochondria simply collaborate with an activated Fas-L/Fas
apoptotic pathway in the usual way or, in case of a primary mitochondrial defect,
amplify Fas-dependent pro-apoptotic signals in a pathological manner.
Eva Hellström-Lindberg and her colleagues in Sweden observed that ery-
throid precursor cells from RA and RARS patients showed constitutive cytochrome
c release from mitochondria, with subsequent activation of downstream caspases
(105,109). While cytochrome c release is an indicator of mitochondria-dependent
apoptosis, it is not clear whether the primary defect is in the mitochondria.
Mitochondrial dysfunction can lead to necrotic and apoptotic cell death,
as detailed in chapter 5 (110). Apoptosis can be induced by specific respiratory
chain inhibitors (111) and by a lack of mtDNA gene expression (112). We found
significantly impaired mitochondrial gene expression in primary human CD34+
cells from MDS patients (113). Comparison with age-matched controls showed
that dysregulated mitochondrial gene expression in MDS goes beyond a simple
age-related effect. However, it is unclear whether the problem arises from primary
alterations in the mitochondria, such as mtDNA mutations, or from genetic or
epigenetic changes in the cell nucleus.
Generally, a higher level of apoptosis is observed in early (low-risk) MDS
than in the more advanced subtypes (114,115). This seems peculiar if mitochon-
drial defects are supposed to be more severe in advanced MDS. However, with
more severe mitochondrial dysfunction, genomic instability may come into play,
providing ample opportunity to acquire genetic alterations in the cell nucleus
which help to escape apoptotic control.
Genomic Instability
Mitochondrial defects may contribute to genomic instability. For example,
deficient mitochondrial ATP regeneration may promote chromosomal instabil-
ity. Since the mitotic spindle apparatus strongly depends on GTP- and ATP-
hydrolyzing motor proteins, cells with inadequate ATP supply may have difficulty
in correctly segregating their chromosomes during mitosis. Malfunction of the

spindle apparatus may favor aneuploidy. Notably, most karyotype anomalies in
MDS are characterized by losses or gains of whole chromosomes, or large dele-
tions, rather than specific translocations. Chromosomal instability seems to be
virtually absent in patients with pure sideroblastic anemia, for whom only a mild
respiratory chain defect is postulated which does not block the import of iron into
mitochondria. These patients have the lowest frequency of karyotype anomalies,
the lowest incidence of leukemic transformation, and the best survival among
MDS patients (116,117).
In case of more severe RC defects, impairment of ATP-dependent DNA
repair mechanisms may be another pathway contributing to genomic instability.
V. Van Duppen and colleagues in Leuven, Belgium, observed that CD34+ pro-
genitors from low-risk MDS patients were more susceptible to induced oxidative
damage and had reduced DNA repair compared to healthy control subjects (118).
Jankowska et al. recently found base excision repair dysfunction in a subgroup of
patients with MDS (119).
ATP Deficiency?
Besides causing disturbed mitochondrial iron metabolism and heme synthesis,
mitochondrial dysfunction may contribute to MDS pathogenesis via impaired
ATP synthesis. It is tempting to speculate that impaired energy metabolism may
be pertinent to the proliferation and function of bone marrow cells in MDS. There
is a hierarchy of ATP-consuming processes in mammalian cells (120). If ATP
supply is compromised, processes that are not essential for the immediate needs
of the cell will be given up before those that are more critical for ionic integrity
(and thus for immediate survival). Accordingly, cells shut down their metabolic
activities in a certain order. Pathways of macromolecule synthesis, that is, protein
and RNA/DNA synthesis, are most sensitive to energy supply. Sunitha Wickra-
masinghe in London and colleagues (121) observed an arrest of DNA synthesis,
a depression of RNA synthesis, and a marked depression of protein synthesis in
erythroblasts affected by mitochondrial iron overload. Ramin Tehranchi and his
colleagues in Stockholm found that some MDS patients had decreased levels of
mitochondrial ATP production in mononuclear bone marrow cells; however, there
was no clear difference between the patient and control groups (90). The question
of impaired ATP regeneration in MDS requires further investigation.
Megaloblastic Changes
It is still unclear why bone marrow cells in MDS often show megaloblastic
changes. We propose that these changes may be attributable to impaired pyrimidine
nucleotide synthesis, caused by mitochondrial dysfunction. Dihydroorotate dehy-
drogenase (DHODH), an enzyme necessary for de novo pyrimidine nucleotide syn-
thesis, is located in the inner mitochondrial membrane, with its function depending
on interaction with the respiratory chain at the level of coenzyme Q (122). We
have shown that inhibition of the respiratory chain impairs pyrimidine synthesis
142 Gattermann
and thereby affects DNA precursor pool concentrations (123). The mitochondrial
dysfunction caused a general breakdown of nucleotide triphosphates, owing to
impaired ATP regeneration, and a selective decrease in pyrimidines, as a result of
coimpairment of DHODH. If pyrimidine nucleotides are deficient, the resulting
nucleotide imbalance may not only produce a megaloblastic phenotype but may
also cause a wide range of genetic events, due to aberrant DNA replication or
repair (124).
Autoimmunity
Successful immunosuppressive treatment in a proportion of patients with low-
risk MDS suggests that autoimmune mechanisms can contribute to MDS pathol-
ogy. However, the targets of this immune response are still unknown. In 1990,
a hydrophobic peptide of a mitochondrially encoded protein was identified as a
maternally transmitted histocompatibility antigen in mice (125). The peptide was
part of ND1, a subunit of NADH dehydrogenase (complex I of the respiratory
chain). The authors concluded that cells can display peptides derived from mito-
chondrially encoded proteins, and that such peptides can be histocompatibility
antigens. If this holds true in humans, it is conceivable that an MDS clone express-
ing a mutant subunit of the mitochondrial respiratory chain becomes the target of
an autoimmune response.
SYNOPSIS
Figure 3 summarizes how mitochondrial defects may fit into the picture of MDS
pathogenesis. With advancing age, hematopoietic stem cells accumulate muta-
tions of mtDNA as well as nuclear DNA, both of which can cause mitochondrial
dysfunction. Some of the affected stem cells will not be able to cope with their
mitochondrial damage and die, probably by apoptosis. This may contribute to age-
related decrease of the bone marrow reserve. Other stem cells harboring mutant
mitochondria will continue to proliferate, and their daughter cells may show dys-
function and morphological changes, but this is not MDS, since the bone marrow
is not clonal. It should be emphasized that mitochondrial defects per se are not suf-
ficient to produce a MDS, because they are unlikely to provide the cellular growth
advantage that is necessary for clonal expansion. As a precondition for establish-
ing a clonal marrow—and thus the clinical picture of MDS—a stem cell carrying
mutant mitochondria must gain a relative growth advantage through additional
mutations in the cell nucleus. However, mitochondrial defects may play an impor-
tant role in shaping the phenotype of the MDS clone, as exemplified by the siderob-
lastic phenotype. In addition, they may contribute to genomic instability, thereby
facilitating the transforming event that initiates clonal expansion and also promot-
ing further genomic changes that drive the clonal evolution toward leukemia.
When looking for early pathogenetic events in MDS, one should try to
identify changes which can explain two characteristic and widespread features
Ageing Depleted Clonal Clonal Leukemia

stem cells hematopoiesis evolution
Oncogene activation
Growth advantage
Accumalation Genomic instability Oncogene activation

of mitochondrial defects chromosomal changes
Figure 3 Model of MDS pathogenesis including mitochondrial pathology.
of the disease, namely, dysplasia and genomic instability. Mitochondrial defects

appear to be good candidates, considering their pleiotropic effects (Fig. 4).
THE SPECIAL CASE OF RARS-T

As mentioned above, an MDS clone harboring a mitochondrial defect requires
additional genomic changes in order to gain a growth advantage. RARS-T (refrac-
tory anemia with ringed sideroblasts associated with marked thrombocytosis) is
probably the first “MDS subtype” to reveal the cause of its growth advantage.
RARS-T is classified as “myelodysplastic/myeloproliferative syndrome, unclas-
sifiable” by the WHO. The diagnosis requires at least 15% ringed sideroblasts in
the bone marrow and a platelet count of at least 6 × 109 /L in the peripheral blood.
About 60% of patients with RARS-T carry the JAK2 V617 F mutation (126–
132), which is characteristic of myeloproliferative syndromes. JAK2 V617 F is
not related to the sideroblastic phenotype because the latter is absent in patients
with polycythemia vera, most of whom are positive for JAK2 V617 F. Instead,
the JAK2 mutation appears to be driving the clonal proliferation in RARS-T.
The disorder can therefore be regarded as a clonal bone marrow disease arising
in a hematopoietic stem cell which carries a mitochondrial defect causing the
sideroblastic phenotype. However, the riddle of RARS-T remains partly unsolved
until the exact nature of the mitochondrial defect has been elucidated.
144 Gattermann
Impaired protein Decreased mitochondrial

and DNA synthesis membrane potential
Impaired motoring Increased susceptibility

in the mitotic spindle to apoptosis
Reduced DNA repair
ATP Apoptosis
synthesis
Iron Pyrimidine
handling synthesis
Sideroblastic phenotype Impaired pyrimidine

denovo synthesis
Refractory anemia
Nucleotide imbalance,
Impaired heme synthesis
genomic instability
Impaired erythroid maturation
Megaloblastic changes
Figure 4 Effects of mitochondrial dysfunction in hematopoietic cells.
REFERENCES
1. Jacobs A, Bowen DT. Pathogenesis and evolution of refractory anaemia. In: Mufti
GJ, Galton DAG, eds. The Myelodysplastic Syndromes. Edinburgh: Churchill Liv-
ingstone, 1992:33–53.
2. Juneja SK, Imbert M, Sigaux S, et al. Prevalence and distribution of ringed
sideroblasts in primary myelodysplastic syndromes. J Clin Pathol 1983; 36:566–
569.
3. Maldonado JE, Maigne J, Lecoq D. Comparative electron-microscopic study of the
erythrocytic line in refractory anemia (preleukemia) and myelomonocytic leukemia.
Blood Cells 1976; 2:167–185.
4. Sakura T, Murakami H, Saitoh T, et al. Ultrastructural abnormalities of bone marrow
erythroblasts in refractory anemia. Ultrastruct Pathol 1998; 22:173–180.
5. Cohen AM, Alexandrova S, Bessler H, et al. Ultrastructural observations on bone
marrow cells of 26 patients with myelodysplastic syndromes. Leuk Lymphoma 1997;
27:165–172.
6. van de Loosdrecht AA, Brada SJL, Blom NR, et al. Mitochondrial disruption and
limited apoptosis of erythroblasts are associated with high risk myelodysplasia. An
ultrastructural analysis. Leuk Res 2001; 25:385–393.
7. Rötig A, Cormier V, Koll F, et al. Site-specific deletions of the mitochondrial
genome in the Pearson marrow-pancreas syndrome. Genomics 1991; 10:502–
504.
8. Houwerzijl E, Pol H-W, Blom N, et al. Erythroid precursors from patients with low-
risk myelodysplasia demonstrate ultrastructural features of autophagy. Blood 2007;
110(11, abstract #2444).
9. Aoki Y. Multiple enzymatic defects in mitochondria in hematological cells of patients
with primary sideroblastic anemia. J Clin Invest 1980; 66:43–49.
10. Matthes TW, Meyer G, Samii K, et al. Increased apoptosis in acquired sideroblastic
anaemia. Br J Haematol 2000; 111:843–852.
11. Matthes T, Rustin P, Trachsel H, et al. Different pathophysiological mechanisms
of intramitochondrial iron accumulation in acquired and congenital sideroblastic
anemia caused by mitochondrial DNA deletion. Eur J Haematol 2006; 77:169–174.
12. Bowen D, Peddie C. Mitochondrial oxygen consumption and ineffective
haematopoiesis in patients with myelodysplastic syndromes. Br J Haematol 2002;
18:345–346.
13. Thompson JE, Conlon JP, Yang X, et al. Enhanced growth of myelodysplastic
colonies in hypoxic conditions. Exp Hematol 2007; 35:21–31.
14. Fracchiolla NS, Catena FB, Novembrino C, et al. Possible association between reac-
tive oxygen metabolites and karyotypic abnormalities in myelodysplastic syndromes.
Haematologia (Budap) 2003; 88:594–596.
15. Ghoti H, Amer J, Winder A, et al. Oxidative stress in red blood cells, platelets and
polymorphonuclear leukocytes from patients with myelodysplastic syndromes. Eur
J Haematol 2007; 79:463–467.
16. Peddie C, Wolf CR, McLellan LI, et al. Oxidative DNA damage in CD34+ myelodys-
plastic cells is associated with intracellular redox changes and elevated plasma
tumour necrosis factor concentration. Br J Haematol 1997; 99:625–631.
17. Bowen D, Wang L, Frew M, et al. Antioxidant enzyme expression in myelodysplastic
and acute myeloid leukemia bone marrow: Further evidence of a pathogenetic role
for oxidative stress? Haematologica 2003; 88:1070–1072.
18. Farquhar MJ, Bowen DT. Oxidative stress and the myelodysplastic syndromes. Int J
Hematol 2003; 77:342–350.
19. Craven SE, French D, Weilan Y, et al. Loss of Hspa9b in zebrafish recapitulates the
ineffective hematopoiesis of the myelodysplastic syndromes. Blood 2005; 105:3528–
3534.
20. Chen TH, Walshauser M, Kambal A, et al. Reduced HSPA9B expression, a 5q31.2
candidate gene, in primary human CD34+ cells recapitulates features of ineffective
hematopoiesis observed in MDS. Blood 2007 (ASH 2007, abstract #4564).
21. Sankaran VG, Walkley CR, Spiegelman BM, et al. Rb intrinsically promotes erythro-
poiesis by coupling cell cycle exit with mitochondrial biogenesis. Blood 2007(ASH
2007, abstract #638).
22. Freeman KB, Haldar D. The inhibition of mammalian mitochondrial NADH oxi-
dation by chloramphenicol and its isomers and analogues. Can J Biochem 1968;
46:1003–1008.
23. Bottomley SS. Sideroblastic anemias. In: Lee GR, Foerster J, Lukens JN, Paraskevas
F, Greer JP, Rodgers G, eds. Wintrobe’s Clinical Hematology. 10th ed. Philadelphia,
Baltimore: Lippincott Williams and Wilkins, 1998:1022–1045.
24. Gregg XT, Reddy V, Prchal JT. Copper deficiency masquerading as myelodysplastic
syndrome. Blood 2002; 100:1493–1495.
25. Williams DM, Loukopoulos D, Lee GR, et al. Role of copper in mitochondrial iron
metabolism. Blood 1976; 48:77–85.
146 Gattermann
26. Goodman JR, Dallman PR. Role of copper in iron localization in developing ery-
throcytes. Blood 1969; 34:747–753.
27. Harman D. Ageing: A theory based on free radical and radiation chemistry. J Gerontol
1956; 11:298–300.
28. Harman D. Ageing: Overview. Ann N Y Acad Sci 2001; 928:1–21.
29. Feng J, Bussiere F, Hekimi S. Mitochondrial electron transport is a key determinant
of life span in Caenorhabditis elegans. Developmental Cell 2001; 1:633–644.
30. Lindsay DG. Diet and ageing: The possible relation to reactive oxygen species. J
Nutr Health Aging 1999; 3:84–91.
31. de Grey AD. The reductive hotspot hypothesis of mammalian ageing: Mem-
brane metabolism magnifies mutant mitochondrial mischief. Eur J Biochem 2002;
269:2003–2009.
32. de Grey AD. Three detailed hypotheses implicating oxidative damage to mitochon-
dria as a major driving force in homeotherm aging. Eur J Biochem 2002; 269:1995.
33. Wallace DC. Mitochondrial diseases in man and mouse. Science 1999; 283:1482–
1488.
34. Wallace DC. A mitochondrial paradigm of metabolic and degenerative diseases,
ageing, and cancer: A dawn for evolutionary medicine. Annu Rev Genet 2005;
39:359–407.
35. Cortopassi GA, Shibata DD, Soong NW, et al. A pattern of accumulation of a somatic
deletion of mitochondrial DNA in aging human tissues. Proc Natl Acad Sci USA
1992; 89:7370–7374.
36. Corral-Debrinski M, Shoffner JM, Lott MT, et al. Association of mitochondrial
DNA damage with aging and coronary atherosclerotic heart disease. Mutat Res
1992; 275:169–180.
37. Kadenbach B, Munscher C, Frank V, et al. Human ageing is associated with stochastic
somatic mutations of mitochondrial DNA. Mutat Res 1995; 338:161–172.
38. Brierley EJ, Johnson MA, Lightowlers RN, et al. Role of mitochondrial DNA muta-
tions in human aging: Implications for the central nervous system and muscle. Ann
Neurol 1998; 43:217–223.
39. Murdock DG, Christiacos NC, Wallace DC. The age-related accumulation of a
mitochondrial DNA control region mutation in muscle, but not brain, detected by
a sensitive PNA-directed PCR clamping based method. Nucleic Acids Res 2000;
28:4350–4355.
40. Coller HA, Bodyak ND, Khrapko K. Frequent intracellular clonal expansions of
somatic mtDNA mutations. Ann NY Acad Sci 2002; 959:434–447.
41. Richter C. Reactive oxygen and DNA damage in mitochondria. Mutat Res 1992;
275:249–255.
42. Yakes FM, van Houten B. Mitochondrial DNA damage is more extensive and persists
longer than nuclear DNA damage in human cells following oxidative stress. Proc
Natl Acad Sci USA 1997; 94:514–519.
43. Richter C, Park JW, Ames BN. Normal oxidative damage to mitochondrial and
nuclear DNA is extensive. Proc Natl Acad Sci USA 1988; 85:645–646.
44. Marcelino LA, Thilly WG. Mitochondrial mutagenesis in human cells and tissues.
Mutat Res 1999; 434:177–203.
45. Khrapko K, Coller HA, Andre PC, et al. Mitochondrial mutational spectra in human
cells and tissues. Proc Natl Acad Sci USA 1997; 94:13798–13803.
46. Wang E, Wong A, Cortopassi G. The rate of mitochondrial mutagenesis is faster in

mice than in humans. Mutat Res 1997; 377:157–166.
47. Taylor RW, Turnbull DM. Mitochondrial DNA mutations in human disease. Nat Rev
Genet 2005; 6:389–402.
48. Shin MG, Kajigaya S, McCoy JP, et al. Marked mitochondrial DNA sequence
heterogeneity in single CD34+ cell clones from normal adult bone marrow. Blood
2004; 103:553–561.
49. Ogasawara Y, Nakayama K, Tarnowka M, et al. Mitochondrial DNA spectra of single
human CD34+ cells, T cells, B cells, and granulocytes. Blood 2005; 106:3271–
3284.
50. Yao Y-G, Ellison FM, McCoy JP, et al. Age-dependent accumulationof mtDNA
mutations in murine hematopoietic stem cells is modulated by the nuclear genetic
background. Hum Mol Genet 2007; 16:286–294.
51. Krishnan KJ, Greaves LC, Reeve AK, et al. The ageing mitochondrial genome.
Nucleic Acids Res 2007; 35:7399–7405.
52. Khrapko K, Kraytsberg Y, de Grey AD, et al. Does premature aging of the mtDNA
mutator mouse prove that mtDNA mutations are involved in natural aging? Aging
Cell 2006; 5:279–282.
53. Vermulst M, Bielas JH, Kujoth G, et al. Mitochondrial point mutations do not limit
the natural lifespan of mice. Nat Genet 2007; 39:540–543.
54. Khrapko K, Vijg J. Mitochondrial DNA mutations and aging: A case closed? Nat
Genet 2007; 39:445–446.
55. Gattermann N, Berneburg M, Heinisch J, et al. Detection of the ageing-associated
5-Kb deletion of mitochondrial DNA in blood and bone marrow of hematologically
normal adults. Absence of the deletion in clonal bone marrow disorders. Leukemia
1995; 9:1704–1710.
56. Morrison SJ, Wandycz AM, Akashi K, et al. The aging of hematopoietic stem cells.
Nat Med 1996; 2:1011–1016.
57. Marley SB, Lewis JL, Davidson RJ, et al. Evidence for a continuous decline in
haematopoietic cell function from birth: Application to evaluating bone marrow
failure in children. Br J Haematol 1999; 106:162–166.
58. Geiger H, van Zant G. The aging of lympho-hematopoietic stem cells. Nature
Immunology 2002; 3:329–333.
59. Trifunovic A, Wredenberg A, Falkenberg M, et al. Premature ageing in mice
expressing defective mitochondrial DNA polymerase. Nature 2004; 429:417–
423.
60. Kujoth GC, Hiona A, Pugh TD, et al. Mitochondrial DNA mutations, oxidative
stress, and apoptosis in mammalian aging. Science 2005; 309:481–484.
61. Chen M, Kundu M, Shelat S, et al. Animal model of mitochondrial dysfunction
generating macrocytic anemia and myelodysplastic bone marrow failure. Blood
2007(ASH 2007, abstract #5382).
62. Trifunovic A, Hansson A, Wredenberg A, et al. Somatic mtDNA mutations cause
aging phenotypes without affecting reactive oxygen species production. Proc Natl
Acad Sci USA 2005; 102:17993–17998.
63. Inoue S, Yokota S, Nakada K, et al. Pathogenic mitochondrial DNA-induced res-
piration defects in hematopoietic cells result in anemia by suppressing erythroid
differentiation. FEBS Lett 2007; 581:1910–1916.
148 Gattermann
64. Gattermann N, Retzlaff S, Wang YL, et al. A heteroplasmic point mutation of

mitochondrial tRNALeu(CUN) in non-lymphoid cell lineages from a patient with
acquired idiopathic sideroblastic anaemia. Br J Haematol 1996; 93:845–855.
65. Gattermann N, Retzlaff S, Wang Y-L, et al. Heteroplasmic point mutations of mito-
chondrial DNA affecting subunit I of cytochrome c oxidase in two patients with
acquired idiopathic sideroblastic anemia. Blood 1997; 90:4961–4972.
66. Gattermann N. From sideroblastic anemia to the role of mitochondrial DNA muta-
tions in myelodysplastic syndromes. Leuk Res 2000; 24:141–151.
67. Gattermann N, Wulfert M, Junge B, et al. Ineffective hematopoiesis linked with a
mitochondrial tRNA mutation (G3242 A) in a patient with myelodysplastic syn-
drome. Blood 2004; 103:1499–1502.
68. Bröker S, Meunier B, Rich P, et al. MtDNA mutations associated with sideroblastic
anaemia cause a defect of mitochondrial cytochrome c oxidase. Eur J Biochem 1998;
258:132–138.
69. Wulfert M, Küpper AC, Tapprich C, et al. Analysis of mitochondrial DNA in 104
patients with myelodysplastic syndromes. Exp Hematol 2008; 36:577–586.
70. Wulfert M, Tapprich C, Gattermann N. Optimized mtDNA fragments for heterodu-
plex analysis of the whole human mitochondrial genome with denaturing HPLC.
J Chromatogr B Biomed Appl 2006; 831:236.
71. Shin MG, Kajigaya S, Levin BC, et al. Mitochondrial DNA mutations in patients
with myelodysplastic syndromes. Blood 2003; 101:3118–3125.
72. He L, Luo L, Proctor SJ, et al. Somatic mitochondrial DNA mutations in adult-onset
leukaemia. Leukemia 2003; 17:2487–2491.
73. Linnartz B, Anglmayer R, Zanssen S. Comprehensive scanning of somatic mito-
chondrial DNA alterations in acute leukemia developing from myelodysplastic syn-
dromes. Cancer Res 2004; 64:1966–1971.
74. Grist SA, Lu X-J, Morley AA. Mitochondrial mutations in acute leukemia. Leukemia
2004; 18:1313–1316.
75. Yao Y-G, Ogasawara Y, Kajigaya S, et al. Mitochondrial DNA sequence variation in
single cells from leukemia patients. Blood 2007; 109:756–762.
76. Carew JS, Zhou Y, Albitar M, et al. Mitochondrial DNA mutations in primary
leukemia cells after chemotherapy: Clinical significance and therapeutic implica-
tions. Leukemia 2003; 17:1437–1447.
77. Penta JS, Johnson FM, Wachsman JT, et al. Mitochondrial DNA in human malig-
nancy. Mutat Res 2001; 488:119–133.
78. Salas A, Yao Y-G, Macaulay V, et al. A critical reassessment of the role of mito-
chondria in tumorigenesis. PLoS Medicine 2005; 2:1158–1166 (e296).
79. Porra RJ, Jones OTG. Studies on ferrochelatase. Biochem J 1963; 87:181–
192.
80. Grasso JA, Myers TJ, Hines JD, et al. Energy-dispersive X-ray analysis of the
mitochondria of sideroblastic anaemia. Br J Haematol 1980; 46:57–72.
81. Shirihai OS, Gregory T, Yu C, et al. ABC-me: A novel mitochondrial transporter
induced by GATA-1 during erythroid differentiation. The EMBO J 2000; 19:2492–
2502.
82. Flachs P, Sponarova J, Kopecky P, et al. Mitochondrial uncoupling protein 2 gene
transcript levels are elevated in maturing erythroid cells. FEBS Lett 2007; 581:1093–
1097.
83. Lange H, Kispal G, Lill R. Mechanism of iron transport to the site of heme synthesis
inside yeast mitochondria. J Biol Chem 1999; 274:18989–18996.
84. Mühlenhoff U, Stadler JA, Richhardt N, et al. A specific role of the yeast mito-
chondrial carriers MRS3/4p in mitochondrial iron acquisition under iron-limiting
conditions. J Biol Chem 2003; 278:40612–40620.
85. Shaw GC, Cope JJ, Li L, et al. Mitoferrin is essential for erythroid iron assimilation.
Nature 2006; 440:96–100.
86. Kushner JP, Lee GR, Wintrobe MM, et al. Idiopathic refractory sideroblastic anemia.
Clinical and laboratory investigation of 17 patients and review of the literature.
Medicine (Baltimore) 1971; 50:139–159.
87. Pasanen AVO, Vuopio P, Borgström GH, et al. Heme biosynthesis in refractory
sideroblastic anaemia associated with the preleukaemic syndrome. Scand J Haematol
1981; 27:35–44.
88. Steensma DP, Hecksel KA, Porcher JC, et al. Candidate gene mutation analysis
in idiopathic acquired sideroblastic anemia (refractory anemia with ringed siderob-
lasts). Leuk Res 2007; 31:623–628.
89. Cazzola M, Invernizzi R, Bergamaschi G, et al. Mitochondrial ferritin expression
in erythroid cells from patients with sideroblastic anemia. Blood 2003; 101:1996–
2000.
90. Tehranchi R, Invernizzi R, Grandien A, et al. Aberrant mitochondrial iron dis-
tribution and maturation arrest characterize early erythroid precursors in low-risk
myelodysplastic syndromes. Blood 2005; 106:247–253.
91. della Porta MG, Malcovati L, Invernizzi R, et al. Flow cytometry evaluation of
erythroid dysplasia in patients with myelodysplastic syndrome. Leukemia 2006;
20:549–555.
92. Santambrogio P, Biasiotto G, Sanvito F, et al. Mitochondrial ferritin expression in
adult mouse tissues. J Histochem Cytochem 2007; 55:1129–1137.
93. Nie G, Sheftel AD, Kim SF, et al. Overexpression of mitochondrial ferritin
causes cytosolic iron depletion and changes cellular iron homeostasis. Blood 2005;
105:2161–2167.
94. Muta K, Krantz SB. Inhibition of heme synthesis induces apoptosis in human ery-
throid progenitor cells. J Cell Physiol 1995; 163:38–50.
95. Nakajima O, Takahashi S, Harigaee H, et al. Heme deficiency in erythroid lineage
causes differentiation arrest and cytoplasmic iron overload. EMBO J 1999; 18:6282–
6289.
96. Fuhler GM, Hooijenga F, Drayer AL, et al. Reduced expression of flavocytochrome
b558, a component of the NADPH oxidase complex, in neutrophils from patients
with myelodysplasia. Exp Hematol 2003; 31:752–759.
97. Houwerzijl E, Blom N, Van Der Want JJ, et al. Increased peripheral platelet destruc-
tion and caspase-3-independent programmed cell death of bone marrow megakary-
ocytes in myelodysplastic patients. Blood 2005; 105:3472–3479.
98. Houwerzijl E, Blom N, van der Want JJ, et al. Megakaryocyte dysfunction in
myelodysplastic syndromes and idiopathic thrombocytopenic purpura is in part due
to different forms of cell death. Leukemia 2006:1937–1942.
99. Braun T, Carvalho G, Grosjean J, et al. Differentiating megakaryocytes in myelodys-
plastic syndromes succumb to mitochondrial derangement without caspase activa-
tion. Apoptosis 2007; 12:1101–1108.
150 Gattermann
100. Ott M, Gogvadze V, Orrenius S, et al. Mitochondria, oxidative stress and cell death.
Apoptosis 2007; 12:913–922.
101. Raza A, Gezer S, Mundle S, et al. Apoptosis in bone marrow biopsy samples involv-
ing stromal and hematopoietic cells in 50 patients with myelodysplastic syndromes.
Blood 1995; 86:268–276.
102. Bouscary D, De Vos J, Guesnu M, et al. Fas/Apo-1(CD95) expression and
apoptosis in patients with myelodysplastic syndromes. Leukemia 1997; 11:839–
845.
103. Hellström-Lindberg E, Schmidt-Mende J, Forsblom AM, et al. Apoptosis in refrac-
tory anaemia with ringed sideroblasts is initiated at the stem cell level and associated
with increased activation of caspases. Br J Haematol 2001; 112:714–726.
104. Shetty V, Hussaini S, Alvi S, et al. Excessive apoptosis, increased phagocytosis,
nuclear inclusion bodies and cylindrical confronting cisternae in bone marrow biop-
sies of myelodysplastic patients. Br J Haematol 2002; 116:817–825.
105. Tehranchi T, Fadeel B, Forsblom AM, et al. Granulocyte colony-stimulating factor
inhibits spontaneous cytochrome c release and mitochondria-dependent apoptosis of
myelodysplastic syndrome hematopoietic progenitors. Blood 2003; 101:1080–1086.
106. Testa U. Apoptotic mechanisms in the control of erythropoiesis. Leukemia 2004;
18:1176–1199.
107. Claessens YE, Park S, Dubart-Kupperschmitt A, et al. Rescue of early-stage
myelodysplastic syndrome-deriving erythroid precursors by the ectopic expression
of a dominant-negative form of FADD. Blood 2005; 105:4035–4042.
108. Fontenay M, Cathelin S, Amiot M, et al. Mitochondria in hematopoiesis and hema-
tological diseases. Oncogene 2006; 25:4757–4767.
109. Tehranchi R, Fadeel B, Schmidt-Mende J, et al. Antiapoptotic role of growth fac-
tors in the myelodysplastic syndromes: Concordance between in vitro and in vivo
observations. Clin Cancer Res 2005; 11:6291–6299.
110. Kroemer G, Reed JC. Mitochondrial control of cell death. Nat Med 2000; 6:513–519.
111. Wolvetang EJ, Johnson KL, Krauer K, et al. Mitochondrial respiratory chain
inhibitors induce apoptosis. FEBS Lett 1994; 339:40–44.
112. Wang J, Silva JP, Gustafsson CM, et al. Increased in vivo apoptosis in cells lacking
mitochondrial DNA gene expression. Proc Natl Acad Sci USA 2001; 98:4038–4043.
113. Schildgen V, Junge B, Gattermann N. Impaired mitochondrial gene expression in
MDS patients. Leuk Res 2007; 31(Suppl. 1):S35.
114. Parker JE, Fishlock KL, Mijovic J, et al. ‘Low-risk’ myelodysplastic syndrome
is associated with excessive apoptosis and an increased ratio of pro- versus anti-
apoptotic bcl-2-related proteins. Br J Haematol 1998; 103:1075–1082.
115. Berger G, Hunault-Berger M, Rachieru P, et al. Increased apoptosis in mononucleated
cells but not CD34+ cells in blastic forms of myelodysplastic syndromes. Haematol
J 2001; 2:87–96.
anaemia (AISA). Br J Haematol 1990; 74:45–52.
117. Germing U, Gattermann N, Aivado M, et al. Two types of acquired idiopathic sider-
oblastic anaemia (AISA): A time-tested distinction. Br J Haematol 2000; 108:724–
728.
118. van Duppen V, Raets V, Raeymaekers L, et al. CD34+ progenitors from low-risk
MDS patients are more susceptible to induced oxidative damage and have reduced
DNA repair compared to healthy control subjects. Blood 2003; 102(11), abstract
#3410.
119. Jankowska AM, Gondek LP, Szpurka H, et al. Base excision repair dysfunction in a
subgroup of patients with myelodysplastic syndrome. Leukemia 2008; 22:551–558.
120. Buttgereit F, Brand MD. A hierarchy of ATP-consuming processes in mammalian
cells. Biochem J 1995; 312:163–167.
121. Wickramasinghe SN, Hughes M. Capacity of ringed sideroblasts to synthesize
nucleic acids and protein in patients with primary acquired sideroblastic anaemia.
Br J Haematol 1978; 38:345–352.
122. Löffler M, Jöckel J, Schuster G, et al. Dihydroorotate-ubiquinone oxidoreductase
links mitochondria in the biosynthesis of pyrimidine nucleotides. Mol Cell Biochem
1997; 174:125–129.
123. Gattermann N, Dadak M, Hofhaus G, et al. Severe impairment of nucleotide synthesis
through inhibition of mitochondrial respiration. Nucleosides Nucleotides Nucleic
Acids 2004; 23:1275–1279.
124. Kunz BA, Kohalmi SE, Kunkel TA, et al. Deoxyribonucleoside triphosphate levels:
A critical factor in the maintenance of genetic stability. Mutat Res 1994; 318:1–64.
125. Loveland B, Wang C-R, Yonekawa H, et al. Maternally transmitted histocompatibil-
ity antigen of mice: A hydrophobic peptide of a mitochondrially encoded protein.
Cell 1990; 60:971–980.
126. Renneville A, Quesnel B, Charpentier A, et al. High occurrence of JAK2 V617
F mutation in refractory anemia with ringed sideroblasts associated with marked
thrombocytosis. Leukemia 2006; 20:2067–2070.
127. Szpurka H, Tiu R, Murugesan G, et al. Refractory anemia with ringed siderob-
lasts associated with marked thrombocytosis (RARS-T), another myeloproliferative
condition characterized by JAK2 V617 F mutation. Blood 2006; 108:2173–2181.
128. Remacha AF, Nomdedeu JF, Puget G, et al. Occurrence of the JAK2 V617 F muta-
tion in the WHO provisional entity: Myelodysplastic/myeloproliferative disease,
unclassifiable refractory anemia with ringed sideroblasts associated with marked
thrombocytosis. Haematologica 2006; 91:719–720.
129. Boissinot M, Garand R, Hamidou M, et al. The JAK2-V617 F mutation and essential
thrombocythemia features in a subset of patients with refractory anemia with ringed
sideroblasts (RARS). Blood 2006; 108:1781–1782.
130. Wang SA, Hasserjian RP, Loew JM, et al. Refractory anemia with ringed sideroblasts
associated with marked thrombocytosis harbors JAK2 mutation and shows overlap-
ping myeloproliferative and myelodysplastic features. Leukemia 2006; 20:1641–
1644.
131. Gattermann N, Billiet J, Kronenwett R, et al. High frequency of the JAK2 V617 F
mutation in patients with thrombocytosis (platelet count ⬎ 600 × 109 /L) and ringed
sideroblasts more than 15% considered as MDS/MPD, unclassifiable. Blood 2007;
109:1334–1335.
132. Zipperer E, Wulfert M, Germing U, et al. MPL 515 and JAK2 mutation analysis in
MDS presenting with a platelet count of more than 500 × 109/l. Ann Hematol 2008;
87:413–415.
7
Defects in Iron Metabolism and Iron
Overload in MDS
Luca Malcovati
Department of Hematology, University of Pavia Medical School and
Fondazione IRCCS Policlinico San Matteo, Pavia, Italy
INTRODUCTION
The myelodysplastic syndromes (MDS) are a heterogeneous group of hemato-
logic disorders that are clinically characterized by peripheral cytopenias due to
ineffective hematopoiesis and progressively impaired ability of progenitor cells to
differentiate (1). The erythroid lineage is the most commonly affected and, as a
result, anemia is the most frequent peripheral cytopenia observed. According to the
World Health Organization (WHO) definition of anemia (hemoglobin ⬍13 g/dL
for men and ⬍12 g/dL for women) more than 90% of MDS patients are anemic at
the time of diagnosis, while moderate to severe anemia (hemoglobin ⬍10 g/dL) is
observed in almost 60% of cases (2). Severe anemia is usually symptomatic, and
the majority of these patients eventually need to be treated with regular red blood
cell transfusions, the frequency of which usually increases over time.
The anemia associated with MDS is the result of ineffective erythropoiesis, a
process that is sustained by excessive intramedullary hematopoietic cell apoptosis
(3).
This defect in the maturation of erythroid cells is associated with various
abnormalities in iron metabolism. These abnormalities include increased iron
absorption from the gut, and abnormal iron accumulation in the mitochondrial
matrix of red blood cell precursors in about one-third of patients (4). In addition,
secondary iron overload is observed in patients who develop a need for regular
red cell transfusions (5).
153
154 Malcovati
INCREASED IRON ABSORPTION DUE TO INEFFECTIVE

ERYTHROPOIESIS
Ineffective erythropoiesis can contribute to increased body iron stores, even in
the absence of red cell transfusion. About 60% of those MDS patients in follow-
up at the Division of Hematology of the University of Pavia not receiving red
cell transfusions showed evidence of iron overload as assessed by biochemical
markers. This proportion is significantly higher in MDS patients who have ringed
sideroblasts (L. Malcovati and colleagues, unpublished observation). In a study by
Mario Cazzola in Pavia and his colleagues, plasma iron and transferrin saturation
were simultaneously increased in about 60% of patients with idiopathic refractory
sideroblastic anemia at their initial evaluation, before any blood transfusion (4).
These biochemical abnormalities imply that these patients have increased
iron stores. In fact, all these patients had elevated serum ferritin values and 90%
had increased stainable iron on liver biopsy specimens at clinical onset. Mean
serum ferritin level in that series of patients was around 700 ␮g/L, with a mean
transferrin saturation of about 60%. Ferrokinetic measurement of erythron iron
turnover showed that most patients had an increased erythroid marrow activity of
up to over 10 times the base value. Erythropoietic activity measured by erythron
transferring uptake was found to be on average 130% of the base value in patients
with refractory anemia, and more than 400% of the base value in patients with
idiopathic refractory sideroblastic anemia. The efficiencies of erythropoiesis were
around 50% in refractory anemia and 10% in sideroblastic anemia (6).
The size of the increase of serum ferritin levels in refractory anemias was
correlated with the duration of anemia and the magnitude of increase in the erythron
iron turnover. This suggests that the level of erythroid proliferation directly affects
gastrointestinal iron absorption, which over time contributes to iron overload. In
patients with refractory anemias, the iron overload due to ineffective erythopoiesis
is usually mild and is not usually associated with clinical signs of organ damage,
but it may be significantly worsened by blood transfusions.
Excessive iron absorption and progressive iron loading are also distinctive
characteristics of congenital anemias due to ineffective erythropoiesis, such as
the intermediate and major forms of ß-thalassemia, and other rare anemias,
including congenital dyserythropoietic anemia and X-linked sideroblastic anemia
(XLSA) (7). These constitutional diseases characterized by progressive iron
overload are also defined as iron-loading anemias. In these conditions, the
degree of anemia is a poor predictor of iron loading, which correlates better with
erythroid marrow activity. The progressive iron overload is usually associated
with an accumulation of iron in parenchymal cells, resulting in toxic damage
and organ dysfunction. In MDS, in contrast, parenchymal iron loading is of less
clinical importance because of a shorter disease duration and more effective
erythropoiesis.
A major step forward in the understanding of the pathophysiology of the
iron loading anemias was the identification of hepcidin as the key regulator of
Defects in Iron Metabolism and Iron Overload in MDS 155
systemic iron homeostasis. However, crucial steps in this process must still be
clarified (8,9).
Hepcidin inhibits iron influx into plasma from duodenal enterocytes that
absorb dietary iron, from macrophages that recycle iron from senescent erythro-
cytes and from hepatocytes that store iron (10). Hepcidin acts by binding to the
cellular iron exporter, ferroportin, and causing its internalization and degradation
(11). Hepcidin production is increased by iron and inflammation, and decreased
by anemia and hypoxia (12). However, the molecular mechanisms of hepcidin
regulation by iron, oxygen, and anemia are still not fully understood (13,14).
In iron-loading anemias, hepcidin is regulated by opposing influences of
ineffective erythropoiesis and concomitant iron overload (9). Most of the iron
absorbed from the diet or recycled from hemoglobin is destined to produce ery-
throcytes. It is, therefore, not surprising that hepcidin production is also regulated
by anemia (14). When oxygen delivery is inadequate, the homeostatic response
is to increase the production of erythrocytes. Thus, hepcidin levels decrease and
more iron is made available from intestinal absorption and from the pool stored in
hepatocytes and macrophages. Patients with thalassemia intermedia develop iron
overload even if never transfused, and their urinary hepcidin levels are remarkably
low despite systemic iron overload (15,16). These observations suggest erythro-
poietic activity is the most potent suppressor of hepcidin synthesis, but how this
signal is conveyed from the bone marrow to the liver, the site of hepcidin synthesis,
is not yet known (14,17).
Red blood cell transfusions would be expected to affect hepcidin production
by relieving anemia, which should reduce hepcidin suppression, and by increasing
body iron load, which should increase upregulation of hepcidin synthesis. Indeed,
a comparison of urinary hepcidin before and after transfusion showed that most
patients responded by increasing hepcidin levels. Nevertheless, this response to
iron load was much lower in thalassemia major patients when compared with
normal subjects, indicating a continued regulation of hepcidin by a suppressive
drive (18).
The discovery of hepcidin has broadened our understanding of disturbances
of iron homeostasis in iron-loading anemias. The findings suggest that anemia,
especially when associated with increased and ineffective erythropoiesis, has a
strong effect on hepcidin production which is dominant over iron (10). The conse-
quent low levels of hepcidin may be responsible for increased absorption of iron,
resulting in systemic iron overload and associated organ damage.
Recently, GDF15, a member of the transforming growth factor-beta super-
family secreted during erythroblast maturation, was found to be significantly
increased in patients with thalassemia, resulting in hepcidin suppression (19). It
has been proposed that anemia from ineffective erythropoiesis results in increased
GDF15 levels by erythroblast secretion, driving hepcidin suppression. As a result,
increased amounts of dietary iron overload are absorbed and secondary hemochro-
matosis develops over the patient’s lifetime.
156 Malcovati
Interestingly, high levels of GDF15 were found in a small group of MDS

patients, even though these levels were lower than those detected in patients with
thalassemia syndromes (19). This observation suggests that a similar mechanism
might underlie the mild iron loading due to ineffective hematopoiesis observed in
MDS.
MITOCHONDRIAL IRON ACCUMULATION AND RING SIDEROBLASTS

Ring Sideroblasts
Most of the iron used in erythroblasts for hemoglobin synthesis are obtained from
transferrin internalized on the transferrin receptor (20). Within cells, iron is stored
in ferritin molecules, which in human tissues are composed of variable proportions
of two subunits: L-ferritin (light) and H-ferritin (heavy). H-ferritin has ferroxidase
activity that detoxifies free iron into the less-soluble ferric form. The excess iron
is stored in ferritin molecules that partially aggregate, producing hemosiderin
within specific endosomes called siderosomes. These iron-loaded endosomes can
be detected by Prussian blue staining (Perls reaction) in about one-third of normal
immature erythroid cells, and those erythroblasts with few blue granules scattered
in the cytoplasm are defined as “ferritin sideroblasts” [Fig. 1(A)]. The number
and size of ferritin sideroblasts increase whenever the iron supply to the erythron
exceeds the amount required for hemoglobin synthesis, as typically occurs in
iron-loading anemias.
About one-third of the patients with MDS show the so-called “ring sider-
oblasts”. These are identified as immature red cells in which 10 or more blue
granules form a ring around the nucleus after Prussian blue staining [Figs. 1(A)
and 1(B)] (21). Under electron microscope, these granules were found to be iron-
loaded mitochondria rather than cytoplasmic siderosomes (22). Therefore, ring
(A) (B)
Figure 1 (A) Perls Prussian blue stain of bone marrow smears from a patient with
refractory anemia with ringed sideroblasts showing a ferritin sideroblast with few small iron-
containing granules, and a ringed sideroblast with more numerous granules, surrounding
the nucleus. (B) Perls Prussian blue stain of bone marrow smears from a typical patient with
refractory anemia with ringed sideroblasts, showing a high proportion of ringed sideroblasts.
sideroblasts differ from ferritin sideroblasts in morphological features, Perls pos-

itive granules tending to form a ring surrounding the nucleus. More important,
they differ in pathophysiology, as the stained granules in ring sideroblasts are
iron-loaded mitochondria rather than cytoplasmic siderosomes containing ferritin
and hemosiderin.
Mitochondrial Ferritin
The nature of the excess mitochondrial iron in ring sideroblasts has only recently
been defined (23). Mitochondrial ferritin is a novel ferritin encoded by an intron-
less gene on chromosome 5q23.1 (24). The protein is synthesized as a precursor of
about 30 kDa, which includes a leader sequence of 60 amino acids that targets the
protein to mitochondria and is proteolytically removed inside the mitochondria.
Mitochondrial ferritin has ferroxidase activity and is therefore likely to sequester
potentially harmful free iron (25). This protein has a very restricted expression
in human tissues (erythroid cells, testis) and does not seem to be an essential part
of the transfer of free iron to heme and other iron compounds in mitochondria.
In 2003, Cazzola et al. studied the expression of mitochondrial ferritin in patients
with inherited sideroblastic anemia and refractory anemia with ringed sideroblasts
(RARS), as well as in healthy subjects and patients with refractory anemia with-
out ringed sideroblasts with a double immunocytochemical staining for H- and
mitochondrial ferritin (23). They observed that erythroblasts from patients with
sideroblastic anemia were positive for mitochondrial ferritin, which regularly
appeared as granules ringing the nucleus [Figs. 2(A) and 2(B)], while immature red
cells, mostly proerythroblasts and basophilic erythroblasts, from healthy subjects
and MDS patients without ringed sideroblasts, showed diffuse cytoplasmic pos-
itivity for H-ferritin [Fig. 2(C)]. These findings suggested that the iron deposited
in perinuclear mitochondria of ring sideroblasts is present in the form of mito-
chondrial ferritin and that this latter is a specific marker of sideroblastic anemia.
Classification and Pathophysiology of Sideroblastic Anemias

The biological mechanisms underlying mitochondrial iron accumulation and mito-
chondrial ferritin expression in MDS with ring siderblasts are still unknown.
Mitochondrial iron loading and ring sideroblasts characterize a group of
congenital and acquired disorders that display remarkable clinical and hematologic
heterogeneity, identified as sideroblastic anemias (Table 1) (26,27).
The most common of the inherited forms of anemia is XLSA (OMIM
301300), which is caused by mutations in the erythroid-specific ALA synthase
gene (28). The most frequent acquired sideroblastic anemia is the myelodysplastic
syndrome described for the first time in 1956 by Sven-Erik Björkman in Sweden
(21) and identifed as idiopathic acquired sideroblastic anemia (RARS) in the
1982 FAB classification (29). In 2001, the WHO classification distinguished two
entities within this group: refractory anemia with ringed sideroblasts (RARS)
and refractory cytopenia with multilineage dysplasia and ringed sideroblasts
158 Malcovati
(A) (B)
(C)
Figure 2 (A,B) Bone marrow smear from a patient with refractory anemia with ringed
sideroblasts (immunoalkaline phosphatase staining for mitochondrial ferritin). Erythro-
blasts are positive for mitochondrial ferritin, which appears as granules ringing the nucleus.
(C) Bone marrow smear from a patient with refractory anemia (immunoalkaline phos-
phatase staining for H-ferritin subunit), showing a diffuse pattern consistent with cytosolic
localization of ferritin.
(RCMD-RS) (30). This distinction was based on whether marrow dysplasia is

unilineage (erythroid only) or multilineage, which was previously reported to be
of clinical importance with respect to MDS cases fitting the category of refrac-
tory anemia (31–33). The 2008 WHO classification changed the word “ringed” to
“ring” and merged RCMD-RS into RCMD, because if multilineage dysplasia is
present, the presence of ring sideroblasts does not modify the prognosis.
MDS are frequent hematologic disorders in older persons (see chap. 2): the
crude incidence is about 5:100,000 per year, and about 40:100,000 per year in
individuals over 70 years of age (34–36). RARS represents 10% to 15% of all
MDS and is, therefore, a relatively common hematologic disorder in the elderly.
Table 1 Classification of the Sideroblastic Anemias

Category
Hereditary sideroblastic anemias

X-linked sideroblastic anemia (XLSA)
X-linked sideroblastic anemia with cerebellar ataxia (XLSA/A)
Mitochondrial myopathy, lactic acidosis, and sideroblastic anemia (MLASA)
Pearson marrow-pancreas syndrome (sideroblastic anemia with vacuolization of
marrow precursors and exocrine pancreatic dysfunction)
Acquired sideroblastic anemias
Myelodysplastic syndromes
Refractory anemia with ringed sideroblasts
Refractory cytopenia with multilineage dysplasia and ringed sideroblasts
Myelodysplastic/myeloproliferative disorders
Refractory anemia with ringed sideroblasts associated with marked thrombocytosis
Ethanol-induced sideroblastic anemia
Drug-induced sideroblastic anemia (chloramphenicol, isoniazid)
Different pathophysiological mechanisms operate in the various sideroblas-

tic anemias, but ring sideroblasts, ineffective erythropoiesis, and iron loading are
key features of all these syndromes.
The most well-characterized type of siderobastic anemia at a molecular
level is XLSA, which is caused by defective heme synthesis and secondary mito-
chondrial iron loading (28). Defective heme synthesis is caused by missense
mutations in the erythroid-specific ALA synthase gene (ALAS2) that result in the
defective activity of the mitochondrial enzyme. The ALAS2 protein catalyzes
the rate-limiting step in heme biosynthesis and provides the large quantity of
heme required by erythroid cells for hemoglobin synthesis. The ALAS2 promoter
contains putative binding sites for the transcription factors GATA-1 and NF-E2,
which also play a crucial role in the erythroid-specific activation of globin genes
(37). ALAS2 expression is also modulated translationally, since synthesis of the
ALAS2 mRNA is dependent on an adequate iron supply (38).
There are now over 30 mutations in ALAS2 described in more than 30
XLSA kindreds. Defective ALAS2 enzyme activity in bone marrow erythroid cells
leads to insufficient protoporphyrin IX synthesis, mitochondrial iron overload, and
intramedullary death of red cell precursors (39–41).
Almost all XLSA patients are hemizygous males. Most of the heterozygous
females have only minor red cell abnormalities without significant clinical signs,
since immature red cells expressing the normal ALAS2 are sufficient to sustain
a normal level of red-cell production. As in any X-linked disorder, however,
the clinical phenotype of female carriers may be influenced by the pattern of
X-chromosome inactivation. A few XLSA female heterozygotes may, therefore,
preferentially express the mutant allele in erythroid cells and are anemic (42,43).
160 Malcovati
Other genes mutated in rare inherited sideroblastic anemia syndromes

include ABCB7 at Xq13.1 in the XLSA associated with spinocerebellar ataxia
(MIM #301310) (44), mitochondrial-encoded genes in Pearson marrow-pancreas
syndrome (MIM #557000) (45,46), and PUS1 (pseudouridine synthase-1) at
12q24.33 in mitochondrial myopathy with lactic acidosis and sideroblastic anemia
(MIM #600462) (47,48).
Despite the considerable advances that have been made in defining molec-
ular bases of inherited sideroblastic anemias, the pathophysiology of refractory
anemia with ring sideroblasts is still poorly understood. Evidence obtained from
the congenital forms raised the possibility that RARS could be associated with
somatic mutations in the same pathways where germline mutations cause the
inherited forms of the diseases (49). ALA2 gene has been analyzed for mutation in
RARS cases by several other groups of investigators, and mutations are uncom-
mon. Somatic missense mutations of FECH and its promoter were not observed
in idiopathic acquired sideroblastsic anemia patients. FECH was modestly over-
expressed in progenitor cells from patients with ringed sideroblasts, compared
with MDS patients without sideroblasts and healthy controls. In addition, ABCB7
and PUS1, genes implicated in congenital sideroblastic anemia syndromes, were
analyzed but again no coding mutations in acquired cases were found.
A defect mitochondrial enzyme function in bone marrow cells from patients
with RARS had been suggested more than 20 years ago (50), and there is now
increasing evidence that mitochondria play a central role in the pathobiology of
ineffective erythropoiesis in MDS (51). Electron microscopy demonstrated pro-
nounced ultra-structural mitochondrial changes not only in RARS but also in
other types of MDS—the mitochondria were enlarged with or without disrup-
tion of internal cristae and/or mitochondrial membranes, which was significantly
associated with the accumulation of iron (52,53).
One clue to a possible basis for mitochondrial genomic instability in MDS
may be provided by the constitutional disease called Pearson syndrome in which
sideroblastic anemia accompanies pancreatic abnormalities. In this inherited dis-
order, a number of defects of mitochondrial DNA (mtDNA) were observed. Mito-
chondrial DNA mutations have been reported in a few patients with MDS, but
the true significance of these mutations in disease pathogenesis is still not clear
(54,55). Norbert Gattermann in Düsseldorf and his colleagues initially described
abnormalities in the mitochondrial gene encoding cytochrome c oxidase in patients
with idiopathic acquired sideroblastic anemia, and later in refractory anemia with
excess blasts lacking ring sideroblasts (54,56). Another group has also reported
data supporting a role for mtDNA mutations in MDS (57). They found that mtDNA
cytochrome c oxidase I and II genes contained large numbers of mutations and
evidence of a number of single nucleotide substitutions in a large proportion of
patients. Finally, Neal Young and colleagues at the United States National Institutes
of Health systematically analyzed the entire mitochondrial genome in 10 patients
with MDS (55). Overall, they found no increase in the number of mtDNA genes
harboring polymorphisms or mutations between the patients and healthy controls,
although there were a few more mtDNA changes resulting in aminoacid substitu-
tions in MDS samples. The study did not confirm either the previously described
mutations in sideroblastic anemia or a major role for mitochondrial genomic
instability in MDS.
Studies of an in vitro model of erythroid differentiation showed that early
erythroid progenitor cells from low-risk MDS spontaneously release excessive
amounts of cytochrome c from mitochondria, resulting in activation of caspase-9
and subsequent cell death (51,58). Moreover, inhibition of caspase-9 abrogated
the enhanced sensitivity to Fas ligation, suggesting that the increased sensitivity
of MDS progenitor cells to death receptor stimulation might be due to a consti-
tutive activation of the mitochondrial apoptotic signaling pathway in these cells.
Cytochrome c release is more pronounced in RARS and positively correlates with
mitochondrial ferritin expression, suggesting that the two events are closely linked
(59).
It has been shown that mitochondrial ferritin appears at a very early stage
of erythroid differentiation, without morphologically visible signs of iron accu-
mulation, and that mitochondrial ferritin expression continues to increase during
differentiation (59,60). Since aberrant mitochondrial ferritin accumulation and
cytochrome c release are present in very early erythoblasts, iron is likely to play
an important role in the effective erythropoiesis of these patients. However, the
relationship between these two events must still be clarified.
G-CSF inhibits Fas-induced caspase activation and protects against
cytochrome c release in early progenitor cells from both sideroblastic and refrac-
tory anemia patients (58). The anti-apoptotic effect of G-CSF in mononuclear
bone marrow cells is significantly more pronounced in patients with RARS, pro-
viding evidence for the beneficial clinical effects of growth factor administration in
patients with RARS. However, G-CSF had no effect on mitochondrial respiratory
chain activity or aberrant mitochondrial ferritin expression.
Studies of clonality through X-chromonsome inactivation patterns showed
that nearly all patients with refractory anemia with ring sideroblasts have evidence
of monoclonal hematopoiesis in the stem cell compartment, suggesting that the
initial pathogenetic event in this condition occurs in the multipotent stem cell
compartment (60). Therefore, the genetic defects present in RARS must give rise
to a proliferative advantage leading to expansion of the clone, as well as to a
marked inability to form mature and normal erythrocytes.
Studies of global gene expression in MDS (see chap. 4) showed that patients
with RARS have homogeneous and distinct profiles, whereas patients with other
subtypes of MDS, such as refractory anemia and refractory anemia with excess
blasts show more overlap (61). Upregulation of the expression of several genes in
the heme biosynthesis pathway was observed in RARS; five genes involved in this
pathway (ALAS2, ALAD, HMBS, UROD, FECH) were differentially expressed
with higher than average expression levels in patients with RARS. In particular,
ALAS2, the first enzyme of heme synthesis, showed the highest expression levels
with marked upregulation (an average increase of more than 12-fold, with some
162 Malcovati
cases showing an increase of more than 100-fold) in patients with RARS com-
pared with healthy subjects. The other enzymes in the heme biosynthesis pathway
(ALAD, HMBS, UROD, FECH) showed an average increase of approximately
2-fold in RARS. Other genes which possibly play a role in the pathophysiology
of RARS were found to be expressed at higher levels, including genes involved
in erythropoiesis, such as GATA-1, CA2, and EPO-R, and in mitochondrial (e.g.,
CGI-69, TRAP1, TIMM10, and several mitochondrial ribosomal proteins). How-
ever, the functional significance of these expression changes is still not clear.
Overall, from the available data, it can be concluded that refractory anemia
with ring sideroblasts is a unique stem cell disease with abnormalities in genes of
heme pathway, but the mechanisms underlying mitochondrial iron accumulation
and proliferative advantage of myelodysplastic clone remain unknown.
Refractory Anemia with Ring Sideroblasts and Thrombocytosis

A proportion of patients with refractory anemia and ring sideroblasts have a high
platelet count, which progressively increases during the natural course of the dis-
ease (62). In these patients, thrombocytosis is associated with an increased num-
ber of megakaryocytes in the bone marrow showing morphological abnormalities,
including enlarged or giant forms, or clusters of megakayocytes, similar to changes
observed in essential thrombocythemia. On the basis of the coexistence of both
these myelodysplastic and myeloproliferative features, this disorder—identified as
refractory anemia with ringed sideroblasts associated with marked thrombocytosis
(RARS-T)—has been included as a provisional entity in the WHO classification
of myeloid neoplasms (see chap. 9) within the group of unclassifiable myelodys-
plastic/myeloproliferative disorders (30). Interestingly, this entity was found to be
significantly associated with JAK2 mutation (63), which was recently identified as
a key pathogenetic event in philadelphia-negative myeloproliferative neoplasms,
that is, polycythemia vera, essential thrombocythemia, and primary myelofibrosis
(64–67). JAK2 encodes for a tyrosine kinase that activates STATs pathway signal
transduction from receptors of hematopoietic growth factors including erythropoi-
etin, thrombopoietin, and G-CSF. The mutations so far identified result in a gain of
function which sustains proliferation of hematopoietic progenitors, the hallmark
of chronic myeloproliferative neoplasms. These mutations affect approximately
95% of the patients with polycythemia vera and more than 50% of the patients
with essential thrombocythemia and primary myelofibrosis (68,69). In addition,
about 5% of patients with essential thrombocythemia and primary myelofibrosis
were found to carry gain of function mutations in the thrombopoietin receptor
gene (MPL), resulting in activation of JAK–STAT pathway (70,71).
About 40% to 70% of patients with RARS-T were found to carry the clas-
sical JAK2 (V617 F) mutation, with a percentage of mutant allele similar to that
observed in patients with essential thrombocythemia and primary myelofibrosis
(63,72–75). Recently, homozygous cases, as well as occasional patients with MPL
mutations (76–78), have also been described. Although the molecular basis of
this disease has not yet been clarified, the association between the expression
of mitochondrial ferritin and JAK2 mutation is highly unlike to be a question of
chance given the estimated overall incidence of these two events. In fact, both
JAK2 (V617 F) and ringed sideroblasts can be observed in about 1:200,000 per-
sons per year, resulting in an estimated probability for the coexistence of these
two events of about 1:40,000,000,000. Although the true incidence of RARS-T
has not been precisely defined so far, it is much higher than what this calculation
would suggest. Preliminary studies on the clonality of hematopoiesis estimated
from XCIP and JAK2 mutation analysis, together with clinical and molecular cor-
relates, suggest that JAK2-mutated cells might represent a subclone emerging at a
stem cell level (78). However, the molecular mechanisms underlying the induction
of mitochondrial ferritin expression and JAK–STAT activating mutations must still
be clarified.
SECONDARY IRON OVERLOAD

Clinical Consequences and Prognostic Impact of Acquired Transfusional
Iron Overload
Anemia is common in MDS and nearly all patients with MDS are likely to
receive red blood cell transfusions at some time during their clinical course.
However, it is worth noting that according to evidence- and consensus-based
therapeutic guidelines, in about 40% of cases, regular red cell transfusions are the
only therapeutic option offered to the patient (79,80).
Recently, it has been shown that the onset of a need for regular transfu-
sions in patients with MDS is associated with a significant worsening of survival
(1). Transfusion-dependent patients have a significantly higher risk for both non-
leukemic death, mainly attributable to a higher risk of heart failure–related death,
and leukemic progression compared with those who do not need regular transfu-
sions (5). This suggests that the negative effect of transfusion dependency is at
least partly due to more severe anemia and more aggressive disease.
Patients receiving regular red blood cell transfusions invariably develop iron
overload. In a normal balanced state, 1 to 2 mg of iron enters and leaves the body
every day. Dietary iron is absorbed by duodenal enterocytes and circulates in the
plasma bound to transferrin, the main iron transport protein. Most of the circulating
iron is used by the erythroid bone marrow to generate hemoglobin, and excess iron
is stored in the liver and in reticuloendothelial macrophages. Traces of iron are lost
each day by the loss of mucosal and epithelial cells, and any blood loss, but the
human body has not evolved active physiological mechanisms to clear excess iron.
One unit of blood contains 200–250 mg of iron, and in patients receiving
regular red blood cell transfusions an iron overload can occur after about 20
transfusions. This results in a progressive increase in transferrin saturation, in the
appearance of nontransferrin-bound iron, and, ultimately, in the loading of
parenchymal organs, such as the liver, heart, and endocrine glands (81). In these
164 Malcovati
sites, free iron boosts the generation of free hydroxyl radicals that can result in
cell damage (82).
In patients with thalassemia major who have been inadequately iron chelated,
most clinical manifestations of iron overloading do not appear until the second
decade of life. However, evidence from serial liver biopsies indicates that the
damaging effects begin much earlier. After about 1 year of transfusions, iron is
deposited in parenchymal tissues where it may cause significant toxicity. Hepatic
iron levels exceeding 15 mg/g liver, dry weight, are associated with an increased
risk of cardiac disease and early death, while levels of between 7 and 15 mg/g
liver, dry weight, are associated with an increased risk of other complications,
including hepatic fibrosis (83).
To date, there is limited evidence on the role of iron in organ damage in
patients with MDS. In an autopsy study on 135 subjects with chronic acquired
anemia requiring red cell transfusions, cardiac iron deposits were found in about
60% of the patients receiving more than 100 red cell units, but those without
evidence of myocardial iron had had major bleeding complications, depleting iron
stores (84).
In 1981, Schafer and colleagues reported the clinical consequences of
acquired transfusional iron overload in adult patients with refractory anemia or
aplastic anemia who had received between 60 and 210 units of blood (85). In
this study, 10 out of 15 liver biopsy specimens contained between 7 and 26 times
the normal levels of iron, and portal fibrosis was common. Cardiac left ventric-
ular function was impaired in the most heavily transfused patients and in those
with coexisting coronary artery disease. However, no assessment of cardiac iron
deposits was carried out. All patients had glucose intolerance associated with a sig-
nificantly reduced insulin output. The pituitary reserve of adrenocorticotropin was
limited in 10 out of 12 patients, and that of gonadotropin in 5 out of 13 patients.
The authors concluded that the development of transfusional iron overload in
adults can result in widespread organ dysfunction.
The development of secondary iron overload, assessed as serum ferritin level,
significantly worsens survival of transfusion-dependent MDS patients (5,86). The
effect of iron overload is mainly noticeable among patients with refractory anemia,
who have a median survival of over 100 months and are, therefore, more prone
to develop the toxic effects of iron overload, while there is a borderline effect on
the survival of patients with refractory cytopenia with multilineage dysplasia, who
have a median survival of about 50 months. The negative effect of iron overload
is significantly associated with serum ferritin levels, with an approximately 40%
increase in hazard for every 500 ng/mL increase in serum ferritin.
In addition, it has recently been shown that elevated pre-transplant serum
ferritin levels significantly affect the outcome of patients with acute leukemia
or MDS undergoing myeloablative hematopoietic stem cell transplantation (87).
The lower survival was mainly attributed to a significant increase in treatment-
related mortality of varying etiology. Although this was a retrospective study, these
findings strongly indicate that iron overload plays an important role in transplan-
tation outcome in patients with MDS, as it does in patients with thalassemia.
A significant improvement in erythropoietic function has been reported in
a small proportion of MDS patients receiving effective iron chelation (88). The
available evidence is not sufficient to determine the magnitude of such an effect
and its biological bases, but a broader use of iron chelation therapy in clinical
practice in the future should improve understanding of this phenomenon and its
clinical relevance.
Diagnosis and Monitoring of Iron Overload

The reference method for determining body iron stores is the measurement of
hepatic iron concentration (89). However, assessment of this has traditionally
required a percutaneous liver biopsy, which is difficult to implement in the clin-
ical management of patients with MDS because of concomitant neutropenia or
thrombocytopenia.
Serum ferritin is the most commonly used indirect estimate of iron over-
load. The interpretation of single serum ferritin values can be influenced by various
factors, including infection, inflammation, or liver disease. However, sequential
measurements of serum ferritin level have been shown to be useful to monitor sec-
ondary iron overload in transfusion-dependent patients and both prevent toxicity
and predict outcome (90,91). In patients with MDS, a direct correlation between
serum ferritin levels and the number of red blood cell transfusions received has
been observed, with a level of about 1000 ng/mL reached after a median of 21 red
blood cell transfusions (5).
Assessment of nontransferrin-bound iron is a potentially useful approach
that allows toxic iron levels to be estimated. However, the methods for determining
this fraction of body iron and its precise prognostic significance require further
investigation.
Magnetic resonance imaging (MRI) appears to be a promising noninvasive
tool for measuring tissue iron content (92). Liver iron content measured using
MRI is directly related to liver iron content measured in biopsy samples (93,94).
In addition, cardiovascular MRI could potentially be used not only to determine
myocardial iron content, but also cardiac function (93,95). This makes cardiovas-
cular MRI potentially a very useful tool to investigate the effects of iron-mediated
organ damage in patients with MDS. Recently, two small studies were published
which measured cardiac iron in MDS using this technique (96,97). Neither study
demonstrated cardiac iron accumulation in transfused patients. However, in both
studies, only a minority of heavily transfused patients were no longer undergoing
iron chelation therapy. Therefore, for the moment, the risk of cardiac iron deposits
in transfusion-dependent patients who are not receiving iron chelation therapy
cannot be excluded.
166 Malcovati
REFERENCES
3. Parker JE, Mufti GJ, Rasool F, et al. The role of apoptosis, proliferation, and the Bcl-
2-related proteins in the myelodysplastic syndromes and acute myeloid leukemia
secondary to MDS. Blood 2000; 96:3932–3938.
4. Cazzola M, Barosi G, Gobbi PG, et al. Natural history of idiopathic refractory
5. Malcovati L, Della Porta M, Pascutto C, et al. Prognostic factors and life expectancy
in myelodysplastic syndromes classified according to WHO criteria. A basis for
clinical decision-making. J Clin Oncol 2005; 23:7594–7603.
6. Cazzola M, Barosi G, Berzuini C, et al. Quantitative evaluation of erythropoietic
activity in dysmyelopoietic syndromes. Br J Haematol 1982; 50:55–62.
7. Cazzola M, Barosi G, Bergamaschi G, et al. Iron loading in congenital dysery-
thropoietic anaemias and congenital sideroblastic anaemias. Br J Haematol 1983;
54:649–654.
8. Ganz T. Hepcidin, a key regulator of iron metabolism and mediator of anemia of
inflammation. Blood 2003; 102:783–788.
9. Nemeth E, Ganz T. Hepcidin and iron-loading anemias. Haematologica 2006;
91:727–732.
10. Ganz T. Hepcidin and its role in regulating systemic iron metabolism. Hematology
Am Soc Hematol Educ Program 2006; 507:29–35.
11. Nemeth E, Tuttle MS, Powelson J, et al. Hepcidin regulates cellular iron efflux by
binding to ferroportin and inducing its internalization. Science 2004; 306:2090–
2093.
12. Detivaud L, Nemeth E, Boudjema K, et al. Hepcidin levels in humans are corre-
lated with hepatic iron stores, hemoglobin levels, and hepatic function. Blood 2005;
106:746–748.
13. Lin L, Valore EV, Nemeth E, et al. Iron transferrin regulates hepcidin synthesis in pri-
mary hepatocyte culture through hemojuvelin and BMP2/4. Blood 2007; 110:2182–
2189.
14. Pak M, Lopez MA, Gabayan V, et al. Suppression of hepcidin during anemia requires
erythropoietic activity. Blood 2006; 108:3730–3735.
15. Origa R, Galanello R, Ganz T, et al. Liver iron concentrations and urinary hepcidin
in beta-thalassemia. Haematologica 2007; 92:583–588.
16. Gardenghi S, Marongiu MF, Ramos P, et al. Ineffective erythropoiesis in beta-
thalassemia is characterized by increased iron absorption mediated by down-
regulation of hepcidin and up-regulation of ferroportin. Blood 2007; 109:5027–5035.
17. Cazzola M, Beguin Y, Bergamaschi G, et al. Soluble transferrin receptor as a potential
determinant of iron loading in congenital anaemias due to ineffective erythropoiesis.
Br J Haematol 1999; 106:752–755.
18. Kattamis A, Papassotiriou I, Palaiologou D, et al. The effects of erythropoetic activity
and iron burden on hepcidin expression in patients with thalassemia major. Haema-
tologica 2006; 91:809–812.
19. Tanno T, Bhanu NV, Oneal PA, et al. High levels of GDF15 in thalassemia suppress
expression of the iron regulatory protein hepcidin. Nat Med 2007; 13:1096–1101.
20. Ponka P, Schulman HM. Regulation of heme synthesis in erythroid cells: Hemin
inhibits transferrin iron utilization but not protoporphyrin synthesis. Blood 1985;
65:850–857.
21. Björkman SE. Chronic refractory anemia with sideroblastic bone marrow; a study of
four cases. Blood 1956; 11:250–259.
22. Caroli J, Bernard J, Bessis M, et al. Hémochromatose avec anémie hypochrome et
absence d’hémoglobine anormal. Presse Med 1957; 65:1991–1996.
23. Cazzola M, Invernizzi R, Bergamaschi G, et al. Mitochondrial ferritin expression in
erythroid cells from patients with sideroblastic anemia. Blood 2003; 101:1996–2000.
24. Levi S, Corsi B, Bosisio M, et al. A human mitochondrial ferritin encoded by an
intronless gene. J Biol Chem 2001; 276:24437–24440.
25. Langlois d’Estaintot B, Santambrogio P, Granier T, et al. Crystal structure and bio-
chemical properties of the human mitochondrial ferritin and its mutant Ser144Ala. J
Mol Biol 2004; 340:277–293.
26. Mollin DL. Sideroblasts and sideroblastic anaemia. Br J Haematol 1965; 11:41–48.
27. Cartwright GE, Deiss A. Sideroblasts, siderocytes, and sideroblastic anemia. N Engl
J Med 1975; 292:185–193.
28. Cox TC, Bottomley SS, Wiley JS, et al. X-linked pyridoxine-responsive sideroblastic
anemia due to a Thr388-to-Ser substitution in erythroid 5-aminolevulinate synthase.
N Engl J Med 1994; 330:675–679.
classification of the myeloid neoplasms. Blood 2002; 100:2292–2302.
31. Gattermann N, Aul C, Schneider W. Risk of leukemic transformation in two types of
acquired idiopathic sideroblastic anemia. Haematol Blood Transfus 1990; 33:374–
381.
33. Germing U, Gattermann N, Aivado M, et al. Two types of acquired idiopathic siderob-
lastic anaemia (AISA): a time-tested distinction. Br J Haematol 2000; 108:724–728.
35. Germing U, Strupp C, Kundgen A, et al. No increase in age-specific incidence of
36. Aul C, Bowen DT, Yoshida Y. Pathogenesis, etiology and epidemiology of myelodys-
plastic syndromes. Haematologica 1998; 83:71–86.
37. Kramer MF, Gunaratne P, Ferreira GC. Transcriptional regulation of the murine
erythroid-specific 5-aminolevulinate synthase gene. Gene 2000; 247:153–166.
38. Hofer T, Wenger RH, Kramer MF, et al. Hypoxic up-regulation of erythroid 5-
aminolevulinate synthase. Blood 2003; 101:348–350.
39. Cotter PD, May A, Li L, et al. Four new mutations in the erythroid-specific 5-
aminolevulinate synthase (ALAS2) gene causing X-linked sideroblastic anemia:
Increased pyridoxine responsiveness after removal of iron overload by phlebotomy
and coinheritance of hereditary hemochromatosis. Blood 1999; 93:1757–1769.
168 Malcovati
40. Cazzola M, May A, Bergamaschi G, et al. Absent phenotypic expression of X-linked

sideroblastic anemia in one of 2 brothers with a novel ALAS2 mutation. Blood 2002;
100:4236–4238.
41. Bekri S, May A, Cotter PD, et al. A promoter mutation in the erythroid-specific
5-aminolevulinate synthase (ALAS2) gene causes X-linked sideroblastic anemia.
Blood 2003; 102:698–704.
42. Cazzola M, May A, Bergamaschi G, et al. Familial-skewed X-chromosome inactiva-
tion as a predisposing factor for late-onset X-linked sideroblastic anemia in carrier
females. Blood 2000; 96:4363–4365.
43. Cotter PD, May A, Fitzsimons EJ, et al. Late-onset X-linked sideroblastic anemia.
Missense mutations in the erythroid delta-aminolevulinate synthase (ALAS2) gene
in two pyridoxine-responsive patients initially diagnosed with acquired refractory
anemia and ringed sideroblasts. J Clin Invest 1995; 96:2090–2096.
44. Bekri S, Kispal G, Lange H, et al. Human ABC7 transporter: Gene structure and muta-
tion causing X-linked sideroblastic anemia with ataxia with disruption of cytosolic
iron–sulfur protein maturation. Blood 2000; 96:3256–3264.
45. Muraki K, Sakura N, Ueda H, et al. Clinical implications of duplicated mtDNA in
Pearson syndrome. Am J Med Genet 2001; 98:205–209.
46. Shanske S, Tang Y, Hirano M, et al. Identical mitochondrial DNA deletion in a
woman with ocular myopathy and in her son with Pearson syndrome. Am J Hum
Genet 2002; 71:679–683.
47. Casas K, Bykhovskaya Y, Mengesha E, et al. Gene responsible for mitochondrial
myopathy and sideroblastic anemia (MSA) maps to chromosome 12q24.33. Am J
Med Genet A 2004; 127:44–49.
48. Bykhovskaya Y, Casas K, Mengesha E, et al. Missense mutation in pseudouridine syn-
thase 1 (PUS1) causes mitochondrial myopathy and sideroblastic anemia (MLASA).
Am J Hum Genet 2004; 74:1303–308.
49. Steensma DP, Hecksel KA, Porcher JC, et al. Candidate gene mutation analysis in
idiopathic acquired sideroblastic anemia (refractory anemia with ringed sideroblasts).
Leuk Res 2007; 31:623–628.
50. May A, de Souza P, Barnes K, et al. Erythroblast iron metabolism in sideroblastic
marrows. Br J Haematol 1982; 52:611–621.
51. Hellstrom-Lindberg E, Schmidt-Mende J, Forsblom AM, et al. Apoptosis in refrac-
tory anaemia with ringed sideroblasts is initiated at the stem cell level and associated
with increased activation of caspases. Br J Haematol 2001; 112:714–726.
52. Sakura T, Murakami H, Saitoh T, et al. Ultrastructural abnormalities of bone marrow
erythroblasts in refractory anemia. Ultrastruct Pathol 1998; 22:173–180.
53. van de Loosdrecht AA, Brada SJ, Blom NR, et al. Mitochondrial disruption and
limited apoptosis of erythroblasts are associated with high risk myelodysplasia. An
ultrastructural analysis. Leuk Res 2001; 25:385–393.
54. Gattermann N, Wulfert M, Junge B, et al. Ineffective hematopoiesis linked with a
mitochondrial tRNA mutation (G3242 A) in a patient with myelodysplastic syndrome.
Blood 2004; 103:1499–1502.
55. Shin MG, Kajigaya S, Levin BC, et al. Mitochondrial DNA mutations in patients
56. Broker S, Meunier B, Rich P, et al. MtDNA mutations associated with sideroblastic
anaemia cause a defect of mitochondrial cytochrome c oxidase. Eur J Biochem 1998;
258:132–138.
57. Reddy PL, Shetty VT, Dutt D, et al. Increased incidence of mitochondrial cytochrome
c-oxidase gene mutations in patients with myelodysplastic syndromes. Br J Haematol
2002; 116:564–575.
58. Tehranchi R, Fadeel B, Forsblom AM, et al. Granulocyte colony-stimulating factor
inhibits spontaneous cytochrome c release and mitochondria-dependent apoptosis of
myelodysplastic syndrome hematopoietic progenitors. Blood 2003; 101:1080–1086.
59. Tehranchi R, Invernizzi R, Grandien A, et al. Aberrant mitochondrial iron distribution
and maturation arrest characterize early erythroid precursors in low-risk myelodys-
60. Della Porta MG, Malcovati L, Gallı̀ A, et al. Mitochondrial ferritin expression and
clonality of hematopoiesis in patients with refractory anemia with ringed sideroblasts.
Blood 2005; 106:3444.
62. Shaw GR. Ringed sideroblasts with thrombocytosis: An uncommon mixed myelodys-
plastic/myeloproliferative disease of older adults. Br J Haematol 2005; 131:180–184.
dition characterized by JAK2 V617 F mutation. Blood 2006; 108:2173–2181.
64. Kralovics R, Passamonti F, Buser AS, et al. A gain-of-function mutation of JAK2 in
myeloproliferative disorders. N Engl J Med 2005; 352:1779–1790.
65. James C, Ugo V, Le Couedic JP, et al. A unique clonal JAK2 mutation leading to
constitutive signalling causes polycythaemia vera. Nature 2005; 434:1144–1148.
66. Baxter EJ, Scott LM, Campbell PJ, et al. Acquired mutation of the tyrosine kinase
JAK2 in human myeloproliferative disorders. Lancet 2005; 365:1054–1061.
67. Tefferi A, Lasho TL, Gilliland G. JAK2 mutations in myeloproliferative disorders. N
Engl J Med 2005; 353:1416–1417.
68. Skoda R. The genetic basis of myeloproliferative disorders. Hematology Am Soc
Hematol Educ Program 2007; 2007:1–10.
69. Campbell PJ, Green AR. The myeloproliferative disorders. N Engl J Med 2006;
355:2452–2466.
70. Pardanani AD, Levine RL, Lasho T, et al. MPL515 mutations in myeloproliferative
and other myeloid disorders: A study of 1182 patients. Blood 2006; 108:3472–3476.
71. Pikman Y, Lee BH, Mercher T, et al. MPLW515 L is a novel somatic activating
mutation in myelofibrosis with myeloid metaplasia. PLoS Med 2006; 3:e270.
72. Ceesay MM, Lea NC, Ingram W, et al. The JAK2 V617 F mutation is rare in RARS
but common in RARS-T. Leukemia 2006; 20:2060–2061.
73. Gattermann N, Billiet J, Kronenwett R, et al. High frequency of the JAK2 V617 F
mutation in patients with thrombocytosis (platelet count ⬎ 600 × 109/L) and ringed
109:1334–1335.
74. Boissinot M, Garand R, Hamidou M, et al. The JAK2-V617 F mutation and essential
thrombocythemia features in a subset of patients with refractory anemia with ring
sideroblasts (RARS). Blood 2006; 108:1781–1782.
75. Renneville A, Quesnel B, Charpentier A, et al. High occurrence of JAK2 V617 muta-
tion in refractory anemia with ringed sideroblasts associated with marked thrombo-
cytosis. Leukemia 2006; 20:2067–2070.
170 Malcovati
76. Schnittger S, Bacher U, Haferlach C, et al. Detection of an MPLW515 mutation in

a case with features of both essential thrombocythemia and refractory anemia with
ringed sideroblasts and thrombocytosis. Leukemia 2008; 22:453–5.
77. Schmitt-Graeff AH, Soon-Siong T, Olschewski M, et al. The JAK2-V617 F mutation
status identifies subtypes of refractory anemia with ringed sideroblasts associated
with marked thrombocytosis. Haematologica 2008; 93:34–40.
78. Malcovati L, Della Porta MG, Pietra D, et al. Granulocyte JAK2 (V617 F) mutation
status in myeloid neoplasms with ringed sideroblasts. Blood 2006; 108:854.
79. Alessandrino EP, Amadori S, Barosi G, et al. Evidence- and consensus-based practice
guidelines for the therapy of primary myelodysplastic syndromes. A statement from
the Italian Society of Hematology. Haematologica 2002; 87:1286–1306.
81. Olivieri NF. The beta-thalassemias. N Engl J Med 1999; 341:99–109.
82. Kushner JP, Porter JP, Olivieri NF. Secondary iron overload. Hematology Am Soc
Hematol Educ Program 2001:47–61.
83. Olivieri NF, Brittenham GM. Iron-chelating therapy and the treatment of thalassemia.
Blood 1997; 89:739–761.
84. Buja LM, Roberts WC. Iron in the heart. Etiology and clinical significance. Am J
Med 1971; 51:209–221.
85. Schafer AI, Cheron RG, Dluhy R, et al. Clinical consequences of acquired transfu-
sional iron overload in adults. N Engl J Med 1981; 304:319–324.
86. Malcovati L, Della Porta MG, Cazzola M. Predicting survival and leukemic evolution
in patients with myelodysplastic syndrome. Haematologica 2006; 91:1588–1590.
87. Armand P, Kim HT, Cutler CS, et al. Prognostic impact of elevated pre-transplant
serum ferritin in patients undergoing myeloablative stem cell transplantation. Blood
Prepublished online January 2007:18.
88. Jensen PD, Heickendorff L, Pedersen B, et al. The effect of iron chelation on
haemopoiesis in MDS patients with transfusional iron overload. Br J Haematol 1996;
94:288–299.
89. Angelucci E, Brittenham GM, McLaren CE, et al. Hepatic iron concentration and
total body iron stores in thalassemia major. N Engl J Med 2000; 343:327–331.
90. Olivieri NF, Nathan DG, MacMillan JH, et al. Survival in medically treated patients
with homozygous beta-thalassemia. N Engl J Med 1994; 331:574–578.
91. Porter JB. Practical management of iron overload. Br J Haematol 2001; 115:239–
252.
92. Cohen AR, Galanello R, Pennell DJ, et al. Thalassemia. Hematology Am Soc Hematol
Educ Program 2004:14–34.
93. Anderson LJ, Holden S, Davis B, et al. Cardiovascular T2-star (T2∗ ) magnetic reso-
nance for the early diagnosis of myocardial iron overload. Eur Heart J 2001; 22:2171–
2179.
94. St Pierre TG, Clark PR, Chua-anusorn W, et al. Noninvasive measurement and
imaging of liver iron concentrations using proton magnetic resonance. Blood 2005;
105:855–861.
95. Anderson LJ, Wonke B, Prescott E, et al. Comparison of effects of oral deferiprone
and subcutaneous desferrioxamine on myocardial iron concentrations and ventricular
function in beta-thalassaemia. Lancet 2002; 360:516–520.
96. Konen E, Ghoti H, Goitein O, et al. No evidence for myocardial iron overload
in multitransfused patients with myelodysplastic syndrome using cardiac magnetic
resonance T2 technique. Am J Hematol 2007; 82:1013–1016.
97. Chacko J, Pennell DJ, Tanner MA, et al. Myocardial iron loading by magnetic
resonance imaging T2∗ in good prognostic myelodysplastic syndrome patients on
long-term blood transfusions. Br J Haematol 2007; 138:587–593.
8
Therapy-Related Myelodysplastic
Syndrome and Myeloid Leukemia
Lucy A. Godley and Richard A. Larson

Department of Medicine and Cancer Research Center, University of Chicago,
Chicago, Illinois, U.S.A.
INTRODUCTION
Therapy-related myeloid leukemia (t-MDS/t-AML) is a well-recognized clinical
syndrome occurring as a late complication following cytotoxic therapy (1–4).
The term “therapy-related” leukemia is based on a patient’s history of exposure
to cytotoxic agents. Although a causal relationship is implied, the mechanism
remains to be proven. These neoplasms are thought to be the direct consequence
of mutational events induced by the prior therapy. Table 1 shows the various
primary diagnoses and primary cytotoxic therapies received by 306 patients with
therapy-related myeloid leukemia studied at the University of Chicago (3).
RISK FACTORS FOR THE DEVELOPMENT OF THERAPY-RELATED

LEUKEMIAS
The etiology and specific factors that predispose to the development of therapy-
related leukemia have been difficult to study, and it has not yet been possible to
determine whether the development of t-MDS/t-AML is a stochastic event, occur-
ring by chance, or whether certain individuals are at higher risk. The development
of new technologies that allow the rapid sequencing of a large number of single
nucleotide polymorphisms (SNPs) has been applied to t-MDS/t-AML, and the
results suggest that at least a fraction of patients probably have a heritable predis-
position, such as altered drug metabolism or DNA repair [Table 2; and reviewed
173
174 Godley and Larson
Table 1 Primary Diagnoses and Primary Cytotoxic Therapies in 306 Patients Who
Developed Therapy-Related Myeloid Leukemia: The University of Chicago Series (3)
Combined
No. of Chemotherapy Radiotherapy modality
Primary diagnosis patients only (%) only (%) therapy (%)
No prior malignancy 18 12 (67)a 2 (11) 4 (22)

Hematologic 171 69 (40) 5 (3) 97 (57)
malignancy
Hodgkin lymphoma 77 18 (23) 4 (5) 55 (71)
Non-Hodgkin 70 33 (47) 1 (1) 36 (51)
lymphoma
Myeloma 23 17 (74) 0 6 (26)
Other 1 1 (100) 0 0
Solid tumors 117 40 (35) 36 (32) 38 (33)
Breast 32b 11 (35) 5 (16) 15 (48)
Ovary 15 12 (80) 1 (7) 2 (13)
Prostate 13b 0 11 (100) 0
Lung 9 5 (56) 2 (22) 2 (22)
Cervix 7 0 4 (57) 3 (43)
Other 41 12 (30) 13 (32) 16 (39)
Totals 306b 121 (40) 43 (14) 139 (46)

a Numbers in parentheses are percentages for each row of data according to primary therapy.
b In 3 patients, the treatment for the primary tumor was not fully known.
in Ref. 5]. Widespread identification of underlying preexisting conditions would

help inform the choice of initial therapy as well as the screening and counseling
of patients at the time of treatment for their primary disease.
We have previously reported that an inactivating polymorphism (187Ser)
in the gene encoding NAD(P)H:quinone oxidoreductase, NQO1, is increased
among individuals with t-AML (6). Homozygotes may be particularly vulner-
able to leukemogenic changes induced by carcinogens, and heterozygotes are at
risk for treatment-induced mutation or loss of the remaining wild-type allele in
their hematopoietic stem cells. Bolufer and colleagues examined the prevalence
of several genetic polymorphisms and found that the NQO1-187Ser polymor-
phism conferred a 2-fold increased risk of developing t-AML (7). This risk was
increased when the NQO1-187Ser polymorphism was found in combination with
CYP1A1∗ 2A and del[GSTT1]. However, the risk of developing t-AML was 18-
fold lower among patients lacking all three polymorphisms. Other polymorphisms
involving detoxifying enzymes have also been reported (8). A large Japanese study
of patients with AML de novo and t-AML found that the NQO1 polymorphism
was more strongly associated with t-AML than polymorphisms in GST-M1, GST-
T1, and CYP3A4 (9). Furthermore, patients carrying the NQO1-187Ser allele who
had been exposed to chemotherapy had significantly shorter telomeres in their
Table 2 Factors That May Predispose to the Development of t-MDS/t-AML
Allele(s) predisposing to
Gene/protein t-MDS/t-AML SNP results in Frequency in t-MDS/t-AML patients Reference(s)
NQO1/NAD(P)H:quinone 187Ser Inactivation of NQO1 Increased; confers a 2-fold increased (6,7)

oxidoreductase risk of developing t-MDS/t-AML
MTHFR/methylene 677T/1298A Unknown Increased risk of developing (11)
tetrahydrofolate reductase t-MDS/t-AML after treatment for
breast cancer
677C/1298C Unknown Increased risk of developing (11)
t-MDS/t-AML after treatment for
a hematopoietic malignancy
hMSH2 −6 exon13C SNP within Overrepresented in patients who had (13)
intron splice acceptor received O(6)-guanine alkylating
agents
hMLH1 −93A SNP of the basal Increased in t-AML patients who had (14)
hMLH1 promoter received a methylating
chemotherapy agent
MDM2; TP53 MDM2 SNP 309; TP53 MDM2: alters SP1 binding; Increased risk of t-MDS/t-AML in (16)
Arg72Pro TP53: changes protein patients with MDM2 TT/TP53
Therapy-Related Myelodysplastic Syndrome and Myeloid Leukemia
function ArgArg; MDM2 G/TP53 Pro; and

TP53 ProPro who received
radiotherapy
175
neutrophils and lymphocytes and were more likely to develop clonal hematopoiesis
than patients with wild-type NQO1 alleles (10). These findings provide a molecular
link between NQO1 genotype and an increased risk of developing t-AML.
Guillem and colleagues identified a haplotype in MTHFR, the gene encoding
methylene tetrahydrofolate reductase, which conferred an increased risk in partic-
ular patient populations (11). Two SNPs were included in the haplotype: 677C/T
and 1298A/C. An increased risk of developing t-AML was associated with the
677T/1298A haplotype in breast cancer patients and the 677C/1298C haplotype
in patients with a primary hematopoietic malignancy.
Several groups have examined the genes-encoding components of DNA
repair pathways. Although about half of t-AML cases show microsatellite insta-
bility, promoter hypermethylation and transcriptional silencing of the hMSH2 and
hMLH1 genes are not common events (12). Instead, other genetic alterations of
these genes seem to predispose to t-AML. Patients with t-AML who have been pre-
viously treated with O(6)-guanine alkylating agents, such as cyclophosphamide
and procarbazine, have an increased frequency of a variant C SNP that occurs
within an intron splice acceptor of the hMSH2 gene (13). Additionally, 2 of 13
cases that exhibited microsatellite instability were homozygous for the C allele, a
frequency much higher than that observed in a control population. Furthermore, a
variant SNP at position −93 of the hMLH1 promoter was found in 75% of patients
with t-AML who had received methylating chemotherapy as part of prior therapy
for Hodgkin disease (14). In contrast, this variant SNP was found in only 30%
of patients with t-AML without prior exposure to methylating agents. In patients
who had been treated with a methylating agent, the presence of the variant −93
SNP conferred a significantly increased risk of developing t-AML, with an odds
ratio of 5.3.
The TP53 tumor suppressor gene directs cellular responses to many DNA
damaging agents. Patients with Li-Fraumeni syndrome who have germline muta-
tions in one TP53 allele are at increased risk for developing t-AML (15). Recently,
we reported that constitutional genetic variations in the p53 pathway affect t-
AML risk (16). We tested associations between patients with t-AML (N = 171)
and two common variants within the functional TP53 pathway, a SNP at position
309 within MDM2 a gene that encodes a regulator of p53 function, and the TP53
codon 72 polymorphism. The MDM2 SNP309 polymorphism is located within a
binding site for the SP1 transcription factor in the MDM2 core promoter. SP1 binds
more effectively to the G allele compared to the T allele at position 309 (17–19).
Transcription of MDM2 is increased when the promoter contains the G allele,
resulting in lower basal levels of TP53. The TP53 Arg72Pro polymorphism alters
the ability of the TP53 protein to induce apoptosis versus cell-cycle arrest (20–22).
Although neither polymorphism alone influenced risk of t-AML, an interactive
effect was detected such that MDM2 TT TP53 Arg/Arg double homozygotes,
and individuals carrying both a MDM2 G allele and a TP53 Pro allele are at
increased risk of t-AML (for interaction, p = 0.009). In addition, the risk of devel-
oping t-AML was 2.7-fold higher in TP53 Pro/Pro homozygotes who received
Therapy-Related Myelodysplastic Syndrome and Myeloid Leukemia 177
radiotherapy compared to TP53 Arg/Arg homozygotes (p = 0.04). These data

indicate that the MDM2 and TP53 variants interact to modulate responses to
genotoxic therapy and are determinants of risk in t-AML.
Inherited mutations of other genes are also known to predispose to the
development of t-AML. For example, children with neurofibromatosis type 1
who have germline NF1 mutations that result in altered RAS signaling are at an
increased risk of developing t-MDS/t-AML (23). Furthermore, mice carrying one
inactive Nf1 allele are susceptible to t-AML after treatment with alkylating agents
(24). In addition, patients with Fanconi anemia, who have mutations in 1 of 11
genes that encode proteins involved in DNA damage and repair, are at increased
risk of developing t-MDS/t-AML (25). We look forward to future studies that will
shed additional light on the genetic contribution to t-MDS/t-AML susceptibility.
SUBTYPES OF THERAPY-RELATED MYELOID LEUKEMIA

Neither t-MDS nor t-AML is easily categorized according to the French-American-
British (FAB) classification schema, but the World Health Organization (WHO)
classification recognizes them as a distinct entity (26). Morphologically, t-MDS/t-
AML most closely resembles AML with multilineage dysplasia, also a distinct
form of de novo AML within the WHO classification. t-AML is distinguished
from t-MDS solely on the blast count being ⬎20% in either the peripheral blood
or the bone marrow. Based on clinical, morphological, and genetic features, t-MDS
and t-AML are a single disease, each representing different ends of the leukemic
spectrum. The characteristics of therapy-related leukemia and the timing of its
development after a primary diagnosis depend on the exposure to specific agents
as well as the cumulative dose and dose intensity of the preceding cytotoxic therapy.
In the classic form of therapy-related leukemia that follows treatment with
alkylating agents and/or radiation therapy, the clinical features resemble those seen
in primary MDS, although the degree of dysgranulopoiesis and dysmegakaryocy-
topoiesis is typically greater. Fatigue, weakness, and occasionally fever are the
most frequent initial patient complaints. Anemia, often macrocytic, and throm-
bocytopenia are extremely common, and an increased mean corpuscular volume
(MCV) is often the first clue to the diagnosis. Leukopenia may also be present.
Dysplastic changes are often observed in all the three cell lines (Fig. 1). Mild-to-
marked reticulin fibrosis may be present. Auer rods are rarely seen, and myeloper-
oxidase and nonspecific esterase reactivity are often only weakly expressed.
Clonal chromosome abnormalities, often of a complex nature, are identified
in most cases of classical therapy-related leukemia (1–3,27–29). Loss of part or
all of chromosomes 5 and/or 7 are the characteristic findings, and have been
reported in over 70% of cases in some series (3). The most common single
abnormality is monosomy 7, followed in frequency by deletion of the long arm of
chromosome 5 [del(5q)] and by monosomy 5. These same abnormalities are also
observed in primary MDS and AML de novo, especially in older patients and those
with occupational exposure to potential carcinogens such as benzene. Alkylating
Dysplasia seen in:

White blood cells:
Hypolobation
Hypercondensed chromatin
A B Hypogranular cytoplasm
Red blood cells:

Megaloblastoid changes
Nuclear-cytoplasmic dyssynchrony
Hypercondensed chromatin
Basophilic stippling
C D
Megakaryocytes:
Micromegakaryocytes
Separation of nuclear lobes
Hypogranular cytoplasm
E F
Figure 1 Characteristic but nonspecific cytologic changes are observed in therapy-related

myeloid leukemias. Trilineage dysplasia seen in the peripheral blood and bone marrow from
a 38-year-old male survivor of Hodgkin disease, who had been treated with chemother-
apy, radiation, and an autologous stem-cell transplant. The patient was asymptomatic until
presentation with a macrocytic anemia and thrombocytopenia, 5 years after his diagnosis
of Hodgkin lymphoma and 3 years after autologous stem-cell transplant. A bone mar-
row biopsy revealed trilineage dysplasia and fewer than 5% myeloblasts, consistent with a
diagnosis of therapy-related myelodysplasia. Karyotype analysis revealed a loss of chromo-
some 7, consistent with alkylating agent–induced disease. (A–C) Dysplasia in the periph-
eral blood. (A,B) Dysplastic neutrophils have hypolobated, hypercondensed chromatin,
and hypogranular cytoplasm. The cell shown in (A) has pseudo-Pelger Huet features. (C)
A multinucleated abnormal erythroid precursor in the peripheral blood. (D–G) Dyspla-
sia in the bone marrow. (D) An abnormal multinucleated erythroid cell with basophilic
stippling (arrowheads). (E,F) Micromegakaryocytes with abnormal separation of nuclear
lobes. Source: Courtesy of Dr. Ozden Ozer, Department of Pathology, The University of
Chicago.
agents vary in their likelihood of causing the development of therapy-related

disease (melphalan ⬎ cyclophosphamide) (30,31), and there is a dose–response
relationship between the amount of alkylating agent received and the risk of disease
development (29). This form of t-MDS/t-AML typically occurs within 5 to 7 years
after the initial chemotherapy and/or radiotherapy have been given, and the risk
appears to diminish considerably by 10 years after the last exposure. Response to
treatment is poor, and survival is short.
Therapy-related leukemia following chemotherapy with topoisomerase II
inhibitors is characterized by translocations involving chromosome bands 11q23
or 21q22 (32,33). Balanced translocations may involve the MLL gene at chromo-
some band 11q23, leading to a t(9;11). The fusion gene PML/RARA is observed
in therapy-related acute promyelocytic leukemia (t-APL). Rearrangements of the
core binding factor genes AML1 (RUNX1/CBFA2) at chromosome band 21q22
and CBFB at chromosome band 16q22, as well as the NUP98 gene at chromo-
some band 11p15.5 have also been described. The rearrangements inv(16) and
t(15;17) in particular have been observed in patients who had previously received
only radiotherapy (32). In contrast to alkylating agent–associated t-AML, these
leukemias are rarely preceded by t-MDS. They occur with a shorter latency, often
within 2 to 3 years of the first cytotoxic therapy and, in some cases, within
12 months. These t-AMLs often present with rapidly progressive leukemia and
high white blood cell counts. Although they also have a poor prognosis overall,
they are more responsive to initial remission induction chemotherapy.
In addition to these two major classes of t-MDS/t-AML, patients treated
with fludarabine, an immunosuppressive nucleoside analog, are also at risk for
t-AML (34). This raises the possibility that suppression of the immune system
may play a role in the emergence of t-AML. Out of 521 patients treated for
chronic lymphocytic lymphoma (CLL) with fludarabine alone or in combination
with chlorambucil, 6 (1.2%) developed t-AML, many of whom also had charac-
teristic abnormalities of chromosomes 5 and 7 (35). In another study, 8 of 202
CLL patients (4%) treated with fludarabine, mitoxantrone, and dexamethasone
developed t-AML, often with abnormalities of chromosome 7, between 1 and 5
years after therapy (36). Another recent report describes a patient with grade 2
follicular lymphoma who was treated with fludarabine, cyclophosphamide, and rit-
uximab who developed t-AML with a chromosomal translocation involving 11q23
40 months after the initial therapy (37). New compounds for the treatment of CLL
include radioimmunotherapeutic agents, such as yttrium-90 ibritumomab tiuxetan
and iodine-131 tositumomab, raising concerns about an increased incidence of
t-AML after these agents, which add radiation to chemotherapy and might further
damage DNA within bone marrow cells. Reassuringly, however, Czuczman and
colleagues reported recently that the incidence of t-AML among non-Hodgkin
lymphoma patients treated with yttrium-90 ibritumomab tiuxetan was 2.5% at
4.4 years, consistent with the expected incidence of t-AML in a patient population
heavily pretreated with chemotherapy alone (38).
In addition to cancer patients, t-AML has been seen in patients following
treatment with immunosuppressive therapies not previously thought to cause DNA
damage directly, particularly in patients who have received solid organ transplants
(4). A mechanism for the development of t-AML has been proposed for azathio-
prine, an immunosuppressant widely used in recipients of organ transplantation,
through selection of a mutator phenotype to allow the emergence of AML with
abnormalities of chromosomes 5 and 7 (4). Thus, a growing body of work suggests
that the use of any compound that could damage DNA directly or suppress the
immune system’s ability to detect malignant cells could lead to an increased risk
of t-AML.
GENETIC PATHWAYS AND COOPERATING MUTATIONS IN THE

ETIOLOGY OF THERAPY-RELATED MYELOID LEUKEMIA
Particular mechanisms of DNA damage that lead either to chromosomal deletions
or balanced translocations may underlie the differences in latencies between the
two main forms of therapy-related leukemia (29,39). In the case of chromoso-
mal deletions, one allele of a putative tumor suppressor gene may be inactivated.
Before the affected cell could gain a proliferative advantage, however, the sec-
ond allele might also have to be deleted or mutated, or silenced by an epigenetic
event. In some cases, the first allele may be inactivated in the germline DNA
due to a heritable cause. More recent evidence suggests that haploinsufficiency of
individual genes such as EGR1 on chromosome 5q may also allow for malignant
transformation (40). Haploinsufficiency of RPS14 has been implicated in the eti-
ology of the refractory anemia with 5q− syndrome (41). However, even loss of
both alleles of an individual tumor suppressor may not be sufficient to confer a
malignant phenotype. As described in the model of colorectal tumorigenesis, mul-
tiple tumor suppressor genes or oncogenes may need to be mutated to ultimately
transform a cell. This series of genetic changes may require an extended period
of time, thus explaining the long latency of alkylator-induced t-AML. In contrast,
balanced chromosome translocations result in the activation of cellular oncogenes
in a dominant fashion. These rearrangements, such as those involving the MLL
gene at 11q23, may yield a fusion gene that acts as a dominant oncogene. Whereas
this fusion gene alone may not be sufficient to fully transform a hematopoietic
progenitor cell, relatively fewer genetic events may be required to progress to the
leukemic phenotype.
Pedersen-Bjergaard and his colleagues have proposed eight different genetic
pathways for the multi-step development of t-MDS/t-AML (Fig. 2) (42). There
is growing evidence that mutations in a limited number of molecular pathways
may cooperate in the genesis of leukemia. Gilliland and colleagues have described
AML as a disease arising from the cooperation between two types of alterations
(43). Changes in signaling molecules deregulate proliferative and/or anti-apoptotic
activities, for example, through mutations that constitutively activate receptor
tyrosine kinases like FLT3. In addition, fusion proteins encoded by recurring
chromosomal translocations, such as MLL/AF9 that results from the t(9;11), impair
normal differentiation pathways with little effect on cellular proliferation. Gene
expression array experiments with CD34+ t-AML cells have provided evidence to
support this hypothesis (44). Loss of TAL1, GATA1, and EKLF expression has been
observed, and these might result in impaired differentiation of hematopoietic cells,
whereas overexpression of FLT3, PIK3C2B, and BCL2 results in a proliferative
or survival advantage.
Additional evidence for the notion that t-AML arises from a combination
of mutations comes from studies that identify point mutations plus recurring
chromosomal rearrangements. One study examined three cases of t-AML with
t(8;21), and two of the patients were also found to be positive for the JAK2 V617F
Pathway t-MDS t-AML

p15 promotor
methylation
Alkylating Agents % of cases
Centromeric Breakage 0 50 100%
RAS mutations
AML 1
mutations p53 mutations
7q−/−7
n = 15
I 84%
n = 39
n = 16 p53 mutations 80%

5q−/−5 n = 25
II 17p−/−17
n = 34 7q−/−7 Complex karyotypes
n = 18 MLL-AML1 amplific. 91%
Topoisomerase II Inhibitors
Chimeric Fusion Genes
RAS mutations
11q23
III MLL BRAF mutations 50%
n = 11
cKIT mutations
7q−/−7 21q22 inv(16)
IV AML1 CBFB 83%
t(3;21) n = 10
FLT3 mutations
t(15;17)
V RARA 83%
n=2
11p15
VI NUP98 ? ?%
De Novo Cases?
NPM1 mutations
FLT3 mutations
Normal Normal
RAS mutations RAS mutations
VII karyotype karyotype
AML1 mutations 50%
n=9 n = 15
MLL ITD
Others Others
VIII n = 15 n=5 40%
Figure 2 Genetic pathways identified in 140 cases of t-MDS and t-AML by Pedersen-
Bjergaard et al. (42). At least eight genetic pathways could be distinguished among 140
cases of therapy-related myeloid malignancies. Pathways I and II are characteristic of
patients who had been treated previously with alkylating agents. Mutations of AML1, p53,
and RAS are common in patients with abnormalities of 7q or −7. Pathways III–VI are often
seen in patients who had received topoisomerase II inhibitors. By definition, any myeloid
leukemia arising after prior chemo- or radiotherapy is considered therapy related. However,
because there are few characteristic chromosomal changes or genetic mutations seen in
leukemias described by Pathways VII and VIII, these may actually represent serendipitous
de novo cases, or cases in which the underlying genetic defect has not yet been elucidated.
RAS mutations are commonly seen in the transition to t-AML within Pathway I, in t-AMLs
with MLL rearrangements (Pathway III), and in t-AMLs with a normal karyotype (Pathway
VII). Source: Adapted from Ref. 42.
mutation (45). Both the patients had received chemotherapy with anthracyclines
and topoisomerase inhibitors, one for breast cancer and one for testicular semi-
noma. Another group examined 140 unselected patients with t-MDS/t-AML and
found that 2 of the 89 t-MDS patients had the JAK2 V617F mutation (46). In
these cases, however, both the patients had received radiotherapy, presented with
myeloproliferative features including splenomegaly, and had a normal karyotype.
Notably, these features are unusual for t-MDS. These investigators went on to look
for PTPN11 mutations in this patient cohort and found four patients with such
mutations; three also had deletions or loss of chromosome 7q (47). In addition, all
four had rare balanced chromosomal translocations: two had breakpoints at 3q26
with rearrangement of EVI1, and two had breakpoints at 21q22 and RUNX1 rear-
rangements. Such cases provide additional evidence that t-AML arises through
multiple genetic events.
THERAPY-RELATED LEUKEMIA AFTER BREAST CANCER

Although long-term survivors of Hodgkin disease were among the first patients to
suffer the development of t-AML, this complication is now increasingly observed
in those treated for solid tumors. Several large studies have examined the risks for
women receiving adjuvant chemoradiotherapy for breast cancer. Several groups
have addressed the question as to whether the addition of granulocyte-colony
stimulating factor (G-CSF) to routine therapy regimens contributes to the risk of
developing t-AML out of the concern that G-CSF could promote the proliferation
of damaged cells, which might otherwise undergo apoptosis, and thus drive the
development of t-AML. In six trials completed by the National Surgical Adju-
vant Breast and Bowel Project, the incidence of therapy-related leukemia was
sharply elevated among patients receiving intensified doses of adriamycin and
cyclophosphamide that required G-CSF support—the relative risk (RR) was 6.16
(p = 0.0001) (48). Breast radiotherapy alone increased the RR to 2.38 (p = 0.006).
In a second study, 5510 women over 65-years old who received chemotherapy for
stage I–III breast cancer were analyzed (49). Out of the 906 patients, 16 (1.77%)
patients, who also received G-CSF or granulocyte-macrophage CSF, developed
t-AML compared to 1.04% of those patients who did not receive myeloid growth
factors. Importantly, the incidence of t-MDS was not evaluated. The hazard rate
was 2.59 (95%CI, 1.30–5.15) for the development of leukemia within 4 years.
Finally, a large case-control study from France compared 182 patients who devel-
oped therapy-related leukemia after breast cancer treatment with 534 matched
controls (50). The risk of leukemia was markedly increased after chemotherapy
that included a topoisomerase-II inhibitor (p ⬍ 0.0001), and was higher after
mitoxantrone (RR = 15.6) than after anthracyclines. The risk was increased 3.9-
fold after breast radiotherapy (p = 0.003). The risk was also increased among
those who received G-CSF (RR = 6.3, p = 0.0009), even when controlling for
chemotherapy doses.
THERAPY-RELATED LEUKEMIA AFTER AUTOLOGOUS HEMATOPOIETIC

CELL TRANSPLANTATION (HCT)
Several thousand autotransplants are performed each year for patients with
relapsed lymphoma and other diseases. Estimates of the incidence of therapy-
related leukemia among these lymphoma and Hodgkin disease patients range
between 1% and 14% at 3 to 15 years (51). The risk appears lower in patients
undergoing autologous HCT for breast or germ cell cancers or for myeloma.
Important risk factors include age, extent of prior therapy, and exposure to cer-
tain agents before and during the transplant procedure. Genotoxic damage and
proliferative stresses imposed on hematopoietic stem cells during the priming or
mobilizing chemotherapy and engraftment are also likely to play a role. Kalaycio
and colleagues studied the clinical parameters of 526 patients with lymphoma
who received autologous stem-cell transplants as part of their treatment and found
that requiring 5 or more days of apheresis to collect sufficient autologous stem
cells, prior exposure to radiation, or having received four or more chemotherapy
regimens prior to the autologous transplant were independent risk factors for the
development of t-MDS/t-AML (Fig. 3) (52). As noted above, inherited polymor-
phisms in genes governing drug metabolism and DNA repair may also contribute
to leukemogenesis.
Comparative studies of the latency periods between first cytotoxic exposure,
the autologous HCT itself, and the emergence of therapy-related leukemia suggest
that the initial malignant event occurs prior to HCT in most cases (3,51). In
addition, analyses of cryopreserved stem cells by fluorescence in situ hybridization
have on occasion revealed that clonal abnormalities were already present prior to
the HCT. However, the cytotoxic therapy delivered during the multistep HCT
procedure is likely additive to previous genomic damage and probably contributes
to the etiology by cooperating mutations and selective pressure.
FACTORS THAT INFLUENCE OUTCOME IN t-AML

Therapy-related myeloid leukemia is generally a fatal disease. The life-threatening
complications of this disorder are the result of persistent and profound cytopenias
due to the failure of normal hematopoiesis regardless of the fraction of myeloblasts
accumulating in the bone marrow or blood. There has been general agreement that
patients with t-AML have shorter survivals than patients with de novo AML.
Supportive care is still considered by many to be the standard management.
A number of potential factors explain the poor outcome of patients with
therapy-related leukemia. The persistence of the primary malignant disease, par-
ticularly metastatic cancer or lymphoma, causes morbidity and mortality inde-
pendent of the bone marrow failure caused by leukemia. Injury to organs and
their vascular supply from prior treatment may compromise the ability of these
patients to receive intensive remission induction chemotherapy or successful HCT.
There may be depletion of normal hematopoietic stem cells as a consequence of
(A) 40
p < 0.001
35
30
t-MDS/t-AML (%)
25
20
>5 days (n = 126)
15
10
5 2–3 days (n = 400)
0
0 1 2 3 4 5 6 7 8 9 10 11 12 13
Years post-transplant
(B) 40
p = 0.001
35
30
t-MDS/t-AML (%)
25
20
15 RT (n = 157)
10
No RT (n = 369)
5
0
0 1 2 3 4 5 6 7 8 9 10 11 12 13
(C) 40
p < 0.001
35
30
t-MDS/t-AML (%)
25
4–6 prior regimens (n = 34)
20
15
10
5
1–3 prior regimens (n = 492)
0
0 1 2 3 4 5 6 7 8 9 10 11 12 13
Figure 3 Risk factors for the development of t-AML after autologous stem cell trans-
plantation as a component of lymphoma therapy. Increased risk is seen in patients who
(A) required 5 or more days to collect sufficient autologous cells, (B) received radiation
therapy prior to transplantation, and (C) received 4 or more chemotherapy regimens prior
to transplantation. Source: Adapted from Ref. 52.
previous therapy, so that these patients suffer prolonged cytopenias after induction
chemotherapy. The bone marrow stroma may have been damaged, especially by
therapeutic radiation to fields that include the pelvis or lumbosacral spine, so that
it will not support regeneration of normal hematopoiesis. Patients with t-AML are
often chronically immunosuppressed from prior disease or on-going therapy or
may have dysfunctional phagocytes, and thus are often colonized with pathogenic
or antibiotic-resistant bacteria and fungi. Following prior supportive care, patients
may be refractory to additional transfusion support, and therefore, not good can-
didates for intensive myelosuppressive chemotherapy. Finally, the high frequency
of unfavorable cytogenetic aberrations arising during or after chemoradiotherapy
appears to result in the rapid emergence of chemotherapy resistance in t-AML
stem cells.
TREATMENT OF THERAPY-RELATED MYELOID LEUKEMIA

The survival of patients with therapy-related leukemia is often poor despite prompt
diagnosis and treatment. However, there is a paucity of prospective treatment data
since these patients are most often excluded from frontline clinical trials. There
are no randomized studies comparing standard AML therapy to other forms of
treatment. In a nationwide Japanese study of 256 patients with t-MDS (41%) or
t-AML (59%), a poor prognosis was associated with abnormalities of chromosome
5, hypoproteinemia, high C-reactive protein, thrombocytopenia, and persistence of
the primary malignancy (53). The majority of the Japanese patients (72%) received
antileukemia chemotherapy, either a standard combination using an anthracycline
plus cytarabine, or low-dose cytarabine, or tretinoin (ATRA) in the case of 7
patients with t-APL. The median age was 61 years, and the median survival was
only 9.7 months. A complete remission (CR) was seen in 85 patients (46%). The
median remission duration was 8.2 months.
Poor hematopoietic reserves make the administration of standard AML
therapy difficult. Many patients have poor tolerance for the acute toxicity of treat-
ment. Because therapy-related leukemia evolves in the milieu of chemotherapy,
the malignant cells are relatively drug-resistant. Expression of the multidrug resis-
tance phenotype is common. In a review of 644 t-AML patients treated with a
variety of standard AML chemotherapy regimens, only 182 (28%) achieved a CR
(54). Individual small series report CR rates of 40% to 50%. This is considerably
lower than the 55% to 80% CR rate observed in patients with de novo AML.
In addition, remissions are often short even when confirmed cytogenetically and
consolidated intensively (55).
TREATMENT OF t-AML WITH BALANCED CHROMOSOMAL

REARRANGEMENTS
In marked contrast to the overall poor outcome for t-AML, those patients who
develop t-APL with t(15;17) or those with t(8;21) or inv(16) have response rates
that are similar to patients with de novo AML with the same chromosomal
rearrangements. Nonetheless, overall survival is less because of the comorbid fac-

tors discussed above. In a report on 106 cases of t-APL identified between 1982
and 2001 in France, Spain, and Belgium, the characteristics of the t-APL patients
were similar to those with de novo APL (56). In addition, more than 80% of those
treated with anthracycline-based chemotherapy and/or ATRA achieved a CR. Ten
of the complete responders relapsed, and seven others died from persistent primary
tumor. The actuarial survival was 58% at 8 years, and did not differ between patient
groups analyzed by primary treatment (chemotherapy, radiotherapy, or both) or
prior exposure to particular drugs (alkylating agents, topoisomerase-II inhibitors,
or both).
Among patients analyzed at the International Workshop in Chicago in 2000,
33 of 39 (85%) intensively treated patients with t-AML and inv(16) achieved a
CR (32). Twelve (36%) subsequently relapsed. Five underwent HCT in first CR
(four allogeneic; one autologous), and all of them were alive and leukemia-free at
last follow-up. The responding patients were significantly younger than those who
did not achieve CR (median, 44 years vs. 62 years, p = 0.012). Patients younger
than 55 years of age had improved survival when compared to older patients. The
median survival in the young patient group (n = 26) had not been reached and was
longer than 3 years, but was only 12 months for the 13 older patients (p = 0.006).
Similarly, 24 of 35 (69%) workshop patients who had been treated aggressively
for t-APL achieved a CR. The median overall survival for t-AML patients with
either inv(16) or t(15;17) was 29 months after receiving intensive AML therapy.
Development of both inv(16) t-AML and t-APL was associated with prior exposure
to topoisomerase-II inhibitors. However, 21% of the inv(16) patients and 29% of
the t(15;17) patients had received only radiotherapy previously.
Seventy-two t-AML patients with any t(21q22) were also studied at the
International Workshop (33). Their median survival was 14 months, and 18%
were alive after 5 years. Fifty-three patients with t(21q22) received intensive
AML therapy, and the median survival for the seven who underwent HCT was
31 months compared to 17 months for those who did not. Patients with t(8;21)
had a more favorable outcome than those with other 21q22 rearrangements (p =
0.014). It is noticed that the survivals did not differ for the 11 t-AML patients with
only t(8;21) and for the 33 patients with t(8;21) plus other abnormalities (medians,
17 months and 31 months; p = 0.6). Mutations in KIT were not studied in these
patients.
HEMATOPOIETIC CELL TRANSPLANTATION FOR t-AML

The treatment most likely to cure t-AML is allogeneic HCT. Several small case
series have described the outcomes of these patients, and the survival appears to
be about 20% to 30% (57). However, chronic and cumulative toxicities from prior
chemoradiotherapy impact on the ability to perform HCT and adversely affect
survival. Early deaths from regimen-related toxicity are more common after HCT
for therapy-related leukemia than for primary AML.
In an analysis of 70 patients (31 with t-MDS and 39 with t-AML) who

underwent allogeneic HCT between 1980 and 1998 in France, poor outcomes
were associated with age greater than 37 years, male sex, positive cytomegalovirus
serology in the recipient, absence of CR at the time of HCT, and the use of intensive
conditioning chemotherapy (58). The treatments given were heterogeneous, and
the donors were varied. The estimated 2-year survival rate was 30%, event-free
survival rate 28%, relapse rate 42%, and transplant-related mortality 49%. Thus,
for patients who have chemotherapy-responsive t-AML, allogeneic HCT can be
curative, but it is unfortunately not often successful. Nonmyeloablative, reduced
intensity allogeneic HCT, is under investigation for those who are not eligible for
standard myeloablative HCT.
Similar results have been seen in children who have undergone allogeneic
HCT for t-AML developing after therapy for acute lymphoblastic leukemia
(ALL). Hale et al. reported the outcomes of 21 children who had received
epipodophyllotoxin-containing regimens for ALL and subsequently developed
t-AML (59). Thirteen patients received induction chemotherapy prior to HCT,
whereas seven underwent HCT immediately after diagnosis. One patient received
an autologous HCT in first CR from t-AML, but later relapsed, and was subse-
quently treated at second relapse with an allogeneic HCT. Eleven patients received
bone marrow cells from HLA-matched siblings, eight received bone marrow cells
from matched unrelated donors, and two received haploidentical marrow from
family members. Three-years after HCT, only four patients (19%) were alive.
Seven patients died from transplant-related causes, and 10 patients died from
relapsed t-AML after a median of 5 months.
The European Bone Marrow Transplant registry has reported on 65 t-AML
patients who underwent autologous HCT (60). The median age was 39 years
(range, 3–69). Estimates of overall and disease-free survival at 3 years were 35%
and 32%, respectively. The relapse rate was lower for patients transplanted in first
CR (48% vs. 89%, p = 0.05). Age over 40 years resulted in higher transplant-
related mortality (47% vs. 7%, p = 0.01).
THE IMPORTANCE OF CYTOGENETIC ABNORMALITIES

IN PREDICTING PATIENT OUTCOMES
The most informative data on the prognostic impact of karyotype on outcome in
t-AML were reported by the German AML Cooperative Group (AMLCG) (61).
This group compared karyotype analysis and survival between 93 patients with
t-AML and 1091 patients with de novo AML, all of whom received intensive
treatment. Favorable, intermediate, and unfavorable karyotypes were observed
in 26%, 28%, and 46% of t-AML patients, and in 22%, 57%, and 20% of de
novo AML patients. Overall, the median survival was 10 months for patients with
t-AML compared to 15 months for patients with de novo AML (p = 0.0007).
Armand et al. analyzed the outcomes of 80 patients with t-AML treated at
the Dana Farber Cancer Institute (62). They found that cytogenetic abnormalities
Table 3 Survival of 306 Patients with Therapy-Related Myeloid Leukemia According

to Clinical and Cytogenetic Features: The University of Chicago Series (3)
No. of patients Median survival (mo)

Clinical/cytogenetic subseta (%) (95%CI)
Total group 306 8 (7–9)

Presenting as t-MDS 224 (73) 8.6 (7.6–9.9)
Presenting as t-AML 82 (27) 6.9 (4.0–8.5)
Abnormal chromosome 5 63 (21) 7
Abnormal chromosome 7 85 (28) 9
Abnormalities of both chromosomes 5 and 7 66 (22) 5
Recurring balanced rearrangement 31 (10) 11
Other clonal abnormality 39 (13) 9
Normal karyotype 24 (8) 11
a 139/306 Patients (45%) had complex karyotypes, containing 3 or more unrelated chromosomal
abnormalities.
were the strongest predictors for overall survival. When adjusted for cytogenetic
changes, the patients with t-AML died as well as patients with de novo AML,
further emphasizing the importance of cytogenetic abnormalities in predicting
severity of disease and outcomes.
At the University of Chicago, 306 consecutive patients with t-AML were
analyzed for clinical outcome according to cytogenetic subset as well as other
clinical features, including disease latency (3). In contrast to the German series, not
all of our patients underwent intensive remission induction chemotherapy. Many
received only supportive care. Survival times are shown in Table 3. Only 24 patients
(8%) were alive 3 years after diagnosis. Patients with t-AML who responded to
remission induction therapy but subsequently died from their primary malignancy
were included in the survival analysis. The incidence of unfavorable karyotypes
was greater than 70%. Even patients with a normal karyotype or with a balanced
chromosomal rearrangement did poorly overall. The patients who had the worst
overall survival were those patients with abnormalities of both chromosomes 5
and 7 (p = 0.005).
In an updated analysis of the German AMLCG study, the survival of 121
patients with t-AML was compared to 1511 patients with de novo AML according
to karyotype (63). All received intensive AML therapy. The median survival for
the t-AML patients ranged from 27 months for those with a favorable karyotype to
6 months for those with an unfavorable karyotype (Table 4). Importantly, almost
half of the patients with t-AML (58/121) had an unfavorable karyotype, whereas
only about 20% (302/1511) of the de novo AML patients had an unfavorable
karyotype. For those with a favorable karyotype, the median survival was not yet
reached after 5 years for the 306 de novo AML patients compared to 27 months
for the 29 t-AML patients (p = 0.02). Some of these t-AML patients appeared
to be cured. Within the large intermediate risk cytogenetic group, no significant
Table 4 Survival According to Cytogenetic Risk Group for Patients with t-AML or De
Novo AML Treated by the German AML Cooperative Group (AMLCG) (63)
No. of patients (%) Median survival (mo)
t-AML De novo AML De novo

Karyotype (n = 121) (n = 1511) t-AML AML p
Favorable 29 (24) 306 (20) 27 ⬎60 0.02

Intermediate 34 (28) 903 (60) 12 16 0.19
Unfavorable 58 (48) 302 (20) 6 7 0.006
difference in survival was observed between the t-AML and de novo AML patients.
An unfavorable karyotype predicated a very short survival in both groups of AML
patients.
RECOMMENDATIONS FOR TREATMENT OF t-AML

Figure 4 shows a treatment algorithm for the management of patients who develop
therapy-related myeloid leukemia. Primary considerations are the patient’s perfor-
mance status, which reflects age, comorbidities, the status of the primary disease,
Clinical and morphological diagnosis of t-MDS/t-AML

Cytogenetic analysis should be included in the initial assessment
Good Poor
performance status performance status
t-APL Favorable karyotypes Unfavorable Supportive care

Normal karyotype
karyotypes
includes inv (16) and
includes complex
t(8;21)
karyotypes
ATRA
Standard Standard
+
induction induction Investigational therapy
chemotherapy
chemotherapy chemotherapy
(+As2O3)
Allogeneic HCT Allogeneic HCT

High-dose
or or
cytarabine
consolidation consolidation
consolidation
chemotherapy chemotherapy
Figure 4 Decision tree for the management of t-AML.

and the presence of complications from primary therapy, plus the clonal abnor-
malities detected in the t-AML cells. In general, t-AML patients should be encour-
aged to participate in prospective clinical trials that are appropriately designed for
other AML patients with similar cytogenetic abnormalities. Patients who have an
HLA-matched donor should be considered for allogeneic HCT, although patients
with favorable karyotypes may do reasonably well with conventional intensive
chemotherapy.
ACKNOWLEDGMENT
This work was supported in part by grants CA40046 and CA14599 from the
National Cancer Institute, U.S.A.
REFERENCES
1. Brunning RD, Matutes E, Flandrin G, et al. Acute myeloid leukaemias and myelodys-
plastic syndromes, therapy related. In: Jaffe ES, Harris NL, Stein H, et al., eds. World
Health Organization Classification of Tumours: Pathology and Genetics of Tumours
of Haematopoietic and Lymphoid Tissues. Lyon: IARC Press, 2001:89–91.
2. Rowley JD, Olney HJ. International workshop on the relationship of prior therapy
to balanced chromosome aberrations in therapy-related myelodysplastic syndromes
and acute leukemia: Overview report. Genes Chromosomes Cancer 2002; 33:331–
345.
3. Smith SM, Le Beau MM, Huo D, et al. Clinical-cytogenetic associations in 306
patients with therapy-related myelodysplasia and myeloid leukemia: The University
of Chicago series. Blood 2003; 102:43–52.
4. Offman J, Opelz G, Doehler B, et al. Defective DNA mismatch repair in acute
myeloid leukemia/myelodysplastic syndrome after organ transplantation. Blood
2004; 104:822–828.
5. Seedhouse C, Russell N. Advances in the understanding of susceptibility to treatment-
related acute myeloid leukaemia. Br J Haematol 2007; 137:513–529.
6. Larson RA, Wang Y, Banerjee M, et al. Prevalence of the inactivating 609 C–⬎T
polymorphism in the NAD(P)H:quinone oxidoreductase (NQO1) gene in patients
with primary and therapy-related myeloid leukemia. Blood 1999; 94:803–807.
7. Bolufer P, Collado M, Barragan E, et al. Profile of polymorphisms of drug-
metabolising enzymes and the risk of therapy-related leukaemia. Br J Haematol
2007; 136:590–596.
8. Allan JM, Wild CP, Rollinson S, et al. Polymorphism in glutathione S-transferase P1
is associated with susceptibility to chemotherapy-induced leukemia. Proc Natl Acad
Sci U S A 2001; 98:11592–11597.
9. Naoe T, Takeyama K, Yokozawa T, et al. Analysis of genetic polymorphism in
NQO1, GST-M1, GST-T1, and CYP3A4 in 469 Japanese patients with therapy-
related leukemia/myelodysplastic syndrome and de novo acute myeloid leukemia.
Clin Cancer Res 2000; 6:4091–4095.
10. Fern L, Pallis M, Ian Carter G, et al. Clonal haemopoiesis may occur after con-
ventional chemotherapy and is associated with accelerated telomere shortening and
defects in the NQO1 pathway; possible mechanisms leading to an increased risk of

t-AML/MDS. Br J Haematol 2004; 126:63–71.
11. Guillem VM, Collado M, Terol MJ, et al. Role of MTHFR (677, 1298) haplotype
in the risk of developing secondary leukemia after treatment of breast cancer and
hematological malignancies. Leukemia 2007; 21:1413–1422.
12. Sheikhha MH, Tobal K, Liu Yin JA. High level of microsatellite instability but not
hypermethylation of mismatch repair genes in therapy-related and secondary acute
myeloid leukaemia and myelodysplastic syndrome. Br J Haematol 2002; 117:359–
365.
13. Worrillow LJ, Travis LB, Smith AG, et al. An intron splice acceptor polymorphism
in hMSH2 and risk of leukemia after treatment with chemotherapeutic alkylating
agents. Clin Cancer Res 2003; 9:3012–3020.
14. Worrillow L, Smith A, Scott K, et al. Polymorphic MLH1 and risk of cancer after
methylating chemotherapy for Hodgkin lymphoma. J Med Genet 2008; 45:142–146.
15. Felix CA, Hosler MR, Provisor D, et al. The p53 gene in pediatric therapy-related
leukemia and myelodysplasia. Blood 1996; 87:4376–4381.
16. Ellis NA, Huo D, Yildiz O, et al. MDM2 SNP309 and TP53 Arg72Pro interact to
alter therapy-related acute myeloid leukemia susceptibility. Blood 2008; 112:741–
749.
17. Bond GL, Hu W, Bond EE, et al. A single nucleotide polymorphism in the MDM2
promoter attenuates the p53 tumor suppressor pathway and accelerates tumor forma-
tion in humans. Cell 2004; 119:591–602.
18. Bond GL, Hirshfield KM, Kirchhoff T, et al. MDM2 SNP309 accelerates tumor
formation in a gender-specific and hormone-dependent manner. Cancer Res 2006;
66:5104–5110.
19. Bond GL, Menin C, Bertorelle R, et al. MDM2 SNP309 accelerates colorectal tumour
formation in women. J Med Genet 2006; 43:950–952.
20. Dumont P, Leu JI, Della Pietra AC III, et al. The codon 72 polymorphic variants of
p53 have markedly different apoptotic potential. Nat Genet 2003; 33:357–365.
21. Thomas M, Kalita A, Labrecque S, et al. Two polymorphic variants of wild-type p53
differ biochemically and biologically. Mol Cell Biol 1999; 19:1092–1100.
22. Marin MC, Jost CA, Brooks LA, et al. A common polymorphism acts as an intragenic
modifier of mutant p53 behaviour. Nat Genet 2000; 25:47–54.
23. Maris JM, Wiersma SR, Mahgoub N, et al. Monosomy 7 myelodysplastic syndrome
and other second malignant neoplasms in children with neurofibromatosis type 1.
Cancer 1997; 79:1438–1446.
24. Mahgoub N, Taylor BR, Le Beau MM, et al. Myeloid malignancies induced by
alkylating agents in Nf1 mice. Blood 1999; 93:3617–3623.
25. Mathew CG. Fanconi anaemia genes and susceptibility to cancer. Oncogene 2006;
25:5875–5884.
27. Le Beau MM, Espinosa R III, Davis EM, et al. Cytogenetic and molecular delineation
of a region of chromosome 7 commonly deleted in malignant myeloid diseases. Blood
1996; 88:1930–1935.
28. Zhao N, Stoffel A, Wang PW, et al. Molecular delineation of the smallest commonly
deleted region of chromosome 5 in malignant myeloid diseases to 1–1.5 Mb and
preparation of a PAC-based physical map. Proc Natl Acad Sci U S A 1997; 94:6948–
6953.
29. Godley LA, Larson RA. The syndrome of therapy-related myelodysplasia and
myeloid leukemia. In: Bennett JM, ed. The Myelodysplastic Syndromes: Patho-
biology and Clinical Management, New York: Marcel Dekker, Inc., 2002:139–176.
30. Curtis RE, Boice JD Jr, Stovall M, et al. Risk of leukemia after chemotherapy and
radiation treatment for breast cancer. N Engl J Med 1992; 326:1745–1751.
31. Greene MH, Harris EL, Gershenson DM, et al. Melphalan may be a more potent
leukemogen than cyclophosphamide. Ann Intern Med 1986; 105:360–367.
32. Andersen MK, Larson RA, Mauritzson N, et al. Balanced chromosome abnormal-
ities inv(16) and t(15;17) in therapy-related myelodysplastic syndromes and acute
leukemia: Report from an international workshop. Genes Chromosomes Cancer 2002;
33:395–400.
33. Slovak ML, Bedell V, Popplewell L, et al. 21q22 balanced chromosome aberrations
in therapy-related hematopoietic disorders: report from an international workshop.
Genes Chromosomes Cancer 2002; 33:379–394.
34. Coso D, Costello R, Cohen-Valensi R, et al. Acute myeloid leukemia and myelodys-
plasia in patients with chronic lymphocytic leukemia receiving fludarabine as initial
therapy. Ann Oncol 1999; 10:362–363.
35. Morrison VA, Rai KR, Peterson BL, et al. Therapy-related myeloid leukemias are
observed in patients with chronic lymphocytic leukemia after treatment with fludara-
bine and chlorambucil: Results of an intergroup study, cancer and leukemia group B
9011. J Clin Oncol 2002; 20:3878–3884.
36. McLaughlin P, Estey E, Glassman A, et al. Myelodysplasia and acute myeloid
leukemia following therapy for indolent lymphoma with fludarabine, mitoxantrone,
and dexamethasone (FND) plus rituximab and interferon alpha. Blood 2005;
105:4573–4575.
37. Martin-Salces M, Canales MA, de Paz R, et al. Treatment-related acute myeloid
leukemia with 11q23 translocation following treatment with fludarabine, cyclophos-
phamide and rituximab. Leuk Res 2008; 32:199–200.
38. Czuczman MS, Emmanouilides C, Darif M, et al. Treatment-related myelodysplastic
syndrome and acute myelogenous leukemia in patients treated with ibritumomab
tiuxetan radioimmunotherapy. J Clin Oncol 2007; 25:4285–4292.
39. Fenske TS, McMahon C, Edwin D, et al. Identification of candidate alkylator-induced
cancer susceptibility genes by whole genome scanning in mice. Cancer Res 2006;
66:5029–5038.
110:719–726.
41. Ebert BL, Pretz J, Bosco J, et al. Abstracts of Papers, 49th National Meeting of the
American Society of Hematology, Atlanta, GA, December 8–11, 2007. Washington,
DC: American Society of Hematology, 2007; Abstract #1.
42. Pedersen-Bjergaard J, Christiansen DH, Desta F, et al. Alternative genetic path-
ways and cooperating genetic abnormalities in the pathogenesis of therapy-related
myelodysplasia and acute myeloid leukemia. Leukemia 2006; 20:1943–1949.
43. Gilliland DG, Jordan CT, Felix CA. The molecular basis of leukemia. Hematol Am
Soc Hematol Educ Program 2004:80–97.
44. Qian Z, Fernald AA, Godley LA, et al. Expression profiling of CD34 + hematopoi-
etic stem/progenitor cells reveals distinct subtypes of therapy-related acute myeloid
leukemia. Proc Natl Acad Sci U S A 2002; 99:14925–14930.
45. Schnittger S, Bacher U, Kern W, et al. JAK2 seems to be a typical cooperating muta-
tion in therapy-related t(8;21)/AML1-ETO-positive AML. Leukemia 2007; 21:183–
184.
46. Desta F, Christiansen DH, Andersen MK, et al. Activating mutations of JAK2V617 F
are uncommon in t-MDS and t-AML and are only observed in atypic cases. Leukemia
2006; 20:547–548.
47. Christiansen DH, Desta F, Andersen MK, et al. Mutations of the PTPN11 gene
in therapy-related MDS and AML with rare balanced chromosome translocations.
Genes Chromosomes Cancer 2007; 46:517–521.
48. Smith RE, Bryant J, DeCillis A, et al. Acute myeloid leukemia and myelodysplastic
syndrome after doxorubicin-cyclophosphamide adjuvant therapy for operable breast
cancer: The National Surgical Adjuvant Breast and Bowel Project Experience. J Clin
Oncol 2003; 21:1195–1204.
49. Hershman D, Neugut AI, Jacobson JS, et al. Acute myeloid leukemia or myelodys-
plastic syndrome following use of granulocyte colony-stimulating factors dur-
ing breast cancer adjuvant chemotherapy. J Natl Cancer Inst 2007; 99:196–
205.
50. Le Deley MC, Suzan F, Cutuli B, et al. Anthracyclines, mitoxantrone, radiotherapy,
and granulocyte colony-stimulating factor: risk factors for leukemia and myelodys-
plastic syndrome after breast cancer. J Clin Oncol 2007; 25:292–300.
51. Hake CR, Graubert TA, Fenske TS. Does autologous transplantation directly increase
the risk of secondary leukemia in lymphoma patients? Bone Marrow Transplant 2007;
39:59–70.
52. Kalaycio M, Rybicki L, Pohlman B, et al. Risk factors before autologous stem-
cell transplantation for lymphoma predict for secondary myelodysplasia and acute
myelogenous leukemia. J Clin Oncol 2006; 24:3604–3610.
53. Takeyama K, Seto M, Uike N, et al. Therapy-related leukemia and myelodysplastic
syndrome: A large-scale Japanese study of clinical and cytogenetic features as well
as prognostic factors. Int J Hematol 2000; 71:144–152.
54. Kantarjian HM, Estey EH, Keating MJ. Treatment of therapy-related leukemia
and myelodysplastic syndrome. Hematol Oncol Clin North Am 1993; 7:81–
107.
55. Larson RA, Wernli M, Le Beau MM, et al. Short remission durations in therapy-
related leukemia despite cytogenetic complete responses to high-dose cytarabine.
Blood 1988; 72:1333–1339.
56. Beaumont M, Sanz M, Carli PM, et al. Therapy-related acute promyelocytic leukemia.
J Clin Oncol 2003; 21:2123–2137.
57. Anderson JE, Gooley TA, Schoch G, et al. Stem cell transplantation for secondary
acute myeloid leukemia: Evaluation of transplantation as initial therapy or following
induction chemotherapy. Blood 1997; 89:2578–2585.
58. Yakoub-Agha I, de La Salmoniere P, Ribaud P, et al. Allogeneic bone marrow
transplantation for therapy-related myelodysplastic syndrome and acute myeloid
leukemia: A long-term study of 70 patients-report of the French society of bone
marrow transplantation. J Clin Oncol 2000; 18:963–971.
59. Hale GA, Heslop HE, Bowman LC, et al. Bone marrow transplantation for therapy-
induced acute myeloid leukemia in children with previous lymphoid malignancies.
Bone Marrow Transplant 1999; 24:735–739.
60. Kroger N, Brand R, van Biezen A, et al. Autologous stem cell transplantation for
therapy-related acute myeloid leukemia and myelodysplastic syndrome. Bone Mar-
row Transplant 2006; 37:183–189.
61. Schoch C, Kern W, Schnittger S, et al. Karyotype is an independent prognostic
parameter in therapy-related acute myeloid leukemia (t-AML): An analysis of 93
patients with t-AML in comparison to 1091 patients with de novo AML. Leukemia
2004; 18:120–125.
62. Armand P, Kim HT, DeAngelo DJ, et al. Impact of cytogenetics on outcome of de
novo and therapy-related AML and MDS after allogeneic transplantation. Biol Blood
Marrow Transplant 2007; 13:655–664.
63. Kern W, Haferlach T, Schnittger S, et al. Prognosis in therapy-related acute myeloid
leukemia and impact of karyotype. J Clin Oncol 2004; 22:2510–2511.
9
Diagnosis of Myelodysplastic Syndromes:
Criteria and Challenges
Attilio Orazi
Department of Pathology and Laboratory Medicine, Weill Medical College of
Cornell University, New York, New York, U.S.A.
James W. Vardiman
Department of Pathology, University of Chicago Medical Center, Chicago,
Illinois, U.S.A.
INTRODUCTION
The diagnostic evaluation of myelodysplastic syndromes (MDS) includes morpho-
logic evaluation of peripheral blood, marrow aspirate, and bone marrow biopsy
specimens, which must be interpreted in the context of the complete blood count
results and adequate clinical information (1–5). Correlation with marrow cytoge-
netic results (see chap. 3) is also essential. Importantly, the presence of a normal
karyotype does not exclude a diagnosis of MDS. Conversely, an abnormal kary-
otype may indicate MDS within the appropriate clinical context, even if morpho-
logic findings are inconclusive. In some clinical settings, FISH analysis as well as
other techniques (e.g., spectral karyotyping) may also be successfully employed
(see chaps. 3 and 4).
The previously used 1982 French-American-British (FAB) classification of
MDS (3) (see chap. 1) was based entirely on findings identifiable by cytological
analysis of well-stained smears of peripheral blood and bone marrow aspirate. The
main criteria for subclassification of MDS in the FAB system were the percentage
of blasts in the peripheral blood and in bone marrow aspirates, and the presence
of dysplastic changes in at least one of the three main marrow cell lines. A more
195
196 Orazi and Vardiman
comprehensive approach is used by the WHO system (updated in 2001 and again
in 2008), which stresses the importance of integrating other techniques such as
bone marrow biopsy histology, cytogenetics, and molecular genetics, in the light
of relevant clinical information. However, not all of these techniques are generally
available or can provide useful information at that time when initial treatment
decisions need to be made.
DIAGNOSIS
Morphology
Although the diagnosis of MDS can be suspected from the clinical history and
peripheral blood counts, the diagnosis is customarily made by observing unequiv-
ocal dysplastic features of the erythroid, granulocytic, or megakaryocytic cells
upon morphologic inspection of the peripheral blood, bone marrow aspirate, and
bone marrow biopsy (1–5). Myeloblasts may or may not be increased in number.
Monocytes may also be increased in number, and are best appreciated by using
nonspecific esterase stains (recommended is the ␣-naphtyl butyrate esterase).
Accurate enumeration of monocytes in the blood and bone marrow is necessary
to distinguish between MDS and the MDS/myeloproliferative neoplasm (MPN)
entity, chronic myelomonocytic leukemia (CMML).
Morphologic dysplasia is not necessarily synonymous with MDS; the
important issue of MDS “mimics” will be covered in a later part of this chapter.
To address the issue of “false positive” myelodysplasia (i.e., occasional abnormal
appearing cells in patients without a bona fide clonal myeloid disorder), the WHO
system recommends that at least 10% of the cells in a lineage be morphologically
dysplastic in order to say with confidence that the lineage shows dysplasia (1).
The morphological classification of MDS under the WHO system is principally
based on the percent of blasts in the bone marrow and peripheral blood, the type
and degree of dysplasia, and the presence of ring sideroblasts. In its approach,
the WHO classification of MDS is largely similar to the previously used and
well-validated FAB system (1), with the important exception that the distinction
between unilineage or multilineage dysplasia is an important aspect of MDS
subclassification in the WHO.
Blasts
The enumeration of blasts is important for the diagnosis and classification of MDS,
and in most studies, the blast percentage is one of the most important prognostic
indicators (6–10). Blast proportion is also an integral component of currently used
prognostic scoring systems such as the International Prognostic Scoring System
(IPSS) (9) and the more recent WHO classification-based Prognostic Scoring
System (WPSS) (10) (chap. 1).
Blasts should be carefully enumerated from well-prepared peripheral blood
smears and bone marrow aspirates. Blast cells that should be included in the
blast count include myeloblasts (both those without or with a few fine azurophilic
Diagnosis of Myelodysplastic Syndromes: Criteria and Challenges 197
granules), monoblasts, and megakaryoblasts. Promonocytes are also considered

“blast equivalents” in the WHO classification scheme (1). These cells are morpho-
logically difficult to distinguish from promyelocytes and early myelocytes in the
absence of cytochemistry for esterase, but they are rarely found in MDS anyway.
Their presence suggests a diagnosis of CMML or, in the presence of an increased
blast count, acute myeloid leukemia (AML) with monocytic differentiation.
The count derived from the aspirate smears should be carefully correlated
with the estimate from the biopsy sections (5,11–13). Often, in the marrow biopsy,
blasts may be hard to appreciate, particularly if there is marrow fibrosis. In such
cases, an immunohistochemical stain for CD34+ cells may be helpful (13–18).
However, because not all blasts are CD34+, care must be taken not to use the CD34
result in lieu of careful morphologic assessment. Blast proportion assessment by
flow cytometry is not a substitute for morphological blast enumeration, and can
be misleading, as discussed further below.
Dyserythropoiesis
In the blood, dyserythopoiesis is usually manifested as anemia, which is usually
normocytic or macrocytic. Red blood cell (RBC) anisocytosis, poikilocytosis,
basophilic stippling, and nucleated RBCs can be seen. A “dimorphic” pattern
of normochromic, normocytic, or macrocytic cells with hypochromic cells is
often observed, and may reflect an admixture of clonal and residual normal
hematopoietic cells. In the marrow, megaloblastoid change, multinuclearity,
nuclear fragments, bizarre nuclear shapes, internuclear bridging, and ring
sideroblasts (i.e., erythroid precursors with at least 10 or more perinuclear iron
granules) are among the abnormalities that can be observed [Figs. 1(A) and
1(B) see colour insert]. Erythroid hyperplasia is common in MDS, but erythroid
hypoplasia can also be seen. In addition, a number of intrinsic RBC abnormalities,
such as spherocytic RBCs or alpha thalassemia, can be acquired in patients
with MDS. The latter may be associated with microcytosis, rather than the more
common normocytic or macrocytic anemia.
Dysgranulopoiesis
Dysgranulopoiesis is often (but not inevitably) manifested as neutropenia in the
peripheral blood. Hypogranularity of cytoplasm and nuclear hypolobation are the
most commonly observed morphologic abnormalities in neutrophils [Fig. 1(C)].
Nuclear “sticks” or excrescences, and hypercondensed, abnormally clumped chro-
matin are also common findings. Hypersegmentation of neutrophil nuclei is less
frequent, but still observed regularly. In the marrow, abnormalities in cytoplasmic
granulation, nuclear lobation, and chromatin clumping are the usual findings in
developing granulocytic series. It is particularly important that well-stained slides
be used to evaluate dysgranulopoiesis in the blood or marrow, as a poorly stained
smear may lead to over-interpretation of hypogranularity of the cytoplasm. For
this reason, a diagnosis of MDS should not be based solely on hypogranularity of
the cytoplasm of the neutrophils.
(A) (B)
(C)
(D)
(E)
Figure 1 Bone marrow morphology in patients with MDS [aspirate: (A–D); biopsy: (E)].
(A) Dyserythropoietic changes including nuclear bridging in the cases of RCMD. (B) Ring
sideroblasts characterized a case of RARS. Note the presence of perinuclear distribution
of the siderotic granules (⬎10). (C) Nuclear hypolobation and presence of blasts in a case
of RAEB. (D) Characteristic nonlobated megakaryocytes seen in a case with isolated 5q
abnormality. (E) The same type of cell can also be appreciated in an H&E stained section
of bone marrow biopsy. (see color insert)
Dysmegakaryopoiesis
Just as with neutropenia and dysgranulopoiesis, the most common peripheral
blood manifestation of dysmegakaryopoiesis is thrombocytopenia. Platelets vary
in size in MDS, but abnormally large platelets are frequent, and the cells may
be either hypo- or hypergranulated. In the marrow, megakaryocytes are often
increased in number (especially in association with interstitial deletions of the
long arm of chromosome 5), but megakaryocyte hypoplasia is also seen. The
dysplastic megakaryocytes may be micromegakaryocytes (⬍15 ␮m in diameter),
have monolobation or hypolobation of their nuclei [Figs. 1(D) and 1(E)], have
multiple widely separated nuclei, or be hypogranulated, the latter two anomalies
are less specific for MDS being also seen in reactive conditions associated with
marrow dyspoietic changes.
Bone Marrow Biopsy

The value of bone marrow biopsy in this group of disorders is well established (11–
18). Bone marrow biopsy may help in confirming a suspected diagnosis of MDS
by excluding reactive conditions in which dyshematopoietic changes may also be
observed. Reactive conditions, which may simulate MDS, are discussed below
in the differential diagnosis section. In MDS, bone marrow biopsy supplemented
by immunohistochemistry can be used to increase the diagnostic accuracy and to
refine the IPSS risk evaluation (12).
Bone marrow biopsy in MDS allows an accurate assessment of marrow
cellularity, the predominant cell line(s) involved, presence of fibrosis, and/or other
stromal changes, dysmegakaryopoiesis (which is more easily detected in histology
preparations than in aspirate smears), and, finally, architectural disorganization.
In normal marrow, granulopoietic precursors are mainly found in the paratrabec-
ular region, while erythroid and megakaryocytes are more or less confined to the
central marrow cavities. In MDS, topographical organization is lost, with precur-
sors of different cell lines found in all marrow regions. Other common findings
observed in the marrow biopsy in MDS patients include the presence of areas of
edema, increased number of microvessels, plasmacytosis, increased mast cells,
macrophages with increased cellular debris and hemosiderin, and lymphocytosis
and lymphoid follicles.
Among the architectural alterations detected by bone marrow biopsy, one
that may be prognostically important is the presence of aggregates or clusters
of “abnormally localized” immature myeloid precursor cells (ALIP), that is,
myeloblasts and promyelocytes, in an abnormal central marrow cavity location
(19,20) [Fig. 2(A) see colour insert]. Cases are classified as “ALIP-positive” if at
least three aggregates (⬎5 myeloid precursors) or clusters (3–5 myeloid precur-
sors) are identified in each section (20). ALIP is present in the aggressive MDS
subtypes, and is associated with a poor prognosis and an increased incidence of
progression to acute leukemia. If ALIP is observed in cases diagnosed as lower
(A) (B)
Figure 2 Case of RAEB: (A) several clusters of blasts forming ALIP in an H&E stained
histology section; (B) CD34 immunohistology can help the identification of blasts by
ruling out “pseudo-ALIP”, for example, clusters of megaloblastoid erythroblasts or other
cell types. (see color insert)
risk MDS [i.e., refractory anemia (RA), refractory anemia with ring sideroblasts
(RARS), or refractory cytopenia with multilineage dysplasia (RCMD)], the case
should be reassessed for accuracy of the blast count. Recently, the immature cells
in ALIP have been demonstrated to coexpress Vascular endothelial growth factor
(VEGF), a potent angiogenic peptide, as well as one of the VEGF receptors, FLT1
(21). This finding suggests an autocrine/paracrine cytokine interaction that may
be important in leukemia cell renewal (21). VEGF may also stimulate endothe-
lial cells to release apoptotic factors such as TNF-␣, which can contribute to
hematopoietic failure.
The presence of ALIP is not unique to MDS and has been reported in
the context of regenerating hematopoiesis, such as after bone marrow transplan-
tation or induction chemotherapy. In addition, identification of the presence of
ALIP may be compromised by using paraffin sections of excessive thickness or
otherwise suboptimal preparation that limits morphology assessment. CD34, an
antigen expressed in progenitor and early precursor marrow cells, can be used as
a “surrogate marker” for the presence of ALIP [Fig. 2(B)]. Both an increase in
the percentage of CD34+ cells and a tendency of CD34+ cells to form aggre-
gates have been shown to be reliable predictors of acute leukemic transformation
and of poor survival in MDS cases, irrespective of the MDS subtype (14,16–18).
This approach can be used to identify patients with MDS undergoing transition to
AML, who may be candidates for aggressive therapy.
Most of the CD34+ cells found in MDS morphologically resemble blasts.
However, a proportion of the CD34+ cells show promyelocyte-like cytologi-
cal features. These should be counted as blasts rather than as promyelocytes
for the purpose of blast-cell enumeration. Aberrant expression of CD34 by
megakaryocytes in MDSs has also been reported (22). In marrow from healthy
persons, CD34 positivity is only found on a small subset of megakaryocyte
precursors morphologically identifiable as immature blasts. This suggests that

the CD34+ megakaryocytes seen in MDS represent poorly functional neoplastic
megakaryocytes showing, in addition to morphologic atypia, anomalous pheno-
typic differentiation. However, CD34+ megakaryocytes can also be identified in
other types of neoplastic myeloid disorders, as well as in reactive conditions (e.g.,
megaloblastic anemia).
Immunohistology using antibodies to CD34, CD42b (platelet glycoprotein
Ib alpha polypeptide), or CD61 (platelet glycoprotein IIIa) may be particularly
useful in distinguishing myelodysplastic syndrome with fibrosis (MDS-F) (23)—a
disorder characterized by the presence of blasts that are CD34+ but negative with
antibodies to CD42b (or CD61)—from acute megakaryoblastic leukemia, which
is associated with an increased number of megakaryoblasts that stain positively
with CD42b or CD61, but only variably positive with CD34 (24).
CD117 (c-Kit) has been recently proposed as an additional marker to iden-
tify myeloid precursor cells in marrow biopsies of patients with MDS. However,
in our experience, this antibody is a less reliable marker than CD34, due to its
weak and variable expression in MDS myeloblasts. Other antibodies, which can
be occasionally useful, include antihemoglobin (25) or glycophorin A or C, which
can identify megaloblastic erythroblasts that may occasionally be confused with
myeloid precursors; myeloperoxidase or lysozyme (26,27), which can be use-
ful in differentiating background maturing myeloid precursors that are strongly
myeloperoxidase positive from the leukemic myeloblasts, which are characteris-
tically weakly positive or negative in MDS cases; and CD68 (macrosialin) and
CD163 (macrophage-associated antigen), which can be used to identify back-
ground monocytes and can help rule out CMML (28,29).
Bone marrow biopsy supplemented by immunohistology is especially help-
ful in three subsets of patients with MDS that may not have suitable marrow
aspirate material for analysis: MDS-F (23), therapy-related MDS (TR-MDS) (30),
and MDS with hypocellular marrow (31–34). The presence of reticulin fibrosis
and/or fatty changes in the bone marrow of these MDS patients causes hemodi-
lution or poorly cellular smears, which can make accurate disease characteriza-
tion difficult or impossible. The low cellular yield of the bone marrow aspirate
in these MDS subtypes may also be insufficient to obtain adequate cytogenetic
information.
Loss of p53 function has a major role in the transformation process in
hematologic malignancies. Alterations of p53 are frequent in aggressive MDS,
particularly in therapy-related cases, and in secondary AML (35,36). A proportion
of these cases have abnormalities of chromosome band 17p, site of the TP53 gene.
The p53 gene product can be easily demonstrated in routinely processed bone
marrow biopsy specimens. In our experience, p53 overexpression is almost always
associated with the presence of a complex karyotypes and poor prognosis (37).
It is now well established that angiogenesis is involved in the pathogenesis of
MDS. By immunostaining bone marrow biopsies with CD34 or other endothelium-
reactive markers, such as CD105, CD31, or vWF, microvessel density has been
found to be increased in various leukemic disorders (38–40). Results obtained in
cases of MDS suggest a correlation between increased angiogenesis, aggressive

MDS subtypes, and a high rate of progression to acute leukemia (40).
Other Diagnostic Procedures

Cytogenetics/Molecular Genetic Studies
The discovery of a clonal cytogenetic abnormality within the appropriate clinical
and morphologic context provides confirmatory evidence for a diagnosis of MDS.
Certain cytogenetic abnormalities may even provide presumptive evidence for a
diagnosis of MDS in a case in which the morphologic findings alone are not
sufficient to make an unequivocal diagnosis. The value of cytogenetics in MDS
was recently discussed in a recent analysis based on over 2000 patients (41) and
is reviewed in detail in chapter 3.
Flow Cytometry
The utility of immunophenotyping by flow cytometry in MDS has focused on
several strategies, including determining the number and immunophenotype of
blasts, as well as assessing the maturation pattern of the myeloid cell populations
(42–46). This is discussed in detail in chapter 11. Although there is often a
reasonable good correlation between CD34+ percentage as measured by flow
cytometry and blast count by morphology (1,2,5,34,47), the percentage of blasts
determined by flow cytometry may be influenced by hemodilution of the specimen
at the time of collection, processing artifact, or marrow fibrosis resulting in a poorly
representative specimen. Therefore, analysis of blasts by flow cytometry should
never be used in lieu of morphologic inspection of the peripheral blood and bone
marrow aspirate smears.
Recent attention has been directed towards the flow cytometric analysis of
abnormal maturation patterns of erythroid, granulocytic, and monocytic popula-
tions in MDS. The detection of asynchronous antigen expression on maturing
cells or other patterns of aberrant phenotypes by four-color flow cytometry has
been shown to correlate well with morphologic and cytogenetic abnormalities.
In particular, the presence of decreased myeloblast CD45 (leukocyte common
antigen) density, increased CD13 and CD34 density, and increased expression of
CD11 c (integrin alpha X), and dim CD4 were found to be grade-related (i.e.,
abnormalities were more extreme in aggressive MDS subtypes) (45). However, in
cases with borderline dysplasia by morphology and no cytogenetic abnormalities,
flow cytometry results are considered as suggestive of MDS only if there are three
or more aberrant features in erythropoietic, granulocytic, and monocytic matu-
ration; a single aberrant finding is not considered definitive. Furthermore, more
studies are needed with age-matched controls with various disorders to determine
how specific MDS-associated flow cytometry changes may be. Finally, specific
phenotypic abnormalities in myeloblasts, such as the expression of CD56 (neural
cell adhesion molecule 1) or TdT (terminal deoxynucleotidyl transferase), have
been reported to be associated with a worse prognosis (48,49).
CLASSIFICATION OF MDS
The WHO classification of the MDS (2008 revision, 4th edition) was in the press
at the time of this writing. The updated classification, which largely overlaps the
2001 WHO scheme (50), distinguishes several MDS subtypes (Table 1):
r Refractory cytopenia with unilineage dysplasia (RCUD)
Subcategories: refractory anemia (RA), refractory neutropenia (RN), refractory

thrombocytopenia (RT)
r Refractory anemia with ring sideroblasts (RARS)
r Refractory cytopenia with multilineage dysplasia (RCMD)
r Refractory anemia with excess of blasts (RAEB)
Subcategories: RAEB-1, RAEB-2

r Myelodysplastic syndrome, unclassifiable (MDS, U)
r Myelodysplastic syndrome with isolated del(5q) chromosomal abnormality
Refractory Cytopenia with Unilineage Dysplasia

This designation encompasses those MDS that present a refractory cytopenia (of
at least 6 months duration, in the absence of a confirmatory cytogenetic abnor-
mality) associated with dysplasia in ⬎10% of cells in a single cell line. Subtypes
include refractory anemia (RA) and the rare cases of isolated refractory neutrope-
nia (RN) and refractory thrombocytopenia (RT). Refractory bi-cytopenia may be
included in this category if accompanied by unilineage dysplasia, but refractory
pancytopenia that is associated with unilineage dysplasia should be considered as
MDS, unclassifiable.
Patients’ presenting complaints are, in most cases, related to anemia (these
cases are properly designated as RA). In RA, anemia may be normocytic and
normochromic, but is often macrocytic. In the blood, blasts are absent or represent
⬍1% of circulating leukocytes, and they account for fewer than 5% of the nucleated
marrow cells. The marrow is usually hypercellular due to erythroid hyperplasia,
and dyserythropoiesis is present, but ring sideroblasts account for fewer than 15%
of the erythroid cells. In RA, fewer than 10% of the cells in the granulocytic or
megakaryocytic lineages show dysplasia. Monocytes are ⬍1 × 109 /L in the blood,
and there is no monocytosis (⬍5% monocytes; normal range 0–4%) in the bone
marrow. No Auer rods are present. The diagnosis is one of exclusion—other causes
of anemia with dyserythropoiesis, such as megaloblastic anemia and congenital
dyserythropoiesis, must be carefully excluded. In general, RA can be considered as
a “low-grade” MDS. Median survival times with unilineage dysplasia are reported
to be 6 to 7 years, and only ∼10% progress to overt acute leukemia. RN and RT are
very rare, and likely account for fewer than 1% to 2% of all cases of MDS. Other
causes of neutropenia and thrombocytopenia need to be excluded, and extreme
caution should be used in making such diagnoses.
Table 1 Summary of Findings in the Revised (2008) WHO Classification of MDS

Disease Blood findings Bone marrow findings
Refractory cytopenia with Cytopeniaa,d Dysplasia (⬎10%) in one

unilineage dysplasia lineage only
(RA, RN, RT)
No or rare blasts (⬍1%) ⬍5% blasts
Monocytes ⬍1 × 10 /L ⬍15% ring siderobasts
Refractory anemia with Anemiaa Erythroid dysplasia only
ring sideroblasts
No blastsb ⬍5% blasts
Monocytes ⬍1 × 10/L ⬎15% ringed siderobasts
Refractory cytopenia with Cyotpenia(s) Dysplasia in ⬎10% of the
multilineage dysplasia cells of two or more
myeloid lineages
No or rare blasts (⬍1%)b ⬍5% blasts
Monocytes ⬍1 × 10/L ⬍15% ringed siderobasts
No Auer rodsc No Auer rods
Refractory cytopenia with Cytopenias Dysplasia in ⬎10% of the
multilineage dysplasia cells of two or more
and ring sideroblasts myeloid lineages
No or rare blastsb ⬍5% blasts
Monocytes ⬍1 × 10/L ⬎15% ringed siderobasts
No Auer rods No Auer rods
Refractory anemia with Cytopenias Unilineage or multilineage
excess blasts-1 dysplasia
⬍5% blastsb 5–9% blasts
Monocytes ⬍1 × 10/L No Auer rods
No Auer rods
Refractory anemia with Cytopenias Unilineage or multilineage
excess blasts-2 dysplasia
5–19% blasts 10–19% blasts
Monocytes ⬍1 × 10/L Auer rods ±
Auer rods ±
MDS with isolated del(5q) Anemia NI to incr megakaryocytes
with hypolobated nuclei
Usually nl or sl incr ⬍5% blasts
platelets
⬍5% blasts Del(5q) is sole cytogenetic
abnormality
No Auer rods No Auer rods
a Bi-cytopenia may be occasionally observed.
b If the marrow blast percentage is <5%, but there are 2–4% myeloblasts in the blood, the diagnostic
classification is RAEB-1. If the marrow blast percentage is <5% and there are 1% myeloblasts in the
blood, the case should be considered as MDS-U.
c Cases with Auer rods and <5% blasts in the blood and <10% in the marrow should be classified as
RAEB-2.
d Cases with unilineage dysplasia and pancytopenia are classified as MDS-U.
Refractory Anemia with Ring Sideroblasts (RARS)

RARS is an MDS characterized by unexplained anemia, morphologic dysplasia
in the erythroid lineage, and ring sideroblasts comprising 15% or more of the
erythroid precursors. There is no significant (⬍10%) dysplasia in granulocytic or
megakaryocytic lineages. Anemia is the principal finding. The RBCs often exhibit
a “dimorphic” pattern of normochromic and hypochromic cells, but macrocytosis
is frequently observed as well. The criteria are similar to those described for
RA with unilineage dysplasia, except that in the bone marrow, ring sideroblasts
account for ⬎15% of the erythroid precursors [Fig. 1(B)].
Like RA, RARS is a “low-grade” process. In most series, RARS is reported
to have the best prognosis and lowest rate of conversion to AML of all of the
subtypes of MDS. Median survival times of 7 to 9 years or longer are commonly
reported, and the conversion rate to acute leukemia is ⬍5%.
It must be remembered that ring sideroblasts can be found in any form of
MDS, and also in AML and non-neoplastic conditions. In some congenital sider-
oblastic anemias, such as in the Pearson marrow-pancreas syndrome, ring sider-
oblasts result from germline abnormalities (mutation or deletion) of mitochondrial
DNA that encodes factors important for regulation of iron metabolism. Whether
some patients diagnosed with RARS, who have no evidence of a cytogenetic abnor-
mality, may have ring sideroblasts only due to mitochondrial DNA abnormalities
without clonal abnormalities in nuclear DNA is not clear (see chap. 4).
In rare patients with ring sideroblasts, the platelet count is elevated, some-
times to more than 1000 × 109 /L, and the distinction between RARS and a myelo-
proliferative disorder, such as essential thrombocythemia, becomes problematic.
If the platelet count is greater than 450 × 109 /L, and the megakaryocytes have
features of those described in the MPNs, an analysis for a JAK2 V617F mutation
and assignment to the provisional entity of refractory anemia with ring sideroblasts
and thrombocytosis (RARS-T, see later) should be considered (51–53).
Refractory Cytopenia with Multilineage Dysplasia (RCMD)

This subcategory is an addition in the WHO classification of MDS that was not
recognized in the FAB system. Its identification has improved the prognostic
utility of the morphologic classification of MDS (54–61). RCMD is characterized
by one or more cytopenias in the peripheral blood and dysplastic changes in ⬎10%
of cells in two or more of the myeloid lineages: erythroid, granulocytic, and/or
megakaryocytic series. In RCMD, there are ⬍1% blasts in the blood and ⬍5%
blasts in the bone marrow, and Auer rods are not found. If 15% or more of the
erythroid precursors are ring sideroblasts, the designation of RCMD with ring
sideroblasts (RCMD-RS) can be made; although in the setting of multilineage
dysplasia, there is no clear prognostic impact if ring sideroblasts are found or not.
Cases that meet the criteria for RCMD, but have persistently 1% blasts in the
blood, should be considered as MDS, unclassifiable (see below), while those with
multilineage dysplasia and ⬍5% blasts in the bone marrow but 2% to 4% blasts
in the blood should be classified as RAEB. Patients with RCMD have a worse
outcome (reported median survival times are 17–33 months) than patients with RA.
Refractory Anemia with Excess of Blasts (RAEB)

RAEB is used to describe MDS with 5% to 19% blasts in the bone marrow or
blood. However, if there are ⬍5% blasts in the bone marrow, the finding of 2% to
4% blasts in the peripheral blood is sufficient for the diagnosis. Two subcategories
are recognized. RAEB-1 is defined as having 5% to 9% blasts in the bone marrow
or 2% to 4% in the blood. If blasts are 10% or more in the marrow, or 5% or more
in the blood, the designation should be RAEB-2. Several investigators have shown
that patients with 10% or more blasts in the marrow and 5% or more in the blood
(RAEB-2) have a worse survival and higher rate of transformation to AML than
those with 5% to 9% blasts in bone marrow (RAEB-1). RAEB is a serious disorder,
regardless of whether it transforms to overt acute leukemia. Although 30% to 40%
of patients with RAEB develop AML, more will die from the complications of
neutropenia, thrombocytopenia, or anemia. Median survival times for RAEB-1 is
∼18 months versus 10 months for those with RAEB-2.
MDS Associated with an Isolated Del(5q) Chromosomal Abnormality

This subtype of MDS is characterized by anemia, with or without other cytopenias,
and interstitial deletion of the long arm of chromosome 5q as the sole cytogenetic
abnormality. Myeloblasts comprise ⬍5% of nucleated bone marrow cells and ⬍1%
of peripheral blood leukocytes; Auer rods are not seen. An interstitial deletion of
the long arm of chromosome 5 is a common abnormality in MDS, and may be
seen in isolation or as part of a more complex karyotype. A unique syndrome
occurs in which del(5q) is the sole chromosomal abnormality, associated with a
prevalence in females with a refractory anemia, normal or high platelet count,
megakaryocytes with nonlobated, monolobated or hypolobated nuclei, fewer than
5% blasts, and a relatively good prognosis. The critical gene in this syndrome
may be different than the genes associated with deletions of 5q found in other
types of MDS or AML, and recently, a defect in ribosomal protein function due to
haploinsufficiency of RPS14 has been implicated in the pathogenesis of the “5q−
syndrome” (62). Patients with isolated del(5q) have a high response rate to the
drug lenalidomide (see chap. 16).
Myelodysplastic Syndrome, Unclassifiable (MDS, U)

This subtype encompasses those cases that do not fit easily into the other cate-
gories of MDS. Three possible situations which qualify a patient for this category
include (1) patients with RCUD or RCMD with 1% blasts in the blood found
on two occasions, (2) MDS with morphologic unilineage dysplasia associated
with pancytopenia, and (3) patients with persistent cytopenias lacking diagnostic
morphologic features of MDS or of any specific subgroups of MDS (i.e., ⬍10%

dysplastic cells in any lineage) but with cytogenetic abnormalities considered as
presumptive evidence of MDS (see chap. 19).
DIAGNOSTIC DIFFICULTIES IN MDS

Differentiation of MDS from Non-Neoplastic Disorders
Morphologic changes of myelodysplasia do not necessarily equate with a diag-
nosis of MDS. Reactive conditions associated with non-clonal dysplastic changes
that may resemble MDS are listed in Table 2. Vitamin B12 and folic acid defi-
ciency lead to megaloblastic changes and often severe dyserythropoiesis, and such
deficiencies must always be excluded before making a diagnosis of MDS, particu-
larly if dyserythropoiesis is the only finding. Heavy-metal intoxication, particularly
arsenic poisoning (63), as well as copper deficiency (64), can induce marked, some-
times bizarre morphologic abnormalities in the erythroid and granulocytic series.
A number of drugs and medications, ranging from alcohol to hematopoietic growth
factors [Fig. 3(A) see colour insert], can also induce myelodysplastic features. Con-
genital dyserythropoietic anemias may be difficult to distinguish from RA (65).
In each case in which myelodysplasia is seen, it is important to consider
other possibilities. In general clinical practice, it is important to remember that
Table 2 Dyshematopoietic Changes MDS-Like (Non-Clonal)

Dyserythropoiesis Dysgranulopoiesis Dysmegakaryopoiesis
B12/folate deficiencies Chemotherapy/bone marrow Infections (HIV)

regeneration
Methotrexate/other B12/folate deficiency Chemotherapy
chemotherapy
Alcohol and other toxic CSFs treatment Paraneoplastica
compounds
Autoimmune conditions Infections (HIV, B19) Autoimmunea
Aplastic anemia/PNH Paraneoplastic Post-BM transplant
Post-BM transplan Congenital (Fanconi anemia) TMPD (Down
syndrome)
Infections (parvovirus, HIV) Drug (purine analog)
Hemophagocytic
lymphohistiocytosis
Congenital dyserythropoietic
anemias
aThey have overlapping morphological features including the presence of myelofibrosis, a finding
which may raise a question of MDS-F.
Abbreviations: HIV, human immunodeficiency virus; CSF, colony stimulating factors; B19, parvovirus
B19; PNH, paroxysmal nocturnal hemoglobinuria; BM, bone marrow; TMPD, transient myeloprolif-
erative disease (Down syndrome associated)
(A) (B)
Figure 3 (A) Bone marrow aspirate in a patient receiving G-CSF for neutropenia after
chemotherapy. The transient increase in the number of promyelocytes may suggest a wrong
diagnosis of myeloid neoplasm, perhaps an MDS/MPN. (B) Hypercellular bone marrow in
a patient with chronic anemia which may raise the possibility of low-grade MDS. However,
the clinical history of a metastatic lung carcinoma, the lack of peripheral blood/bone
marrow dysplasia, and the absence of cytogenetic abnormalities strongly favor a diagnosis
of paraneoplastic marrow hyperplasia. (see color insert)
anemia associated with high MCV and low reticulocytes may be seen in patients
with hypothyroidism or liver failure. In most of these cases, the anemia is not
megaloblastic, and the only thing in common with MDS is the presence of ery-
throcyte macrocytosis. Among the anemias and cytopenias most commonly seen
in clinical practice, it is worthwhile mentioning autoimmune conditions (these
can also cause myelofibrosis, i.e., autoimmune myelofibrosis) as well as other
hematopoietic neoplasms (e.g., lymphomas) and nonhematopoietic malignancies
(paraneoplastic myelodysplasia) [Fig. 3(B)]. All of these can, to a certain extent,
mimic the presentation of MDS.
Viral disorders, particularly HIV infection, may be associated with
myelodysplastic features in the blood and marrow. In HIV infection, cytopenia(s)
are common, and the marrow is often hypercellular with evidence of apoptosis
and a pink background of cellular debris (66). Increased marrow plasma cells
and lymphocytes and variable degrees of fibrosis are also common (67). Dysery-
thropoiesis is the most frequently reported myelodysplastic change, and has been
described in up to 70% of patients with HIV infection, but dysmegakaryopoiesis
as well as dysgranulopoiesis have also been described. There are often several
potential causes for the dyspoietic features in patients with HIV infection, includ-
ing direct effects of the HIV virus, concomitant infections, medications such
as AZT, and autoimmune phenomena. Other viruses may also cause dysplastic
changes, such as parvovirus infection. Cytomegalovirus infection has been asso-
ciated with cytopenias and dysplastic features in immunocompromised as well as
non-immunocompromised patients.
In summary, the diagnosis of MDS has serious implications, and every

effort to substantiate it with clinical, morphologic, and laboratory evidence, such
as detection of cytogenetic abnormalities, must be made.
Hypocellular MDS (h-MDS)

In 90% to 95% of patients with MDS, the marrow cellularity is normal or increased
relative to that expected for the patient’s age, but the marrow is hypocellular in
about 5% to 10% cases (10,31–34,68). Hypocellularity is defined as bone marrow
cellularity of ⬍30% in a patient less than 60 years of age, or ⬍20% in patients
older than 60 years of age (69). In MDS, it is rare that the marrow cellularity falls
below 10%, but such extreme hypocellularity is regularly seen in cases of severe
aplastic anemia AA (70). In patients with markedly fatty marrow with a normal
karyotype, the distinction of hypoplastic MDS from AA may be difficult and is
sometimes impossible [Fig. 4(A) see colour insert].
Hypoplastic MDS (h-MDS) is more frequent in women, and the incidence
is proportional to age, just as with other forms of MDS. Previous genotoxic expo-
sure or therapy needs to be excluded, since hypocellular marrows can be seen
in cases of therapy related–MDS (30). Generally, h-MDS is associated with pro-
nounced cytopenias, a finding which may suggest a diagnosis of AA. Fatty bone
marrows usually cause poor-quality aspirates, compounding diagnostic difficul-
ties. A careful inspection of peripheral blood and bone marrow touch preparation
smears together with a detailed analysis of adequately representative bone marrow
biopsy sections must supplement examination of the poorly cellular aspirate. In
this diagnostic context, megaloblastic features alone have insufficient specificity
(A) (B)
Figure 4 Hypoplastic MDS. (A) In cases in which the bone marrow cellularity is markedly
reduced, h-MDS may be morphologically indistinguishable from aplastic anemia. (B)
CD34, by identifying a normal to increased frequency of positive precursor cells, suggests
a diagnosis of h-MDS in this case. The low frequency of precursor cells seen in aplastic
anemia (usually ⬍30% of the normal CD34 value) often causes a CD34 “completely
negative” immunohistochemical result. (see color insert)
to establish a diagnosis of MDS, because megaloblastic features and macrocytosis

are common findings in AA as well.
Within the subgroup of MDS with hypoplastic bone marrow, classification
according to FAB criteria reveals a majority of patients with RA [66.7% in one
series (32)]. The presence of ⬎20% myeloblasts in a hypocellular marrow supports
a diagnosis of hypocellular AML instead of MDS. Dysplastic features other than
dyserythropoiesis occur less frequently in h-MDS, and are of lower grade in
comparison to MDS with normocellular or hypercellular marrow. Since h-MDS
patients do not have an increased proportion of blasts in the aspirate or in the
bone marrow biopsy, the separation from AA may be problematic. This is further
compounded by the high proportion of cases showing a mesenchymal reaction,
especially an increase of mast cells, and reactive lymphoid follicles—features
similar to those observed in bone marrow biopsies obtained from patients with AA.
Nevertheless, the differential diagnosis is still largely based on bone mar-
row histology and immunohistology. The detection within a markedly hypocel-
lular bone marrow of “residual” megakaryopoiesis, particularly in association
with dysmegakaryopoiesis and dysgranulopoiesis (the latter in peripheral blood or
marrow smears/touch preparations), would support the diagnosis of h-MDS rather
than AA. If reticulin fibrosis is detected, this would also favor h-MDS over AA.
Immunohistochemistry can help in distinguishing h-MDS from acquired AA, the
former disorder being characterized by higher CD34 expression as compared to
AA (32,34) [Fig. 4(B)]. Flow cytometry may be less helpful (71). It has also been
reported that immunohistochemical stains for PCNA (proliferating cell nuclear
antigen) may be helpful in distinguishing h-MDS from AA (32,33). In h-MDS,
there are more CD34+ cells and more PCNA+ cells than in AA. Finally, bone
marrow cellularity does not appear to be an important prognostic factor in MDS,
because patients with h-MDS have a similar prognosis to those cases of MDS
with normo/hypercellular marrows. The distinction from AA, however, is prog-
nostically significant because the risk of progression to acute leukemia is much
greater in h-MDS (72). Cytogenetic studies may be of particular value in this
group, as abnormalities of chromosomes 5 and/or 7 are frequently observed in
h-MDS and support a diagnosis of MDS over AA. Nevertheless, AA and h-MDS
have a number of overlapping features, such as the appearance of a clone of
PNH cells (73,74) or evidence of T-cell mediated myelosuppression (75) that may
respond to immunosuppressive therapy (76), suggesting that they share a common
pathophysiologic pathway (77) (see chap. 19).
MDS with Fibrosis (MDS-F)

About 5% to 10% of cases of MDS have significantly increased marrow reticulin
fibers or even collagen fibrosis (23,78–80). MDS-F is an MDS variant which is
characterized by a marked increase in bone marrow reticulin fibers and presents
with pancytopenia and minimal or absent organomegaly. Such cases are generally
easily recognized if the peripheral smear and bone marrow touch preparations
Figure 5 MDS-F. An aggressive MDS subset which can be separated by paying attention
to the histological features including the presence of small/dwarf megakaryocytes within
the context of multilineage dysplasia and of an increased number of blasts. These features
are not seen in cases of myelofibrotic MPN (e.g., PMF). Clinical correlation is also recom-
mended to rule out splenomegaly and any previous relevant history (e.g., pre-existing other
myeloid neoplasm, exposure to chemotherapy, and/or radiotherapy). (see color insert)
are carefully examined, and if the trephine biopsy sections are well prepared
(Fig. 5 see colour insert). Overall, patients with MDS and marrow fibrosis are
reported to have shorter survival times than those without this feature. The marrow
shows trilineage dysplasia with prominent dysmegakaryopoiesis. In most cases an
increased number of blasts are seen. The blasts are more easily documented in
the marrow biopsy than in the aspirate, the latter being frequently suboptimal due
to the presence of myelofibrosis. For a case to qualify as MDS-F, a silver-stained
reticulin fibrosis score of ⬎2 on the basis of the system proposed by Manoharan
and colleagues (81) should be obtained.
MDS-F accounts for 10% to 15% of primary MDS cases and up to 50% of
therapy related–MDS. Within the subgroup of MDS-F, the classification according
to FAB criteria reveals a majority of patients with RAEB, especially RAEB-2. Only
rare cases of RA with fibrosis have been reported. These rare cases of RA with
fibrosis seem to share the poor prognosis associated with the RAEB with fibrosis
subtypes. In MDS-F, increased CD34 expression in the marrow is often observed,
and this marker can be used to assess the blast count. Often, cases with of MDS-F
have prominent megakaryopoiesis, which is characterized by a wide spectrum
of sizes of the megakaryocytes, ranging from micro-megakaryocytes to enlarged
forms. The use of antibodies reactive with megakaryocytes has confirmed the
higher number of these cells in MDS-F cases than either in normal subjects or
patients affected by MDS without fibrosis (23). Immunohistology can be useful in
detecting the small forms (micromegakaryocytes) and confirming the diagnosis.
The differential diagnosis of MDS-F includes several myelofibrotic myeloid
neoplasms. Cases of MDS-F may morphologically overlap acute panmyelosis with
fibrosis (APMF) (acute myelofibrosis), and these disorders may be very difficult
if not impossible to distinguish. Clinically, APMF usually presents with an abrupt
onset with fever and bone pain, and a very aggressive course. In APMF, the histol-
ogy of the marrow shows marked fibrosis (⬎2 according to the Manoharan grading
system), severe trilineage dysplasia associated with numerous dwarf megakary-
ocytes, and an increased number of blasts. Based on bone marrow biopsy, a median
blast value of 22.5% (range of ∼20–25%) was reported (24). Most of the blasts
express CD34 and are negative for megakaryocyte associated markers (24). The
differential diagnosis includes acute megakaryocytic leukemia, which can also
morphologically overlap APMF, as well as other types of AML associated with
a fibrotic bone marrow. Classic MPN, such as primary myelofibrosis (chronic
idiopathic myelofibrosis), can be usually easily distinguished by their character-
istic morphologic features (e.g., large to giant megakaryocytes with cloud-like
nuclei) and, in myelofibrotic primary myelofibrosis, the presence of significant
splenomegaly (82,83).
MDS-F patients show an unfavorable prognosis mainly attributable to
complications deriving from pancytopenia and continuous transfusions, with
a life expectancy of 9.6 months, compared with 17.4 months in MDS without
fibrosis (78).
MYELODYSPLASTIC/MYELOPROLIFERATIVE NEOPLASMS (MDS/MPN)

The disorders that fall into this group, chronic myelomonocytic leukemia
(CMML), juvenile myelomonocytic leukemia (JMML), and atypical chronic
myeloid leukemia (aCML), have features of both MPN as well as of MDS (84),
and belong to their own separate group in the WHO system (50). In some patients,
myelodysplastic features may be prominent, whereas in others, the process may
be more myeloproliferative. The distinction of these disorders from MDS has been
addressed in several recent review articles (85,86). The principal entities in the
MPN/MDS overlap group are found listed in Table 3.
Chronic Myelomonocytic Leukemia (CMML)

CMML is a heterogeneous disorder. In a majority of the cases, particularly those
with a high white blood count (WBC), splenomegaly may be present, and the
disorder resembles an MPN rather than MDS (87). In other cases, the WBC is low,
Table 3 MDS/MPN and Related Disorders

Chronic myelomonocytic leukemia
Atypical chronic myeloid leukemia
Juvenile myelomonocytic leukemia
Myelodysplastic/myeloproliferative neoplasm, unclassifiable
Refractory anemia with ring sideroblasts associated with thrombocytosis
5q−/JAK2 positive atypical myeloproliferative neoplasm
Myelodysplastic/myeloproliferative neoplasm, unclassifiable associated with isolated
isochromosome 17q
there is no splenomegaly, dysplasia is prominent, and the disorder has more in com-
mon with MDS (87,88). Some authors have suggested that dividing CMML into a
myeloproliferative type and a myelodysplastic type based on the WBC, but there is
currently no evidence that such a separation is clinically or biologically meaning-
ful (87,88). Furthermore, patients may exhibit variable leukocyte and monocyte
counts throughout the course of their disease. Some patients may first present with
the features of RA without an absolute monocytosis, and later develop sufficient
numbers of monocytes in the blood or marrow to meet the criteria for CMML.
The diagnostic criteria for CMML are similar in both the WHO classification
and the FAB classification (84). The diagnosis of CMML is made when there is a
persistent, unexplained absolute monocytosis in the blood equal to or greater than
1 × 109 /L. In the blood, the monocytes are often normal morphologically, but they
may be mildly atypical and have hyperlobulated nuclei, increased cytoplasmic
basophilia or prominent cytoplasmic granularity. There may also be immature
forms (promonocytes) present. There is often evidence of dysgranulopoiesis, but
this is often less prominent in patients with higher WBC counts. In the marrow, the
findings are often similar to RAEB, but with monocytosis (⬎5%; normal marrow
range 0–4% monocytes) with a median number of monocytes, based on naphtyl
butyrate esterase positivity, of 10% (29) blasts range from ⬍5% to 19%. Reticulin
fibrosis is present in up to 30% of patients, and even collagen fibrosis may be
observed. Rare patients present with eosinophilia ⬎1.5 × 109 /L and are found
to have rearrangement of PDGFRB, usually in association with a translocation
between chromosomes 5 and 12; such cases should be classified as a myeloid
neoplasm with rearrangement of PDGFRB, rather than merely as CMML because
they are likely to respond to imatinib therapy.
CMML is further divided into two subgroups, depending on the number of
blasts in the blood and marrow. These include CMML-1 with blasts ⬍5% in the
blood or ⬍10% in the bone marrow; and CMML-2, which is characterized by
blasts 5% to 19% in the blood or 10% to 19% in the bone marrow, or when Auer
rods are present. The finding of ⬎20% blasts (promonocytes are considered as
blast equivalent and added to the blast count) in the blood or the bone marrow
indicates AML rather than CMML (20).
The separation of CMML from MDS is aided by the demonstration of
monocytic differentiation in the former. This can be accomplished by cytochemical
staining with esterase (aspirate smears) or immunohistochemistry of bone marrow
biopsy (29). Rare cases of CML that are associated with the BCR–ABL fusion
variant that results in a p190 fusion protein may have monocytosis and resemble
CMML, so that cytogenetic studies to exclude this possibility should always be
performed (89).
Atypical Chronic Myeloid Leukemia (aCML)

Although a marked degree of dysgranulopoiesis “MDS-like” characterizes the
rare MDS/MPN entity termed atypical chronic myeloid leukemia (aCML), its
distinctive hallmark is the presence of “effective” granulopoiesis (84–86,90,91).
In aCML, the WBC is increased (median values 35 × 109 /L to 96 × 109 /L)

and shifted to the left with immature granulocytes, including blast cells (usually
⬍5%; blasts are always ⬍20%), promyelocytes, and myelocytes that account for
10% to 20% of peripheral blood white cells. Dysplastic neutrophils are typically
seen. Granulocytes may show typical pseudo-Pelger-Huët changes, mononuclear
cell-like morphology due to nuclear hypolobation/condensation, and cytoplas-
mic hypogranularity. Hyperlobulation and/or abnormalities in the nuclear chro-
matin pattern (abnormal chromatin clumping) can also be observed. Although
the absolute number of monocytes may rarely be high, their relative number is
always ⬍10%. In contrast to CML, basophilia is not prominent in aCML, usually
accounting ⬍2% of peripheral blood white cells.
The bone marrow is hypercellular, with an elevated myeloid-to-erythroid
ratio (but usually less than 10:1), due to granulocytic hyperplasia. In contrast,
the amount of erythropoiesis and megakaryopoiesis varies from case to case.
Megakaryocytes can be reduced in number, particularly in cases of aCML associ-
ated with thrombocytopenia. Although trilineage dysplasia can be seen, the dys-
plastic changes are usually more prominent in the granulocytic lineage. Megakary-
ocytic dysplasia can be similar in appearance to that seen in MDS displaying a
predominance of small/dwarf forms. However, this may be less apparent in cases
in which the megakaryocytes are reduced in number. An increased proportion of
blasts (always ⬍20%) can be seen. The identification of these cells in bone marrow
biopsies may be facilitated by immunostaining with CD34 (29,86). Mild fibrosis
is also frequently seen at presentation or with disease progression.
When interpreting a bone marrow examination, the main differerential diag-
nosis is not with MDS but rather with CMML. The latter can be excluded by
observing the lack of peripheral blood monocytosis and by paying careful attention
to the WHO guidelines (86). Finally, it cannot be overemphasized that cytogenetic
and molecular genetic studies are essential in the diagnosis of aCML in order
to exclude t(9;22)(q34;q22) or the BCR/ABL fusion product of usual-type CML.
Some cases reported in the literature as “atypical CML” have been associated with
abnormalities of PDGFRA; such cases usually show eosinophilia and should be
classified as myeloid neoplasm with rearrangement of PDGFRA. Other ancillary
studies, particularly immunophenotyping studies, do not usually help unless blast
cell numbers are elevated.
MDS/MPN, Unclassifiable (MDS/MPN,U) and Miscellaneous Entities

with MDS/MPN-Like Features
Some cases demonstrate features of both myelodysplasia and myeloproliferative
syndromes and do not fit well into any of the previously mentioned categories.
Many of these cases have typical features of a myelodysplastic syndrom as well
as an atypical finding more suggestive of a MPN, such as marked marrow fibrosis,
leukocytosis, or organomegaly. Such cases may be termed MDS/MPN, unclassi-
fiable, with a comment describing the atypical findings (86,92). MDS/MPN,U are
rare—only 2% of the cases classified as MDS in one series (92). The prognosis is
poor, with a median survival of 21 months (92).
Refractory Anemia with Ring Sideroblasts and Thrombocytosis

A provisional entity within the group of MDS/MPN,U is a syndrome which has
features of refractory anemia with ring sideroblasts and an elevated platelet count,
called refractory anemia with ringed sideroblasts and thrombocytosis (RARS-T)
(51–53). This rare entity shows no sex predilection or specific cytogenetic abnor-
mality, and must be differentiated from the other conditions which also display
thrombocytosis, such as 5q− syndrome, MDS with abnormality of chromosome
3q21q26, and classical MPN (93). The hallmark of RARS-T is the presence
of an abnormal erythropoiesis similar to that seen in cases of RARS, which is
present in a hypercellular marrow characterized also by increased megakary-
opoiesis with large MPN-like forms. Their overlapping dyserythropoietic features
include a tendency to abnormal accumulation of iron in mitochondria producing
ring sideroblasts. Interestingly, both RARS and RARS-T share a high frequency
of hemochromatosis-associated gene mutations, a molecular anomaly originally
described in cases of RARS, which has been also shown to occur in the majority
of cases of RARS-T (53).
The demonstration in RARS-T of a high frequency of positivity for JAK2
mutation, possibly acquired as a secondary genetic event (94), can perhaps explain
its relationship to a classical myeloproliferative disorder. In addition, MPL W515
mutations have been described in rare patients with RARS-T. The prognosis
of RARS-T subgroup is better than that of the whole group of patients with
MDS/MPN,U (95,96). In several studies, the RARS-T showed a median survival
comparable or better than RARS (96), a prognosis which is definitely better
than that of CMML-1 or CMML-2 patients. The better prognosis may be related
to the presence of JAK2 positivity (96,97), as RARS-T patients without JAK2
mutations do more poorly. While the JAK2 positive cases of RARS-T have a
better prognosis than the rest of the MDS/MPN group, it should not be forgotten
that they still have a worse outcome than those with essential thrombocythemia,
with which this condition may be easily confused. In particular, RARS-T patients
are more likely to develop AML (98), and therefore a correct diagnosis is of
fundamental importance for defining the prognosis of the individual patient. The
importance of performing an iron stain on a bone marrow aspirate smear while
working up a patient suspected to have a myeloid neoplasm, cannot therefore be
overemphasized (98).
5q−/JAK2 Positive Atypical Myeloproliferative Disease

The simultaneous presence of a 5q− abnormality and a JAK2 mutation identify
this rare disease with myeloproliferative characteristics (86,99). In contrast with
the classical 5q− syndrome, the anemia seen in these cases is more often nor-
mochromic. Higher platelet counts and higher WBCs are also more often seen
in these cases (as opposed to cases of 5q− with unmutated JAK2). In 5q−/JAK2
positive cases, the bone marrow shows a prominent myeloid proliferation. The
megakaryocytes show the characteristic hypolobated or nonlobated nuclei asso-
ciated with the 5q− syndrome. Similar cases of “5q− syndrome” presenting
with chronic myeloproliferative disorders–like manifestations, thrombocytosis in
particular, have been described in the past (100). More recently, it has been docu-
mented that these cases are characterized by a frequency of JAK2 mutations that
is higher than that seen in the “nonmyeloproliferative” 5q− syndrome (99). This
hybrid condition must be differentiated from the 5q− syndrome, a well-defined
subtype of MDS with a relatively benign prognosis. The clinical behavior of cases
with both isolated 5q abnormality and myeloproliferative characteristics is more
variable. However, lenalidomide treatment may be effective in both subtypes.
Other cases characterized by the simultaneous presence of an isolated 5q− abnor-
mality and a JAK2 mutation have not shown myeloproliferative features (101).
The existence of this entity as a distinct disorder thus remains controversial.
MDS/MPDU Associated with Isolated Isochromosome 17q

Additionally, a mixed MDS/MPN disorder associated with isochromosome 17q
has been described (102); it occurs with a male predominance in adults and is
associated with severe hyposegmentation of neutrophil nuclei, variable monocy-
tosis, and a high rate of transformation to AML. The marrow is hypercellular and
may show severe dysmegakaryopoiesis frequently associated with myelofibrosis.
In some instances, the bone marrow features may even resemble those seen in pri-
mary myelofibrosis (Orazi A, unpublished observation). It is still unclear whether
cases of isochromosome 17q with monocytosis compatible with a diagnosis of
CMML should still be considered as part of that disease spectrum for purposes of
classification.
Other poorly characterized MDS or MDS/MPN neoplasms may occur in

association with systemic mast cell disease, where these neoplasms represent one
of the types of clonal nonmast cell lineage hematological diseases associated with
systemic mastocytosis (103).
Therapy-Related Myelodysplastic Syndrome

Although it might seem as if therapy-related MDS (t-MDS) and therapy-related
AML (t-AML) belong to a separate classification categories, in practice, they
are often considered together because the underlying biology of their disease is
similar, particularly for those cases that have followed alkylating agent and/or
radiation therapy (104). Additionally, in many centers, their clinical management
is quite similar. Some studies have shown that classification of t-MDS by the same
criteria used for de novo MDS does not identify prognostic groups well, nor does
the blast count have the same prognostic importance in t-MDS as it does in de
novo MDS (105). Rather, the median survival times in these patients are often
extremely poor, and much shorter than for the corresponding de novo subtypes
(105).
Two major types of t-MDS/t-AML are recognized based of the causative
agents: alkylating agent/radiation related and topoisomerase II inhibitor related
(see chap. 8). The alkylating agent/radiation type of t-MDS typically occurs late,
usually ⬎5 years after use of treatment and is associated with −7/del7q and/or
−5/del5q. By the FAB system, most cases of t-MDS fall within the RAEB category
(30). The second type of t-MDS/t-AML occurs relatively early, in 2 to 3 years
after the use of agents targeted at topoisomerase II (epipodophyllotoxins, e.g.,
etoposide, teniposide, doxorubicin), and presents with translocations involving
chromosomal bands 11q23 and 21q22, or less frequently, other translocation which
are also typical of the de novo AML (106). Many of these patients progress directly
to AML without a previously documented MDS phase. t-MDSs with 17p deletions
have morphologic features similar to those described before for the de novo cases
with the same cytogenetic abnormality and are similarly characterized by the
presence of p53 mutations (107,108).
The clinical features of t-MDS are similar to those seen in aggressive primary
MDS cases except for the more pronounced pancytopenia and anisopoikilocytosis,
which are usually associated with the former group. Occasional nucleated RBCs
are seen in the peripheral blood. The bone marrow is, on average, less cellular
than in the primary cases, and significant reticulin fibrosis (≥2+) is also more
common (30). In spite of the similar degrees of fibrosis in both primary MDS-F
and t-MDS, the latter group differs in terms of the number of megakaryoblasts
and megakaryocytes which are significantly higher in primary MDS (23). CD34
expression is almost always increased in these aggressive MDS subtypes (23,30).
In addition, p53 protein overexpression can be frequently observed, particularly in
cases associated with prior alkylating agent chemotherapy (35,37). p53 expression
was found to be associated with increased apoptosis in marrow hematopoietic cells
and severe ineffective hematopoiesis (37).
“Early” MDS and Idiopathic Cytopenia of Unknown Significance (ICUS)

Occasionally, one is confronted with patients who have persistent cytopenia (usu-
ally anemia) for which no underlying cause can be found, but in whom there is
insufficient morphologic evidence to support the clinical suspicion of MDS. For
some of these patients, a “working diagnosis” of idiopathic cytopenia of unknown
significance (ICUS) may be considered (109), although it is important to point
out that this is not synonymous with a diagnosis of MDS. In such cases, it is
necessary to explore every possible cause for the cytopenia(s), and to rule out
disorders such as autoimmune disorders, lymphoproliferative diseases (particu-
larly large granular lymphocyte proliferations and hairy cell leukemia), infection,
etc. Re-examination of the morphologic features with review by a colleague is
also recommended. Immunohistochemical studies on biopsies may uncover an
increase in the number of blasts that was not appreciated previously, and should
always be performed. Cytogenetic studies are strongly indicated in such cases, and
FISH may show a clonal abnormality when karyotype studies do not. If a clonal
cytogenetic abnormality is found, the diagnosis of MDS becomes more likely,
and indeed, if certain abnormalities are present, a presumptive diagnosis of MDS-
U may be made. Still, such patients should be followed carefully over time for
morphologic evidence of dysplasia before an unequivocal diagnosis can be made.
CONCLUSION
The diagnosis of MDS requires integration between morphology, immunopheno-
type, genetic features, and all of these must be interpreted in the light of the patient’s
history and clinical manifestations. Although in this chapter much emphasis has
been placed on the morphologic interpretation of the bone marrow, it is impor-
tant to perform a comprehensive evaluation in each case. This multiparametric
approach forms the basis for the WHO classification of tumors of hematopoietic
and lymphoid tissues. Only by following the principles of the classification can
the hematopathologist reliably and reproducibly identify MDS.
Particularly challenging areas include the separation of “low-grade” MDS
from reactive conditions, distinction of hypoplastic MDS and AA, and correct
identification of the fibrotic and other atypical subtypes of MDS and their sepa-
ration from atypical myeloproliferative disorders or subsets of AML. Additional
difficulties can be experienced when dealing with the important subsets of patients
with therapy-related MDS or MDS arising from a background of various congen-
ital or acquired marrow conditions or after another myeloid disorder. A sizable
proportion of these cases display myelofibrosis the diagnosis may have to rely
largely on a bone marrow biopsy whose interpretation is often greatly helped by
a careful comparison with previously obtained bone marrow samples and by its
integration with peripheral blood findings (FISH included), and clinical evidence.
Novel cytogenetics and molecular genetics approaches are likely to revolu-
tionize the way we classify these diseases in the near future. However, it is highly
unlikely that these new techniques will be capable, on their own, of adequately
stratifying patients for purpose of treatment. Even with the most sophisticated tech-
nology at one’s disposal, an adequate characterization of bone marrow morphology
will represent the best “reality check” available to us for many years to come.
REFERENCES
1. Brunning RD, Orazi A, Germing U et al. Myelodysplastic syndroms/neoplasms,
overview. In: Swerdlow SH, Canpo E, Harris NL et al, eds. World Health Organiza-
tion Classification of Tumors: Pathology and Genetics of Tumors of Haematopoietic
and Lymphoid Tissues. Lyon: IARC Press, 2008:88–93.
2. Vardiman JW. Hematopathological concepts and controversies in the diagnosis and
classification of myelodysplastic syndromes. Hematology Am Soc Hematol Educ
Program 2006:199–204; Review.
myelodysplastic syndromes. Br J Haematol 1982; 51:189.
classification of the myeloid neoplasms. Blood 2002; 100:2292–2302; Review.
5. Orazi A, O’Malley DP, Arber D. Myelodysplastic syndromes. In: Illustrated Pathol-
ogy of the Bone Marrow. New York: Cambridge University Press, 2006:43–57.
6. Fenaux P. Myelodysplastic syndromes: From pathogenesis and prognosis to treat-
ment. Semin Hematol 2004; 41(Suppl. 4):6–12.
7. Aul C, Giagounidis A, Heinsch M, et al. Prognostic indicators and scoring systems
for predicting outcome in patients with myelodysplastic syndromes. Rev Clin Exp
Hematol 2004; 8:E1; Review.
8. Steensma DP, Bennett JM. The myelodysplastic syndromes: Diagnosis and treat-
ment. Mayo Clin Proc 2006; 81:104–130.
J Clin Oncol 2007; 25:3503–3510.
11. Delacrétaz F, Schmidt PM, Piguet D, et al. Histopathology of myelodysplastic syn-
dromes. The FAB classification (proposals) applied to bone marrow biopsy. Am J
Clin Pathol 1987; 87:180–186.
12. Verburgh E, Achten R, Maes B, et al. Additional prognostic value of bone marrow
histology in patients subclassified according to the International Prognostic Scoring
System for myelodysplastic syndromes. J Clin Oncol 2003; 21:273–282.
13. Orazi A. Histopathology in the diagnosis and classification of acute myeloid
leukemia, myelodysplastic syndromes, and myelodysplastic/myeloproliferative dis-
eases. Pathobiology 2007; 74:97–114; Review.
14. Soligo DA, Oriani A, Annaloro C, et al. CD34 immunohistochemistry of bone
marrow biopsies: prognostic significance in primary myelodysplastic syndromes.
Am J Hematol 1994; 46:9–17.
15. Horny HP, Wehrmann M, Schlicker HU, et al. QBEND10 for the diagnosis of
myelodysplastic syndromes in routinely processed bone marrow biopsy specimens.
J Clin Pathol 1995; 48:291–294.
16. Lambertenghi-Deliliers G, Annaloro C, Soligo D, et al. The diagnostic and prog-
nostic value of bone marrow immunostaining in myelodysplastic syndromes. Leuk
Lymphoma 1998; 28:231–239.
17. Baur AS, Meugé-Moraw C, Schmidt PM, et al. CD34/QBEND10 immunostaining in
bone marrow biopsies: An additional parameter for the diagnosis and classification
of myelodysplastic syndromes. Eur J Haematol 2000; 64:71–79.
18. Horny HP, Sotlar K, Valent P. Diagnostic value of histology and immunohistochem-
istry in myelodysplastic syndromes. Leuk Res 2007; 31:1609–1616.
19. Tricot G, De Wolf-Peeters C, Vlietinck R, et al. Bone marrow histology in myelodys-
plastic syndromes. II. Prognostic value of abnormal localization of immature pre-
cursors in MDS. Br J Haematol 1984; 58:217–225.
20. De Wolf-Peeters C, Stessens R, Desmet V, et al. The histological characteriza-
tion of ALIP in the myelodysplastic syndromes. Pathol Res Pract 1986; 181:402–
407.
21. Bellamy WT, Richter L, Sirjani D, et al. Vascular endothelial cell growth fac-
tor is an autocrine promoter of abnormal localized immature myeloid precursors
and leukemia progenitor formation in myelodysplastic syndromes. Blood 2001;
97:1427–1434.
22. Pellegrini W, Facchetti F, Marocolo D, et al. Expression of CD34 by megakaryocytes
in myelodysplastic syndromes. Haematologica 2000; 85:1117–1118.
23. Lambertenghi-Deliliers G, Orazi A, Luksch R, et al. Myelodysplastic syndromes
with increased marrow fibrosis: A distinct clinico-pathologic entity. Br J Haematol
1991; 78:161–166.
24. Orazi A, O’Malley DP, Jiang J, et al. Acute panmyelosis with myelofibrosis: An
entity distinct from acute megakaryoblastic leukemia. Mod Pathol 2005; 18:603–
614.
25. Neiman RS. Erythroblastic transformation in myeloproliferative disorders: Confir-
mation by an immunohistologic technique. Cancer 1980; 46:1636–1640.
26. Pinkus GS, Pinkus JL. Myeloperoxidase: A specific marker for myeloid cells in
paraffin sections. Mod Pathol 1991; 4:733–741.
27. Werner M, Kaloutsi V, Walter K, et al. Immunohistochemical examination of rou-
tinely processed bone marrow biopsies. Pathol Res Pract 1992; 188:707–713.
28. Manaloor EJ, Neiman RS, Heilman DK, et al. Immunohistochemistry can be used
to subtype acute myeloid leukemia in routinely processed bone marrow biopsy
specimens. Comparison with flow cytometry. Am J Clin Pathol 2000; 113:814–822.
29. Orazi A, Chiu R, O’Malley DP, et al. Chronic myelomonocytic leukemia: The role
of bone marrow biopsy immunohistology. Mod Pathol 2006; 19:1536–1545.
30. Orazi A, Cattoretti G, Soligo D,et al. Therapy-related myelodysplastic syndromes:
FAB classification, bone marrow histology, and immunohistology in the prognostic
assessment. Leukemia 1993; 7:838–847.
31. Maschek H, Kaloutsi V, Rodriguez-Kaiser M, et al. Hypoplastic myelodysplastic
syndrome: Incidence, morphology, cytogenetics, and prognosis. Ann Hematol 1993;
66:117–122.
32. Orazi A, Albitar M, Heerema NA, et al. Hypoplastic myelodysplastic syndromes can
be distinguished from acquired aplastic anemia by CD34 and PCNA immunostaining
of bone marrow biopsy specimens. Am J Clin Pathol 1997; 107:268–274.
33. Kitagawa M, Kamiyama R, Kasuga T. Expression of the proliferating cell nuclear
antigen in bone marrow cells from patients with myelodysplastic syndromes and
aplastic anemia. Hum Pathol 1993; 24:359–363.
34. Matsui WH, Brodsky RA, Smith BD, et al. Quantitative analysis of bone mar-
row CD34 cells in aplastic anemia and hypoplastic myelodysplastic syndromes.
Leukemia 2006; 20:458–462.
35. Orazi A, Kahsai M, John K, et al. p53 overexpression in myeloid leukemic disorders
is associated with increased apoptosis of hematopoietic marrow cells and ineffective
hematopoiesis. Mod Pathol 1996; 9:48–52.
36. Merlat A, Lai JL, Sterkers Y, et al. Therapy-related myelodysplastic syndrome and
acute myeloid leukemia with 17p deletion. A report on 25 cases. Leukemia 1999;
13:250–257.
37. Orazi A, Cattoretti G, Heerema NA, et al. Frequent p53 overexpressionin therapy-
related myelodysplastic syndromes and acute myeloid leukemias: An immunohisto-
chemical study of bone marrow biopsies. Mod Pathol 1993; 6:521–525.
38. Ponzoni M, Savage DG, Ferreri AJM, et al. Chronic idiopathic myelofibrosis: Inde-
pendent prognostic importance of bone marrow microvascular density evaluated by
CD105 (endoglin) immunostaining. Mod Pathol 2004; 17:1513–1520.
39. Molica S, Vacca A, Levato D, et al. Angiogenesis in acute and chronic lymphoblastic
40. Pruneri G, Bertolini F, Soligo D, et al. Angiogenesis in myelodysplastic syndromes.
Br J Cancer 1999; 81:1398–1401.
2124 patients. Blood 2007; 110:4385–4395.
42. Stetler-Stevenson M, Arthur DC, Jabbour N, et al. Diagnostic utility of flow cyto-
metric immunophenotyping in myelodysplastic syndrome. Blood 2001; 98:979–
987.
43. Benesch M, Deeg HJ, Wells D, et al. Flow cytometry for diagnosis and assessment
of prognosis in patients with myelodysplastic syndromes. Hematology 2004; 9:171–
177.
45. Pirrucello SJ, Young KH, Aoun P. Myeloblast phenotypic changes in myelodysplasia.
CD34 and CD117 expression abnormalities are common. Am J Clin Pathol 2006;
125:884–894.
prognostic subgroups in low- and intermediate-1-risk myelodysplastic syndromes
by flow cytometry. Blood 2008; 111:1067–1077.
47. Hanson CA, Ross CW, Schnitzer B. Anti-CD34 immunoperoxidase staining in paraf-
fin sections of acute leukemia: Comparison with flow cytometric immunophenotyp-
ing. Hum Pathol 1992; 23:26–32.
48. Font P, Subirá D, Mtnez-Chamorro C, et al. Evaluation of CD7 and terminal deoxynu-
cleotidyl transferase (TdT) expression in CD34+ myeloblasts from patients with
myelodysplastic syndrome. Leuk Res 2006; 30:957–963.
49. Mann KP, DeCastro CM, Liu J, et al. Neural cell adhesion molecule (CD56)-
positive acute myelogenous leukemia and myelodysplastic and myeloproliferative
syndromes. Am J Clin Pathol 1997; 107:653–660.
50. Jaffe ES, Harris NL, Stein H, et al., eds. World Health Organization Classification
Tissues. Lyon, France: IARC Press, 2001.
51. Szpurka H, Tiu R, Murugesan G, et al. Refractory anemia with ringed siderob-
lasts associated with marked thrombocytosis (RARS-T), another myeloproliferative
condition characterized by JAK2 V617 F mutation. Blood 2006; 108:2173–2181.
52. Soupir CP, Vergilo J, Wang SA, Freeman J, Fend F, Hasserjian RP, JAK2 V617 F
Mutation Defines Two Subsets of Refractory Anemia with Ringed Sideroblasts and
Marked Thrombocytosis (RARS-T) with Destinctive Clinical and Morphological
Features. Mod Pathol 2008: 21 Suppl; p275A
53. Nearman ZP, Szpurka H, Serio B, et al. Hemochromatosis-associated gene mutations
in patients with myelodysplastic syndromes with refractory anemia with ringed
sideroblasts. Am J Hematol 2007; 82:1076–1079.
54. Rosati S, Mick R, Xu F, et al. Refractory cytopenia with multilineage dysplasia:

Further characterization of an “unclassifiable” myelodysplastic syndrome. Leukemia
1996; 10:20–26.
55. Balduini CL, Guarnone R, Pecci A, et al. Multilineage dysplasia without increased
blasts identifies a poor prognosis subset of myelodysplastic syndromes. Leukemia
1998; 12:1655–1656 (letter).
728.
new classification of primary myelodysplastic syndromes: A retrospective analysis
of 1600 patients. Leuk Res 2000; 24:983–992.
does make a difference. Blood 2004; 103;3265–3270.
59. Malcovati L, Porta MG, Pascutto C, et al. Prognostic factors and life expectancy
in myelodysplastic syndromes classified according to WHO criteria: A basis for
clinical decision making. J Clin Oncol 2005; 23:7594–7603.
60. Matsuda A, Germing U, Jinnai I, et al. Improvement of criteria for refractory cytope-
nia with multlineage dysplasia according to the WHO classification based on prog-
nostic significance of morphological features of patients with refractory anemia
according to the FAB classification. Leukemia 2007; 21:678–686.
61. Verburgh E, Achten R, Louw VJ, et al. A new disease categorization of low-grade
myelodysplastic syndromes based on the expression of cytopenia and dysplasia in
one versus more than one lineage improves on the WHO classification. Leukemia
2007; 21:668–677.
62. Ebert BL, Pretz J, Bosco J, et al. Identification of RPS14 as a 5q− syndrome gene
by RNA interference screen. Nature 2008; 451:335–340.
63. Westhoff DD, Samaha RJ, Barnes A Jr. Arsenic intoxication as a cause of mega-
loblastic anemia. Blood 1975; 45:241–246.
64. Gregg XT, Reddy V, Prchal JT. Copper deficiency masquerading as myelodysplastic
syndrome. Blood 2002; 100:1493–1495.
65. Wickramasinghe SN. Congenital dyserythropoietic anaemias: clinical features,
haematological morphology and new biochemical data. Blood Rev 1998; 12:178–
200.
66. Thiele J, Zirbes TK, Bertsch HP, et al. AIDS-related bone marrow lesions—
Myelodysplastic features or predominant inflammatory-reactive changes (HIV-
myelopathy)? A comparative morphometric study by immunohistochemistry with
special emphasis on apoptosis and PCNA-labeling. Anal Cell Pathol 1996; 11:141–
157.
67. O’Malley DP, Sen J, Juliar BE, et al. Evaluation of stroma in human immunod-
eficiency virus/acquired immunodeficiency syndrome-affected bone marrows and
correlation with CD4 counts. Arch Pathol Lab Med 2005; 129:1137–1140.
68. Yoshida Y, Oguma H, Maekawa T. Refractory myelodysplastic anaemias with
hypocellular bone marrow. J Clin Pathol 1995; 41:763.
New proposals. Br J Haematol 1995; 91:612–617.
70. Brodsky RA, Jones RJ. Aplastic anaemia. Lancet 2005; 365:1647–1656
71. Guyotat D, Campos L, Thomas X, et al. Myelodysplastic syndromes: A study of

surface markers and in vitro growth patterns. Am J Hematol 1990; 34:26–31.
72. Barrett J, Saunthararajah Y, Molldrem J. Myelodysplastic syndrome and aplastic
anemia: Distinct entities or diseases linked by a common pathophysiology? Semin
Hematol 2000; 37:15–29.
73. Karadimitris A, Luzzatto L. The cellular pathogenesis of paroxysmal nocturnal
haemoglobinuria. Leukemia 2001; 15:1148–1152.
74. Kaiafa G, Papadopoulos A, Ntaios G, et al. Detection of CD55- and CD59-deficient
granulocytic populations in patients with myelodysplastic syndrome. Ann Hematol
2007; Dec 19; [Epub ahead of print].
75. Wang SA, Yue G, Hutchinson L, et al. Myelodysplastic syndrome with pure red cell
aplasia shows characteristic clinicopathological features and clonal T-cell expansion.
Br J Haematol 2007; 138:271–277.
76. Lim ZY, Killick S, Germing U, et al. Low IPSS score and bone marrow hypocellu-
larity in MDS patients predict hematological responses to antithymocyte globulin.
Leukemia 2007; 21:1436–1441.
77. Sloand EM, Rezvani K. The role of the immune system in myelodysplasia: Impli-
cations for therapy. Semin Hematol 2008; 45:39–48.
78. Maschek H, Georgii A, Kaloutski V, et al. Myelofibrosis in primary myelodysplastic
syndromes: A retrospective study of 352 patients. Eur J Haematol 1992; 48:208–
214.
79. Pagliuca A, Layton DM, Manoharan A, et al. Myelofibrosis in primary myelodys-
plastic syndromes: A clinico-morphological study of 10 cases. Br J Haematol 1989;
71:499–504.
80. Steensma DP, Hanson CA, Letendre L, et al. Myelodysplasia with fibrosis: A distinct
entity? Leuk Res 2001; 25:829–838.
81. Manoharan A, Horsley R, Pitney WR. The reticulin content of bone marrow in acute
leukaemia in adults. Br J Haematol 1979; 43:185–190.
82. Thiele J, Kvasnicka HM, Orazi A. Bone marrow histopathology in myeloprolifer-
ative disorders—Current diagnostic approach. Semin Hematol 2005; 42:184–195;
Review.
83. Tefferi A, Thiele J, Orazi A, et al. Proposals and rationale for revision of the World
Health Organization diagnostic criteria for polycythemia vera, essential thrombo-
cythemia, and primary myelofibrosis: Recommendations from an ad hoc interna-
tional expert panel. Blood 2007; 110:1092–1097.
84. Bennett JM, Catovsky D, Daniel MT, et al. The chronic myeloid leukemias: Guide-
lines for distinguishing chronic granulocytic, atypical chronic myeloid, and chronic
myelomonocytic leukemia. Proposals by the French-American-British Cooperative
Leukemia Group. Br J Haematol 1994; 87:746–754.
85. Tefferi A, Elliott MA, Pardanani A. Atypical myeloproliferative disorders: Diagnosis
and management. Mayo Clin Proc 2006; 81:553–563; Review.
86. Orazi A, Germing U. The myelodysplastic/myeloproliferative neoplasms: Myelo-
proliferative diseases with dysplastic features. Leukemia 2008; 22:1308–19. Review.
87. Germing U, Gatterman N, Minning H, et al. Problems in the classification of
CMML—Dysplastic versus proliferative type. Leuk Res 1998; 22:871–878.
88. Germing U, Kundgen A, Gatterman N. Risk assessment in chronic myelomonocytic
leukemia (CMML). Leuk Lymphoma 2004; 45:14311–14318.
89. Melo JV, Myint H, Galton DA, et al. P190BCR-ABL chronic myeloid leukaemia:
The missing link with chronic myelomonocytic leukaemia? Leukemia 1994; 8:208–
211.
90. Vardiman JW, Imbert M, Pierre R, et al. Atypical chronic myeloid leukaemia. In: Jaffe
ES, Harris NL, Stein H, Vardiman JW, eds. World Health Organization Classification
of Tumors: Pathology and Genetics of Tumours of Haematopoietic and Lymphoid
Tissues. Lyon: IARC Press, 2001:53–54.
91. Fend F, Horn T, Koch I, Vela T, Orazi A. Atypical chronic myeloid leukemia as
defined in the WHO classification is a JAK2 V617F negative neoplasm. Leuk Res.
2008; 32:1931–5.
92. Bain B, Vardiman, JW, Imbert M, et al. Myelodysplastic/myeloproliferative dis-
ease, unclassifiable. In: Jaffe ES, Harris NL, Stein H, Vardiman JW, eds. World
Health Organization Classification of Tumors: Pathology and Genetics of Tumours
of Haematopoietic and Lymphoid Tissues. Lyon: IARC Press, 2001:58–59.
93. Schmitt-Graeff A, Thiele J, Zuk I, et al. Essential thrombocythemia with ringed
sideroblasts: A heterogeneous spectrum of diseases, but not a distinct entity. Haema-
tologica 2002; 87:392–399.
94. Kralovics R, Teo SS, Li S, et al. Acquisition of the V617 F mutation of JAK2 is
a late genetic event in a subset of patients with myeloproliferative disorders. Blood
2006; 108:1377–1380.
95. Cannella L, Breccia M, Latagliata R, et al. Clinical and prognostic features of
patients with myelodysplastic/myeloproliferative syndrome categorized as unclassi-
fied (MDS/MPD-U) by WHO classification. Leuk Res 2008; 32:514–516.
96. Atallah E, Nussenzveig R, Yin CC, et al. Prognostic interaction between throm-
bocytosis and JAK2 V617F mutation in the WHO subcategories of melodysplas-
tic/myeloproliferative disease-unclassifiable and refractory anemia with ringed sider-
oblasts and marked thrombocytosis. Leukemia. 2008; 22:1295–8. [Epub ahead of
print].
97. Schmitt-Graeff AH, Teo SS, Olschewski M, et al. JAK2V617F mutation status
identifies subtypes of refractory anemia with ringed sideroblasts associated with
marked thrombocytosis. Haematologica 2008; 93:34–40.
98. Malcovati L, Cazzola M. Myelodysplastic/myeloproliferative disorders. Haemato-
logica 2008; 93:4–6.
99. Ingram W, Lea NC, Cervera J, et al. The JAK2 V617F mutation identifies a subgroup
of MDS patients with isolated deletion 5q and a proliferative bone marrow. Leukemia
2006; 20:1319–1321.
100. Takahashi H, Furukawa T, Hashimoto S, et al. 5q– syndrome presenting chronic
myeloproliferative disorders-like manifestation: A case report. Am J Hematol 2000;
64:120–123.
101. Steensma DP, Tefferi A. JAK2 V617F and ringed sideroblasts: Not necessarily
RARS-T. Blood 2008; 111:1748.
102. McClure RF, Dewald GW, Hoyer JD, et al. Isolated isochromosome 17q: A dis-
tinct type of mixed myeloproliferative disorder/myelodysplastic syndrome with an
aggressive clinical course. Br J Haematol 1999; 106:445–454.
103. Horny HP, Sotlar K, Sperr WR, et al. Systemic mastocytosis with associated clonal
haematological non-mast cell lineage diseases: A histopathological challenge. J Clin
Pathol 2004; 57:604–608.
104. Rowley JD, Olney HJ. International workshop on the relationship of prior therapy to
balanced chromosome aberrations in therapy-related leukemia and myelodysplastic
syndromes and acute leukemia: Overview report. Genes Chromosomes Cancer 2002;
33:331–345.
105. Singh ZN, Huo D, Anastasi J, et al. Therapy-related myelodysplastic syndrome:
Morphologic subclassification may not be clinically relevant. Am J Clin Pathol
2007; 127:197–205.
106. Quesnel B, Kantarjian H, Bjergaard JP, et al. Therapy-related 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; 11:2370–2379.
107. Lai JL, Preudhomme C, Zandecki M, et al. Myelodysplastic syndromes and acute
myeloid leukemia with 17-p deletion. An entity characterized by specific dysgranu-
lopoiesis and a high incidence of p53 mutations. Leukemia 1995; 9:370–381.
108. Sugimoto K, Hirano N, Toyoshima H, et al. Mutations of the p53 gene in myelodys-
plastic syndrome (MDS) and MDS-derived leukemia. Blood 1993; 81:3022–3026.
31:1461–1468.
10
Hypocellular Myelodysplastic Syndrome:
Relationship to Aplastic Anemia and
Hypocellular Acute Myeloid Leukemia
Tomoko Hata and Masao Tomonaga

Department of Hematology, Atomic Bomb Disease Institute, Nagasaki University
Graduate School of Biomedical Sciences, Sakamoto, Nagasaki City, Japan
INTRODUCTION
The myelodysplastic syndromes (MDS) and acute myeloid leukemia (AML) are
usually manifest by a bone marrow that is normocellular or hypercellular, consist-
ing of abnormally proliferating hematopoietic clones (1,2). In contrast, aplastic
anemia (AA) is characterized by hypocellular marrow, and hematopoiesis in AA is
usually polyclonal (though occasionally it can be oligoclonal or even monoclonal).
A minority of patients with MDS (8–20%) (3–9) or with AML (9–14%)
(10–15), however, present with a bone marrow that is hypocellular, compared
to age-matched healthy persons. MDS cases with a hypocellular marrow can be
difficult for clinicians to distinguish from AA. It is also difficult to distinguish
refractory anemia with excess blasts (RAEB) with a hypocellular marrow from
AML with a hypocellular marrow, because the low cellularity makes enumerating
the proportion of blast cells quite difficult. Due to the lack of biological markers
that are specific for either AA or MDS, differentiating between these clinical
entities in the setting of a hypocellular marrow is still fundamentally based on
morphological evaluation of the extent of blood cell dysplasia and the blast cell
percentage (16).
In this chapter, we describe clinical and hematological features of these
overlapping disease categories based on our own series and international reports.
227
228 Hata and Tomonaga
In addition, we discuss recent molecular analyses in hypocellular MDS and AA that

have employed newly developed single nucleotide polymorphism (SNP) arrays,
in order to clarify the margin between these disease entities. This chapter is
complementary to chapter 19, which describes immunotherapy for MDS more
broadly—because of the success of immunotherapy in AA, MDS patients with a
hypocellular marrow are considered by some investigators to be good candidates
for immunsuppressive or immunomodulatory therapy.
PATIENT PROFILE: CASES SELECTED FOR IMMUNOSUPPRESSIVE

THERAPY
Since 1992, we have treated 23 MDS cases and 31 AA cases with immuno-
suppressive agents—of the MDS cases, 9 (39%) were hypocellular for age
and 8 (35%) were normocellular. Patients’ clinical features are summarized in
Tables 1 and 2, respectively. Although hypocellular MDS cases are not well clas-
sified by existing systems, we attempted to make a diagnosis of MDS based on both
the FAB classification and, more recently, the WHO classification (see chap. 1).
Aplastic anemia was diagnosed when pancytopenia was present in the peripheral
blood, and bone marrow was histologically evaluated as hypocellular according
to the age-adjusted criteria of Robert J. Hartsock and colleagues (17), and Nukhet
Tuzuner and John Bennett (18,19) (discussed later in the chapter).
When attempting to diagnose AA, it is important to recognize that mild
dyserythropoiesis (Dys-E) is often observed, but dysgranulopoiesis (Dys-G) is
absent or rare (⬍10% of cells), and megakaryocytes are almost always absent
(16). HLA-DR15 was identified in 10 out of 24 (42%) AA cases examined for
this abnormality. Only one AA case had a clonal karyotypic abnormality, trisomy
8. Most of the AA patients were treated with a combination of anti-thymocyte
globulin (ATG) and cyclosporin A (CyA).
All MDS patients in this series, best fit the FAB category of refractory
anemia (RA); using the WHO classification, 11 (48%) cases were still classified
as RA while 12 (52%) were reclassified as refractory cytopenia with multilineage
dysplasia (RCMD). Most cases of MDS, regardless of marrow cellularity, showed
Dys-G and dysmegakaryocytopoieis (Dys-Meg) in more than 10% of cells in
addition to Dys-E. In seven cases, clonal karyotypic abnormalities were observed.
HLA-DR15 was found in 7 out of 20 (35%) MDS cases examined. A minor
Paroxysmal Nocturnal Hemoglobinuria (PNH) clone was detected in a few patients
(data not presented). In contrast to AA, most MDS cases were treated with CyA
alone or CyA plus FK506; ATG or ALG were used in only five cases.
METHODS TO EVALUATE BONE MARROW CELLULARITY

Histological and Cytological Evaluation
There are several methods to evaluate bone marrow cellularity in clinical samples.
Histological sections of bone marrow core biopsies obtained from the iliac crest are
essential, while clot specimens of bone marrow aspirates may provide additional
Table 1 Patient Profile: Myelodysplastic Syndromes (MDS)
Diagnosis
BM cellularity Response
FAB IPSS HLA- IST
No. Age/sex (WHO) score Histology MRI DR15 Karyotype Regimen of IST Hb PMN Plt Outcome depend
1 47/F RA (RCMD) 0.5 Hypo (III) − Normal CyA→FK506 No No No Dead
2 59/F RA (RCMD) 1.0 Hypo (III) − 46,XX,+1,der(1;7) ATG+CyA No No No Alive
(q10,p10)[19]
3 71/M RA (RA) 0.5 Hypo (III) + 45,X,-Y[9/20] / CyA No No No AMLa
47,XY,+8[3/20]
4 63/F RA (RA) 0.5 Hypo (III) + Normal CyA→FK506 Major No Major Alive +
5 58/M RA (RCMD) 0.5 Hypo (III) + Normal ATG+CyA→ Major Major Major Dead +
FK506
46,XX,add(3)(p21)
[2/40]/
6 64/F RA (RA) 1.0 Hypo (IV) − 46,XX,add(11)(q13) CyA→FK506 Minor Major Minor Dead +
[1/40]
47,XY,+8[13]/45,X,
−Y[2]/46,XY
7 60/M RA (RCMD) 0.5 Hypo (III) − [5] CyA Major Major No Alive +
8 68/M RA (RA) 0.5 Hypo (III) − Normal CyA Major Major Major Alive +
9 59/M RA (RA) 1.0 Hypo (III) + 46,XY;t(12,15) ALG+CyA Major Major Major Alive +
(q23,q12)[3]
10 69/M RA (RA) 0.5 Normo (III) + Normal CyA→ATG→ Minor No No Alive +
FK506
(Continued)
Table 1 Patient Profile: Myelodysplastic Syndromes (MDS) (Continued)
Diagnosis
FAB IPSS HLA- IST
No. Age/sex (WHO) score Histology MRI DR15 Karyotype Regimen of IST Hb PMN Plt Outcome depend
11 67/F RA (RCMD) 0.5 Normo ND ND Normal CyA No No No Alive
12 67/F RA (RA) 0.5 Normo (IV) − Normal CyA No No No Alive
13 47/F RA (RCMD) 0.5 Normo (IV) − Normal CyA No No No Alive
14 32/M RA (RCMD) 0.5 Normo ND − Normal CyA No No No Alive
15 40/M RA (RA) 0.5 Normo (III) ND 45,X,−Y[3] CyA Major Major Major Alive +
16 58/M RA (RA) 0.5 Normo (III) − Normal CyA Minor No No Alive +
17 63/F RA (RA) 0.5 Normo (III) − Normal CyA No No No Alive
18 41/M RA (RCMD) 0.5 Hyper (III) − Normal CyA No No No Dead
19 66/M RA (RCMD) 0.5 Hyper ND − Normal CyA No No No Dead
20 52/M RA (RCMD) 1.5 Hyper (III) − 47,XY,+1, ATG+CyA No No No Dead
der(1;7)(q10;p10),
+8 [9/20]
21 31/M RA (RCMD) 0.5 Hyper (IV) + Normal CyA Major Major Major Alive +
22 48/F RA (RCMD) 0.5 Hyper (III) − Normal CyA Major Minor Major Alive +
23 50/F RA (RA) 0.5 Hyper (IV) + Normal CyA Major Major Major Alive +
aDeath.
Abbreviations: FAB, 1982 French-American-British MDS classification; WHO, 2001 World Health Organization classification of hematopoietic and lymphoid
malignancies; RA, refractory anemia; RCMD, refractory cytopenias with multilineage dysplasia; hypo, hypocellular; hyper, hypercellular; ATG, anti-thymocyte
globulin; ALG, anti-lymphocyte globulin; CyA, cyclosporine A; BM, bone marrow; MRI, magnetic resonance image; HLA, human leukocyte antigens; IST,
immunosuppressive therapy; Hb, hemoglobin; PMN, polymorphonuclear leukocytes; Plt, platelets.
Table 2 Patient Profile: Aplastic Anemia (AA)
HLA- IST
No. Age/sex Diagnosis Histology MRI DR15 Karyotype Regimen of IST Hb PMN Plt Outcome depend
1 65/F AA Hypo (I) ND ND ALG, CyA No No No Dead
2 54/F AA Hypo ND ND ND CyA, ALG No No No Alive
3 60/F AA Hypo (I) − ND ALG, CyA No No No Dead
4 16/M AA Hypo ND − Normal CyA, ATG Major Major Major MDS −
5 72/F AA Hypo (I) + Normal CyA, ATG No No No Dead
Hypocellular Myelodysplastic Syndrome
6 73/F AA Hypo (I) − Normal CyA Major Major Major Alive +

7 62/F AA Hypo (I) + Normal CyA, ATG Major Major Major Alive +
8 75/F AA Hypo (I) + Normal CyA, ATG,FK506 No Major No Alive −
9 25/M AA Hypo (I) − Normal CyA, ATG Major Major Major Alive −
10 16/F AA Hypo ND − Normal CyA, ALG Major Major Major PNH† +
11 69/F AA Hypo ND + Normal CyA, ATG Major Major Major Alive −
12 58/M AA Hypo (I) − Normal ALG, CyA→ATG No No No Dead
15 48/F AA Hypo (I) − Normal CyA, ATG Major Major Major Alive +
(Continued)
231
232
Table 2 Patient Profile: Aplastic Anemia (AA) (Continued)

HLA- IST
No. Age/sex Diagnosis Histology MRI DR15 Karyotype Regimen of IST Hb PMN Plt Outcome depend
16 42/M AA Hypo (I) − Normal ALG, CyA→CyA Major Major Major MDS −
17 61/F AA Hypo ND ND ND ALG, CyA Major Major Major Alive −
18 17/F AA Hypo (I) − Normal ALG, CyA Major Major Major Alive −
19 25/M AA Hypo (I) − Normal CyA, ATG Major Major Major Alive +
20 65/F AA Hypo (I) + Normal ALG, CyA Major Major Major Dead +
21 58/F AA Hypo (I) + Normal CyA, ATG Major Major Major Alive +
22 61/F AA Hypo ND ND ND ATG, PSL →CyA No No No Dead
23 64/M AA Hypo (I) − Normal ALG, CyA Major Major Major Alive +
24 54/F AA Hypo (I) + Normal CyA Major Major Major Alive +
25 64/F AA Hypo (I) − Normal ALG, CyA Major Major Major Alive
26 53/F AA Hypo ND ND ND ALG, CyA Major Major Major Dead −
27 59/F AA Hypo (I) + Normal ALG, CyA Major Major Major Alive −
28 54/M AA Hypo (III) + 47,XY, +8 ALG, CyA Major Major Major Alive −
29 66/F AA Hypo (III) ND Normal ALG, CyA→FK506 Major Major Major MDS +
30 70/M AA Hypo (I) ND 46,XY ALG, CyA Major Major Major PNH −
31 59/M AA Hypo ND + 46,XY ALG, CyA Major Major Major PNH −
Abbreviations: BM, bone marrow; MRI, magnetic resonance image; HLA, human leukocyte antigens; IST, imuunosuppressive therapy; Hb, hemoglobin; PMN,
polymorphonuclear leukocytes; Plt, platelets.
Hata and Tomonaga
Hypocellular Myelodysplastic Syndrome 233
information and confirmation of biopsy findings. Experienced hematologists can

also estimate cellularity by observing tissue particles on an edge smear or a
squash smear of bone marrow aspirates, but aspirate smears are less reliable when
marrow aspiration is difficult, as may be the case in MDS if there is accompanying
myelofibrosis.
In healthy persons, bone marrow cellularity steadily declines with aging,
especially after the age 70 (17). Taking into consideration this age effect, it is usu-
ally recommended that a marrow cellularity less than 30% is defined as hypocel-
lular for patients less than 70-year old, and less than 20% for patients older than
70 years.
During a review of many histological samples from patients with clonal
myeloid disorders, Nukhet Tuzuner from Turkey and his colleagues found that,
while a certain proportions of MDS and AML cases have hypocellular bone
marrow, this is never true in chronic myeloproliferative disorders (CMPD) (19).
Interestingly, those investigators observed that the AML cases with hypocellu-
lar bone marrow were mostly elderly patients, whereas hypocellular MDS were
observed in all age groups.
Most of the histological diagnoses in our own series were made on core
biopsies at iliac crest, supplemented by clot sections of bone marrow aspirates.
Bone marrow histology of AA patients is not infrequently inhomogeneous in
distribution of hematopoietic tissue in a single-core biopsy specimen, but can be
determined as hypocellular in total when cellular and acellular areas are summed
up. Likewise, in most hypocellular MDS cases, histology show inhomogeneous
distribution of hematopoiesis, and some cases require bilateral sampling in order
to be confident about the diagnosis.
Magnetic Resonance Imaging

Magnetic resonance imaging is the most powerful radiological method for visualiz-
ing the systemic distribution of hematopoietic tissues in flat bones especially verte-
brae. Short TI inversion recovery (STIR) imaging is particularly useful. Kusumoto
and colleagues have proposed a system for classifying patients with MDS and AA
into four groups, based on STIR images of vertebral bone marrow (Fig. 1) (20).
Type I [Fig. 1(A)] is characterized by a diffuse, low-signal intensity, and
is mostly representative of hypocellular bone marrow, as seen in typical cases of
AA and in some cases of MDS. Type II [Fig. 1(B)] represents a normal bone
marrow cellularity and is characterized by marginally high-signal intensity with
a small lower intensity area in the center of a vertebra. Type III [Fig. 1(C)] has
an inhomogeneous signal intensity that results from a mixture of cellular and
hypocellular (fatty) tissues, and is most frequently seen in MDS subtype RA,
and less commonly observed in AA. Type IV [Fig. 1(D)] is characterized by a
diffuse, high-signal intensity, and is most compatible with a diagnosis of hypercel-
lular acute leukemia or CMPD, including proliferative chronic myelomonocytic
Figure 1 Four MRI–STIR signal intensity patterns classified according to Kusumoto

et al. (20). (A) Type I: diffuse low signal, typically indicating hypocellular bone marrow.
(B) Type II: marginally high intensity with a small lower intensity in the center of a vertebrae,
representative of normocellular bone marrow. (C) Type III: inhomogeneous signal intensity
suggestive of uneven distribution of cellular bone marrow, seen in a variety of settings.
(D) Type IV: diffuse high-signal intensity, representative of hypercellular bone marrow.
leukemia. In our series, Type III MRI findings were also frequently observed in
MDS-RA (Table 1) and less frequently in AA (Table 2).
Comparison of Histology with MRI

We compared biopsy-proven histological cellularity with MRI–STIR typing. In
AA (Table 2), 23 patients underwent MRI–STIR imaging—21 (91%) cases were
Type I and 2 (9%) cases were Type III. In MDS (Table 1), 20 cases had MRI–STIR
performed; Type III results were the most common (15/20, 75%) and were seen
in all histological patterns. Although 4 of the 5 Type IV cases were hypercellular
or normocellular, 1 was hypocellular.
Unfortunately, there is still no good computer-based summing technology
for Type III inhomogeneous pattern to quantitatively evaluate hematopoietic tis-
sue. Clinicians should take into account the possibility of discrepancy between
histology and MRI images. In conclusion, from a clinical view point, core biopsy
at the iliac crest should be routinely performed and can be combined with cellu-
larity evaluation based on observation of particles on an edge smear or a squash
smear of aspirates in a given case. MRI imaging can add additional information,
but should be considered a secondary test, especially given the uncertain clini-
cal importance of cellularity in MDS (see section, “Clinical and Hematological
Feature of Hypo-MDS”).
HYPOCELLULAR MYELODYSPLASTIC SYNDOMES (Hypo-MDS):

A DISTINCT CLINICAL ENTITY?
Clinical and Hematological Feature of Hypo-MDS
The significance of marrow cellularity in MDS is unclear. Tuzuner et al. reviewed
a series of 100 cases of MDS with varying degrees of cellularity (28 hypocel-
lular and 72 normocellular/hypercellular) to see if any clinical differences were
present (19). However, they could not find any differences in the degree of cytope-
nias for the three lineages, cytogenetic results, overall survival, or the rate of
leukemic transformation. These findings suggested that histological discrimina-
tion of cellularity does not necessarily imply that there is a basic difference in the
pathophysiology of MDS among different cellularity groups. Likewise, our small
series shown in Table 1 also suggested that there is no remarkable difference in
clinical features and hematological features.
However, more recent reports based on a larger series have suggested some
differences in clinical and hematological features related to marrow cellularity.
Gang Yue and colleagues at the University of Massachusetts identified 163 (15.5%)
hypo-MDS cases out of 1049 consecutive MDS patients, based on the histological
criteria mentioned above (8). The frequency of hypo-MDS in each WHO sub-
type was as follows: RA (21.5%), RARS (2.5%), RCMD (16.6%), RCMD-RS
(2.5%), RAEB-1 (15.3%), and RAEB-2 (9.8%). Compared to normo- or hyper-
cellular MDS patients, hypo-MDS patients were younger, less anemic, but more
neutropenic and thrombocytopenic; in contrast, cytogenetic results and the distri-
bution of International Prognostic Scoring System (IPSS) risk categories (see chap.
1) were similar. The investigators also detected a small PNH clone in 10% of hypo-
MDS patients and a similar proportion of normo/hypercellular MDS—much less
commonly than is observed in AA patients. Most importantly, hypo-MDS groups
showed a favorable overall survival compared to cellular group (56 months vs.
28 months). This survival preference for the hypo-MDS group was demonstrated
in all IPSS risk groups and cytogenetic risk groups. Data on treatment response
[including immunosuppressive therapy (IST)] were not reported.
This study is the largest series of hypo-MDS to date and strongly suggests
that hypo-MDS does indeed have special features. Their findings need to be
confirmed by others, and more information on the response to specific drugs for
MDS is also necessary.
IST for MDS Patients

Several pilot studies have assessed the role of IST in MDS [e.g., (21–24)]; these
studies are reviewed in more detail in chapter 19. Jeff Molldrem and colleagues
reported a preferential response in hypocellular MDS in their initial report of ATG
therapy in 1998 (21). More recently, Lim and colleagues reported a retrospec-
tive multicenter analysis of IST for 96 patients with MDS. Hypo-MDS patients,
defined as less than 20% cellularity, comprised 55% of all patients in this ret-
rospective series. This high value is likely a bias caused by higher recruitment
rate of hypo-MDS patients to IST by doctors at each institute (9). A total of
40 (42%) patients achieved a hematological response, of which 30 (75%) had a
durable response lasting a median duration of 31.5 months. Multivariate analy-
sis suggested that low IPSS score and hypocellularity were independent factors
predicting a hematological response to ATG.
As shown in Figure 2, our retrospective analysis to compare response rate
and pattern between MDS and AA patients at a single institution suggested a sta-
tistically significant higher response rate in AA patients (80% trilineage response)
than in MDS (50%). There was a tendency for hypo-MDS patients to respond
better (around 60% response rate) that did not achieve statistical significance,
possibly due to low numbers.
100 Erythroid Neutrophil Platelets
p = 0.05 p < 0.01 p < 0.01 major

minor
80
* * : not significant
* *
60
(%)
40
20
0
All Hypocellular All Hypocellular All Hypocellular
AA AA AA
MDS MDS MDS
Figure 2 Hematologic response of patients with myelodysplastic syndomes (MDS) to

immunosuppressive therapy in comparison to patients with aplastic anemia (AA). Com-
pared to AA, response rate in trilineage was significantly lower in MDS. Although statis-
tically not significant, hypocellular MDS showed higher response rates compared to other
MDS.
1.0 1.0
Cumulative survival
Cumulative survival
0.8 0.8
AA Responder
0.6 0.6 (IST-dependent : 12/12)
MDS
0.4 0.4
Nonresponder
0.2 0.2
p < 0.001
0 0
0 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 0 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15
Years Years
(A) (B)
1.0
Cumulative survival
0.8
Responder
0.6 (IST-dependent : 10/24)
0.4
0.2 Nonresponder
p < 0.001
0
0 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15
(C) Years
Figure 3 Overall survival of patients with myelodysplastic syndromes (MDS) treated with
immunosuppressive therapy (IST). (A) Overall survival of patients with MDS treated with
IST compared to patients with AA treated with IST. There was no statistically significant
difference between outcomes in the two diseases. (B) Overall survival of patients with MDS
compared between IST responders and nonresponders. There was a statistically significant
difference. Among 12 MDS IST responders, all the 12 cases remained immunosuppressive
therapy (IST)-dependent. There were five cases with hypocellular MDS among responders.
Four of them showed almost normal Hb level after treatment, but erythroid hyperplasia
and mild trilineage dysplasia persisted. (C) Overall survival of patients with AA, with
comparison between responders and nonresponders. As for MDS, there was a statistically
significant difference. Among 24 responders, 10 cases were IST-dependent and 14 patients
eventually became IST-independent, suggesting a long-term complete remission potentially
compatible with cure of AA.
Recent analysis of our series based on long-term follow-up (median follow-

up of 8 years) revealed almost similar overall survival between MDS and AA,
as shown in Figure 3(A). Unlike the findings of Yue and colleagues, overall sur-
vival was not different between hypo-MDS and other MDS in our series (data not
shown). There were remarkable differences in survival between IST responders
and nonresponders in both diseases as shown in Figures 3(B) and 3(C), respec-
tively. All 12 responders remained IST (CyA) dependent, whereas only 10 out of
24 responders of AA were IST dependent. Thus, it is more common to achieve
an IST-independent long-durable complete hematologic remission in AA com-
pared to MDS, suggesting a basic difference in response-maintaining mechanism
between the two stem cell disorders. All literature suggests that a substantial num-
ber of AA patients who responded to IST became IST-independent and might
have been cured of disease (24,25). In contrast, there has been no report on
long-term outcome of IST for MDS. Our observation suggests that IST may result
in hematological improvements in some MDS cases but is not curative.
When IST is chosen for MDS, the optimal regimen is unknown, as there
has been no study to compare efficacy of ATG and CSA when used alone or in
combination for these diseases. Combinations are frequently employed in AA, but
MDS series have been more variable. This is an important research priority. IST
could also be combined with other agents, including immunomodulatory drugs
like lenalidomide (see chap. 20) or hypomethylating agents like decitabine and
azacitidine (see chap. 21).
Mechanism of Hematologic Remission by Immunosuppressive Therapy

The mechanism of IST in MDS is unclear. Recently, Elaine Sloand and her col-
leagues at the National Institutes of Health in the United States reported a series
of immunological studies on MDS patients, focusing especially on patients with
distinct chromosomal abnormalities such as trisomy 8, monosomy 7 and 5q− (26).
They observed that the T-cell receptor V␤profile study suggested that there were
clonally expanded T-cell populations prior to IST in patients with trisomy 8 but not
in those with monosomy 7 or 5q−. Those patients with trisomy 8 preferentially
responded to IST (about 65%). After responding to IST, the clonal expansion of
T-cell declined and the V␤profiles returned to normal polyclonal patterns, suggest-
ing an important role of one or more cytotoxic T-cell clones in the development
of bone marrow failure in MDS, especially RA with trisomy 8 (27). This finding
is similar to the expansion of clonal cytotoxic T-cells in patients with AA prior to
IST, and the return to a polyclonal state with response to IST (16). Whether these
T-cell clones are attacking distinct antigen(s) on hematopoietic stem cells is still
unclear in both disorders.
Sloand et al. also observed a paradox in their series of trisomy 8 patients:
CD34 cells with trisomy 8 expressed more Fas antigen than CD34 cells from
patients with monosomy 7 or from normal controls, yet were partially resistant to
Fas-mediated apoptosis, probably due to increased expression of c-myc, cyclin D1,
and suvivin (28). By using siRNA-mediated downregulation of suvivin in trisomy
8 cells, they observed increased apoptosis of trisomy 8 cells (29). They speculated
that trisomy 8 cells, despite expressing increased amounts of Fas antigen, acquire
partial resistance to Fas-mediated apoptosis due to trisomy 8–specific upregulation
of anti-apoptotic mechanism involving survivin pathway.
The sophisticated experiments by Sloand et al. on CD34 cells from MDS
patients with distinct chromosome abnormalities may explain the general obser-
vation that MDS patients with trisomy 8 are highly responsive to IST and acquire
hematologic remission; however, paradoxically they observed that the trisomy 8
clones persisted in hematologic remission and even showed an expansion. They
speculated that the trisomy 8 cells (stem cells) may have growth advantage over
normal stem cells.
Whether this kind of abnormality suggests malignant nature of trisomy
8 cells is an important question because this chromosome abnormality is not
infrequently observed, although low percent, in AA patients either remaining in

AA status or progressed to MDS.
In terms of pathophysiology of hypo-MDS, mechanism of bone marrow
hypocellularity consisting of trisomy 8 clone was not mentioned by Sloand et al.
They suggested that T-cell clone(s) against trisomy 8 cells may injure normal stem
cells as bystanders in a milieu of immunological effector molecule(s), and trisomy
8 clones expands by their growth advantage over normal hematopoiesis. It is well
studied that distribution of chromosome abnormalties in hypo-MDS patients is
not particularly different from normo- or hypercellular MDS (5,8). Therefore, it is
still unknown whether hypocellulatrity is the result of immunological suppression
to both MDS clone and normal stem cells.
CLINICAL OVERLAP BETWEEN AA AND MDS

Transition from AA to MDS
Among severe AA patients who were treated with IST (usually combined with
ATG and CSA), 10% to 15% patients eventually progressed to MDS (6,25).
Typically, AA cases that progess to MDS often carry monosomy 7 or trisomy
8. In the NIH series, patients with monosomy 7 tended to show worse prognosis
including higher rates of leukemic change, while patients with trisomy 8 had a more
benign prognosis (30). In contrast, very few AA patients treated with allogeneic
stem cell transplant showed such a clonal progression (25), suggesting a complete
eradication of premalignant stem cells existing as a cryptic population in AA
with stem cell transplant. Notably, childhood cases of severe AA that developed
after acute hepatitis (probably due to viral infection) have also developed MDS
(31), suggesting that either stem cell injury in AA patients involves not only
autoimmune injury but also virus-related injury to hematopoietic stem cells, or
the autoimmune injury may promote emergency of a latent clonal disorder, such
as PNH, MDS, and AML. Another possibility is that the same patients who are
predisposed to develop AA after viral infections have “fragile” stem cells and also
have a predisposition to develop hematopoietic neoplasia.
Cryptic MDS in Severe AA as an Extreme Case of Hypo-MDS

Sergej Konoplev and his colleagues at the M.D. Anderson Cancer Center in
Houston, Texas, recently performed a retrospective analysis of 128 consecutive
patients with an initial diagnosis of AA seen at that center from 1993 to 2004
(32). A total of 12 (10%) patients diagnosed with AA eventually developed MDS.
Mean initial bone marrow cellularity among these 12 patients was 5%, and all
patients had a normal karyotype. Median time to develop MDS after the diagnosis
of AA was 9 months, and the median cellularity at the time of MDS diagnosis
was 30%. At the time of MDS evolution, 9 out of 12 (75%) patients showed
appearance of chromosome abnormalities, including monosomy 7 in 5 patients.
Fluorescent in situ hybridization (FISH) also detected monosomy 7 in six samples
at the time MDS was diagnosed, while a retrospective study of the 12 patients’
marrows at the time of AA diagnosis with FISH revealed monosomy 7 in two

cases. This observation suggests that among patients with a clinical diagnosis of
AA, there exist a few cryptic cases of hypo-MDS (⬍5%) that may not be detected
with routine karyotyping. Therefore, careful follow-up studies after a diagnosis
of AA are warranted to detect emergence of dysplastic clones with chromosome
abnormalities.
Comparison of Long-Term Remission After IST Between AA and MDS

As in our series, the majority of AA patients who respond to IST eventually are
able to stop taking their immunosuppressive agents, and many will maintain a
long-term remission. In our previous study on colony formation by bone marrow
cells obtained from AA patients in long-term remission, we observed almost
complete recovery to normal levels in both CFU-GM and BFU-E progenitors
(33). In contrast, MDS patients tended to show incomplete recovery in progenitor
levels after a response to IST (data not shown).
In five hypo-MDS cases surviving more than 10 years under maintenance
of IST, we have performed morphologic and cytogenetic analyses. Cytogenetics
showed normal karyotypes in all cases and morphological evaluation of dysplasia
revealed persistence of trilineage dyplasia in all cases, although the dyplasia was
mild in all. Moreover, the bone marrows were normo- or hypercellular in four cases
with erythroid hyperplasia as shown by low M/E ratio, suggesting a possibility of
an increased or compensated erythropoiesis to keep the Hb level in normal range.
IST seems to reduce the degree of ineffective hematopoiesis by MDS clones
through as yet unknown mechanism.
SINGLE NUCLEOTIDE POLYMORPHISM (SNP) ARRAYS:

HIGH-RESOLUTION CYTOGENETICS FOR MDS BIOLOGY
Detection of Chromosomal Microlesions and Uniparental Disomy (UPD)
by SNP Arrays
As described in chapter 4, several groups (at Cleveland Clinic, in London and in
Japan) have used high-resolution SNP arrays to study MDS (34–36) and these
tools have also now been employed in series of AA patients (37). In MDS, pre-
viously unrecognized lesions can be found in those who had a normal karyotype
by conventional G-banded metaphase cytogenetics. Most (about 75%) of cytoge-
netically occult microlesions detected by SNP arrays have been micro-deletions
or micro-gains of chromosomal material, while about 20% of microlesions were a
UPD (20%). UPD is a copy-number neutral abnormality with loss of heterozygos-
ity (LOH). Whereas metaphase cytogenetics revealed 60% abnormal karyotypes
in this series, SNP array cytogenetics provided abnormal results up to 80% of
patients. Unfortunately, trisomy 8 patients were not studied.
These cryptic lesions at chromosome level detected by the high-resolution
SNP array seemed to have prognostic significance. Patients with microlesions and
normal karyotype showed a poorer prognosis compared to those patients with

normal SNP array patterns as well as a normal karyotype (35).
The Cleveland Clinic Group has also used SNP arrays to identify cryptic
microlesions and UPD in patients with AA, PNH, and hypo-MDS (37). They
found a relatively high frequency of these cryptic lesions, not only in hypo-MDS
patients but also in AA and PNH patients. These observers found that the response
rate to IST was lower (57%) in patients with AA, who found to have such cryptic
lesions, than in those with normal SNP array pattern (72%).
Although SNP arrays are a high-resoluton cytogenetic technique, there is
a limitation, as Gondek and colleagues pointed out (34). Detection sensitivity
of such cryptic lesions in blood cells depends on the proportion of cells in the
hematopoietic tissue examined that are part of a clonal population. In order for
SNP arrays to detect karyotypically occult lesions with any confidence, at least
50% of bone marrow cells should be clonal.
Benign Clonal Hematopoiesis Versus Malignant Clonal Hematopoiesis

Based on the results obtained by using SNP array, Jaroslaw Maciejewski’s group
at the Cleveland Clinic group recently proposed a schema in which most cases
of AA are polyclonal with a normal pattern in SNP arrays, but in some patients
one or a few stem cells tend to restore hematopoiesis, mimicking a clonal disease
(38). In contrast, MDS has already a malignant clone with chromosome instability
detectable by SNP array at apparently higher rates than by metaphase cytogenetics.
In order to move closer to clinical utility of SNP arrays, further prospective clinical
studies are needed to detect SNP array–based microlesions and UPD for patients
with AA and hypo-MDS prior to therapy and after response to therapy.
Nevertheless, it is important to recognize that SNP array may not be sensitive
enough to detect a small population of clonal hematopoiesis in severe hypocellular
bone marrow at initial diagnosis of AA. Therefore, the most suitable samples for
SNP array may be the remission bone marrow with normo- or hypercellularity
obtained in patients with hypo-MDS (as above-described long-term survivors),
other MDS or AA after IST. If microlesions and UPD were found in some AA
patients in clinical remission, it might provide evidence that clonal hematopoiesis
is emerging with chromosome instability. Thus, existing diagnostic criteria for
AA and MDS may be challenged by this new technology.
HYPOCELLULAR ACUTE MYELOID LEUKEMIA (Hypo-AML)

Clinical and Hematologic Feature of Hypo-AML
As mentioned above hypo-AML comprises about 15% of adult AML (10,15).
In one series, the age of these patients ranged from 45 to 82 with a median age
of 65 years. Peripheral blood examination in hypo-AML reveals almost always
severe pancytopenia with relatively few circulating blast cells. In a hypocellular
bone marrow, it is difficult to count blast cells correctly, and a diagnosis of AML
may be difficult. In our series of 32 cases, the blast percent ranged from 20%
to 70% (mean 34%) of total mononuclear cells (15). By excluding lymphocytes,
which are relatively increased due to hypocellularity, the blast percentage ranged
from 38% to 94% (mean 58%). Exclusion of erythroblasts further increased the
blast percent up to a mean of 78%. Therefore, in hypo-AML, there is a definite
maturation arrest in myeloid differentiation that is quite similar to overt AML with
cellular bone marrow and distinct from RAEB in which myeloid differentiation is
partially maintained.
Hypo-AML can be differentiated from RAEB when blast cells were counted
by excluding lymphocytes and erythroblasts. RAEB cases manifest not infre-
quently hypocellular bone marrow and confused with hypo-AML. Complicating
matters, trilineage dysplasia is also seen in some cases with hypo-AML. In sum-
mary, hypo-AML can be diagnosed for cases with bone marrow cellularity less
than 30% in patients with age less than 70 years and less than 20% in those with
age 70 years or more, and blast percentage at more than 20% (based on WHO
classification for AML) of the bone marrow mononuclear cells when lymphocytes
and erythrocytes are excluded. In addition, the blast cells of hypo-AML are usually
poorly differentiated like overt AML-M0 or -M1. When patients with AA or MDS
progress to AML, the type of AML is usually not hypo-AML but overt AML
with cellular bone marrow, suggesting that there might actually be no common
pathophysiology for a hypocellular bone marrow.
Good Response to Chemotherapy But High Relapse Rate

Conventional chemotherapy or low-dose cytarabine therapy can provide around
60% complete remission rate for hypo-AML patients (14). In patients with a
karyotypic abnormality, the abnormal clone usually disappears after getting into
remission (39). This means that hypocellularity in this AML subtype is not related
to immunological suppression as seen in AA and MDS, but suppression of normal
hematopoiesis by hypo-AML clone, most probably due to slow growing of blast
population. However, relapse is inevitable in almost all cases, indicating that hypo-
AML is a form of elderly AML with higher resistance to conventional chemother-
apy, as is also often seen in patients with AML-M0 (15). Only allogeneic stem
cell transplantation may be a curable therapy for this subtype of AML, but clinical
data on reduced intensity transplantation in hypo-AML are not yet available.
In conclusion, hypo-AML seems to have a different pathophysiology in
inducing hypocellular bone marrow that is distinct from AA and hypo-MDS.
Whether there is overlap, and hematopoietic failure in hypo-MDS might some-
times have a similar mechanism, is unclear.
SUMMARY
The distinction between AA and MDS based on histology and morphology alone
is difficult. New technologies such as SNP arrays may become an important
mechanism to overcome ambiguity. Concepts of hematopoiesis in AA versus

MDS are changing based on the new findings obtained by SNP arrays. Emergence
of a single clone in a setting of impaired hematopoiesis was previously proposed
to explain the expansion of PNH clone (16). Whether a group of patients with
a clinical diagnosis of AA or hypo-MDS based on bone marrow hypocellularity
and presence or absence of morphological dysplasia will be clearly reclassified by
high-resolution SNP array, is an interesting subject to investigate prospectively.
If cryptic MDS cases can be diagnosed among AA patients at initial pre-
sentation or in remission after IST, the boundary between AA and MDS becomes
clearer, and patient management will improve. In the era of new drug develop-
ment such as immunomodulatory drugs and hypo-methylating agents for patients
with MDS, it is extremely important to establish a method to make a distinc-
tion between real AA with either nonclonal or clonal (due to decreased stem cell
number) benign disease with autoimmune nature and real hypo-MDS as a clonal
premalignant disease with chromosome instability.
In conclusion, hypo-MDS remains an interesting stem cell disorder in
the borderlands between AA and MDS, and still suffers from considerable
ambiguity.
REFERENCES
2. Harris NL, Jaffe ES, Diebold J, et al. World Health Organization classification of
neoplastic diseases of the hematopoietic and lymphoid tissues. Report of the clinical
advisory committee meeting—Airlie House, Virginia, November 1997. J Clin Oncol
1999; 17:3835–3849.
3. Yoshida Y, Oguma S, Uchino H, et al. Refractory myelodysplastic anaemias with
hypocelluar bone marrow. J Clin Pathol 1988; 41:763–767.
4. Toyama K, Ohyashiki K, Yoshida Y, et al. Clinical and cytogenetic findings of
myelodysplastic syndromes showing hypocellular bone marrow or minimal dyspla-
sia, in comparison with typical myelodysplastic syndromes. Int J Hematol 1993;
58:53–61.
5. Tuzuner N, Cox C, Rowe JM, et al. Hypocelluler myelodysplastic syndromes(MDS):
6. Elghetany MT, Hudnall SD, Gardner FH. Peripheral blood picture in primary hypocel-
lular refractory anemia and idiopathic acquired aplastic anemia: An additional tool
for differential diagnosis (review). Haematologica 1997; 82:21–24.
7. Goyal R, Qawi H, Ali I, et al. Biologic characteristics of patients with hypocellular
myelodysplastic syndromes. Leuk Res 1999; 23:357–364.
8. Yue G, Hao S, Fadare O, et al. Hypocellularity in myelodysplastic syndrome is
independent factor which predicts a favorable outcome. Leukemia Res 2007; 32:553–
558.
9. Lim ZY, Killick S, Germing U, et al. Low IPSS score and bone marrow hypocellularity
in MDS patients predict hematological responses to antithymocyte globulin. Leukemia
2007; 21:1436–1441.
10. Beard MEJ, Bateman CJT, Croether DC, et al. Hypoplastic acute myelogenous
leukemia. Br J Haematol 1975; 31:167–176.
11. Howe RB, Bloomfield CD, McKenna RW. Hypocellular acute leukemia. Am J Med
1982; 72:391–395.
12. Needleman SW, Burns CP, Dick FR, et al. Hypoplastic acute leukemia. Cancer 1981;
48:1410–1414.
13. Gladson CL, Naeim F. Hypocellular bone marrow with increased blasts. Am J Hematol
1986; 21:15–22.
14. Nagai K, Kohno T, Chen YX, et al. Diagnostic criteria for hypocellular acute leukemia:
A clinical entity distinct from overt acute leukemia and myelodysplastic syndrome.
Leuk Res 1996; 20:563–574.
15. Tuzuner N, Cox C, Rowe JM, et al. Hypocellular acute myeloid leukemia: The
Rochester (New York) experience. Hematol Pathol 1995; 9:195–203.
16. Young NS. Acquied aplastic anemia. In: Young NS, Gerson SL, High KA, eds. Clinical
Hematology. Philadelphia: Mosby, 2006; 136–157.
17. Hartsock RJ, Smith EB, Petty CS. Normal variations with aging of the amount of
hematopoietic tissue in bone marrow from the anterior iliac crest. Am J Clin Pathol
1965; 43:326–331.
18. Tuzuner N, Bennett JM. Reference standards for bone marrow cellularity. Leuk Res
1994; 18:645–647.
19. Tuzuner N, Cox C, Rowe JM, et al. Bone marrow cellularity in myeloid stem cell
disorders: Impact of age correction. Leuk Res 1994; 18:559–564.
20. Kusumoto S, Jinnai I, Matsuda A, et al. Bone marrow patterns in patients with aplas-
tic anaemia and myelodysplastic syndrome: Observations with magnetic resonance
imaging. Eur J Haematol 1997; 59:155–161.
21. Molldrem JJ, Caples M, Mavroudis D, et al. Antithymocyte globulin for patients with
22. Molldrem JJ, Jiang YZ, Stetler-Stevenson M, et al. Haematological response of
patients with myelodysplastic syndrome to antithymocyte globulin is associated with a
loss of lymphocyte-mediated inhibition of CFU-GM and alterations in T-cell receptor
V␤ profiles. Br J Haematol 1998; 102:1314–1322.
23. Jonasova A, Neuwirtova R, Cermak J, et al. Cyclosporin A therapy in hypoplastic
MDS patients and certain refractory anaemias without hypoplastic bone marrow. Br
J Haematol 1998; 100:304–309.
Hematol 2000; 37:15–29.
25. Socie G, Henry- Amar M, Bacigalupo A, et al. Malignant tumors occurring after
treatment of aplastic anemia. European bone marrow transplantation—Severe aplastic
anaemia working party. N Eng J Med 1993; 329:1152–1157.
26. Sloand EM, Kim S, Fuhrer M, et al. Fas-mediated apoptosis is important in regulating
cell replication and death in trisomy 8 hematopoietic cells but not in cells with other
cytogenetic abnormalities. Blood 2002; 100:4427–4432.
27. Sloand EM, Mainwaring L, Fuhrer M, et al. Preferential suppression of tri-
somy 8 compared with normal hematopoietic cell growth by autologous lympho-
cytes in patients with trisomy 8 myelodysplastic syndrome. Blood 2005; 106:841–
851.
abnormalities. Blood 2004; 104:4210–4218.
29. Sloand EM, Pfannes L, Chen G, et al. CD34 cells from patients with trisomy 8
myelodysplastic syndrome (MDS) express early apoptotic markers but avoid pro-
grammed cell death by up-regulation of antiapoptotic proteins. Blood 2007; 109:2399–
2405.
30. Maciejewski JP, Risitano A, Sloand EM, et al. Distinct clinical outcomes for cytoge-
netic abnormalities evolving from aplastic anemia. Blood 2002; 99:3129–3135.
31. Ohara A, Kojima S, Hamajima N, et al. Myelodysplastic syndrome and acute myel-
ogenous leukemia as a late clonal complication in children with acquired aplastic
anemia. Blood 1997; 90:1009–1013.
32. Konoplev S, Medeiros L, Lennon P, et al. Therapy may unmask hypoplastic myelodys-
plastic syndrome that mimics aplastic Anemia. Cancer 2007; 110:1520–1526.
33. Matsuo Y, Iwanaga M, Mori H, et al. Recovery of hematopoietic progenitor cells in
patients with severe aplastic anemia who obtained good clinical response with a com-
bination therapy of immunosuppressive agents and recombinant human granulocyte
colony-stimulating factor. Int J Hematol 2000; 72:37–43.
plastic syndromes. Exp Hematol 2007; 35:1728–1738.
35. Gondek LP, Tiu R, O’keefe CL, et al. Chromosomal lesions and uniparental disomy
detected by SNP arrays in MDS, MDS/MPD, and MDS-derived AML. Blood 2008;
111:1534–1542.
36. Ogawa S, Nanya Y, Yamamoto G. Genome-wide copy number analysis on GeneChip
platform using copy number analyzer for affymetrix GeneChip 2.0 software. Methods
Mol Biol 2007; 396:185–206.
37. Wlodarski M, O’Keefe C, Gondek L, et al. High density SNP arrays reveal that
distinct clonal lesions including uniparental disomy can be detected in a proportion
of patients with aplastic anemia with normal metaphase cytogenetics. Blood 2006;
108:ASH abstract 12.
38. Tiu R, Gondek L, O’keefe C, et al. Clonality of the stem cell compartment during
evolution of myelodysplastic syndromes and other bone marrow failure syndromes.
Leukemia 2007; 21:1648–1657.
39. Tagawa M, Shibata J, Tomonaga M, et al. Low-dose cytosine arabinoside regimen
induced a complete remission with normal karyotypes in a case with hypoplastic acute
myeloid leukaemia with No. 8-trisomy: In vivo evidence for normal haematopoietic
recovery. Br J Haematol 1985; 60:449–455.
Down Up
RA RARS a
Figure 4.1 See page 91.

Down Up
Del(5q) Normal karyotype Other karyotype
B
Untreated Lenalidomide
Figure 4.3B See page 95.

A
Figure 4.4A See page 97.
1
12S CyB
ND6
16S
FAB-type ND5
RA
ND1
RARS
RAEB
RAEB-t
AML
ND2 CMML ND4
ND4L
ND3
COX1
COX3
COX2 AP6

(A) (B)
(C)
(D)
(E)

(A) (B)
(A) (B)
(A) (B)

(A) (B)
(C) (D)

(A) (B)
(C) (D)
(E) (F)

(A) (B)

11
Diagnostic and Prognostic Utility of Flow
Cytometry in MDS
Denise A. Wells and Michael R. Loken

Hematologics, Inc., Seattle, Washington, U.S.A.
INTRODUCTION
Diagnosis and classification of myelodysplastic syndromes (MDS) have tradition-
ally been based on histopathologic and clinical features consisting of cytopenias,
dysplastic morphologic features, abnormal marrow cellularity, an increased pro-
portion of blast cells, and clonal karyotypic abnormalities (see chaps. 1 and 9). The
combination of obvious marrow dysplasia and clonal karyotypic abnormalities is
considered diagnostic for MDS, with each technology confirming the other.
However, not all patients with MDS will have this combination of findings.
Because MDS-associated karyotypic abnormalities are only present in a subset
of cases of suspected MDS (see chap. 3); the diagnosis must often be based
exclusively on morphological criteria. In addition, dysplastic changes that mimic
MDS can be seen in nonclonal disorders as well, such as vitamin deficiencies,
heavy metal exposures, postchemotherapy or post–stem-cell transplant, and in
viral infections, alcoholism, and drug effects. While the 2001 WHO classification
of MDS (1) included karyotyping in addition to morphologic dysplastic features,
immunophenotyping was considered less relevant.
Flow cytometric evaluation of marrow from patients with MDS has been
studied for more than 15 years, yet the technology has not been yet incorporated
into major classification or prognostic systems. A thorough immunophenotypic
analysis has been performed on normal marrow, leading to an excellent understand-
ing of the antigenic patterns associated with the maturational stages of developing
myeloid cells, monocytes, myeloblasts, and erythroid cells (2–9). Flow cytometry
247
248 Wells and Loken
patterns of cells from MDS patients may differ from normal in multiple aspects
(10–17), and observing such abnormalities can supply crucial additional informa-
tion for accurate diagnosis and prognosis of these diseases. Moreover, the aberrant
immunophenotypes observed in MDS suggest disrupted coordination of gene reg-
ulation, since cell surface antigens are direct or indirect gene products. These gene
product abnormalities accumulate over time in neoplastic cells in MDS, often
culminating in the catastrophic event of post-MDS acute leukemia.
Historical Perspectives
Early Flow Cytometric Analysis of Antigen Expression
in Hematopoietic Cells
The seminal manuscripts detailing multidimensional flow cytometric analysis of
antigenic combinations on normal marrow were published in the late 1980s and
early 1990s. These studies identified the different cell compartments present in the
marrow and analyzed the cell maturational distribution, so that each maturational
stage was characterized by its expression of various cell-surface antigens (2–
9). A crucial finding of these early studies was that changes of expression of
multiple antigens often occurred together, delineating steps or stages of normal
hematopoietic development. These observations led to the hypothesis that precise
coregulation of gene product expression is required for normal development of
hematopoietic cells.
Cells at various immunophenotypically defined stages of maturation were
additionally analyzed by fluorescence-activated cell sorting and then morpholog-
ically evaluated (9). These sorted Wright–Giemsa stained cell populations were
found to highly correspond to accepted morphologic definitions of marrow inhab-
itants. In addition, abnormalities in myeloblasts were found to be sufficient to
identify small residual populations of AML postchemotherapy (18,19). The neo-
plastic myeloblasts were not only different from their normal counterparts, but
they were also different from each other, with each leukemic phenotype display-
ing a unique pattern of gene products. It soon became clear that abnormalities in
the expression of antigens could also be observed in maturing myeloid populations
in MDS, as described in greater detail below.
DNA Content Measurement by Flow Cytometry

In other early studies using flow cytometry, investigators were able to measured
cellular DNA content (20,21), and these techniques were later applied to MDS
marrows as a complement to the study of multiple antigen changes. The prolifer-
ation rate of hematopoietic cell populations can be estimated by the measurement
of the proportion of cells that are in S-phase or S-G2 M-phases of the cell cycle. In
studying cell-cycle activity, a higher percentage of cells in S-phase can be found
in MDS marrows as compared to normal (22,23), with decreased proliferative
activity associated with disease progression.
Diagnostic and Prognostic Utility of Flow Cytometry in MDS 249
While these early flow cytometric studies were useful in determining the
proportion of cycling cells in whole marrow, they did not adequately dissect
cell-cycle dynamics in specific neoplastic cell populations such as in maturing
myeloid cells. However, newer technologies using 5-bromodeoxyuridine or far
red–fluorescing DNA probes such as DRAQ5 (24), in combination with mon-
oclonal antibodies, promise to further enhance understanding of the biology of
simultaneous proliferation and apoptosis in MDS (25,26).
Single Antigen Studies in MDS
The early works defining antigenic maturational patterns in normal marrow were
not followed up in MDS for a number of years. The first studies of flow cytometric
analysis in MDS that did emerge were based on measurement of single disease-
associated immunophenotypic abnormalities. These studies described changes in
MDS marrows based on the proportion of cells expressing a given antigenic marker
(“percent positive”) or on the intensity of fluorescence (e.g., dim vs. bright), and
were reviewed by Mohamed Elghetany in 1998 (27). A more recent study by Luca
Malcovati and his colleagues in Italy (28) used a percent positive approach, and
found that while no single antigenic abnormalities could distinguish MDS from
pathologic controls, a more complex discriminant analysis based on erythroid and
myeloid antigen expression, percent of bone marrow CD34+ cells, myeloid to
lymphoid ratio, and the ratio of immature to mature myeloid cells, could reliably
differentiate MDS from other conditions. However, an analysis of only percent
positive cells ignores the important qualitative differences in intensity relationships
between multiple antigens. Studies based on the binary logic of expression or
nonexpression of any particular gene product are often inadequate, and can be
difficult to replicate.
Multiple Antigenic Studies in MDS
Marrows from patients with MDS are a study of continuous antigen expression,
containing complex cell mixtures where the same antigen is expressed on both
normal and neoplastic cells at multiple developmental stages. Complicating mat-
ters, neoplastic cells do not obey normal genetic regulatory processes, so that the
discrimination between immature cells from more mature cells in normal devel-
opmental processes is often lost. This type of variable intensity in the amount of
antigen expressed is termed continuous, rather than discrete (29). Therefore, the
percent positive number generated is highly dependent on placement of the posi-
tive/negative threshold. A multiple antigenic discrimination between immature and
mature cells is required, which cannot be assessed by predefined positive/negative
thresholds. Furthermore, a determination of fluorescence intensity of any sin-
gle antigen in a heterogeneous population may oversimplify these heterogeneous
populations and may give an inadequate assessment of disease progression (30).
A large study conducted by Marc Maynadie and his colleagues in France (31)
applied multicolor flow cytometry with CD45 (i.e., leukocyte common antigen)
gating as described below, but assessed each antigen separately, calculating mean
250 Wells and Loken
fluorescence intensities for heterogeneous populations of cells. While hierarchical

antigenic clustering described in this study found no differences for antigens
expressed on monocytes, data for granulocytes was correlated with progression
of MDS, a finding most likely reliant on increases in myeloblasts and immature
myeloid cells in later stages of the disease.
The Utility of CD45 and Right-Angle Light Scatter

At the same time that changes in hematopoietic multiple antigens were being
observed, investigators observed that CD45 in combination with right-angle light
scatter (“side scatter”, SSC) provides additional critical information regarding mat-
uration of hematopoietic cells in normal bone marrow, distinguishing all major
lineages as well as blasts (32). All marrow cell populations can be simultaneously
viewed on the basis of a single antigen, CD45, and the physical light-scattering
properties of the cell. CD45 intensity is differentially expressed on lymphoid,
myeloid, monocyte, and blasts cell populations, while SSC is a measure of cellular
intra-cytoplasmic granularity. The use of CD45 with additional antibody combi-
nations (commonly referred to as CD45 gating) provided a consistent means of
identifying each lineage separately (33).
Multicolor Flow Cytometric Analysis in MDS

Multicolor flow cytometric analysis began in the 1990s and was applied to MDS
marrows early on. Using this approach, K.L. Bowen and Bruce Davis (34) found
that in some MDS marrows, there was an abnormal pattern of low granulocyte
expression of CD16 (Fc-gamma receptor IIIb) or CD16 and CD11b (integrin,
alpha M). Later, Maryalice Stetler-Stevenson and her colleagues at the National
Institutes of Health (14) used a multiple antigen pattern recognition approach in
45 straightforward MDS cases and various control groups in order to define MDS-
associated immunophenotypic abnormalities, and then applied their findings to 20
more difficult cases. These investigators found that while there was no significant
correlation between flow cytometry patterns and either cytogenetic results or FAB
morphologic classification (35), abnormalities in two or more lineages detected by
flow cytometry were specific for MDS, and such abnormalities were not observed
in aplastic anemia, healthy controls, or remission bone marrow aspirates from
patients with nonmyeloid neoplasia (14). The rate of detection of abnormalities in
two or more lineages in straightforward MDS patients was 88% for flow cytometric
immunophenotyping, compared with 93% by morphology. Multiparameter flow
cytometry was supportive of a diagnosis of MDS in most of the 20 difficult cases,
and flow results were abnormal in 6 of 8 cases that could not be classified as
MDS despite repeated marrow morphologic and cytogenetic assessment. Steven
Kussick and his colleagues (11,12) also evaluated known MDS patients by flow
cytometry and found that the overall sensitivity of the analysis for identifying
abnormalities was 95%, whereas the overall specificity for diagnosis was 67%.
Our own studies on MDS began in 1993. We used multidimensional flow

cytometric analysis to develop a scoring system for diagnosis and prognosis in
MDS, quantifying “distance from normal,” which could predict the likelihood of
post stem cell transplant relapse (17). Multiple phenotypic aberrances reflected
accumulation of abnormalities and progression of disease and correlated with
classification schemes in the pretransplant setting.
FLOW CYTOMETRIC SCORING IN DIAGNOSIS AND PROGNOSIS

Diagnosis
Heterogeneity of Bone Marrow at Diagnosis
A bone marrow analysis is an observation at a particular time point, and reflects
the kinetics involved in maintaining the correct proportion of cells in the periph-
ery. In MDS marrows, the immunophenotype observed may be that of a mixture
of normal cells, neoplastic cells, or cells undergoing programmed cell death (see
chap. 4). MDS are dynamic processes in which clonal populations may be abnor-
mally sensitive to the cytokines and growth factors secreted by lymphocytes and
macrophages. The suppression or replacement of normal cells may be the result
of an immortal neoplastic population or a population that has escaped immune
surveillance, and the proportion of abnormal cells may increase over time by a
process of selection and adaptation (36).
At later time points, any permutation of populations observed previously
may be detected. Any measurement of these populations can be significant or
insignificant, reflecting population dynamics and predicting eventual outcome.
Therefore, single time-point measurements may be misleading. Progression of
MDS occurs with additional genetic mutations (37) with immunophenotypic
abnormalities reflecting accumulated gene product aberrancies. The enumera-
tion of abnormalities in flow cytometric scoring systems (FCSS) has been shown
to have diagnostic specificity, prognostic value for patients undergoing stem cell
transplant (SCT) and for disease progression, and correlates with transfusion
dependence (11,14,17,38).
Sensitivity and Specificity of Immunophenotyping

The observation that antigenic abnormalities are associated with cases of MDS
defined by other means (i.e., morphology and cytogenetics) was reported in several
studies (11,14,17,39). In these studies, for a flow cytometric diagnosis of MDS
with two or more abnormalities, the sensitivities ranged from 70% to 98% and the
specificities from 78% to 93%. Based on phenotypic and scatter characteristics,
we characterized marrow cells from 115 patients with MDS, and derived a FCSS
that categorized patients as normal/mildly abnormal (0–1), moderately abnormal
(2,3), or severely abnormal (≥4). We found that with a flow score of 3 or greater,
the specificity for a diagnosis of MDS was 100%; however, sensitivity was only
252 Wells and Loken
55%. Reducing the score to 2 increased the sensitivity to 76%, but reduced the
specificity to 91%.
In addition, a decrease in precursor B cells as assessed by flow cytometry
may also accompany MDS (40,41). Although the significance of this nonspecific
finding is unclear, it may be incorporated into future scoring systems.
The importance of these findings is that a diagnosis of MDS by flow cytom-
etry alone, without definitive morphology and karyotyping studies, should only
be made if the antigenic abnormalities meet sufficient criteria to avoid “overcall-
ing” of patients presenting with unexplained cytopenias. The combination of these
three technologies (morphology, karyotyping, and immunophenotyping) provides
a better overall picture of the status of the marrow than any technique alone. Taken
together, abnormal gene product expression has been consistently observed in
multiple laboratories by flow cytometry using a variety of reagent combinations.
But at the present time, there is no consensus on how to integrate these studies in
making a diagnosis of MDS.
Recently, the term “idiopathic cytopenia of undetermined significance”
(ICUS) was ascribed to patients with otherwise inexplicable cytopenias with-
out the required morphologic or karyotypic criteria for MDS (42). Some of these
patients may have a requisite number of flow cytometric abnormalities to distin-
guish them from patients who will not progress to obvious MDS. This suggests that
ICUS patients should be closely followed and re-evaluated for not only morpho-
logic, karyotypic, and hematologic indices, but also by repeated flow cytometric
evaluation.
A multidimensional flow cytometric analysis that incorporates an extensive
immunophenotypic screening may, additionally, exclude a diagnosis of MDS.
Other causes of cytopenias such as lymphoma, myeloma, or hairy cell leukemia
may, in some cases, not be identified by morphology but may be detected by flow.
Indeed, one study identified several patients with cytopenias who were referred
for SCT with a presumptive diagnosis of MDS, with subsequent deferral of SCT
upon correct diagnosis of a lymphoid neoplasm (43).
Flow cytometric analysis of dysplasia compliments morphologic assessment
because of the additional information regarding maturing myeloid cells, mono-
cytes, and abnormal myeloblasts. Arjan van de Loosdrecht and his colleagues
in the Netherlands found that flow cytometric scoring was more sensitive than
morphology for detecting abnormalities in progenitor cells, granulocytes, and
monocytes (44). In that study, in almost all MDS cases classified by morphology
as unilineage dysplasia (refractory anemia with or without ringed sideroblasts
or unclassifiable MDS), flow cytometry detected myelomonocytic lineage abnor-
malities. Flow cytometry can potentially reclassify patients previously grouped
as refractory anemia to those with refractory anemia with multilineage dysplasia,
although the prognostic importance of this reclassification by flow alone has not
been formally demonstrated. Therefore, the use of a flow scoring system may be
critical for accurate diagnosis when combined with morphology, karyotyping, and
additional clinical diagnostic factors such as transfusion dependence.
Prognosis
International Prognostic Scoring System (IPSS)
Prognostic scoring systems for MDS are based on several laboratory and clinical
factors. The IPSS risk score is based on bone marrow blast cell percentage, number
of peripheral cytopenias, and karyotype (45), and multiple studies have suggested
that the IPSS score correlates well with patient outcomes. However, there is still
variability in patient outcomes within a given risk category. Our study described
above showed that myeloid and monocytic dyspoiesis as determined by flow-
cytometric scoring in MDS correlates with IPSS and helps predict outcome after
hematopoietic stem cell transplantation (17). More importantly, the flow score
could discriminate risk of post-SCT relapse within the IPSS intermediate-1 group,
adding additional prognostic power beyond the IPSS alone.
WHO Classification-Based Prognostic Scoring System (WPSS)

A more recent proposal, the WPSS scoring system, takes into account WHO sub-
groups, karyotype, and transfusion requirement for risk of leukemic evolution in
MDS patients at any time during the course of their disease (46). van de Loos-
drecht and colleagues found that while flow scoring significantly correlated with
WHO subgroups and the WPSS, there was no correlation with WHO cytogenet-
ics or IPSS subgroups, suggesting that different disease entities may be present
within specific WHO and IPSS subgroups that may be further defined by flow
scoring. In addition, when patients were followed over time, the flow score was
able to predict both progression to high-risk MDS and transfusion dependence, as
previously mentioned. Moreover, these studies found that in those patients with
good-risk cytogenetics but with nonlineage expression of lymphoid antigens on
myeloblasts (predominantly the presence of CD7, typically found on T lympho-
cytes but not in myeloid series), there was a higher rate of MDS disease progression
and transfusion dependence.
Monitoring MDS Posttreatment by Flow Scoring

New therapies for treatment of MDS are promising and include lenalidomide
and drugs altering gene expression or signal transduction (47–49). Flow cytomet-
ric scores have been applied to patients posttreatment in an attempt to monitor
response. A study by H. Joachim Deeg in Seattle and his colleagues found that
treatment of MDS patients with antithymocyte globulin (ATG) followed by inter-
mittent treatment with etanercept, a chimeric antibody that binds tumor necrosis
factor, resulted in strikingly improved flow scores (50). Westers and colleagues
in the Netherlands found that low-risk patients treated with erythropoietin and
G-CSF could also be monitored by flow scoring (38). Thus, flow cytometry may
provide additional laboratory assessment of response to effective treatment.
254 Wells and Loken
FLOW CYTOMETRIC APPROACHES IN MDS

Reagent Combinations
The maturation of normal bone marrow populations can be divided into stages
corresponding to the gain or loss of various antigens related to a particular lin-
eage as well as with changes in antigenic amounts. As discussed above, these
stages identified by changes in antigen expression also coincide with morphologic
appearance (9). Reagent antibody combinations to study MDS are selected to
measure abnormalities at developmental transition time-points associated with the
steps of maturation with the goal of separating out each transition.
As mentioned above, the use of CD45 gating in combination with SSC is
a powerful means to demarcate various populations present in marrow aspirates
(Fig. 1). This simple projection of data can give a complete marrow differential
(51), assess hypogranularity as measured by decreased SSC, determine decreased
myelopoiesis as determined by the lymphoid to myeloid ratio, and determine
abnormalities in intensities of CD45, especially in abnormal myeloblasts of MDS
(17) [Fig. 1(B)]. In addition, SSC displayed on a logarithmic rather than linear
scale can more easily identify hypogranularity in maturing myeloid cells and
monocytes, so that the very brightest and very dimmest cells can be easily observed
and compressed.
Gating Strategies
Gating strategies must be capable of distinguishing antigen expression on multiple
lineages and on maturational stages within those lineages. The software used in
flow cytometric analysis of MDS samples must be capable of flexibility so that
multiple projections of data are simultaneously displayed. A combination of gates
using logical operations such as Boolean logic are required to discriminate the
various populations at various stages of maturation (13). This type of multidimen-
sional analysis is distinguished from multiparameter analysis, which set thresholds
and calculates percent positive and forms a single observation of two-dimensional
pattern recognition without observing all other patterns and populations present
in a simultaneous display. A multidimensional analysis is crucial for the study
of MDS, given the multiple lineages involved, the multiple maturational changes
observed, and the multiple clones that may occur in the marrow of a patient
with dysplasia. In contrast to immunophenotyping an acute leukemia, where a
relatively homogeneous blast population is present, is analyzed MDS-associated
changes can affect the entire heterogeneous maturational sequence, which must
therefore be assessed in its entirety.
Analysis of Myeloblasts
The combination of CD45 and SSC provides a mean to identify myeloblasts
before they change in CD45 expression or light-scatter properties as they mature
(33). Myeloblasts in human bone marrow express relatively low levels of CD45
Figure 1 CD45 and right-angle light scattering (SSC): (A) CD45 and SSC pattern from
a normal marrow showing representative granularity and CD45 expression of all non-
erythroid lineages and (B) abnormal CD45/SSC pattern in an MDS marrow with hypogran-
ularity in the myeloid compartment and decreased CD45 in the myeloblasts. Note the
distinct differences in the displayed log scale for these parameters as compared to normal.
as compared to mature lymphocytes and their maturing progeny. On a flow dia-

gram, this low-CD45 region [Fig. 1(A)], includes not only myeloblasts, but also B
precursor cells (lymphoblasts/“hematogones”), monoblasts, basophils, erythrob-
lasts, dendritic cell precursors, hematopoietic stem cells (HSC), and all CD34+
cells, in various proportions depending on the composition of the specimen. The
neoplastic blasts observed in MDS are usually restricted to this region; however,
256 Wells and Loken
maturation along the monocytic or neutrophilic lineages may place the abnormal
blasts outside this gate. Abnormal myeloblasts in MDS can have a log decrease
in CD45 expression (17) [Fig. 1(B)] as compared to normal myeloblasts. In rare
instances, abnormal myeloblasts can lack CD45 expression altogether (52) and
must be carefully distinguished from nonhematopoietic cell malignancies (53).
Marrow aspirate cells lacking CD45 may also represent aggregated platelets and
nucleated red cells, leukemic blasts in acute lymphoblastic leukemia, lymphoid
blast crisis of chronic myelogenous leukemia, malignant plasma cell populations,
and rarely, small-cell lung carcinoma.
The identification of immature cells in marrow has been facilitated by the use
of CD34, an antigen expressed on all hematopoietic precursors but lost relatively
early in the development of marrow cells (it should be noted that CD34 is dimly
expressed on mature basophils, which also lie within the CD45/SSC blast region
and must be distinguished from immature CD34+ precursor cells). The CD34+
cells within the marrow must not be considered a homogeneous population, since
this group includes precursors of B lymphoid, monocyte, neutrophil, erythroid,
dendritic, and megakaryocytic lineages. The actual composition of the CD34+ cell
fraction depends on the requirements of the individual responses to hematopoietic
stress. Each of these early committed cells has different antigenic expression pat-
terns; therefore, changes in antigenic expression of the CD34 cell compartment
may be the result of differences in composition of normal cells rather than the
presence of abnormal cells. Results are highly dependent on the CD34 antibody
used, since there are three different epitopes of CD34 antigen, and different chro-
mophores coupled to the antigen can also yield distinct results (e.g., phycoerythrin
yields a signal approximately 20 times brighter than fluorescein) (13).
In MDS, CD34+ cells have been studied extensively (31,40,54–56).
Kikoyuki Ogata and colleagues in Japan applied mean fluorescence intensity to
CD34+ myeloblasts and found increased intensity of several antigens, in particu-
lar CD4, CD15, and CD117, in low-grade MDS marrows (56). Mariela Monreal
and colleagues in Argentina found that the proportion of hematopoietic stem cells
(HSC), characterized by bright CD34, intermediate SSC, and dim CD38, was
significantly increased in high-risk MDS and secondary AML, but not in low-risk
MDS (55), suggesting an increase in immature CD34+ cells in higher grade MDS.
In addition, CD34 may also be aberrantly expressed on mature monocytes or neu-
trophils in some cases of MDS (17). Moreover, as discussed above, myeloblasts in
MDS can be CD34−, so simple enumeration of CD34+ cells must be interpreted
with caution in studies of this disease. Often in MDS, the percentage of all marrow
CD34+ cells is within the normal range, but the composition of CD34+ cells may
be profoundly altered.
Because of the variability of myeloblast position in CD45/SSC plots and
the heterogeneity of the CD34+ fraction, multiple, or redundant methods must be
applied to identify and enumerate the blast cells present.
Blasts committed to the myeloid lineage present in the CD45/SSC blast
region can be defined by using a different combination of reagents combining
the Class II histocompatibility antigen HLA-DR, and the leukocyte adhesion

molecule, CD11b. While myeloblasts express HLA-DR, they do not express
CD11b. This additionally allows the exclusion of basophils and immature mono-
cytes (both CD11b positive). The combination of CD34 with HLA-DR can also
differentiate CD34+ myeloblasts from dimly CD34+ basophils (HLA-DR nega-
tive). Erythroblasts can be distinguished from myeloblasts by the former’s expres-
sion of CD36 (thrombospondin receptor), bright CD71 (transferrin receptor), lack
of CD64 (Fc fragment of IgG, high-affinity receptor Ia), and a decrease in CD45
within the blast region (4,57). B-cell precursors within the region are identified
by their characteristic CD19 expression and decreased SSC (58,59). In some
MDS cases, myeloblasts may abnormally lack expression of HLA-DR, or express
CD11b to varying degrees and must be distinguished from their normal immature
counterparts. Any single antigen may be lost or aberrantly expressed in MDS.
Therefore, reliance on a single expression pattern to identify blasts can lead to
inappropriate conclusions.
CD117, a type III receptor tyrosine kinase (c-kit receptor), is an additional
antigen expressed on myeloblasts, mast cells, a subset of HSC, and other tissues
(60), and like other single antigens, can be variably expressed on neoplastic
myeloblasts in MDS. CD117 is lost during early monocyte development, and
is lost at the promyelocyte stage during neutrophil development. It can serve as
an alternative or additional antigenic dimension to further distinguish myeloblasts
from other cells that appear in the “blast region.”
Analysis of Monocytes
Monocytes and dendritic cells are most certainly part of the abnormal processes
that occur in MDS, with their notable increased secretion of cytokines, abnormal
phagocytic functions, and possibly disordered antigen presentation (61). Mono-
cytes may be proportionally increased or decreased in MDS marrows. Morpho-
logic monocytic abnormalities in MDS marrows are not always readily discern-
able while flow cytometric analyses can detect distinct antigenic differences from
normal. These include abnormal antigenic relationships between HLA-DR and
CD11b, decreased CD45 expression, expression of CD56, lack of CD14, CD13,
or CD33 expression, the presence of CD34, the presence of nonlineage lymphoid
antigens, and hypogranularity (17).
Analysis of Maturing Myeloid Cells

The heterogeneic immunophenotypic patterns of myeloid maturation remarkably
correspond with the changes observed by morphology in each developmental
stage, as illustrated in Figure 2. As maturation progresses from the myeloblast to
the promyelocyte stage in normal, the cell enlarges and abundant primary gran-
ules are synthesized (62). The phenotypic events associated with promyelocyte
development include increased granularity as evidenced by increased SSC, loss
of HLA-DR, CD34, and CD117, increased intensity of CD15 (myeloid-specific
258 Wells and Loken
104 Monos
Neut
103
Eos
CD13 PE
MB
Bands
Baso
102
MC MT
Pros
101
101 102 103 104
CD16 FITC
(A)
104
103
CD13 PE
102
101
101 102 103 104
CD16 FITC
(B)
Figure 2 Depiction of a normal immunophenotypic pattern for CD13 and CD16 with
corresponding maturational stages by morphology. (A) The characteristic “sickle” pattern
of CD13 and CD16 as observed in flow cytometric analysis, gated on myeloblasts (MB) and
basophils, shown in red, with monocytes, maturing myeloid cells [promyelocytes (Pros),
myelocytes (MC), metamyelocytes (MT), bands, neutrophils, and eosinophils in green, with
lymphocytes and erythroid cells eliminated]. Basophils have slightly decreased intensity
of CD13. Eosinophils are easily identified by their characteristic degree of CD13/CD16
intensity, between maturing myeloid cells and monocytes. Note that the maturing myeloid
cells lose CD13 and then increase CD13 intensity as they simultaneously gain CD16
intensity to become myelocytes. (B) Illustration of the morphologic stages portrayed along
their corresponding antigenic expression. Also note that at the stage when promyelocytes
lose their primary granules to become myelocytes, they also gain CD13 and CD16 intensity
in a coordinated fashion.
fucosyl transferase), and decreased intensity of CD13. The appearance of specific

or secondary granules demarcates progression to a myelocyte, delineated pheno-
typically by decreased SSC reflecting a decrease in the light refraction of these
granules, loss of CD4, and acquisition of CD11b, while CD13 intensity remains
essentially the same. As maturation and cell division progress, cellular granularity
decreases, CD45 and CD11b intensity increases, and there are coordinated
increases in CD13 and CD16 intensity reflecting coordinated gene product
synthesis and expression. The characteristic sickle-shaped pattern of CD13 and
CD16 is a hallmark of normal expression of these two antigens (Fig. 2). In
carefully gated multidimensional flow cytometric analyses using Boolean gating,
perturbations of this complex maturational coordination have been observed
and described in MDS (11,13,14,17,39,44). However, as mentioned above,
single-pattern recognition may be misleading. For example, the absence of CD16
expression may be due to a marked shift to the left as the result of a recovering
marrow post-insult, in paroxysmal nocturnal hemoglobinemia (CD16 is a GPI-
linked protein) (63,64), or in genetic polymorphisms (familial absence of CD16)
(65,66).
Additional immunophenotypic abnormalities observed in maturing myeloid
populations in MDS include increases or decreases in CD64 and CD10 (14,39),
CD56 on myeloblasts and maturing myeloid cells, an increased lymphoid to
myeloid ratio indicating decreased myelopoiesis, lack of CD33 expression, and
abnormal decreases in CD45 expression (11,14,17,39,44,67).
Analysis of the Erythroid Compartment

Erythroid development is relatively simple as compared to leukocyte maturation.
Only two steps are clearly identifiable with most antigens being lost, rather than
acquired, with maturation. The antigens used to study erythroid development that
are lost include HLA-DR, HLA Class I antigens, CD34, and CD45, while CD71
and CD36 expression precedes CD235 (glycophorin) and all blood-group antigens.
The simplicity of the development contrasts with the daunting technical difficul-
ties involved with removing red blood cells, while maintaining the nucleated cell
compartment. The most promising phenotypic changes relating to genetic dys-
regulation among erythroid cells in MDS are intracytoplasmic. Hemoglobin F is
markedly increased in patients with MDS. Fetal hemoglobin containing erythrob-
lasts were found to be incapable of maturing hemoglobin F cells to end stage, and
correlated with extramedullary hematopoiesis (68–70). In addition, Matteo Della
Porta and colleagues (71) evaluated the proteins of erythroid iron metabolism and
mitochondrial ferritin to quantitatively assess erythroid dysplasia, and found that
this correlated with a diagnosis of MDS with ringed sideroblasts.
Example of Multiple Phenotypic Abnormalities in a Case of MDS

Figure 3 is an example of a multidimensional analysis (which can only be dis-
played as bivariate histograms) in an analysis of a MDS marrow from an elderly
260
Figure 3 Display of multiple bivariate immunophenotypic analyses of a MDS marrow: (A) light scatter gate
eliminating erythroid cells, cellular debris, and platelets; (B) CD45 and SSC histogram of lymphoid (blue),
maturing myeloid cells (green) and myeloblasts (red), with arrows showing hypogranularity of maturing myeloid
cells and decreased CD45 expression of myeloblasts (red), similar to Figure 1; (C–F) myeloblasts express
heterogeneous HLA-DR and CD13, while maturing myeloid cells show aberrant asynchronous maturation with
presence of the mature antigen CD11b, but virtually no maturation to CD16 (grey ovals); (G–P) myeloblasts
Wells and Loken
completely lack CD34 while expressing heterogenous CD117 (arrows), the remaining antigens are expressed
normally. Note the decreased proportion of CD14 positive monocytes (J).
Figure 3 Continued.
262 Wells and Loken
patient presenting with anemia and thrombocytopenia, with normal karyotyping

and FISH studies, demonstrating several maturational derangements. The abnor-
malities observed include hypogranularity of maturing myeloid cells, an asyn-
chronous shift to the left with virtually no maturation of CD16. The myeloblasts
were abnormal and proportionally increased (15%), had decreased CD45 expres-
sion with lack of CD34 expression and simultaneous CD117 expression. While
the marrow was classified as RAEB in the WHO classification (35), and the IPSS
category for this patient was INT-2 (45), the flow cytometric score for this marrow
aspirate was in the severe category. These findings could arguably categorize this
MDS as high for risk of progression to acute leukemia, even though the blast
count was not above 20% and CD34+ cells are not increased. This marrow study
stresses the importance of using multiple data projections and antigenic definitions
to accurately characterize an MDS marrow and does not rely on single parameters
such as the presence of CD34+ myeloblasts.
SUMMARY AND CONCLUSION

Flow cytometry can detect phenotypic abnormalities in patients with MDS with
acceptable sensitivity and specificity to be useful in diagnosis, especially in diffi-
cult cases without concurrent abnormalities in karyotyping and morphology. The
abnormalities are identified as inappropriate intensity and/or relationships of anti-
genic gene products, and can be identified on both immature and mature cells.
Flow scoring systems have been developed to quantify abnormalities as “distance
from normal” and reflect the degree of dysregulated gene coordination. Pheno-
typically aberrant myeloblasts, even when they comprise less than 5% marrow
cellularity, are highly suggestive of progressive disease and may indicate a need to
re-classify MDS patients. Consensus on reagent panels, flow scoring, and analysis
methods and software and large clinical double blinded studies with outcome data
are needed, in order to define new MDS subgroups and to further refine diagnosis
and prognosis in this disease.
REFERENCES
1. Brunning RD, Bennett JM, Flandrin, et al. Vardiman myelodysplastic syndromes.
In: Jaffe ES, Harris NL, Stein H, and Vardiman JW, eds. World Health Organization
Classification of Tumours Pathology & Genetics Tumours of Haematopoietic and
Lymphoid Tissues. Lyon, France: IARC Press, 2001:62–73.
2. Civin CI, Banquerigo ML, Strauss LC, et al. Antigenic analysis of hematopoiesis. VI.
Flow cytometric characterization of My-10-positive progenitor cells in normal human
bone marrow. Exp Hematol 1987; 15:10–17.
3. Civin CI, Loken MR. Cell surface antigens on human marrow cells: Dissection of
hematopoietic development using monoclonal antibodies and multiparameter flow
cytometry. Int J Cell Cloning 1987; 5:267–288.
4. Loken MR, Shah VO, Dattilio KL, et al. Flow cytometric analysis of human bone
marrow: I. Normal erythroid development. Blood 1987; 69:255–263.
5. Terstappen LW, Levin J. Bone marrow cell differential counts obtained by multidi-
mensional flow cytometry. Blood Cells 1992; 18:311–30; discussion 31–32.
6. Terstappen LW, Loken MR. Five-dimensional flow cytometry as a new approach for
blood and bone marrow differentials. Cytometry 1988; 9:548–556.
7. Terstappen LW, Loken MR. Myeloid cell differentiation in normal bone marrow and
acute myeloid leukemia assessed by multi-dimensional flow cytometry. Anal Cell
Pathol 1990; 2:229–240.
8. Terstappen LW, Safford M, Loken MR. Flow cytometric analysis of human bone
marrow. III. Neutrophil maturation. Leukemia 1990; 4:657–663.
9. Terstappen LWMM, Segers-Nolten I, Safford M, et al. Monomyeloid cell differen-
tiation in normal bone marrow assessed by multidimensional flow cytometry. In:
Burger G, Oberholzer M, Vooijs GP, eds. Advances in Analytical Cellular Pathology.
Amsterdam: Excerpta Medica, 1990:211–212.
10. Benesch M, Deeg HJ, Wells D, et al. Flow cytometry for diagnosis and assessment of
prognosis in patients with myelodysplastic syndromes. Hematology 2004; 9:171–177.
12. Kussick SJ, Wood BL. Using 4-color flow cytometry to identify abnormal myeloid
populations. Arch Pathol Lab Med 2003; 127:1140–1147.
13. Loken MR, van de Loosdrecht A, Ogata K, et al. Flow cytometry in myelodysplastic
syndromes: Report from a working conference. Leuk Res 2007; 32:5–17.
14. Stetler-Stevenson M, Arthur DC, Jabbour N, et al. Diagnostic utility of flow cytometric
immunophenotyping in myelodysplastic syndrome. Blood 2001; 98:979–987.
15. Terstappen LW, Konemann S, Safford M, et al. Flow cytometric characterization of
acute myeloid leukemia. Part 1. Significance of light scattering properties. Leukemia
1991; 5:315–321.
16. Terstappen LW, Safford M, Konemann S, et al. Flow cytometric characterization of
acute myeloid leukemia. Part II. Phenotypic heterogeneity at diagnosis. Leukemia
1992; 6:70–80.
the IPSS and with outcome after hematopoietic stem cell transplantation. Blood 2003;
102:394–403.
18. Sievers EL, Lange BJ, Alonzo TA, et al. Immunophenotypic evidence of leukemia
after induction therapy predicts relapse: Results from a prospective Children’s Cancer
Group study of 252 patients with acute myeloid leukemia. Blood 2003; 101:3398–
3406.
19. Sievers EL, Radich JP. Detection of minimal residual disease in acute leukemia. Curr
Opin Hematol 2000; 7:212–216.
20. Gonchoroff NJ, Katzmann JA, Currie RM, et al. S-phase detection with an antibody to
bromodeoxyuridine. Role of DNase pretreatment. J Immunol Methods 1986; 93:97–
101.
21. Houck DW, Loken MR. Simultaneous analysis of cell surface antigens, bromod-
eoxyuridine incorporation and DNA content. Cytometry 1985; 6:531–538.
22. Jensen IM, Hokland M, Hokland P. A quantitative evaluation of erythropoiesis in
myelodysplastic syndromes using multiparameter flow cytometry. Leuk Res 1993;
17:839–846.
264 Wells and Loken
23. Jensen IM, Hokland P. The proliferative activity of myelopoiesis in myelodysplasia

evaluated by multiparameter flow cytometry. Br J Haematol 1994; 87:477–482.
24. Smith PJ, Wiltshire M, Davies S, et al. A novel cell permeant and far red-fluorescing
DNA probe, DRAQ5, for blood cell discrimination by flow cytometry. J Immunol
Methods 1999; 229:131–139.
25. Holm M, Hoyer M, Jensen I, et al. Dynamic cell cycle kinetics in vitro and in vivo in
myelodysplastic syndromes with special reference to the influence of hematopoietic
growth factors. Leuk Res 2000; 24:999–1008.
26. Holm M, Thomsen M, Hoyer M, et al. Optimization of a flow cytometric method
for the simultaneous measurement of cell surface antigen, DNA content, and in vitro
BrdUrd incorporation into normal and malignant hematopoietic cells. Cytometry
1998; 32:28–36.
27. Elghetany M. Surface marker abnormalities in myelodysplastic syndromes (Review).
Haematologica 1998; 83:1104–1115.
28. Malcovati L, Della Porta MG, Lunghi M, et al. Flow cytometry evaluation of erythroid
and myeloid dysplasia in patients with myelodysplastic syndrome. Leukemia 2005;
19:776–783.
29. Owens M, Loken MR. Flow Cytometry Principles for Clinical Laboratory Prac-
tice: Quality Assurance for Quantitative Immunophenotyping. New York: Wylie-Liss,
1995.
30. Wells DA, Loken MR. Flow cytometric mean fluorescence intensity: The biophysics
behind the number. Leuk Res 2008; 32(6):845–6.
31. Maynadie M, Picard F, Husson B, et al. Immunophenotypic clustering of myelodys-
32. Shah VO, Civin CI, Loken MR. Flow cytometric analysis of human bone marrow.
IV. Differential quantitative expression of T-200 common leukocyte antigen during
normal hemopoiesis. J Immunol 1988; 140:1861–1867.
33. Stelzer GT, Shults KE, Loken MR. CD45 gating for routine flow cytometric analysis
of human bone marrow specimens. Ann NY Acad Sci 1993; 677:265–280.
34. Bowen KL, Davis BH. Abnormal patterns of expression of CD16 (FcR-III) and CD11b
(CRIII) in the bone marrow of patients with myelodysplastic syndrome. Lab Hematol
1997; 3:292–298.
36. Audeh M. MDS and secondary AML: An evolutionary perspective. In: Raza A,
Mundel SD, eds. Myelodysplastic Syndromes and Secondary Acute Myelogenous
Leukemia: Directions for a New Millennium. Boston: Kluwer Academic Publishers,
2001.
37. Bacher U, Haferlach T, Kern W, et al. A comparative study of molecular mutations in
381 patients with myelodysplastic syndrome and in 4130 patients with acute myeloid
leukemia. Haematologica 2007; 92:744–752.
38. Westers T, Chamuleau MED, Westra AH. Changes in flow cytometric dysplasia score
during treatment with erythropoietin/G-CSF reflects responses in low/int-I myelodys-
plastic syndromes. Leuk Res 2007; 31(Suppl. 1):S90.
39. Stachurski D, Smith BR, Pozdnyakova O, et al. Flow cytometric analysis of
myelomonocytic cells by a pattern recognition approach is sensitive and specific
in diagnosing myelodysplastic syndrome and related marrow diseases: Emphasis on a
global evaluation and recognition of diagnostic pitfalls. Leuk Res 2008; 32(2):215–24.
40. Ogata K, Nakamura K, Yokose N, et al. Clinical significance of phenotypic features

of blasts in patients with myelodysplastic syndrome. Blood 2002; 100:3887–3896.
31:1461–1468.
43. Wells DA, Hall MC, Shulman HM, et al. Occult B cell malignancies can be detected
by three-color flow cytometry in patients with cytopenias. Leukemia 1998; 12:2015–
2023.
prognostic subgroups in low and intermediate-1 risk myelodysplastic syndromes by
flow cytometry. Blood 2008; 111(3):)1067–1077.
J Clin Oncol 2007; 25:3503–3510.
47. Giagounidis AA, Germing U, Aul C. Current treatment strategies in low-risk
myelodysplastic syndromes. Cancer Treat Rev 2007; 33:S19–S24.
48. List AF. New approaches to the treatment of myelodysplasia. Oncologist 2002;
7(Suppl. 1):39–49.
49. Shadduck RK, Latsko JM, Rossetti JM, et al. Recent advances in myelodysplastic
syndromes. Exp Hematol 2007; 35:137–143.
50. Deeg HJ, Jiang PY, Holmberg LA, et al. Hematologic responses of patients with
MDS to antithymocyte globulin plus etanercept correlate with improved flow scores
of marrow cells. Leuk Res 2004; 28:1177–1180.
51. Rainer RO, Hodges L, Seltzer GT. CD 45 gating correlates with bone marrow differ-
ential. Cytometry 1995; 22:139–145.
52. Khalidi HS, Medeiros LJ, Chang KL, et al. The immunophenotype of adult acute
myeloid leukemia: High frequency of lymphoid antigen expression and comparison
of immunophenotype, French-American-British classification, and karyotypic abnor-
malities. Am J Clin Pathol 1998; 109:211–220.
53. Etzell JE, Keet C, McDonald W, et al. Medulloblastoma simulating acute myeloid
leukemia: Case report with a review of “myeloid antigen” expression in non-
hematopoietic tissues and tumors. J Pediatr Hematol Oncol 2006; 28:703–710.
54. Haase D, Feuring-Buske M, Schafer C, et al. Cytogenetic analysis of CD34+ sub-
populations in AML and MDS characterized by the expression of CD38 and CD117.
Leukemia 1997; 11:674–679.
55. Monreal MB, Pardo ML, Pavlovsky MA, et al. Increased immature hematopoietic
progenitor cells CD34+/CD38 dim in myelodysplasia. Cytometry B Clin Cytom
2006; 70:63–70.
56. Ogata K, Kishikawa Y, Satoh C, et al. Diagnostic application of flow cytometric
characteristics of CD34+ cells in low-grade myelodysplastic syndromes. Blood 2006;
108:1037–1044.
57. Loken MR, Civin CI, Bigbee WL, et al. Coordinate glycosylation and cell sur-
face expression of glycophorin A during normal human erythropoiesis. Blood 1987;
70:1959–1961.
266 Wells and Loken
58. Hollander Z, Shah VO, Civin CI, et al. Assessment of proliferation during maturation
of the B lymphoid lineage in normal human bone marrow. Blood 1988; 71:528–531.
59. Loken MR, Shah VO, Hollander Z, et al. Flow cytometric analysis of normal B
lymphoid development. Pathol Immunopathol Res 1988; 7:357–370.
60. Escribano L, Ocqueteau M, Almeida J, et al. Expression of the c-kit (CD117) molecule
in normal and malignant hematopoiesis. Leuk Lymphoma 1998; 30:459–466.
61. Castoldi G, Rigolin GM. The monocytic component in myelodysplastic syndromes.
Cancer Treat Res 2001; 108:81–92.
62. Kapff C, Jandl JH. Blood Atlas and Sourcebook of Hematology. 2nd ed. Boston:
Little, Brown and Company, 1991.
63. Dunn DE, Tanawattanacharoen P, Boccuni P, et al. Paroxysmal nocturnal hemoglobin-
uria cells in patients with bone marrow failure syndromes. Ann Intern Med 1999;
131:401–408.
64. Schrezenmeier H, Hertenstein B, Wagner B, et al. A pathogenetic link between aplastic
anemia and paroxysmal nocturnal hemoglobinuria is suggested by a high frequency
of aplastic anemia patients with a deficiency of phosphatidylinositol glycan anchored
proteins. Exp Hematol 1995; 23:81–87.
65. de Haas M, Kleijer M, van Zwieten R, et al. Neutrophil Fc gamma RIIIb deficiency,
nature, and clinical consequences: A study of 21 individuals from 14 families. Blood
1995; 86:2403–2413.
66. Wang L, Wells DA, Deeg HJ, et al. Flow cytometric detection of nonneoplastic
antigenic polymorphisms of donor origin after allogeneic marrow transplant: A report
of two cases. Am J Clin Pathol 2004; 122:135–140.
67. Mann KP, DeCastro CM, Liu J, et al. Neural cell adhesion molecule (CD56)-positive
acute myelogenous leukemia and myelodysplastic and myeloproliferative syndromes.
Am J Clin Pathol 1997; 107:653–660.
68. Choi JW, Kim Y, Fujino M, et al. Significance of fetal hemoglobin-containing
erythroblasts (F blasts) and the F blast/F cell ratio in myelodysplastic syndromes.
Leukemia 2002; 16:1478–1483.
69. Craig JE, Sampietro M, Oscier DG, et al. Myelodysplastic syndrome with kary-
otype abnormality is associated with elevated F-cell production. Br J Haematol 1996;
93:601–605.
70. Mendek-Czajkowska E, Slomkowski M, Zdebska E, et al. Hemoglobin F in primary
myelofibrosis and in myelodysplasia. Clin Lab Haematol 2003; 25:289–292.
71. Della Porta MG, Malcovati L, Invernizzi R, et al. Flow cytometry evaluation of ery-
throid dysplasia in patients with myelodysplastic syndrome. Leukemia 2006; 20:549–
555.
12
Molecular Pathogenesis of the
5q− Syndrome
Jacqueline Boultwood and James S. Wainscoat

Leukaemia Research Fund Molecular Haematology Unit, Nuffield Department
of Clinical Laboratory Sciences, John Radcliffe Hospital, Oxford, U.K.
INRODUCTION
The 5q− syndrome was first described as a “distinct haematological disorder with
deletion of long arm of no. 5 chromosome” by Herman van Den Berghe and
his colleagues in Leuven, Belgium, in 1974 (1). The syndrome was described
in more detail in a subsequent report in 1975 by Gerard Sokal and colleagues,
also in Leuven (2). The principal features described in these early reports were
a moderate-to-severe macrocytic anemia, with slight leukopenia but normal or
elevated platelet count. The bone marrow showed a depressed erythroid series
with an occasional excess of myeloblasts. Most of the megakaryocytes had a
nonlobulated nucleus. It was recognized early that patients with the 5q− syndrome
had a female preponderance, and a good prognosis with a low transformation rate
to acute myeloid leukemia (AML) (3). The 5q− syndrome quickly became linked
with the myelodysplastic syndromes (MDS). The World Health Organization
(WHO) definition of the 5q− syndrome (see chap. 9) is part of the WHO’s MDS
classification and includes a more stringent requirement for the patient to be in the
refractory anemia (RA) subgroup of MDS, excluding patients with over 5% blasts.
Those patients who fall outside the WHO definition of the 5q− syndrome may
have some of the hematological features of the 5q− syndrome, but nevertheless
do not share its good prognosis (4).
The specificity of the hematological indicators for the 5q− syndrome was
tested by L. Teerenhovi in Helsinki (5). Teerenhovi analyzed 83 patients with MDS
267
268 Boultwood and Wainscoat
for the presence of three hematological features: macrocytic anemia, normal or

high platelet count, and megakaryocytic hypolobulation in most megakaryocytes.
Nine patients displayed all the three features, and in eight of these patients,
the 5q− chromosome was the only clonal aberration. This study confirms and
extends the original observations on the 5q− syndrome that there is indeed a
strong phenotype/genotype relationship in this disorder. This phenotype/genotype
relationship has been a foundation for all our research on the 5q− syndrome.
Notably, a strong phenotype/genotype relationship is generally not seen in other
examples of large chromosomal deletions in leukemia.
In this review, we focus on the 5q− syndrome, but it should be noted that
although the 5q deletion is found commonly in MDS, it also occurs across a spec-
trum of hematological disorders. For example, in a single institution study from the
Mayo Clinic of 358 consecutive cases associated with chromosome 5q deletions,
the following diagnoses were identified: MDS (53%), AML (22%), plasma cell
proliferative disorder (PCPD; 9%), myeloproliferative disorder (MPD; 7%), acute
lymphoblastic leukemia (ALL; 2%), PCPD with MDS (2%), MDS/MPD (2%),
and malignant lymphoma (ML; 2%) (6). The relationship between these various
disorders is not known, although it would seem logical that at least in MDS and
AML the 5q deletion may play some overlapping pathogenetic role, because of
the pathological similarities of those entities.
CLINICAL, LABORATORY, AND PROGNOSTIC FEATURES

The most recent study to examine the features of the 5q− syndrome and related
disorders in detail is that of Ari Giagounidis in Dusseldorf (7). Giagounidis exam-
ined 76 consecutive patients with MDS and either isolated del(5q) (n = 66) or
del(5q) plus one additional chromosomal abnormality (n = 10). Nearly half of the
study population showed erythroid hypoplasia in the bone marrow and 22% of the
total patient group showed significant dysplasia in nonmegakaryocytic cell lines.
The projected median survival of patients with isolated del(5q) was 146 months
with a median follow-up of 67 months. Patients with an increased medullary blast
count and those with an additional chromosome abnormality had a significantly
shorter overall survival (24 and 45 months, respectively). Neither different cytoge-
netic breakpoints nor the relative amount of dysplasia had an impact on survival.
In total 29 patients died, mostly from transformation to acute leukemia, infection,
or cardiac failure.
This study, in addition to confirming earlier studies concerning the hemato-
logical features of the 5q− syndrome, answers a number of other questions. For
example, although it was already clear that the addition of several cytogenetic aber-
rations to a del(5q) had a negative impact on survival, it was not clear that this was
also true for just a single extra abnormality, which the study clearly demonstrated.
Another question concerns the prognosis of those patients with an increased blast
count but otherwise with features typical of the 5q− syndrome. The study by
Giagounidis demonstrated the difference in survival between patients with del(5q)
Molecular Pathogenesis of the 5q− Syndrome 269
and less than 5% medullary blasts and those with more than 5% medullary blasts,
supporting the exclusion of the latter group of patients from the WHO classification
of the 5q− syndrome.
The data also confirm the previous observations of a female preponderance
of females in the 5q− syndrome. However, in contrast to the observation of Bent
Pedersen in Aarhus, Denmark, and his colleagues (8), the Giagounidis data show
neither a prolonged survival of women nor a higher median age of female patients
at diagnosis. Therefore, these data offer no support for the hypothesis advanced by
Pedersen that the female preponderance in the 5q− syndrome is merely an effect
of these phenomena.
This study is the first large-scale study of the bone marrow morphology in the
5q− syndrome (7). An important feature to emerge is the erythroid hypoplasia of
the bone marrow–observed in nearly half of the patients. This finding is in contrast
to other early forms of MDS (RA with or without ring sideroblasts or refractory
cytopenia with multilineage dysplasia) where the marrow is typically hypercellular
or normocellular (see chap. 10). A previous study of 43 cases from the Mayo Clinic
found that only 25% of their cases showed erythroid hypocellularity (9).
CELL OF ORIGIN
There is much evidence for involvement of multiple myeloid lineages in the 5q−
syndrome suggesting the possibility of a myeloid-restricted progenitor. However,
various lines of evidence have emerged consistent with the view that a pluripotent
(lympho-myeloid) stem cell is the primary target. One study of cell subsets from
patients with del(5q) found no evidence of peripheral blood T-cell and only one
case of B-cell involvement (10). The same study reported that a minimum of
94% of cells in the minor CD34+CD38− HSC compartment were 5q deleted as
determined by FISH and that in 3 of 5 patients 5q aberrations were detected in
a large fraction of purified CD34+CD19+ pro-B cells. Our study of three 5q−
cases using combined immunophenotyping and FISH, identified the involvement
of B-cells in one case (11).
Another study reported that the 5q31 deletion was present in the megakary-
ocytic cells in patients with the 5q− syndrome (12). A subsequent paper examined
the multilineage involvement in the various myeloid cell lineages (13). The per-
centages of cells found to be carrying the 5q31 deletion were as follows: erythrob-
lasts 40% to 50%, promyelocytes 45%, neutrophils 28%, and megakaryocytes
52% to 68%. Interestingly, when the megakaryocytes were divided up between
the large multilobular megakaryocytes and hypolobular megakaryocytes charac-
teristic of the syndrome, the scores were 19% and 93%, respectively. The authors
of this report concluded that there is a preferential distribution of the 5q31 deletion
within immature cells and morphologically abnormal megakaryocytes.
A recent gene expression study of 5q− syndrome using purified 5q− deleted
CD34+CD38−Thy1+ cells also gives further support for the early progenitor or
stem cell origin of 5q MDS (14).
CYTOGENETICS
The early cytogenetic studies of the interstitial 5q deletion showed that the distal
breakpoint by conventional cytogenetics was usually in band q32 and the proximal
breakpoint in q12 or q14, with variant breakpoints in around 10% of the cases
(3). Although the 5q deletion is typically large, encompassing most of the long
arm of chromosome 5, cases with smaller deletions have been reported (15). There
appears to be no difference in the patterns of reported breakpoints of the 5q deletion
between MDS and AML.
GENETIC STABILITY IN THE 5Q− SYNDROME

By definition, the del(5q) occurs as the sole karyotypic abnormality in the 5q−
syndrome and the relative clinical stability of this syndrome may be due to the
absence of other additional abnormalities. A study by Crescenzi in Perugia and
her colleagues reported the deletion of several tumor suppressor genes in MDS
patients with disease characterized by both the del(5q) and additional karyotypic
changes. However, no additional deletions were observed in patients with the 5q−
syndrome (16). In a screen of patients with MDS and the del(5q) for mutation
of three genes known to be associated with disease progression in MDS: FLT3,
NRAS and p53, we found no cases of the 5q− syndrome with mutations in these
genes (17).
Most recently, we have undertaken a genome wide single nucleotide poly-
morphism (SNP) analysis of a spectrum of 42 MDS patients with del(5q) in order
to investigate whether additional genomic abnormalities occur. SNP array analysis
has been shown to detect not only gene deletions but also regions of uniparental
disomy (UPD). Copy number changes in addition to the del(5q) were observed in
approximately half of the 21 del(5q) MDS patients, but in none of the 21 patients
with the 5q− syndrome using the 50 K Affymetrix SNP arrays. Moreover, this
study showed that while small regions of UPD (⬎2 Mb) are common both in the
5q− syndrome and del(5q) MDS, large regions of UPD (⬎10 Mb) are found in
del(5q) MDS, but not in the 5q− syndrome (18).
MOLECULAR MAPPING OF THE CDR IN THE 5Q− SYNDROME

The consistent loss of genetic material from the long arm of chromosome 5 in
association with a specific syndrome is suggestive of a recessive mechanism of
tumorigenesis. The del(5q) in the 5q− syndrome is widely believed to give the
location for a gene(s), the loss of which may affect important processes such
as growth control and normal hematopoiesis (19,20). Using FISH and molecular
mapping techniques, our group identified the commonly deleted region (CDR)
or critical region of gene loss of the 5q− syndrome (11,21). We subsequently
narrowed the CDR to approximately 1.5 Mb interval at 5q31−q32 flanked by
D5S413 and the GLRA1 gene (22). Defining a minimal CDR for a disease allows
for the identification of candidate genes within the interval. We have been involved
in the generation of a transcription map of the CDR (23–25) and have used the
Ensembl gene prediction program for the complete genomic annotation of this
region (22). The CDR is gene rich and contains 43 genes (22) (Table 1). We have
suggested that one or more of the 43 candidate genes mapping within the CDR
of the 5q− syndrome represents the gene or genes critical to the development
of this disorder (22). Several promising candidate genes map within the CDR
including the tumor suppressor genes SPARC and MEGF1, RPS14, a component
of the 40S ribosomal subunit and several microRNA genes (22,26). Whether these
act according to Knudson’s “two hit” mechanism (27) or through the “one hit”
mechanism of haploinsufficiency (a dosage effect resulting from the loss of a
single allele of a gene) (28) is an important question. We have performed mutation
analysis of all the 44 genes mapping to the CDR in a group of patients with
the 5q− syndrome and no mutations have been identified (26). These data offer
strong support for the proposal that haploinsufficiency for one or more of the genes
mapping to the CDR may be the pathogenetic basis of the 5q− syndrome (26).
ANALYSIS IN HEMATOPOIETIC STEM CELLS FROM PATIENTS WITH

THE 5Q− SYNDROME
The identification of genes and gene pathways deregulated in leukemia using gene
expression profiling can lead to a better understanding of the molecular pathogen-
esis of the disease. We have determined the transcriptome of the CD34+ cells of
a group of patients with the 5q− syndrome using a comprehensive array platform
(26). Many genes that are significantly differentially expressed in the CD34+ cells
from patients with the 5q− syndrome compared to both normal individuals and to
patients with RA and a normal karyotype were identified in this study. Many of the
genes downregulated in the 5q− syndrome map to the long arm of chromosome 5,
consistent with a gene-dosage effect. Using hierarchical clustering, we have shown
that MDS patients with the 5q− syndrome have a distinct gene expression profile,
suggesting a common underlying pathophysiologic basis to the disease (26).
The majority of the genes mapping to the CDR of the 5q− syndrome at
5q31−q32 show a reduction in expression levels in patients with the 5q− syn-
drome, consistent with the deletion of one allele (26). Candidate genes identified
as showing haploinsufficiency in the hematopoietic stem cells of patients with
the 5q− syndrome in this study include the tumor suppressor gene SPARC and
RPS14, a component of the 40 S ribosomal subunit (26). Soren Lehmann and his
colleagues in the Koeffler laboratory at the University of California, Los Angeles,
recently demonstrated downregulation of SPARC and RPS14 (as well as several
other genes mapping to the CDR) in the mononuclear cells of patients with the
5q− syndrome when compared to mononuclear cells MDS from patients without
the del(5q) (29).
Importantly, we found that two genes mapping to the CDR, RMB22, and
CSNK1A1, showed a reduction in gene expression of ⬎50% in some patients with
Table 1 Genes Mapping to the Commonly Deleted Region of the 5q− Syndrome (22) (unpublished data)
272
Expression status
Ensembl
no. Contig Gene/description PBL CD34+
169247 AC011364.4.1.122394 SH3TC2, SH3 domain and tetratricopeptide repeats containing protein-2 − −
173210 AC091940.3.1.143702 ABLIM3, actin-binding LIM protein family member 3 + +
157510 AC012613.7.1.164550 AFAP1L1, actin filament associated protein 1-like 1 + +
164284 AC131025.1.1.179633 GRPEL2, GrpE protein homolog 2, mitochondrial precursor + +
145882 AC131025.1.1.179633 PCYOX1 L, phenylcysteine oxidase-like precursor + +
127743 AC131025.1.1.179633 IL-17␤, interleukin-17B precursor + +
113712 AC021078.5.1.145050 CSNK1A1, casein kinase I isoform alpha + +
183111 AC021078.5.1.145050 Novel, hypothetical protein − −
155846 AC022100.7.1.149603 PPARGC1B, peroxisome proliferator-activated receptor gamma coactivator 1-beta + +
132915 AC008427.9.1.175501 PDE6 A, Rod cGMP-specific 3 ,5 -cyclic phosphodiesterase subunit alpha − −
155850 AC008427.9.1.175501 SLC26A2, sulfate transporter, diastrophic dysplasia protein + +
164296 AC011406.5.1.89012 TIGD6, tigger transposable element-derived protein 6 + +
113716 AC011406.5.1.89012 SMF, SMF protein, contains HMG1 box + +
182578 AC011382.4.1.86829 CSF1R, macrophage colony-stimulating factor I receptor precursor + −
214485 AC011382.4.1.86829 Novel, no description + +
113721 AC005895.1.1.87857 PDGFR␤, beta platelet-derived growth factor receptor precursor + +
113722 AC005895.1.1.87857 CDX1, caudal-type homeobox protein-1 − −
011083 AC005895.1.1.87857 SLC6A7, sodium-dependent proline transporter + −
070808 AC011372.7.1.183330 CAMK2 A, calcium/calmodulin-dependent protein kinase type II alpha chain + +
183876 AC011372.7.1.183330 ARSI, arylsulfatase I precursor − −
070814 AC011372.7.1.183330 TCOF1, Treacle protein, Treacher Collins syndrome protein + +
019582 AC011388.7.1.73201 CD74, HLA class II histocompatibility antigen gamma chain + +
164587 AC011388.7.1.73201 RPS14, 40 S ribosomal protein S14 + +
Boultwood and Wainscoat
070614 AC008472.8.1.107733 NDST1, bifunctional heparan sulfate N-deacetylase/N-sulfotransferase I + +
171992 AC011383.6.1.93407 SYNPO, synaptopodin, actin-associated protein + +
164591 AC008453.6.1.145064 MY0Z3, myozenin-3 + +
086589 AC008453.6.1.145064 RBM22, pre-mRNA-splicing factor RNA binding motif protein 22 + +
132912 AC008450.5.1.92949 DCTN4, dynactin subunit 4 + +
181368 AC010441.6.1.157076 NM032947.3, MSTP150, small putative membrane protein + +
145908 AC010441.6.1.157076 ZNF300, zinc finger protein 300 + −
197083 AC022106.5.1.149790 Q17R26, no description − −
211445 AC0008641.7.1.176629 GPX3, glutathione peroxidase 3 precursor + +
145901 AC0008641.7.1.176629 TNIP1, Nef-associated factor 1 + +
197043 AC0008641.7.1.176629 ANX6, Annexin VI + +
198624 AC008385.7.1.151712 CCDC69, coiled-coil domain containing 69 + +
Molecular Pathogenesis of the 5q− Syndrome
196743 AC008385.7.1.151712 GM2 A, ganglioside GM2 activator precursor + +

186334 AC008385.7.1.151712 SLC36A3, solute carrier family 36 (proton/amino acid symporter), member 3 + +
186335 AC008385.7.1.151712 SLC36A2, solute carrier family 36 (proton/amino acid symporter), member 2 + +
123643 AC034205.4.1.159562 SLC36A1, proton-coupled amino acid transporter 1 + +
086570 AC011374.6.1.203961 FAT2, protocadherin FAT2 precursor (multiple epidermal growth factor-like + +
domains 1)
113140 AC011374.6.1.203961 SPARC, secreted protein acidic and rich in cysteine + +
177556 AC091982.3.1.170368 ATOX1, copper transport protein + +
145907 AC091982.3.1.170368 G3BP1, Ras GTPase-activating protein-binding protein 1 + +
273
the 5q− syndrome consistent with the downregulation of the remaining allele (26).
The RMB22 gene, encoding a highly conserved RNA-binding protein, is the most
significantly downregulated gene mapping to the CDR of the 5q− syndrome (26).
There is evidence to suggest that RBM22 plays a role in the regulation of gene
splicing and apoptosis (30). CSNK1A1, a serine/threonine kinase, plays a role in
the regulation of Hedgehog (Hh) signaling (31) and the Wnt pathway (32,33). The
markedly reduced expression levels of RMB22 and CSNK1A1 may play a role in
the molecular pathogenesis of the 5q− syndrome.
We have identified several differentially expressed genes in the 5q− syn-
drome mapping to chromosomes other than chromosome 5 including SPAG6,
WIG-1, BMI1 (all upregulated) and DPH5 (downregulated) (26). Little is known
about the function of SPAG6, but interestingly it is also markedly overexpressed in
paediatric AML (34). Wig-1, a p53-induced gene, encodes a growth inhibitory pro-
tein (35) and Bmi-1 is required for maintenance of adult self-renewing hematopoi-
etic stem cells (36). SPAG6, WIG-1 and BMI1 are upregulated in the majority of
patients with the 5q− syndrome and may play a role in the pathogenesis of this
disorder (26).
Using pathway analysis, we identified several significantly deregulated
canonical gene pathways in patients with the 5q− syndrome including the
Wnt/catenin signaling, protein ubiquitination, aminoacyl-tRNA biosynthesis, cell-
cycle regulation, and actin cytoskeleton signaling pathways (in which SPARC
plays a role) (26).
CANDIDATE GENES FOR THE 5Q− SYNDROME

We have suggested that one or more of the 43 candidate genes mapping within the
CDR of the 5q− syndrome represents the gene or genes critical to the development
of this disorder (22). There is a good evidence that haploinsufficiency of RPS14
plays a critical role in the development of the characteristic hematological
abnormalities observed in the 5q− syndrome such as diminished erythropoiesis
(37), and there is also evidence suggesting that SPARC plays a role in the
mode of action of the drug lenalidomide in the treatment of the 5q− syndrome
(38).
Candidate Gene SPARC

Lenalidomide Inhibits the Malignant Clone and Upregulates the
SPARC Gene Mapping to the Commonly Deleted Region in Patients
with the 5Q− Syndrome
The identification of the molecular targets of drug treatments in cancer can illu-
minate the molecular basis of the disease. The molecular basis of the hypere-
osinophilic syndrome, for example, was identified after this disorder was found to
be responsive to imatinib mesylate (39). We have recently investigated the in vitro
effects of the drug lenalidomide on growth, maturation, and global gene expression
in differentiating erythroblasts from MDS patients with del(5)(q31) in order to gain

insight into the molecular targets of lenalidomide (38). In this study, we showed
that lenalidomide inhibited the growth of del(5q) erythroid progenitors, while not
affecting normal cells. Using gene expression profiling, we found that the pattern
of gene expression in del(5q) erythroid progenitors was significantly altered fol-
lowing treatment with lenalidomide, with the VSIG4, PPIC, TPBG, Activin A, and
SPARC genes upregulated by more than 2-fold in all samples (38).
Upregulation of the tumor suppressor gene SPARC by lenalidomide is
intriguing since it maps within the CDR of the 5q− syndrome. The protein encoded
by SPARC has multiple known functions: it is anti-proliferative, anti-adhesive, and
anti-angiogenic (40,41), all principle effects of immunomodulatory drugs (42,43).
SPARC is considered to act as a tumor suppressor in several human cancers. The
downregulation of SPARC has been reported in several malignancies, including
primary leukemia cells and cell lines derived from AML (44). In vitro, SPARC
inhibits growth of several malignant cell lines, including AML–MLL cell lines
(44). SPARC also plays a major role in the regulation of the extra-cellular matrix
(ECM) interactions (40,45), and interestingly, we found that the ECM receptor
interaction pathway was most significantly deregulated pathway following treat-
ment of del(5q) erythroid progenitors with lenalidomide (38).
We have shown that lenalidomide inhibits growth of del(5q) erythroid pro-
genitors and we suggest that the upregulation of SPARC may underlie some of the
potent effects of lenalidomide in MDS with del(5q) (38). Given that SPARC is the
only gene mapping to the CDR of the 5q− syndrome showing significant upreg-
ulation following treatment of del(5q) erythroid progenitors with lenalidomide, it
is possible that SPARC plays a role in the pathogenesis of the 5q− syndrome.
Marked upregulation of SPARC has been recently reported in several non-
Hodgkin’s lymphoma (NHL) cell lines following treatment with lenalidomide
(46). Importantly, lenalidomide-induced upregulation of SPARC was shown to
correlate with sensitivity to the drug in NHL. These data further suggest that
SPARC plays a key role in the mode of action of lenalidomide in hematological
malignancies.
Abnormal Hematology in SPARC-Null Mice
Lehmann and colleagues have recently demonstrated that SPARC-null mice show
some hematological abnormalities (29). SPARC-null mice show significantly lower
platelet counts compared to wild-type animals and the hemoglobin, hematocrit,
and mean corpuscular volume (MCV) are lower in mice lacking SPARC, although
these differences were not statistically significant (29). Moreover, it was found
that SPARC-null mice show a significantly impaired ability to form erythroid
burst–forming units (BFU-E), supporting a role for SPARC in early erythroid
differentiation. While SPARC-null mice do not exactly replicate the phenotype
of the human 5q− syndrome, the hematologic abnormalities observed in these
animals do suggest that low expression levels of SPARC could play a role in the
5q− syndrome (29).
Candidate Gene RPS14

Haploinsufficiency for RPS14 in the 5q− Syndrome and Analogy
with the Molecular Basis of Diamond–Blackfan Anemia
The ribosomal proteins are encoded by complex gene families (47). Ribosomal
proteins act to establish ribosomal RNA (rRNA) structure during ribosome assem-
bly and to maintain this structure in functioning ribosomes (48). In the yeast Sac-
charomyces cerevisiae, RPS14 is an essential protein necessary for the assembly
of 40S ribosomal subunits (49,50). RPS14 plays a key role in ribosome biogen-
esis interacting with helix 23 of 18S rRNA (51). Following depletion of RPS14,
ribosomal proteins, and rRNA destined for 40S subunits are rapidly degraded, but
60S subunits assemble at normal rates (50).
RPS14 maps within the CDR of the 5q− syndrome (22) and we have shown
haploinsufficiency of RPS14 in the hematopoietic stem cells of patients with the
5q− syndrome (26). Intriguingly, haploinsufficiency of a closely related ribosomal
protein, RPS19, similarly required for the maturation of 40S ribosomal subunits
(52), is one of the causative genes for Diamond–Blackfan anemia (DBA) (53).
DBA is a rare pure red blood cell aplasia of childhood characterized by the
absence or decreased numbers of erythroid precursors in the bone marrow but an
otherwise normal cellularity (54). Patients with DBA have an increased incidence
of malignancy (54). It has been established that 25% of patients with DBA bear a
mutated allele of the gene encoding RPS19 (55). A wide range of mutations have
been identified in DBA patients, from missense to nonsense mutations and from
partial to complete deletion of one allele (53,56). Some missense mutations affect
both the stability and the intracellular transport of RPS19 (57). It has also been
suggested that missense mutations in RPS19 in DBA affect the capacity of the
protein to be incorporated into pre-ribosomes, thus blocking maturation of the pre-
40S particles (58). Mutations in RPS19 in DBA patients lead to haploinsufficiency
of RPS19 (55). RPS19 protein and mRNA expression analysis in patients with
RPS19-mutated DBA confirm that expression from the healthy RPS19 allele is not
sufficient to compensate for the defective allele (59). siRNA generated deficiency
of RPS19 in CD34+ cells has been shown to block erythroid development and
mimic the defects seen in DBA (60).
While DBA is the only genetic disease linked to mutation of an auto-
somal ribosomal protein gene to date, a number of other bone marrow failure
symptoms (dyskeratosis congenita, cartilage-hair hypoplasia, and Shwachman–
Diamond syndrome) involve genes encoding putative ribosome biogenesis factors
(61).
The anemia in DBA and the 5q− syndrome is due to a failure of
erythropoiesis and both diseases show haploinsufficiency for ribosomal proteins
(RPS19 and RPS14, respectively) required for the maturation of 40S ribosomal
subunits. By analogy with DBA, haploinsufficiency of RPS14 may lead to defec-
tive erythropoiesis and thus play a critical role in the pathogenesis of the 5q−
syndrome.
Gene Silencing of RPS14 in Hematopoietic Stem Cells

In the course of writing this chapter, important new data emerged from Ebert
and colleagues in the laboratory of Golub in Boston strongly supporting a role
for haploinsufficiency of RPS14 in the pathogenesis of the 5q− syndrome (37).
Ebert and colleagues employed systematic RNA interference (RNAi) technology
to determine the role of each of the genes mapping to the CDR of the 5q−
syndrome (37). Three to five unique, lentivirally expressed short hairpin RNAs
(shRNAs) targeting each of the 41 genes mapping to the CDR (22) were introduced
into normal human bone marrow CD34+ cells and the effects of each shRNA on
hematopoietic differentiation were investigated. It was found that the knock-down
of RPS14 recapitulated the phenotype of the 5q− syndrome—a block in erythroid
differentiation (leading to erythroid cell apoptosis) with relative preservation of
megakaryocyte differentiation as measured by FACS analysis was observed (37).
Critically, Ebert et al. then showed that forced expression of an RPS14 cDNA
in primary bone marrow cells from patients with the 5q− syndrome rescued the
phenotype (37). It was also found that RPS14 haploinsufficiency caused a block
in the processing of pre-rRNA and in the formation of the 40 S ribosomal subunit;
a ribosomal processing defect highly analogous to the functional defect was seen
in DBA. These compelling data highlight an important role for the RPS14 gene in
the pathogenesis of the 5q− syndrome and indicate that the 5q− syndrome may
be caused by a defect in ribosomal protein function. Ebert et al. suggest that the
5q− syndrome may therefore represent a new member of the emerging class of
ribosomal disorders.
Haploinsufficiency of RPS14 in 5q− Syndrome Causes Marked

Deregulation of Ribosomal- and Translation-Related Genes
In RPS19-deficient DBA, the impaired 40 S ribosomal subunit biogenesis sug-
gests impaired translation as the mechanism that causes anemia in DBA (55). In
accord with this hypothesis, global gene expression profiles in the hematopoietic
progenitor cells of patients with DBA are characterized by downregulation of mul-
tiple ribosomal genes, as well as several genes which are required for translation
initiation and elongation (62). Intriguingly, we have previously shown that MDS
patients with the del(5q) show deregulation of genes involved in translation initia-
tion when compared to MDS without the del(5q) (63). By analogy with DBA, we
have recently performed a detailed investigation of the expression levels of a large
group of ribosomal- and translation-related genes in the CD34+ cells of patients
with 5q− syndrome (64). The hypothesis under consideration is that DBA and the
5q− syndrome share a related molecular basis in that they are both disorders of
defective ribosomal biogenesis.
The expression profiles of a large group of ribosomal- and translation-related
genes were determined in the CD34+ cells of 15 MDS patients with 5q− syn-
drome, 18 MDS patients with RA and a normal karyotype, and 17 healthy controls.
A total of 55 of these 579 ribosomal- and translation-related probe sets were
significantly differentially expressed in the three-way comparison of the two

patient and control groups with approximately 90% of the significantly differ-
entially expressed genes showing lower expression levels in the 5q− syndrome
patient group. Using hierarchical clustering, patients with the 5q− syndrome could
be separated both from other patients with RA and healthy controls solely on the
basis of the deregulated expression of ribosomal- and translation-related genes
(64). We found that the most significantly deregulated gene ontology category
in patients with the 5q− syndrome compared to patients with RA and a normal
karyotype and healthy controls was protein biosynthesis, with protein metabolism
and translation also significantly deregulated (64).
We have shown that patients with the 5q− syndrome have a striking defect
in the expression of genes involved in ribosome biogenesis and in the control of
translation. By analogy with DBA, we suggest that the deregulation that we have
observed in ribosomal gene expression and translation-related gene expression in
the HSC of patients with the 5q− syndrome are secondary to RPS14 haploinsuf-
ficiency. These abnormalities may lead to impairment of ribosome biogenesis and
subsequent reduction of protein translation capacity. These data strongly support
the proposal that the 5q− syndrome represents a disorder of aberrant ribosome
biogenesis (64).
OTHER MDS AND AML WITH THE DEL(5Q)

Del(5q) is the most commonly reported deletion in de novo MDS and is found
in 10% to 15% of all patients (see chap. 3). The relationship between the 5q−
syndrome and the other myeloid malignancies with the del(5q) is a complex
question. When the del(5q) occurs as the sole karyotypic abnormality in RA (5q−
syndrome), it carries a good prognosis, otherwise it is among the worst prognostic
indicators (7,65). Moreover, the del(5q) in AML, particularly secondary AML,
invariably occurs together with other karyotypic abnormalities and frequently
as part of a complex karyotype (20,66). Interestingly, mutations with loss of
function of p53 are significantly associated with the del(5q) in t-MDS and t-AML
after previous treatment with alkylating agents and are associated with genetic
instability (67).
The 1 to 1.5 Mb CDR at 5q31, identified in AML and the more aggressive
forms of MDS by Michelle Le Beau in Chicago and her co-workers, is distinct from
the CDR of the 5q− syndrome (68). The AML CDR at 5q31 is flanked by D5S479
and D5S500 and contains 18 genes including the candidate gene EGR1 (68). No
inactivating mutations have been described in any candidate genes mapping to
this interval (69). However, it has been shown that Egr1-deficient mice treated
with a DNA alkylating agent to induce secondary cooperating mutations develop
immature T-cell lymphomas or a MPD at increased rates and with shorter latencies
than that of wild-type littermates (70). Low or absent expression of CTNNA1, also
mapping to 5q31, has been recently reported in a proportion of MDS/AML patients
with a del(5q), and it has been suggested that markedly reduced expression levels of
CTNNA1 may play an important role in MDS/AML with the del(5q) (71). However,
it is important to note that in the 5q− syndrome, patients show expression levels
consistent with the loss of one CTNNA1 allele only (14,63).
The identification of more than one CDR of the del(5q) in myeloid malig-
nancies suggests the existence of more than one pathogenetically relevant gene.
This might be expected given the very different clinical features and prognosis
observed in patients with the 5q syndrome and patients with the more aggressive
forms of MDS or AML and a del(5q). However, the del(5q) in the 5q− syndrome
is cytogenetically indistinguishable from the del(5q) found in AML, and it should
be recognized that in the majority of patients with the del(5q) and a myeloid
malignancy, both CDRs mapping to distal 5q will be deleted.
TREATMENT
The cornerstone of treatment of the 5q− syndrome has been supportive care with
most patients becoming transfusion dependent and hence needing iron chelation
therapy. Over recent years, a major advance in the treatment of the 5q− syn-
drome has been the introduction of lenalidomide into the drug armamentarium.
Lenalidomide is a potent analogue of thalidomide that was designed to reduce the
risk of teratogenicity and neurotoxicity. The earliest clinical studies investigated
lenalidomide in multiple myeloma, with promising results in terms of its ability
to overcome drug resistance and its satisfactory tolerability (72). The first study
of lenalidomide in MDS found that it had a remarkable efficacy in patients with
the 5q− syndrome (83% of such patients responded as compared to 57% among
those MDS patients with a normal karyotype and 12% among those with other
karyotypes) (73). A subsequent paper investigated this response in more detail: of
148 patients with MDS and a 5q31 deletion, 112 had a reduced need for transfu-
sion and 99 became transfusion independent (74). Among 85 patients who could
be evaluated, 62 had cytogenetic improvement and 38 had a complete cytogenetic
remission. There was complete resolution of cytologic abnormalities in 38 of 106
patients whose serial bone marrow samples could be evaluated. Lenalidomide has
been approved by the United States Food and Drug Administration for the treat-
ment of International Prognostic Scoring System low or intermediate-1 risk MDS
patients with chromosome 5q deletion, with or without additional cytogenetic
abnormalities.
The mechanism of action of lenalidomide remains uncertain although it
is known to have effects on T-cell co-stimulation, angiogenesis inhibition and
modulation of apoptosis (75). We have provided evidence in favour of the SPARC
gene as a candidate target gene for lenalidomide (38). A better understanding of
the action of lenalidomide would facilitate the development of further thalidomide
derivatives for the treatment of MDS.
FUTURE PERSPECTIVES
There are two major aspects of the 5q− syndrome on which substantive progress
has been made over recent years—the molecular basis of the disorder and its
successful treatment with lenalidomide. Our research has focused on the molec-
ular genetic basis for the 5q− syndrome and on the target genes involved in the
hematological response of patients with the 5q− syndrome treated with lenalido-
mide. We have established the critical region of gene loss on 5q and have suggested
that haploinsuffiency for one or more of the genes mapping to this interval is the
pathogenetic basis of the 5q− syndrome. We have suggested that the ribosomal
gene RPS14 represents a good candidate gene for the 5q− syndrome, based on
analogies to DBA and recent gene expression profiling data. The significance of
RPS14 haploinsuffiency in the 5q− syndrome has recently been highlighted by
the functional studies of Ebert et al. (37). However, it remains possible that hap-
loinsuffiency of other genes mapping to the CDR also plays a role in the molecular
pathogenesis of the 5q− syndrome. It may be speculated that haploinsufficiency of
RPS14 causes the characteristic hematological abnormalities observed in the 5q−
syndrome, while haploinsufficiency of tumor suppressor gene(s) such as SPARC,
leads to defects in cell growth. Animal models could be illuminating in this regard,
and we are generating mice with chromosomal deletions which mimic those seen in
the 5q− syndrome (A McKenzie et al., unpublished data). It is important to under-
stand the mechanism of action of lenalidomide in the 5q− syndrome. Our finding
that SPARC is upregulated following treatment of del(5q) cells with lenalidomide
provides one new direction for such studies. It is now probable that rapid progress
will be made in all these fundamental aspects of the biology of the 5q− syndrome.
REFERENCES
1. Van den Berghe H, Cassiman JJ, David G, et al. Distinct haematological disorder with
deletion of long arm of no. 5 chromosome. Nature 1974; 251(5474):437–438.
2. Sokal G, Michaux JL, Van Den Berghe H, et al. A new hematologic syndrome with a
distinct karyotype: The 5q− chromosome. Blood 1975; 46(4):519–533.
3. Van den Berghe H, Vermaelen K, Mecucci C, et al. The 5q− anomaly. Cancer Genet
Cytogenet 1985; 17(3):189–255.
4. Lewis S, Oscier D, Boultwood J, et al. Hematological features of patients with
myelodysplastic syndromes associated with a chromosome 5q deletion. Am J Hematol
1995; 49(3):194–200.
5. Teerenhovi L. Specificity of haematological indicators for ‘5q− syndrome’ in patients
with myelodysplastic syndromes. Eur J Haematol 1987; 39(4):326–330.
6. Santana-Davila R, Holtan SG, Dewald GW, et al. Chromosome 5q deletion: Spe-
cific diagnoses and cytogenetic details among 358 consecutive cases from a single
institution. Leuk Res 2008; 32(3):407–11.
7. Giagounidis AA, Germing U, Haase S, et al. Clinical, morphological, cytogenetic, and
prognostic features of patients with myelodysplastic syndromes and del(5q) including
band q31. Leukemia 2004; 18(1):113–119.
8. Pedersen B. 5q−: Does longer survival of female patients explain the preponderance.
Anticancer Res 1997; 17(5 A):3281–3285.
9. Mathew P, Tefferi A, Dewald GW, et al. The 5q− syndrome: A single-institution
study of 43 consecutive patients. Blood 1993; 81(4):1040–1045.
10. Nilsson L, Astrand-Grundstrom I, Arvidsson I, et al. Isolation and characterization

of hematopoietic progenitor/stem cells in 5q-deleted myelodysplastic syndromes:
Evidence for involvement at the hematopoietic stem cell level. Blood 2000;
96(6):2012–2021.
11. Jaju RJ, Jones M, Boultwood J, et al. Combined immunophenotyping and FISH
identifies the involvement of B-cells in 5q− syndrome. Genes Chromosomes Cancer
2000; 29(3):276–280.
12. Godon C, Talmant P, Garand R, et al. Deletion of 5q31 is observed in megakaryocytic
cells in patients with myelodysplastic syndromes and a del(5q), including the 5q−
syndrome. Genes Chromosomes Cancer 2000; 29(4):350–352.
13. Bigoni R, Cuneo A, Milani R, et al. Multilineage involvement in the 5q− syndrome: A
fluorescent in situ hybridization study on bone marrow smears. Haematologica 2001;
86(4):375–381.
14. Nilsson L, Eden P, Olsson E, et al. The molecular signature of MDS stem cells sup-
ports a stem-cell origin of 5q myelodysplastic syndromes. Blood 2007; 110(8):3005–
3014.
15. Yunis JJ, Rydell RE, Oken MM, et al. Refined chromosome analysis as an independent
prognostic indicator in de novo myelodysplastic syndromes. Blood 1986; 67(6):1721–
1730.
16. Crescenzi B, La Starza R, Romoli S, et al. Submicroscopic deletions in 5q− associated
malignancies. Haematologica 2004; 89(3):281–285.
17. Fidler C, Watkins F, Bowen DT, et al. NRAS, FLT3 and TP53 mutations in patients
with myelodysplastic syndrome and a del(5q). Haematologica 2004; 89(7):865–866.
18. Li W, Fidler C, Giagounidis A, et al. Genome-wide analysis of copy number changes
and loss of heterozygosity in myelodysplastic syndrome with del(5q) using high-
density SNP arrays. Haematologica 2008; 93(7):994–1000.
19. Giagounidis AA, Germing U, Wainscoat JS, et al. The 5q− syndrome. Hematology
2004; 9(4):271–277.
20. Boultwood J, Lewis S, and Wainscoat JS. The 5q− syndrome. Blood 1994;
84(10):3253–3260.
21. Boultwood J, Fidler C, Lewis S, et al. Molecular mapping of uncharacteristically
small 5q deletions in two patients with the 5q− syndrome: Delineation of the critical
region on 5q and identification of a 5q− breakpoint. Genomics 1994; 19(3):425–432.
22. Boultwood J, Fidler C, Strickson AJ, et al. Narrowing and genomic annotation of the
commonly deleted region of the 5q− syndrome. Blood 2002; 99(12):4638–4641.
23. Boultwood J, Fidler C, Strickson AJ, et al. Transcription mapping of the 5q− syndrome
critical region: Cloning of two novel genes and sequencing, expression, and mapping
of a further six novel cDNAs. Genomics 2000; 66(1):26–34.
24. Boultwood J, Strickson AJ, Jabs EW, et al. Physical mapping of the human ATX1
homologue (HAH1) to the critical region of the 5q− syndrome within 5q32, and
immediately adjacent to the SPARC gene. Hum Genet 2000; 106(1):127–129.
25. Boultwood J, Fidler C, Soularue P, et al. Novel genes mapping to the critical region
of the 5q− syndrome. Genomics 1997; 45(1):88–96.
in patients with the 5q− syndrome. Br J Haematol 2007; 139(4):578–589.
27. Knudson A. Alfred Knudson and his two-hit hypothesis. (Interview by Ezzie Hutchin-
son). Lancet Oncol 2001; 2(10):642–645.
28. Largaespada DA. Haploinsufficiency for tumor suppression: The hazards of being
single and living a long time. J Exp Med 2001; 193(4):F15–F18.
29. Lehmann S, O’Kelly J, Raynaud S, et al. Common deleted genes in the 5q− syndrome:
Thrombocytopenia and reduced erythroid colony formation in SPARC null mice.
Leukemia 2007; 21(9):1931–1936.
30. Montaville P, Dai Y, Cheung CY, et al. Nuclear translocation of the calcium-binding
protein ALG-2 induced by the RNA-binding protein RBM22. Biochim Biophys Acta
2006; 1763(11):1335–1343.
31. Jia J, Tong C, Wang B, et al. Hedgehog signalling activity of Smoothened
requires phosphorylation by protein kinase A and casein kinase I. Nature 2004;
432(7020):1045–1050.
32. Dejmek J, Safholm A, Kamp Nielsen C,et al. Wnt-5 a/Ca2+-induced NFAT activ-
ity is counteracted by Wnt-5 a/Yes-Cdc42-casein kinase 1alpha signaling in human
mammary epithelial cells. Mol Cell Biol 2006; 26(16):6024–6036.
33. Hammerlein A, Weiske J, and Huber O. A second protein kinase CK1-mediated step
negatively regulates Wnt signalling by disrupting the lymphocyte enhancer factor-
1/beta-catenin complex. Cell Mol Life Sci 2005; 62(5):606–618.
34. Steinbach D, Schramm A, Eggert A, et al. Identification of a set of seven genes for
the monitoring of minimal residual disease in pediatric acute myeloid leukemia. Clin
Cancer Res 2006; 12(8):2434–2441.
35. Hellborg F, Qian W, Mendez-Vidal C, et al. Human wig-1, a p53 target gene that
encodes a growth inhibitory zinc finger protein. Oncogene 2001; 20(39):5466–5474.
36. Park IK, Qian D, Kiel M, et al. Bmi-1 is required for maintenance of adult self-
renewing haematopoietic stem cells. Nature 2003; 423(6937):302–305.
37. Ebert BL, Pretz J, Bosco J, et al. Identification of RPS14 as a 5q− syndrome gene by
RNA interference screen. Nature 2008; 451(7176):335–9.
5q− syndrome patients. Proc Natl Acad Sci U S A 2007; 104(27):11406–11411.
39. Cools J, DeAngelo DJ, Gotlib J, et al. A tyrosine kinase created by fusion of the
PDGFRA and FIP1L1 genes as a therapeutic target of imatinib in idiopathic hypere-
osinophilic syndrome. N Engl J Med 2003; 348(13):1201–1214.
40. Framson PE, Sage EH. SPARC and tumor growth: Where the seed meets the soil? J
Cell Biochem 2004; 92(4):679–690.
41. Bradshaw AD, Sage EH. SPARC, a matricellular protein that functions in cellular
differentiation and tissue response to injury. J Clin Invest 2001; 107(9):1049–1054.
42. Bartlett JB, Dredge K, Dalgleish AG. The evolution of thalidomide and its IMiD
derivatives as anticancer agents. Nat Rev Cancer 2004; 4(4):314–322.
43. Settles B, Stevenson A, Wilson K, et al. Down-regulation of cell adhesion molecules
LFA-1 and ICAM-1 after in vitro treatment with the anti-TNF-alpha agent thalido-
mide. Cell Mol Biol (Noisy-le-grand) 2001; 47(7):1105–1114.
44. DiMartino JF, Lacayo NJ, Varadi M, et al. Low or absent SPARC expression in acute
myeloid leukemia with MLL rearrangements is associated with sensitivity to growth
inhibition by exogenous SPARC protein. Leukemia 2006; 20(3):426–432.
45. Barker TH, Baneyx G, Cardo-Vila M, et al. SPARC regulates extracellular matrix
organization through its modulation of integrin-linked kinase activity. J Biol Chem
2005; 280(43):36483–36493.
46. Zang LH, Schafer P, Muller G, et al. Lenalidomide displays direct anti-non-Hodgkin’s
lymphoma (NHL) cell activity in association with enhanced SPARC expression but
independent of its ability to strongly inhibit NHL cell VEGF production in vitro.
Blood 2007; 110:3473 Abstract.
47. Chen IT, Dixit A, Rhoads DD, et al. Homologous ribosomal proteins in bacteria,
yeast, and humans. Proc Natl Acad Sci U S A 1986; 83(18):6907–6911.
48. Antunez de Mayolo P, Woolford JL Jr. Interactions of yeast ribosomal protein rpS14
with RNA. J Mol Biol 2003; 333(4):697–709.
49. Larkin JC, Thompson JR, Woolford JL Jr. Structure and expression of the Saccha-
romyces cerevisiae CRY1 gene: A highly conserved ribosomal protein gene. Mol Cell
Biol 1987; 7(5):1764–1775.
50. Moritz M, Paulovich AG, Tsay YF, et al. Depletion of yeast ribosomal proteins
L16 or rp59 disrupts ribosome assembly. J Cell Biol 1990; 111(6 Pt 1):2261–
2274.
51. Fewell SW, Woolford JL Jr. Ribosomal protein S14 of Saccharomyces cerevisiae
regulates its expression by binding to RPS14B pre-mRNA and to 18 S rRNA. Mol
Cell Biol 1999; 19(1):826–834.
52. Flygare J, Aspesi A, Bailey JC, et al. Human RPS19, the gene mutated in Diamond–
Blackfan anemia, encodes a ribosomal protein required for the maturation of 40 S
ribosomal subunits. Blood 2007; 109(3):980–986.
53. Draptchinskaia N, Gustavsson P, Andersson B, et al. The gene encoding ribosomal
protein S19 is mutated in Diamond–Blackfan anaemia. Nat Genet 1999; 21(2):169–
175.
54. Lipton JM, Atsidaftos E, Zyskind I, et al. Improving clinical care and elucidating
the pathophysiology of Diamond Blackfan anemia: An update from the Diamond
Blackfan anemia registry. Pediatr Blood Cancer 2006; 46(5):558–564.
55. Flygare J, Karlsson S. Diamond–Blackfan anemia: Erythropoiesis lost in translation.
Blood 2007; 109(8):3152–3154.
56. Willig TN, Draptchinskaia N, Dianzani I, et al. Mutations in ribosomal protein S19
gene and Diamond Blackfan anemia: Wide variations in phenotypic expression. Blood
1999; 94(12):4294–4306.
57. Da Costa L, Tchernia G, Gascard P, et al. Nucleolar localization of RPS19 protein in
normal cells and mislocalization due to mutations in the nucleolar localization signals
in 2 Diamond–Blackfan anemia patients: Potential insights into pathophysiology.
Blood 2003; 101(12):5039–5045.
58. Gregory LA, Aguissa-Toure AH, Pinaud N, et al. Molecular basis of Diamond–
Blackfan anemia: Structure and function analysis of RPS19. Nucleic Acids Res 2007;
35(17):5913–5921.
59. Gazda HT, Zhong R, Long L, et al. RNA and protein evidence for haplo-insufficiency
in Diamond–Blackfan anaemia patients with RPS19 mutations. Br J Haematol 2004;
127(1):105–113.
60. Ebert BL, Lee MM, Pretz JL, et al. An RNA interference model of RPS19 deficiency
in Diamond–Blackfan anemia recapitulates defective hematopoiesis and rescue by
dexamethasone: Identification of dexamethasone-responsive genes by microarray.
Blood 2005; 105(12):4620–4626.
61. Liu JM, Ellis SR. Ribosomes and marrow failure: Coincidental association or molec-
ular paradigm? Blood 2006; 107(12):4583–4588.
62. Gazda HT, Kho AT, Sanoudou D, et al. Defective ribosomal protein gene expression
alters transcription, translation, apoptosis, and oncogenic pathways in Diamond–
Blackfan anemia. Stem Cells 2006; 24(9):2034–2044.
63. Pellagatti A, Cazzola M, Giagounidis A, et al. Gene expression profiles of CD34+
correlation to FAB subtype and karyotype. Blood 2006; 108(1):337–345.
64. Pellagatti A, Hellström-Lindberg E, Giagounidis A, et al. Haploinsufficiency of
RPS14 in 5q− syndrome is associated with deregulation of ribosomal- and translation-
related genes. Br J Haematol 2008; 142(1):57–64.
65. Nimer SD. Clinical management of myelodysplastic syndromes with interstitial dele-
tion of chromosome 5q. J Clin Oncol 2006; 24(16):2576–2582.
66. Dann EJ, Rowe JM. Biology and therapy of secondary leukaemias. Best Pract Res
Clin Haematol 2001; 14(1):119–137.
67. Christiansen DH, Andersen MK, Pedersen-Bjergaard J. Mutations with loss of het-
erozygosity of p53 are common in therapy-related myelodysplasia and acute myeloid
leukemia after exposure to alkylating agents and significantly associated with dele-
tion or loss of 5q, a complex karyotype, and a poor prognosis. J Clin Oncol 2001;
19(5):1405–1413.
68. Lai F, Godley LA, Joslin J, et al. Transcript map and comparative analysis of the
1.5-Mb commonly deleted segment of human 5q31 in malignant myeloid diseases
with a del(5q). Genomics 2001; 71(2):235–245.
69. Shannon KM, Le Beau MM, Largaespada DA, et al. Modeling myeloid leukemia
tumor suppressor gene inactivation in the mouse. Semin Cancer Biol 2001; 11(3):191–
200.
110(2):719–726.
71. Liu TX, Becker MW, Jelinek J, et al. Chromosome 5q deletion and epigenetic suppres-
sion of the gene encoding alpha-catenin (CTNNA1) in myeloid cell transformation.
Nat Med 2007; 13(1):78–83.
72. Richardson PG, Schlossman RL, Weller E, et al. Immunomodulatory drug CC-5013
overcomes drug resistance and is well tolerated in patients with relapsed multiple
myeloma. Blood 2002; 100(9):3063–3067.
N Engl J Med 2005; 352(6):549–557.
with chromosome 5q deletion. N Engl J Med 2006; 355(14):1456–1465.
75. Ortega J, List A. Immunomodulatory drugs in the treatment of myelodysplastic syn-
dromes. Curr Opin Oncol 2007; 19(6):656–9.
13
Chronic Myelomonocytic Leukemia and
Myelodysplastic Syndrome/
Myeloproliferative Overlap Syndromes
Phuong L. Nguyen and Curtis A. Hanson

Department of Laboratory Medicine and Pathology, Mayo Clinic, Rochester,
Minnesota, U.S.A.
INTRODUCTION
Practicing hematologists and hematopathologists have long recognized that
patients with myelodysplastic syndromes (MDS) typically present with cytope-
nia(s), and those with chronic myeloproliferative disorders (CMPD) typically
present with elevated blood counts and organomegaly, but there are also patients
who have clinical, laboratory, and/or pathologic features of both MDS and CMPD,
and those patients are not easily assigned to either category. Thus, prior to the pub-
lication of the World Health Organization (WHO) classification of hematopoietic
neoplasms (see chaps. 1 and 9), some patients with left-shifted neutrophilia and
absolute monocytosis and morphologic features of dysgranulopoiesis were clas-
sified as chronic myelomonocytic leukemia (CMML) under the family of MDS,
whereas others who also had left-shifted neutrophilia and morphologic features of
dysgranulopoiesis but less monocytosis were classified as atypical chronic myeloid
leukemia (aCML) under the CMPD umbrella.
Given the diagnostic confusion about these misfit conditions, it is difficult
to estimate the exact incidence of overlap myelodysplastic/myeloproliferative
(MDS/MPD) disorders. For example, in a retrospective study of 566 consecutive
patients who were originally given a diagnosis of MDS, Radana Neuwirtová in
Prague and colleagues found that 19 (3%) of such patients had presented with both
285
286 Nguyen and Hanson
morphologic features of MDS and laboratory features of a CMPD, defined in that

study as a platelet count ⬎400 × 109 /L and/or a leukocyte count ⬎9.5 × 109 /L (1).
In a retrospective review of 271 patients with MDS who had received decitabine
treatment in the setting of a phase 3 or phase 2 clinical trial, Pierre Wijermans in
the Netherlands and his colleagues identified 31 patients (11%) with CMML (2).
This reported frequency of CMML is similar to the 13% frequency found among
2050 patients in the Düsseldorf MDS Registry (3). Conversely, in their review of
76 patients at the M. D. Anderson Cancer Center with Philadelphia chromosome–
negative and BCR/ABL-negative chronic myeloid disorders, Francesco Onida
in Milan and his colleagues reported 36 patients (47%) who had an absolute
monocytosis ⬎1.0 × 109 /L, possibly representing patients with CMML (4).
In the current (2001) WHO classification of hematopoietic neoplasms,
this overlap MDS/MPD category includes the following: CMML; juvenile
myelomonocytic leukemia (JMML); aCML; and MDS/MPD, unclassifiable (5).
JMML is discussed in chapter 14 with other pediatric MDSs. This chapter discusses
CMML, atypical CML, and unclassifiable MDS/MPD, as well as the provisional
WHO entity of “refractory anemia with ringed sideroblasts associated with marked
thrombocytosis” (RARS-T). Given our improving understanding of the molecular
genetics of neoplastic myeloid disorders, as well as the excitement that discov-
eries such as JAK2 V617F in CMPD have engendered, molecular analysis may
help us understand better the defining biologic characteristics of CMPD versus
MDS in the near future. Therefore, the authors anticipate that this area of overlap
MDS/CMPD disorders is likely to undergo critical redefinition in the near future.
CHRONIC MYELOMONOCYTIC LEUKEMIA (CMML)

Diagnosis
Definition
The WHO diagnostic criteria for CMML include: (a) peripheral blood monocytosis
(absolute monocyte count ⬎1 × 109 /L), (b) less than 20% undifferentiated blasts
in the blood and bone marrow, (c) absence of the BCR/ABL fusion gene, and (d)
morphologic features of dysplasia in one or more hematopoietic lineages (6). It is
important to exclude nonclonal causes of monocytosis. In the absence of definitive
morphological features of myelodysplasia, the WHO recommends that there be
cytogenetic evidence of an acquired clonal abnormality, or that the monocytosis
be persistent for at least 3 months without any evidence of a reactive cause. It
should be noted that by these criteria, no distinction is made between cases with a
leukocytosis [e.g., total white blood count (WBC) ⬎13 × 109 /L] and those cases
with a WBC of 13 × 109 /L or less (vide infra).
Clinical Features
The median age of patients with CMML in most of the reported series is between
65 and 75 years (range, 30 to ⬎90 years) and there is a slight male predominance
Chronic Myelomonocytic Leukemia and Myelodysplastic Syndrome 287
(M:F, 1.5–2.5:1). Splenomegaly is described in 30% to 50% of patients (2); when

present, it is usually associated with leukocytosis (7,8). At this time, there is
no known environmental or genetic predisposition to CMML. There have been
sporadic case reports of CMML arising as a consequence of prior cytotoxic therapy,
but no consistent pattern of exposure has been identified.
Laboratory and Pathologic Findings

Laboratory evaluation in CMML typically reveals anemia and thrombocytopenia,
while a leukocytosis (⬎13 × 109 /L) has been reported in 30% to 60% of patients
with CMML.
In the blood, the monocytosis (median 2 × 109 /L to 3 × 109 /L) may include
cytologically atypical forms with unusually fine nuclear chromatin or abnormal
nuclear lobulations. While not formally a diagnostic requirement, in practice, a
marrow monocytosis is also typically present (monocytes and precursors represent
a median of 10% marrow cells in CMML cases). Since morphologic determination
of monocytic differentiation can be difficult in the face of cellular immaturity or
cytologic atypia, staining for ␣-naphthyl butyrate esterase by cytochemistry can
help establish the presence of blood and marrow monocytosis (Fig. 1 see colour
insert).
Morphologic features of dysgranulopoiesis are often seen in CMML, while
dysmegakaryopoietic and dyserythropoietic features are more variable. The bone
marrow is typically hypercellular with increased granulopoiesis and monocytes;
megakaryocytes may be normal, increased, or decreased, while erythropoiesis is
usually decreased. In some cases, nodules of cells staining for CD123 (interleukin-
3 receptor alpha chain) and Tcl-1 (T cell leukemia/lymphoma 1) can be identified in
bone marrow biopsy sections and correspond to nodules of plasmacytoid dendritic
cells (formerly known as plasmacytoid monocytes). This finding has been reported
in CMML but not in CML or aCML (9), although it was present in only 15% to
40% of the cases.
Blood and marrow blasts number are less than 20% by definition. In our
experience, because immature monocytic precursors including monoblasts and
promonocytes tend to be negative for CD34, immunophenotypic analysis of CD34
expression by paraffin-section immunohistochemistry or by flow cytometry may
underestimate the blast population.
Genetics
Chromosomal abnormalities occur in approximately 35% patients with CMML.
In a series of 205 patients at a tertiary care center, monosomy 7 and trisomy 8
were seen in 7.8% and 6.3% of patients, respectively (10); approximately 6%
of patients had a complex karyotype with three or more abnormalities. Other
reported cytogenetic abnormalities included −5, −7, −21, and 20q− (8,10).
Point mutations involving N- or K-RAS oncogenes were described in approxi-
mately 40% patients with CMML (10), with approximately two-thirds of the muta-
tions involving codon 12 of N-RAS gene and the remaining one-third involving
(A) (B)
(C) (D)
Figure 1 Chronic myelomonocytic leukemia. (A) Peripheral blood smear showing a

leukocytosis and a spectrum of monocyte maturation. (B) Hypercellular bone marrow
aspirate with a marked granulocytic and monocytic expansion. (C) A combined chloroac-
etate esterase/butyrate esterase stain showing: chloroacetate esterase positive granulocytic
precursors (blueish stain), butyrate esterase positive monocytes (reddish-brown stain),
and combined chloroacetate/butyrate esterase positive hybrid myelomonocytic cells (both
blueish and reddish-brown stains). These latter cells are commonly identified in CMML,
thus contributing to the difficulty in morphologic cell identification. (D) A packed bone
marrow biopsy with a granulocytic and monocytic hyperplasia. There is monocytic nodule
present (outlined by arrows). Source: From Ref. 51. (see color insert)
K-RAS (11). In contrast, mutations involving the JAK2 gene are uncommon among
patients with CMML (12).
Rare patients with the laboratory and pathologic features of CMML have the
additional finding of eosinophilia, defined as ⬎1.5 × 109 eosinophils/L. When this
finding is associated with a fusion gene involving platelet-derived growth factor-␤

receptor (PDGF␤(R), such as in the t(5;12)(q33;p13) chromosomal abnormality
with the TEL(ETV6) gene as the partner or in the t(5;10)(q33;q22) chromosomal
abnormality with the H4/D10S170 gene as the partner, such patients have been
reported to respond to imatinib mesylate therapy (13–15).
Prognosis
As a group, patients with CMML have a widely variable reported natural history,
with estimated median survival ranging from 7 months (range 0–84 months)
in one retrospective study of 60 patients (16) to 60 months in another study
(17). Transformation to acute myeloid leukemia (AML) occurs in approximately
20% to 60% of patients with CMML (10,18–21), with the median time to AML
transformation estimated at approximately 13 months (range 1–125 months) (18).
The risk of AML transformation appears to be associated with the proportion
of medullary blast count at diagnosis and with duration of disease, with 14% of
patients with CMML and ⬍10% medullary blasts developing AML after 2 years
versus 24% of patients with 10% to 19% medullary blasts and/or 5% to 19%
circulating blasts; after 5 years, 18% and 63% of patients in these two groups had
developed AML, respectively (20). The AML seen in patients with CMML who
have transformed typically has myelomonocytic or monocytic differentiation.
In an effort to identify meaningful prognostic and biologic predictors of
patient outcome, there have been several attempts at subclassification and prog-
nostication, as described below.
Myeloproliferative Versus Myelodysplastic-Type CMML

Previously, the French-American-British (FAB) classification schema (see chap.
1) recognized two types of CMML, one thought to be closer to MDS (CMML-
MDS) in which the total leukocyte count is 13 × 109 /L or lower, and the other
more akin to CMPD (CMML-MPD) where the total leukocyte count exceeds
13 × 109 /L (22). The latter has been estimated to constitute between 30% and
60% of all patients with CMML. However, over time, no consistent clear-cut
differences between these two types of CMML have been demonstrated (i.e., there
is no difference in overall median survival or leukemic progression). Whereas some
investigators have reported a shorter median survival among patients with CMML
and a leukocytosis ⬎13 × 109 /L compared to those with a leukocyte count ⬍13 ×
109 /L (8,16,19), others have found no statistically significant differences (6,10). A
trend towards a shortened survival was noted among the CMML-MPD group after
16 months in a 213-patient retrospective study from the M. D. Anderson Cancer
Center (10). None of these studies showed a statistically significant difference
between CMML-MDS and CMML-MPD with respect to risk of progression to
AML. Moreover, some patients with features of CMML-MDS type at initial
presentation have been shown to develop the CMML-MPD picture over time. In a
single-institutional study, the presence of RAS mutations appeared to be associated
with the myeloproliferative type of CMML with higher leukocyte and monocyte
counts (11).
CMML-1 Versus CMML-2
In the current WHO classification of myeloid neoplasms, CMML is subclassified
into CMML-1 and CMML-2, with fewer than 5% blasts in the blood and less
than 10% blasts in the bone marrow in CMML-1, against 5% to 19% circulating
blasts and/or 10% to 19% blasts in the bone marrow in CMML-2 (6). For the
purpose of enumerating blasts in CMML, blasts include myeloblasts, monoblasts,
and promonocytes. In addition to classifying a process as CMML-2 according to
the blast proportion in the blood and/or marrow, the WHO also recommends that
the diagnosis of CMML-2 be made when Auer rods are present, regardless of the
blood or bone marrow blast proportion. However, we are not aware of studies to
date that have demonstrated biologic and clinical equivalency between patients
with CMML-2 based on blast proportions and patients with CMML-2 based solely
on the presence of Auer rods.
Prognostic Scoring Systems
Several groups have attempted to improve prognostication in CMML by proposing
scoring systems predictive of patient outcomes. These scoring systems have been
based on various combinations of number and severity of cytopenia(s), marrow
blast proportion, serum lactate dehydrogenase (LDH) levels, absolute lymphocyte
count, karyotype, presence or absence of circulating immature granulocytic pre-
cursors, and others (Table 1) (10,16,18,23). The M. D. Anderson Cancer Center
recently updated their experience with a prognostic score for CMML in which the
median follow-up of the initial “learning” patient cohort (n = 213) was extended
to 25.4 months (range 0–156 months) compared to the original median follow-up
of 10 months (range 0–154 months), and in which a “validation” cohort of 250
CMML patients was added. Beran and colleagues reported that, based on the pres-
ence of a hemoglobin value ⬍120 g/L, absolute lymphocyte count ⬎2.5 × 109 /L,
circulating immature granulocytic cells ⬎0%, and LDH ⬎700 U/L, patients with
CMML could be stratified into four risk groups, with a median survival of 26.3,
15.7, 9.6, and 4.3 months, according to the presence of 1, 2, 3, or all 4 adverse
variables, respectively. It was noted that in this update, LDH value replaced bone
marrow blast proportion as a prognostic variable (24).
In contrast, in their more recent published analyses of over 300 patients with
CMML in the Düsseldorf MDS Registry, Ulrich Germing and his colleagues re-
affirmed their earlier observation of the independent prognostic value of medullary
blast counts, with 63% of their patients with CMML-2 (with 10% or greater marrow
blasts) developing AML after 5 years compared to 18% of patients with CMML-
1 (p = 0.001) (3,20). The median survival of patients with CMML-2 was also
shorter compared to that of patients with CMML-1 (15 months vs. 20 months;
p = 0.005). In both updates, the adverse prognostic value of elevated LDH levels,
low hemoglobin values (⬍120 g/L in the MDAPS and ⬍100 g/L in the Düsseldorf
Table 1 Prognostic Scoring Systems for Chronic Myelomonocytic Leukemia
Bournemouth score (23)a Spanish CMML score (16) MDAPS (10) Dusseldorf score (18)
Factors
Gender – – – Male
Hemoglobin (g/L) ⬍100 – ⬍120 ⬍120
Leukocyte count – ⬎10 × 109 /L – –
ANC 2.5 × 109 /L vs. ⬎26 × 109 /L – – –
PB IMC (%) – – ⬎0 –
ALC – – ⬎2.5 × 109 /L ⬎ 2.5 × 109 /L
Platelets ⬍100 × 109 /L – – –
LDH – 1.5 × normal ⬎700 U/L “Elevated”
Medullary blasts (%) ⬎5 ⬎5 – ⬎10
Stratification
Low-risk 0–1 factor 0–1 factor 1 factor 0 factor
Intermediate-risk – – – 1–2 factor(s)
Inter. 1 – – 2 factors –
Inter. 2 – – 3 factors –
High-risk 2–4 factors 2–3 factors 4 factors 3–4 factors
Median survival
Low-risk 27 mo vs. 44 mo vs. 26.3 mo 93 mo
Intermediate-risk – – – 26 mo
Inter. 1 – – 15.7 mo –
Chronic Myelomonocytic Leukemia and Myelodysplastic Syndrome
Inter. 2 – – 9.6 mo –
High-risk 17 mo 7 mo 4.9 mo 11 mo
p-Values ⬍0.01 ⬍0.0001 0.0000 ⬍0.00005
a Bournemouth score, with modification.
Abbreviations: MDAPS, M. D. Anderson Prognostic Score; ANC, absolute neutrophil count; PB IMG, peripheral blood immature myeloid cells; ALC, absolute
291
lymphocyte count; LDH, serum lactate dehydrogenase.

study), and absolute lymphocyte count ⬎2.5 × 109 /L was re-confirmed. As a side
note, the International Prognostic Scoring System (IPSS) for MDS (see chaps. 1
and 16) appears to be of limited value in CMML, perhaps due to the exclusion
of patients with a leukocyte count ⬎12 × 109 /L in the original analysis, which
resulted in exclusion of a substantial portion of patients with CMML (21,25).
Differential Diagnosis
CMML Versus MDS
The presence of a blood monocytosis should exclude MDS, especially as patients
with MDS typically have cytopenia(s) including a leukopenia. Theoretically,
patients with MDS may present with a blood monocytosis as a manifestation
of a concurrent subacute infectious or inflammatory process. In such cases, the
degree of the blood monocytosis is typically mild; there is not an accompanying
marrow monocytosis; and, most importantly, the monocytosis should not persist
for more than a month or two. In our experience, a hematologist or a pathologist
is more likely to encounter a patient who originally presents with an MDS picture
and with a relative but not absolute blood monocytosis, but who over time develops
a clear-cut blood and marrow monocytosis that fulfill the diagnosis of CMML.
Such patients remind us of the need for careful considerations of all clinical, lab-
oratory, and morphologic data with appropriate follow-up to arrive at the correct
diagnosis.
CMML Versus Chronic Myelogenous Leukemia (CML) and
Other CMPD
The presence of the Philadelphia chromosome and/or BCR/ABL fusion gene
defines CML and distinguishes it from other myeloid neoplasms. The occurrence
of the Philadelphia chromosome as a secondary event is rare and is often dis-
cernible with a careful review of the clinical history and other concurrent genetic
features. While patients with other CMPD such as polycythemia vera, essential
thrombocythemia, and the cellular phase of primary myelofibrosis may present
with a leukocyte count sufficiently high to result in a blood monocytosis (and at
times perhaps even marrow monocytosis), the monocytosis in such scenarios is
usually modest, and other features such as JAK2 V617F mutations and megakary-
ocyte morphology and cluster distribution typical of a CMPD can help identify
these disorders.
CMML-2 Versus Acute Myelomonocytic Leukemia
As mentioned above, for myelomonocytic neoplasms, for the purpose of blast
enumeration and subclassification, myeloblasts, monoblasts, and promonocytes
are included together. Yet, it can be difficult to distinguish by morphology between
a blast or promonocyte and an immature and cytologically atypical monocytic
precursor. Immunophenotypic analysis for CD34 expression by flow cytometry or
by paraffin-section immunohistochemistry is of limited value given the frequency
with which monoblasts and promonocytes do not express CD34 (vide supra).
Such uncertainties are especially unnerving when the combined blast count hovers
around 20%, so that the distinction between CMML-2 and acute myelomonocytic
leukemia cannot be made with confidence. In our experience, a review of the
clinical history as well as close follow-up are often necessary to assess the tempo
of the disease, in addition to a consideration of the risks and benefits of treatment
with curative intent versus a “watchful wait” approach versus palliation for that
particular patient.
REFRACTORY ANEMIA WITH RINGED SIDEROBLASTS ASSSOCIATED

WITH MARKED THROMBOCYTOSIS
Diagnosis
Definition
When the final version of the 2001 update of the WHO Classification of
Tumours of Haematopoietic and Lymphoid Tissues was published, it listed “refrac-
tory anemia with ringed sideroblasts associated with marked thrombocytosis”
(RARS-T) as a provisional entity within the category of unclassifiable myelodys-
plastic/myeloproliferative diseases (MDS/MPD-U). RARS-T was used to describe
patients whose hematologic parameters suggest both a myeloproliferative com-
ponent in the form of a thrombocytosis with platelet counts exceeding 600 ×
109 /L and with marrow megakaryocytic proliferation, but whose bone marrow
specimens also show features of dyserythropoiesis with more than 15% ringed
sideroblasts (RS) (26). (Also see chap. 9 for a slight revision of the designation,
to “Refractory Anemia with Ring Sideroblasts and Thrombocytosis”.)
Studies thus far have used different thresholds of thrombocytosis, from
⬎400 × 109 /L to ⬎600 × 109 /L, to define RARS-T. Moreover, the number of
patients in all of the studies to date has been small. In aggregate, these studies
described 38 patients with RARS and marked thrombocytosis (Table 2) and 112
patients with RARS and milder thrombocytosis (Table 3). In this section, for
clarity, the authors use the abbreviation “RARS-T” to refer to cases that meet
the WHO criterion of marked thrombocytosis, that is, platelet counts ⬎600 ×
109 /L; we use the designation “RARS-T/mild” to refer to cases in which the
thrombocytosis is lower.
Clinical Features
When the WHO criterion of marked thrombocytosis with platelet counts exceeding
600 × 109 /L is used, RARS-T accounts for 0.7% of 2986 consecutive MDS
diagnoses in the Düsseldorf MDS Registry and 1% of 1510 consecutive MDS
diagnoses in the combined Rush University Medical Center and University of
Massachusetts Medical Center MDS databases (27,28). When thrombocytosis
was defined as a platelet count between 400 × 109 /L and 600 × 109 /L, RARS-T/
mild accounted for 6 of the 317 (1.9%) consecutive MDS diagnoses at a single
294
Table 2 Reported Characteristics of Patients with Refractory Anemia with Ring Sideroblasts Associated with Marked Thrombocytosis
(Platelets ⬎600 × 109 /L)
Wang et al. (28) Gattermann et al. (27) Renneville et al. (38) Remacha et al. (39) Gupta et al. (40)
Patients (n) 16 10 7 3 2
Age (yr) (range) 66a (43–80) 65a (54–87) 69a (48–74) 67, 70, 82 47, 52
Sex (M:F) 12:4 4:6 3:4 0:3 1:1
Splenomegaly (yes/no) – – 0/7 1b /2 2/0
Hgb (g/L) (range) 99a (76–129) 92a (52–144) 96a (82–101) 95, 105, 124 122, 96
MCV (fL) (range) – – 97a (86–108) 88, 91, 91 88, 88
WBC (× 109 /L) (range) 9.3a (1.7–17.6) 8.9a (2.9–32.5) 7.3a (4.0–9.1) 6.7, 7.4, 9.5 13.3, 10.9
ANC (× 109 /L) (range) 6.8a (0.9–10.2) – – – –
Plt (× 109 /L) (range) 784a (602–1236) 773a (664–2100) 784a (600–1386) 621, 739, 1260 1157, 1265
PB smear stippling – – – – 2
Bone marrow
Cellularity (%) (range) 72 (25–95) – – “Hypercellular” “Hypercellular”
Megakyryocytes
ET-like 2 – 0 3 1
MDS-like 3 – 2 0 1
Mixed 10 – 1 0 0
Normal 0 – 4 0 0
RS (%) (range) (20 to ⬎75) 43a (15–57) 21a (15–54) 20, 40, 70 32, 15–20
Blasts (range) 1.6a (0–5) – ⬍5% – ⬍2%
Nguyen and Hanson
Retic fibrosis
Normal – – 1 – –
1+ to 2+ – – 7 – –
≥3+ – – 0 – Diffuse (1 of 1)
Cytogenetics
Abnormal 4 1 1 0 –
Normal 11 6 6 2 1
Not done 1 3 1 1
JAK2 V617F
Mutated 6 9 5 3 –
Wild-type 6 1 2 0 –
AML 1 0 1 – 1
Median overall survival (mo) 88 (median 66) (−233) – – –
(observation period)
a Median values.
b The splenomegaly was described as mild.
Abbreviations: Hgb, hemoglobin; MCV, mean corpuscular volume; WBC, white blood cell count; ANC, absolute neutrophil count; Plt, platelet count; PB, peripheral
blood; ET, essential thrombocythemia; MDS, myelodysplastic syndrome; RS, ringed sideroblasts; l Retic fibrosis, reticulin fibrosis syndrome; AML, acute myeloid
leukemia.
295
Table 3 Reported Characteristics of Patients with RARS-T/Mild
296
Wang et al. Boissinot et al. Szpurka et al. Ceesay et al. Schmitt-Graeff

Shaw (34) (28) (37) (36) (30) et al. (35)
Thrombocytosis criterion ⬎450 400–600 ⬎500 ⬎500 ⬎500 ⬎500

(× 109 /L)
Patients, n 16 27 16 9 6 38
Age, median (range) 75.4 (65.5–86.7) 68 (44–77) 79, 73 (59–88) 73 (61–85) 62 (52–71) 72 (54–86)
Sex (M:F) 8:8 14:13 9:7 – 2:2 21:17
Splenomegaly (yes/no) 2(slight)/14 – – 4/5 – 7/31 (15f/u)a
Hgbb (g/L) (range) 112 (62–156) 89 (63–109) 100, 88(75–133) 94 (86–124) 96 (74–119) 98(F), 110(M)
(60–130,F;
67–146,M)
MCV (fL) (range) 94.1 – – 103 (86.1– – 95 (76–127)
(79.7–102.5) 110.8)
WBC (× 109 /L) (range) 7.7 (2–15.7) 6.5 8, 10.9 5.384 (2.34– 8.39 (5.2–16.2) 10.7 (3.6–34)
(3.4–12.1) (5.3–17.5) 17.11)
ANC (× 109 /L) (range) – 3.9 (0.8–10) – 5.38 (2.34– – –
17.11)
Plt (× 109 /L) (range) 754.5 446 563, 943 664 (512– 699 (525–1099) 956 (533–2115)
(325–1389) (402–566) (450–1478) 2100)
PB smear stippling 8/14 – – – – –
Bone marrow
Cellularity (%) (range) 70 (40–90) 76 (40–100) – 85 (20–95) 90 –
M:E ratio 2.0 (0.1–6.5) – – – – –
Megakaryocytes – – – – – –
ET-like 11 3 – 5 – 6
MDS-like 1 6 – 1 – 11
Nguyen and Hanson
Mixed 1 1 – – – 0
Normal 3 16 – 3 – 0
CIMF-like 0 0 – 0 – 21
RS (range) – – 37,70 (15–91) 72 (42–88) 58 (45–70) –
Blasts – 1.7 (0–4) – 1 (0–2) – ⬎5% in 2, 19%
in 1
Retic fibrosis
Normal 3 – – 2 – 19
1+ to 2+ 10 – – 7 – 6
≥3+ 1 – – 0 – 2
Cytogenetics
Abnormal 0 5 – 3 (5q− × 1) 0 1
Normal 13 18 – 6 6 –
Not done 3 4 – 0 0 –
JAK2 V617F
Mutated – 0 5 6 4 –
Wild-type – 19 11 3 2 –
AML 0 2 – – 1/4 4 (none in ET
patients)
Median overall survival 71 (41;7–104) 101 (66;–) – – 3/4 alive (33; 100(81.5;
(mo) (observation period 3–58) 3–157)
in median; range in
months)
a 7 of 31 patients had splenomegaly at presentation/diagnosis and 15 additional patients developed splenomegaly with follow-up.
Abbreviations: Hgb, hemoglobin; MCV, mean corpuscular volume; WBC, white blood cell count; ANC, absolute neutrophil count; Plt, platelet count; PB,
peripheral blood; M:E, myeloid:erythroid ratio; ET, essential thrombocythemia; MDS, myelodysplastic syndrome; CIMF, chronic idiopathic myelofibrosis
(primary myelofibrosis); RS, ringed sideroblasts; Retic fibrosis, reticulin fibrosis; AML, acute myeloid leukemia.
297
European institution (29). In smaller retrospective studies involving fewer patients

(n = 17–110) in which variably low thresholds for thrombocytosis were employed
(range ⬎400 × 109 /L to 600 × 109 /L), patients with RARS-T/mild accounted for
approximately 9% to 35% of all patients with RARS (30–34). In their review of
patients who fulfilled their inclusion criteria for essential thrombocythemia (ET)
including a platelet count ⬎500 × 109 /L, Annette Schmitt-Graeff in Freiburg and
her colleagues reported that 2.7% of their patients with ET also had ⬎15% RS
(35).
Patients with RARS-T or RARS-T/mild tend to be older, with the median
age ranging from 65 years in large series, where the thrombocytosis is marked
(27,28), to 76 years in smaller series or in which the threshold for thrombocytosis
is lower (34,36,37). Although one group found a male predominance in a study
of 16 patients with RARS-T (38), others have found an approximately equal
male-to-female distribution among their patients with RARS-T or RARS-T/mild
(27,28,34,36,38). In one study, splenomegaly was noted in 4/9 patients with RARS-
T/mild (35); other studies with a combined number of 61 patients with RARS-T or
RARS-T/mild reported no splenomegaly in the majority of patients at presentation
(34,38). Venous thromboembolic events appeared to be very infrequent among the
small number of patients in whom this feature was studied (34,38).
Most patients with RARS-T or RARS-T/mild present with anemia, with reported
median hemoglobin values ranging from 88 to 112 g/L (range 52–156), although
occasional patients reportedly presented with normal hemoglobin levels and
who developed anemia only on follow-up (27,28,34,36,37). The anemia was
normocytic (34,35) or macrocytic (36,38). Although most series noted normal
to high-normal median leukocyte counts, the reported ranges have also indicated
occasional patients with leukopenia and absolute neutropenia (38), as well
as others with moderate-to-marked leukocytosis (27,35). With respect to the
thrombocytosis, apart from the variable definitions noted above, in some patients,
it was noted that the thrombocytosis was apparent only on follow-up or upon
review of prior records, with normal platelet counts at presentation (34,36,39).
Moderate-to-marked poikilocytosis was noted in the blood films of 8 of
the 12 patients with RARS-T/mild and 2 of the 2 patients with RARS-T, often
with basophilic stippling and/or a dimorphic pattern (34,40). The bone marrow
is typically hypercellular; occasional patients had normal to slightly decreased
marrow cellularity (28,34). Gene Shaw at Marshfield Clinic in Wisconsin found
no apparent erythroid hyperplasia in his report of 16 patients with RARS-T/mild
(34). By definition, ringed sideroblasts number ⬎15% and can reach 90% (39).
Although there has been little direct comment on the presence or absence of
dysgranulopoiesis, the WHO designation RARS implies that dysmyelopoietic
features are limited only to the erythroid lineage (Fig. 2 see colour insert).
With respect to megakaryocyte morphology in RARS-T, there is a spec-
trum of megakaryocytic morphologic features with a possible predominance of
(A) (B)
(C) (D)
(E) (F)
Figure 2 Refractory anemia with ring sideroblasts and thrombocytosis. (A) Peripheral
blood smear showing a thrombocytosis with occasional atypical platelets and a neu-
trophilia. Source: From Ref. 51. (B) Iron stain of bone marrow aspirate showing two
ringed sideroblasts. Source: From Ref. 51. (C) The bone marrow aspirate smear shows
a dysplastic megakaryocyte with a hypolobate nucleus. (D) Megakaryocytes with abnor-
mally disconnected nuclear lobes, so-called multinucleated megakaryocytes (arrows). (E,F)
Hypercellular bone marrow biopsy showing a panhyperplasia and prominent large atypical
megakaryocytes. The megakaryocytes form subtle clusters in the marrow biopsy. Source:
From Ref. 51. (see color insert)
ET-type megakaryocytes in which the megakaryocytes are increased, clustered,

and large to giant with hyperlobulated nuclei (34,36). Among the 38 patients in
their ET registry who also had ⬎15% RS, Schmitt-Graeff and colleagues similarly
found a predominance of myeloproliferative disease (MPD)-type megakaryocyte
morphology with further breakdowns into 6 patients with ET-like and 21 with
chronic idiopathic myelofibrosis-like megakaryocytes; 11 patients had MDS-type
megakaryocytes in which the megakaryocytes were hypolobulated or multinu-
cleated (35). It should be noted, however, that other studies have not found a
clear-cut predominant pattern of megakaryocyte morphology, and mixtures of ET-
or MPD-type megakaryocytes with MDS-type megakaryocytes as well as normal
megakaryocytes have been described (28,34,36,38). Both Shaw and Szpurka et
al. found slight-to-moderate reticulin fibrosis in the majority of their patients with
RARS-T/mild and RARS-T, respectively; marked reticulin fibrosis appeared to be
rare; a small number of patients had normal reticulin fibers (34,36).
Genetics
Karyotype
Karyotypic abnormalities occurr in fewer than 30% of patients with RARS-T
or RARS-T/mild (27,28,34,36,38). Reported abnormalities included trisomy 8,
11q−, and del12p (1.1–1.3), as well as more complex abnormalities, but no kary-
otypic abnormality has been identified as specifically associated with RARS-T or
RARS-T/mild.
JAK2 V617F Mutation

Attempts to determine whether RARS-T represents a distinct syndrome with both
MPD and MDS features, a variant of RARS with thrombocytosis, a variant of
ET with concurrent ringed sideroblasts, or (much less likely) simply a chance
concurrence of both ET and RARS in an elderly patient, have largely centered
on analyses to calculate the frequency of JAK2 V617F mutation and on in vitro
clonogenic assays of BFU-E to determine the presence or absence of endogenous
erythroid colony formation. The JAK2 V617F mutation is a somatic point mutation
at nucleotide 1849 of the gene encoding the Janus kinase 2 (JAK2) tyrosine
kinase; this mutation is a G-to-T transversion that is translated into a substitution
of phenylalanine for valine at codon 617 (JAK2 V617F) and that leads to a
constitutively activated JAK2 tyrosine kinase and phosphorylation of STAT5 (41).
Observations of factor-independent growth in animal models expressing the
JAK2 V617F mutated kinase, as well as reports of this mutation in up to 95% to
98% of patients with polycythemia vera and approximately 50% of patients with
ET or myelofibrosis (41,42), in contrast to its infrequent (⬍5%) occurrence in
MDS (43,44), suggest that this mutation may contribute to the myeloproliferative
phenotype of these disorders and may thus serve as a relatively specific diagnostic
marker of MPD. Using the more sensitive allele-specific PCR assay, this mutation
was reported in 31% to 90% of patients with RARS-T (27,28,36–38). In con-
trast, Wang et al. found no JAK2 V617F mutation among their 27 patients with
RARS-T/mild or among their 16 patients with MDS/MPD-U with platelets

⬎600 × 109 /L (28). M.M. Ceesay in London and colleagues from several Euro-
pean centers similarly found that a positive JAK2 V617F mutation result seemed
to discriminate between RARS with thrombocytosis and RARS without thrombo-
cytosis, with the mutation detected in 4 of the 6 patients with RARS and a platelet
count ⬎500 × 109 /L, in contrast to none of the 66 patients with RARS (30). These
relatively frequent rates of JAK2 V617F mutation among patients with RARS-
T may thus support the position that there is an underlying myeloproliferative
component in the pathogenesis of RARS-T.
On the other hand, when semisolid clonogenic assays of BFU-E were per-
formed on bone marrow cells obtained from JAK2 V617F–mutated RARS-T
patients, the results were more typical of those seen in MDS, with two separate
laboratories reporting no endogenous erythroid colony formation in the absence
of erythropoietin supplement in 2 of the 3 and 4 of the 4 patients, respectively
(38,39). Analogously, assays to assess megakaryocytic colony formation on cells
obtained from four patients with JAK2 V617F–mutated RARS-T yielded simi-
larly negative results (37). Moreover, in the presence of erythropoietin, two groups
reported that the erythroid colonies were rare to poorly developed and even exhib-
ited increased cell death, features which are more compatible with MDS and
indicative of ineffective erythropoiesis (37,38). Thus, at least in the small number
of patient samples with JAK2 V617F–positive RARS-T who have been studied
by both methods, there appear to be both myeloproliferative and myelodysplastic
biological features.
Although the number of patients studied in aggregate was small (n = 37),
thus far, no statistically significant differences have been found between those
RARS-T patients with and those without the JAK2 V617F mutation with respect
to hematologic parameters, bone marrow morphology including megakaryocyte
morphology, and cytogenetic findings (28,36,37).
Other Genetic Aberrations

A W515L mutation of the MPL gene was recently described in a 63-year-old
patient with 1222 × 109 /L platelets and 65% RS; JAK2 V617F mutation was neg-
ative (45). Zachary Nearman and colleagues at Cleveland Clinic reported a signif-
icantly higher frequency of hemochromatosis-associated gene mutations (C282Y
or H63D variants) among patients with RARS compared to healthy controls, and
especially among patients with RARS-T (10/14) compared to patients with MDS
or those with other JAK2 V617F–positive MPD (46).
Prognosis
Interestingly, despite using different definitions of thrombocytosis (platelet count
⬎450 × 109 /L and ⬎600 × 109 /L, respectively), Shaw and Wang et al. reported
similar outcomes for patients with RARS-T/mild and those with RARS-T.
With median follow-ups of 41 months (range 7–104 months) and 66 months
(range 6–129 months), respectively, they found that the median survival for patients
with RARS-T was not statistically different from that of patients with RARS
(71 months vs. 64 months) or from that of patients with RARS-T/mild (88 months
vs. 101 months) (28,34). These investigators did observe that the median overall
survival of patients with RARS-T was inferior to that of patients with ET, where
over the same observation period, the median overall survival had not been reached
in the latter. These investigators also found that the patients with RARS-T/mild had
significantly more prolonged survival compared to those with MDS (71 months
vs. 20 months; p ⬍ 0.01) (34). There was a trend towards more prolonged survival
among patients with RARS-T compared to those with MDS/MPD-U and marked
thrombocytosis, 88 months versus 44 months, but the difference did not reach
statistical significance (p = 0.09) (28).
These results are consistent with the observation by Schmitt-Graeff and
colleagues that MDS-type megakaryocyte morphology appears to confer a worse
prognosis, with a median overall survival of 57 months among their 11 patients
with RARS-T/mild and MDS-type megakaryocytes compared to those with ET-
like megakaryocyte morphology in whom the median overall survival was not
reached during the observation period of 3–157 months (35). Evolution to high-
grade MDS or AML appears to be infrequent among patients with RARS-T or
RARS-T/mild as compared to patients with MDS or as compared to those with
MDS/MPD-U and marked thrombocytosis (28,34). Although the difference was
not statistically significant, Wang et al. noted that their 16 patients with RARS-T
had shorter median overall survival compared to the 14 patients with 5q− syn-
drome and platelets ⬎400 × 109 /L (88 months vs. 125 months; p = 0.18) (28).
Differential Diagnosis
Given their different prognoses, the diagnosis of RARS-T should be carefully
weighed against that of ET and against that of other types of MDS or MDS/MPD-U
with thrombocytosis.
RARS-T Versus ET with Ringed Sideroblasts

It is important to exclude ET with concurrent RS that are due to other non-
neoplastic (reactive) causes of RS, such as zinc toxicity, copper deficiency, anti-
tuberculosis medications, etc. Because it is still unclear whether or not in a few
elderly patients RARS-T may represent the chance coincident superimposition of
both ET and RARS, a review of prior history and hematologic and/or blood and
bone marrow findings may be helpful. While it seems intuitive that the presence of
⬎15% RS should segregate patients with RARS-T from those with ET, Schmitt-
Graeff and colleagues reported six patients who fulfilled their criteria for inclusion
in their ET database but who otherwise also had ⬎15% RS (35). In cases with both
thrombocytosis and ⬎15% RS, Schmitt-Graeff and colleagues suggested that a
rigorous microscopic examination of megakaryocyte morphology should sort out
ET from MDS cases and others. However, as noted above, in several studies of
patients with RARS-T or RARS-T/mild, megakaryocyte morphology was mixed,

and it might not be possible to distinguish definitively RARS-T from ET with
ringed sideroblasts by morphology alone. Although some studies have noted that
in comparison to patients with ET, patients with RARS-T had lower hemoglobin
values, lower leukocyte counts, lower platelet counts, and higher bone marrow
cellularity (28,34), Shaw also noted that there was a large overlap between the
two groups (34). The presence of JAK2 V617F mutation and of ET-associated
megakaryocyte morphology did not appear to be sufficiently discriminatory, since
these two findings have been described in up to 90% of the patients in each group.
RARS-T Versus RARS with Thrombocytosis

Although the prognoses for patients with RARS-T, RARS, and RARS-T/mild
appear similar according to Shaw and Wang et al., it is important to distinguish
RARS-T from RARS with mild or marked thrombocytosis in which the throm-
bocytosis is of a secondary/reactive nature so as not to overlook comorbidities
such as inflammation, hemorrhage, or iron deficiency. Although not completely
discriminatory, lower hemoglobin values and lower leukocyte counts had been
reported among patients with RARS compared to those in patients with RARS-T
or RARS-T/mild (28,34), with Shaw reporting a significantly higher mean cor-
puscular volume (102.9 fL vs. 94 fL; p ⬍ 0.0001) among patients with RARS
(34). When strict diagnostic criteria for RARS are applied, as prescribed by the
WHO and as further restricted by Shaw (26,34), the bone marrow of patients
with RARS should not display significant megakaryocytic atypia, although with
the caveat that a small minority of patients with RARS-T themselves may have
normal megakaryocyte number, distribution, and morphology. There is no statisti-
cally significant difference in the frequencies of karyotypic abnormalities among
patients with RARS-T, RARS-T/mild, or RARS (28). JAK2 V617F mutation is
rare among patients with RARS or RARS-T/mild, which suggests that detection
of the mutation may favor a diagnosis of RARS-T (28,36,39).
RARS-T Versus Other MDS or MDS/MPD-U with Thrombocytosis

Dysmyelopoietic features involving the granulocytic and megakaryocytic lineages
are uncommon in RARS-T, and by definition patients with RARS-T have fewer
than 5% blasts in the peripheral and bone marrow (26,34). Furthermore, the
presence of blood and/or marrow monocytosis and/or dysgranulopoiesis should
prompt one to consider a diagnosis of CMML instead, since thrombocytosis and
ringed sideroblasts may be seen in CMML. Because JAK2 V617F mutation is rare
to infrequent among patients with MDS or other MDS/MPD besides RARS-T,
the presence of such a mutation may help favor the diagnosis of RARS-T. Lastly,
as discussed below, the designation “MDS/MPD-U” is one of default, to be consid-
ered only when other MDS/MPD entities including the provisional entity RARS-T
under discussion have been carefully excluded.
JAK2 V617F–Positive RARS-T Versus JAK2 V617–Negative RARS-T

As noted above, although JAK2 V617F mutation appears to be rare (≤5%) among
true MDS (43,47) and among MDS/MPD-U with or without thrombocytosis
(28,36), this mutation has been reported to be frequent among patients who fulfill
the current WHO definition of RARS-T, ranging from approximately 30% to 90%
(27,28,36–39). Although sample sizes have been small with up to 16 patients in
the largest reported study (28), thus far, no statistically significant differences have
been reported between patients with RARS-T with and those without the JAK2
V617F mutation with respect to their clinical, hematological, morphological, or
cytogenetic features (28,30,36,37).
SUMMARY
When strictly defined according to WHO criteria with thrombocytosis set at
⬎600 × 109 /L, RARS-T is a rare disease in large MDS registries, accounting
for approximately 1% of all MDS. Survival data obtained retrospectively in small
series suggest that the prognosis of patients with RARS-T, as a group, is similar to
that of patients with RARS with or without mild thrombocytosis, inferior to that of
patients with ET, and superior to that of patients with MDS. While reports of the
presence of JAK2 V617F mutation and of MPD-type megakaryocyte morphology
among 30% to 90% of patients with RARS-T support the notion that there is
a “proliferative” component to the biology of this entity, results of endogenous
clonogenic assays in a small number of JAK2 V617F–mutated RARS-T patients
still hint at a possible myelodysplastic element. In the absence of a consistent or
specific genetic, morphologic, and clinical profile, it appears that at this time, this
RARS-T category likely encompasses more than one disease, and care should be
exercised to distinguish this entity from others in the differential diagnosis.
ATYPICAL CHRONIC MYELOID LEUKEMIA (aCML)

Diagnosis
Definition
The WHO and FAB definitions of aCML are similar (22,48). Both refer to
a myeloid disorder in which there is a leukocytosis due to a left-shifted neu-
trophilia, but aCML differs from chronic myelogenous leukemia in that there is
no t(9;22)(q34;q11) Philadelphia chromosomal abnormality or BCR/ABL fusion
gene, and absolute basophilia is absent or minimal. In further contrast with CML,
dysgranulopoiesis is prominent in aCML, and unlike CMML, there is no absolute
monocytosis with aCML. Blood and marrow blasts number are less than 20% in
patients with aCML by definition, while the presence of ⬎10% circulating neu-
trophilic precursors should favor a diagnosis of aCML over chronic neutrophilic
leukemia.
Clinical Features
It is estimated that aCML occurs at approximately 1/100th the frequency of
Philadelphia chromosome–positive, BCR/ABL-positive CML, although the exact
incidence is not known. In previous reports of aCML, the reported median age at
diagnosis was in the seventh or eighth decade, with a male:female ratio between
1:1 and 2.5:1. In a more recent report of 55 patients who fulfilled the WHO defi-
nition of aCML, Massimo Breccia and colleagues in Rome reported a median age
of 62 years (range 46–81 years), with a slight female predominance (M:F, 24:31)
(49). Splenomegaly and/or hepatomegaly was reported in approximately one-half
of the patients in this study.

Similar to prior studies, Breccia et al. described a median leukocyte count of
23.7 × 109 /L (range 14–150) with 10% to 20% circulating immature granulocytic
precursors (median 13%) (49). Dysgranulopoiesis is pronounced and typically
involves nuclear abnormalities such as peculiarly abnormal condensation of
nuclear chromatin and pseudo-Pelger-Huet anomalies; neutrophils with cyto-
plasmic hypogranularity or with bizarrely hypersegmented nuclei can also be
seen (Fig. 3 see colour insert). Anemia is common, whereas the platelet count is
variable (range, 44 × 109 /L to 2675 × 109 /L; median 319) (48,49).
The bone marrow is hypercellular due to increased granulopoiesis with
prominent dysgranulopoiesis. The number of megakaryocytes is variable, with
dysmegakaryopoiesis reported in approximately half of the cases in the recent
report by Breccia et al. (49). Reticulin fibrosis was seen in approximately 20%
(A) (B)
Figure 3 Atypical chronic myeloid leukemia. Peripheral blood smear showing: (A) left-
shifted granulocytosis and (B) dysplastic neutrophils with hypolobulation and hypogranu-
lation. Source: From Ref. 51. (see color insert)
of the cases and was described to be of “trace” amounts in the same study. By
definition, blasts number less than 5%.
Cytochemistry and flow cytometry immunophenotyping analysis are of no
known value in the diagnosis of aCML.
Genetics
Although cytogenetic abnormalities were reported in up to 80% of the cases of
aCML in prior studies, they were seen in only 11 of the 55 patients (20%) in the
report by Breccia and coworkers (20%) (49). This discrepancy is likely due to the
small numbers of patients studied in the different series. A variety of chromosomal
aberrations has been described in patients with aCML and has included trisomy
8, del(20q), i(17q), and del(12p). However, to date, no specific chromosomal or
genetic abnormality has been identified in association with aCML.
Prognosis
Patients with aCML have a poor prognosis overall, with a median survival ranging
from just about 1 year in older reports to about 2 years in the more recent study
by Breccia et al. (49) and in the group of patients with Philadelphia chromosome–
negative/BCR/ABL-negative myeloid disorders reported by Onida et al. from M. D.
Anderson Cancer Center (4). Attempts to construct prognostic predictors are few
and clearly limited by the small number of patients. With that caveat, Breccia and
colleagues reported older age (⬎65 years), female gender, and a leukocyte count
⬎50 × 109 /L to be features associated with a shorter survival after multivariate
analysis. In this retrospective study in which the majority of patients were reported
to have received conservative therapy, 22 of 55 (40%) patients transformed into
acute leukemia. Factors associated with leukemic transformation were palpable
hepato- or splenomegaly, a monocyte proportion between 3% and 8%, medullary
blasts ⬎5%, and transfusional requirement (49).
MYELODYSPLASTIC/MYELOPROLIFERATIVE DISEASE, UNCLASSIFIABLE

Before assigning patients to a category of MDS/MPD, unclassifiable (MDS/MPD-
U), a prior history of an underlying MDS or MPD should be excluded, there should
be no t(3;3)(q21;q26) or inv(3)(q21q26) abnormalities, and there should be no his-
tory of recent exposure to cytotoxic or growth factor therapy that might account
for the cytopenia(s) or cytosis(es) observed, respectively. The 2001 WHO classi-
fication reserves the MDS/MPD-U designation for those patients who exhibit (a)
the clinical, laboratory, and morphologic features of one of the MDS categories
as well as features of myeloproliferation, such as thrombocytosis ≥600 × 109 /L
associated with megakaryocytic proliferation and/or leukocytosis ≥13.0 × 109 /L;
and (b) whose constellation of clinical, hematologic, pathologic, and genetic fea-
tures does not fit into any of the recognized MDS, MPD, or overlap MDS/MPD
categories (50). Splenomegaly may or may not be present. The presence of 20%
blasts or more in the blood or marrow should lead to the diagnosis of acute
leukemia. It is likely that by following this process of exclusion rigorously, with a
careful review of the patient’s prior history including prior laboratory hematologic
parameters, there remain very few patients in this category. The pathobiology of
these disorders is likely to be variable.
REFERENCES
1. Neuwirtova R, Mocikova K, Musilova J, et al. Mixed myelodysplastic and myelopro-
liferative syndromes. Leuk Res 1996; 20(9):717–726.
2. Wijermans PW, Ruter B, Baer MR, et al. Efficacy of decitabine in the treat-
ment of patients with chronic myelomonocytic leukemia (CMML). Leuk Res 2008;
32(4):587–591.
3. Germing U, Strupp C, Alvado M, et al. New prognostic parameters for chronic
myelomonocytic leukemia? Blood 2002; 100(2):731–732 (letter).
4. Onida F, Ball G, Kantarjian HM, et al. Characteristics and outcome of patients with
philadelphia chromosome negative, bcr/abl negative chronic myelogenous leukemia.
Cancer 2002; 95(8):1673–1684.
5. Vardiman JW. Myelodysplastic/myeloproliferative diseases: Introduction. In: Jaffe
Tissues. Lyon: IARC Press, 2001:47–48.
6. Vardiman JW, Pierre R, Bain B, et al. Chronic myelomonocytic leukaemia. In: Jaffe
Tissues. Lyon: IARC Press; 2001:49–52.
7. Germing U, Gattermann N, Minning H, et al. Problems in the classification of CMML
dysplastic versus proliferative type. Leuk Res 1998; 22:871–878.
8. Nosslinger T, Reisner R, Gruner H, et al. Dysplastic versus proliferative CMML a
retrospective analysis of 91 patients from a single institution. Leuk Res 2001; 25:741–
747.
9. Orazi A, Chiu R, O’Malley DP, et al. Chronic myelomonocytic leukemia: The role of
bone marrow biopsy immunohistology. Modern Pathol 2006; 19:1536–1545.
10. Onida F, Kantarjian HM, Smith TL, et al. Prognostic factors and scoring systems in
chronic myelomonocytic leukemia: A retrospective analysis of 213 patients. Blood
2002; 99(3):840–849.
11. Onida F, Beran M. Chronic myelomonocytic leukemia: Myeloproliferative variant.
Curr Hematol Rep 2004; 3:218–226.
12. Pardanani A, Levine RL, Lasho TL, et al. MPL515 mutations in myeloproliferative
and other myeloid disorders: A study of 1182 patients. Blood 2006; 108(10):3475–
3476.
13. Magnusson MK, Meade KE, Nakamura R, et al. Activity of STI571 in chronic
myelomonocytic leukemia with a platelet-derived growth factor ␤ receptor fusion
oncogene. Blood 2002; 100(3):1088–1091.
14. Pitini V, Arrigo C, Teti D, et al. Response to STI571 I chronic myelomonocytic
leukemia with platelet derived growth factor beta receptor involvement: A new case
report. Haematologica 2003; 88(6):e78–e79.
15. Drechsler M, Hildebrandt B, Kundgen A, et al. Fusion of H4/D10S170 to PDGRFb

in a patient with chronic myelomonocytic leukemia and long-term responsiveness to
imatinib. Ann Hematol 2007; 86:353–354.
16. Gonzalez-Medina I, Bueno J, Torrequebrada A, et al. Two groups of chronic
myelomonocytic leukaemia: Myelodysplastic and myeloproliferative. Prognostic
implications in a series of a single center. Leukemia Res 2002; 26:821–824.
17. Kerkhofs H, Hermans J, Haak HL, et al. Utility of the FAB classification for myelodys-
plastic syndromes: Investigation of prognostic factors in 237 cases. Br J Haematol
1987; 65(1):73–81.
18. Germing U, Kundgen A, Gattermann N. Risk assessment in chronic myelomonocytic
leukemia (CMML). Leuk Lymphoma 2004; 45(7):1311–1318.
19. Breccia M, Latagliata R, Mengarelli A, et al. Prognostic factors in myelodysplastic
and myeloproliferative types of chronic myelomonocytic leukemia: A retrospective
analysis of 83 patients from a single institution. Haematologica 2004; 89(7):866–
868.
20. Germing U, Strupp C, Knipp S, et al. Chronic myelomonocytic leukemia in the light
of the WHO proposals. Haematologica 2007; 92(7):974–977.
21. Fenaux P, Beuscart R, Lai JL, et al. Prognostic factors in adult chronic myelomonocytic
leukemia: An analysis of 107 cases. J Clin Oncol 1988; 6(9):1417–1424.
22. Bennett JM, Catovsky D, Daniel MT, et al. The chronic myeloid leukaemias: Guide-
lines for distinguishing chronic granulocytic, atypical chronic myeloid and chronic
myelomonocytic leukaemia. Proposals by the French-American-British Cooperative
Leukaemia Group. Br J Haematol 1994; 84:746–754.
23. Worsley A, Oscier DG, Stevens J, et al. Prognostic features of chronic myelomonocytic
leukaemia: A modified Bournemouth score gives the best prediction of survival. Br J
Haematol 1988; 68:17–21.
24. Beran M, Wen S, Shen Y, et al. Prognostic factors and risk assessment in chronic
myelomonocytic leukemia: Validation study of the M. D. Anderson Prognostic Scor-
ing System. Leuk Lymphoma 2007; 48(6):1150–1160.
25. Oscier D, Chapman R. The classification of chronic myelomonocytic leukaemia. Leuk
Res 1998; 22:879–880.
26. Bain B, Vardiman JW, Imbert M, et al. Myelodysplastic/myeloproliferative disease,
unclassifiable. In: Jaffe ES, Harris NL, Stein H, Vardiman JW, eds. World Health
Organization Classification of Tumours: Pathology and Genetics of Tumours of
Haematopoietic and Lymphoid Tissues. Lyon: IARC Press, 2001:59.
27. Gattermann N, Billiet J, Kronenwett R, et al. High frequency of the JAK2 V617F
mutation in patients with thrombocytosis (platelet count ⬎600 × 109/L) and ringed
109(3):1334–1335.
28. Wang SA, Hasserjian RP, Loew JM, et al. Refractory anemia with ringed sideroblasts
associated with marked thrombocytosis harbors JAK2 mutation and shows over-
lapping myeloproliferative and myelodysplastic features. Leukemia 2006; 20:1641–
1644.
29. Cabello AI, Collado R, Ruiz MA, et al. A retrospective analysis of myelodysplastic
syndromes with thrombocytosis: Reclassification of the cases by WHO proposals.
Leuk Res 2005; 29:365–370.
30. Ceesay MM, Lea NC, Ingram W, et al. The JAK2 V617F mutation is rare in RARS
but common in RARS-T. Leukemia 2006; 20:2060–2061.
31. Kushner JP, Lee GR, Wintrobe MM, et al. Idiopathic refractory sideroblastic ane-
mia: Clinical and laboratory investigation of 17 patients and review of the literature.
Medicine 1971; 50:139–159.
32. Streeter RR, Presant CA, Reinhard E. Prognostic significance of thrombocytosis in
idiopathic sideroblastic anemia. Blood 1977; 50:427–432.
33. Juneja SK, Imbert M, Jouault H, et al. Haematological features of primary myelodys-
plastic syndromes (PMDS) at initial presentation: A study of 118 cases. J Clin Pathol
1983; 36:1129–1135.
34. Shaw GR. Ringed sideroblasts with thrombocytosis: An uncommon mixed myelodys-
plastic/myeloproliferative disease of older adults. Br J Haematol 2005; 131:180–184.
35. Schmitt-Graeff A, Thiele J, Zuk I, et al. Essential thrombocythemia with ringed sider-
oblasts: A heterogeneous spectrum of diseases, but not a distinct entity. Haematologica
2002; 87:392–399.
dition characterized by JAK2 V617F mutation. Blood 2006; 108:2173–2181.
37. Boissinot M, Garand R, Hamidou M, et al. The JAK2-V617F mutation and essential
thrombocythemia features in a subset of patients with refractory anemia with ring
sideroblasts (RARS). Blood 2006; 108(5):1781–1782.
38. Renneville A, Quesnel B, Charpentier A, et al. High occurrence of JAK2 V617F
mutation in refractory anemia with ringed sideroblasts associated with marked throm-
bocytosis. Leukemia 2006; 20:2067–2070.
39. Remacha AF, Nomdedeu JF, Puget G, et al. Occurrence of the JAK2 V617F muta-
tion in the WHO provisional entity: Myelodysplastic/myeloproliferative disease,
unclassifiable—refractory anemia with ringed sideroblasts associated with marked
thrombocytosis. Haematologica 2006; 91:719–720.
40. Gupta R, Abdalla SH, Bain BJ. Thrombocytosis with sideroblastic erythropoiesis: A
mixed myeloproliferative myelodysplastic syndrome. Leuk Lymphoma 1999; 34(5–
6):615–619.
41. James C, Ugo V, Le Couedic JP, et al. A unique clonal JAK2 mutation leading to
constitutive signaling causes polycythaemia vera. Nature 2005; 434:1144–1148.
42. Levine RL, Wadleigh M, Cools J, et al. Activating mutation in the tyrosine kinase
JAK2 in polycythemia vera, essential thrombocythemia, and myeloid metaplasia with
myelofibrosis. Cancer Cell 2005; 7:387–397.
43. Steensma DP, Dewald GW, Lasho TL, et al. The JAK2 V617F activating tyrosine
and myelodysplastic syndromes. Blood 2005; 106:1207–1209.
44. Ingram W, Lea NC, Cervera J, et al. The JAK2 V617F mutation identifies a subgroup
of MDS patients with isolated deletion 5q and a proliferative bone marrow. Leukemia
2006; 20:1319–1321.
45. Schnittger S, Bacher U, Haferlach C, et al. Detection of an MPLW515 mutation in
a case with features of both essential thrombocythemia and refractory anemia with
ringed sideroblasts and thrombocytosis. Leukemia 2008; 22(2):453–455.
46. Nearman ZP, Szpurka H, Serio B, et al. Hemochromatosis-associated gene muta-
tions in patients with myelodysplastic syndromes with refractory anemia with ringed
sideroblasts. Am J Hematol 2007; advance online publication.
47. Nelson ME, Steensma DP. JAK2 V617F in myeloid disorders: What do we know now,
and where are we headed? Leuk Lymphoma 2006; 47:177–194.
48. Vardiman JW, Imbert M, Pierre R, et al. Atypical chronic myeloid leukemia. In: Jaffe
Tissues. Lyon: IARC Press, 2001:53–54
49. Breccia M, Biondo F, Latagliata R, et al. Identification of risk factors in atypical
chronic myeloid leukemia. Haematologica 2006; 91(11):1566–1568.
50. Bain B, Vardiman JW, Imbert M, et al. Myelodysplastic/myeloproliferative disease,
unclassifiable. In: Jaffe ES, Harris NL, Stein H, Vardiman JW, eds. World Health
Organization Classification of Tumours: Pathology and Genetics of Tumours of
Haematopoietic and Lymphoid tissues. Lyon: IARC Press, 2001:58–59.
51. Bennett JM, Hanson CA. Myelodysplastic/myeloproliferative diseases (overlap syn-
dromes). In: Silver RT, Tefferi A, eds. Myeloproliferative Disorders: Biology and
Management. New York: Informa Healthcare, 2008:211–218.
14
MDS in Children
Henrik Hasle
Department of Pediatrics, Aarhus University Hospital Skejby, Aarhus, Denmark
Charlotte M. Niemeyer
Division of Paediatric Hematology and Oncology, Department of Pediatrics
and Adolescent Medicine, Albert-Ludwigs-University, Mathildenstrasse 1,
Freiburg, Germany
INTRODUCTION
Myelodysplastic syndrome (MDS) is much rarer in children than in adults and
most of the literature on MDS is based upon studies in elderly patients; however,
there are significant differences between MDS in children and adults (Table 1).
The morphologic features and cytogenetic findings at diagnosis differ significantly
between children and adults. Many children have associated abnormalities, that is,
preexisting bone marrow (BM) failure or congenital abnormalities. The therapeutic
aim in children with MDS is primarily a cure whereas this possibility is often not
realistic in adults. The rarity of MDS in children and the lack of an overall accepted
classification have contributed to the paucity of MDS in the pediatric literature.
However, an increasing number of larger series on childhood MDS patients have
been published over the last decade (1–10).
Myeloid leukemia of Down syndrome (ML-DS) and juvenile myelomono-
cytic leukemia (JMML) have often been included under the heading of MDS;
today, there is consensus that these disorders are distinct from MDS (11). ML-DS
is discussed at the end of this chapter, whereas JMML is not discussed in any
details in the present paper.
311
312 Hasle and Niemeyer
Table 1 Major Differences Between MDS in Children and Adults
Children Adults
Incidence/million 1–2 ⬎30

RA with ringed sideroblasts (%) 2 25
Constitutional abnormalities (%) 30 ⬍5
Cytogenetic aberrations (%) 50 40
−7/del(7q) (%) 30 10
−5/del(5q) (%) 1–2 20
Mutation of NRAS Rare Common
Hypermethylation (%) ⬎ 50 ⬎ 50
Principal aim of treatment Curative Palliative
Historical Background of the Classification of Childhood MDS

The rarity and the heterogeneous nature of the disease have contributed to the
difficulties in classifying childhood MDS. MDS was for the first time included in
the 2005 revised version of the classification of childhood malignancies (12). A
variety of names was previously used for childhood MDS reflecting the concep-
tual and diagnostic difficulties. The term preleukemia was introduced in pediatrics
in the early 1970s (13), but the number of reported cases remained extremely
sparse. Only 11 cases were identified in a review published in 1980 (14). The term
preleukemia is misleading since it gives the impression that the patient has yet to
develop leukemia. MDS is a clonal malignant disease in itself and not a precursor
of a malignant condition. This confusion was also increased by using the term
preleukemia for conditions with an increased risk of malignant hematological dis-
eases, that is, aplastic anemia and Fanconi anemia. Furthermore, in the 1980s, MDS
in children was often presented together with cases of transient pancytopenia pre-
ceding ALL (pre-ALL) and collectively described as preleukemic states (15,16).
These two conditions are very distinctive and should be considered separately (17).
Monosomy 7 in children has often been considered to represent a distinct
hematological disorder described as the monosomy 7 syndrome (18,19) and
included as a separate entity in the classification of childhood MDS (19,20). It was
later shown that complete loss of chromosome 7 occurs in all MDS subgroups (21,
22) and there is no evidence to consider monosomy 7 as a discrete entity (22,23).
The French-American-British (FAB) group classification from 1982 MDS
comprised the five subgroups: refractory anemia (RA), RA with ringed sideroblasts
(RARS), RA with excess of blasts (RAEB), RAEB in transformation (RAEB-t),
and chronic myelomonocytic leukemia (CMML) (24). The FAB classification was
developed based upon review of smears from adult patients. The FAB classifica-
tion of Acute myeloid leukemia (AML) became rapidly generally accepted in
pediatrics, whereas the FAB classification of MDS was only slowly and partly
accepted among pediatricians (25,26). The FAB classification gradually became
MDS in Children 313
used by most researchers and clinicians, although some investigators experienced

problems in applying the classification in a pediatric population (3). The FAB clas-
sification had prognostic impact in children (1,22) and facilitated communication
about pediatric MDS but lacked to address the specific diseases and morphological
features in children and the frequent occurrence of associated anomalies.
The WHO classification on neoplastic diseases of the hematopoietic tissues
from 2001 (27) incorporated both morphology and cytogenetics. The threshold for
distinguishing AML from MDS was lowered from 30% to 20% blasts. Also the
WHO classification was based upon review of adult cases, and although JMML
was recognized as a separate entity, the classification of MDS did not acknowledge
the special features of MDS in children. The WHO subtype RARS is extremely
rare in children and the unique 5q− syndrome is absent in pediatrics. Furthermore,
the importance of multilineage dysplasia is unknown in children. There are no data
to indicate whether a blast threshold of 20% is better than the traditional 30% to
distinguish MDS from AML in children. The unique features of Down syndrome
are not appropriately addressed in the WHO classification. Accordingly, the WHO
classification had limited use in pediatrics.
Current Approach to the Classification of Childhood MDS

International consensus has been achieved on the classification of MDS in
childhood (11). Myelodysplastic and myeloproliferative disorders in children are
separated into three main groups: JMML, MDS, and Down syndrome diseases
(Table 2).
MDS is subdivided into refractory cytopenia (RC), RAEB, and RAEB-t.
The classification is used for both de novo and secondary MDS. The change in
nomenclature from RA to RC reflects that anemia is not a prerequisite for the
diagnosis (further details under the section, “Clinical and Laboratory Features”).
It is suggested to retain the RAEB-t entity but to emphasize that the blast count
Table 2 Diagnostic Categories of Myelodysplastic and

Myeloproliferative Diseases in Children According to the Pediatric
Approach to the WHO Classification (11)
Myelodysplastic/myeloproliferative disease
Juvenile myelomonocytic leukemia (JMML)
Down syndrome (DS) disease
Transient abnormal myelopoiesis (TAM)
Myeloid leukemia of DS
Myelodysplastic syndrome (MDS)
Refractory cytopenia (RC) (PB blasts ⬍2% and BM blasts ⬍5%)
Refractory anemia with excess blasts (RAEB) (PB blasts 2–19% or
BM blasts 5–19%)
RAEB in transformation (RAEB-t) (PB or BM blasts 20–29%)
is insufficient to differentiate AML from MDS (further details under the section,
“Differential Diagnosis”).
Myeloid leukemia in children with Down syndrome has unique features and
is kept separate as a distinct entity (see the section, “Myeloid Leukemia and Down
Syndrome,” for more details).
The Toronto group has proposed a descriptive system designed to assess
children with MDS according to category, cytology, and cytogenetics (CCC) (28).
The system excludes JMML but include patients with Down syndrome. Cytology
is used to subdivide both RC and RAEB into three subgroups based upon level
of dysplasia. The system records associated abnormalities and cytogenetic abnor-
malities. The CCC system has an infinite number of possible subgroups making it
difficult to use in clinical practice or research. The two pediatric classification sys-
tems are both superior to classify MDS in children than the adult FAB and WHO
classifications with the pediatric WHO system being the most exclusive (29).
The pediatric modification of the classification (11) emphasizes the
subtypes of pediatric MDS and eliminates adult subtypes that are rare or unseen.
However, there will still be borderline cases difficult to fit into the classification
but the pediatric WHO classification allows classification of more than 95% of
the patients (10).
Primary and Secondary MDS

MDS can arise in a previously healthy child and is conformingly named “de novo”
or “primary.” It may also develop in a child with a known predisposing condition
and referred to as “secondary.” Secondary MDS is seen in patients (a) after
chemo- or radiation therapy (therapy-related MDS), (b) with inherited BM failure
disorders, (c) with acquired aplastic anemia, and (d) with familial MDS. It is to be
recognized, however, that children with so-called “primary” MDS may have an
underlying yet unknown genetic defect predisposing them to MDS at young age.
Therefore, the distinction between primary and secondary disease may become
arbitrary.
Myeloid neoplasia in patients with predisposing conditions almost always
shares the biologic characteristics of MDS regardless of the presenting blast count.
The prognosis appears to depend primarily on the cytogenetic profile.
The natural history and therapy are different when MDS occurs in Fanconi
anemia (30) and should be reported separately. There are no solid data documenting
whether MDS occurring in patients with constitutional abnormalities other than
Down syndrome and Fanconi anemia differ from MDS in other children. Patients
should be diagnosed according to the MDS guidelines and included in series of
MDS reporting type and frequency of predisposing conditions.
EPIDEMIOLOGY
The epidemiological literature on childhood MDS is sparse, some of the rea-
sons for this is as follows: (1) The lack of an overall accepted classification.
MDS in Children 315
(2) The indolent nature of the disease may not lead to referral to a tertiary center.
(3) Cancer registries generally do not register MDS. (4) Many epidemiological
data are derived from multi-institutional studies to which MDS patients may be
referred only after progression of their disease. Some authors have tried to esti-
mate the frequency of MDS by searching for a preceding “preleukemic” phase
among children with AML. The earliest of these reports found MDS in 17% of
childhood AML corresponding to 2.9% of all children with leukemia (31). Other
studies have confirmed that 12% to 20% of childhood AML are preceded by a
recognized preleukemic phase (32,33). The studies underestimate the incidence of
MDS because AML does not develop in all cases of MDS. Some children die from
complications of cytopenia or are treated before progression to AML. The exclu-
sion of various constitutional abnormalities also contributes to an underestimation
of the frequency of MDS (1,34–36).
Incidence, Sex, Age, and Subtype Distribution

Combined population-based data from Denmark and British Columbia (BC) in
Canada identified 38 cases of MDS representing 4% of all hematological malig-
nancies in children (Table 3) corresponding to an annual incidence of MDS of
1.8 per million children aged 0 to 14 years (2,5). MDS and JMML combined
constituted 7.7% of childhood leukemia in Japan (7) with a high proportion of
therapy-related cases (23%). Data from the United Kingdom suggest a consider-
ably lower annual incidence of MDS of 0.8 per million (Table 3) (9). The United
Kingdom study excluded secondary MDS partly explaining the lower incidence.
The incidence of RC in United Kingdom, Denmark, and BC is very similar but the
incidence of advanced MDS and JMML differ significantly. Possible differences
Table 3 Annual Incidence of Hematological Malignancies in Children 0–14 years.

Denmark and BC U.K.
N % Incidence/million Incidence/million
ALL 815 79 38.5 ND

AMLa 115 11 5.4 5.8
MDSa 38 4 1.8 0.8
Myeloid leukemia of DS 19 2 0.9 0.6
JMML 25 2 1.2 0.6
CML 13 1 0.6 0.5
PV/ETb 3 0 0.1 ND
Unclassified 3 0 0.1
Total 1030 100 48.7 -
Combined data from Denmark 1980–1991 and British Columbia 1982–1996 (2,5) and data from the
United Kingdom 1990–1999 (9).
a Excluding Down syndrome (DS).
b PV, polycythemia vera; ET, essential thrombocythemia.
in classification practice can only explain a smaller part of the variation (9) and it
is possible that there are genuine regional differences in incidence.
The male/female distribution in pediatric MDS is equal (144/146) with
a median age at presentation of 6.8 years (2,4,5,7,9). Application of the FAB
classification in 239 patients showed RA (n = 90, 38%), RARS (n = 1, 0%),
RAEB (n = 88, 37%), and RAEB-t (n = 60, 25%) (2,4,5,7,9). Many earlier
studies have included patients with Down syndrome. Myeloid leukemia in Down
syndrome is now considered a separate entity no longer considered as MDS (11).
The exclusion of Down syndrome from patient series of MDS decreases the
number of cases diagnosed as RAEB-t by approximately 50% (37). Considering
that the great majority of children with myeloid leukemia in DS are under 5 years
of age, exclusion of DS from MDS cohorts will result in an even greater reduction
of RAEB-t in young age.
Associated Abnormalities
Constitutional abnormalities are present in about 30% of childhood MDS (1–
5,7,9) (Table 4). Down syndrome has been reported in about 25% of those with a
morphologic diagnosis of MDS but should no longer be included in MDS series.
MDS has been reported in a number of constitutional cytogenetic abnormalities
other than trisomy 21, but only for trisomy 8 mosaicism is there solid evidence
for an increased risk of MDS (38). Trisomy 8 in the leukemic cells may be due
to constitutional trisomy 8 in 15% to 20% of the cases (39). Reports of MDS in
patients with Klinefelter and Turner syndrome have appeared sporadically but no
increased risk has been documented (40). Congenital malformations have been
Table 4 Abnormalities Associated with MDS in Children

Risk of MDS/
Constitutional conditions Genes involved AML (%)
Bone marrow failure syndromes

Fanconi anemia 12 autosomal, 1 X-linked 50
Severe congenital neutropenia HAX1, ELA2, GFI1, WASP 20–40
Shwachman–Diamond syndrome SBDS 30
Blackfan–Diamond anemia RPS19, RPS24, RPS17 2
Dyskeratosis congenita DKC1, TERC, TERT, NOP10 5
Amegakaryocytic thrombocytopenia MPL 10
Familial thrombocytopenia RUNX1 ?
Trisomy 8 mosaicism ? ?
Familial MDS ? ?
Acquired conditions
Prior chemotherapy/radiation
Aplastic anemia
Down syndrome is considered separately and not included in the Table.

MDS in Children 317
reported in a relatively high number of children with MDS. The multiplicity of

reported malformations is considerable and no consistent pattern is evident.
Inherited Bone Marrow Failure

MDS or AML develop in up to 50% of the patients with Fanconi anemia before
the age of 40 years (41) and is often associated with monosomy 7 and duplications
of 1q (30). It is difficult to diagnose RC in a patient with Fanconi anemia and the
definition of clonality is problematic and cytogenetic aberrations may be temporary
or reappear as new clones (42). Gain of 3q material is strongly associated with
high risk of MDS (43).
The survival of patients with severe congenital neutropenia (SCN) has
improved significantly following the introduction of granulocyte-colony stimu-
lating factor (G-CSF) treatment. Studies from the SCN International Register
have shown a 10-year cumulated risk of MDS development of 21% (44). Partial
or complete loss of chromosome 7 is found in more than half of the patients who
develop MDS. The development of MDS is mostly, but not always, preceded by
acquired mutations in the G-CSF receptor gene (45). There is no direct cause-and-
effect relationship between the development of MDS and G-CSF therapy but the
risk of MDS is highest in the less-response group of patients—40% at 10 years
(44). MDS is not seen in cyclic or idiopathic neutropenia treated with G-CSF
in a similar schedule. It is recommended to perform yearly BM examination in
SCN to search for morphological signs of MDS, cytogenetic, or molecular aber-
rations. Early hematopoietic allogeneic stem cell transplantation (HSCT) seems
promising (46) but whether it is beneficial in larger group of patients remain to be
seen.
MDS develops in 30% of those with Shwachman–Diamond syndrome (47).
MDS in Shwachman–Diamond syndrome is often associated with chromosome
7 abnormalities of which isochromosome 7q may represent a separate entity
with a long stable clinical course (48). Microarray studies have identified several
abnormal gene expression patterns in SDS BM mononuclear cells, which might
be involved in the evolution of malignant clones (49).
MDS has occasionally been described in patients with Diamond–Blackfan
anemia (5,22,50,51) and dyskeratosis congenita (52). At least two different con-
genital thrombocytopenia have been associated with MDS/AML (53,54).
BM examinations including cytogenetic studies should be performed in
all patients with BM failures syndromes and altering cytopenia to ensure early
identification of progression towards MDS (55).
Acquired Aplastic Anemia

MDS develops in 10% to 15% of those patients with aplastic anemia not treated
with HSCT (56,57). Patients diagnosed as nonsevere aplastic anemia may be over-
represented among those in whom clonal evolution is observed (58), suggesting
that some cases of MDS are misdiagnosed at presentation. MDS appears to occur
at the same rate in idiopathic- and hepatitis-associated aplastic anemia (59). MDS
may occur earlier in children than in adults, and in most cases, diagnosed within
the first 3 years from presentation (56,57). Whether prolonged treatment with the
combination of G-CSF and cyclosporine is associated with development of MDS
is a controversial issue (56,60). Shorter period of G-CSF treatment is associated
with a lower incidence of MDS (61) contrasting with the high risk of MDS in those
on long-term treatment with G-CSF and those with a poor response to G-CSF (62).
Familial MDS
Families with several members affected with MDS have been described (34,63–
65). A large proportion of the patients showed monosomy 7 or deletion 7q. Larger
studies found familial MDS in 0% to 10% of childhood MDS with monosomy 7
(1,22,66). Familial MDS does also occur without −7/7q− (1,67). Some families
show discordance for −7 (22), therefore it is uncertain whether −7 per se increases
the risk for familial cases. There are no conspicuous clinical characteristics of the
familial cases (22,65). The putatively inherited predisposing locus in familial
MDS with −7/7q− does not seem to be located on chromosome 7 (63) or to the
commonly deleted portions of 5q (68). This is in accordance with the lack of
leukemia among persons with constitutional aberrations of chromosome 7 (69)
and the different parental origin of the remaining chromosome 7 in siblings with
monosomy 7 (70).
Therapy-Related MDS
Children previously treated for another malignancy are at risk of therapy-related
MDS from 2 to 10 years (peaking at 4–5 years) following the leukemogenic ther-
apy (71–74). There seems to be two different types of therapy-related malignant
transformation (75). One is related to treatment with alkylating agents leading to
MDS after a latency period of 3 to 5 years and characterized cytogenetically by
deletions or loss of whole chromosomes. The other type is associated with the
use of epipodophyllotoxins and acquired translocations involving chromosome
11q23. Most children follow the epipodophyllotoxin model of disease with short
latency period and accelerated clinical course regardless of the involved etiologic
agent (76). Studies of polymorphism in drug-metabolizing enzymes may identify
individuals with a high genetic susceptibility to therapy-related MDS (77,78). A
higher frequency of therapy-related MDS in Japan (7) may be related to therapeutic
or ethnic factors.
Therapy-related MDS constitutes less than 5% of the patients in series of
MDS and JMML (2,5,9) with the Japanese study as a notable exception with
therapy-related disease representing 23% (79). New intensive treatment protocols
may lead to an increased risk of therapy-related diseases in the future (80).
MDS in Children 319
PATHOPHYSIOLOGY
MDS is a clonal disease arising in a progenitor cell restricted to myelopoiesis, ery-
thropoiesis, and megakaryopoiesis (81,82). The initiating events may infrequently
occur in a more immature cell involving the lymphoid cell line resulting in the
very rare progression of MDS to ALL (6,83,84). The initiating events of MDS
remain obscure, in children like in adults. Because MDS is very heterogeneous,
different mechanisms of initiation and progression of the disease are likely to
exist. Genetic damage in a pluripotent hematopoietic progenitor cell may give
rise to genetic instability with subsequent acquisition of numerous molecular and
cellular abnormalities (85). Congenital disorders with DNA repair defects like
Fanconi anemia or acquired mutations in genes maintaining genetic stability may
result in a mutator phenotype predisposing to MDS (86,87). About 30% of chil-
dren with MDS have a known constitutional disorder. It may be speculated that an
even higher proportion of the children have a congenital abnormality predisposing
them to the acquisition of genetic changes. Subsequent events, that is, mutations
in proto-oncogenes like ras, p53 or WT1, and karyotypic changes like monosomy
7, may be part of a final common pathway of disease progression (65,88,89).
Methylation studies in children with RAEB or RAEB-t have demonstrated that at
least half of the patients had hypermethylation of the p15 gene (90) or CALCA
and CDKN2B genes (91)—a frequency similar to adult MDS. The functional con-
sequences of hypermethylation and the correlation with clinical features are still
unknown.
Mutations in TP53 and FMS are found in 30% of adult MDS but the
mutations are lacking in children with MDS (92). Mutations of the NRAS proto-
oncogene represent the most frequent molecular changes in adult MDS but are
rare in children (88,93).
Mutations in TERC, the gene coding for the RNA component of telomerase,
result in autosomal dominant dyskeratosis congenita. Two large studies showed
TERC mutations in only 3 of 217 children with MDS (94,95).
Using cDNA microarray assays, a clear difference in the gene expression
pattern is observed between BM stoma cells obtained from healthy children and
from pediatric patients with either MDS or AML. The global gene function pro-
filing analysis indicated that in the pediatric MDS microenvironment the disease
stages may be characterized mainly by underexpression of genes associated with
biological processes such as transport. Furthermore, a subset of downregulated
genes related to endocytosis and protein secretion may be able to discriminate
MDS from MDS-AML (96).
CLINICAL AND LABORATORY FEATURES

The presenting features in almost all cases of MDS are those of pancytopenia.
Single lineage cytopenia may occasionally be the presenting characteristic. In
a few cases, the cytopenia is an incidental finding during a routine work-up. A
few patients have been diagnosed during evaluation as possible sibling stem cell
donor. Not all children with RC have anemia, but macrocytosis [elevated mean
corpuscular volume (MCV)] is a characteristic finding (8). Fetal hemoglobin
(HbF) is frequently moderately elevated. WBC is low to normal. Leukocytosis is
generally not a feature of MDS and in the case of increased WBC, the diagnosis
should be reconsidered. Some patients present with slight hepatosplenomegaly
but most have no organomegaly.
Extramedullary myeloid tumor may be the presenting feature of MDS
(22,97), but blasts in the cerebrospinal fluid is not seen in MDS.
BONE MARROW FEATURES

The BM may be hypo-, normo-, or hypercellular. Decreased cell content has
been observed in up to 40% of childhood RC (8). Both the PB and BM display
characteristic dysplastic features with megaloblastic erythropoiesis, bizarre small
or unusual large megakaryocytes, and dysgranulopoiesis (98). The presence of the
characteristic dysplastic features is suggestive of MDS but not diagnostic (11).
Interobserver variation in the evaluation of dysplasia exists (99) and centralized
review is recommended (98). The degree of dysplasia has prognostic relevance in
adults with RA (100), but it remains uncertain whether the same is true in children.
CYTOGENETICS
An abnormal karyotype is found in about 50% of the children (4,9,101–103).
The numerical abnormalities dominate with only 10% showing a translocation,
a derivative, or a deletion as the sole abnormality. Structural abnormalities are
frequently a part of a complex karyotype with numerical abnormalities. This is
in contrast to AML where structural abnormalities are by far the most frequent
findings (104,105).
Monosomy 7 is the most common cytogenetic abnormality in childhood
MDS seen in approximately 30% of the cases (1,101–103). After monosomy 7,
trisomy 8 and trisomy 21 are the most common numerical abnormalities. Con-
stitutional trisomy 21 is clinically obvious when present, whereas constitutional
trisomy 8 mosaicism may be clinically unrecognized (38) and should be tested for
when trisomy 8 is found in the BM.
There are only very few data on the prognostic value of cytogenetic abnor-
malities in children. Monosomy 7 as the only cytogenetic aberration has in most
studies not been an unfavorable feature in childhood MDS (4,22,23,106), whereas
complex abnormalities with or without chromosome 7 involvement are associated
with a poor outcome (22,103). This is in contrast to adults where −7/7q− is
associated with a very poor prognosis (107). However, monosomy 7 is associated
with a shorter time to progression in children with RC (8). Favorable cytogenetic
aberrations have been identified in adults as −Y, 20q−, and 5q−, these aberrations
are so infrequent in children that they are of no practical importance.
MDS in Children 321
AML-specific translocations, that is, t(8;21)(q22;q22), t(15;17)(q22;q12), or

inv(16)(p13q22) may be seen with a low blast cell count but should be considered
as AML regardless of the blast count (27,108,109).
IMMUNOPHENOTYPE
Flow cytometry immunophenotyping has not resulted the diagnostic yield in MDS
as it has in acute leukemia. A normal flow cytometry examination does not pre-
clude MDS. Flowcytometry using a pattern-recognition–based concept may serve
as a useful adjunct in the diagnostic process of difficult cases with nondiagnostic
morphology and cytogenetics (110). Immunophenotypic clustering partly discrim-
inates patients with RA from RAEB/RAEB-t (111). There is a lack of reported
data on the immunophenotype characteristics of MDS in children.
DIFFERENTIAL DIAGNOSIS
The two main diagnostic challenges are to distinguish MDS with a low blast count
from aplastic anemia and other nonclonal disorders, and to differentiate MDS with
excess of blasts from AML. The traditional classification has been based on pure
morphology but a number of additional factors need to be considered.
Refractory Cytopenia Versus Aplastic Anemia

Reduced BM cellularity is found in 40% of patients with RC (8). BM fibrosis,
dilution, and sampling variation make it difficult to assess the cellularity from
an aspirate. A trephine biopsy is therefore essential for the evaluation of a child
with suspected aplastic anemia or MDS. Hypoplastic MDS may be difficult to
discriminate from aplastic anemia. The presenting MCV is higher in MDS than
in aplastic anemia (112). Careful sequential morphologic studies including BM
biopsies will almost always establish a distinction between MDS and AA (113–
115). The biopsy in hypoplastic MDS shows scarcely scattered granulopoietic
cells, patchy islands of immature erythropoiesis and micromegakaryocytes (115).
Marked erythroid hypoplasia may also be a feature of childhood MDS (116).
Clonal hematopoiesis is strongly suggestive of MDS, when standard cytogenetics
fail FISH or HUMARA assays may be useful to establish clonality. Point mutations
of the N-RAS oncogene are frequent in MDS but have not been observed in aplastic
anemia (117). Overexpression of p53 is a useful marker of MDS (118).
MDS Versus Other Nonclonal Disorders

Dysplasia in the BM may occur in a variety of disorders of very different etiologies,
that is, infection, drug therapy, and chronic disease. RC is a diagnosis of exclu-
sion after ruling out infectious diseases like parvovirus (119,120), herpesvirus 6
(121), HIV (122), and visceral leishmaniasis (123), vitamin B12 deficiency (124),
drug therapy (125,126), rheumatoid arthritis (127), metabolic disorders (128),
Table 5 Minimal Diagnostic Criteria for MDS
At least 2 of the following:

Sustained unexplained cytopenia (neutropenia, thrombocytopenia, or anemia)
At least bilineage morphologic myelodysplasia
Acquired clonal cytogenetic abnormality in hematopoietic cells
Increased blasts (≥5%)
and other causes of cytopenia and dysplasia (129–131). It should be empha-

sized that serology may be unreliable and only PCR can exclude viral infection
(119). Nonclonal chronic disorders with dysplastic features, that is, mitochondrial
disorders like Pearson syndrome, should not be considered as MDS. RARS is
extremely rare in children. The finding of sideroblastic anemia should prompt
investigation for possible mitochondrial cytopathy or disorders of heme synthesis
(132).
It may be difficult to diagnose MDS in children who have a low blast cell
count and no clonal marker. The minimal diagnostic criteria listed in Table 5 may
help in this situation (11). The WHO classification has a definition of “minimal
dysplasia being at least 10% of the cell lineage (27),” it is a less useful criterion for
children where marked dysplasia may be observed in reactive conditions and MDS
may present with discrete dysplasia. The length of persistent cytopenia should be
at least 1 month.
Since hematopoiesis is often dysplastic in patients with congenital BM-
failure disorders, it is recommended to diagnose MDS in these patients only if
the BM blast count is increased, a persistent clonal chromosomal abnormality
is present, or the BM becomes hypercellular in the presence of persistent PB
cytopenia (11). The finding of sideroblastic anemia should prompt investigation
for possible mitochondrial cytopathy or disorders of heme synthesis (133).
Separating MDS from AML

AML is the major differential diagnosis of MDS. There are significant differences
in clinical features, cytogenetics, and in response to therapy between MDS
and AML (101,134,135) reflecting fundamental biologic differences (136),
thus making the morphologically based classification a surrogate marker for
the distinction between biological entities (Table 6). Whether the proposed
redefinition of AML by 20% blasts may be useful in pediatrics remains untested.
A British study suggested a better outcome following AML therapy in patients
with RAEB-t compared with RAEB (137), however, this was not found in an
American study (23). The European Working Group on MDS in childhood
(EWOG-MDS) experiences showed poor response to chemotherapy in RAEB-t
and no benefit from chemotherapy before HSCT (138). The conflicting data
reflect that the morphologically defined RAEB-t group is heterogeneous and blast
MDS in Children 323
Table 6 Major Differences Between MDS and AML in Children

MDS AML
WBC Low-normal Low-normal-high

Hepatomegaly Infrequent Common
Cytogenetic aberrations Numerical (−7) Structural
Dysplasia Multilineage Infrequent
Hematopoiesis Clonal (including CR) Nonclonal
Cell of origin Stem cell Lineage-restricted
Response to chemotherapy Poor Moderate
Iatrogenic model Alkylating agents Epipodophyllotoxins
count in a single specimen is insufficient to differentiate MDS from AML. It may

be useful to retain the RAEB-t category for the group of patients with clinical and
biological MDS, recognizing, however, that biological features rather than any
arbitrary cut-off in blast count may be more important in distinguishing MDS from
(chemosensitive) AML (139). An algorithm to facilitate the distinction between
MDS and AML is presented in Figure 1. Monosomy 7 is strongly suggestive
of MDS (22) and patients presenting with monosomy 7 and a blast count above
30% may share many features with MDS rather than with true de novo AML
(140).
In borderline cases with BM blasts of 20% to 30% and no cytogenetic clues
to the diagnosis, it is recommended to repeat the BM examination after 2 weeks.
If the blast count has increased to more than 30%, the case should be regarded
as AML (Fig. 1). Significant organomegaly or increased WBC are suggestive of
a diagnosis of AML. It should be emphasized that most children with myeloid
malignancies have clear-cut AML, some have MDS with low blast count and only
AML
t(8;21),
t(15;17), PB/BM BM WBC >15–20
blasts Repeat BM
–7 blasts
inv(16), Organomegaly after 2 wk
>30% <20%
t(9;11)
MDS
Figure 1 Algorithm for distinguishing MDS from AML.
a few percentages have borderline features. The major diagnostic pitfall may be
associated with undue haste in starting therapy.
PROGNOSIS AND NATURAL COURSE

Progression of MDS to MDS-related AML may occur at different rates. The
percentage of blast cells in the BM may increase abruptly after a set of transforming
events. Some patients show a steady rate of progression and the 20% or 30%
threshold is passed only after a period of months or years. Although any case
above the threshold is conventionally defined as AML, cases with dysplasia and
progression to a higher percentage of blasts are better described as persistent MDS
or MDS-related AML rather than AML Head, 2002 2687 /id}.
Children with RC and RAEB or even RAEB-t may show a long and stable
clinical course without treatment. Blood transfusions are only required infre-
quently and severe infections are rarely seen. The condition may smolder with
unchanged cytopenia for months or even years but will eventually progress in
virtually all patients. In a series of 67 children with primary RC, four died from
complications of pancytopenia prior to therapy or progression and 20 progressed
to more advanced MDS at a median of 1.7 years from presentation (8). Although
RC with monosomy 7 is associated with a higher risk of progression, both RC and
RAEB patients with monosomy 7 may show stable disease without treatment for
several years (8,22). Once progression has occurred, the outcome is inferior even
after HSCT (8,141).
The International Prognostic Scoring System (IPSS) for MDS weighted data
on BM blasts count, cytopenia, and cytogenetics, and separated patients into four
prognostic groups (107). Children have more often high-risk features compared
with adults (Table 7) (142). Thrombocytopenia and BM blasts ⬎5% correlated
with poor survival in children, whereas 2 to 3 lineage cytopenia and cytogenetics
did not provide prognostic information (142). This contrasts with data from Japan
(7) and the United Kingdom (9), showing the cytogenetic component of the IPSS
to be significantly associated with outcome due to a poor prognosis in patients
with monosomy 7. Similar to EWOG-MDS, the Japanese study showed a poor
Table 7 Distribution and Overall Survival of Children (142) and Adults (107) with
MDS in the four IPSS Groups
Children Adults
IPSS group N = 142 (%) Median survival (yr) N = 816 (%) Median survival (yr)
Low 7 ⬎10 33 5.7

Intermediate-1 47 9.7 38 3.5
Intermediate-2 25 4.5 22 1.2
High 21 2.2 7 0.4
MDS in Children 325
outcome in those presenting with BM blasts ⬎5%. Overall, the IPSS provides
little diagnostic information in children but identifies a very small group (7%) of
the patients with low-risk disease and a very favorable outcome (142).
A pediatric prognostic scoring system (FPC) proposed by the British group1
assigned one point each for HbF ⬎10%, platelets ⬍40 × 109 /L, and two or more
cytogenetic abnormalities. A significantly higher survival was found in children
with MDS and a score of zero. Application of the scoring system in other series
has been hampered by HbF being available in only a minority of the MDS patients.
Data from EWOG-MDS was used to evaluate the FPC score in 65 patients with
complete data and showed that complex karyotype was the only factor associated
with a poor survival (142) and confirmed in a larger series (103).
Spontaneous regression of MDS has occasionally been reported in the litera-
ture (143–146). The frequency of spontaneous remission is unknown, but estimated
to occur in well below 5% of the patients.
TREATMENT
MDS is a clonal early stem cell disorder with very limited residual nonclonal
stem cells. Myeloablative therapy is therefore the only treatment option with a
realistic curative potential in a significant portion of the patients. A diversity
of therapy strategies like hematopoietic growth factors, differentiating agents,
hormones, amifostine, low-dose cytotoxic drugs, or experimental agents have
been investigated in adults and in the elderly not candidates for HSCT. None of
these approaches have been documented to prolong survival and they are generally
not indicated in children and adolescents. Given the lack of recurrent molecular
abnormalities in MDS, rational drug development aiming at molecular targeted
therapy is problematic.
Immunosuppressive therapy has been successful in some adults with MDS
and low blast count, especially in patients with BM hypoplasia and HLA-DR15
(DR2) (147). Other studies have been less optimistic reporting a significant burden
of side effects (148). Immunosuppressive therapy with antithymocyte globulin
and cyclosporine in 31 children with hypoplastic RC resulted in a complete or
partial response in 22 of 29 evaluable patients at 6 months. Overall and failure-
free survival rates at 3 years were 88% and 57%, respectively (149). The long-
term outcome of immunosuppressive therapy in MDS is not known. High-dose
methylprednisolone has been used with some success by the Turkish group (150)
but the approach has not been studied in other series.
DNA methyltransferase inhibitors, azacitidine, and decitabine, have shown
clinical efficacy in adults with MDS. Hypermethylation may occur at a similar
frequency in children and adults (90,91), thus making children potential candidates
for methyltransferease inhibitor therapy, however, so far treatment results from
pediatrics are lacking. Children with MDS are at high risk of cytopenia-related
complications and optimal supportive care should be the primary focus during all
phases of the disease course.
AML-Type Chemotherapy
Conventional intensive chemotherapy without HSCT is unlikely to eradicate the
primitive pluripotent cells involved in MDS, rendering the therapy noncurative in
most patients, although reported results are somewhat conflicting. Most studies
found a significant morbidity and mortality of induction chemotherapy with a
complete remission rate of less than 60%, many relapses, and overall survival less
than 30% (7,23,108,135). The treatment-related mortality rate has been between
10% and 30% (23,108,135,137). A few studies have reported an outcome in
MDS patients not significantly different from that in AML (137,151) especially
in patients with RAEB-t or AML following MDS (23,137). The results reflect
the heterogeneous nature of RAEB-t and emphasis that a single morphologically
evaluation is insufficient for relevant treatment stratification (11).
Autologous SCT is often used in younger adults (152) but has only infre-
quently been reported in children. Two studies included eight children receiving
autologous SCT with one long-term survivor (23,137).
Hematopoietic Allogeneic Stem Cell Transplantation

HSCT is the therapy of choice for virtually all forms of MDS in childhood.
Studies specifically addressing the question of HSCT in children have indicated
a probability of disease-free survival (DFS) following transplantation with an
HLA-matched family donor (MFD) of about 50% (153–159). Children receiving
a graft from an HLA-matched unrelated donor (MUD) have previously suffered a
higher transplant-related mortality (TRM) and lower DFS, but more recent studies
have shown survival following MUD-HSCT comparable to that of MFD-HSCT
(160–162).
The European Group for Blood and Marrow Transplantation (EBMT)
reported 3-year DFS, TRM, and relapse risk of 45%, 30%, and 36%, respectively
for MDS patients less than 20 years of age who received MFD-HSCT between
1983 and 1998 (163). EWOG-MDS has successfully studied a large number of
patients with a preparative regimen of busulfan, cyclophosphamide, and melpha-
lan (155). The regimen has been studied in a large number of patients under the
auspices of EWOG-MDS and EBMT showing DFS of 87% in MFD-HSCT and
41% in MUD-HSCT (164). Other large studies have confirmed a favorable out-
come in patients conditioned with a myeloablative busulfan–based regimen (165).
TBI can generally be omitted since it is known to have no superior an-leukemic
efficacy compared with busulfan (141,166) and is associated with more long-term
effects in children.
Stage of disease as indicated by FAB type has a significant effect on relapse
and outcome following HSCT (141,167–169) with a low relapse rate in RC. HSCT
early in the course of the disease has therefore been recommended for all children
and adolescents with MDS. However, in children with RC and absence of profound
cytopenia postponement of HSCT with a watch-and-wait strategy may be justified
MDS in Children 327
especially in patients with a normal karyotype (8). Following a myeloablative

preparative regimen in RC, EWOG-MDS reported a 5-year DFS of 78% and
76% for children receiving a MFD-HSCT or MUD-HSCT, respectively (166).
Reduced-intensity conditioning regimen followed by allogeneic PB HSCT showed
a promising 1-year progression-free survival of 66% in adults (170). A fludarabine-
based preparative regimen in 19 children with RC and normal karyotype resulted
in an overall survival and DFS at 3 years of 84% and 74%, respectively (171),
being comparable to those of patients treated with myeloablative HSCT (166).
Long-term follow-up is needed to demonstrate the expected reduction in long-
term sequelae with reduced-intensity condition.
For advanced MDS intensive GvHD prophylaxis including in vitro T-cell-
depleted (169) is associated with an increased risk of relapse. Whether PB as a
source of stem cells source is followed by improved long-term survival remains
a matter of discussion (162,172). It also remains unknown whether AML-type
induction chemotherapy prior to HSCT for advanced MDS can reduce relapse and,
thus improve DFS. An analysis from EWOG-MDS on 53 children with primary
advanced MDS showed no benefit of intensive AML-type therapy preceding HSCT
(138). Small series of patients transplanted as first line therapy have shown survival
of 65% to 70% (135,156,173). Considering the significant morbidity and mortality
of induction chemotherapy and the high rate of TRM following HSCT (174), most
children with MDS may benefit from HSCT as first-line therapy sparing the
toxicity related to induction chemotherapy. Children without a matched donor and
progressive disease should be considered for haploidentical HSCT (175).
Relapse following HSCT is associated with a very grave outcome. Success-
ful donor leukocyte infusions have occasionally been reported (176). Especially
early relapse detected by increasing mixed chimerism may benefit from with-
drawal of immunosuppressive therapy and donor leukocyte infusion (177–179).
Close analyses of chimerism status post-SCT may be indicated to initiate pre-
emptive immunotherapy (180). Increasing WT1 expression may also be a useful
marker to monitor impending relapse post-SCT (181).
Secondary MDS
Children with MDS secondary to chemo- or radiation therapy generally have a
very poor survival. AML-type therapy may induce remission but very few patients
remain in remission and even HSCT has been reported to offer cure to only 20% to
30% of patients (76,159,182–185). The Children’s Cancer Group (CCG) reported
a superior, although still poor, outcome on intensive-timing versus standard-timing
induction (32% vs. 0%) (76). The frequency of severe treatment-related toxicity is
increased (183,186), while the risk of relapse may be similar to that observed for
patients with primary MDS (163). Recent data document a significant improve-
ment over time and survival being strongly related to cytogenetics (187).
The few published cases of HSCT in MDS arising from congenital BM
failure disorders or acquired aplastic anemia indicate a poor outcome for this
heterogeneous group of patients. Early HSCT before neoplastic transformation or

during less-advanced MDS may be associated with improved survival (46,188).
Cooperative studies like those of EWOG-MDS and the EBMT are needed to
provide further information on the appropriate timing, conditioning regimen, and
GvHD prophylaxis for the different subtypes of MDS in childhood.
MYELOID LEUKEMIA AND DOWN SYNDROME

Individuals with Down syndrome (DS) have a strong age-related increased risk of
leukemia with a more than 50-fold increased risk during the first 5 years of life
and a 10-fold increased risk from 5 to 29 years of age. After age 30, the risk of
leukemia is close to that seen in non-DS individuals (189). The cumulated risk for
leukemia by the age of 5 years is 2.1% and that by 30 years is 2.7%. Almost half
the leukemias are myeloid and most of them occur before 5 years of age where
the relative risk of myeloid leukemia is increased by more than 150 fold (189). In
addition to the myeloid leukemia, a myeloproliferative disorder indistinguishable
from leukemia may occur in infants with DS. Children with DS who have been
cured from leukemia appear to have a reduced risk of secondary malignancies
(190) but the increased risk of both myeloid and lymphoid leukemia may result in
both leukemias in the same patient as independent events (191).
Transient Abnormal Myelopoiesis

Increased WBC with circulating blasts often accompanied by anemia and throm-
bocytopenia may be seen in up to 10% of newborns with DS. The blast cells almost
invariably have cell-surface antigens characteristic of megakaryoblasts (192). The
percentage of blasts is often higher in blood than in BM, and a BM aspiration
in most cases is of limited additional diagnostic value. Clonal abnormalities are
observed in 35% of the patients (193). The condition is referred to as transient
abnormal myelopoiesis (TAM) or transient myeloproliferative disorder. The pre-
sentation is indistinguishable from leukemia and some have therefore favored
the name transient leukemia (192,193). Transient leukemia may occasionally
occur in an infant with normal phenotype and trisomy 21 in the blast cells (194).
An unknown portion of these infants may have low-level constitutional trisomy
21 mosaic detectable by FISH only (195). TAM involves selectively the trisomic
cells in individuals with DS mosaic (196).
Life-threatening complications, mainly progressive hepatic dysfunction,
may occur in 10% to 20% of the patients with TAM, but spontaneous remission
appears in the majority within 1 to 3 months (193). Generally, no chemotherapy
is indicated in TAM, however, in those with progressive hepatic or pulmonary
problems, a short course of low-dose cytarabine may be very effective (197).
Myeloid leukemia develops 1 to 3 years later in about 25% of the children who
have recovered from TAM (192,193). The development of subsequent myeloid
MDS in Children 329
leukemia is associated with acquired clonal cytogenetic abnormalities (193) and

persistent elevated WT1 expression (198).
Myeloid Leukemia
Myeloid leukemia in DS has often been classified as AML (199) despite BM blasts
less than 30% in many patients (5). Of those with a morphological diagnosis of
RC, RAEB, or RAEB-t, 25% have DS (2,5,37). In contrast to non-DS children,
there are no biological or therapeutic differences between MDS and AML in DS.
Recognizing the unique biological features the disease may be best described by
the unifying term myeloid leukemia of Down syndrome (11). ML-DS is preferred
to acute megakaryoblastic leukemia (AKML) because other phenotypes in DS
are observed sharing the same biologic and clinical characteristics. It is no longer
appropriate to use the terms MDS or AML (AKML) in young children with DS.
Myeloid leukemia in older DS children (4 years or older) tend to be GATA1
negative and has a higher risk of relapse (200,201). Such patients may represent
spontaneous AML not fulfilling the criteria for ML-DS.
EPIDEMIOLOGY
ML-DS develops in 1% to 2% of children with DS (189) corresponding to an
annual incidence of 0.6 to 1.0 per million children (2,5,9) (Table 3). The age
distribution is very unusual with 49% being 1 year of age at diagnosis, 34%
2 years of age, and only 2% more than 4 years of age (190). Only very few present
before 1 year of age and there appears to be no age overlap between TAM and
ML-DS (199,202–207).
PATHOBIOLOGY
Leukemia in children with trisomy 21 mosaicism selectively involves the trisomic
cells (208,209) pointing at the etiological role of the additional chromosome 21 as
the first hit in the multistep process leading to leukemia. All patients with ML-DS
have an acquired mutation in the GATA1 gene (210). The mutation is not found
in AML-M7 in non-DS or in other AML patients. The GATA1 mutation is also
found in patients with TAM (211,212). The GATA1 gene encodes a transcription
factor essential for the normal erythroid and megakaryocytic differentiation in
accordance with the selective involvement of these two lineages rather than granu-
locytic lineage in myeloid leukemia of DS (213). The studies of GATA1 mutations
support the notion of myeloid leukemia of DS as a separate entity.
A model of the pathogenic steps in myeloid leukemia of DS is presented
in Figure 2. The trisomy 21 is the first event that may predispose the cells to
a proliferative advantage or further mutations. GATA1 mutation is found in the
majority of patients with TAM and may be present in 3% to 4% of newborns with
DS and normal hematology (214). The mechanisms of the regression of TAM
1st event 2nd event 3rd event
25% Myeloid
TAM Regression
leukemia
GATA1
5% ?
mutation
1% Myeloid
Normal
leukemia
Liver hematopoiesis Bone marrow hematopoiesis
Conception Birth Age 1–3 years
Figure 2 Pathogenetic model for myeloid leukemia in Down syndrome.
remain unexplained but may be associated with a decreased telomerase activity in

TAM not found in ML-DS (215). About 25% of those with TAM and about 1%
of DS with normal hematology in the newborn period develop ML-DS (189).
CLINICAL AND LABORATORY FEATURES

Isolated thrombocytopenia is often the presenting feature of ML-DS. At diagnosis,
both platelet count and WBC are lower than in non-DS patients (199), in contrast
to the very high WBC seen in TAM. The blast cells have in most cases morphologic
and antigen features of megakaryoblasts, although other morphological variants
may occur. Many patients have a relatively indolent course characterized by a
period of thrombocytopenia and dysplasia with relatively few blasts in the BM.
CYTOGENETICS
Numerical aberrations, mainly trisomy 8 and an extra chromosome 21 (tetrasomy
21), are the most common acquired cytogenetic abnormalities (216). Monosomy
7 is very common in MDS but very uncommon in patients with DS (4,22,213).
Karyotype is not known to be a prognostic factor in DS. The clonal cells in children
with DS are myeloid progenitors with the potential for differentiation along the
megakaryocytic and erythroid lineages (213), the granulocytic lineage is in most
cases, in contrast to non-DS children, not involved in the leukemic process.
TREATMENT
In contrast to TAM, ML-DS is fatal if untreated but responds well to AML treat-
ment with a very favorable outcome (199,217,218). The prognosis of myeloid
MDS in Children 331
leukemia in DS was considered very poor before 1990. Reports from the Nordic
Society of Paediatric Haematology and Oncology (NOPHO) (219) and the Pedi-
atric Oncology Group (POG) (217) and later the CCG (199) showed a surprisingly
high survival rate for DS patients receiving AML treatment. DS was later shown
to be the most important prognostic factor in AML (218). Several groups have
reported long-term survival in DS patients well above 80% (203,205–207,220).
The prognosis for ML-DS has improved significantly during the last 10 to 15
years. The main explanation for the improved survival is the relatively large frac-
tion of patients, diagnosed before 1990, not treated adequately. DS patients treated
on AML protocols have a significantly better outcome than those receiving mini-
mal treatment (197), however, intensive timing of induction is associated with an
increased mortality (199,205). DS children are at a low risk for relapse, and due
to the high risk for treatment-related toxicity, they benefit from less time-intensive
therapy allowing recovery prior to initiation of the next chemotherapeutic course
(199,205). HSCT is associated with excess toxicity without therapeutic gain and
is not indicated in ML-DS (199,221). It is recommended to start therapy when the
myeloid disorder is diagnosed and not to await progression (197).
DS myeloblasts are 10 fold more sensitive to cytarabine in vitro than non-
DS cells (222,223). The increased sensitivity of DS blasts may be related to the
expression of chromosome 21 localized genes like cystathionine-␤-synthetase
and superoxide dismutase (222). An elevated cystathionine-␤-synthetase activ-
ity may modulate cytarabine metabolism by decreasing levels of deoxycytidine
triphosphate or decreasing generation of S-adenosyl-methionine and hypomethy-
lation of the deoxycytidine kinase gene (222). The cystathionine-␤-synthetase
gene polymorphism (844ins68) is more frequently observed in DS myeloblasts
than in non-DS myeloblasts and those DS patients with the polymorphism have
an increased cytarabine sensitivity compared with those with the wild-type gene
(224).
It is remarkable that only the constitutional and not the acquired trisomy
21 is associated with a superior outcome (199). Further studies of the molecular
mechanism of the increased sensitivity to chemotherapy in DS may lead to new
approaches in the treatment of AML.
REFERENCES
children and a new prognostic scoring system. Blood 1995; 85:1742–1750.
2. Hasle H, Kerndrup G, Jacobsen BB. Childhood myelodysplastic syndrome in Den-
mark: Incidence and predisposing conditions. Leukemia 1995; 9:1569–1572.
3. Bader-Meunier B, Mielot F, Tchernia G, et al. Myelodysplastic syndrome in child-
hood: Report of 49 patients from a French multicenter study. Br J Haematol 1996;
92:344–350.
4. Luna-Fineman S, Shannon KM, Atwater SK, et al. Myelodysplastic and myelopro-
liferative disorders of childhood: A study of 167 patients. Blood 1999; 93:459–466.
5. Hasle H, Wadsworth LD, Massing BG, et al. A population-based study of child-

hood myelodysplastic syndrome in British Columbia, Canada. Br J Haematol 1999;
106:1027–1032.
6. Lopes LF, Lorand Metze I. Childhood myelodysplastic syndromes in a Brazilian
population. Pediatr Hematol Oncol 1999; 16:347–353.
7. Sasaki H, Manabe A, Kojima S, et al. Myelodysplastic syndrome in childhood: A
retrospective study of 189 patients in Japan. Leukemia 2001; 15:1713–1720.
8. Kardos G, Baumann I, Passmore SJ, et al. Refractory anemia in childhood: A
retrospective analysis of 67 patients with particular reference to monosomy 7. Blood
2003; 102:1997–2003.
9. Passmore SJ, Chessells JM, Kempski H, et al. Paediatric MDS and JMML in the
UK: A population based study of incidence and survival. Br J Haematol 2003;
121:758–767.
10. Elghetany MT. Myelodysplastic syndromes in children: A critical review of issues in
the diagnosis and classification of 887 cases from 13 published series. Arch Pathol
Lab Med 2007; 131(7):1110–1116.
classification of myelodysplastic and myeloproliferative diseases. Leukemia 2003;
17(2):277–282.
12. Steliarova-Foucher E, Stiller C, Lacour B, et al. International Classification of Child-
hood Cancer, third edition. Cancer 2005; 103(7):1457–1467.
13. Humbert JR, Hathaway WE, Robinson A, et al. Pre-leukemia in children with
a missing bone marrow C chromosome and a myeloproliferative disorder. Br J
Haematol 1971; 21:705–716.
14. Kleihauer E. The preleukemic syndromes (hematopoietic dysplasia) in childhood.
Eur J Pediatr 1980; 133:5–10.
15. Bernard J, Schaison G. Transitory bone marrow failure. A series of 13 preleukemic
cases in children. Am J Pediatr Hematol Oncol 1980; 2:141–144.
16. Wegelius R. Bone marrow dysfunctions preceding acute leukemia in children: A
clinical study. Leuk Res 1992; 16:71–76.
17. Hasle H, Heim S, Schroeder H, et al. Transient pancytopenia preceding acute lym-
phoblastic leukemia (pre-ALL). Leukemia 1995; 9:605–608.
18. Weiss K, Stass S, Williams D, et al. Childhood monosomy 7 syndrome: Clinical and
in vitro studies. Leukemia 1987; 1:97–104.
19. Evans JP, Czepulkowski B, Gibbons B, et al. Childhood monosomy 7 revisited. Br
J Haematol 1988; 69:41–45.
20. Hann IM. Myelodysplastic syndromes. Arch Dis Child 1992; 67:962–966.
21. Baranger L, Baruchel A, Leverger G, et al. Monosomy-7 in childhood hemopoietic
disorders. Leukemia 1990; 4:345–349.
22. Hasle H, Aricò M, Basso G, et al. Myelodysplastic syndrome, juvenile myelomono-
cytic leukemia, and acute myeloid leukemia associated with complete or partial
monosomy 7. Leukemia 1999; 13:376–385.
23. Woods WG, Barnard DR, Alonzo TA, et al. Prospective study of 90 children requiring
treatment for juvenile myelomonocytic leukemia or myelodysplastic syndrome: A
report from the Children’s Cancer Group. J Clin Oncol 2002; 20:434–440.
MDS in Children 333
25. Cantù-Rajnoldi A, Porcelli P, Cattoretti G, et al. Myelodysplastic syndromes in

children: Observation on five cases. Eur Paediatr Haematol Oncol 1984; 1:71–
75.
26. van Wering ER, Kamps WA, Vossen JM, et al. Myelodysplastic syndromes in child-
hood: Three case reports. Br J Haematol 1985; 60:137–142.
27. Jaffe ES, Harris NL, Stein H, Vardiman JW, eds. World Health Organization Clas-
sification of Tumours. Pathology and Genetics of Tumours of Haematopoietic and
Lymphoid Tissues. Lyon: IARC Press, 2001.
28. Mandel K, Dror Y, Poon A, et al. A practical, comprehensive classification for
pediatric myelodysplastic syndromes: The CCC System. J Pediatr Hematol Oncol
2002; 24(7):596–605.
29. Occhipinti E, Correa H, Yu L, et al. Comparison of two new classifications for
pediatric myelodysplastic and myeloproliferative disorders. Pediatr Blood Cancer
2005; 44(3):240–244.
30. Maarek O, Jonveaux P, Le Coniat M, et al. Fanconi anemia and bone marrow clonal
chromosome abnormalities. Leukemia 1996; 10:1700–1704.
31. Blank J, Lange B. Preleukemia in children. J Pediatr 1981; 98:565–568.
32. Lie SO, Berglund G, Gustafsson G, et al. High-dose Ara-C as a single-agent con-
solidation therapy in childhood acute myelogenous leukemia. Hamatol Bluttransfus
1990; 33:215–221.
33. Woods WG, Kobrinsky N, Buckley J, et al. Intensively timed induction therapy
followed by autologous or allogeneic bone marrow transplantation for children with
acute myeloid leukemia or myelodysplastic syndrome: A Childrens Cancer Group
pilot study. J Clin Oncol 1993; 11:1448–1457.
34. Brandwein JM, Horsman DE, Eaves AC, et al. Childhood myelodysplasia: Sug-
gested classification as myelodysplastic syndromes based on laboratory and clinical
findings. Am J Pediatr Hematol Oncol 1990; 12:63–70.
35. Creutzig U, Cantù-Rajnoldi A, Ritter J, et al. Myelodysplastic syndromes in child-
hood. Report of 21 patients from Italy and West Germany. Am J Pediatr Hematol
Oncol 1987; 9:324–330.
36. Tuncer MA, Pagliuca A, Hicsönmez G, et al. Primary myelodysplastic syndrome
in children: The clinical experience in 33 cases. Br J Haematol 1992; 82:347–
353.
37. Stary J, Baumann I, Creutzig U, et al. Getting the numbers straight in pediatric MDS:
Distribution of subtypes after exclusion of Down syndrome. Pediatr Blood Cancer
2008; 50(2):435–436.
38. Hasle H, Clausen N, Pedersen B, et al. Myelodysplastic syndrome in a child with
constitutional trisomy 8 mosaicism and normal phenotype. Cancer Genet Cytogenet
1995; 79:79–81.
39. Maserati E, Aprili F, Vinante F, et al. Trisomy 8 in myelodysplasia and acute leukemia
is constitutional in 15–20% of cases. Genes Chromosome Cancer 2002; 33:93–97.
40. Hasle H. Turner syndrome and myelodysplastic syndrome. No reason to alert. J
Pediatr Hematol Oncol 1997; 19:179.
41. Butturini A, Gale RP, Verlander PC, et al. Hematologic abnormalities in Fanconi
anemia: An international Fanconi anemia registry study. Blood 1994; 84:1650–1655.
42. Alter BP, Caruso JP, Drachtman RA, et al. Fanconi anemia: myelodysplasia as a
predictor of outcome. Cancer Genet Cytogenet 2000; 117(2):125–131.
43. Tonnies H, Huber S, Kuhl JS, et al. Clonal chromosome aberrations in bone marrow
cells of Fanconi anemia patients: Gains of the chromosomal segment 3q26q29 as an
adverse risk factor. Blood 2003; 101(10):3872–3874.
44. Rosenberg PS, Alter BP, Bolyard AA, et al. The incidence of leukemia and mortality
from sepsis in patients with severe congenital neutropenia receiving long-term G-
CSF therapy. Blood 2006; 107(12):4628–4635.
45. Germeshausen M, Ballmaier M, Welte K. Incidence of CSF3R mutations in severe
congenital neutropenia and relevance for leukemogenesis: Results of a long-term
survey. Blood 2007; 109(1):93–99.
46. Zeidler C, Welte K, Barak Y, et al. Stem cell transplantation in patients with severe
congenital neutropenia without evidence of leukemic transformation. Blood 2000;
95(4):1195–1198.
47. Smith OP. Shwachman–Diamond syndrome. Semin Hematol 2002; 39(2):95–
102.
48. Cunningham J, Sales M, Pearce A, et al. Does isochromosome 7q mandate bone
marrow transplant in children with Shwachman–Diamond syndrome? Br J Haematol
2002; 119(4):1062–1069.
49. Rujkijyanont P, Beyene J, Wei K, et al. Leukaemia-related gene expression in bone
marrow cells from patients with the preleukaemic disorder Shwachman–Diamond
syndrome. Br J Haematol 2007; 137(6):537–544.
50. van Dijken PJ, Verwijs W. Diamond–Blackfan anemia and malignancy. A case report
and a review of the literature. Cancer 1995; 76:517–520.
51. Lipton JM, Atsidaftos E, Zyskind I, et al. Improving clinical care and elucidating
the pathophysiology of Diamond Blackfan anemia: An update from the Diamond
Blackfan Anemia Registry. Pediatr Blood Cancer 2006; 46(5):558–564.
52. Dokal I. Dyskeratosis congenita in all its forms. Br J Haematol 2000; 110:768–
779.
53. Ganly P, Walker LC, Morris CM. Familial mutations of the transcription factor
RUNX1 (AML1, CBFA2) predispose to acute myeloid leukemia. Leuk Lymphoma
2004; 45(1):1–10.
54. Rheingold SR. Acute myeloid leukemia in a child with hereditary thrombocytopenia.
Pediatr Blood Cancer 2007; 48(1):105–107.
55. Gohring G, Karow A, Steinemann D, et al. Chromosomal aberrations in congen-
ital bone marrow failure disorders—An early indicator for leukemogenesis? Ann
Hematol 2007; 86(10):733–739.
56. Ohara A, Kojima S, Hamajima N, et al. Myelodysplastic syndrome and acute myel-
ogenous leukemia as a late clonal complication in children with acquired aplastic
anemia. Blood 1997; 90:1009–1013.
57. Führer M, Rampf U, Burdach S, et al. Immunosuppresive therapy and bone marrow
transplantation for aplastic anemia in children: Results of the study SAA 94. Blood
1998; 92:156a.
58. Führer M, Rampf U, Baumann I, et al. Immunosuppressive therapy for aplastic
anemia in children: A more severe disease predicts better survival. Blood 2005;
106(6):2102–2104.
59. Ohara A, Kojima S, Okamura J, et al. Evolution of myelodysplastic syndrome and
acute myelogenous leukaemia in children with hepatitis-associated aplastic anaemia.
Br J Haematol 2002; 116(1):151–154.
MDS in Children 335
60. Locasciulli A, Arcese W, Locatelli F, et al. Treatment of aplastic anaemia with

granulocyte-colony stimulating factor and risk of malignancy. Lancet 2001; 357:43–
44.
61. Kojima S, Hibi S, Kosaka Y, et al. Immunosuppressive therapy using antithy-
mocyte globulin, cyclosporine, and danazol with or without human granulocyte
colony-stimulating factor in children with acquired aplastic anemia. Blood 2000;
96(6):2049–2054.
62. Kojima S, Ohara A, Tsuchida M, et al. Risk factors for evolution of acquired aplastic
anemia into myelodysplastic syndrome and acute myeloid leukemia after immuno-
suppressive therapy in children. Blood 2002; 100(3):786–790.
63. Shannon KM, Turhan AG, Chang SS, et al. Familial bone marrow monosomy 7.
Evidence that the predisposing locus is not on the long arm of chromosome 7. J Clin
Invest 1989; 84:984–989.
64. Gilchrist DM, Friedman JM, Rogers PC, et al. Myelodysplasia and leukemia syn-
drome with monosomy 7: A genetic perspective. Am J Med Genet 1990; 35:437–441.
65. Luna-Fineman S, Shannon KM, Lange BJ. Childhood monosomy 7: Epidemiology,
biology, and mechanistic implications. Blood 1995; 85:1985–1999.
66. Hasle H, Olsen JH. Cancer in relatives of children with myelodysplastic syndrome,
acute and chronic myeloid leukaemia. Br J Haematol 1997; 97:127–131.
67. Mijovic A, Antunovic P, Pagliuca A, et al. Familial myelodysplastic syndromes: A
key to understanding leukaemogenesis? Leuk Res 1997; 21(Suppl. 1):S6.
68. Gao Q, Horwitz M, Roulston D, et al. Susceptibility gene for familial acute myeloid
leukemia associated with loss of 5q and/or 7q is not localized on the commonly
deleted portion of 5q. Genes Chromosom Cancer 2000; 28(2):164–172.
69. Hasle H, Olsen JH, Hansen J, et al. Occurrence of cancer in a cohort of 183 persons
with constitutional chromosome 7 abnormalities. Cancer Genet Cytogenet 1998;
105:39–42.
70. Minelli A, Maserati E, Giudici G, et al. Familial partial monosomy 7 and myelodys-
plasia. Different parental origin of the monosomy 7 suggests action of a mutator
gene. Cancer Genet Cytogenet 2001; 124(2):147–151.
71. Pui CH, Ribeiro RC, Hancock ML, et al. Acute myeloid leukemia in children treated
with epipodophyllotoxins for acute lymphoblastic leukemia. N Engl J Med 1991;
325:1682–1687.
72. Rubin CM, Arthur DC, Woods WG, et al. Therapy-related myelodysplastic syn-
drome and acute myeloid leukemia in children: Correlation between chromosomal
abnormalities and prior therapy. Blood 1991; 78:2982–2988.
73. Hayani A, Mahoney DHJ, Taylor LD. Therapy-related myelodysplastic syndrome in
children with medulloblastoma following MOPP chemotherapy. J Neurooncol 1992;
14:57–62.
74. Levine EG, Bloomfield CD. Leukemias and myelodysplastic syndromes secondary
to drug, radiation, and environmental exposure. Semin Oncol 1992; 19:47–84.
75. Pedersen-Bjergaard J, Christiansen DH, Andersen MK, et al. Causality of myelodys-
plasia and acute myeloid leukemia and their genetic abnormalities. Leukemia 2002;
16(11):2177–2184.
76. Barnard DR, Lange B, Alonzo TA, et al. Acute myeloid leukemia and myelodysplas-
tic syndrome in children treated for cancer: Comparison with primary presentation.
Blood 2002; 100:427–434.
77. Thomsen JB, Schroeder H, Kristinsson J, et al. Possible carcinogenic effect of 6-

mercaptopurine on bone marrow stem cells: Relation to thiopurine metabolism.
Cancer 1999; 86(6):1080–1086.
78. Davies SM. Therapy-related leukemia: Is the risk life-long and can we identify
patients at greatest risk. J Pediatr Hematol Oncol 2000; 22(4):302–305.
79. Sasaki H, Manabe A, Kojima S, et al. Myelodysplastic syndromes in childhood: A
retrospective study in Japan. Leukemia 2000; 14:968.
80. Rodriguez-Galindo C, Poquette CA, Marina NM, et al. Hematologic abnormalities
and acute myeloid leukemia in children and adolescents administered intensified
chemotherapy for the Ewing sarcoma family of tumors. J Pediatr Hematol Oncol
2000; 22(4):321–329.
81. Tefferi A, Thibodeau SN, Solberg LA. Clonal studies in the myelodysplastic syn-
drome using X-linked restriction fragment length polymorphisms. Blood 1990;
75:1770–1773.
82. Busque L, Gilliland DG. X-inactivation analysis in the 1990 s: Promise and potential
problems. Leukemia 1998; 12(2):128–135.
83. Aktas D, Tuncbilek E. Myelodysplastic syndrome associated with monosomy 7 in
childhood: A retrospective study. Cancer Genet Cytogenet 2006; 171(1):72–75.
84. Goel R, Kumar R, Bakhshi S. Transformation of childhood MDS-refractory anemia
to acute lymphoblastic leukemia. J Pediatr Hematol Oncol 2007; 29(10):725–727.
85. Head DR. Revised classification of acute myeloid leukemia. Leukemia 1996;
10:1826–1831.
86. Loeb LA. A mutator phenotype in cancer. Cancer Res 2001; 61:3230–3239.
87. Maserati E, Minelli A, Pressato B, et al. Shwachman syndrome as mutator pheno-
type responsible for myeloid dysplasia/neoplasia through karyotype instability and
chromosomes 7 and 20 anomalies. Genes Chromosome Cancer 2006; 45(4):375–
382.
88. Sheng XM, Kawamura M, Ohnishi H, et al. Mutations of the RAS genes in childhood
acute myeloid leukemia, myelodysplastic syndrome and juvenile chronic myelocytic
89. Tamaki H, Ogawa H, Ohyashiki K, et al. The Wilms’ tumor gene WT1 is a
good marker for diagnosis of disease progression of myelodysplastic syndromes.
Leukemia 1999; 13(3):393–399.
90. Hasegawa D, Manabe A, Kubota T, et al. Methylation status of the p15 and p16 genes
in paediatric myelodysplastic syndrome and juvenile myelomonocytic leukaemia.
Br J Haematol 2005; 128(6):805–812.
91. Vidal DO, Paixao VA, Brait M, et al. Aberrant methylation in pediatric myelodys-
plastic syndrome. Leuk Res 2007; 31:175–181.
92. Jekic B, Novakovic I, Lukovic L, et al. Lack of TP53 and FMS gene mutations in chil-
dren with myelodysplastic syndrome. Cancer Genet Cytogenet 2006; 166(2):163–
165.
93. Jekic B, Novakovic I, Lukovic L, et al. Low frequency of NRAS and KRAS2 gene
mutations in childhood myelodysplastic syndromes. Cancer Genet Cytogenet 2004;
154(2):180–182.
94. Field JJ, Mason PJ, An P, et al. Low frequency of telomerase RNA mutations among
children with aplastic anemia or myelodysplastic syndrome. J Pediatr Hematol Oncol
2006; 28(7):450–453.
MDS in Children 337
95. Ortmann CA, Niemeyer CM, Wawer A, et al. TERC mutations in children with
refractory cytopenia. Haematologica 2006; 91(5):707–708.
96. Roela RA, Carraro DM, Brentani HP, et al. Gene stage-specific expression in
the microenvironment of pediatric myelodysplastic syndromes. Leuk Res 2007;
31(5):579–589.
97. Hicsönmez G, Cetin M, Yenicesu I, et al. Evaluation of children with myelodysplastic
syndrome: Importance of extramedullary disease as a presenting symptom. Leuk
Lymphoma 2001; 42(4):665–674.
98. Cantù-Rajnoldi A, Fenu S, Kerndrup G, et al. Evaluation of dysplastic features in
myelodysplastic syndromes: Experience from the morphology group of the Euro-
pean Working Group of MDS in Childhood (EWOG-MDS). Ann Hematol 2005;
84(7):429–433.
99. Barnard DR, Kalousek DK, Wiersma SR, et al. Morphologic, immunologic, and
cytogenetic classification of acute myeloid leukemia and myelodysplastic syndrome
in childhood: A report from the Childrens Cancer Group. Leukemia 1996; 10:5–12.
100. Rosati S, Anastasi J, Vardiman J. Recurring diagnostic problems in the pathology of
the myelodysplastic syndromes. Semin Hematol 1996; 33:111–126.
101. Hasle H. Myelodysplastic syndromes in childhood. Classification, epidemiology,
and treatment. Leuk Lymphoma 1994; 13:11–26.
102. Groupe Francais de Cytogénétique Hématologique. Forty-four cases of child-
hood myelodysplasia with cytogenetics, documented by the Groupe Francais de
Cytogénétique Hématologique. Leukemia 1997; 11:1478–1485.
103. Gohring G, Michalova K, Beverloo B, et al. A complex karyotype but not monosomy
7 is an independent prognostic factor in advanced childhood MDS. Blood 2007;
110(11):Abstract #2452.
104. Grimwade D, Walker H, Oliver F, et al. The importance of diagnostic cytogenetics
on outcome in AML: Analysis of 1612 patients entered into the MRC AML 10 trial.
The Medical Research Council Adult and Children’s Leukaemia Working Parties.
Blood 1998; 92:2322–2333.
105. Martinez-Climent JA, Garcı́a-Conde J. Chromosome rearrangements in childhood
acute myeloid leukemias and myelodysplastic syndromes. J Pediatr Hematol Oncol
1999; 21:91–102.
106. Woods WG, Neudorf S, Gold S, et al. A comparison of allogeneic bone marrow trans-
plantation, autologous bone marrow transplantation, and aggressive chemotherapy
in children with acute myeloid leukemia in remission. Blood 2001; 97(1):56–62.
107. Greenberg P, Cox C, Le Beau MM, et al. International scoring system for evaluating
108. Chan GCF, Wang WC, Raimondi SC, et al. Myelodysplastic syndrome in children:
differentiation from acute myeloid leukemia with a low blast count. Leukemia 1997;
11:206–211.
109. Latger-Cannard V, Buisine J, Fenneteau O, et al. Dysgranulopoiesis, low blast count
and t(8;21): An unusual presentation of t(8;21) AML according to the WHO classi-
fication: A pediatric experience. Leuk Res 2001; 25(11):1023–1024.
110. Stetler-Stevenson M, Arthur DC, Jabbour N, et al. Diagnostic utility of flow cytomet-
ric immunophenotyping in myelodysplastic syndrome. Blood 2001; 98(4):979–987.
111. Maynadie M, Picard F, Husson B, et al. Immunophenotypic clustering of myelodys-
plastic syndromes. Blood 2002; 100(7):2349–2356.
112. Führer M, Rampf U, Bender-Götze C. Mean corpuscular volume (MCV) in patients

with aplastic anemia does this parameter select patients with clonal disease? A
retrospective analysis of data of the SAA 944 study. Leukemia 2000; 14:961.
113. Fohlmeister I, Fischer R, Mödder B, et al. Aplastic anemia and the hypocellu-
lar myelodysplastic syndrome: Histomorphological, diagnostic, and prognostic fea-
tures. J Clin Pathol 1985; 38:1218–1224.
114. Elghetany MT, Hudnall SD, Gardner FH. Peripheral blood picture in primary
hypocellular refractory anemia and idiopathic acquired aplastic anemia: An addi-
tional tool for differential diagnosis. Haematologica 1997; 82:21–24.
115. Baumann I, Führer M, Bender-Götze C, et al. Histopathological features of hypoplas-
tic myelodysplastic syndrome and comparison with severe aplastic anemia in child-
hood. Leuk Res 1999; 23(Suppl. 1):S41.
116. Goyal R, Varma N, Marwaha RK. Myelodysplastic syndrome with erythroid
hypoplasia. J Clin Pathol 2005; 58(3):320–321.
117. Shinohara K, Yujiri T, Kamei S, et al. Absence of point mutation of N-ras oncogen
in bone marrow cells with aplastic anemia. Int J Cell Cloning 1992; 10:94–98.
118. Elghetany MT, Vyas S, Yuoh G. Significance of p53 overexpression in bone marrow
biopsies from patients with bone marrow failure: Aplastic anemia, hypocellular
refractory anemia, and hypercellular refractory anemia. Ann Hematol 1998; 77:261–
264.
119. Hasle H, Kerndrup G, Jacobsen BB, et al. Chronic parvovirus infection mimicking
myelodysplastic syndrome in a child with subclinical immunodeficiency. Am J
Pediatr Hematol Oncol 1994; 16:329–333.
120. Yarali N, Duru F, Sipahi T, et al. Parvovirus B19 infection reminiscent of myelodys-
plastic syndrome in three children with chronic hemolytic anemia. Pediatr Hematol
Oncol 2000; 17(6):475–482.
121. Kagialis-Girard S, Durand B, Mialou V, et al. Human herpesvirus 6 infection
and transient acquired myelodysplasia in children. Pediatr Blood Cancer 2006;
47(5):543–548.
122. Mueller BU, Tannenbaum S, Pizzo PA. Bone marrow aspirates and biopsies in
children with human immunodeficiency virus infection. J Pediatr Hematol Oncol
1996; 18(3):266–271.
123. Yarali N, Fisgin T, Duru F, et al. Myelodysplastic features in visceral leishmaniasis.
Am J Hematol 2002; 71(3):191–195.
124. Wollman MR, Penchansky L, Shekhter Levin S. Transient 7q- in association with
megaloblastic anemia due to dietary folate and vitamin B12 deficiency. J Pediatr
Hematol Oncol 1996; 18:162–165.
125. Brichard B, Vermylen C, Scheiff JM, et al. Haematological disturbances during
long-term valproate therapy. Eur J Pediatr 1994; 153:378–380.
126. Gesundheit B, Kirby M, Lau W, et al. Thrombocytopenia and megakaryocyte dys-
plasia: An adverse effect of valproic acid treatment. J Pediatr Hematol Oncol 2002;
24(7):589–590.
127. Yetgin S, Ozen S, Saatci U, et al. Myelodysplastic features in juvenile rheumatoid
arthritis. Am J Hematol 1997; 54:166–169.
128. Hinson DD, Rogers ZR, Hoffman GF, et al. Hematological abnormalities and
cholestatic liver disease in two patients with mevalonate kinase deficiency. Am J
Med Genet 1998; 78:408–412.
MDS in Children 339
129. Hirose M, Taguchi Y, Makimoto A, et al. New variant of congenital dyserythropoietic

anemia with trilineage myelodysplasia. Acta Haematol 1995; 94:102–104.
130. Bader-Meunier B, Rieux-Laucat F, Croisille L, et al. Dyserythropoiesis associated
with a Fas-deficient condition in childhood. Br J Haematol 2000; 108(2):300–304.
131. Kratz CP, Rogge T, Kopp M, et al. Myelodysplastic features in an infant with cystic
fibrosis presenting with anaemia, oedema and failure to thrive. Eur J Pediatr 2005;
164(1):56–57.
132. Knerr I, Metzler M, Niemeyer CM, et al. Hematologic features and clinical course of
an infant with Pearson syndrome caused by a novel deletion of mitochondrial DNA.
J Pediatr Hematol Oncol 2003; 25(12):948–951.
133. Bader-Meunier B, Rötig A, Mielot F, et al. Refractory anaemia and mitochondrial
cytopathy in childhood. Br J Haematol 1994; 87:381–385.
134. Haas OA, Gadner H. Pathogenesis, biology, and management of myelodysplastic
syndromes in children. Semin Hematol 1996; 33:225–235.
135. Hasle H, Kerndrup G, Yssing M, et al. Intensive chemotherapy in childhood
myelodysplastic syndrome. A comparison with results in acute myeloid leukemia.
Leukemia 1996; 10:1269–1273.
136. Albitar M, Manshouri T, Shen Y, et al. Myelodysplastic syndrome is not merely
“preleukemia”. Blood 2002; 100(3):791–798.
137. Webb DKH, Passmore SJ, Hann IM, et al. Results of treatment of children with
refractory anaemia with excess blasts (RAEB) and RAEB in transformation (RAEBt)
in Great Britain 1990–99. Br J Haematol 2002; 117:33–39.
138. Niemeyer CM, Duffner U, Bender-Götze C, et al. AML-type intensive chemotherapy
prior to stem cell transplantation (SCT) does not improve survival in children and
adolescents with primary myelodysplastic syndromes (MDS). Blood 2000; 96:521a.
classification of the myeloid neoplasms. Blood 2002; 100(7):2292–2302.
140. Hasle H, Alonzo TA, Auvrignon A, et al. Monosomy 7 and deletion 7q in children
and adolescents with acute myeloid leukemia: An international retrospective study.
Blood 2007; 109(11):4641–4647.
141. Anderson JE, Appelbaum FR, Schoch G, et al. Allogeneic marrow transplantation
for refractory anemia: A comparison of two preparative regimens and analysis of
prognostic factors. Blood 1996; 87:51–58.
142. Hasle H, Baumann I, Bergsträsser E, et al. The International Prognostic Scor-
143. Scheurlen W, Borkhardt A, Ritterbach J, et al. Spontaneous hematological remission
in a boy with myelodysplastic syndrome and monosomy 7. Leukemia 1994; 8:1435–
1438.
144. Benaim E, Hvizdala EV, Papenhausen P, et al. Spontaneous remission in monosomy
7 myelodysplastic syndrome. Br J Haematol 1995; 89:947–948.
145. Mantadakis E, Shannon KM, Singer DA, et al. Transient monosomy 7. A case series
in children and review of the literature. Cancer 1999; 85:2655–2661.
146. De Simone A, Cantù-Rajnoldi A, Sainati L, et al. Spontaneous remission from RAEB
in a child. Leukemia 2001; 15(5):856–857.
147. Saunthararajah Y, Nakamura R, Nam JM, et al. HLA-DR15 (DR2) is overrepre-
sented in myelodysplastic syndrome and aplastic anemia and predicts a response
to immunosuppression in myelodysplastic syndrome. Blood 2002; 100(5):1570–

1574.
148. Steensma DP, Dispenzieri A, More SB, et al. Antithymocyte globulin has limited
efficacy and substantial toxicity in unselected anemic patients with myelodysplastic
syndrome. Blood 2003; 101:2156–2158.
149. Yoshimi A, Baumann I, Fuhrer M, et al. Immunosuppressive therapy with anti-
thymocyte globulin and cyclosporine A in selected children with hypoplastic refrac-
tory cytopenia. Haematologica 2007; 92(3):397–400.
150. Hicsönmez G, Tuncer AM, Sayli T, et al. High-dose methylprednisolone, low-dose
cytosine arabinoside, and mitoxantrone in children with myelodysplastic syndromes.
Hematol Pathol 1995; 9:185–193.
151. Creutzig U, Bender-Götze C, Ritter J, et al. The role of intensive AML-specific
therapy in treatment of children with RAEB and RAEB-T. Leukemia 1998; 12:652–
659.
152. de Witte T, Suciu S, Verhoef G, et al. Intensive chemotherapy followed by allogeneic
or autologous stem cell transplantation for patients with myelodysplastic syndromes
(MDSs) and acute myeloid leukemia following MDS. Blood 2001; 98(8):2326–
2331.
153. Guinan EC, Tarbell NJ, Tantravahi R, et al. Bone marrow transplantation for children
154. Uderzo C, Locasciulli A, Cantù-Rajnoldi A, et al. Allogeneic bone marrow trans-
plantation for myelodysplastic syndromes of childhood: Report of three children
with refractory anemia with excess of blasts in transformation and review of the
literature. Med Pediatr Oncol 1993; 21:43–48.
155. Locatelli F, Pession A, Bonetti F, et al. Busulfan, cyclophosphamide and melphalan
as conditioning regimen for bone marrow transplantation in children with myelodys-
plastic syndromes. Leukemia 1994; 8:844–849.
156. Nichols K, Parsons SK, Guinan E. Long term follow-up of 12 pediatric patients with
primary myelodysplastic syndrome treated with HLA-identical sibling donor bone
marrow transplantation. Blood 1996; 87:4020–4022.
157. Locatelli F, Zecca M, Niemeyer C, et al. Role of allogeneic bone marrow transplan-
tation for the treatment of myelodysplastic syndromes in childhood. The European
Working Group on Childhood Myelodysplastic Syndrome (EWOG-MDS) and the
Austria-Germany-Italy (AGI) Bone Marrow Transplantation Registry. Bone Marrow
Transplant 1996; 18(Suppl. 2):63–68.
158. Rubie H, Attal M, Demur C, et al. Intensified conditioning regimen with busulfan
followed by allogeneic BMT in children with myelodysplastic syndromes. Bone
159. Leahey A, Friedman DL, Bunin NJ. Bone marrow transplantation in pediatric patients
with therapy-related myelodysplasia and leukemia. Bone Marrow Transplant 1999;
23:21–25.
160. Davies SM, Wagner JE, Defor T, et al. Unrelated donor bone marrow transplantation
for children and adolescents with aplastic anaemia or myelodysplasia. Br J Haematol
1997; 96:749–756.
161. Hongeng S, Krance RA, Bowman LC, et al. Outcomes of transplantation with
matched-sibling and unrelated- donor bone marrow in children with leukaemia.
Lancet 1997; 350:767–771.
MDS in Children 341
162. Deeg HJ, Storer B, Slattery JT, et al. Conditioning with targeted busulfan and
cyclophosphamide for hemopoietic stem cell transplantation from related and unre-
lated donors in patients with myelodysplastic syndrome. Blood 2002; 100(4):1201–
1207.
163. de Witte T, Hermans J, Vossen J, et al. Haematopoietic stem cell transplantation for
patients with myelo-dysplastic syndromes and secondary acute myeloid leukaemias:
A report on behalf of the Chronic Leukaemia Working Party of the European Group
for Blood and Marrow Transplantation (EBMT). Br J Haematol 2000; 110(3):620–
630.
164. Locatelli F, Zecca M, Duffner U, et al. Busulfan, cyclophosphamide and melphalan
as pretransplant conditioning regimen for children with MDS and JMML. Interim
analysis of the EWOG-MDS/EBMT prospective study. Leukemia 2000; 14:971.
165. Castro-Malaspina H, Harris RE, Gajewski J, et al. Unrelated donor marrow trans-
plantation for myelodysplastic syndromes: Outcome analysis in 510 transplants
facilitated by the National Marrow Donor Program. Blood 2002; 99(6):1943–1951.
166. Locatelli F, Noellke P, Fischer A, et al. Hematopoietic stem cell transplantation
(HSCT) after a myeloablative conditioning regimen in children with refractory
cytopenia (RC): Results of a retrospective analysis from the EWOG-MDS group.
Blood 2007; 110(11):Abstract#251.
for myelodysplastic syndrome with advanced disease morphology: A phase II study
of busulfan, cyclophosphamide, and total-body irradiation and analysis of prognostic
factors. J Clin Oncol 1996; 14:220–226.
168. Sutton L, Chastang C, Ribaud P, et al. Factors influencing outcome in de novo
myelodysplastic syndromes treated by allogeneic bone marrow transplantation: A
long-term study of 71 patients Societe Francaise de Greffe de Moelle. Blood 1996;
88:358–365.
169. Sierra J, Perez WS, Rozman C, et al. Bone marrow transplantation from HLA-
identical siblings as treatment for myelodysplasia. Blood 2002; 100(6):1997–2004.
170. Martino R, Caballero MD, Simon JA, et al. Evidence for a graft-versus-leukemia
effect after allogeneic peripheral blood stem cell transplantation with reduced-
intensity conditioning in acute myelogenous leukemia and myelodysplastic syn-
dromes. Blood 2002; 100(6):2243–2245.
171. Strahm B, Locatelli F, Bader P, et al. Reduced intensity conditioning in unrelated
donor transplantation for refractory cytopenia in childhood. Bone Marrow Transplant
2007; 40(4):329–333.
172. Couban S, Simpson DR, Barnett MJ, et al. A randomized multicenter comparison
of bone marrow and peripheral blood in recipients of matched sibling allogeneic
transplants for myeloid malignancies. Blood 2002; 100(5):1525–1531.
173. Nagatoshi Y, Okamura J, Ikuno Y, et al. Therapeutic trial of intensified conditioning
regimen with high- dose cytosine arabinoside, cyclophosphamide and either total
body irradiation or busulfan followed by allogeneic bone marrow transplantation for
myelodysplastic syndrome in children. Int J Hematol 1997; 65:269–275.
174. Copelan EA, Penza SL, Elder PJ, et al. Analysis of prognostic factors for allogeneic
marrow transplantation following busulfan and cyclophosphamide in myelodysplas-
tic syndrome and after leukemic transformation. Bone Marrow Transplant 2000;
25(12):1219–1222.
175. Kalwak K, Wojcik D, Gorczynska E, et al. Allogeneic hematopoietic cell transplan-

tation from alternative donors in children with myelodysplastic syndrome: Is that an
alternative? Transplant Proc 2004; 36(5):1574–1577.
176. Okumura H, Takamatsu H, Yoshida T. Donor leucocyte transfusions for relapse
in myelodyplastic syndrome after allogeneic bone marrow transplantation. Br J
Haematol 1996; 93:386–388.
177. Beck JF, Klingebiel T, Kreyenberg H, et al. Relapse of childhood ALL, AML
and MDS after allogeneic stem cell transplantation can be prevented by donor
lymphocyte infusion in a critical stage of increasing mixed chimerism. Klin Padiatr
2002; 214(4):201–205.
178. Tamura K, Kanazawa T, Suzuki M, et al. Successful rapid discontinuation of
immunosuppressive therapy at molecular relapse after allogeneic bone marrow trans-
plantation in a pediatric patient with myelodysplastic syndrome. Am J Hematol 2006;
81(2):139–141.
179. Skinner R, Velangi M, Bown N. Donor lymphocyte infusions for post-transplant
relapse of refractory anemia with excess blasts and monosomy 7. Pediatr Blood
Cancer 2008; 50(3):670–672.
180. Bader P, Niemeyer C, Willasch A, et al. Children with myelodysplastic syndrome
(MDS) and increasing mixed chimaerism after allogeneic stem cell transplantation
have a poor outcome which can be improved by pre-emptive immunotherapy. Br J
Haematol 2005; 128(5):649–658.
181. Bader P, Niemeyer C, Weber G, et al. WT1 gene expression: Useful marker for min-
imal residual disease in childhood myelodysplastic syndromes and juvenile myelo-
monocytic leukemia? Eur J Haematol 2004; 73(1):25–28.
182. Ballen KK, Gilliland DG, Guinan EC, et al. Bone marrow transplantation for therapy-
related myelodysplasia: Comparison with primary myelodysplasia. Bone Marrow
Transplant 1997; 20:737–743.
183. Yakoub-Agha I, de La Salmoniere P, Ribaud P, et al. Allogeneic bone marrow
marrow transplantation. J Clin Oncol 2000; 18(5):963–971.
184. Hale GA, Heslop HE, Bowman LC, et al. Bone marrow transplantation for therapy-
induced acute myeloid leukemia in children with previous lymphoid malignancies.
Bone Marrow Transplant 1999; 24(7):735–739.
185. Tsurusawa M, Manabe A, Hayashi Y, et al. Therapy-related myelodysplastic syn-
drome in childhood: A retrospective study of 36 patients in Japan. Leuk Res 2005;
29(6):625–632.
186. Longmore G, Guinan EC, Weinstein HJ, et al. Bone marrow transplantation for
myelodysplasia and secondary acute nonlymphoblastic leukemia. J Clin Oncol 1990;
8:1707–1714.
187. Chang C, Storer BE, Scott BL, et al. Hematopoietic cell transplantation in patients
with myelodysplastic syndrome or acute myeloid leukemia arising from myelodys-
plastic syndrome: Similar outcomes in patients with de novo disease and disease fol-
lowing prior therapy or antecedent hematologic disorders. Blood 2007; 110(4):1379–
1387.
188. Cesaro S, Oneto R, Messina C, et al. Haematopoietic stem cell transplantation for
Shwachman–Diamond disease: A study from the European group for blood and
marrow transplantation. Br J Haematol 2005; 131(2):231–236.
MDS in Children 343
189. Hasle H, Clemmensen IH, Mikkelsen M. Risks of leukaemia and solid tumours in
individuals with Down’s syndrome. Lancet 2000; 355:165–169.
190. Hasle H. Pattern of malignant disorders in individuals with Down’s syndrome. Lancet
Oncol 2001; 2(7):429–436.
191. Hellebostad M, Carpenter E, Hasle H, et al. GATA1 mutation analysis demonstrates
two distinct primary leukemias in a child with Down syndrome; implications for
leukemogenesis. J Pediatr Hematol Oncol 2005; 27(7):408–409.
192. Zipursky A, Brown E, Christensen H, et al. Leukemia and/or myeloproliferative
syndrome in neonates with Down syndrome. Semin Perinatol 1997; 21:97–101.
193. Massey G, Zipursky A, Doyle JJ, et al. A prospective study of the natural history
of transient leukemia (TL) in neonates with Down syndrome (DS): A Pediatric
Oncology Group (POG) study. Blood 2002; 100:87a.
194. Slayton WB, Spangrude GJ, Chen Z, et al. Lineage-specific trisomy 21 in a neonate
with resolving transient myeloproliferative syndrome. J Pediatr Hematol Oncol 2002;
24:224–226.
195. Wu SQ, Loh KT, Chen XR, et al. Transient myeloproliferative disorder in a phe-
notypically normal infant with i(21q) mosaicism. Cancer Genet Cytogenet 2002;
136(2):138–140.
196. Hayashi Y, Eguchi M, Sugita K, et al. Cytogenetic findings and clinical features
in acute leukemia and transient myeloproliferative disorder in Ddwn’s syndrome.
Blood 1988; 72:15–23.
197. Lange B. The management of neoplastic disorders of haematopoiesis in children
with Down syndrome. Br J Haematol 2000; 110:512–524.
198. Hasle H, Lund B, Nyvold CG, et al. WT1 gene expression in children with Down
syndrome and transient myeloproliferative disorder. Leuk Res 2006; 30:543–546.
199. Lange BJ, Kobrinsky N, Barnard DR, et al. Distinctive demography, biology, and
outcome of acute myeloid leukemia and myelodysplastic syndrome in children with
Down syndrome: Children’s Cancer Group studies 2861 and 2891. Blood 1998;
91:608–615.
200. Gamis AS, Alonzo TE, Lange B, et al. Acute myelogenous leukemia (AML) in
Downs syndrome (DS) patients: Outcome, toxicities, and prognostic factors from
the CCG 2891 trial. Blood 2001; 98:720a.
201. Hasle H, Abrahamsson J, Arola M, et al. Myeloid leukemia in children 4 years or
older with Down syndrome often lacks GATA1 mutation and cytogenetics and risk
of relapse are more akin to sporadic AML. Leukemia 2008; 22(7):1428–1430.
202. Craze JL, Harrison G, Wheatley K, et al. Improved survival of acute myeloid
leukaemia in Down’s syndrome. Arch Dis Child 1999; 81:32–37.
203. Creutzig U, Reinhardt D, Diekamp S, et al. AML patients with Down syndrome have
a high cure rate with AML-BFM therapy with reduced dose intensity. Leukemia
2005; 19(8):1355–1360.
204. Zeller B, Gustafsson G, Forestier E, et al. Acute leukaemia in children with Down
syndrome: A population-based nordic study. Br J Haematol 2005; 128(6):797–804.
205. Abildgaard L, Ellebæk E, Gustafsson G, et al. Optimal treatment intensity in children
with Down syndrome and myeloid leukaemia: Data from 56 children treated on
NOPHO-AML protocols and a review of the literature. Ann Hematol 2006; 85:275–
280.
206. Rao A, Hills RK, Stiller C, et al. Treatment for myeloid leukaemia of Down
syndrome: Population-based experience in the UK and results from the Medical
Research Council AML 10 and AML 12 trials. Br J Haematol 2006; 132(5):576–

583.
207. Kudo K, Kojima S, Tabuchi K, et al. Prospective study of a pirarubicin, intermediate-
dose cytarabine, and etoposide regimen in children with Down syndrome and acute
myeloid leukemia: The Japanese Childhood AML Cooperative Study Group. J Clin
Oncol 2007; 25(34):5442–5447.
208. Ferster A, Verhest A, Vamos E, et al. Leukemia in a trisomy 21 mosaic: Specific
involvement of the trisomic cells. Cancer Genet Cytogenet 1986; 20:109–113.
209. Simon JH, Tebbi CK, Freeman AI, et al. Acute megakaryoblastic leukemia associated
with mosaic Down’s syndrome. Cancer 1987; 60:2515–2520.
210. Wechsler J, Greene M, McDevitt MA, et al. Acquired mutations in GATA1 in the
megakaryoblastic leukemia of Down syndrome. Nat Genet 2002; 32:148–152.
211. Mundschau G, Gurbuxani S, Gamis AS, et al. Mutagenesis of GATA1 is an initiating
event in Down syndrome leukemogenesis. Blood 2003; 101:4298–4300.
212. Rainis L, Bercovich D, Strehl S, et al. Mutations in exon 2 of GATA1 are early
events in megakaryocytic malignancies associated with trisomy 21. Blood 2003;
102:981–986.
213. Zipursky A, Wang H, Brown EJ, et al. Interphase cytogenetic analysis of in vivo
differentiation in the myelodysplasia of Down syndrome. Blood 1994; 84:2278–
2282.
214. Pine SR, Guo Q, Yin C, et al. Incidence and clinical implications of GATA1 mutations
in newborns with Down syndrome. Blood 2007; 110(6):2128–2131.
215. Holt SE, Brown EJ, Zipursky A. Telomerase and the benign and malignant megakary-
oblastic leukemias of Down syndrome. J Pediatr Hematol Oncol 2002; 24(1):14–17.
216. Forestier E, Izraeli S, Beverloo B, et al. Cytogenetic features of acute lymphoblastic
and myeloid leukemias in pediatric patients with Down syndrome an iBFM-SG
study. Blood 2008; 111(3):1575–1583.
217. Ravindranath Y, Abella E, Krischer JP, et al. Acute myeloid leukemia (AML) in
Down’s syndrome is highly responsive to chemotherapy: Experience on pediatric
oncology group AML study 8498. Blood 1992; 80:2210–2214.
218. Lie SO, Jonmundsson G, Mellander L, et al. A population-based study of 272 children
with acute myeloid leukaemia treated on two consecutive protocols with different
intensity: Best outcome in girls, infants, and children with Down’s syndrome. Nordic
Society of Paediatric Haematology and Oncology (NOPHO). Br J Haematol 1996;
94:82–88.
219. Slørdahl SH, Smeland EB, Holte H, et al. Leukemic blasts with markers of four cell
lineages in Down’s syndrome (“megakaryoblastic leukemia”). Med Pediatr Oncol
1993; 21:254–258.
220. Gamis AS, Woods WG, Alonzo TA, et al. Increased age at diagnosis has a sig-
nificantly negative effect on outcome in children with Down syndrome and acute
myeloid leukemia: A report from the Children’s Cancer Group study 2891. J Clin
Oncol 2003; 21(18):3415–3422.
221. Creutzig U, Ritter J, Vormoor J, et al. Myelodysplasia and acute myelogenous
leukemia in Down’s syndrome. A report of 40 children of the AML-BFM study
group. Leukemia 1996; 10:1677–1686.
222. Taub JW, Huang X, Matherly LH, et al. Expression of chromosome 21-localized
genes in acute myeloid leukemia: Differences between Down syndrome and non-
MDS in Children 345
Down syndrome blast cells and relationship to in vitro sensitivity to cytosine arabi-
noside and daunorubicin. Blood 1999; 94(4):1393–1400.
223. Zwaan CM, Kaspers GJ, Pieters R, et al. Different drug sensitivity profiles of acute
myeloid and lymphoblastic leukemia and normal peripheral blood mononuclear cells
in children with and without Down syndrome. Blood 2002; 99(1):245–251.
224. Ge Y, Jensen T, James SJ, et al. High frequency of the 844ins68 cystathionine-beta-
synthase gene variant in Down syndrome children with acute myeloid leukemia.
Leukemia 2002; 16(11):2339–2341.
15
Prognostic Factors in the Assessment of
Patients with Myelodysplastic Syndromes
Aristoteles Giagounidis and Carlo Aul

Medizinische Klinik II (Hämatologie, Onkologie und Immunologie), St. Johannes-
Hospital Duisburg, Duisburg, Germany
Ulrich Germing
Klinik für Hämatologie, Onkologie und klinische Immunologie, Heinrich-Heine
Universität Düsseldorf, Düsseldorf, Germany
INTRODUCTION
One of the hallmarks of the myelodysplastic syndromes (MDS) is their heterogene-
ity, which is reflected not only by the variety of hematological manifestations, but
also by the great differences in length of survival and incidence of acute myeloid
leukemia (AML) between individual patients. Whereas some patients with MDS
succumb to complications of bone marrow failure or AML development within a
few months of diagnosis, others show a relatively stable course and may survive
for many years. These marked differences in prognosis complicate therapeutic
decisions.
Figure 1 shows the estimated overall survival of 3310 patients with MDS,
diagnosed at the University of Düsseldorf over a period of 25 years and treated with
any available therapy. In agreement with data from other centers, median survival
of the entire patient population was about 2 years. About 20% of patients exhibited
a relatively benign disease course, surviving at least 6 years after diagnosis. Given
the advanced age of most patients with the problem of attendant comorbidities,
these patients should usually not receive intensive treatment for MDS. On the other
hand, about the same proportion of patients is threatened by an aggressive course
347
348 Giagounidis et al.
1.0
0.8
Cumulative survival
0.6
0.4
0.2
0.0
0 48 96 144 192 240 288 336 384
Months
Figure 1 Cumulative survival of 3310 patients with all subtypes of MDS of the Düsseldorf
bone marrow registry. All therapies included. Median survival is 25 months.
of disease, which may be similar to that of de novo AML. The markedly reduced
life expectancy often justifies the use of intensive therapeutic modalities, such as
AML-type chemotherapy and autologous or allogeneic stem cell transplantation
in these patients.
The divergent life expectancies of newly diagnosed patients reflect the
dynamic multistep pathogenesis of MDS, which is characterized by a series of
genetic alterations affecting the myeloid progenitor cells in the bone marrow.
Considering the marked heterogeneity of MDS, it was not surprising that many
authors tried to identify prognostic parameters for assigning patients to different
risk groups. Now that new treatments such as immunomodulatory agents, epige-
netic treatment approaches, intensive chemotherapy or stem cell transplantation
are becoming increasingly available, there is an urgent need for better and reliable
risk stratification of MDS patients.
In the past, interpretation and comparison of therapeutic trials was largely
hampered by nonuniform patient groups. A classical example is the outcome of
treatment with low-dose cytosine arabinoside, which yielded markedly differing
response rates between 26% and 71% in various trials (1). It is likely that the differ-
ent treatment results were due to enrollment of patients with different risk factors.
Prognostic Factors in the Assessment of Patients 349
Thus, prognostic stratification is critical not only for selection of individualized

therapy but also for clinical trial purposes.
This review expands on chapter 1, summarizing our current knowledge of
prognostic parameters in MDS, and discussing the clinical benefit of evaluating
patients’ survival probability and risk of AML by scoring systems constructed
from multivariate analyses of clinical, hematological, and cytogenetic factors.
Most of the studies reviewed reported results on adult patients with MDS. In
recent years, increasing information has been compiled on pediatric cases of MDS
(see chap. 14), which have been estimated to account for 1% to 3% of childhood
malignancies. Prognostic factors in pediatric MDS appear to be different from
those in adult MDS. For further information on pediatric MDS, which is beyond the
scope of the present review, the reader is referred to several excellent publications
(2–4).
It is noticed that the available prognostic models have been designed to
predict survival of patients that have not undergone any kind of intensive treatment.
They are built on information from large databases and allow discussion on the
natural course of the disease, but they do not account for prognostic changes
that occur due to therapeutic intervention. Predicitve models regarding individual
treatment approaches are less reliable because they are based on small patient
numbers from phase II and phase III clinical studies.
PREREQUISITES FOR USING CLINICAL AND BIOLOGICAL VARIABLES

AS RISK FACTORS
Of course, parameters used for risk evaluation in MDS should be predictive of
patients’ life expectancy. It should be emphasized that large patient populations are
usually required for identifying statistically significant risk parameters. Reported
patient series are often heterogeneous with regard to sample size, period of inves-
tigation, time of diagnosis, minimal diagnostic criteria of MDS, morphological
subgroups, percentage of patients with secondary MDS, inclusion of patients who
have undergone intensive treatment, length of follow-up, and application of partic-
ular statistical methods. These differences may explain, for example, why earlier
studies failed to demonstrate significant differences between the survival curves
of MDS patients with and without abnormal karyotypes (5), whereas more recent
investigations have clearly established the prognostic importance of cytogenetic
analysis. Another disadvantage of almost all studies performing prognostic factor
analyses is their retrospective design.
To be widely applicable, prognostic parameters should be readily available
to all practicing hematologists. Therefore, molecular genetic findings, despite their
importance for understanding the varied natural course of MDS, are unlikely to
be incorporated into risk analysis at anytime, because such parameters are usually
available only to specialists. Second, the independent statistical significance of the
prognostic variable should be confirmed by multivariate analysis. For example,
the independent prognostic value of the 1982 French-American-British (FAB)
classification of MDS (see chap. 1) diminishes if the medullary blast cell count is
included in a multivariate analysis. Third, prognostic factors also need to be con-
firmed by independent investigators in different patient series. Finally, determina-
tion of the prognostic parameter should be inexpensive. Even the most significant
variables such as chromosomal analysis are increasingly subjected to financial
constraints.
Although MDS are an important field of research at the University of
Düsseldorf and there is a close cooperation between this department and all hospi-
tals and private practices in the area, only 1200 (36%) out of 3300 patients recorded
in the Düsseldorf bone marrow register have been karyotyped. The patients who
had cytogenetics done were on average 10 years younger than the patients in whom
no chromosome studies were performed. It is likely that the practicing clinicians
did not perform a chromosomal analysis in elderly patients because they felt that
their patients were too frail to obtain any therapeutic benefit from the result of kary-
otyping. It is encouraging to note that in 2007, the proportion of newly diagnosed
patients in the registry, who underwent cytogenetic examination, rose to 85%.
It is not clear whether there is truly a need for prognostic parameters that
provide information on the patient’s risk of transformation to acute leukemia.
There are no studies to demonstrate that MDS patients with transformation to acute
leukemia die of other causes than patients whose medullary blast count remains
below 20%. In 1992, Ghulam Mufti in London reported an unpublished analysis
by Guido Tricot in which the median survival of the patients who progressed to
acute leukemia was not significantly different from those patients who did not
transform (6). Obviously, complications due to peripheral blood cytopenias and
functional defects of circulating blood elements are the leading causes of death in
both the patient groups.
A similar conclusion can be derived from our own data. Figure 2 shows the
cumulative survival of patients with high-risk MDS (i.e., medullary blast count
of ⬎10%), who were treated with supportive care only. The median survival for
patients with or without AML development was almost identical, amounting to 10
and 12 months, respectively.
However, predicting the risk of AML transformation may be relevant for
patients initially belonging to low-risk groups (7,8). The analysis of the Düsseldorf
database shows that in patients with a blast count of ⬍10% at diagnosis, the median
overall survival differs significantly according to the fact that the patient’s disease
did or did not transform to AML (Fig. 3). The assessment of AML evolution
risk is becoming increasingly important to international approval agencies in the
face of new drug developments in MDS (e.g., data on azacitidine that emerged in
2007 that showed this agent delays progression to AML compared with supportive
care—see chap. 10). To determine the safety of drugs used in low-risk MDS, the
agencies focus on long-term data on AML evolution both in randomized and
nonrandomized trials.
Although a large number of prognostic factors have been described in MDS
by various investigators (Table 1), only a few variables have gained practical
1.0
0.8
Cumulative survival
0.6
0.4
0.2
AML− (n = 718)
AML+ (n = 399)
0.0
0 48 96 144 192 240 288
Months
Figure 2 Cumulative survival of patients with MDS and ≥10% bone marrow blasts.
AML+ patients have eventually developed acute myeloid leukemia (AML), AML− patients
have not. Median overall survival for AML+ patients was 12 months and for AML− patients
it was 10 months. p = nonsignificant.
acceptance, and these usually suffice for predicting the natural course of the
disease in general terms. It should be emphasized that no single variable can
identify precisely the clinical course and final outcome in the patient. Scoring
systems try to enhance the predictive power by combining several features of
the disease that have demonstrated independent prognostic value on multivariate
analysis. But even these scores may wrongly judge the clinical course in the
individual patient. This is largely due to the fact that none of the present factors
or prognostic scoring systems, which refer to disease characteristics at the time of
diagnosis, is capable of correctly modeling the dynamic process of clonal evolution
in MDS. The recently published scoring system WPSS (WHO-based prognostic
scoring system) has been validated to allow the reappraisal of a patient’s prognosis
during the disease course (9). Although this does not result in a more precise initial
calculation of the total disease duration for an individual patient, the score enables
to review the patient’s condition during the course of the disease. This is in contrast
to other scoring systems that were designed explicitly for evaluation of patients’
prognoses at the diagnosis of their disease only.
1.0
0.8
Cumulative survival
0.6
0.4
0.2
AML− (n = 1677)
AML+ (n = 235)
0.0
0 48 96 144 192 240 288 336 384
Months
Figure 3 Cumulative survival of MDS patients with ⬍10% bone marrow blasts according
to acute myeloid leukemia (AML) evolution. AML+ patients transformed eventually to
AML, median overall survival 20 months; AML− patients did not transform to AML,
median overall survival 40 months. All therapies included. p = 0.00005.
Although prognostic factors were established to identify the natural course

of MDS, there are several recent reports indicating that such factors are also
helpful to predict the response to specific treatments. For example, cytogenetic
abnormalities that are usually associated with an unfavorable clinical course are
highly predictive of outcome after AML-type chemotherapy or allogeneic bone
marrow transplantation (10,11). W. J. Hirst and Ghulam Mufti in London reported
that the rate of disease progression predicts the quality of remissions following
intensive chemotherapy for MDS (12). Furthermore, several reports underline
that demethylating therapies are especially effective in patients with poor-risk
karyotypes in MDS, while these patients do not respond to low-dose cytarabine
treatment (13,14). Fred Appelbaum and Anderson in Seattle evaluated the influ-
ence of the International Prognostic Scoring System (IPSS) risk categorization on
the outcomes of allogeneic bone marrow transplantation in 250 patients with MDS
and found out that the IPSS score is closely correlated to the patients’ disease-free
survival and relapse rate after transplantation (15). Also, the pretransplantational
ferritin value seems to correlate with patient outcome after allogeneic stem cell
transplantation (16,17).
Table 1 Prognostic Factors in Myelodysplastic Syndromes

Clinical parameters DNA ploidy
Age, sex Hypodiploid marrow cells
Previous exposure to radiochemotherapy Molecular genetic findings
Transfusion requirements RAS and FMS mutations
Iron overload C-MPL and DCC expression
Laboratory features P53 mutations
Degree of peripheral blood cytopenias Abnormalities of CDK inhibitors
LDH activity (p15INK4B )
Serum deoxythymidine kinase (TK) Bmi-1 expression
Interleukin-2 receptor (sIL-2R) Immunophenotyping
Ferritin CD13, CD33, and CD34 expression
Bone marrow morphology Monoclonal antibody ratios
Blast cell percentage Lymphoid markers
FAB and WHO subtype Fas/APO-1 (CD95) and Fas-ligand
Degree of hematopoietic dysplasia expression
Cellularity Markers of cell proliferation (e.g.,
Myelofibrosis MIB-1)
Abnormal localization of immature P-170 glycoprotein expression
precursors (ALIP) Bone marrow culturing
Chromosomal status Abortive myeloid cluster formation
Numerical abnormalities Defective maturation within colonies
Structural abnormalities Presence of clonogenic leukemic cells
Karyotypic evolution (CFU-L)
Clonal status (NN, AA, AN) Long-term bone marrow cultures
Abbreviation: CDK, cyclin-dependent kinases.
CLINICAL PARAMETERS THAT PREDICT OUTCOME

In several studies (18–22), older age has been associated with poorer prognosis
with regard to overall survival in MDS. Guido Tricot and his colleagues (18)
reported that younger patients (⬍50 years) die less frequently of infectious and/or
hemorrhagic complications than elderly patients (⬎70 years), while the percentage
of patients transforming to AML was similar in both the age groups. Guillermo
Sanz and colleagues (19) confirmed the prognostic value of age by univariate
and multivariate analyses and incorporated this parameter in their scoring system.
Considering age as an important parameter, it must be realized that the vast
majority of patients with MDS (⬎75%) are older than 60 years at diagnosis. Other
authors have questioned the prognostic influence of age, explaining the poorer life
expectancy of older persons by a higher incidence of deaths unrelated to MDS
(23).
Andrea Kuendgen in Düsseldorf and her colleagues (24) analyzed overall
survival of patients younger than 50 years at the time of MDS diagnosis compared
to patients over the age of 50. The overall survival advantage of younger patients is
confined to the low- and intermediate-1–risk categories, and gets lost in the higher
risk disease states. In low- and intermediate-1–risk patients ⬍50 years of age,
median survival was not reached, while it accounted for 45 months for patients
≥50. In contrast, survival time was identical for both the age groups (8 months) in
the intermediate-2 and high-risk categories due to the overriding impact of disease
biology. Controversies also exist about the apparently adverse prognosis of male
sex. Although survival was significantly shorter in men, Morel and colleagues in
France found no difference in standard mortality ratios between male and female
patients (23). This has been confirmed in other studies, including the investigations
leading to the IPSS (25). A number of other clinical parameters such as weight
loss or presence of anemic symptoms, infections, and hemorrhages at diagnosis
have been associated with an adverse prognosis in MDS (19,26).
Compared to primary MDS, secondary MDS, occurring after exposure of
patients to chemotherapy and/or radiotherapy, show an unfavorable clinical course
with a high rate of leukemic transformation (see chap. 8). Therapy-induced MDS
differ from primary MDS in several respects, including younger age of patients,
more severe cytopenias, more pronounced bone marrow dysplasia (“trilineage
MDS”), and a higher incidence of clonal karyotype anomalies (27). Reduced mar-
row cellularity, often combined with fibrosis (25–50% of cases), makes it difficult
to obtain adequate material for cytological evaluation. Patients with secondary
MDS usually present with an advanced stage of disease. Kuendgen et al. (28)
showed that the overall survival of secondary MDS cases was dramatically short-
ened, independent of the age of the studied MDS population. Young patients with
secondary MDS (all IPSS risk groups included) lived for a median of 11 months
versus 176 months for patients with primary MDS. In patients ≥50 years, median
survival was 8 months versus 25 months, respectively.
In a study of the M.D. Anderson Cancer Center (29), 73% of patients with
secondary MDS were classified as having RAEB or RAEB-t, whereas series of
patients with primary MDS usually show an even distribution between early and
advanced stages. Elihu Estey stated that the poor prognosis of secondary MDS is
largely explained by its assocation with unfavorable chromosome abnormalities
such as deletions or monosomies of chromosomes 5 and/or 7 and complex aber-
rations (29). Once cytogenetics are taken into consideration, there remains only a
relatively small difference in survival between primary and secondary MDS. This
has recently been corroborated by data from a collaborative study of the German-
Austrian-Swiss MDS group showing that in multivariate analysis, only karyotype
and elevation of LDH were significant prognostic parameters in secondary MDS
(28).
In an analysis of their MDS database, Mario Cazzola and Luca Malcovati
in Pavia (30) were able to show that patients who had transfusion-dependent
MDS fared significantly worse than those patients who were transfusion free.
This was also true for leukemia-free survival (LFS) (31). There was a significant
difference in overall survival and LFS according to the number of transfusions
received by an individual whether considered as a total number of transfusions
(20, 20–40, ⬎40 RBC) or by transfusions per month. These effects were main-
tained after accounting for cytogenetics. In higher risk MDS (RAEB-I and
RAEB-II), the difference between nontransfusion dependent and transfused
patients was not significant (31).
The reason for the survival difference seen in the Pavia study has not been
entirely elucidated. It can be speculated that transfusion dependence in MDS is
merely reflecting a more advanced disease stage and a more serious affection
of the bone marrow, partly because the number of functional residual healthy
stem cells is lower. On the other hand, chronic transfusion dependence will in
the long run lead to cardiac remodeling (32), a factor that may precipitate in
heart failure, that is, non-MDS–related morbidity and mortality. Finally, chronic
iron overload from long-term transfusions may also exert a negative effect on
overall survival (see chap. 7). Iron overload inevitably leads to formation of labile
plasma iron, which is a strong inductor of dangerous reactive oxygen species
(ROS) (33). These ROS have been implicated in the deleterious effects of chronic
iron overload on the functioning of, among others, heart, liver, and endocrine
tissues. It is unclear whether the obvious negative effects of iron overload as
demonstrated in thalassemia patients can be applied to the elderly MDS patient
population. However, it would be surprising if the elderly MDS population
with a higher number of comorbidities would be more resistant to ROS than
adolescents.
These questions about iron overload are more than just academic, as iron
depletion (chelation) therapy would be indicated if iron overload had a significant
negative effect on MDS patients. However, the available data show that iron
overload shortens survival in MDS patients is not strong, and is partly based on
the assumptions that ferritin correctly mirrors body iron stores in MDS. These
data include nonrandomized studies showing surprisingly long median overall
survival times for highly selected MDS patients treated with iron chelators. The
two studies reporting these lengthy survival times (34,35) have only been presented
in abstract form and have been criticized because of patient selection bias. The
observed median overall survival times for transfusion-dependent MDS patients
under chelation therapy were ⬎160 months (34) and 115 months (35). The decision
for chelation in both the studies was not based on randomization, but rather on
the subjective assessment of the attending physician. The important message of
those studies therefore is that physicians obviously have a very powerful predictive
ability in choosing the patients for chelation. The patients eventually chosen for
chelation had a median overall survival higher than that of the best subgroup in
current prognostic scoring systems (IPSS, WPSS).
LABORATORY FEATURES WITH PROGNOSTIC IMPORTANCE

Despite a close relationship between medullary blast cell percentage and periph-
eral blood cytopenias, hemoglobin levels and platelet counts are independent risk
factors for predicting survival in MDS. Peripheral blood cytopenias not only reflect
the extent of blast cell infiltration, but also the impaired maturation of hematopoi-
etic progenitors in the bone marrow (dyshematopoiesis). The platelet count seems
to be the most important prognostic factor among the various cytopenias. The
predictive value of neutrophil counts is less certain. Mufti et al. (36) hold that even
mild granulocytopenia is of prognostic value, and they integrated a cut-off value of
2.5 × 109 /L in their scoring system (Bournemouth score). Other authors found that
neutrophil counts become prognostic indicators only when falling below certain
limits (0.5 × 109 /L, 1.0 × 109 /L, or 2.0 × 109 /L) (5,19,37). In our own studies
(38,39), we were unable to demonstrate shortened survival and increased risk of
AML transformation among patients with reduced neutrophil counts. Kerkhofs
et al. and Mufti et al. were the first to note that combinations of peripheral cytope-
nias (bi- and tricytopenia) carry a worse prognosis than isolated cytopenia (21,36).
The presence of circulating blast cells carries a poor prognosis, both in
terms of survival and risk of leukemic transformation, and has been confirmed by
numerous studies (5,19,38,40–42). In the Spanish series, patients with circulating
blast cells had an actuarial median survival of only 6 months (19). The close
association between the proportion of blast cells in the peripheral blood and bone
marrow explains why circulating blasts have no independent prognostic value
when analyzed by multivariate analysis. Noteworthy, the percentage of blast cells
in the periperal blood is a major criterion of both the FAB and WHO classifications.
Besides peripheral blood findings, several other laboratory parameters have
been evaluated for their prognostic value in MDS. Surprisingly, serum lactate dehy-
drogenase (LDH) activity has received little attention in the literature, although
LDH levels have been identified as useful prognostic factors in a variety of other
hematological malignancies (43,44). In two independent patient series, we could
demonstrate that LDH activity is strongly correlated with survival (38,39). Median
survival for patients with normal enzyme levels at diagnosis was 37 months, as
compared to 11 months for patients with increased LDH activity (Fig. 4). By
reflecting increased cell turnover, the elevated enzyme activity may represent a
measure of ineffective hematopoiesis. We were unable to show that LDH levels
depend on the percentage of bone marrow blasts. In the CMML group, LDH
activity was the only parameter besides the medullary blast cell count that was
significantly related to the prognosis of patients (38). The independent prognostic
value of LDH was confirmed by multivariate survival analysis. When cytogenetic
data were included in the Cox regression model, the prognostic importance of LDH
was maintained. Other groups have confirmed the prognostic importance of LDH
in MDS (19,42,45). It is shown in a recent analysis by our group that the addition
of LDH to the IPSS was very helpful in further refining the prognostic power of
the IPSS (46). As a rule of thumb, for any specific risk group, an increased LDH
level upgrades an affected individual to the next higher risk group (see below).
Another interesting parameter that has been incorporated into prognostic fac-
tor analysis in MDS is deoxythymidine kinase, a key enzyme of the salvage path-
way for pyrimidine deoxyribonucleotide synthesis. Serum levels of deoxythymi-
dine kinase (sTK) have been shown to provide prognostic information in AML,
1.0
0.8
Cumulative survival
0.6
0.4
0.2
LDH normal (n = 1269)

LDH elevated (n = 1123)
0.0
0 48 96 144 192 240 288 336 384
Months
Figure 4 Cumulative overall survival of 2392 MDS patients from the Düsseldorf MDS
registry according to their lactate dehydrogenase activity (LDH) in the serum. Median
overall survival for patients with normal LDH was 34 months versus 14 months in patients
with elevated LDH. p ⬍ 0.00005.
multiple myeloma, Hodgkin’s disease, and non-Hodgkin’s lymphomas (47). Stud-

ies in patients with AML have found that serum TK is closely related to the degree
of medullary blast cell infiltration and may thus serve as a measure of leukemic cell
burden (48). Investigations in MDS have shown that about 80% of patients present
with increased enzyme levels (⬎5 U/␮L) at the time of diagnosis. Although TK
activities tended to increase with progression of MDS, there was no correlation
between serum TK and medullary blast cell count. For patients with TK levels
⬍10 U/␮L, actuarial survival 2 years after diagnosis was 65%, compared with
33% for those with enzyme values ≥10 U/␮L. The 5-year cumulative survival
rates were 34% and 14%, respectively (49). Whereas Carlo Aul and colleagues
in Germany found that sTK levels at diagnosis did not predict transformation
to AML, Pelligrino Musto and colleagues in Italy reported a strong correlation
between sTK and risk of AML development (50). In the Italian study, the inde-
pendent prognostic value of sTK for predicting overall and leukemia-free survival
was corroborated by multivariate analysis (50).
Ferritin is the main intracellular iron storage protein in eukaryote cells. It
is a reasonably good predictor of body iron load, but may be falsely elevated
in acute phase reactions due to infection or inflammation, and in chronic alco-

hol abuse and malnourishment. Serially measured, ferritin adequately reflects the
iron stores of MDS patients, as has been shown by several studies (33,51). Luca
Malcovati and his coworkers presented data on the prognostic relevance of fer-
ritin in transfusion-dependent MDS patients. They showed that in very low-risk
populations (i.e., refractory anemia, refractory anemia with ring sideroblasts, and
5q-disease according to WHO), elevated ferritin values above 1000 ng/mL had a
significantly deleterious effect on median overall survival with an increased haz-
ard ratio of 30% for every 500 ng/mL of ferritin above the 1000 ng/mL threshold
(52). The effect vanished in higher risk populations (e.g., refractory cytopenia
with multilineage dysplasia with or without ring sideroblasts).
BONE MARROW MORPHOLOGY

Cytological bone marrow smears are available in all cases of MDS and can
therefore be used for prognostic purposes. Among quantitative parameters, the
medullary blast cell count is by far the most important prognostic parameter.
Whereas some authors found that distinction between cases with ⬍5% and ≥5%
blast cells in the bone marrow is sufficient for prognostic purposes. Sanz et al. (19)
showed that the patients’ life expectancy progressively shortens with increasing
medullary blast cell percentage. They also demonstrated that RAEB is a heteroge-
neous disease that, by defining an extra cut-off point at 10% bone marrow blasts,
can be separated into two subgroups with significantly different prognoses.
Figures 5 and 6 show the estimated overall survival as well as the cumulative
rates of AML transformation in 1494 untreated patients with primary MDS sepa-
rated according to the bone marrow blast cell percentage at the time of diagnosis.
As expected, the medullary blast cell burden is not only indicative of survival, but
also possesses a strong predictive weight for the incidence of AML transformation.
Cumulative risk of AML 2 years after diagnosis was 12% for patients with bone
marrow blasts ⬍5% and climbed up to 61% in the group with the highest blast
cell percentage. Some authors found no difference in survival between patients
with 10% to 19% blasts and those with 20% to 29% blasts in the bone marrow
(19,53).
The degree of bone marrow dysplasia has also been associated with an
adverse prognosis in MDS. In Japanese patients, qualitative abnormalities of
hematopoietic cells appeared to have higher prognostic significance than quanti-
tative changes (41,54). Jurgen Thiele in Cologne and his colleagues described a
linkage of severe dyshematopoesis to leukemic evolution of MDS (54). Whereas
abnormalities of the erythroid series were of only minor importance, dysgran-
ulopoiesis and dysmegakaryopoiesis had a strong prognostic weight. There is a
long-lasting debate on the prognostic importance of Auer rods in MDS. How-
ever, as only a minority of MDS patients (less than 3% in our series) carries this
morphological feature, this question is of academic interest. Contrary to previous
Figure 5 Cumulative survival of 1494 primary MDS patients treated with supportive care
only and separated according to blast cell percentage in the bone marrow at the time of
diagnosis. Abbreviations: p, probability of error; MDS, myelodysplastic syndromes.
assumptions, the presence of Auer rods does not indicate a rapidly progressive
disorder, unless the bone marrow blast count is increased (55).
The increasing use of bone marrow biopsies (not just aspirates) has provided
new possibilities for the prognostic evaluation of MDS patients and led to the
definition of two additional variants of the syndrome (hypoplastic myelodysplasia
and myelodysplasia with bone marrow fibrosis). Besides, a precise assessment
of bone marrow cellularity and myelofibrosis, histological specimens allow an
investigation of the relationship between hematopoietic tissue and bony trabeculae.
Krause et al., as cited in a paper by Marc Boogaerts and colleagues (56),
first described the presence of clusters of myeloid blast cells or promyelocytes
aberrantly localized in the intertrabecular areas of the trephine sections (abnormal
localization of immature precursors, or ALIP). The prognostic impact of ALIPs
was demonstrated by Tricot and his colleagues (5) and appeared to be confined
to patients with RA and RARS. The survival time of ALIP-positive patients in
the RA and RARS categories was significantly shorter than that of ALIP-negative
cases (416 days vs. 1465 days). ALIP-positive patients also had an increased
risk of AML transformation (44% vs. 5%). These findings were confirmed by
Marc Boogaerts and his colleagues in Leuven, who reported a median survival of
28 months for ALIP-positive cases, as compared to 65 months for ALIP-negative
patients with RA/RARS (56). Mangi and Mufti emphasized the importance of
performing combined histopathological, cytochemical, and immunological studies
Figure 6 Cumulative risk of transformation to AML in 1494 primary MDS patients

treated with supportive care only and separated according to blast cell percentage in the
bone marrow at the time of diagnosis. Abbreviations: p, probability of error; AML, acute
myeloid leukemia; MDS, myelodysplastic syndromes.
to clearly distinguish between true ALIPs and pseudoclusters of proerythroblasts

and micromegakaryocytes in the bone marrow (57). The size of the aggregates
also appears to be important because only patients with large blastic infiltrates
were invariably characterized by short survival and early evolution to AML (58).
Maschek in Hannover and colleagues performed a multivariate analysis of several
histological parameters in a large cohort of MDS cases and were able to corroborate
the independent prognostic value of ALIPs (59).
Hyperfibrotic MDS was first described by Claude Sultan in Paris and his
colleagues in a series of eight patients (60). Hyperfibrotic MDS that is now con-
sidered a distinct MDS entity is often associated with hyperplasia and marked
dysplasia of megakaryopoiesis and must be distinguished from acute megakary-
oblastic leukemia and myeloproliferative disorders with accompanying fibrosis.
Myelofibrosis occurs rarely in MDS. Maschek et al. observed a marked diffuse
myelofibrosis in only 21 (3.7%) out of 569 samples from patients with primary
MDS (59). The median survival of these patients was markedly reduced in com-
parison to patients without an increase in reticulin fibres (8 months vs. 18 months).
In other series, marked myelofibrosis was also found to be associated with an unfa-
vorable clinical course, the median survival ranging from 7 to 18 months (61,62).
In some series, the negative prognostic influence of myelofibrosis on survival
and AML transformation was confirmed by multivariate analysis (59,63). In the
Hannover series, myelofibrosis had the strongest impact on prognosis in the oth-
erwise good prognosis group of MDS patients presenting with ⬍5% blast cells in
the bone marrow (59).
Kazuma Ohyashiki and colleagues in Japan reported 7 out of 82 patients
with primary MDS who developed severe myelofibrosis during the course of MDS
and expired 1 to 9.5 months after the onset of myelofibrosis. Six of the 7 patients
had presented with chromosomal abnormalities, often complex aberrations, at the
time of diagnosis of MDS (64). In a recent review of 349 bone marrow biopsies
of 200 patients with primary MDS, Guntram Büsche in Hannover and colleagues
evaluated marrow fibrosis for its prognostic significance in the context of the
WHO classification. Marrow fibrosis correlated with multilineage dysplasia, more
severe thrombocytopenia, higher probability of a clonal karyotype abnormality,
and higher percentages of blasts in the peripheral blood. Its frequency varied
markedly between different MDS types ranging from 0% in RARS subtype to
16% in RCMD and RAEB subtypes. Patients with marrow fibrosis suffered from
marrow failure significantly earlier with shortening of the survival time down to 0.5
(RAEB-1/-2), and 1 to 2 (RCMD, RA) years in median. The prognostic relevance
of myelofibrosis was independent of the IPSS and the WHO classification of
disease (65).
Although not included in the original proposals of the FAB group, hypoplas-
tic MDS is now accepted as separate entity, occurring in 8% to 19% of all MDS
patients (20,66,67). Tuzuner et al. proposed criteria for the diagnosis of hypoplastic
MDS (see chap. 10), based on age-corrected or absolute bone marrow cellularity
(68). Distinguishing hypoplastic MDS from aplastic anemia can be challenging,
because trilineage dysplasia is sometimes difficult to assess and karyotypic abnor-
malities can also be discovered in patients with aplastic anemia (69). Irrespective
of divergent diagnostic criteria, the bone marrow hypoplasia is obviously not an
adverse factor, because patients with hypocellular MDS have no worse prognosis
to those with normocellular or hypercellular MDS (68,70). In the Hannover series
referring to 352 patients, median survival was 22 months for hypoplastic MDS,
27 months for normocellular MDS, and 14 months for hypercellular MDS (66),
while Huang et al. reported superior outcome for hypoplastic MDS patients with
low- and intermediate-1 risk according to the IPSS (70). Some authors reported that
the incidence of leukemic transformation in hypoplastic MDS is slightly lower
than in typical MDS (71). Others have claimed that certain karyotype abnor-
malities involving chromosomes 6 and 7 are associated with hypoplastic MDS
(72,73).
FAB AND WHO CLASSIFICATIONS

The prognostic usefulness of the 1982 FAB classification of MDS has been demon-
strated by numerous studies (5,19–23,26,36–41) and its predictive power was the
basis for its widespread acceptance among hematologists in the 1980s. In gen-
eral, patients with RA and RARS survive significantly longer than patients with
Table 2 Natural History of Primary Myelodysplastic Syndromes
Median survival (mo)

Patients
Study (n) RA RARS RAEB RAEB-t CMML
(36) 141 32 76 10.5 5 22

(21) 237 50 ⬎60 9 6 ⬎60
(19) 370 26 34 9 5 12
(146) 503 32 45 19 11 15
(88) 408 51 ⬎84 18 15 21
(59) 569 26.5 42 8.5 4.5 12.5
(38) 235 30 21 8 4 19
(91) 240 57 35 10 4 32
(41) 838 65 58 16 10 20
(25) 814 50 83 18 7 29
(142) 386 68 65 18 9 24
Median survival of FAB subgroups in 11 studies published between 1985 and 1999.
RAEB, CMML, and RAEB-t. Despite its undisputed value, the FAB classification
has certain limitations which result from the arbitrary demarcation of subgroups,
overlapping features of patients in different FAB categories, overemphasis of mor-
phological findings (e.g., Auer rods) and exclusion of other important prognostic
variables such as peripheral blood cell counts and karyotype anomalies. A com-
parison between different studies shows that there is considerable heterogeneity
in both survival and risk of AML transformation within defined FAB sub-groups,
particularly in patients with RARS and CMML (Table 2).
We have previously proposed to distinguish two forms of sideroblastic ane-
mia: (i) “pure” sideroblastic anemia (PSA), which shows only signs of dysery-
thropoiesis; and (ii) refractory anemia with ring sideroblasts (RARS), which is
not only characterized by erythroid hyperplasia and ineffective erythropoiesis, but
also by disturbed granulopoiesis and megakaryopoiesis. On retrospective analy-
sis of 94 patients with sideroblastic anemia, we found that both types differed
considerably in terms of survival and indicence of AML (74).
Almost identical results have been obtained through a prospective study of
232 new patients with acquired idiopathic sideroblastic anemia, followed-up for
a period of more than 10 years (75). Figure 7 shows the Kaplan–Meier survival
curves for the PSA and RARS patients. The 5-year cumulative survival was
53% and 37%, respectively. Predominant causes of death in the PSA group were
infections and cardiovascular or cerebrovascular diseases, whereas in the RARS
group nearly two-thirds of deaths were attributable to complications of bone
marrow failure, including nine patients transforming to AML. The cumulative
probability of leukemic progression 5 years after diagnosis was 0% for the PSA
group and 15% for patients with RARS.
Figure 7 Cumulative survival of 232 patients with acquired idiopathic sideroblastic

anemia. Patients with pure sideroblastic anemia (PSA) showing only signs of dysery-
thropoiesis have significantly better life expectancy than patients with refractory anemia
with ringed sideroblasts (RARS) in whom the bone marrow smear also shows disturbed
granulopoiesis and megakaryopoiesis. Abbreviations: N, number; p, probability of error.
Similar results were obtained in a study by Richard Garand and his col-
leagues at several centers in France, who looked at the natural course of 84
patients with sideroblastic anemia (76). Patients with AISA-E (equivalent to our
PSA group) survived significantly longer than patients with the myelodysplastic
variant AISA-M (equivalent to our RARS group). The French study was also able
to demonstrate a statistical difference in the risk of AML development between
both the patient groups. It thus appears that cytomorphological distinction between
PSA and RARS provides valuable and reproducible prognostic information.
Chronic myelomonocytic leukemia, characterized by absolute monocytosis
(⬎1 × 109 /L) in the peripheral blood, is also a heterogeneous disease. The consid-
erable interstudy variation in survival among patients with CMML is in marked
contrast to findings in RAEB and RAEB-t. Based on the presence or absence of
increased peripheral leukocyte counts (cut-off level 13 × 109 /L), the FAB coop-
erative group (77) later proposed to separate two subtypes of CMML (myelodys-
plastic and myeloproliferative variant). However, distinguishing between these
two subtypes provides little prognostic information, because their survival curves
are almost identical (78).
By combining morphological, clinical, immunological, and cytogenetic
methods, the World Health Organization (WHO) panel has updated and revised
the FAB classification of MDS (79). By defining a medullary blast count of 20%
as the dividing line between MDS and AML, the RAEB-t category was eliminated
from the MDS classification. CMML was also not included into the revised clas-
sification. Refractory anemia (with or without ring sideroblasts) was defined as a
disorder involving the erythroid lineage only. As new category, a group of patients
with limited blast cell numbers (⬍5%), but with significant multilineage dyspla-
sia was defined (refractory cytopenia with multilineage dysplasia, RCMD). With
regard to its distinctive morphological and clinical features, MDS with isolated
5q-deletion was considered a separate category within MDS. Finally, the WHO
panel proposed to categorize MDS patients who did not fulfill the criteria of the
other categories as “myelodysplastic syndrome, unclassifiable.”
As previously mentioned, multilineage dysplasia in early MDS and a blast
count cut-off value of 10% were shown to be significant prognostic variables in
different studies. Not surprisingly, the WHO classification, which incorporates
these features as opposed to the FAB proposals, has valuable prognostic signifi-
cance. In a retrospective study of 1600 patients, Ulrich Germing and colleagues
(80) showed that the WHO classification allowed to prognostically separate the
5q− syndrome patients efficiently from patients with ⬍5% blasts with unilineage
erythroid dysplasia, and from those with multilineage dysplasia. Furthermore,
there was a statistically significant difference between the overall survival of
patients with less than or more than 10% bone marrow blasts. Interestingly, from a
prognostic point of view, no difference was seen between patients displaying ring
sideroblasts or not. This was true both for patients with unilineage or multilin-
eage dysplasia, that is, RA and RARS patients as well as RCMD and RCMD-RS
patients had a similar overall survival (Fig. 8).
These results were validated in a prospective way in 1095 patients from the
Düsseldorf registry (81). The relative risk for death from any cause was set to 1 for
patients with RA, RARS, and isolated del(5q). The relative risk for patients with
RCMD/RCMD-RS was 1.62, for RAEB-I 1.97, and 3.62 for RAEB-II. The risk for
AML evolution rose from 1 for RA/RARS/del(5q) to 2.51 for RCMD(RS), 3.71 for
RAEB-I, and 15.34 in RAEB-II. Updated Kaplan–Meier plots of the prospective
cohort are shown in Figure 9. The 2008 WHO classification revision has omitted
the distinction between RCMD and RCMD-RS, because those diseases, although
morphologically clearly different, do not have a different outcome.
Regarding the 5q-deletion, it is important to realize that only the isolated
del(5q) population with a blast count of ⬍5% belongs to the very good prognostic
category. Once additional chromosomal abnormalities or an increase in medullary
blasts is present, the overall survival is reduced. Patients with a complex kary-
otype in addition to a del(5q) have a particularly ominous prognosis with median
survival rates of 7 to 8 months. Patients with an increase in bone marrow blasts
from 5% to 10% also have a seriously reduced median overall survival of only 23
months (82). A preliminary analysis of 50 patients shows that these factors retain
their prognostic strength even when the patients are being treated with lenalido-
mide. The AML evolution rates are dramatically different for patients with the
1.0
0.8
Cumulative survival
0.6
RA
0.4
RCMD
0.2 RAEB-I Isolated del(5q)
RARS
RCMD-RS
RAEB-II
0.0
0 48 96 144 192 240 288 336 384
Months
Figure 8 Retrospective analysis of cumulative survival of patients with myelodysplas-

tic syndromes from the Düsseldorf registry according to the WHO classification (2001).
RA, unilineage refractory anemia, n = 107, median overall survival (mOS) = 66 months;
RARS, unilineage refractory anemia with ring sideroblasts, n = 138, mOS = 75 months;
RCMD, refractory cytopenia with multilineage dysplasia, n = 306, mOS = 41 months;
RCMD-RS, refractory cytopenia with multilineage dysplasia and ring sideroblasts, n =
183, mOS = 33 months; RAEB-I, refractory anemia with excess of blasts type I, n =
256, mOS = 25 months; RAEB-II, n = 235, mOS = 11 months; isolated del(5q), n = 28,
mOS = 61 months.
5q− syndrome (isolated del(5q) and blasts ⬍5%; 12.5% 3 year AML transforma-
tion), patients with additional chromosomal abnormalities (33% AML transfor-
mation at 3 years), and those with an increase in blasts (57% 3 year transformation
rate) (83).
Refractory anemia with ring sideroblasts and thrombocytosis (RARS-T)
is being recognized as a distinct entity of myelodysplastic/myeloproliferative
syndrome in the WHO classification (see chap. 9). These patients present
with megakaryocytic hyperplasia, thrombocytosis, and ring sideroblasts in more
than 15% of erythroid precursors. About two-thirds of these patients exhibit
the JAK2 V617F mutation. Interestingly, these JAK2-mutated patients do not
have a different outcome from patients with RARS without thrombocythemia
(Fig. 10).
1.0
0.8
Cumulative survival
0.6
del(5q)
0.4 RCMD/RCMD-RS
RA/RARS
RAEB-I
0.2 RAEB-II
0.0
0 12 24 36 48 60 72 84 96
Months
Figure 9 Prospctively assessed cumulative overall survival for patients according to the
WHO classification (2001). Del(5q) patients, n = 52, median overall survival (mOS) =
61 months; refractory anemia with or without ring sideroblasts, n = 120, mOS = 66 months;
refractory cytopenia with multilineage dysplasia with or without ring sideroblasts, n = 432,
mOS = 40 months; refractory anemia with excess blasts (RAEB)-I, n = 142, mOS =
25 months; RAEB-II, n = 149, mOS = 15 months.
As previously mentioned, the median overall survival for chronic

myelomonocytic leukemia varies strongly between different studies. In the WHO
classification, CMML is categorized as myelodysplastic/myeloproliferative syn-
drome and further subdivided into CMML I and CMML II according to the
medullary blast count (⬍10% vs. 10–20%). The Düsseldorf database analysis for
CMML patients according to the WHO classification is presented in Figure 11.
CHROMOSOMAL ANALYSIS
Cytogenetic investigations are increasingly used in the diagnostic work-up of
patients with MDS (see chap. 3). Clonal karyotypic abnormalities are found in
32% to 76% of patients with primary MDS (84–94) and are encountered in almost
all cases of secondary MDS (95).
The most common cytogenetic abnormalities in MDS are del(5q), mono-
somy 7/del(7q), and trisomy 8 which account for more than 60% of cases with
1.0
0.8
Cumulative survival
0.6
0.4
0.2 RARS-T
RARS
0.0
0 48 96 144 192 240 288 336 384
Months
Figure 10 Cumulative overall survival of patients with refractory anemia with ring sider-
oblasts (RARS) according to WHO classification, n = 174, median overall survival 66
months, and with RARS with thrombocythemia (RARS-T), n = 23, median overall sur-
vival 54 months. There is no statistical difference in survival, p = 0.50.
abnormal metaphases (92,94). Other recurrent anomalies are translocations involv-

ing 1q, rearrangements of 3q, trisomy 6, del(11q), del(12p), involvement of 12q,
del(13q), deletions of 17p, isochrome 17q, del(20q), trisomy 21, monosomy 21,
and loss of sex chromosomes.
Some defects are associated with characteristic hematological and morpho-
logical findings. For example, the 5q− syndrome, which was first described by van
den Berghe et al. in Belgium (96), preferentially occurs in female patients and is
associated with macrocytic anemia, normal or elevated platelet counts, erythroid
hypoplasia and presence of hypolobulated megakaryocytes in the bone marrow.
Complex aberrations, often involving chromosomes 5 and 7, are three times as
frequent in therapy-induced MDS as in primary MDS (95,97–99). Other unfavor-
able findings such as monosomy 7 also occur more frequently in secondary MDS
than in primary MDS (100).
As the frequency of chromosomal abnormalities generally increases with
progression of MDS (94), it was for years questioned whether the karyotype
constitutes an independent prognostic factor for survival and leukemic transfor-
mation. These early uncertainties were partly due to methodological and statistical
1.0
0.8
Cumulative survival
0.6
0.4
0.2
CMML-I
CMML-II
0.0
0 24 48 72 96 120 144 168
Months
Figure 11 Cumulative overall survival of patients with chronic myelomonocytic leukemia

according to the WHO classification. CMML-1, patients with ⬍10% bone marrow blasts,
n = 296, median cumulative survival 20 months; CMML-2, patients with 10–19% bone
marrow blasts, n = 76, median cumulative survival 14 months. p = 0.0002.
issues (e.g., analysis of insufficient numbers of metaphases, poor quality of band-

ing pattern resolution, small patient samples, inclusion of treated patients, use of
cytogenetic data obtained subsequent to the actual diagnosis of MDS) or careless
grouping of cytogenetic findings. For example, the generally negative influence of
single defects on the natural course of MDS may be masked by including patients
with del(5q) who often show a favorable prognosis.
Christian Steidl in Göttingen and his colleagues evaluated the significance
of the number of metaphases analyzed for the cytogenetic result in 529 MDS
patients. They found that less than 20 analyzed metaphases significantly increased
the likelihood to miss an abnormal clone, and that analysis of more than 25
metaphases improved the sensitivity of the analysis, leading to the identification
of clinically relevant abnormal clones or subclones in a substantial proportion of
patients (101).
A number of studies have clearly shown that chromosomal analysis adds
independent prognostic value to the medullary blast count and other conventional
hematological parameters. Morel et al. (88) performed a prognostic factor analysis
in 408 cases of de novo MDS and found that patients with abnormal karyotype had
Figure 12 Cumulative survival of 262 untreated, primary MDS patients separated accord-
ing to chromosomal status at the time of diagnosis. Abbreviations: N, number; p, probability
of error; MDS, myelodysplastic syndromes.
shorter survival than patients with normal karyotype, and that among patients with
abnormal cytogenetics, prognosis was poorer in patients with complex aberrations.
The very poor prognosis of patients with complex cytogenetic abnormalities was
confirmed in another large study (87) and shown to be a consistent finding in all
FAB subtypes. The survival of these patients rarely exceeds a few months (102).
In our series of 269 untreated patients with primary MDS, patients with complex
aberrations had a median survival of only 8 months, whereas it was 40 months
for patients presenting with single or double defects (Fig. 12). Two years after
diagnosis, the cumulative risk of AML transformation in these subgroups was
37% and 17%, respectively. In comparison, patients with normal chromosomes
had a median survival of 55 months, and their actuarial risk of AML development
2 years after diagnosis was 13%.
A number of studies have shown that an isolated 5q− defect has a more
favorable prognosis than any other karyotype in MDS, including a normal kary-
otype (103–105). Despite longer survival, patients with isolated del(5q) and a
normal bone marrow blast count have a comparably low risk of progression to
AML (about 15% in 5 years).
The independent prognostic significance of chromosomal analysis in MDS
has been confirmed by multivariate analysis (88,106). By evaluating 816 patients
with primary MDS, the International MDS Risk Consensus Group was able to
establish the importance of cytogenetic subgroups regarding survival and risk
of evolution to AML (25). This study defined three subgroups with differing
prognosis: (i) a favorable group including patients with normal karyotype, loss
of the Y chromosome, isolated del(5q) or isolated del(20q); (ii) an unfavorable
group including patients with complex chromosomal defects (≥3 anomalies), −7
or del(7q); and (iii) an intermediate-risk group including other abnormalities such
as trisomy 8 and double cytogenetic defects. Seventy percent of patients were
segregated into the good-risk group, 16% into the poor-risk group, and 14% into
the intermediate-risk group. The median survival times of patients in these three
cytogenetic subgroups were 3.8, 0.8, and 2.4 years, respectively, and the times
for 25% of the patients to undergo AML transformation were 5.6, 0.9, and 1.6
years, respectively (Fig. 13). In a study of 640 patients with de novo MDS, Solé et
al. confirmed the prognostic usefulness of the IPSS cytogenetic subgroups (94).
On univariate analysis, they identified additional subgroups with either excellent
prognosis (12p deletions) or very poor prognosis (single 1q abnormalities). In
addition, they showed that patients with trisomy 8, one of the most frequent
defects in MDS, experienced short median survival (about 1 year) and a high risk
of leukemic transformation (34% at 1 year after diagnosis).
The largest series to date that examined the prognostic impact of cytogenetics
in MDS has been published by Detlef Haase and colleagues in Germany and
Austria (92) and included 2124 patients. A total of 52.3% of patients had an
abnormal karyotype. The outcome was studied in those patients who had never
received other than supportive care treatment (n = 1237). Median survival was 53.4
months for patients with normal karyotypes (n = 612), 54 months for those with
good-risk cytogenetics (n = 767), 31 months with intermediate cytogenetics (n =
210), and 8.7 months for those with complex abnormalies (n = 166). The authors
observed several abnormalities in MDS that, albeit occurring rarely, had a very
favorable median overall survival exceeding 5 years. Those included noncomplex
abnormalities involving del(9q), del(15q), t(15q), del(12p), trisomy 21, isolated
del(5q), and isolated or noncomplex changes at del(20q). However, except for
del(5q) and del(20q), the number of patients involved was less than 10, and
therefore, the prognostic significance must be considered with caution. Trisomy 8
occurred in about 100 patients and, whether isolated or in a noncomplex karyotype,
showed a median overall survival of 23 months, nearly twice as long as in the
Solé publication. Isolated or noncomplex −7 anomaly showed an overall survival
of 14 months. Apart from complex aberrations (overall survival 8.7 months),
involvement of t(5q) in a noncomplex karyotype conferred a very bad prognosis
with overall survival times of only 4.4 months.
Acquisition of new karyotypic anomalies during the course of MDS
(karyotypic evolution) has been shown to herald poor prognosis and is often
associated with an abrupt shift to AML (107,108). The relevance of residual
normal metaphases in a patient with cytogenetic defects is still a matter of
debate. In their large series, Haase et al. reported that patients with coexistence
of normal and abnormal metaphases (AN) have a significantly longer overall
survival (27 months) than those patients in whom all metaphases are abnormal
Figure 13 Cumulative survival (upper graph) and freedom from AML transformation
(lower graph) of 816 patients with primary MDS according to cytogenetic risk categoriza-
tion proposed by the International MDS Risk Consensus Conference. Good risk: normal
karyotype, –Y, del(5q), or del(20q); poor risk: complex chromosomal abnormalities (≥3
anomalies, –7, or del(7q); intermediate risk: other cytogenetic defects. Abbreviations: pts,
patients; AML, acute myeloid leukemia; MDS, myelodysplastic syndromes. Source: From
Ref. (25).
(AA, 18 months; p = 0.014). Kerkhofs et al. (21) reported that AN patients have
a prognosis identical to AA patients, and Mecucci and La Starza (109) even
claimed that patients with residual normal metaphases carry a worse prognosis
than cases in which only aberrant metaphases are found.
Cellular DNA content, which can be measured easily by high-resolution flow
cytometry, also appears to be a prognostic factor. In one study, it was demonstrated
that the presence of hypoploid marrow cells (correlating with loss of chromosomal
material) is associated with shorter survival (110).
In summary, cytogenetics have proved to be highly prognostically rel-
evant. Within the most common chromosomal abnormalities (i.e., those with
an incidence of more than 4%), isolated or noncomplex del(5q), isolated or
noncomplex −Y, isolated or noncomplex del(20q), and a normal karyotype have

the best prognosis. Trisomy 8 confers an intermediate prognosis, and complex
karyotypes and chromosome 7 abnormalities are devastating. Global cooperative
investigations are under way to try and identifiy other good-risk cytogenetic
abnormalities like del(9q), del(12p), and others.
MOLECULAR GENETIC FINDINGS

A number of proto-oncogene mutations have been described in patients with
MDS. It has been suggested that mutations in the ras gene family are relevant to
the pathogenesis of MDS and may be used as prognostic indicators. Ras genes code
for GTP-binding proteins that participate in signal transduction from membrane
receptors to the cell nucleus. Impaired GTPase activity of defective ras proteins
leads to increased intracellular GTP levels which confer a growth advantage upon
affected cells. The incidence of ras mutations in patients with MDS is variable,
ranging from 3% to 33% in different studies (111,112). Ras mutations are par-
ticularly common in CMML and MDS with accompanying monocytosis (113).
Reviewing published data from 624 MDS patients, Parker and Mufti found an
overall incidence of 16%, whereas 39% of samples from CMML patients scored
positive (114). Mutations were preferentially reported in the N-ras gene. Paquette
et al. (115) retrospectively analyzed a large group of MDS patients to determine
the effects of N-ras mutation on prognosis and risk of AML transformation. Of
220 evaluable patients, 20 (9%) had point mutations of N-ras involving codon 12.
Individuals with N-ras mutations had a significantly shorter median survival than
those who were N-ras negative (19 months vs. 39 months). An increased risk of
AML was also found in patients with N-ras mutations. These associations were
independent of the percentage of bone marrow blasts and the presence of cytoge-
netic abnormalities. Whereas several more recent trials reported similar findings
(116,117), other investigators found no correlation between ras mutations and
adverse prognosis, both in terms of survival and transformation to AML (118,119).
Thus, the prognostic value of ras mutations is still controversially discussed.
The fms proto-oncogene encodes the receptor for colony-stimulating fac-
tor 1 (M-CSF). Although relatively infrequent in MDS (117,120), fms mutations
appear to have prognostic value. Padua et al. observed an increased frequency of
transformation to AML in MDS patients harboring mutations of the fms proto-
oncogene. Patients with oncogene mutations also had a significantly poorer sur-
vival compared with those without mutations (117). Bouscary et al. (121) reported
overexpression of the c-mpl gene in 42% of patients with RAEB or RAEB-t and
44% of patients with CMML, while no expression was found in patients with
the early stage MDS. In patients with RAEB and RAEB-t, increased expression
of c-mpl was associated with poor prognosis due to an increased risk of AML
development. Forty-five percent of the c-mpl positive patients evolved towards
AML with a mean follow-up of 10.5 months, while 13% of the c-mpl negative
patients developed secondary leukemia with a mean follow-up of 21.1 months.
However, the independent prognostic significance of these observations has not

yet been demonstrated.
The p53 tumor suppressor gene that is located on the short arm of chro-
mosome 17 (band 17p13) plays a key role in the regulation of the cell cycle.
Mutations of p53 have been reported in a wide range of human malignancies,
including hematopoietic neoplasms (122). Mutant p53 genes have been found
in 0% to 25% of MDS patients, dependent on the detection method and com-
position of the patient sample (123–127). Their presence is strongly correlated
with complex chromosome changes including −5/5q− −7/7q−, and 17p−. It has
therefore been speculated that p53 mutations reflect previous exposure to known
or unknown carcinogens (127). Consistent with this assumption is the relatively
high frequency of p53 mutations in patients with therapy-related MDS (128). As
expected, p53 mutations have been shown by some authors to predict short survival
and an increased risk of AML transformation. In the study reported by Kaneko
et al. (126), 4 of 7 patients with p53 mutations progressed to AML within 7 months
after diagnosis of MDS, and the remaining 3 patients died within 7 months with-
out developing AML. In addition, it appears that p53 mutations indicate a poor
response of patients to aggressive AML-type chemotherapy (129).
Several studies have reported that the cyclin-dependent kinase inhibitor gene
p15INK4B , which is involved in the regulation of the G1/S transition of the cell
cycle, is frequently inactivated by methylation in advanced stage MDS (130,131).
Methylation of p15INK4B was observed in 20 (38%) of 53 MDS patients and was
seen only in patients with a bone marrow blast count ⬎10%. No correlation was
found between karyotype and methylation status. On univariate analysis, patients
with methylated p15INK4B had a median actuarial survival of 18 months as com-
pared with 48 months in patients with unmethylated CDK inhibitor. However, the
prognostic significance of p15INK4B methylation was lost in multivariate analysis
due to its strong correlation with the bone marrow blast percentage (130). These
findings were confirmed in another study reported by Uchida et al. (131) who
concluded that inactivation of genes by promoter hypermethylation may be late
events in the MDS development.
Bmi-1, a gene required to regulate the self-renewal in CD34+ cells, has
been studied in 51 patients with MDS or AML preceded by MDS (132). Bmi-1
expression was higher in RAEB, RAEB-t, and MDS-AML patients than in RA
and RARS. In those patients who progressed from RA or RARS to advanced
stages of MDS, the expression at diagnosis was higher than in patients having a
stable clinical course. Further studies with more patients are necessary in order to
confirm those findings.
Gene expression analysis has increasingly been used to improve diagnosis or
identify prognostic subgroups in MDS. Hofmann et al. suggested that a selection
of 11 genes expressed in CD34+ marrow cells can distinguish between low-risk
MDS, high-risk MDS, and healthy controls (133). By using global mRNA profiling
of CD34+ cells, Sternberg et al. found that in early MDS with normal karyotypes,
B-cell lineage–affiliated gene expression was reduced compared to healty controls
and patients with non-MDS–related anemia (134). Finally, Chen et al. studying
MDS-patient samples with monosomy 7 or trisomy 8 showed dysregulation of
genes important to progenitor cell proliferation and blood cell function. In tri-
somy 8, genes involved in immune and inflammatory responses were upregulated,
while anti-apoptotic genes were downregulated. In monosomy 7, CD34+ cells
showed upregulation of genes inducing leukemia transformation, tumorigenesis,
and apoptosis. Downregulated were genes controlling cell growth and differentia-
tion (135). The results of gene expression analysis so far have not been consistent.
This may partly be due to that the different studies did not use comparable patient
samples: Chen et al. studied patients with distinctive chromosomal abnormalities,
Hofmann et al. included 50% of patients with chromosome aberrations in their
low-risk patient sample, and Sternberg et al. investigated only on patients with
normal karyotypes. Also, because diagnosis of patients with early MDS without
blast cell increase may be difficult in the absence of karyotypic abnormalities,
some patients in the analyzed series may not have suffered from MDS. Third, the
analyzed MDS patient samples are small, and also the number of healthy controls
was low. Further investigations with larger patient samples will be necessary to
allow prognostication by gene array analysis.
In summary, a number of new molecular genetic findings with prognos-
tic significance have been reported in the literature. However, outside research
centers, such markers will rarely be available, which limits their clinical useful-
ness. For prognostic purposes, they can be widely replaced by more conventional
hematological and cytogenetic parameters.
IMMUNOPHENOTYPING
Immunophenotyping has primarily been used to improve the diagnostic accuracy
in MDS. However, over the past few years, evidence has also accumulated that
both immunohistochemistry and flow cytometry (see chap. 11) might be impor-
tant prognostic tools. The diagnosis of MDS involves enumeration of blasts in
peripheral blood and bone marrow. For a number of reasons, flow cytometric blast
identification has not replaced morphological blast counts. This is partly due to
the fact that cytometric reliability of the aspirate becomes poor if the bone marrow
sample is being diluted by peripheral blood. Therefore, for best results, after assur-
ing that an adequate morphology sample with plenty of bone marrow spicules can
be obtained, a second puncture at a different site but on the same iliac crest is
useful.
It should be ascertained that the aspirated bone marrow volume is not exces-
sive to prevent dilution with peripheral blood. Usually, 1 mL of bone marrow is
sufficient for reliable diagnosis. Furthermore, the CD34+ cell cut-off value for
normal levels in flow cytometry is not equivalent to the morphological blast cell
cut-off defined by FAB or WHO. In fact, the flow cytometry cut-off values for
CD34+ cells are a matter of experience, as is counting blast cells in morphology.
Therefore, those cut-offs are best being established by close cooperation with
an expert morphologist. Rigorous standardization of flow cytometric procedures

and instruments is another prerequisite for reliable results. For example, delays in
processing bone marrow samples may cause alteration of antigen expression.
To establish the diagnosis of MDS, abnormal maturation of both mono-
cytic and granulocytic antigens can be useful. This includes abnormal expression
of lymphatic antigens on myeloid precursors as well as altered expression of
myeloid or monocytic antigens. Generally speaking, the abnormal expression of
lymphatic antigens carries more weight than an altered expression of myeloid
or monocytic antigens (136). Also, the combination of several abnormalitites is
of higher predictive value than any single abnormality. However, non-neoplastic
disorders like bone marrow regeneration or growth factor use can mimic abnormal
antigen patterns. Thus, the flow cytometrist must be familiar with changes that
can occur in normal bone marrow during those conditions.
Given that the overall survival of patients with more proliferative MDS (i.e.,
higher blast counts in the bone marrow, RAEB-I and RAEB-II) is significantly
shortened compared to lower risk MDS, it is not surprising to see that several
studies have shown that an increased percentage of CD34+ cells or the presence
of CD34+ cell aggregates in the bone marrow is associated with an adverse
prognosis (137–139). It is noticed that Oriani and colleagues in Milan found
no correlation between the results of CD34 immunostaining and FAB subtype.
In addition, neither the percentage of CD34+ cells nor the number of CD34+
aggregates was correlated with the presence of ALIPs (138,139). Using a cut-
off level of 1% CD34+ cells in the bone marrow, patients with RAEB could be
separated into two prognostic groups with significantly different survival times
(138). The presence of circulating CD34+ cells was also found to be associated
with an unfavorable prognosis. In the study of Sullivan et al. (140), 23 of 62
patients with newly diagnosed MDS had ⬎1% CD34+ cells in the peripheral
blood mononuclear cell fraction. CD34 expression was preferentially observed in
patients with advanced MDS and predicted shorter survival and higher rates of
leukemic transformation. In this study, the presence of circulating CD34+ cells
was a better prognostic indicator than cytogenetics or CFU-GM colony growth.
Up to now, however, the prognostic importance of CD34 expression has not been
confirmed by multivariate analysis in a large patient population.
Arjan van de Loosdrecht and colleagues in the Netherlands (136) examined
50 MDS patient samples for flow cytometric abnormalities. Several abnormali-
ties were noted—28% of MDS patients had an increase of progenitor cells. The
median progenitor cell count was 2.4% for MDS and 1.2% for healthy controls.
The percentage of myeloid progenitor cells in RCMD (with or without ring sider-
oblasts), RAEB-I or RAEB-II was significantly higher than in healthy subjects.
Also, progenitor B-cells were markedly reduced in the bone marrow of MDS
patients compared with controls (0.05% vs. 0.26%, p ⬍ 0.001). Furthermore, the
authors found markers of lineage infidelity on myeloid precursors in MDS patient
samples—38% of patients expressed such markers on the CD34+ cells, namely
expression of CD5 or CD7 on myeloid blasts, or coexpression of more than one
lineage infidelity marker (CD5 and CD7 or CD5 and CD56). Other abnormali-
ties included evidence of abnormal relations between CD13, CD16, CD11c, and
CD15 in maturing granulocytes; hypogranularity as evidenced by side scatter
activity in granulocytes; low or overexpression of CD34, HLA-DR, CD13, and
CD117 on myeloid progenitors; Accounting for all immunophenotypic abnor-
malities in the granulopoetic lineage, 92% of patients displayed some aberrant
immunophenotype. Regarding the monocytic compartment, 40% of MDS patients
had unusual monocytopenia, 12% had monocytosis, and the side scatter activity
in MDS monocytes was significantly lower than in healthy controls. In all, 92% of
patients had some flow cytometric changes in the monocytic compartment. Inter-
estingly, aberrant immunophenotypes in the granulocytic and monocytic series
was also detected in the unilineage dysplastic MDS samples from patients with
RA or RARS according to WHO.
These abnormalities were included into a flow score that was previously
published by Wells and colleagues (141). This scoring system was able to predict
progression to higher risk WHO subtype and/or transfusion dependency. Denise
Wells and her colleagues had reported that this flow cytometric score was correlated
with the IPSS and was of prognostic significance in patients undergoing allogeneic
stem cell transplantation. Patients with mild flow score abnormalities had a 3%
cumulative incidence of relapse at 5 years as opposed to 15% in patients with
moderate flow scores and 33% in patients with high scores (141). If these results
are confirmed in larger series of patients, the use of immunophenotyping will
become a second cornerstone of MDS diagnosis and should be included in new
prognostic scoring systems.
SCORING SYSTEMS
Considering the marked heterogeneity of MDS, it was not surprising that several
authors devised scoring systems to assign patients to different risk groups. These
scores tried to refine the prognostic value of the FAB classifiation by combining
several patient and disease characteristics that were shown by multivariate analysis
to provide independent prognostic information. Whereas initial scoring systems
were based on relatively simple hematological parameters, more recent scores
incorporated cytogenetic findings into risk analysis. Table 3 gives an overview on
different scoring systems and the parameters on which these scores were based
(9,19,25,36,38).
The first scoring system for risk evaluation of MDS patients was proposed in
1985 by the Bournemouth group (36). It was based on univariate analysis of con-
ventional hematological parameters (bone marrow blast count ⬎5%, hemoglobin
concentration ⬍10 g/dL, neutrophil count ⬍2.5 × 109 /L, and platelet count
⬍100 × 109 /L) in a cohort of 141 patients and defined three risk groups which
differed significantly in both survival and rates of leukemic transformation. The
predictive value of the Bournemouth score was confirmed in several independent
studies (19,21,38,142–144). Later, the Bournemouth score was modified by adding
Table 3 Comparison of Different Scoring Systems for Evaluating the Natural Course of
Patients with Myelodysplastic Syndromes Bullets indicate the parameters included in the
scoring system
Prognostic Bournemouth, Sanz, Düsseldorf, IPSS, WPSS,

variable 1985 (27) 1989 (18) 1992 (30) 1997 (96) 2007
Age r
Hemoglobin r r r
Neutrophils r r
Platelets r r r r
LDH r
BM blasts r r r r r
Cytogenetics r r
Transfusion r
dependence
Maximum score 4 5 4 3.5 6
No. of risk groups 3 3 3 4 5
leukocytosis as an unfavorable prognostic factor in order to allow its application

to patients with CMML who often present with neutrophilia rather than neutrope-
nia (145). A number of investigators have proposed modifications of the original
Bournemouth score, employing different combinations and threshold values of
hematological and clinical variables. Based on multivariate analysis and validated
in an independent series of patients, Sanz et al. (19) published another score which
employed bone marrow blast percentage, platelet count, and age as prognostic fac-
tors. By identifying additional cut-off values for the medullary blast cell proportion
and peripheral platelet count, the Spanish group was able to separate patients with
an increased medullary blast percentage (⬎5%) into three prognostically different
risk groups with a median survival of 51 months (score 1), 15 months (score 2
or 3), and 4 months (score 4 or 5), respectively. However, the inclusion of age in
the Sanz score has been critized by other authors who failed to demonstrate its
prognostic impact in univariate and multivariate analyses (36,146). In addition, it
was felt that the inclusion of age complicated clinical decision-making, because
patients who are less able to tolerate aggressive chemotherapy scored higher than
younger patients (147). The Düsseldorf score is another simple scoring system that
differs from the Bournemouth score by including lactate dehydrogenase (LDH)
enzyme levels instead of neutrophil counts (38). Compared with the Bournemouth
score, the Düsseldorf score was found to be advantageous in two respects. Firstly,
it was able to identify those patients with RA and RARS who, without showing
an excess of bone marrow blasts at diagnosis, were bound for an unfavorable clin-
ical course. Secondly, the inclusion of LDH levels qualified this scoring system
for an accurate assessment of patients with CMML, whose prognosis is viewed
too favorably when rated by the Bournemouth score. Another advantage of the
Düsseldorf score is the identification of a true low-risk group (score 0) that is
Figure 14 Cumulative survival of 1147 untreated, primary MDS patients separated

according to the Düsseldorf score, where A represents the low-risk group, B represents
the intermediate-risk group, and C represents the high-risk group. Survival curves for risk
groups A, B, and C are significantly different. Abbreviations: N, number; p, probability of
error; MDS, myelodysplastic syndromes.
characterized by a 5-year survival probability of about 70%. The predictive value

of the Düsseldorf score (Fig. 14) was confirmed in an independent sample of
patients (39) and further validated by other investigators (142,144).
Morel et al. (88) were the first to recognize the strong and independent
prognostic value of chromosomal abnormalities in MDS. On the basis of the
analysis of 408 cases with de novo MDS, they proposed a new scoring system
which, besides bone marrow blast percentage and platelet count, incorporated
cytogenetic data as risk factor. The International MDS Risk Analysis Workshop
substantially advanced the prognostic categorization of patients with MDS. In
this workshop, clinical, morphological, and cytogenetic data from 816 patients
with primary MDS were retrospectively evaluated and subjected to univariate and
multivariate risk analysis (25). Patients who had previously been treated with
short courses of low-dose oral chemotherapy or hematopoietic growth factors
were included in this analysis. Patients with proliferative-type CMML (defined as
patients with a white blood count ⬎12 × 109 /L) were excluded from the analysis
because these patients were thought to represent myeloproliferative diseases rather
than MDS. Cytopenias were defined as hemoglobin ⬍10 g/dL, neutrophils ⬍1.8 ×
109 /L, and platelets ⬍100 × 109 /L. Multivariate analyses identified age ⬎60
years, gender (male sex), number of peripheral cytopenias, bone marrow blast
Table 4 International Prognostic Scoring System (IPSS) for Evaluating Prognosis in

Patients with Myelodysplastic Syndromes
Points
Prognostic variable 0 0.5 1 1.5 2.0
BM blasts (%) 0–4 5–10 – 11–20 21–29

No. of cytopeniasa 0–1 2–3 –
Cytogenetic categoryb Good Int Poor – –
Risk group Score
Low 0
Intermediate-I 0.5–1.0
Intermediate-II 1.5–2.0
High ≥2.5
aPlatelets <100.000/µL, hemoglobin <10 g/dL, neutrophils <1.800/µL.
bGood: normal karyotype, isolated 5q−, isolated 20q−, isolated −Y; intermediate: other anomalies;
poor: complex (≥3 abnormalities), chromosome 7 anomalies.
count, and cytogenetics as the most important variables for disease outcome.
Forty percent of the Workshop patients presented with cytogenetic anomalies,
with del(5q) and trisomy 8 being the most common single abnormalities. Based
on these parameters, a consensus scoring system was developed that distinguished
four prognostic subgroups: low-risk (score 0), intermediate-1 (score 0.5–1.0),
intermediate-2 (score 1.5–2.0), and high-risk (score ⬎2.5) (Table 4). Median
survival for the four risk groups were 5.7, 3.5, 1.2, and 0.4 years, respectively.
The corresponding time intervals for 25% of the patients to undergo evolution to
AML were 9.4, 3.3, 1.1, and 0.2 years, respectively. For patients ⬍60 years of
age, median survival were 11.8 years (low-risk), 5.2 years (intermediate-1), 1.8
years (intermediate-2), and 0.3 years (high-risk), reflecting the influence of age
and disease biology on the disease outcome (25).
The importance of a refined cytogenetic categorization of patients for risk
assessment in MDS was convincingly shown by the Workshop. When cytogenetics
were omitted from the analysis, discrimination of clinical outcome was poorer,
causing inaccurate risk assessment of a substantial proportion of intermediate-1
and intermediate-2 patients (148). As part of the consensus conference, the IPSS
was compared to other prognostic systems (FAB classification, Spanish score, Lille
score) to determine their relative discriminatory abilities for assessing the natural
course of MDS. In this analysis, the IPSS demonstrated greater discriminating
power for both survival and AML evolution than the other scores. The prognos-
tic value of the IPSS has been confirmed in several independent patient series
(94,144). In addition, direct comparison of different scoring systems performed
by Pfeilstöcker et al. (142), showed that the IPSS carried the highest predictive
Table 5 Prognosis of Patients with Myelodysplastic Syndromes from the Düsseldorf

Database According to IPSS + LDH
Median survival
Median survival including LDH
IPSS (mo) LDH (mo) p-Value
Low 88 Normal 107 0.0011

Elevated 63
Intermediate-1 55 Normal 66 0.0006
Elevated 36
Intermediate-2 23 Normal 26 0.01
Elevated 16
High 13 Normal 16 0.01
Elevated 11
value among all scores examined. Therefore, the IPSS is an important tool for risk
evaluation of newly diagnosed MDS patients.
Because the Düsseldorf score had shown the importance of LDH in the
prognostication of MDS, Germing et al. addressed the question whether the IPSS
could be further improved by including LDH as an additional prognostic marker
to the parameters blast percentage, cytogenetics, and number of cytopenias (46).
First, the IPSS was calculated for all patients of the German-Austrian MDS registry
containing 1247 untreated patients. Then, the survival of the patients within each
IPSS-defined subgroup was recalculated according to the fact whether LDH was
elevated or not. Thus, the four subgroups were further subdivided into eight
prognostic groups: low-risk without LDH elevation, low-risk with LDH elevation,
intermediate-1 risk without LDH elevation, intermediate-1 with LDH elevation,
and so forth. The updated results of this score are shown in Table 5. By adding
LDH to the IPSS, it was possible to identify a very low-risk population with an
excellent overall survival of in median 107 months. Interestingly, elevation of LDH
increased the risk of each subgroup significantly. In fact, the calculated overall
survival of a low-risk patient with LDH elevation equalled that of an intermediate-
1–risk patient with normal LDH. Likewise, for each other subgroup, the patients
with elevated LDH were having a median overall survival equivalent to the next
higher rated subgroup. This is an important information, as it allows us to define
both very low-risk populations, as well as patients that would be underrated by
the classic IPSS (Fig. 15). By multivariate analysis, LDH was shown to be of
independent prognostic value.
The latest proposal for prognosis evaluation is the WHO-adapted prognos-
tic scoring system (WPSS) (9). Given that the WHO classification has gained
increased acceptance in the field of MDS, calculation of the IPSS had to be
adapted. The IPSS includes the subforms chronic myelomonocytic leukemia
and RAEB-t, both of which have been omitted in the WHO classification. This
1.0 1.0
0.8 0.8
Cumulative survival
Cumulative survival
0.6 0.6
0.4 0.4
LDH−
0.2 0.2
LDH−
LDH+
LDH+
0.0 0.0
0 48 96 144 192 240 288 336 0 48 96 144 192 240 288 336
(A) Months (B) Months
1.0 1.0
0.8 0.8
Cumulative survival
Cumulative survival
0.6 0.6
0.4 0.4
0.2 0.2
LDH−
LDH−
LDH+
LDH+
0.0 0.0
0 48 96 144 192 240 0 48 96 144 212 260
(C) Months (D) Months
Figure 15 (A) Median survival of 258 patients from the Düsseldorf MDS database with
low-risk IPSS including serum lactate dehydrogenase (LDH). LDH−, patients with non-
elevated LDH, n = 208, median overall survival 108 months. LDH+, patients with elevated
LDH, n = 50, median overall survival 63 months. p = 0.0011. (B) Median survival of
288 patients from the Düsseldorf MDS database with intermediate-1–risk IPSS including
serum lactate dehydrogenase (LDH). LDH−, patients with non-elevated LDH, n = 180,
median overall survival 66 months. LDH+, patients with elevated LDH, n = 108, median
overall survival 36 months. p = 0.0006. (C) Median survival of 150 patients from the
Düsseldorf MDS database with intermediate-2–risk IPSS including serum lactate dehydro-
genase (LDH). LDH−, patients with non-elevated LDH, n = 92, median overall survival
26 months. LDH+, patients with elevated LDH, n = 58, median overall survival 16 months.
p = 0.01. IPSSLDH int21risk IPSS. (D) Median survival of 169 patients from the Düsseldorf
MDS database with high-risk IPSS including serum lactate dehydrogenase (LDH). LDH−,
patients with non-elevated LDH, n = 82, median overall survival 16 months. LDH+,
patients with elevated LDH, n = 87, median overall survival 11 months. p = 0.01.
Table 6 The WHO-Adapted Prognostic Scoring System, WPSS

Points
Variable 0 1 2 3
WHO category RA, RARS, isolated RCMD(RS) RAEB-I RAEB-II

del(5q)
Cytogeneticsa Good Intermediate Poor
RBC transfusion None Regularb
dependence
Score for risk groups: very low, 0; low, 1; intermediate, 2; high, 3–4; very high, 5–6.
a Cytogenetic categories: good, normal karyotype, isolated 5q−, isolated 20q−, isolated −Y; interme-
diate, other anomalies.

b RBC-transfusion dependence defined as 1 RBC every 8 weeks for at least 4 months. RBC, red blood
cell concentrate.
truncation removed a significant part of high-risk patients from the calculation

basis. Furthermore, a new prognostic scoring system on the basis of the WHO
classification would have the advantage to include dysplasia as a prognostic fac-
tor. In actual fact, the WPSS assigns points to the different subgroups as shown in
Table 6. The cytogenetic risk groups have remained unchanged compared to the
IPSS. Finally, instead of rating peripheral blood cytopenias, the WHO classifica-
tion rates transfusion dependence. This has been criticized, because the adminis-
tration of transfusions is dependent on multiple cofactors like age, comorbidities,
local policies, blood product availability, and its evaluation needs throughout a
patient history. On the other hand, this may also be the great advantage of the
score: A patient in good physical health without significant medical antecedents
will have a lower transfusion threshold than an elderly patient with heart dis-
ease. The physician’s decision to transfuse an MDS patient or not includes a very
powerful decision making process that cannot be accounted for by any seemingly
objective assessment method like blood counts, immunophenotyping, or months
since diagnosis.
In the WPSS, regular transfusion is being assumed when there is more
than one packed red cell concentrate given per 8 weeks. By combining three
parameters—transfusions, cytogenetics, and WHO subtype—the score discrim-
inates between five risk groups (updated Figures 16 from the Düsseldorf bone
marrow registry). It also predicts AML evolution for those patient subgroups.
Furthermore, the WPSS has been designed to allow reevaluation of patients that
progress during the course of the disease. The authors showed that the disease
course of patients who were upgraded to another risk group due to disease pro-
gression could be effectively predicted by recalculation of the score according
to the novel parameters (e.g., begin of transfusion dependency, increase in blast
count, additional chromosomal abnormalities). This is an important difference to
1.0 1.0
0.8 0.8 E
Cumulative risk of AML evolution

Cumulative survival
D
0.6 0.6
0.4 0.4 C
A
B B
0.2 0.2
E D A
0.0 0.0
0 48 96 144 192 240 288 336 384 0 48 96 144 192 240 288 336 384
Months Months
(A) (B)
Figure 16 (A) Median overall survival (mOS) for patients with different risks according
to the WHO adapted prognostic scoring system WPSS. A—very low risk group, n = 73,
mOS = 145 months; B—low risk, n = 159, mOS = 65 months; C—intermediate risk,
n = 172, mOS = 48 months; D—high risk, n = 247, mOS = 26 months; E—very high
risk, n = 86, mOS = 9 months. p = 0.00005. (B) Cumulative risk of AML transformation
for the risk groups: A—very low, B—low, C—intermediate, D—high, and E—very high,
according to the WHO adapted prognostic scoring system WPSS. p = 0.00005.
the IPSS that was exclusively tested for risk predicition at diagnosis. The WPSS
has therefore been called time-dependent risk model.
CONCLUDING REMARKS
Since the identification of the MDS as a distinct but diverse group of hemato-
logical disorders, considerable efforts have been made to define factors that are
prognostically important. To estimate disease progression and survival, the prac-
ticing clinician can recur to clinical, morphological, laboratory, and cytogenetic
findings. Clinically speaking, patients over 60 years of age fare less well than
younger patients. Concurrent infections and a bad ECOG status also point toward
a complicated clinical course. Treatment-related MDS are usually predictors of
poor prognosis. Increased LDH levels impact adversely on survival, as do cir-
culating blast cells in the peripheral blood. Accumulating evidence points to the
fact that iron overload is a negative prognostic marker in low-risk patients. Bone
marrow morphology helps identify patients with an aggressive clinical course.
The bone marrow blast count is a prominent poor prognostic feature, with patients
having more than 10% blasts doing worse than those with a limited blast count.
On the other hand, identification of pure sideroblastic anemia (PSA) patients

in contrast to refractory anemia with ringed sideroblasts (RARS) is important, as
PSA shows significantly better survival than RARS. The skilled morphologist
might also predict the 5q− syndrome from the bone marrow morphology and
therefore will identify another prognostically favorable subgroup. Flow cytometry
can depict both abnormal antigen expression as well as dyplasia and can help to
increase diagnostic accuracy and add prognostic information.
Complex cytogenetic abnormalities are particularly devastating. This holds
true for patients with −7 or del(7q) anomalies. In contrast, patients with isolated
del(5q), del(20q), or −Y will do much better. Other small subgroups with very
good or very poor outcome have been recently identified. Because of the low inci-
dence of those chromosomal aberrations, their importance in prognosis is marginal,
though. If cytogenetic analysis is available, the IPSS score is a good approach to
the prognostic evaluation of MDS patients, and has been extensively validated.
To adapt to the new WHO classification, clinicians should use the WPSS that
provides them with the possibility to reevaluate the patient in case of progression.
Additional molecular biological factors are under investigation. Especially
for the emerging field of gene array analysis, it will be exciting to evaluate future
developments.
While prognostic factors focus on the overall survival of a given patient
population, predictive factors aim at calculating the probability of response to
a certain treatment option. These predictors of response will be covered in the
chapters related to the different treatment options. They become more important
as the management of MDS evolves. Some of them are already well established,
others need additional evaluation by large multicenter trials to identify powerful
determinants of clinical outcome after therapy.
Acknowledgment
This publication was supported by the Bundesministerium für Bildung und
Forschung (BMBF), Kompetenznetz “Akute und chronische Leukämien.”
REFERENCES
1. Cazzola M, Anderson JE, Ganser A, et al. A patient-oriented approach to treatment
2. Sasaki H, Manabe A, Kojima S, et al. Myelodysplastic syndrome in childhood: A
retrospective study of 189 patients in Japan. Leukemia 2001; 15:1713–1720.
3. Hasle H, Baumann I, Bergstrasser E, et al. The International Prognostic Scor-
4. Niemeyer CM, Kratz CP, Hasle H. Pediatric myelodysplastic syndromes. Curr Treat
Options Oncol 2005; 6:209–214.
5. Tricot G, Vlietinck R, Boogaerts MA, et al. Prognostic factors in the myelodysplastic
syndromes: Importance of initial data on peripheral blood counts, bone marrow
cytology, trephine biopsy and chromosomal analysis. Br J Haematol 1985; 60:19–
32.
6. Mufti GJ. A guide to risk assessment in the primary myelodysplastic syndrome.

Hematol Oncol Clin North Am 1992; 6:587–606.
7. Vadhan-Raj S, Keating M, LeMaistre A, et al. Effects of recombinant human
granulocyte-macrophage colony-stimulating factor in patients with myelodysplastic
syndromes. N Engl J Med 1987; 317:1545–1552.
8. Ganser A, Volkers B, Greher J, et al. Recombinant human granulocyte-macrophage
colony-stimulating factor in patients with myelodysplastic syndromes–A phase-I/II
trial. Onkologie 1988; 11:53–55.
J Clin Oncol 2007; 25:3503–3510.
10. de Witte T, Suciu S, Peetermans M, et al. Intensive chemotherapy for poor prognosis
myelodysplasia (MDS) and secondary acute myeloid leukemia (sAML) following
MDS of more than 6 months duration. A pilot study by the Leukemia Cooperative
Group of the European Organisation for Research and Treatment in Cancer (EORTC-
LCG). Leukemia 1995; 9:1805–1811.
11. Nevill TJ, Fung HC, Shepherd JD, et al. Cytogenetic abnormalities in primary
myelodysplastic syndrome are highly predictive of outcome after allogeneic bone
12. Hirst WJ, Mufti GJ. The rate of disease progression predicts the quality of remissions
following intensive chemotherapy for myelodysplastic syndromes. Leuk Res 1994;
18:797–804.
13. Burnett AK, Milligan D, Prentice AG, et al. A comparison of low-dose cytarabine
and hydroxyurea with or without all-trans retinoic acid for acute myeloid leukemia
and high-risk myelodysplastic syndrome in patients not considered fit for intensive
treatment. Cancer 2007; 109:1114–1124.
14. Raj K, John A, Ho A, et al. CDKN2B methylation status and isolated chromosome
7 abnormalities predict responses to treatment with 5-azacytidine. Leukemia 2007;
21:1937–1944.
15. Appelbaum FR, Anderson J. Allogeneic bone marrow transplantation for myelodys-
plastic syndrome: Outcomes analysis according to IPSS score. Leukemia 1998;
12(Suppl. 1):S25–S29.
16. Armand P, Kim HT, Cutler CS, et al. Prognostic impact of elevated pretransplantation
2007; 109:4586–4588.
17. Platzbecker U, Ehninger G, Bornhauser M. Prognostic impact of elevated pretrans-
plantation serum ferritin in patients undergoing myeloablative stem-cell transplan-
tation. Blood 2007; 110:3083; author reply 4.
18. Tricot G, Vlietinck R, Verwilghen RL. Prognostic factors in the myelodysplastic
syndromes: A review. Scand J Haematol Suppl 1986; 45:107–113.
19. Sanz GF, Sanz MA, Vallespi T, et al. Two regression models and a scoring system
for predicting survival and planning treatment in myelodysplastic syndromes: A
multivariate analysis of prognostic factors in 370 patients. Blood 1989; 74:395–
408.
20. Coiffier B, Adeleine P, Gentilhomme O, et al. A multiparametric study of prognostic
factors in 336 patients. Cancer 1987; 60:3029–3032.
21. Kerkhofs H, Hermans J, Haak HL, et al. Utility of the FAB classification for
myelodysplastic syndromes: Investigation of prognostic factors in 237 cases. Br
J Haematol 1987; 65:73–81.
22. Cunningham I, MacCallum SJ, Nicholls MD, et al. The myelodysplastic syndromes:
An analysis of prognostic factors in 226 cases from a single institution. Br J Haematol
1995; 90:602–606.
23. Morel P, Declercq C, Hebbar M, et al. Prognostic factors in myelodysplastic syn-
dromes: Critical analysis of the impact of age and gender and failure to identify a
very-low-risk group using standard mortality ratio techniques. Br J Haematol 1996;
94:116–119.
26. Foucar K, Langdon RM II, Armitage JO, et al. Myelodysplastic syndromes. A
clinical and pathologic analysis of 109 cases. Cancer 1985; 56:553–561.
27. Aul C, Gattermann N, Schneider W. Epidemiological and etiological aspects of
myelodysplastic syndromes. Leuk Lymphoma 1995; 16:247–262.
28. Kuendgen A, Mende C, Haase D, et al. Prognostic factors in treatment-related
myelodysplastic syndromes (t-MDS) and acute myeloid leukemia (t-AML). ASH
Ann Meet Abstr 2007; 110:2451.
29. Estey EH. Prognosis and therapy of myelodysplastic syndromes. Cancer Treat Res
1993; 64:233–267.
32. Oliva EN, Dimitrov BD, Benedetto F, et al. Hemoglobin level threshold for car-
diac remodeling and quality of life in myelodysplastic syndrome. Leuk Res 2005;
29:1217–1219.
33. Porter JB. Concepts and goals in the management of transfusional iron overload.
Am J Hematol 2007; 82:1136–1139.
34. Leitch HA, Wong DHC, Leger CS, et al. Improved leukemia-free and overall sur-
vival in patients with myelodysplastic syndrome receiving iron chelation therapy: A
subgroup analysis. ASH Annu Meet Abstr 2007; 110:1469.
35. Rose C, Brechignac S, Vassilief D, et al. Positive impact of iron chelation therapy
(CT) on survival in regularly transfused MDS patients. A prospective analysis by
the GFM. ASH Annu Meet Abstr 2007; 110:249.
36. Mufti GJ, Stevens JR, Oscier DG, et al. Myelodysplastic syndromes: A scoring
system with prognostic significance. Br J Haematol 1985; 59:425–433.
37. Coiffier B, Adeleine P, Viala JJ, et al. Dysmyelopoietic syndromes. A search for
prognostic factors in 193 patients. Cancer 1983; 52:83–90.
38. Aul C, Gattermann N, Heyll A, et al. Primary myelodysplastic syndromes: Analysis
of prognostic factors in 235 patients and proposals for an improved scoring system.
Leukemia 1992; 6:52–59.
39. Aul C, Gattermann N, Germing U, et al. Risk assessment in primary myelodysplastic
syndromes: Validation of the Dusseldorf score. Leukemia 1994; 8:1906–1913.
40. Morra E, Lazzarino M, Castello A, et al. Risk assessment in myelodysplastic syn-
dromes: Value of clinical, hematologic and bone marrow histologic findings at
presentation. Eur J Haematol 1990; 45:94–100.
41. Oguma S, Yoshida Y, Uchino H, et al. Clinical characteristics of Japanese patients

with primary myelodysplastic syndromes: A co-operative study based on 838 cases.
Anemia Study Group of the Ministry of Health and Welfare. Leuk Res 1995; 19:219–
225.
42. Maddox AM, Keating MJ, Smith TL, et al. Prognostic factors for survival of 194
patients with low infiltrate leukemia. Leuk Res 1986; 10:995–1006.
43. Keating MJ, Smith TL, Gehan EA, et al. Factors related to length of complete
remission in adult acute leukemia. Cancer 1980; 45:2017–2029.
44. A predictive model for aggressive non-Hodgkin’s lymphoma. The International Non-
Hodgkin’s Lymphoma Prognostic Factors Project. N Engl J Med 1993; 329:987–994.
45. Wimazal F, Sperr WR, Kundi M, et al. Prognostic significance of serial determina-
tions of lactate dehydrogenase (LDH) in the follow-up of patients with myelodys-
plastic syndromes. Ann Oncol 2008; 19:970–976.
46. Germing U, Hildebrandt B, Pfeilstocker M, et al. Refinement of the International
Prognostic Scoring System (IPSS) by including LDH as an additional prognos-
tic variable to improve risk assessment in patients with primary myelodysplastic
syndromes (MDS). Leukemia 2005; 19:2223–2231.
47. Luoni R, Ucci G, Riccardi A, et al. Serum thymidine kinase in hematologic malig-
nancies. Haematologica 1988; 73:31–35.
48. Hagberg H, Gronowitz S, Killander A, et al. Serum thymidine kinase in acute
leukaemia. Br J Cancer 1984; 49:537–540.
49. Aul C, Gattermann N, Germing U, et al. Serum deoxythymidine kinase in myelodys-
plastic syndromes. Cancer 1994; 73:322–327.
50. Musto P, Bodenizza C, Falcone A, et al. Prognostic relevance of serum thymidine
kinase in primary myelodysplastic syndromes: Relationship to development of acute
myeloid leukaemia. Br J Haematol 1995; 90:125–130.
51. Cappellini MD, Cohen A, Piga A, et al. A phase 3 study of deferasirox (ICL670),
a once-daily oral iron chelator, in patients with beta-thalassemia. Blood 2006;
107:3455–3462.
52. Malcovati L, Porta MG, Pascutto C, et al. Prognostic factors and life expectancy
in myelodysplastic syndromes classified according to WHO criteria: A basis for
clinical decision making. J Clin Oncol 2005; 23:7594–7603.
53. van der Weide M, Sizoo W, Nauta JJ, et al. Myelodysplastic syndromes: Analysis of
clinical and prognostic features in 96 patients. Eur J Haematol 1988; 41:115–122.
54. Thiele J, Vykoupil KF, Georgii A. Myeloid dysplasia (MD): A hematological dis-
order preceding acute and chronic myeloid leukemia. A morphological study on
sequential core biopsies of the bone marrow in 27 patients. Virchows Arch A Pathol
Anat Histol 1980; 389:343–367.
55. Seymour JF, Estey EH. The contribution of Auer rods to the classification and
prognosis of myelodysplastic syndromes. Leuk Lymphoma 1995; 17:79–85.
56. Boogaerts MA, Verhoef GE, Demuynck H. Treatment and prognostic factors in
myelodysplastic syndromes. Baillieres Clin Haematol 1996; 9:161–183.
57. Mangi MH, Mufti GJ. Primary myelodysplastic syndromes: Diagnostic and prog-
nostic significance of immunohistochemical assessment of bone marrow biopsies.
Blood 1992; 79:198–205.
58. Lambertenghi-Deliliers G, Annaloro C, Oriani A, et al. Prognostic relevance of
histological findings on bone marrow biopsy in myelodysplastic syndromes. Ann
Hematol 1993; 66:85–91.
59. Maschek H, Gutzmer R, Choritz H, et al. Life expectancy in primary myelodysplas-

tic syndromes: A prognostic score based upon histopathology from bone marrow
biopsies of 569 patients. Eur J Haematol 1994; 53:280–287.
60. Sultan C, Sigaux F, Imbert M, et al. Acute myelodysplasia with myelofibrosis: A
report of eight cases. Br J Haematol 1981; 49:11–16.
61. Verhoef GE, De Wolf-Peeters C, Ferrant A, et al. Myelodysplastic syndromes with
bone marrow fibrosis: A myelodysplastic disorder with proliferative features. Ann
Hematol 1991; 63:235–241.
62. Takahashi M, Koike T, Nagayama R, et al. Myelodysplastic syndrome with myelofi-
brosis: Myelodysplastic syndrome as a major primary disorder for acute myelofi-
brosis. Clin Lab Haematol 1991; 13:17–23.
63. Cassano E, Giordano M, Riccardi A, et al. Myelodysplastic syndromes: A multipara-
metric study of prognostic factors and a proposed scoring system. Haematologica
1990; 75:141–145.
64. Ohyashiki K, Sasao I, Ohyashiki JH, et al. Clinical and cytogenetic characteristics
of myelodysplastic syndromes developing myelofibrosis. Cancer 1991; 68:178–183.
65. Buesche G, Teoman H, Wilczak W, et al. Marrow fibrosis predicts early fatal marrow
failure in patients with myelodysplastic syndromes. Leukemia 2008; 22:313–322.
66. Maschek H, Kaloutsi V, Rodriguez-Kaiser M, et al. Hypoplastic myelodysplastic
syndrome: Incidence, morphology, cytogenetics, and prognosis. Ann Hematol 1993;
66:117–122.
67. Yoshida Y, Oguma S, Uchino H, et al. Refractory myelodysplastic anaemias with
hypocellular bone marrow. J Clin Pathol 1988; 41:763–767.
69. Appelbaum FR, Barrall J, Storb R, et al. Clonal cytogenetic abnormalities in patients
with otherwise typical aplastic anemia. Exp Hematol 1987; 15:1134–1139.
70. Huang TC, Ko BS, Tang JL, et al. Comparison of hypoplastic myelodysplastic syn-
drome (MDS) with normo-/hypercellular MDS by International Prognostic Scoring
System, cytogenetic and genetic studies. Leukemia 2008; 22:544–550.
71. Kitagawa M, Kamiyama R, Takemura T, et al. Bone marrow analysis of the
myelodysplastic syndromes: Histological and immunohistochemical features related
to the evolution of overt leukemia. Virchows Arch B Cell Pathol Incl Mol Pathol
1989; 57:47–53.
72. Moormeier JA, Rubin CM, Le Beau MM, et al. Trisomy 6: A recurring cytogenetic
abnormality associated with marrow hypoplasia. Blood 1991; 77:1397–1398.
73. de Planque MM, Kluin-Nelemans HC, van Krieken HJ, et al. Evolution of acquired
severe aplastic anaemia to myelodysplasia and subsequent leukaemia in adults. Br J
Haematol 1988; 70:55–62.
728.
76. Garand R, Gardais J, Bizet M, et al. Heterogeneity of acquired idiopathic siderob-
lastic anaemia (AISA). Leuk Res 1992; 16:463–468.
78. Germing U, Gattermann N, Minning H, et al. Problems in the classification of

CMML—Dysplastic versus proliferative type. Leuk Res 1998; 22:871–878.
79. Harris NL, Jaffe ES, Diebold J, et al. World Health Organization classification
of neoplastic diseases of the hematopoietic and lymphoid tissues: Report of the
Clinical Advisory Committee meeting—Airlie House, Virginia, November 1997. J
Clin Oncol 1999; 17:3835–3849.
new classification of primary myelodysplastic syndromes: A retrospective analysis
of 1600 patients. Leuk Res 2000; 24:983–992.
81. Germing U, Strupp C, Kuendgen A, et al. Prospective validation of the WHO
proposals for the classification of myelodysplastic syndromes. Haematologica 2006;
91:1596–1604.
82. Giagounidis AA, Germing U, Wainscoat JS, et al. The 5q– syndrome. Hematology
2004; 9:271–277.
83. Giagounidis AA, Haase S, Platzbecker U, et al. Risk factors for disease progression
in del(5q) MDS patients treated with lenalidomide. ASH Annu Meet Abstr 2007;
110:1460.
84. Billstrom R, Thiede T, Hansen S, et al. Bone marrow karyotype and prognosis in
primary myelodysplastic syndromes. Eur J Haematol 1988; 41:341–346.
85. Pierre RV, Catovsky D, Mufti GJ, et al. Clinical-cytogenetic correlations in
myelodysplasia (preleukemia). Cancer Genet Cytogenet 1989; 40:149–161.
86. Suciu S, Weh HJ, Hossfeld DK. Prognostic significance of the karyotype in patients
with primary myelodysplastic syndrome. Haematol Blood Transfus 1990; 33:166–
168.
87. Toyama K, Ohyashiki K, Yoshida Y, et al. Clinical implications of chromosomal
abnormalities in 401 patients with myelodysplastic syndromes: A multicentric study
in Japan. Leukemia 1993; 7:499–508.
88. Morel P, Hebbar M, Lai JL, et al. Cytogenetic analysis has strong independent
prognostic value in de novo myelodysplastic syndromes and can be incorporated in
a new scoring system: A report on 408 cases. Leukemia 1993; 7:1315–1323.
89. Bernasconi P, Klersy C, Boni M, et al. Incidence and prognostic significance of
karyotype abnormalities in de novo primary myelodysplastic syndromes: A study
on 331 patients from a single institution. Leukemia 2005; 19:1424–1431.
90. White AD, Hoy TG, Jacobs A. Extended cytogenetic follow-up and clinical progress
in patients with myelodysplastic syndromes (MDS). Leuk Lymphoma 1994; 12:401–
412.
91. Musilova J, Michalova K, Zemanova Z, et al. Karyotype at diagnosis, subsequent
leukemic transformation and survival in myelodysplastic syndrome. Czechoslovak
MDS Cooperative Group. Leuk Res 1995; 19:303–308.
2124 patients. Blood 2007; 110:4385–4395.
93. Parlier V, van Melle G, Beris P, et al. Hematologic, clinical, and cytogenetic analysis
in 109 patients with primary myelodysplastic syndrome. Prognostic significance of
morphology and chromosome findings. Cancer Genet Cytogenet 1994; 78:219–231.
94. Solé F, Espinet B, Sanz GF, et al. Incidence, characterization and prognostic signif-

tol 2000; 108:346–356.
95. Le Beau MM, Albain KS, Larson RA, et al. Clinical and cytogenetic correlations in 63
patients with therapy-related myelodysplastic syndromes and acute nonlymphocytic
leukemia: Further evidence for characteristic abnormalities of chromosomes no. 5
and 7. J Clin Oncol 1986; 4:325–345.
96. Van den Berghe H, Cassiman JJ, David G, et al. Distinct haematological disorder
with deletion of long arm of no. 5 chromosome. Nature 1974; 251:437–438.
97. Estey EH. Prognosis and therapy of secondary myelodysplastic syndromes. Haema-
tologica 1998; 83:543–549.
98. Heim S. Cytogenetic findings in primary and secondary MDS. Leuk Res 1992;
16:43–46.
99. Pedersen-Bjergaard J, Philip P, Larsen SO, et al. Chromosome aberrations and prog-
nostic factors in therapy-related myelodysplasia and acute nonlymphocytic leukemia.
Blood 1990; 76:1083–1091.
100. Kantarjian HM, Estey EH, Keating MJ. Treatment of therapy-related leukemia and
myelodysplastic syndrome. Hematol Oncol Clin North Am 1993; 7:81–107.
101. Steidl C, Steffens R, Gassmann W, et al. Adequate cytogenetic examination in
myelodysplastic syndromes: Analysis of 529 patients. Leuk Res 2005; 29:987–993.
102. Fenaux P, Morel P, Lai JL. Cytogenetics of myelodysplastic syndromes. Semin
Hematol 1996; 33:127–138.
103. Van den Berghe H, Vermaelen K, Mecucci C, et al. The 5q– anomaly. Cancer Genet
Cytogenet 1985; 17:189–255.
104. Mathew P, Tefferi A, Dewald GW, et al. The 5q– syndrome: A single-institution
study of 43 consecutive patients. Blood 1993; 81:1040–1045.
105. Boultwood J, Lewis S, Wainscoat JS. The 5q– syndrome. Blood 1994; 84:3253–
3260.
106. Sanz GF, Sanz MA. Prognostic factors in myelodysplastic syndromes. Leuk Res
1992; 16:77–86.
107. Tricot G, Boogaerts MA, De Wolf-Peeters C, et al. The myelodysplastic syndromes:
Different evolution patterns based on sequential morphological and cytogenetic
investigations. Br J Haematol 1985; 59:659–670.
108. Benitez J, Carbonell F, Sanchez Fayos J,et al. Karyotypic evolution in patients with
myelodysplastic syndromes. Cancer Genet Cytogenet 1985; 16:157–167.
109. Mecucci C, La Starza R. Cytogenetics of myelodysplastic syndromes. Forum (Gen-
ova) 1999; 9:4–13.
110. Clark R, Peters S, Hoy T, et al. Prognostic importance of hypodiploid hemopoietic
precursors in myelodysplastic syndromes. N Engl J Med 1986; 314:1472–1475.
111. van Kamp H, de Pijper C, Verlaan-de Vries M, et al. Longitudinal analysis of
point mutations of the N-ras proto-oncogene in patients with myelodysplasia using
archived blood smears. Blood 1992; 79:1266–1270.
112. Yunis JJ, Boot AJ, Mayer MG, et al. Mechanisms of ras mutation in myelodysplastic
syndrome. Oncogene 1989; 4:609–614.
113. Bartram CR. Molecular genetic aspects of myelodysplastic syndromes. Semin Hema-
tol 1996; 33:139–149.
114. Parker J, Mufti GJ. Ras and myelodysplasia: Lessons from the last decade. Semin
Hematol 1996; 33:206–224.
115. Paquette RL, Landaw EM, Pierre RV, et al. N-ras mutations are associated with poor
prognosis and increased risk of leukemia in myelodysplastic syndrome. Blood 1993;
82:590–599.
116. Constantinidou M, Chalevelakis G, Economopoulos T, et al. Codon 12 ras mutations
in patients with myelodysplastic syndrome: Incidence and prognostic value. Ann
Hematol 1997; 74:11–14.
892.
118. Neubauer A, Greenberg P, Negrin R, et al. Mutations in the ras proto-oncogenes in
patients with myelodysplastic syndromes. Leukemia 1994; 8:638–641.
119. Lyons J, Janssen JW, Bartram C, et al. Mutation of Ki-ras and N-ras oncogenes in
120. Ridge SA, Worwood M, Oscier D, et al. FMS mutations in myelodysplastic,
leukemic, and normal subjects. Proc Natl Acad Sci U S A 1990; 87:1377–1380.
121. Bouscary D, Preudhomme C, Ribrag V, et al. Prognostic value of c-mpl expression
in myelodysplastic syndromes. Leukemia 1995; 9:783–788.
122. Harris CC, Hollstein M. Clinical implications of the p53 tumor-suppressor gene. N
Engl J Med 1993; 329:1318–1327.
123. Mori N, Wada M, Yokota J, et al. Mutations of the p53 tumour suppressor gene in
haematologic neoplasms. Br J Haematol 1992; 81:235–240.
124. Tsushita K, Hotta T, Ichikawa A, et al. Mutation of p53 gene does not play a critical
role in myelodysplastic syndrome and its transformation to acute leukaemia. Br J
Haematol 1992; 81:456–457.
125. Sugimoto K, Hirano N, Toyoshima H, et al. Mutations of the p53 gene in myelodys-
plastic syndrome (MDS) and MDS-derived leukemia. Blood 1993; 81:3022–3026.
126. Kaneko H, Misawa S, Horiike S, et al. TP53 mutations emerge at early phase of
myelodysplastic syndrome and are associated with complex chromosomal abnor-
malities. Blood 1995; 85:2189–2193.
127. Jonveaux P, Fenaux P, Quiquandon I, et al. Mutations in the p53 gene in myelodys-
plastic syndromes. Oncogene 1991; 6:2243–2247.
128. Orazi A, Cattoretti G, Heerema NA, et al. Frequent p53 overexpression in therapy
related myelodysplastic syndromes and acute myeloid leukemias: An immunohisto-
chemical study of bone marrow biopsies. Mod Pathol 1993; 6:521–525.
129. Wattel E, Preudhomme C, Hecquet B, et al. p53 mutations are associated with
resistance to chemotherapy and short survival in hematologic malignancies. Blood
1994; 84:3148–3157.
130. Quesnel B, Guillerm G, Vereecque R, et al. Methylation of the p15(INK4b) gene
in myelodysplastic syndromes is frequent and acquired during disease progression.
Blood 1998; 91:2985–2990.
131. Uchida T, Kinoshita T, Nagai H, et al. Hypermethylation of the p15INK4B gene in
132. Mihara K, Chowdhury M, Nakaju N, et al. Bmi-1 is useful as a novel molecular
marker for predicting progression of myelodysplastic syndrome and patient progno-
sis. Blood 2006; 107:305–308.
133. Hofmann WK, de Vos S, Komor M, et al. Characterization of gene expression
of CD34+ cells from normal and myelodysplastic bone marrow. Blood 2002;
100:3553–3560.
135. Chen G, Zeng W, Miyazato A, et al. Distinctive gene expression profiles of CD34
cells from patients with myelodysplastic syndrome characterized by specific chro-
mosomal abnormalities. Blood 2004; 104:4210–4218.
prognostic subgroups in low- and intermediate-1-risk myelodysplastic syndromes
by flow cytometry. Blood 2008; 111:1067–1077.
137. Baur AS, Meuge-Moraw C, Schmidt PM, et al. CD34/QBEND10 immunostaining in
bone marrow biopsies: An additional parameter for the diagnosis and classification
of myelodysplastic syndromes. Eur J Haematol 2000; 64:71–79.
138. Soligo D, Delia D, Oriani A, et al. Identification of CD34+ cells in normal and
pathological bone marrow biopsies by QBEND10 monoclonal antibody. Leukemia
1991; 5:1026–1030.
139. Oriani A, Annaloro C, Soligo D, et al. Bone marrow histology and CD34 immunos-
taining in the prognostic evaluation of primary myelodysplastic syndromes. Br J
Haematol 1996; 92:360–364.
140. Sullivan SA, Marsden KA, Lowenthal RM, et al. Circulating CD34+ cells: An
adverse prognostic factor in the myelodysplastic syndromes. Am J Hematol 1992;
39:96–101.
the IPSS and with outcome after hematopoietic stem cell transplantation. Blood
2003; 102:394–403.
142. Pfeilstocker M, Reisner R, Nosslinger T, et al. Cross-validation of prognostic scores
in myelodysplastic syndromes on 386 patients from a single institution confirms
importance of cytogenetics. Br J Haematol 1999; 106:455–463.
143. Aul C, Giagounidis A, Germing U, et al. Evaluating the prognosis of patients with
myelodysplastic syndromes. Ann Hematol 2002; 81:485–497.
144. Muller-Berndorff H, Haas PS, Kunzmann R, et al. Comparison of five prognostic
scoring systems, the French-American-British (FAB) and World Health Organization
(WHO) classifications in patients with myelodysplastic syndromes: Results of a
single-center analysis. Ann Hematol 2006; 85:502–513.
145. Worsley A, Oscier DG, Stevens J, et al. Prognostic features of chronic myelomono-
cytic leukaemia: A modified Bournemouth score gives the best prediction of survival.
Br J Haematol 1988; 68:17–21.
146. Goasguen JE, Garand R, Bizet M, et al. Prognostic factors of myelodysplastic
syndromes—A simplified 3-D scoring system. Leuk Res 1990; 14:255–262.
147. Sanz GF, Sanz MA, Greenberg PL. Prognostic factors and scoring systems in
148. Greenberg PL. Risk factors and their relationship to prognosis in myelodysplastic
syndromes. Leuk Res 1998; 22(Suppl. 1):S3-S6.
16
Therapeutic Strategies: The Approach to
Care of Patients with MDS, and Criteria
for Treatment Response
Pierre Fenaux, Lionel Ades and Claude Gardin

Hôpital Avicenne (Assistance Publique-Hôpitaux de Paris),
INTRODUCTION
The treatment of MDS remains difficult and, apart from allogeneic stem cell
transplantation (SCT), is generally not curative. However, some major improve-
ments have been made in the last few years. They arise from improvements in
allogeneic SCT, including better donor–recipient matching by allelic HLA typ-
ing, nonmyeloablative conditioning that allows transplanting older patients, and
more effective anti-infective treatments during the posttransplant period, includ-
ing antifungal agents (see chap. 21). They also arise from the advent of new
drugs like hypomethylating agents (see chap. 20) that, at least for one of them—
azacytidine—can prolong survival in higher risk MDS, and lenalidomide, which
appears to be a “targeted” drug in MDS with del(5q) (chap. 19). Further progress
results from more rationale use of erythroid stimulating agents (ESAs) including
EPO and darbepoietin (chap. 17), and from better recognition of iron overload due
to RBC transfusions and its reversal by adequate chelation therapy (chap. 7).
Therapeutic strategy in MDS, remains largely based on prognostic factors
derived from the International Prognostic Scoring System (IPSS) (1). Therapeutic
guidelines for MDS have been published in the United Kingdom (2) and Italy (3),
and online guidelines are available in the United States (created by an organiza-
tion called the National Comprehensive Cancer Network; www.nccn.org), Nordic
393
394 Fenaux et al.
countries (www.nordicmds.org), and France (www.gfmgroup.org). Response cri-

teria to treatments have been adapted to MDS by an International Working Group,
making the evaluation of new therapies more consistent (Tables 1 and 2) (4).
Here, we will only summarize the different treatment options in MDS
patients, which will be described in detail in chapters 17 through 21, and focus
more on overall treatment approaches, that is, which therapies are most appropriate
for which patients.
IPSS-BASED STRATEGY AND RESPONSE CRITERIA IN MDS

IPSS-Based Strategy
Most clinical groups largely base the therapeutic options they propose to their
MDS patients on IPSS results (see chap. 1) (1–3). Because patients with high and
intermediate-2 IPSS risk categories have a median survival of only 0.4 and 1.2
years, respectively; therapeutic options aimed at modifying the disease course,
avoiding progression to AML, and improving survival are clearly required in this
population. On the contrary, patients fitting into low and intermediate-1 IPSS
risk groups have a median survival of 5.7 and 3.4 years, respectively, and they
often die from other causes than MDS. Therefore, their treatment mainly aims
at ameliorating the consequences of cytopenias and transfusions, and improving
quality of life.
This schematic dichotomy may, of course, suffer exceptions, such as the use
of lenalidomide in lower risk MDS with del(5q), as this drug appears to be able
to suppress the del(5q) clone and may thereby have a positive effect on long-term
survival (4,5).
Response Criteria in MDS

In order to take into account the fact that some treatments can improve their
cytopenias in MDS, although they do not appear to change the disease course,
specific response criteria in addition to those used “classically” for diseases like
AML (including complete or partial remission, etc.) had to be defined for MDS.
In 2000, an international working group (IWG) designed a set of response criteria
divided into two parts (Table 1): responses linked to a possible change in disease
course, including complete remission (CR), partial remission (PR), stable disease
(SD), and progressive disease (PD); and cytogenetic response, partial or complete,
in patients with an abnormal clone at the onset of treatment (6). Those applied
mainly to higher risk MDS, and intensive treatments (like allogeneic SCT or
chemotherapy) and were largely defined by percentages of marrow blasts and
normalization or not of cytopenias after treatment (6).
On the other hand, especially for lower risk MDS and treatments that mainly
aimed at improving cytopenias, responses of “hematological improvement” (HI)
in either of myeloid lineages were defined: HI erythroid (HI-E), HI platelets
(HI-P), HI neutrophils (HI-N). Responses could be “major” or “minor” for each
Therapeutic Strategies 395
Table 1 IWG 2000: Measurement of Response/Treatment Effect in MDS (6)

Altering disease natural history
1. Complete remission (CR)

(a) Bone marrow evaluation: Repeat bone marrow showing less than 5% myeloblasts
with normal maturation of all cell lines, with no evidence for dysplasia. a When
erythroid precursors constitute less than 50% of bone marrow nucleated cells, the
percentage of blasts is based on all nucleated cells; when there are 50% or more
erythroid cells, the percentage blasts should be based on the nonerythroid cells.
(b) Peripheral blood evaluation (absolute values must last at least 2 mo)b :
Hemoglobin greater than 11 g/dL (untransfused, patient not on erythropoietin)
Neutrophils 1500/mm3 or more (not on a myeloid growth factor)
Platelets 100,000/mm3 or more (not on a thrombopoetic agent)
Blasts, 0%
No dysplasiaa
2. Partial remission (PR) (absolute values must last at least 2 mo): All the CR criteria (if
abnormal before treatment), except: Bone marrow evaluation: Blasts decreased by 50%
or more over pretreatment, or a less advanced MDS FAB classification than
pretreatment. Cellularity and morphology are not relevant.
3. Stable disease
Failure to achieve at least a PR, but with no evidence of progression for at least 2 mo.
4. Failure: Death during treatment or disease progression characterized by worsening of
cytopenias, increase in the percentage bone marrow blasts, or progression to an MDS
FAB subtype more advanced than pretreatment.
5. Relapse after CR or PR: one or more of the following:
(a) Return to pretreatment bone marrow blast percentage.
(b) Decrement of 50% or greater from maximum remission/response levels in
granulocytes or platelets.
(c) Reduction in hemoglobin concentration by at least 2 g/dL or transfusion
dependence.c
6. Disease progression
(a) For patients with less than 5% blasts: a 50% or more increase in blasts to more than
5% blasts.
(b) For patients with 5% to 10% blasts: a 50% or more increase to more than 10% blasts.
(c) For patients with 10% to 20% blasts: a 50% or more increase to more than 20%
blasts.
(d) For patients with 20% to 30% blasts: a 50% or more increase to more than 30%
blasts.
(e) One or more of the following: 50% or greater decrement from maximum
remission/response levels in granulocytes or platelets, reduction in hemoglobin
concentration by at least 2 g/dL, or transfusion dependence.c
7. Disease transformation: Transformation to AML (30% or more blasts).
(Continued)
396 Fenaux et al.

(Continued)
8. Survival and progression-free survival

Cytogenetic response: Requires 20 analyzable metaphases using conventional cytogenetic
techniques.
(1) Major: No detectable cytogenetic abnormality, if preexisting abnormality was
present.
(2) Minor: 50% or more reduction in abnormal metaphases.
(3) Fluorescent in situ hybridization may be used as a supplement to follow a
specifically defined cytogenetic abnormality.
Quality of life: Measured by an instrument such as the FACT Questionnaire. Clinically
useful improvement in specific domains:
Physical
Functional
Emotional
Social
Spiritual
Hematologic improvement (HI): Improvements must last at least 2 mo in the absence of
ongoing cytotoxic therapyb . Hematologic improvement should be described by the
number of individual, positively affected cell lines (e.g., HI-E; HI-E + HI-N; HI-E +
HI-P + HI-N).
1. Erythroid response (HI-E)
(a) Major response: For patients with pretreatment hemoglobin less than 11 g/dL,
greater than 2 g/dL increase in hemoglobin; for RBC transfusion–dependent
patients, transfusion independence.
(b) Minor response: For patients with pretreatment hemoglobin less than 11 g/dL,
1–2 g/dL increase in hemoglobin; for RBC transfusion–dependent patients,
50% decrease in transfusion requirements.
2. Platelet response (HI-P)
(a) Major response: For patients with a pretreatment platelet count less than
100,000/mm3 , an absolute increase of 30,000/mm3 or more; for platelet
transfusion–dependent patients, stabilization of platelet counts and platelet
transfusion independence.
(b) Minor response: For patients with a pretreatment platelet count less than
100,000/mm3 , a 50% or more increase in platelet count with a net increase greater
than 10,000/mm3 but less than 30,000/mm3 .
3. Neutrophil response (HI-N)
(a) Major response: For absolute neutrophil count (ANC) less than 1500/mm3 before
therapy, at least a 100% increase, or an absolute increase of more than 500/mm3 ,
whichever is greater.
(b) Minor response: For ANC less than 1500/mm3 before therapy, ANC increase of at
least 100%, but absolute increase less than 500/mm3 .

(Continued)
4. Progression/relapse after HI: One or more of the following: a 50% or greater decrement
from maximum response levels in granulocytes or platelets, a reduction in hemoglobin
concentration by at least 2 g/dL, or transfusion dependence.c
For a designated response (CR, PR, HI), all relevant response criteria must be noted on at least 2
successive determinations at least 1 wk apart after an appropriate period following therapy (e.g., 1 mo
or longer).
a The presence of mild megaloblastoid changes may be permitted if they are thought to be consistent
with treatment effect. However, persistence of pretreatment abnormalities (e.g., pseudo-Pelger-Hüet

cells, ringed sideroblasts, dysplastic megakaryocytes) are not consistent with CR.
b In some circumstances, protocol therapy may require the initiation of further treatment (e.g., consol-
idation, maintenance) before the 2-mo period. Such patients can be included in the response category
into which they fit at the time the therapy is started.
c In the absence of another explanation such as acute infection, gastrointestinal bleeding, hemolysis,
and so on.
lineage based on precise thresholds, and included, for red cells, and platelets, the
importance of transfusion requirement and its modification with treatment.
A recent update of IWG 2000 response criteria has recently been published
(IWG 2006) (Table 2) (7). Modifications brought by this classification are in
responses that may modify disease course, the introduction of a category of “mar-
row response” that includes patients who reduce their percentage marrow blasts
to less than 5%, but remain cytopenic. HI was redefined for each lineage, gen-
erally with more stringent criteria. In particular, HI-E, HI-P, and HI-N “minor”
disappeared—these patients being reclassified as responders or nonresponders
based on the importance of cytopenia and/or transfusion requirement improve-
ment.
In spite of some critics which can be made to those IWG response criteria
(for example the fact that the new IWG 2006 criteria consider no more as response
some minor response per IWG 2000, that however were relatively prolonged) they
constitute a “common language” for publication of clinical trials in MDS.
THERAPEUTIC STRATEGY IN HIGHER RISK MDS (IPSS INT2 OR HIGH)

Treatments Available
Allogeneic SCT
It is described in detail in chapter 22. Because allogeneic SCT remains the only
curative treatment in MDS, there is a consensus over the fact that it should be
envisaged in all patients with int2 or high-risk MDS who have a donor and no
contra-indication to the procedure (8). The age limit for SCT is generally consid-
ered to be around 70 years in most centers (with nonmyeloablative allogeneic SCT
398 Fenaux et al.
Table 2 Proposed Modified International Working Group Response Criteria for

Altering Natural History of MDS in 2006 (7)
Response criteria (responses must last at least 4 wk)
Complete remission Bone marrow: 5% myeloblasts with normal maturation of all

cell linesa
b
Persistent dysplasia will be noteda
c
Peripheral blood
Hgb 11g/dL
Platelets 100 × 109 /L
Neutrophils 1.0 × 109 /L
Blasts 0%
Partial remission All CR criteria if abnormal before treatment except:
Bone marrow blasts decreased by 50% over pretreatment but
still ⬎5%
Cellularity and morphology not relevant
b
Marrow CR Bone marrow: 5% myeloblasts and decrease by 50% over
pretreatment
b
Peripheral blood: if HI responses, they will be noted in
addition to marrow CR
Stable disease Failure to achieve at least PR, but no evidence of progression
for ⬎8 wk
Failure Death during treatment or disease progression characterized by
worsening of cytopenias, increase in percentage of bone
marrow blasts, or progression to a more advanced MDS FAB
subtype than pretreatment
Relapse after CR or At least one of the following:
PR Return to pretreatment bone marrow blast percentage
Decrement of 50% from maximum remission/response
levels in granulocytes or platelets
Reduction in Hgb concentration by 1.5 g/dL or transfusion
dependence
Cytogenetic Complete: disappearance of the chromosomal abnormality
response without appearance of new ones
Partial: at least 50% reduction of the chromosomal abnormality
Disease progression For patients with:
Less than 5% blasts: 50% increase in blasts to ⬎5% blasts
5–10% blasts: 50% increase to ⬎10% blasts
Any of the following:
At least 50% decrement from maximum remission/response
in granulocytes or platelets
Reduction in Hgb by 2 g/dL
Transfusion dependence
Table 2 Proposed Modified International Working Group Response Criteria for

Altering Natural History of MDS in 2006 (7) (Continued)

Survival Endpoints:
Overall: death from any cause
Event free: failure or death from any cause
PFS: disease progression or death from MDS
DFS: time to relapse
Cause-specific death: death related to MDS
a Dysplastic changes should consider the normal range of dysplastic changes (modification) (41).
b Modification to IWG response criteria.
c In some circumstances, protocol therapy may require the initiation of further treatment (e.g., consol-
idation, maintenance) before the 4-wk period. Such patients can be included in the response category
into which they fit at the time the therapy is started. Transient cytopenias during repeated chemotherapy
courses should not be considered as interrupting durability of response, as long as they recover to the
improved counts of the previous course.
Hematologic improvementa b
Erythroid response Hgb increase by 1.5 g/dL

(pretreatment ⬍11 g/dL) Relevant reduction of units of RBC transfusions by an
absolute number of at least 4 RBC transfusions/8 wk
compared with the pretreatment transfusion number
in the previous 8 wk. Only RBC transfusions given
for a Hgb of 9.0 g/dL pretreatment will count in the
RBC transfusion response evaluation
Platelet response (pretreatment Absolute increase of 30 × 109 /L for patients starting
⬍100 × 109 /L) with ⬎20 × 109 /L platelets
b
Increase from ⬍20 × 109 /L to ⬎20 × 109 /L and by
at least 100%
b
Neutrophil response At least 100% increase and an absolute increase
(pretreatment ⬍1.0 × 109 /L) ⬎0.5 × 109 /L
c
Progression or relapse after HI At least one of the following:
At least 50% decrement from maximum response
levels in granulocytes or platelets
Reduction in Hgb by 1.5 g/dL
Transfusion dependence
Deletions to the IWG response criteria are not shown. To convert hemoglobin levels from grams per
deciliter to grams per liter, multiply grams per deciliter by 10.
a Pretreatment counts averages of at least 2 measurements (not influenced by transfusions) 1 wk apart
(modification).
b Modification to IWG response criteria.
c In the absence of another explanation, such as acute infection, repeated courses of chemotherapy
(modification), gastrointestinal bleeding, hemolysis, and so forth. It is recommended that the 2 kinds
of erythroid and platelet responses be reported overall as well as by the individual response pattern.
Abbreviations: Hgb, hemoglobin; RBC, red blood cell; HI, hematologic improvement.
400 Fenaux et al.
generally considered above 50–55 years). Allogeneic SCT is generally restricted

to high- and int2-risk patients, largely based on results from an important IBMTR
study (9). This retrospective work analyzed the outcome of MDS transplanted at
diagnosis of MDS, after a certain delay or upon progression to AML. The authors
found in terms of “survival years,” a benefit for transplanting rapidly patients with
high and int2 IPSS, and on the contrary waiting for progression in low-risk MDS,
while there was more discussion for int1 patients. In all cases, however, transplant
should be performed prior to over AML progression (10).
Many issues, that will be addressed in the chapter devoted to allogeneic
SCT, however, remain with this treatment. First, whereas classical myeloablative
allogeneic SCT can give long-term cure in a substantial proportion of MDS, long-
term results of nonmyeloablative allogeneic SCT, the only possible procedure in
many MDS patients due to their age, are less certain in terms of clonal eradica-
tion. It is also uncertain whether allogeneic SCT transplant should be preceded
by cytoreductive treatment (11). The high relapse rate after transplant in patients
who had an excess of marrow blasts at the time of transplant (12), especially in
the case of nonmyeloablative SCT, suggests the usefulness of previous cytore-
ductive treatment, but this attitude has not been tested prospectively. The use of
hypomethylating agents prior to transplant may be another option, currently tested
by several roups.
Whatever be the current debates over allogeneic SCT in higher risk MDS,
it is, however, clear that it is currently the best therapeutic option, and that HLA
typing of patients and their siblings should be rapidly performed, and unrelated
donors (and probably compatible cord blood grafts) searched in the absence of
HLA identical siblings, unless the patient has a contra indication to allogeneic
SCT.
Chemotherapy
r Classical anthracycline–Ara C chemotherapy
Many large reports using classical anthracycline–Ara C chemotherapy have shown

that this approach yielded lower CR rates (40–60%) in MDS (having or not
progressed to AML) than in de novo AML (see chap. 21). In addition, median
CR duration is only 10 to 12 months, and very few patients obtain prolonged CR
irrespective of the type of consolidation therapy (1 or 2 intensive consolidation
courses or only “1+5” moderate consolidation courses, etc.) (13). Especially
because the period of myelosuppression, in MDS treated with classical intensive
chemotherapy, tends to be longer than for de novo AML—this treatment remains
restricted to patients aged less than 60 or 65 (14). Cytogenetics are the main
prognostic factor of response and of its duration, with monosomy 7 and complex
karyotype being associated with particularly low CR rates and very short CR
duration (13).
No drug combination has proven superior to classical anthracycline–Ara
C combinations. In fact, combinations like VP16–Ara C, fludarabine–Ara C, or
Topotecan–Ara C appear to give inferior results, a finding similar to that gener-

ally found in AML (15–17). New chemotherapeutic agents, like clofarabine and
cloretazine, are, however, being tested in this indication (18).
r Low-dose chemotherapy
Low-dose Ara C (generally 20 mg/m2 /day, 14 days every month) has been used
for many years in higher risk MDS, with CR and PR rates of about 15% and 15%,
respectively, and short responses (median around 6 months) (19,20). Although less
intensive than anthracycline–Ara C regimens, it is still quite myelosuppressive,
and toxic deaths have been reported in 10% to 15% of the patients. Like for
anthracycline–Ara C chemotherapy, karyotype is the strongest prognostic factors,
and almost no responses are obtained in patients with −7 or complex karyotype. In
a randomized trial low-dose Ara C showed no survival benefit over best supportive
care in MDS (21). However, in that trial, patients received only one course of low-
dose Ara C. On the other hand, in the phase III randomized trial that compared
azacytidine and three therapeutic approaches including low-dose Ara C, survival
with low-dose AraC was inferior to that observed with azacytidine, although
patients, in the low-dose Ara C subgroup, had received a median of 4.5 courses of
the drug (22).
Few other strategies with low-dose chemotherapy have been used in MDS.
Low-dose melphalan (2 mg/day for several weeks) has been reported to give about
one-third of CR or PR, with median response duration of several months, in two
small phase II studies (23,24).
Hypomethylating Agents
Use of hypomethylating agents in MDS will be detailed in chapter 40. Their grow-
ing interest in MDS has come from basic work showing the importance of gene
methylation in MDS, especially in higher risk MDS and in the progression from
MDS to AML, but more importantly from clinical studies supporting a survival
benefit for one of the currently available hypomethylating agents—azacytidine,
over best supportive care in a phase III randomized trial in MDS (22). Because,
possibly due to the crossover desingn of the study, the survival advantage obtained
with azacytidine was not significant, another phase III trial was performed. It was
restricted to higher risk MDS, without crossover design, and, in the control group
clinicians could choose between intensive chemotherapy, low-dose Ara C or best
supportive care. Results of this trial, recently obtained and so far only presented at
meetings, show a significant survival with azacytidine, not only over BSC but also
over low-dose Ara C and intensive chemotherapy, and support the use of this drug
as first-line treatment in higher risk MDS. Results obtained with the other available
hypomethylating agent, decitabine, show that this drug may be particularly active
in patients with unfavorable karyotype, that is, patients with monosomy 7 and/or
complex karyotype, and preliminary findings suggest that the same holds true for
azacytidine (25–27).
402 Fenaux et al.
Another important finding with hypomethylating agents is that they appear

to work slowly, so that response is often obtained only after 3 to 4 cycles or even
more. One should generally therefore not stop treatment before four courses of
treatment. Importantly, too, at least for azacytidine, survival improvement is seen
in spite of only modest CR or even PR rates (28,29). This is different from the
usual results seen with chemotherapy, especially in AML, where CR achievement
is generally required to obtain any survival improvement. Hypomethylating agents
appear to prolong survival by improving cytopenias and by preventing progression
to AML or even going back to an earlier MDS stage, than by “clonal eradication.”
Also for this reason, it remains uncertain for how long they should be administered
once a response has been obtained.
Investigational Agents
Investigational agents include in particular histone deacytylase inhibitors, farnesyl
transferase inhibitors, arsenic trioxide, and are currently being tested alone or in
combination with hypomethylating agents (30–32). They, however, cannot for the
moment be advocated in the first-line treatment of higher risk MDS.
Therapeutic Strategy in Higher Risk MDS (IPSS High and Int2)

Proposed therapeutic strategies are summarized in Figure 1.
(a) Patients with a donor: Allogeneic SCT should be performed whenever possi-
ble, in “intensive therapy candidates” (patients aged generally less than 65–70
years without major comorbidities). As said above, it is uncertain whether
allogeneic SCT should be performed as frontline treatment, or should be pre-
ceded by treatment aimed at reducing the tumor burden. In spite of the absence
of prospective studies, it is generally considered that the marrow blast per-
centage should be reduced to less than 10% prior to classical allogeneic SCT,
and perhaps to less than 5% in nonmyeloablative SCT, in order to reduce the
otherwise important relapse rate posttransplant. Most authors applied, until
recently, classical anthracycline–Ara C regimens when they considered that
blast reduction should be obtained prior to transplant. Although this remains
almost a consensus in patients with normal karyotype, many authors would
now rather use hypomethylating agents prior to transplant in patients with
abnormal karyotype, although this attitude has not been validated prospec-
tively in clinical trials.
(b) Patients without a donor candidates for intensive treatment: Intensive
anthracycline-chemotherapy was considered, in spite of its generally short-
lived results, as the mainstay in case of normal cytogenetics, whereas
hypomethylating agents were more and more used in the presence of
abnormal (especially unfavorable) karyotype. Results of the azacytidine ran-
domized phase III study, suggesting a survival benefit of azacytidine over
intensive chemotherapy (although numbers of patients were small for com-
parison), and historical survival comparison between decitabine and intensive
Int-2 and high-risk
patients
Candidate Non candidate

to intensive therapy to intensive therapy
Donor available for HSCT

and less than 5–10% blasts* Hypomethylating agents
Therapeutic Strategies
HSCT
If failure,clinical trial
Donor available for HSCT No donor available for HSCT

and more than 5–10% blasts*
Abnormal cytogenetics Normal cytogenetics Normal cytogenetics: Abnormal cytogenetics:

intensive chemotherapy? hypomethylating agents
hypomethylating agents?
Hypomethylating agents Intensive chemotherapy If failure, If failure,

(MTI) (IC) clinical trial intensive chemotherapy?
clinical trial
HSCT if response IC if failure HSCT if response MTI if failure
HSCT if response
Figure 1 Treatment of int2 or high-risk MDS patients.

403
404 Fenaux et al.
chemotherapy favor the larger use of hypomethylating agents in the absence

of unfavorable karyotype (33). It is possible that intensive chemotherapy will
remain only for younger patients with rapidly evolving disease and normal
karyotype.
(c) In patients considered unfit for intensive treatment, who account for the
majority of higher risk MDS, hypomethylating agents would now be con-
sidered as first-line treatment by most groups. As said before, failure should
generally not be considered before the patient has received four treatment
courses. Once response has been achieved, the optimal duration of mainte-
nance treatment is unknown, but should probably not be less than 6 months.
Many centers still use, in this situation, low-dose chemotherapy, mainly
low-dose Ara C. As noted above, however, this drug can give responses only in
the absence of unfavorable karyotype, and leads to more profound cytopenias than
hypomethylating agents. In addition, recent results of the phase III randomized
trial showed a survival advantage of azacytidine over low-dose Ara C. Results of
this study may lead to complete replacement of low-dose Ara C in those patients,
even in the the case of normal karyotype.
TREATMENT OF LOWER RISK MDS (IPSS LOW OR INT1)

The major aims of treatment, in lower risk MDS, are to improve cytopenias,
thereby improving quality of life. Anemia is the most prominent cytopenia in
most cases of lower risk MDS, and focuses most of the attention.
Treatment of Anemia in Lower Risk MDS: RBC Transfusions or Drugs

Aiming at Preventing Anemia Recurrence?
A first issue in the treatment of anemia in lower risk MDS is whether clinicians
should just try to improve it by RBC transfusions or try to increase Hb level by
drug therapy, thereby avoiding transfusions. More and more physicians tend to
use agents aimed at preventing anemia recurrence, for several reasons. Risks of
viral infection or of alloimmunization due to RBC transfusions, which are now
very low, do not constitute the main reasons. Main reasons include the fact that
patients regularly transfused in RBC spend most, if not all, of their time with Hb
levels below 10 g/dL, which are associated to permanent fatigue, lower quality of
life and possibly more cardiovacular events than in patients where Hb level can
be maintained above 10 to 11 g/dL (34). In addition, regularly transfused patients
require at some point iron chelation therapy. The repetition of hospital visits
for transfusions is time consuming and induces some “dependance” of patients
towards the medical system. The cost associated with transfusions (including
transportation to and from the hospital, serum testing, iron chelation, and days
off work for MDS patients who may still be working) almost amounts to that
of ESAs. Finally, we have recently found that in lower risk MDS with anemia,
receiving ESAs had no impact on progression to AML but was an independant
favorable prognostic factor for survival, strongly supporting the assumption that
ESA treatment and/or maintaining hemoglobin level higher than 10 g/L and/or
avoiding iron overload could reduce the risk of nonleukemic deaths in lower risk
MDS (35).
Treatments That Can Prevent Anemia Recurrence in MDS

Erythrocytic Stimulating Agents (ESAs)
The use of ESAs, which include in clinical practice EPO alfa, EPO beta, and
darbepoietin will be detailed in chapter 17. They remain the first choice for the
treatment of anemia in lower risk MDS without del(5q), and which do not have
a high “nordic” score, that is, the combination of low serum EPO level and high-
transfusion requirement. Their efficacy can be improved by the addition of G-CSF
(36). Clinicians may therefore either combine both agents initially and try to stop
G-CSF after response has been achieved or, on the contrary, start with an ESA
alone and add G-CSF only in case of failure. Contrary to previous findings, we did
not find in a large patient series that RARS responded less well to an ESA alone
than RA, or that patients with multilineage dysplasia had lower response rates,
although their responses to ESAs were shorter (35).
Most responders to ESA respond within 12 weeks of treatment. Optimal
doses used to obtain a response are 30,000 to 60,000 UI/wk for EPO alfa and beta
and 150 to 300 ␮g/wk for darbepoietin, which can be reduced in case of response,
in order to avoid an increase in Hb level to more than 12 g/dL, according to health
authorities guidelines issued by most countries.
Median duration of response to ESA is about 2 years (35). Interestingly,
at least one half of the relapses are not associated to overt disease progression
(i.e., increase in marrow blasts, occurence of other cytopenias, etc.) and appear to
correspond to simple loss of sensitivity of erythroid progenitors to ESAs.
Thalidomide and Lenalidomide

Low-dose thalidomide (100–200 mg/day) can yield 20% to 30% major erythroid
response in lower risk MDS resistant to or relapsing after EPO, while higher doses
are generally not tolerated in those patients due to usual side effects of thalido-
mide (37–39). Responses appear similar in del(5q) and non-del(5q) patients (40).
Even at low dose, debilitating side effects may, however, be observed (short-term
somnolence, constipation and dizziness, and long-term peripheral neuropathy).
By contrast lenalidomide, a thalidomide derivative, has dramatic activity on
anemia in lower risk MDS with del(5q), that will be analyzed in detail in chapter 19
(4,5). In lower risk MDS without del(5q), generally resistant to ESAs, transfusion
independance and ≥50% reduction in transfusion requirement were achieved in
27% and 17% patients, respectively, but median response duration was less than a
year (41).
406 Fenaux et al.
Antithymocyte Globulin (ATG)

The use of ATG, with or without ciclosporine in lower risk MDS, will be detailed
in chapter 18. ATG may be active not only on anemia but also on other cytopenias,
especially thrombocytopenia. Its results are better in relatively younger patients
with a RBC transfusion history of less than 2 years, in patients with normal
karyotype or trisomy 8, and possibly in patients who are HLA DR15, who have a
small PNH clone or have marrow hypocellularity (42).
Hypomethylating Agents
Although hypomethylating agents have been more systematically tested in higher
risk MDS, they are also active in lower risk MDS, especially on anemia where
transfusion independence may be achieved in 30% to 40% of patients resistant or
relapsing after ESAs (28,29).
Iron Chelation Therapy

Although this has not been as well established as for children (e.g., children with
thalassemia major), it is very probable that iron overload, especially in the heart and
liver, reduces survival through cardiac failure or liver cirrhosis. Likewise, although
no randomized studies with and without iron chelators have been performed in
MDS, it has been strongly suggested that adequate chelation in highly transfused
patients improves their survival, especially in a prospective study reported at ASH
2007 by the Groupe Francophone des Myélodysplasies (GFM) (43).
Recommendation for iron chelation therapy, resulting from a consensus
meeting, have been published (44). Gattermann et al. (44) advocate starting
chelation in patients with relatively favorable prognosis (i.e., low- or int1-risk
MDS, or higher risk MDS when “active” treatment like allogeneic SCT intensive
chemotherapy or hypomethylating agents are envisaged), who have received at
least 20 to 30 RBC concentrates, and/or if their serum ferritin level raises above
1000 to 1500 U/L. Adequate iron chelation is now made easier by the recent avail-
ability of oral iron chelators (especially deferasirox), in addition to the classical
parenteral desferoxamine.
Treatment of Neutropenia and Thrombocytopenia in Lower Risk MDS

Neutropenia
G-CSF and GM-CSF can improve neutropenia in 60% to 75% of MDS, but their
prolonged use appears to have no impact on survival. Therefore, they are generally
used for transient periods, in patients who experience sepsis especially if severe.
One should recommend to neutropenic MDS patients immediate use of broad
spectrum antibiotics in case of fever or other signs of infection.
Thrombocytopenia
Androgens and interleukin 11 (IL-11) have shown to improve thrombocytopenia in
about one-thirds of thrombocytopenic lower risk MDS, but response to androgen is
generally transient, while IL-11 is not widely available and associated to relatively
important side effects (45,46).
Because TPO itself is immunogenic, leading to thrombocytopenia, TPO
receptor agonists including AMG 531 and Eltrombopag have been designed to
treat thrombocytopenias of different origins. AMG 531 has been shown to give
55% platelet responses in a phase II trial in lower risk MDS with thrombocytopenia
(47). ATG and hypomethylating agents appear to give platelet response in 35% to
40% of the cases of lower risk MDS, in addition to erythroid responses (48–52).
Treatment Strategy in Lower Risk MDS

It is summarized in Figure 2.
(a) In non-del(5q) patients, with symptomatic anemia (this generally occurs

when Hb level is below 10 g/dL, with or without transfusion requirement),
ESAs, with or without G-CSF, are generally considered as the first choice in
patients who do not have a high “Nordic score” (i.e., both serum EPO level
⬎500 U/L and RBC transfusion requirement ⬎2 units/mo).
In the absence of response to ESAs after 12 weeks of treatment hypomethylat-

ing agents, thalidomide, lenalidomide, and ATG can be used, alone or possibly
combined to an ESA. Each of those approaches has its advantages (e.g., activity
of ATG and hypomethylating agents on both erythroid and platelet lineages) and
disadvantages (cytopenias or myelosuppression associated to hypomethylating
agents and ATG, specific side effects of thalidomide). Those treatments unfortu-
nately often have to be used successively in the same patient, and the proportion of
lower risk MDS patients who fail all those approaches and in whom regular RBC
cell transfusions are required, remains high. Intensive iron overload prevention is
required in those patients as it appears to be associated with survival improvement
(43).
In case of relapse after response to ESAs, disease progression should first
be investigated with a bone marrow examination. In our experience, folate or iron
deficiency due to increased erythroid production with ESAs are rarely a cause of
secondary ESA failure, but their supplementation may be advocated when patients
appear to start relapsing after response, while EPO antibodies are extremely rare.
(b) In lower risk MDS with del(5q), lenalidomide is the only drug capable of
leading to durable transfusion independance in a large proportion of patients
and it probably should be used as first-line treatment. The attitude in patients
who fail Revlimid or relapse after Revlimid is not clearly determined. Patients
should probably follow the pathway indicated in patients without del(5q).
ESAs, possibly combined to Revlimid, may be the first option to test.
408
Low and int-1 MDS patients

if clinically significant cytopenia
Supportive care
as adjunct to treatment
including G-CSF if severe infection and neutropenia
Thrombocytopenia Anemia Del 5q ±others
Immunosuppressive therapy (IST) in selected patients* If symptomatic or less than 100G/L Lenalidomide
clinical trial of TPD receptor agonists
androgens
hypomethylating agents If failure follow appropriate non–del (5q) pathways
Serum EPD level more than 500U/L Serum EPO level less than 500U/l
and more than 2 RBC units per mo... or less than 2 RBC units per month
Hypomethylating agents immunosuppressive therapy (IST) ESA ± G-CSF

clinical trial** in selected patients* for 8–12 weeks
lenalidomide?
No response No response If failure, follow

clinical trial clinical trial serum EPO level more than 500 U
consider HSC, if donor available consider HSC, if donor available and more than 2 RBC units per month
in selected int1 patients in selected int1 patients pathway
Figure 2 Treatment of low and int1-risk patients.

Fenaux et al.
(c) Isolated neutropenia is rare in lower risk MDS. G-CSF can be indicated in
case of severe infection. Rare patients who suffer from recurrent infections
may benefit from long-term low-dose G-CSF treatment. Rapid administration
of broad spectrum antibiotics at the least sign of infection is probably the
most important measure in those patients.
(d) Isolated thrombocytopenia is also rare in lower risk MDS. One should first
carefully check the diagnosis of MDS, as opposed to chronic ITP. Although
this is not performed in many countries, our experience is that platelet lifespan
studies by isotopic methods (in labeled platelets) can be useful in some
difficult cases (53).
Preliminary results with TPO receptors agonists appear promising, but those drugs
are still not widely available. In their absence and if thrombocytopenia is severe (in
platelets ⬍30,000/mm3 ), our first-line choice remains danazol, which has limited
virilizing side effects. In patients who do not respond or relapse and have severe
thrombocytopenia, ATG or hypomethylating agents could be an option (the former
in patients aged less than about 65 years).
Finally, younger patients who have IPSS int1, anemia that requires a high
transfusion rate, or life threatening thrombocytopenia not responding to any treat-
ment may be candidates for allogeneic SCT.
REFERENCES
1. Greenberg, P, Cox C, LeBeau MM, et al. International scoring system for evaluating
2. Alessandrino EP, Amadori S, Barosi G, et al. Evidence- and consensus-based practice
guidelines for the therapy of primary myelodysplastic syndromes. A statement from
the Italian Society of Hematology. Haematologica 2002; 87:1286–1306.
5. List A, Kurtin S, Roe DJ, et al. Efficacy of lenalidomide in myelodysplastic syn-
dromes. N Engl J Med 2005; 352:549–557.
6. Cheson BD, Bennett JM, Kantarjian H, et al. Report of an International Working
Group to standardize response criteria for myelodysplastic syndromes. Blood 2000;
96:3671–3674.
7. Cheson BD, Greenberg PL, Bennett JM, et al. Clinical application and proposal
for modification of the International Working Group (IWG) response criteria in
myelodysplasia. Blood 2006; 108:419–425.
8. Deeg HJ. Optimization of transplant regimens for patients with myelodysplas-
tic syndrome (MDS). Hematology Am Soc Hematol Edu Program 2005:167–
173.
9. Sierra J, Perez WS, Rozman C, et al. Bone marrow transplantation from HLA-
identical siblings as treatment for myelodysplasia. Blood 2002; 100:1997–2004.
10. Cutler CS, Lee SJ, Greenberg P, et al. A decision analysis of allogeneic bone marrow
transplantation for the myelodysplastic syndromes: Delayed transplantation for low-
410 Fenaux et al.
risk myelodysplasia is associated with improved outcome. Blood 2004; 104:579–

585.
11. Scott BL, Storer B, Loken MR, et al. Pretransplantation induction chemotherapy and
posttransplantation relapse in patients with advanced myelodysplastic syndrome.
Biol Blood Marrow Transplant 2005; 11:65–73.
12. Ho AY, Pagliuca A, Kenyon M, et al. Reduced-intensity allogeneic hematopoietic
stem cell transplantation for myelodysplastic syndrome and acute myeloid leukemia
with multilineage dysplasia using fludarabine, busulphan, and alemtuzumab (FBC)
conditioning. Blood 2004; 104:1616–1623.
13. Wattel E, De Botton S, Luc Lai J, et al. Long-term follow-up of de novo myelodys-
plastic syndromes treated with intensive chemotherapy: Incidence of long-term sur-
vivors and outcome of partial responders. Br J Haematol 1997; 98:983–991.
14. Knipp S, Hildebrand B, Kundgen A, et al. Intensive chemotherapy is not recom-
mended for patients aged ⬎60 years who have myelodysplastic syndromes or acute
myeloid leukemia with high-risk karyotypes. Cancer 2007; 110:345–352.
15. Beran M, Shen Y, Kantarjian H, et al. High-dose chemotherapy in high-risk
myelodysplastic syndrome: Covariate-adjusted comparison of five regimens. Cancer
2001; 92:1999–2015.
16. Ossenkoppele GJ, Graveland WJ, Sonneveld P, et al. The value of fludarabine in
addition to ARA-C and G-CSF in the treatment of patients with high-risk myelodys-
plastic syndromes and AML in elderly patients. Blood 2004; 103:2908–2913.
17. Beran M, Estey E, O’Brien S, et al. Topotecan and cytarabine is an active combina-
tion regimen in myelodysplastic syndromes and chronic myelomonocytic leukemia.
J Clin Oncol 1999; 17:2819–2830.
18. Faderl S, Garcia-Manero G, Gandhi V, et al. Results of an exploratory study of oral
(po) and intravenous (iv) clofarabine in patients with myelodysplastic syndrome.
ASH Annu Meet Abstr 2007; 110:1455.
19. Cheson BD, Simon R. Low-dose ara-C in acute nonlymphocytic leukemia and
myelodysplastic syndromes: A review of 20 years’ experience. Semin Oncol 1987;
14:126–133.
20. Gerhartz HH, Marcus R, Delmer A, et al. A randomized phase II study of low-dose
cytosine arabinoside (LD-AraC) plus granulocyte-macrophage colony-stimulating
factor (rhGM-CSF) in myelodysplastic syndromes (MDS) with a high risk of devel-
oping leukemia. EORTC Leukemia Cooperative Group. Leukemia 1994; 8:16–23.
21. Miller KB, Kim K, Morrison FS, et al. The evaluation of low-dose cytarabine in the
treatment of myelodysplastic syndromes: A phase-III intergroup study. Ann Hematol
1992; 65:162–168.
22. Fenaux, P, Mufti GJ, Santini V, et al. Azacitidine (AZA) treatment prolongs overall
survival (OS) in higher-risk MDS patients compared with conventional care regimens
(CCR): Results of the AZA-001 phase III study. ASH Ann Meet Abstr 2007; 110:817.
23. Ontachi Y, Yamauchi H, Takami A, et al. Low dose melphalan therapy was effective
in an elderly patient with MDS-AML. Nippon Ronen Igakkai Zasshi 2001; 38:405–
408.
24. Anargyrou K, Vaiopoulos G, Terpos E, et al. Low dose melphalan is a treatment
option in elderly patients with high risk myelodysplastic syndrome or secondary
acute myeloblastic leukaemia. Haematologia (Budap) 2002; 32:169–173.
25. Lim Z, Ho AYL, Samuel J, et al. Outcomes of MDS patients with chromosome
7 abnormalities treated with 5-azacytidine. ASH Ann Meet Abstr 2007; 110:
1449.
26. Kantarjian H, Oki Y, Garcia-Manero G, et al. Results of a randomized study of

3 schedules of low-dose decitabine in higher-risk myelodysplastic syndrome and
chronic myelomonocytic leukemia. Blood 2007; 109:52–57.
27. Ruter B, Wijermans, P, Claus R, et al. Preferential cytogenetic response to continuous
intravenous low-dose decitabine (DAC) administration in myelodysplastic syndrome
with monosomy 7. Blood 2007; 110:1080–1082; author reply 1083.
azacitidine in patients with the myelodysplastic syndrome: A study of the cancer
and leukemia group B. J Clin Oncol 2002; 20:2429–2440.
29. Silverman LR, McKenzie DR, Peterson BL, et al. Further analysis of trials with
azacitidine in patients with myelodysplastic syndrome: Studies 8421, 8921, and
9221 by the Cancer and Leukemia Group B. J Clin Oncol 2006; 24:3895–3903.
30. Vey N, Bosly A, Guerci A, et al. Arsenic trioxide in patients with myelodysplastic
syndromes: A phase II multicenter study. J Clin Oncol 2006; 24:2465–2471.
31. Garcia-Manero G, Yang H, Bueso-Ramos C, et al. Phase I study of the his-
tone deacetylase inhibitor vorinostat (suberoylanilide hydroxamic acid, SAHA) in
patients with advanced leukemias and myelodysplastic syndromes. Blood 2008;
11:1060–66.
32. Fenaux, P, Raza A, Mufti GJ, et al. A multicenter phase 2 study of the farnesyltrans-
ferase inhibitor tipifarnib in intermediate to high-risk myelodysplastic syndrome.
Blood 2007; 109:4158–4163.
33. Kantarjian HM, O’Brien S, Huang X, et al. Survival advantage with decitabine ver-
sus intensive chemotherapy in patients with higher risk myelodysplastic syndrome:
Comparison with historical experience. Cancer 2007; 109:1133–1137.
34. Crawford J, Cella D, Cleeland CS, et al. Relationship between changes in hemoglobin
level and quality of life during chemotherapy in anemic cancer patients receiving
epoetin alfa therapy. Cancer 2002; 95:888–895.
35. Park S, Grabar S, Kelaidi C, et al. Predictive factors of response and survival in
myelodysplastic syndrome treated with erythropoietin and G-CSF: The GFM expe-
rience. Blood 2008; 111:574–582.
36. Hellstrom-Lindberg E, Gulbrandsen N, Lindberg G, et al. A validated decision
model for treating the anaemia of myelodysplastic syndromes with erythropoietin
+ granulocyte colony-stimulating factor: Significant effects on quality of life. Br J
Haematol 2003; 120:1037–1046.
37. Raza A, Meyer, P, Dutt D, et al. Thalidomide produces transfusion independence
in long-standing refractory anemias of patients with myelodysplastic syndromes.
Blood 2001; 98:958–965.
38. Strupp C, Germing U, Aivado M, et al. Thalidomide for the treatment of patients
with myelodysplastic syndromes. Leukemia 2002; 16:1–6.
39. Bouscary D, Legros L, Tulliez M, et al. A non-randomised dose-escalating phase
II study of thalidomide for the treatment of patients with low-risk myelodysplastic
syndromes: The Thal-SMD-2000 trial of the Groupe Francais des Myelodysplasies.
Br J Haematol 2005; 131:609–618.
40. Kelaidi C, Park S, Brechignac S, et al. Treatment of myelodysplastic syndromes
with 5q deletion before the lenalidomide era; The GFM experience with EPO and
thalidomide. Leuk Res 2008.
41. Raza A, Reeves JA, Feldman EJ, et al. Phase 2 study of lenalidomide in transfusion-
dependent, low-risk, and intermediate-1 risk myelodysplastic syndromes with kary-
otypes other than deletion 5q. Blood 2008; 111:86–93.
412 Fenaux et al.
42. Saunthararajah Y, Nakamura R, Wesley R, et al. A simple method to predict response

to immunosuppressive therapy in patients with myelodysplastic syndrome. Blood
2003; 102:3025–3027.
43. Rose C, Brechignac S, Vassilief D, et al. Positive impact of iron chelation therapy
(CT) on survival in regularly transfused MDS patients. A prospective analysis by
the GFM. ASH Annu Meet Abstr 2007; 110:249.
44. Gattermann N, Porter J, Lopes L, et al. Consensus statement on iron overload in
myelodysplastic syndromes. Hematol Oncol Clin North Am 2005; 19:18–25.
45. Montero AJ, Estrov Z, Freireich EJ, et al. Phase II study of low-dose interleukin-11
in patients with myelodysplastic syndrome. Leuk Lymphoma 2006; 47:2049–2054.
46. Wattel E, Cambier N, Caulier MT, et al. Androgen therapy in myelodysplastic
syndromes with thrombocytopenia: A report on 20 cases. Br J Haematol 1994;
87:205–208.
47. Kantarjian H, Fenaux, P, Sekeres MA, et al. Phase 1/2 Study of AMG 531 in
thrombocytopenic patients (pts) with low-risk myelodysplastic syndrome (MDS):
Update including extended treatment. ASH Annu Meet Abstr 2007; 110:250.
48. Broliden PA, Dahl IM, Hast R, et al. Antithymocyte globulin and cyclosporine A
as combination therapy for low-risk non-sideroblastic myelodysplastic syndromes.
Haematologica 2006; 91:667–670.
49. Stadler M, Germing U, Kliche KO, et al. A prospective, randomised, phase II
study of horse antithymocyte globulin vs rabbit antithymocyte globulin as immune-
modulating therapy in patients with low-risk myelodysplastic syndromes. Leukemia
2004; 18:460–465.
50. Yazji S, Giles FJ, Tsimberidou AM, et al. Antithymocyte globulin (ATG)-based
therapy in patients with myelodysplastic syndromes. Leukemia 2003; 17:2101–
2106.
51. Passweg JR, Giagounidis A, Simcock M, et al. Immunosuppression for patients
with low and intermediate risk myelodysplastic syndrome: A prospective random-
ized multicenter trial comparing antithymocyte globulin + cyclosporine with best
supportive care: SAKK 33/99. ASH Annu Meet Abstr 2007; 110:1461.
52. Lim ZY, Killick S, Germing U, et al. Low IPSS score and bone marrow hypocellu-
larity in MDS patients predict hematological responses to antithymocyte globulin.
Leukemia 2007; 21:1436–1441.
53. Bourgeois E, Caulier MT, Rose C, et al. Role of splenectomy in the treatment of
myelodysplastic syndromes with peripheral thrombocytopenia: A report on six cases.
Leukemia 2001; 15:950–953.
17
Management of Cytopenias in MDS
Luca Malcovati
Department of Hematology, University of Pavia Medical School and Fondazione
IRCCS Policlinicc San Matteo, Pavia, Italy
David T. Bowen
Department of Hematology, St James’s Institute of Oncology, Leeds, U.K.
Eva Hellström-Lindberg
Karolinska Institutet and Karolinska University Hospital Huddinge,
Stockholm, Sweden
INTRODUCTION
The central clinical characteristic of myelodysplastic syndromes (MDS) is periph-
eral blood cytopenias (1). Hematopoietic progenitors in MDS have impaired ability
to differentiate and increased susceptibility to apoptosis, resulting in ineffective
hematopoiesis.
Anemia is the major clinical problem for the majority of patients with
MDS. If one uses the World Health Organization (WHO) definition of anemia
(i.e., hemoglobin ⬍13 g/dL for men and ⬍12 g/dL for women), more than 90%
of MDS patients are anemic at the time of diagnosis, and moderate to severe
anemia—defined as hemoglobin lower than 10 g/dL— is observed in about
60% of cases (2). Neutropenia, thrombocytopenia, or both may be found on
presentation, or may appear later as their disease progresses. The incidence of
thrombocytopenia (defined as platelets ⬍100 × 109 /L) varies from 20% for the
IPSS lowest risk patients to 82% for patients in the IPSS highest risk category.
Severe thrombocytopenia (i.e., platelets ⬍20 × 109 /L) varies from 6% to 25% in
the low and high IPSS risk categories, respectively (3).
413
414 Malcovati et al.
For many patients with MDS, supportive care is currently the most appro-
priate management strategy (4,5). However, with a better understanding of the
mechanisms of cytopenia in this heterogeneous group of diseases, interventional
therapy will become increasingly available for patient subgroups as determined by
biological criteria (6). Even in patients for whom interventional therapy is chosen,
good supportive care will remain essential.
MECHANISMS OF INEFFECTIVE HEMATOPOIESIS IN MDS

Hematopoietic progenitor cells in MDS are characterized by reduced ability to
differentiate and mature, increased apoptosis, and preferential growth of immature
clonal precursors. The pathogenic mechanisms of MDS and the resultant cellular
abnormalities differ considerably between patient subgroups.
There are two broad pathological processes that impair red cell production
in MDS and are considered to be the major contributors to anemia. The first is
ineffective erythropoiesis, characterized by an increased percentage of marrow
erythroblasts, many of which undergo apoptosis before terminal maturation. The
other is hypoproliferative erythropoiesis, characterized by a decreased relative
number of erythroid progenitors in the bone marrow. These processes are not
mutually exclusive, and the anemia of MDS is often multifactorial.
Ineffective hyperproliferative erythropoiesis dominates in refractory anemia
with ringed sideroblasts (RARS). It is also common in a subset of refractory
anemia (RA) and is sometimes also present in those RA with excess blasts (RAEB)
with a moderate increase of marrow blasts. Hypoproliferative erythropoiesis, in
contrast, is typically seen in more advanced MDS cases, as well as in hypoplastic
MDS, where the overall marrow cellularity is reduced compared to healthy
persons of a similar age. Other settings where hypoproliferative erythropoiesis
dominates include some cases of 5q− syndrome, and in MDS with severe marrow
fibrosis.
Ineffective Hematopoiesis
Studies of the mechanisms of cytopenia in MDS have evolved from the ini-
tial morphological observation of the paradoxically expanded bone marrow with
peripheral blood cytopenias (7). This association can be accounted for by the
proliferation of immature hematopoietic progenitors in an attempt to compensate
for defective cell maturation. Successive erythrokinetic studies proved unequivo-
cally that ineffective erythropoiesis is most prominent in the patients with RA and
RARS in which estimated relative efficiencies of erythropoiesis are around 50%
in RA and only 10% in RARS (8).
The reduced efficiency of erythropoiesis is mainly explained by an increased
rate of intramedullary erythroid cell death, probably due to augmented apoptosis,
as discussed in chapter 5 (9–11). Two main pathways leading to apoptotic cell
death have been identified. The intrinsic pathway involves mitochondria and is
Management of Cytopenias in MDS 415
regulated by proteins of the Bcl-2 family (12). The extrinsic pathway involves cell
surface death receptors, resulting in the activation of the caspase cascade, either
directly or through connection to the intrinsic pathway (13). Deregulation of both
the intrinsic and the extrinsic pathways has been reported in MDS cells (14–17).
Hypoproliferative Hematopoiesis
Hypoproliferative erythropoiesis is the more common erythropoietic defect in
MDS, mainly occurring when blasts exceed 10% of the bone marrow cells,
and is less well understood than ineffective erythropoiesis (8). Morphologically,
hypoproliferative erythropoiesis has a reduced percentage of marrow erythroid
precursors, while nuclear and cytoplasmic morphological abnormalities are less
striking.
In contrast to RA and RARS, erythrokinetic studies indicate that patients
with an excess of blasts or chronic myelomonocytic leukemia mainly have a
hypoproliferative erythroid defect (8). Poor progenitor growth does not allow
any further distinction between the processes of ineffective or hypoproliferative
erythropoiesis.
Hypoproliferative erythropoiesis can be caused by intrinsic failure of ery-
throid differentiation at a very early progenitor level, progression of ineffective
erythropoiesis to “burnout”, dominance of myeloid commitment within the multi-
potential progenitor compartment (often related to blast proliferation), or erythroid
failure due to immunological mechanisms as seen also in aplastic anemia (18–
20). Occasionally, an almost total red cell aplasia can be seen in both early and
advanced MDS (21).
Additional Contributors to Cytopenia in MDS

Peripheral red cell destruction or loss may also contribute to anemia in MDS.
Red cell loss may result from bleeding associated with thrombocytopenia and/or
platelet functional defects. Red cell life span is shortened in some MDS patients,
and this may be due to hemolysis (often with a positive direct antiglobulin test) (22)
or hypersplenism. Many different intrinsic red cell abnormalities are also described
in MDS, although their clinical significance is less clear (23–25). Finally, many
patients with long-standing transfusion therapy show progressively increasing
transfusion needs without other obvious changes in hematopoietic parameters—
the mechanism of this is poorly understood.
THERAPEUTIC APPROACHES TO THE ANEMIA OF MDS

In patients with MDS, anemia may be ameliorated by three different but often
concurrent mechanisms: expansion of effective erythropoiesis at the expense of
the ineffective clone, conversion of ineffective to effective erythropoiesis within
the MDS clone, or elimination of the MDS clone to allow reconstitution of normal
polyclonal hematopoiesis. While chemotherapy and stem cell transplantation can
clearly obtain the elimination of the dysplastic clone, whether either of the first
two routes can be achieved by the use of other therapeutic modalities is unclear.
A considerable problem when reviewing therapeutic studies in MDS is the
extensive variation in clinical response criteria used by different study groups.
It is important to distinguish between responses that are merely “biological”
(often transient and/or incomplete), such as improvement in blood counts, and
those that are meaningful to the patient, such as symptom control and avoidance
of transfusions. This issue of varying response criteria has been addressed by
large cooperative groups active in the field of MDS (26), and more recently by
an international working group specifically committed to achieve more uniform
criteria (27,28). Importantly, this international working group included quality of
life (QOL) as a response parameter.
Anemia in MDS: What Needs Treating?

It is only recently that the impact of chronic anemia upon patient outcomes and
QOL has begun to be studied in MDS (29–32). Anemia is associated with a decline
in physical performance (33,34) and an increased risk of death in community-
dwelling older adults (35), and with a poorer outcome in patients with heart
failure (36). Nonetheless, the contribution of anemia to comorbidity in elderly
MDS patients with chronic anemia and transfusion need has received limited
attention. Recently, low hemoglobin levels have been found to be associated with
cardiac remodeling and an increased incidence of heart failure in MDS (29,37,38).
However, several critical issues remain unsolved.
Hemoglobin concentration is reported to decline in older subjects (39), with
a greater reduction among men than among women and substantial individual
variation (40,41). Consequently, a precise definition of the severity of anemia
may be difficult. Furthermore, no study has systematically addressed the question
of target levels for erythrocyte transfusion to maintain outpatient life without
increasing morbidity.
Treatment, whether this consists of transfusion therapy, recombinant ery-
thropoietin (EPO), or other proerythrogenic therapies, should aim to achieve the
best possible QOL and should be adjusted in relation to close observation of comor-
bid conditions. Adopting symptomatic anemia as a major criterion for estimating
the severity of anemia and for making clinical decisions in the MDS popula-
tion appears to be the most reliable approach. Therefore, any decision about target
hemoglobin levels should be individualized following detailed discussion between
the doctor and the patient.
Red Cell Transfusion Therapy

Once anemia is symptomatic, red cell transfusion is the mainstay of therapy.
Although nearly all MDS patients are destined to receive a red cell transfusion at
some point during their clinical course, according to evidence-based therapeutic
guidelines, RBC transfusion will be offered to approximately 40% of patients
as the sole therapeutic option (4,5). This group includes patients with good-
prognosis MDS, as well as those with poor-prognosis disease whose age, comorbid
conditions, or performance status preclude them from receiving more intensive
forms of therapy.
The onset of a regular transfusion requirement in MDS patients is associated
with a worsening of prognosis (29,42). Transfusion-dependent patients have a
significantly higher risk for both nonleukemic death and leukemic evolution,
compared with those who do not need regular transfusions. The detrimental effect
of transfusion dependency is more noticeable in patients with low-risk MDS than
in those with an excess of bone marrow blasts, and is proportional to the severity
of the transfusion requirement, with an approximate increase in risk of death of
30% for every unit increase per month in the patient’s transfusion requirement.
These data suggest that the negative effect of a regular transfusion require-
ment is at least partly due to a more aggressive disease in terms of both more
severe bone marrow failure and increased risk of leukemic progression. However,
transfused patients have a significantly higher risk of cardiac death (29,38). This
might be explained by the fact that transfusion-dependent patients have on average
lower hemoglobin levels over time than those who do not require regular transfu-
sions. Furthermore, indirect evidence suggests that secondary iron overload may
also play a role in the poorer outcomes associated with chronic transfusion.
Taken together, these data suggest that preventing anemia-related morbidity
through adequate transfusion regimens is very important in order to minimize the
impact on the patient’s life expectancy. Red cell transfusion should be considered
in any patient with symptomatic anemia. As discussed above, the decision to
start a regular transfusion regimen should be based on a clinical evaluation of
the patient’s anemia-related symptoms and comorbid conditions rather than on a
single hemoglobin level (4,5). The frequency of red cell transfusion is variable,
from as often as once every 1 to 2 weeks to less frequently (i.e., once every 6–12
weeks). For patients with very short transfusion intervals (more than once every
2 weeks), bleeding and hemolysis should be taken into consideration, but most
frequently these high-transfusion requirements reflect profound erythroid failure
(severe reticulocytopenia) with or without peripheral consumptive processes such
as hypersplenism.
The main goals of RBC transfusion therapy are to preserve QOL and physical
function. Given this, individual patient needs and lifestyle must be taken into
account when defining a transfusion program, and objective assessments of QOL
in clinical practice are important in order to monitor the long-term efficacy of
these programs. For example, the Nordic MDS Study Group is now investigating
the effect of intensive transfusion regimens that aim to reach a target hemoglobin
level of 12 g/dL on cardiac function, QOL, and health care costs in MDS patients.
Although the concepts of transfusion dependence and independence in MDS
may appear intuitive, the definition of clinically relevant transfusion dependence is
anything but simple. In two recent studies, “transfusion dependency” was defined
as requiring a transfusion of at least 2 units of red cells within 8 weeks (6), and as
having at least 2 RBC transfusion episodes every 16 weeks (43). The definition of
transfusion dependency has an important impact not only in the clinical manage-
ment of patients but also in clinical trials evaluating treatment approaches aimed
at reducing transfusion requirement or gaining transfusion independence. Fluctu-
ations in transfusion needs during the study period might be erroneously judged
as a response to therapy (false positive response), particularly in uncontrolled
single-arm studies. Varying transfusion triggers can also confound assessment of
response rates. Although somewhat arbitrary, RBC transfusions administered for a
hemoglobin value of 9 g/dL or less would be an appropriate criterion for entry onto
clinical trials and could reasonably serve as the baseline for response evaluation
in terms of defining RBC transfusion independence (28).
Despite the significant progress which has been made, the risks associated
with red cell transfusion have not been completely eliminated, and many risks
are still unknown. Although chronic infections associated with red cell trans-
fusion appear to be of little relevance to the majority of MDS patients whose
life expectancy is ⬍6 years, there is a group of long-term transfused patients
for whom this can be a major issue. Red cell alloimmunization is common and
increases along with the numbers of units transfused. Despite this, reactions to red
cell transfusion are just as frequent in patients without red cell alloantibodies (44).
The practicalities and expense of obtaining compatible blood for alloimmunized
patients are significant. In countries practicing universal leukodepletion of red
cell products, the risks of red cell alloimmunization have now been reduced, but
there are still many rare yet potentially fatal complications of transfusion, such as
posttransfusion purpura and transfusion-associated graft-versus-host disease.
Iron Overload and Iron Chelation Therapy

Each unit of red cells contains approximately 200–250 mg of elemental iron,
and there is no efficient physiological mechanism for excretion of excess iron.
Therefeore, patients with MDS may develop iron overload as a result of repeated
red cell transfusions. Additionally, in some patients with sideroblastic anemia,
there may be excessive absorption of food iron by the gut. Iron overload due to
increased intestinal absorption is usually mild and is not generally associated with
clinical signs of organ damage. It may, however, worsen secondary overload due
to blood transfusions.
There is little direct evidence at present for the role of iron in organ damage
and the impact of iron overload on the outcome of patients with MDS. Recommen-
dations for iron chelation therapy in MDS have therefore mainly been based on a
comparison with the effects of iron overload and its treatment in other transfusion-
dependent conditions. In patients with thalassemia major who are inadequately
iron chelated, most clinical manifestations of iron overloading do not appear until
the second decade of life. However, after about 1 year of transfusions, iron begins
to be deposited in parenchymal tissues where it may cause significant toxicity
(45).
In an autopsy study on subjects with chronic acquired anemia requiring

red cell transfusions, cardiac iron deposits were found in patients receiving more
than 100 red cell units (46). Several cohort series of multitransfused patients with
MDS or aplastic anemia, document biochemical organ toxicity in association with
iron overload. Death from cardiac failure is reported in approximately one quarter
of patients with MDS, although the contribution of iron to this is unclear (47–
49). Clinical endocrine dysfunctions including glucose intolerance, diabetes, and
pituitary dysfunction have also been reported and could be related to iron-induced
injury of the anterior pituitary or pancreatic beta cells, as is well described in
hereditary hemochromatosis (47–50).
In a retrospective study of prognostic variables in MDS, the development of
secondary iron overload, assessed by means of the serum ferritin level, was asso-
ciated with an adverse effect on survival of transfusion-dependent MDS patients
(29,42). The association of adverse prognosis with iron overload was mainly
noticeable among patients with lower risk forms of MDS, that is, refractory ane-
mia (i.e. RA and RARA) and MDS with isolated del(5q), according to the WHO
classification (51), who have a median survival of over 100 months and are, there-
fore, more prone to developing the toxic effects of iron overload. There was also
a quantitative association between serum ferritin levels and survival, with higher
values associated with a worse prognosis (29,42).
Furthermore, elevated pre-transplant serum ferritin levels were shown to
adversely affect the outcome of MDS patients undergoing allogeneic hematopoi-
etic stem cell transplantation by increasing treatment-related mortality (52). This
implies that although iron overload may not by itself often produce overt clinical
organ failure, it may increase the susceptibility of target organs such as liver or
heart to further insult.
Who Should be Treated with Iron Chelation?
Although the topic of iron chelation in MDS in general remains controversial,
according to the available evidence-based guidelines, the patients recognized as
the best candidates to receive iron chelation therapy are those with sideroblastic
anemia, 5q− syndrome, or other forms of refractory anemia, in whom long-term
transfusion therapy and long survival are likely (4,5). Patients who are candi-
dates for allogeneic stem cell transplantation may also benefit from chelation
therapy, since a high serum ferritin pre-transplant is associated with increased
transplantation-related complications (52).
Iron chelation or phlebotomy (for patients in remission prior to transplant)
should be offered in particular to those who are destined to follow delayed
transplantation strategies. Indeed, the results of a clinical decision analysis from
the International Bone Marrow Transplant Registry demonstrated that the life
expectancy of MDS patients with low and intermediate-1 IPSS, who have a HLA-
identical sibling, was longer when transplantation was delayed by some period, but
performed prior to the development of acute leukemia (53). In fact, although there
is considerable evidence to suggest that the earlier the transplantation is performed
the better the outcome (54), for many patients with low-risk MDS who may expe-
rience a long period without signs of disease progression, the risks of immediate
morbidity and mortality associated with transplantation are unacceptably high.
However, implementing such a strategy in clinical practice requires careful clin-
ical follow-up. Optimizing supportive therapy and avoiding the onset of anemia-
or transfusion-related comorbidities are essential in order not to preclude these
patients from transplantation.
When to Consider Initiating Iron Chelation Therapy?
Except for patients with RARS, iron overload in MDS is uncommon at the time of
diagnosis. Iron chelation may be considered by the time a patient has received 5 g of
excess infused iron (approximately 25 units of red cells) and when continued long-
term transfusion therapy is likely (5). The reference method for determining body
iron stores is the measurement of hepatic iron concentration (55). However, this has
traditionally been assessed through a percutaneous liver biopsy, which (in contrast
to patients with hereditary hemochromatosis or hemoglobinopathies) is difficult
to implement in the clinical management of MDS patients due to concomitant
neutropenia or thrombocytopenia. Sequential measurements of serum ferritin level
have been shown to be a useful surrogate marker for monitoring of secondary iron
overload in thalassemic patients (56,57), and a correlation between serum ferritin
levels and transfusion burden has been observed in MDS (29). Magnetic resonance
imaging (MRI) scanning of the liver may also provide a noninvasive assessment
of iron stores and is increasingly used in clinical practice.
Treatment Options for Transfusion-Related Iron Overload
Deferoxamine, administered as regular subcutaneous infusions, can reduce serum
ferritin and liver iron concentration in MDS patients (58). Observational studies
have also suggested that deferoxamine therapy is occasionally associated with
improved marrow function and reduced transfusion requirements (59,60). How-
ever, the magnitude of such a phenomenon and its biological bases remain unclear.
The recommended iron chelating therapy has been based on deferoxamine
administered at a dose of 20 to 40 mg/kg/day subcutaneously over 12 hours (4,5).
However, this schedule of administration is rather uncomfortable and frequently
results in poor patient compliance. Different schedules have been investigated,
and a continuous infusion through an indwelling intravenous delivery device or
twice-daily subcutaneous bolus injections may be considered where prolonged
subcutaneous infusions are not tolerated (61). One schedule that is clearly not
effective, but continues to be used nevertheless by some physicians, is infusion of
deferoxamine solely as a posttransfusion bolus whenever a unit of red cells has
been administered.
Few observational studies have been reported on the effect of an orally active
iron chelator, deferiprone, on iron excretion in MDS patients (62). Although this
drug promotes increased urinary iron excretion, questions remain as to its safety
(agranulocytosis and possibly hepatotoxicity are concerns) and efficacy (63). The
available scientific evidence cannot support a recommendation for its routine use
in the setting of MDS, for which it remains unlicensed (5).
Recently, deferasirox has been proven to be an effective oral iron chelator
in patients with thalassemia major, and studies are ongoing in MDS (64,65). This
drug is approved in the United States for the treatment of chronic iron overload
due to multiple blood transfusions in patients aged ≥2 years; in Europe, it is
approved for the treatment of chronic iron overload due to blood transfusions in
patients with thalassemia, and in patients with other anemias when deferoxamine
is contraindicated or inadequate. Controlled studies of deferasirox in MDS have
not been yet published.
How to Monitor Iron Chelation Therapy?

Monitoring iron chelation therapy is essential in order to avoid both under- or
overtreatment, and to evaluate patient adherence to the treatment regimen. Patients
undergoing iron chelation should receive periodic monitoring of serum ferritin
levels. Based on the evidence from thalssemia, a concentration of ⬍1000 ng/mL
has been recommended as a target for iron chelation therapy (5). However, more
studies are warranted to identify appropriate goals for iron chelation in MDS
patients.
Magnetic susceptometry using a superconducting quantum interference
device (SQUID) magnetometer provides a direct measure of hepatic storage iron.
However, magnetic susceptometry is available in only a few centers, and the mea-
surements of hepatic iron concentration do not appear to be equivalent among
different sites.
In contrast, as mentioned above MRI is emerging as a more promising
noninvasive tool for measuring tissue iron content. Liver iron content measured
using MRI is directly related to liver iron content in biopsy samples (66). In
addition, cardiovascular MRI could be used to determine myocardial iron content
and its consequences on cardiac function (67). Recently, two small studies failed
to demonstrate cardiac iron accumulation in heavily transfused MDS patients,
but only a minority of transfused patients in both studies were not receiving iron
chelation therapy (68,69). Therefore, for the moment, the risk of cardiac iron
deposits in transfusion-dependent patients who are not receiving iron chelation
therapy cannot be excluded.
Hematopoietic Growth Factors

Erythropoietin in MDS
The therapeutic efficacy of recombinant EPO in the treatment of anemia is now
well established for selected patients with MDS. Cohort studies have clearly
demonstrated responses to EPO, and two randomized placebo-controlled prospec-
tive studies have confirmed the superior response rate of EPO over placebo
(70,71).
Two meta-analyses have been published covering trials of EPO used as a

single agent in MDS. One included 205 patients from 17 trials published in 1994
(72), the other 115 patients from 10 trials (73). The overall response rates, adopting
a 100% reduction of transfusion need as minimal response criteria (or, for patients
who were not transfusion dependent, a sustained rise in hemoglobin of 1.5 g/dL or
more), were 16% and 23.5% in the two analyses, respectively. Positive predictors
of response in the larger analysis were FAB subtype without ringed sideroblasts,
pretreatment serum EPO levels of less than 200 U/L, and the lack of transfusion
need. Patients with RARS responded less well to EPO therapy alone, with an
overall response rate of only 8% (72). The smaller study identified only RAEB
FAB subtype as a negative predictor.
Two Phase 3 randomized controlled trials have evaluated the use of EPO
versus placebo (70,71). The first randomized trial enrolled 20 patients with RA
or RARS and treated them with a weekly dose of EPO of between 1600 and
3200 U/kg intravenously. A response occurred in 12.5% of the evaluable patients
(70). In the second trial, including 87 patients with hemoglobin levels of ⬍9 g/dL
and bone marrow blasts of less than 10%, an overall benefit for EPO over placebo
(p = 0.007) was shown. However, analysis of patient subgroups demonstrated a
significant effect of treatment only in nontransfused patients and in patients with
RA. Again, basal serum EPO levels of less than 200 U/L predicted for response
(71). Overall, these data suggest that patients having RARS and a transfusion need
will respond poorly to EPO as monotherapy, whereas the serum EPO level helps
predict response in other groups.
The optimal dose and schedule of EPO therapy in MDS is unclear. Doses in
the most uncontrolled studies have ranged from 30,000 to 60,000 U/wk, adminis-
tered in single or multiple subcutaneous injections (74–76). The number of doses
per week could probably be reduced but studies exploring this have still not been
undertaken in MDS.
There is no consistent information as to the optimal time to assess response
to treatment. While existing data indicate that treatment should be stopped at 10
to 12 weeks in the absence of response, a few additional responses have been
observed when treatment is prolonged for as long as 36 weeks (77).
Only limited data are available confirming the durability of responses to
therapy with EPO alone. One small study suggested that one-third of responders
were able to either sustain the initial response or achieve a new response at
reintroduction of EPO following recurrence of their anemia while off therapy.
Responses at reintroduction could also be maintained on lower doses of EPO
than those used at induction. Patients who lost their response did so largely for
predictable reasons, including MDS transformation to acute myeloid leukemia
and the development of a new malignancy (78).
EPO therapy is generally well tolerated, with the most common side effects
being flu-like symptoms and occasional splenic pain and enlargement. Between
1998 and 2004, severe pure red cell aplasia developed in some patients treated
with EPO in Europe. This complication was apparently caused by a defective
formulation of EPO syringes and is now rarely observed (79). New concerns
about EPO contributing to poorer outcomes or disease progression in patients
with solid tumors are of uncertain applicability in the MDS setting.
EPO Combined with G-CSF
The synergistic effect of the combination of G-CSF and EPO in the treatment of
anemia has now been demonstrated in vivo in several Phase 2 studies (26,80–83)
and in two randomized Phase 3 trials (30,84). A third trial of EPO with or without
G-CSF, the Eastern Cooperative Oncology Group E1996 trial, was presented in
abstract form in 2004 but has not been yet published (85).
The first randomized study compared treatment with EPO and G-CSF
versus supportive care alone; patients who were considered responders after
12 weeks of combined treatment were given EPO alone for 40 additional weeks.
The study showed that the combined growth factor treatment led to responses in
about 40% of MDS patients. In addition, in patients who experienced a relapse in
the maintenance phase with EPO alone, responses were restored by reintroducing
G-CSF (30). A more recent randomized trial comparing the effect of EPO versus
EPO and G-CSF given for a minimum of 8 weeks showed a significantly higher
erythroid response rate in patients receiving the combination of growth factors
(73.3% vs. 40%) (84). The addition of G-CSF also induced erythroid response in
about half the patients who were unresponsive to EPO alone. Overall, response
rates to the combination therapy were higher than the EPO alone, and the
improvement in response rate induced by the combination of growth factors was
the most pronounced in patients with RARS. Furthermore, the addition of G-CSF
produced responses in a proportion of patients who had not responded to EPO
alone.
A recent analysis from the French MDS group on a large series of MDS
patients failed to confirm these results, and found similar response rate in patients
treated with EPO alone, and those treated with EPO along with G-CSF. This was
particularly the case for RARS and RCMD-RS (62% response with EPO alone
and 64% with EPO and G-CSF). However, this was a retrospective analysis and
patient selection bias in the choice of single agent or combined therapy cannot be
excluded (86).
A Phase 2 study examined the potential benefit of a prolonged combination
treatment with EPO and G-CSF (at least 36 weeks) in MDS patients. The authors
reported an increased response rate with prolonged treatment, 80% of patients
showing a response at 36 weeks compared with a 60% response rate at 12 weeks
(77).
In 2005, the Nordic MDS study group reported the long-term follow-up of
patients treated with EPO and G-CSF in clinical trials carried out in the last decade.
Their results suggest that the median response duration for patients treated with the
combination is 24 months—29 months in patients achieving a normal hemoglobin
level and 12 months in those with a partial response. Some patients maintained
a response for up to 10 years (87). No increase in the risk of leukemic evolution
was noticed when the long-term outcome of these patients was compared with
that of untreated individuals from the cohort of patients used by the International
Myelodysplasia Risk Assessment Workshop (IMRAW) to generate the IPSS (see
chap. 1).
More recently, patients included in Nordic MDS Group studies were com-
pared with a large, untreated cohort of patients from Pavia, Italy. This analysis
showed that patients with no or a mild transfusion requirement treated with EPO
and G-CSF had a significant survival benefit over those that were not treated.
This suggests that the treatment of anemia with hematopoietic growth factors in
selected MDS patients has a positive impact on outcome (88).
These results have been recently confirmed by a retrospective multicentric
study from the French MDS Group comparing the outcome of patients treated
with EPO or darbepoetin ± G-CSF with that of the IPSS/IMRAW cohort (86).
Although the analysis was not adjusted for all clinical variables known to affect the
response to growth factors, such as endogenous EPO level, multilineage dysplasia
and severity of transfusion requirement, EPO treatment was found to be associated
with better patient survival.
As mentioned above, recent reports on the use of EPO in patients with solid
tumor have raised some concerns regarding the potential adverse effects of EPO
treatment on survival and risk of disease relapse (89–92). An updated clinical
guideline on the use of epoetin and darbepoetin in patients with cancer was pub-
lished under the auspices of the American Society of Clinical Oncology/American
Society of Hematology in the late 2007 (93). After a systematic review of the exist-
ing literature, the expert committee concluded that treatment with erythroid growth
factors can still be recommended for patients with low-risk MDS, while it is not
indicated for other cancer patients with disease-related anemia in the absence of
concomitant chemotherapy (93).
Predictive Factors for Response-to-Growth Factors and Parameters

of Early Erythroid Response
The response rate of unselected MDS patients to treatment with growth factors
is relatively low (approximately 15%). Therefore, the identification of subsets
of patients most likely to respond to such treatment is mandatory for clinical
decision-making, and has profound implications for healthcare costs.
Using pretreatment serum EPO (⬍100, 100–500, or ⬎500 U/L) and RBC
transfusion requirement (⬍2 or ≥2 U/mo), a predictive model for response was
developed from the data of patients treated in two multicenter studies (Table 1)
(26). Three groups of responders were identified with predicted response rates of
74%, 23%, and 7% (high-, intermediate-, and low-response groups, respectively).
This decision model has been widely confirmed, and represents a basis from which
to select optimal candidates for treatment with growth factors (31,82).
In addition, patients with isolated erythroid bone marrow dysplasia
according to WHO classification have been found to have a significantly higher
Table 1 A Predictive Algorithm for EPO+G-CSF Response in Patients with MDS

Total score IWG 2000 erythroid response (patients)
Good: ⬎+1 74% (n = 34)

Intermediate: −1 to +1 23% (n = 31)
Poor: ⬍−1 7% (n = 39)
Serum EPO (U/L) Transfusions
⬍100 = +2 pts ⬍2 Units/mo = +2 pts

100–500 = +1 pt ≥2 Units/mo = −2 pts
⬎500 = −3 pts
Response was defined as >1.5 g/dL hemoglobin increment in the absence of transfusions.
Abbreviations: EPO, erythropoietin; IWG, International Working Group; MDS, myelodysplastic
syndromes; G-CSF, granulocyte colony-stimulating factor.
Source: From Ref. 31.
probability of response to treatment with the combination of EPO and G-CSF

than those with refractory cytopenia with multilineage dysplasia (94). In a large
retrospective analysis from the French MDS Group, no impact of unilineage ver-
sus multilineage dysplasia was found on response rate. However, the presence of
multilineage dysplasia was significantly associated with shorter response duration
(86).
Besides a clear identification of parameters predictive of high rates of
response-to-growth factor therapy, early identification of responders and non-
responders is highly desirable, in order to avoid unnecessary patient inconvenience,
toxicity, and cost. An algorithm using baseline serum EPO levels and incremental
increases in serum transferrin receptor (TfR), and hemoglobin concentration was
proposed by an Italian group in 1996, in order to identify early response to EPO
therapy in hematological malignancies (95). For MDS, an increase in serum TfR
of ⬍18% at 2 weeks predicted nonresponse to EPO therapy in a randomized study
(71). More recently, a small, Phase 2 randomized study monitoring erythropoi-
etic response over 7 days in MDS patients receiving a single subcutaneous bolus
of EPO with or without G-CSF showed that the reticulocyte response at day 7
was highly predictive of subsequent response to therapy (96). In fact, reticulocyte
increment at day 7 represents the early product of more effective erythroid output,
and is not merely a shift of reticulocytes into the circulation, as can be observed
at earlier time points in EPO-treated MDS patients (97).
Mechanism of Action of EPO Alone or in Combination with G-CSF

The biological mechanisms underlying the efficacy of treatment with hematopoi-
etic growth factors in MDS have not been completely clarified. Endogenous serum
EPO levels are usually normal or elevated in MDS patients. Nevertheless, a propor-
tion of severely anemic patients show inappropriately low levels of serum EPO,
although the mechanisms underlying the insufficiency of the RBC homeostatic

loop are still unclear (98).
A response to EPO and G-CSF combination is associated with a decrease
in bone marrow apoptosis, more effective bone marrow erythropoiesis, and, in
RARS, a reduced number of ringed sideroblasts (10). EPO prevents erythroid
progenitors from undergoing apoptosis and promotes erythroid differentiation by
initiating intracellular events (99). These events include activation of the STAT5
transcription factor and expression of antiapoptotic genes, and genes associated
with erythroid differentiation (e.g., GATA-1) (100,101). Signaling via the G-CSF
receptor induces erythroid differentiation in vitro, even in the absence of the EPO
receptor or STAT5 activity, and this may explain some of the synergy between
G-CSF and EPO therapy in MDS patients in whom STAT5 activation is often
compromised (102).
Data from an in vitro model of erythropoiesis have shown that dysplastic
erythroid progenitor cells spontaneously release cytochrome c from mitochon-
dria with subsequent activation of caspase-9. The increased sensitivity of MDS
progenitor cells to death receptor stimulation is due to constitutive activation of
the mitochondrial axis of the apoptotic signaling pathway in these cells (16,17).
G-CSF significantly inhibits cytochrome c release from mitochondria by mod-
ulating proapoptotic Bax, a protein that promotes the release of cytochrome c
from the intermembrane space of mitochondria, and altering the expression of
of anti-apoptotic members of the Bcl-2 family. The in vitro anti-apoptotic effect
of G-CSF has been shown to be more pronounced in patients with sideroblastic
anemia, a subset of patients in whom the improvement in response rate induced
by combining growth factors was more striking (83).
Many studies have investigated whether the hematological response to EPO
in MDS patients might be sustained by the differentiation of the dysplastic clone
or by the stimulation of residual normal hematopoiesis (103–105). One study
demonstrated that erythroid response in patients responding to EPO and G-CSF
is associated with an increased proportion of cytogenetically normal CD34+
cells, which may represent residual normal hematopoietic cells (104). However,
another study assessing functional iron deficiency in MDS patients treated with
hematopoietic growth factors, showed an increased proportion of hypochromic
erythrocytes in RARS patients responding to EPO treatment, indicating that growth
factors also promote survival and maturation of dysplastic erythroblasts (105). It
is likely that different mechanisms are active in different types of MDS.
New Erythropoietic-Stimulating Agents
In the last few years, new erythropoietic-stimulating agents have been investigated
in patients with MDS and other hypoproliferative anemias. The efforts in this
area have been mainly addressed to develope agents with better pharmacokinetic
properties and to define more effective treatment schedules.
Darbepoetin alfa is a molecule derived from the EPO polypeptide by adding
N-linked glycosylated sites, which results in a 3-fold increase in plasma half-life
and slightly diminished receptor binding. Darbepoetin was approved in the United
States for chemotherapy-associated anemia in 2002, and has been assessed in Phase
2 clinical trials in low-risk MDS patients and found to be at least as effective as
EPO (85,106–109).
Just as with EPO, a low endogenous serum EPO level and no or low
RBC transfusion requirement were confirmed as favorable predicting factors
for response to treatment with darbepoetin. Larger randomized trials and longer
follow-ups are needed to corroborate the promising preliminary results of these
studies.
More recently, a new agent, continuous erythropoietin receptor activator
(CERA), has been developed, with an even more prolonged half-life allowing
subcutaneous administration every 3 weeks. This agent has not yet been approved
by regulatory agencies for general use. Preliminary results from Phase 2 studies
in anemic patients with hematological malignancies suggest that CERA is well
tolerated and might be active in the treatment of anemia (110,111). Further trials
including patients with MDS are warranted to define the possible role of CERA
in the treatment of anemia in MDS.
MANAGEMENT OF NEUTROPENIA AND THROMBOCYTOPENIA

IN MDS
Neutropenia
Severe neutropenia in MDS is usually paralleled by pancytopenia, but may occa-
sionally also be an isolated clinical problem. A retrospective multicentric study
from the French MDS Group showed that 22 out of 1009 (2%) MDS patients newly
diagnosed over a 5-year period presented with neutropenia (absolute neutrophil
count ⬍1.5 × 109 /L) as the only cytopenia (112).
The use of myeloid growth factors in the treatment of neutropenia in MDS
was investigated in a randomized controlled trial from the European Organization
for the Research and Treatment of Cancer (EORTC). The study compared the
outcomes of 82 MDS patients using two different daily doses of GM-CSF (either
108 or 216 mg daily for 8 weeks), and showed that the drug increased circulating
neutrophil counts in a high proportion of patients. However, these results did not
translate into increased patient survival rates (113). On the basis of the existing
evidence, there is no indication for prophylactic use of G-CSF or GM-CSF in
MDS. However, the use of these agents may be indicated in individual patients
with severe neutropenia and recurrent infections (4,114).
The use of pegfilgrastim, a long-acting form of G-CSF, has been associated
with leukemoid reactions and spontaneous splenic rupture in MDS (115). In
addition, it is unclear whether the pharmacokinetics of this agent allows the same
synergestic effect with EPO as has been observed with G-CSF. This agent offers
increased convenience compared with multiple doses of G-CSF each week, but
should be used only with caution in patients with MDS.
Thrombocytopenia
Severe thrombocytopenia with chronic bleeding can be a serious clinical problem
in MDS. Bleeding in patients with platelet counts not normally associated with
hemorrhage is usually attributed with a platelet functional defect. While a variety
of defects of platelet aggregation are described, the most prevalent is defective
epinephrine-induced aggregation.
Severe thrombocytopenia in MDS is usually associated with pancytopenia,
but according to the above-cited report from the French MDS Group, about 2%
of newly diagnosed MDS patients may present with isolated thrombocytopenia
(platelet count ⬍100 × 109 /L) (112). Between 10% and 30% deaths in MDS
patients include hemorrhage as a component of the cause of death (3,116).
Platelet transfusions may relieve bleeding tendency on a short-term basis,
but they do not prevent symptoms in the long term and are usually avoided as pro-
phylactic therapy in patients with chronic thrombocytopenia, reserving platelet
transfusion for episodes of hemorrhage or during active treatment (4). Alloim-
munization is problematic. Danazol may be effective with short-term increases in
platelet counts in up to 72% of patients (117,118), while antifibrinolytic therapy
(e.g., with epsilon amino–caproic acid) may help mucosal bleeding.
Megakaryocytopoiesis has proven difficult to stimulate in vivo. The
development of thrombopoietic agents was delayed due to immunogenicity of
early compounds tested in the late 1990s [i.e., recombinant thrombopoietin and
megakaryocyte growth and differentiation factor (MGDF)]. Recently, investiga-
tors have begun exploring a number of second-generation thrombopoietin agonists
that are less immunogenic. One of these, romiplostim (formerly AMG 531), is a
peptibody which binds and activates the thrombopoietin receptor and has been
shown to be effective in refractory idiopathic thrombocytopenic purpura. This
agent is currently being investigated in patients with MDS and severe thrombocy-
topenia (119).
Preliminary data from a Phase 1/2 study in MDS patients suggest that
romiplostim appears to be well tolerated, with headache as the most common
adverse event, and the agent resulted in increased and sustained platelet counts
in 18 out of 44 patients (41%). The mean duration of the platelet response was
23 weeks. These preliminary results suggest that AMG 531 might have a role
in the treatment of low-risk MDS patients who are thrombocytopenic or have a
history of bleeding and support the need for further investigations in this patient
population (120). However, potential safety issues from thrombopoietin agonists
include excessive platelet response, promotion of marrow fibrosis (thrombopoi-
etin overexpression in murine models leads to a disorder resembling idiopathic
myelofibrosis) or acceleration of disease progression to leukemia. Several patients
treated with romiplostim in the Phase 1/2 study in MDS, developed circulating
blasts that disappeared after the drug was stopped, and some leukemic cells do
express functional thrombopoietin receptors, indicating that this latter concern
may be valid.
A number of other megakaryopoietic agents are in clinical trials, including

the thrombopoietin receptor agonists eltrombopag and AKR-501. Data are not yet
available on these agents in MDS.
CLOSING COMMENTS
Anemia is one of the most important factors affecting the prognosis of MDS
patients, especially those with a low risk of disease progression who have a
relatively long-life expectancy. There is emerging evidence to suggest that the
effective treatment of anemia in these subjects with EPO may result in a survival
benefit. Once symptomatic anemia occurs, optimal management is mandatory
in order to avoid a worsening of outcomes due to anemia-related morbidity.
In addition, neutropenia and thrombocytopenia remain difficult management
issues.
In order to improve treatment for patients with MDS, better tools are needed
to identify the pathogenetic mechanisms underlying cytopenias in individual
patients. Cytopenias in MDS may have several different causes, each of which
may show varying responses to different treatment options. Before any treatment
for MDS is planned, it is important to estimate the overall prognosis, and it is key
to evaluate patients’ comorbidities and performance status.
New biological information may help develop new treatment modalities.
In the meantime, carefully designed clinical trials incorporating standardized
response criteria and QOL should allow statistically sound predictive models
to be developed that will, in turn, improve the selection of patients for existing
therapeutic options.
REFERENCES
3. Kantarjian H, Giles F, List A, et al. The incidence and impact of thrombocytopenia
in myelodysplastic syndromes. Cancer 2007; 109:1705–1714.
4. Alessandrino EP, Amadori S, Barosi G, et al. Evidence- and consensus-based prac-
tice guidelines for the therapy of primary myelodysplastic syndromes. A state-
ment from the Italian Society of Hematology. Haematologica 2002; 87:1286–
1306.
8. Cazzola M, Barosi G, Berzuini C, et al. Quantitative evaluation of erythropoietic

activity in dysmyelopoietic syndromes. Br J Haematol 1982; 50:55–62.
9. Raza A, Gezer S, Mundle S, et al. Apoptosis in bone marrow biopsy samples involv-
ing stromal and hematopoietic cells in 50 patients with myelodysplastic syndromes.
Blood 1995; 86:268–276.
10. Hellstrom-Lindberg E, Kanter-Lewensohn L, Ost A. Morphological changes and
apoptosis in bone marrow from patients with myelodysplastic syndromes treated
with granulocyte-CSF and erythropoietin. Leuk Res 1997; 21:415–425.
11. Lepelley P, Campergue L, Grardel N, et al. Is apoptosis a massive process in
myelodysplastic syndromes? Br J Haematol 1996; 95:368–371.
12. Vander Heiden MG, Thompson CB. Bcl-2 proteins: Regulators of apoptosis or of
mitochondrial homeostasis? Nat Cell Biol 1999; 1:E209–E216.
13. Wajant H. The Fas signaling pathway: More than a paradigm. Science 2002;
296:1635–1636.
14. Claessens YE, Bouscary D, Dupont JM, et al. In vitro proliferation and differentiation
of erythroid progenitors from patients with myelodysplastic syndromes: Evidence
for Fas-dependent apoptosis. Blood 2002; 99:1594–1601.
15. Claessens YE, Park S, Dubart-Kupperschmitt A, et al. Rescue of early-stage
myelodysplastic syndrome-deriving erythroid precursors by the ectopic expression
of a dominant-negative form of FADD. Blood 2005; 105:4035–4042.
16. Tehranchi R, Fadeel B, Forsblom AM, et al. Granulocyte colony-stimulating factor
inhibits spontaneous cytochrome c release and mitochondria-dependent apoptosis
of myelodysplastic syndrome hematopoietic progenitors. Blood 2003; 101:1080–
1086.
17. Tehranchi R, Invernizzi R, Grandien A, et al. Aberrant mitochondrial iron dis-
tribution and maturation arrest characterize early erythroid precursors in low-risk
Hematol 2000; 37:15–29.
19. Molldrem JJ, Caples M, Mavroudis D, et al. Antithymocyte globulin for patients
with myelodysplastic syndrome. Br J Haematol 1997; 99:699–705.
sented in myelodysplastic syndrome and aplastic anemia and predicts a response to
immunosuppression in myelodysplastic syndrome. Blood 2002; 100:1570–1574.
21. Williamson PJ, Oscier DG, Bell AJ, et al. Red cell aplasia in myelodysplastic
syndrome. J Clin Pathol 1991; 44:431–432.
22. Sokol RJ, Hewitt S, Booker DJ. Erythrocyte autoantibodies, autoimmune haemoly-
sis, and myelodysplastic syndromes. J Clin Pathol 1989; 42:1088–1091.
23. Higgs DR, Wood WG, Barton C, et al. Clinical features and molecular analysis of
acquired hemoglobin H disease. Am J Med 1983; 75:181–191.
24. Steensma DP, Gibbons RJ, Higgs DR. Acquired alpha-thalassemia in association
with myelodysplastic syndrome and other hematologic malignancies. Blood 2005;
105:443–452.
25. Steensma DP, Viprakasit V, Hendrick A, et al. Deletion of the alpha-globin gene
cluster as a cause of acquired alpha-thalassemia in myelodysplastic syndrome. Blood
2004; 103:1518–1520.
26. Hellstrom-Lindberg E, Negrin R, Stein R, et al. Erythroid response to treatment

with G-CSF plus erythropoietin for the anaemia of patients with myelodysplastic
syndromes: Proposal for a predictive model. Br J Haematol 1997; 99:344–351.
27. Cheson BD, Bennett JM, Kantarjian H, et al. Report of an international working
group to standardize response criteria for myelodysplastic syndromes. Blood 2000;
96:3671–3674.
28. Cheson BD, Greenberg PL, Bennett JM, et al. Clinical application and proposal
for modification of the international working group (IWG) response criteria in
myelodysplasia. Blood 2006; 108:419–25.
29. Malcovati L, Della Porta M, Pascutto C, et al. Prognostic factors and life expectancy
in myelodysplastic syndromes classified according to WHO criteria. A basis for
clinical decision-making. J Clin Oncol 2005; 23:7594–7603.
30. Casadevall N, Durieux P, Dubois S, et al. Health, economic, and quality-of-life effects
of erythropoietin and granulocyte colony-stimulating factor for the treatment of
myelodysplastic syndromes: A randomized, controlled trial. Blood 2004; 104:321–
327.
31. Hellstrom-Lindberg E, Gulbrandsen N, Lindberg G, et al. A validated decision
model for treating the anaemia of myelodysplastic syndromes with erythropoietin
+ granulocyte colony-stimulating factor: Significant effects on quality of life. Br J
Haematol 2003; 120:1037–1046.
32. Jansen AJ, Essink-Bot ML, Beckers EA, et al. Quality of life measurement in
patients with transfusion-dependent myelodysplastic syndromes. Br J Haematol
2003; 121:270–274.
33. Izaks GJ, Westendorp RG, Knook DL. The definition of anemia in older persons.
JAMA 1999; 281:1714–1717.
34. Penninx BW, Guralnik JM, Onder G, et al. Anemia and decline in physical perfor-
mance among older persons. Am J Med 2003; 115:104–110.
35. Patel KV, Harris TB, Faulhaber M, et al. Racial variation in the relationship of
anemia with mortality and mobility disability among older adults. Blood 2007;
109:4663–4670.
36. Wu WC, Rathore SS, Wang Y, et al. Blood transfusion in elderly patients with acute
myocardial infarction. N Engl J Med 2001; 345:1230–1236.
37. Oliva EN, Dimitrov BD, Benedetto F, et al. Hemoglobin level threshold for car-
diac remodeling and quality of life in myelodysplastic syndrome. Leuk Res 2005;
29:1217–1219.
38. Della Porta MG, Malcovati L, Travaglino E, et al. A prognostic model for predicting
the impact of comorbidities on survival of patients with myelodysplastic syndromes.
39. Ferrucci L, Corsi A, Lauretani F, et al. The origins of age-related proinflammatory
state. Blood 2005; 105:2294–2299.
40. Nilsson-Ehle H, Jagenburg R, Landahl S, et al. Blood haemoglobin declines in the
elderly: Implications for reference intervals from age 70 to 88. Eur J Haematol 2000;
65:297–305.
41. Beutler E, Waalen J. The definition of anemia: What is the lower limit of normal of
the blood hemoglobin concentration? Blood 2006; 107:1747–1750.

J Clin Oncol 2007; 25:3503–10.
44. Fluit CR, Kunst VA, Drenthe-Schonk AM. Incidence of red cell antibodies after
multiple blood transfusion. Transfusion 1990; 30:532–535.
45. Kushner JP, Porter JP, Olivieri NF. Secondary iron overload. Hematol Am Soc
Hematol Educ Program 2001:47–61.
46. Buja LM, Roberts WC. Iron in the heart. Etiology and clinical significance. Am J
Med 1971; 51:209–221.
47. Jaeger M, Aul C, Sohngen D, et al. Secondary hemochromatosis in polytrans-
fused patients with myelodysplastic syndromes. Beitr Infusionsther 1992; 30:464–
468.
48. Takatoku M, Uchiyama T, Okamoto S, et al. Retrospective nationwide survey of
Japanese patients with transfusion-dependent MDS and aplastic anemia highlights
the negative impact of iron overload on morbidity/mortality. Eur J Haematol 2007;
78:487–494.
49. Schafer AI, Cheron RG, Dluhy R, et al. Clinical consequences of acquired transfu-
sional iron overload in adults. N Engl J Med 1981; 304:319–324.
50. Cazzola M, Barosi G, Gobbi PG, et al. Natural history of idiopathic refractory
52. Armand P, Kim HT, Cutler CS, et al. Prognostic impact of elevated pre-transplant
2007; 109:4586–8.
585.
54. Sierra J, Perez WS, Rozman C, et al. Bone marrow transplantation from
HLA-identical siblings as treatment for myelodysplasia. Blood 2002; 100:1997–
2004.
55. Angelucci E, Brittenham GM, McLaren CE, et al. Hepatic iron concentration and
total body iron stores in thalassemia major. N Engl J Med 2000; 343:327–331.
56. Olivieri NF, Nathan DG, MacMillan JH, et al. Survival in medically treated patients
with homozygous beta-thalassemia. N Engl J Med 1994; 331:574–578.
57. Porter JB. Practical management of iron overload. Br J Haematol 2001; 115:239–
252.
58. Jensen PD, Jensen FT, Christensen T, et al. Evaluation of transfusional iron overload
before and during iron chelation by magnetic resonance imaging of the liver and
determination of serum ferritin in adult non-thalassaemic patients. Br J Haematol
1995; 89:880–889.
59. Haines ME, Wainscoat JS. Relapsing sideroblastic anaemia. Br J Haematol 1991;
78:285–286.
60. Jensen PD, Heickendorff L, Pedersen B, et al. The effect of iron chelation on
haemopoiesis in MDS patients with transfusional iron overload. Br J Haematol
1996; 94:288–299.
61. Franchini M, Gandini G, de Gironcoli M, et al. Safety and efficacy of subcutaneous

bolus injection of deferoxamine in adult patients with iron overload. Blood 2000;
95:2776–2779.
62. Kersten MJ, Lange R, Smeets ME, et al. Long-term treatment of transfusional iron
overload with the oral iron chelator deferiprone (L1): A Dutch multicenter trial. Ann
Hematol 1996; 73:247–252.
63. Pippard MJ, Weatherall DJ. Oral iron chelation therapy for thalassaemia: An uncer-
tain scene. Br J Haematol 2000; 111:2–5.
64. Cappellini MD, Cohen A, Piga A, et al. A phase 3 study of deferasirox (ICL670),
a once-daily oral iron chelator, in patients with beta-thalassemia. Blood 2006;
107:3455–3462.
65. Porter J, Galanello R, Saglio G, et al. Relative response of patients with
myelodysplastic syndromes and other transfusion-dependent anaemias to deferasirox
(ICL670): A 1-yr prospective study. Eur J Haematol 2008; 80:168–176.
66. St Pierre TG, Clark PR, Chua-anusorn W, et al. Noninvasive measurement and
imaging of liver iron concentrations using proton magnetic resonance. Blood 2005;
105:855–861.
67. Anderson LJ, Holden S, Davis B, et al. Cardiovascular T2-star (T2∗) magnetic
resonance for the early diagnosis of myocardial iron overload. Eur Heart J 2001;
22:2171–2179.
68. Konen E, Ghoti H, Goitein O, et al. No evidence for myocardial iron overload
in multitransfused patients with myelodysplastic syndrome using cardiac magnetic
resonance T2 technique. Am J Hematol 2007; 82:1013–1016.
69. Chacko J, Pennell DJ, Tanner MA, et al. Myocardial iron loading by magnetic
resonance imaging T2∗ in good prognostic myelodysplastic syndrome patients on
long-term blood transfusions. Br J Haematol 2007; 138:587–593.
70. Stein RS, Abels RI, Krantz SB. Pharmacologic doses of recombinant human erythro-
poietin in the treatment of myelodysplastic syndromes. Blood 1991; 78:1658–1663.
71. A randomized double-blind placebo-controlled study with subcutaneous recombi-
nant human erythropoietin in patients with low-risk myelodysplastic syndromes.
Italian Cooperative Study Group for rHuEpo in myelodysplastic syndromes. Br J
Haematol 1998; 103:1070–1074.
72. Hellstrom-Lindberg E. Efficacy of erythropoietin in the myelodysplastic syndromes:
A meta-analysis of 205 patients from 17 studies. Br J Haematol 1995; 89:67–71.
73. Rodriguez JN, Dieguez JC, Muniz R, et al. Human recombinant erythropoietin in
the treatment of myelodysplastic syndromes anemia. Meta-analytic study. Sangre
(Barc) 1994; 39:435–439.
74. Hellstrom E, Birgegard G, Lockner D, et al. Treatment of myelodysplastic syndromes
with recombinant human erythropoietin. Eur J Haematol 1991; 47:355–360.
75. Bowen D, Culligan D, Jacobs A. The treatment of anaemia in the myelodysplastic
syndromes with recombinant human erythropoietin. Br J Haematol 1991; 77:419–
423.
76. Aloe Spiriti MA, Petti MC, Latagliata R, et al. Recombinant human erythropoietin
in the treatment of myelodysplastic syndromes. An interim report. Haematologica
1993; 78:123–126.
77. Mantovani L, Lentini G, Hentschel B, et al. Treatment of anaemia in myelodys-
plastic syndromes with prolonged administration of recombinant human granulo-
cyte colony-stimulating factor and erythropoietin. Br J Haematol 2000; 109:367–

375.
78. Hast R, Wallvik J, Folin A, et al. Long-term follow-up of 18 patients with myelodys-
plastic syndromes responding to recombinant erythropoietin treatment. Leuk Res
2001; 25:13–18.
79. Bennett CL, Luminari S, Nissenson AR, et al. Pure red-cell aplasia and epoetin
therapy. N Engl J Med 2004; 351:1403–1408.
80. Negrin RS, Stein R, Vardiman J, et al. Treatment of the anemia of myelodysplas-
tic syndromes using recombinant human granulocyte colony-stimulating factor in
combination with erythropoietin. Blood 1993; 82:737–743.
81. Negrin RS, Stein R, Doherty K, et al. Maintenance treatment of the ane-
mia of myelodysplastic syndromes with recombinant human granulocyte colony-
stimulating factor and erythropoietin: Evidence for in vivo synergy. Blood 1996;
87:4076–4081.
82. Remacha AF, Arrizabalaga B, Villegas A, et al. Erythropoietin plus granulocyte
colony-stimulating factor in the treatment of myelodysplastic syndromes. Identifi-
cation of a subgroup of responders. The Spanish Erythropathology Group. Haema-
tologica 1999; 84:1058–1064.
83. Hellstrom-Lindberg E, Ahlgren T, Beguin Y, et al. Treatment of anemia in myelodys-
plastic syndromes with granulocyte colony-stimulating factor plus erythropoietin:
Results from a randomized phase II study and long-term follow-up of 71 patients.
Blood 1998; 92:68–75.
84. Balleari E, Rossi E, Clavio M, et al. Erythropoietin plus granulocyte colony-
stimulating factor is better than erythropoietin alone to treat anemia in low-risk
myelodysplastic syndromes: Results from a randomized single-centre study. Ann
Hematol 2006; 85:174–180.
85. Miller KB, Kim HT, Greenberg P, et al. Phase III prospective randomized trial of
EPO with or without G-CSF versus supportive therapy alone in the treatment of
myelodysplastic syndromes (MDS): Results of the ECOG- CLSG trial (E1996).
Blood 2004; 104(11):24a.
86. Park S, Grabar S, Kelaidi C, et al. Predictive factors of response and survival in
myelodysplastic syndrome treated with erythropoietin and G-CSF: The GFM expe-
rience. Blood 2008; 111:574–582.
87. Jadersten M, Montgomery SM, Dybedal I, et al. Long-term outcome of treatment of
anemia in MDS with erythropoietin and G-CSF. Blood 2005; 106:803–811.
88. Jadersten M, Malcovati L, Dybedal I, et al. Erythropoietin and granulocyte-colony
stimulating factor treatment associated with improved survival in myelodysplastic
syndrome. J Clin Oncol 2008; 26:3607–13.
89. Henke M, Laszig R, Rube C, et al. Erythropoietin to treat head and neck cancer
patients with anaemia undergoing radiotherapy: Randomised, double-blind, placebo-
controlled trial. Lancet 2003; 362:1255–1260.
90. Leyland-Jones B, Semiglazov V, Pawlicki M, et al. Maintaining normal hemoglobin
levels with epoetin alfa in mainly nonanemic patients with metastatic breast cancer
receiving first-line chemotherapy: A survival study. J Clin Oncol 2005; 23:5960–
5972.
91. Wright JR, Ung YC, Julian JA, et al. Randomized, double-blind, placebo-controlled
trial of erythropoietin in non-small-cell lung cancer with disease-related anemia.
J Clin Oncol 2007; 25:1027–1032.
92. Khuri FR. Weighing the hazards of erythropoiesis stimulation in patients with cancer.
N Engl J Med 2007; 356:2445–2448.
93. Rizzo JD, Somerfield MR, Hagerty KL, et al. Use of epoetin and darbepoetin
in patients with cancer: 2007 American Society of Clinical Oncology/American
Society of Hematology clinical practice guideline update. J Clin Oncol 2008; 26:132–
149.
does make a difference. Blood 2004; 103:3265–3270.
95. Cazzola M, Ponchio L, Pedrotti C, et al. Prediction of response to recombinant
human erythropoietin (rHuEpo) in anemia of malignancy. Haematologica 1996;
81:434–441.
96. Bowen D, Hyslop A, Keenan N, et al. Predicting erythroid response to recombinant
erythropoietin plus granulocyte colony-stimulating factor therapy following a single
subcutaneous bolus in patients with myelodysplasia. Haematologica 2006; 91:709–
710.
97. Bowen D, Ehmer B, Neubert P, et al. The clearance of a single i.v. bolus of recom-
binant human erythropoietin from the serum of patients with myelodysplastic syn-
dromes and its effects on erythropoiesis. Exp Hematol 1991; 19:613–616.
98. Aul C, Arning M, Runde V, et al. Serum erythropoietin concentrations in patients
with myelodysplastic syndromes. Leuk Res 1991; 15:571–575.
99. Kelley LL, Green WF, Hicks GG, et al. Apoptosis in erythroid progenitors deprived
of erythropoietin occurs during the G1 and S phases of the cell cycle without growth
arrest or stabilization of wild-type p53. Mol Cell Biol 1994; 14:4183–4192.
100. Kelley LL, Koury MJ, Bondurant MC, et al. Survival or death of individual proery-
throblasts results from differing erythropoietin sensitivities: A mechanism for con-
trolled rates of erythrocyte production. Blood 1993; 82:2340–2352.
101. Sui X, Krantz SB, Zhao ZJ. Stem cell factor and erythropoietin inhibit apoptosis
of human erythroid progenitor cells through different signalling pathways. Br J
Haematol 2000; 110:63–70.
102. Millot GA, Svinarchuk F, Lacout C, et al. The granulocyte colony-stimulating fac-
tor receptor supports erythroid differentiation in the absence of the erythropoietin
receptor or Stat5. Br J Haematol 2001; 112:449–458.
103. Verhoef G, van den Berghe H, Boogaerts M. Cytogenetic effects on cells derived
from patients with myelodysplastic syndromes during treatment with hemopoietic
growth factors. Leukemia 1992; 6:766–769.
104. Rigolin GM, Porta MD, Ciccone M, et al. In patients with myelodysplastic syndromes
response to rHuEPO and G-CSF treatment is related to an increase of cytogenetically
normal CD34 cells. Br J Haematol 2004; 126:501–507.
105. Ljung T, Back R, Hellstrom-Lindberg E. Hypochromic red blood cells in low-risk
myelodysplastic syndromes: Effects of treatment with hemopoietic growth factors.
Haematologica 2004; 89:1446–1453.
106. Mannone L, Gardin C, Quarre MC, et al. High-dose darbepoetin alpha in the treat-
ment of anaemia of lower risk myelodysplastic syndrome results of a phase II study.
Br J Haematol 2006; 133:513–519.
107. Musto P, Lanza F, Balleari E, et al. Darbepoetin alpha for the treatment of anaemia
in low-intermediate risk myelodysplastic syndromes. Br J Haematol 2005; 128:204–
209.
108. Stasi R, Abruzzese E, Lanzetta G, et al. Darbepoetin alfa for the treatment of anemic
patients with low and intermediate-1-risk myelodysplastic syndromes. Ann Oncol
2005; 16:1921–1927.
109. Rose S, Ali Y, Maffei B, et al. Treatment of anemia in myelodysplastic syndrome
with darbepoetin and granulocyte colony stimulating factor. Am J Hematol 2007;
82:245–246.
110. Osterborg A, Steegmann JL, Hellmann A, et al. Phase II study of three dose levels
of continuous erythropoietin receptor activator (C.E.R.A.) in anaemic patients with
aggressive non-Hodgkin’s lymphoma receiving combination chemotherapy. Br J
Haematol 2007; 136:736–744.
111. Dmoszynska A, Kloczko J, Rokicka M, et al. A dose exploration, phase I/II study of
administration of continuous erythropoietin receptor activator once every 3 weeks
in anemic patients with multiple myeloma receiving chemotherapy. Haematologica
2007; 92:493–501.
112. Park S, Andrieu V, Sapena R, et al. MDS with Isolated neutropenia or thrombocytope-
nia. Incidence and characteristics in the Groupe Francophone des Myelodysplasies
(GFM) Registry. ASH Annu Meet Abstr 2007; 110:2456.
113. Willemze R, van der Lely N, Zwierzina H, et al. A randomized phase-I/II multicen-
ter study of recombinant human granulocyte-macrophage colony-stimulating factor
(GM-CSF) therapy for patients with myelodysplastic syndromes and a relatively low
risk of acute leukemia. EORTC Leukemia Cooperative Group. Ann Hematol 1992;
64:173–180.
114. Smith TJ, Khatcheressian J, Lyman GH, et al. 2006 update of recommendations
for the use of white blood cell growth factors: An evidence-based clinical practice
guideline. J Clin Oncol 2006; 24:3187–3205.
115. Arshad M, Seiter K, Bilaniuk J, et al. Side effects related to cancer treatment: CASE
2. Splenic rupture following pegfilgrastim. J Clin Onc 2005; 25:8533–8534.
117. Chan G, DiVenuti G, Miller K. Danazol for the treatment of thrombocytopenia in
patients with myelodysplastic syndrome. Am J Hematol 2002; 71:166–171.
118. Chabannon C, Molina L, Pegourie-Bandelier B, et al. A review of 76 patients with
myelodysplastic syndromes treated with danazol. Cancer 1994; 73:3073–3080.
119. Newland A, Caulier MT, Kappers-Klunne M, et al. An open-label, unit dose-finding
study of AMG 531, a novel thrombopoiesis-stimulating peptibody, in patients with
immune thrombocytopenic purpura. Br J Haematol 2006; 135:547–553.
120. Kantarjian H, Fenaux P, Sekeres MA, et al. Phase 1/2 study of AMG 531 in throm-
bocytopenic patients (pts) with low-risk myelodysplastic syndrome (MDS): Update
including extended treatment. ASH Annu Meet Abstr 2007; 110:250.
18
Immune Dysregulation and the Role for
Immunotherapy in Myelodysplastic
Syndrome (MDS)
Kebede Hussein and David P. Steensma

Division of Hematology, Department of Medicine, Mayo Clinic, Rochester,
Minnesota, U.S.A.
INTRODUCTION
While the pathogenesis of MDS is not yet fully defined, hematopoietic failure
resulting in peripheral blood cytopenias is the central characteristic of the dis-
ease. Several preclinical and clinical observations point to the body’s intrinsic
cellular defense mechanism as an important contributor to MDS-associated inef-
fective hematopoiesis. These observations have clinical relevance because in some
patients it is possible to use therapies that alter the number and activity of specific
immune cell populations, thereby modifying a disordered immune response and
restoring normal hematopoiesis. Indeed, the current (2008 v.2) iteration of the
National Comprehensive Cancer Network (NCCN) guideline for MDS therapy
(www.nccn.org) recommends an early assessment of whether a patient is a poten-
tial immunotherapy candidate as part of the suggested management algorithm,
attesting to the considerable promise of this treatment modality.
In this chapter, we review available information with regard to immune
dysregulation as a cause of MDS-associated failed hematopoiesis, and we discuss
the potential role for therapies directed at the immune system in patients suffering
from MDS (for recent reviews on this topic, see Refs. 1–4).
437
438 Hussein and Steensma
MDS AS A MARROW FAILURE SYNDROME

Understanding the rationale behind immune-based therapies requires an under-
standing of MDS in the context of other acquired bone marrow failure syndromes
(5–9), since it is lessons learned from treating patients with other syndromes that
initially informed current approaches to MDS.
In addition to MDS, acquired bone marrow failure syndromes include aplas-
tic anemia (AA), paroxysmal nocturnal hemoglobinuria (PNH), and large granular
lymphoproliferative (LGL) disorders. The distinction between these disorders is
not always clear, and some patients have features of more than one marrow
failure syndrome simultaneously (e.g., the presence of a PNH clone in an MDS
patient, or markedly reduced cellularity in MDS marrow that is reminiscent of AA)
(Fig. 1). There is also potential for one disorder to evolve into another.
In each of these marrow failure syndromes, but especially in AA, laboratory
and clinical evidence point to immune destruction or suppression of hematopoietic
stem or progenitor cells as a central component of the pathobiology. However, it
is not clear what initiates the aberrant immune response to hematopoietic cells,
nor are the mechanisms by which loss of normal self-tolerance is maintained
understood in any detail (10).
Overlap among acquired bone marrow failure states
Figure 1 There is substantial clinical and pathobiological overlap among the various
acquired bone marrow failure disorders. Hypocellular MDS resembles aplastic anemia,
while PNH and LGL clones can appear by themselves or in association with AA or MDS.
The Role for Immunotherapy in Myelodysplastic Syndrome 439
Aplastic Anemia
Among the acquired marrow failure syndromes, the centrality of an immunologi-
cal mechanism is best accepted for AA (11). Evidence supporting this hypothesis
has been accumulating since the early 1970s. Key findings have included the clin-
ical utility of antilymphocyte serum in patients with AA undergoing allogeneic
transplantation, even when the graft itself proved relatively unsuccessful (12);
the observation that AA patients treated with syngeneic hematopoietic stem cell
transplantion quickly relapse unless they also receive immunosuppression (13);
reports of striking clinical responses to immunosuppressive drugs directed specif-
ically at T lymphocytes, such as antithymocyte globulin and cyclosporine (14);
and detection of aberrant autoreactive T cells that can mediate hematopoietic cell
destruction in vitro (Fig. 2).
Paroxysmal Nocturnal Hemoglobinuria

In PNH, patients’ cell-mediated immunity is usually intact (unless they have con-
comitant AA, in which case the above considerations apply), but clonal hematopoi-
etic cells with abnormal properties manage to proliferate without being recognized
by the immune system. PNH is an uncommon form of hemolytic anemia, but has
been highly informative biologically (15–17). The disorder results from acquired
clonal expansion of hematopoietic stem/progenitor cells that have somatic muta-
tions in the X-linked gene, PIGA, which encodes a phosphatidylinositol glycan
anchor. This glycosylphosphatidylinositol (GPI) anchor molecule is responsible
for tethering many proteins to the cell surface. Consequently, a proportion of the
blood cells in patients with PNH have a partial (type II) or complete deficiency
(type III) of GPI-linked proteins. Since some of these GPI-linked proteins—
specifically, CD55 (decay accelerating factor) and CD59 (complement regula-
tory protein)—normally protect cells against complement-mediated lysis, their
absence increases cells’ susceptibility to complement-mediated lysis, resulting in
the hemolysis characteristic of the disease.
PNH cells often arise in marrow of AA, presumably because autoreactive
immune destruction of normal hematopoietic cells in AA provides the “stealth”
PNH cells with a selective survival advantage. The specific molecules recognized
by the immune system in AA and lacking in PNH are unknown. Similar mecha-
nisms may also explain the occasional presence of a PNH clone in patients with
MDS. The frequency and clinical importance of PNH clones in MDS is currently
being assessed in a large registry-based study by the manufacturer of eculizumab, a
monoclonal anti-C5 complement antibody that prevents formation of the terminal
complement complex and subsequent erythrocyte lysis.
Large Granular Lymphocyte Disorders

LGL disorders are usually chronic and indolent conditions, and are character-
ized by clonal proliferation of CD8+ T cells, or, more rarely, natural killer cells.
Figure 2 A model for how an interleukin-2 driven expanded cytotoxic (CD8+) T-cell
clone may suppress hematopoiesis. Ligand–receptor interactions alter intracellular signal-
ing, ultimately resulting in altered gene expression and impaired growth and differentiation
of hematopoietic progenitor cells; cell–cell interactions between lymphocytes and progen-
itor cells (not shown) also play an important role. Secreted cytokines such as interferon-␥
(IFN-␥ ), Fas ligand, tumor necrosis factor (TNF-␣), and TNF-related apoptosis-inducing
ligand (TRAIL) are important signaling molecules. There is extensive interaction among
the apoptosis-inducing death receptor pathways (TNF-␣, Fas/Fas ligand, and TRAIL) and
between these pathways IFN-␥ ; for instance, and IFN-␥ and TNF-␣ can upregulate each
other’s cellular receptors and also upregulate the Fas receptor. The intracellular interferon
regulatory factor 1 (IRF-1) inhibits the transcription of cellular genes and entry of the
cell into the cell cycle, and mediates some of the effects of IFN-␥ . Source: Adapted from
Ref. 11. (see color insert)
Neutropenia is the most frequent clinical manifestation of LGL disorders, and

autoimmune disorders are often present as well (18). Patients with LGL disorders
often respond to immunomodulating therapies such as methotrexate or corticos-
teroids (18). LGL clones and MDS have also been detected simultaneously (19).
In some patients with MDS, an apparent excess of large granular lymphocytes
in blood or marrow is observed, yet clonality studies (e.g., T-cell receptor gene
rearrangement assessment by Southern blotting or PCR) are negative or equivocal
(20). The significance of this finding is unclear.
Comparisons Between Marrow Failure States

In contrast to AA, where hematopoiesis in insufficient but remains polyclonal (or
at least oligoclonal), the other three acquired marrow failure states (MDS, LGL,
and PNH) are clonal disorders, although some residual normal hematopoietic
clones usually remain.
Several hypotheses have been forwarded to explain the high incidence of
MDS and PNH arising in patients originally diagnosed with AA (21). Often,
sensitive techniques (e.g., flow cytometry with antibodies to GPI-linked proteins)
allow detection of PNH clone at the time of AA diagnosis, and the same may be
true of clones with dysplastic features (e.g., those with chromosomal abnormalities
consistent with MDS). Prolonged survival of patients with AA might then allow
the PNH and/or the MDS clone to become the dominant clone, changing the nature
of the disease.
Another possibility is related to the “field cancerization theory”: Just as
cigarette smoking predisposes to cancer throughout the aerodigestive tract and may
contribute to multiple simultaneous neoplasms, abnormal bone marrow clones of
different types may arise following a similar genotoxic insult (22,23). While clonal
PNH cells can be found at diagnosis of AA or MDS (as mentioned above), some-
times they are found months or years after successful treatment of AA and MDS
with immunosuppressive therapy (24,25). The latter raises questions about whether
the two disorders are directly related, or simply both present in a susceptible host.
Diagnostic Challenges
Although 80% to 85% of patients with MDS have a bone marrow that is either
hypercellular or normocellular for age, the rest have hypocellular bone marrow,
which may predict a more favorable clinical course (26). When hypocellularity is
extreme, the marrow blast proportion is normal (⬍5%), and dysplastic cells rare,
it can be difficult to differentiate AA from MDS (27). Several markers have been
proposed to aid in this distinction. For instance, patients with overall MDS have an
increased number of marrow CD34+ cells (28) and HbF-containing erythroblasts,
as well as a higher percentage of Ki-67+ cells compared to AA, and the presence
of a clonal chromosomal abnormality also strongly favors a diagnosis of MDS
(28–31). Dysplastic neutrophils with pseudo-Pelger-Huët nuclei or hypogranular
cytoplasm are atypical in AA, and their presence suggests MDS (32,33). To some
extent, the distinction may be arbitrary, as cytopenias are the major clinical problem
in both settings, and evidence of T-cell mediated immunosuppression can be found
in each disorder (27).
IMMUNOLOGICAL CONTRIBUTION TO THE HEMATOPOIETIC

FAILURE IN MDS
Hematopoiesis and Hematopoietic Failure in MDS
As discussed in chapter 5, early MDS is characterized by increased apoptosis
of normal and/or abnormal progenitor cells. There is an experimental evidence
indicating that cytopenias may also be related to abnormal cytokine levels or cell-
mediated suppression of hematopoiesis (34,35). A subset of MDS patients have a
decreased CD4 to CD8 ratio, increased absolute numbers of CD8+ T cells (which
may be LGL cells), or increased levels of inhibitory cytokines like interferon-

␥ , tumor necrosis factor-␣, and Fas ligand (36–39). Cytokine release by clonal
cells may injure “innocent bystander” normal hematopoietic cells, worsening
cytopenias.
Some have divided the immune pathogenesis of MDS into “intrinsic”
(cytokine-mediated) and “extrinsic” (cell-mediated) defects, although there are
extensive areas of overlap between the two (40). Inhibitory cytokines may be
released as a result of complex interactions between stromal elements, transformed
stem, or progenitor cells that are part of the MDS clone, and normal stem cells,
with the final end point of ineffective hematopoiesis. In addition, cytotoxic T cells
may be directed at the MDS clone in the bone marrow, and normal hematopoietic
cells may be eliminated as a result of their proximity or similarity to transformed
cells (Fig. 2).
One major piece of evidence supporting the importance of these mecha-
nisms in MDS is that anticytokine and immunosuppressive or immunomodulatory
agents have had positive clinical effects in the management of a subset of patients
with MDS. While these agents are not useful in all patients with MDS, no
treatment has proven effective in all patients with MDS; this is a consequence
of the complex and heterogenous nature of MDS. Antithymocyte globulin (ATG)
and cyclosporine (CSA) play a significant role in the management of AA, which
resulted in their use in hypocellular MDS and subsequently other forms of
MDS, regardless of marrow cellularity; ATG and CSA will be discussed later.
Thalidomide and lenalidomide are touted as “immunomodulatory” agents, and
indeed, these drugs dramatically alter cytokine levels and can change the balance
of lymphocyte subsets in MDS patients. These medications are discussed further
in chapter 19. Finally, neutralizing the effect of TNF-␣ with a soluble Fc:TNF-␣
receptor fusion protein (etanercept) or the chimeric anti-TNF-␣ monoclonal
antibody cA2 (infliximab) may ameliorate the cytopenias of MDS and improve
constitutional symptoms in some cases (41,42). While these anti-TNF agents are
rarely used for patients with MDS in clinical practice and are not approved by
the Food and Drug Administration for MDS-related indications, this anecdotal
success emphasizes the pathobiological importance of cytokine excess in MDS.
The mechanisms underlying abnormal cytokine levels in MDS are unclear.
One hypothesis relates to disordered activity of the mitogen-activated protein
kinase (MAPK) family, as these factors are important in regulating the growth
inhibitory signals of TNF-␣, TGF-␤, and interferon family members on human
hematopoiesis (43). In particular, p38 MAPK is found to be overactivate in a
subset of MDS bone marrow samples, and regulates hematopoietic progenitor
cell apoptosis (44,45).
Clinical Observations of Immune Defects and Autoimmunity

in Patients with MDS
Several groups have observed that ∼10% to 15% of patients with MDS exhibit clin-
ical or serological evidence of autoimmunity (46–50). Autoimmune manifestations
may precede MDS, may be present at diagnosis, or can develop at any time during
the clinical course. Because autoimmunity is so common in the general popula-
tion, when an autoimmune disorder and MDS coexist, it is often unclear whether
the two are pathologically related or merely a coincidence. Supporting the idea
of coincidence, autoimmune phenomena in MDS are very nonspecific and het-
erogenous, correlate poorly with MDS disease features, and have little effect on
the clinical course (50).
Described autoimmune diseases associated with MDS include various
forms of vasculitis, pyoderma gangrenosum, Coombs-positive hemolytic ane-
mia, immune thrombocytopenic purpura, chronic inflammatory demyelinating
polyneuropathy, and classic connective tissue disorders including rheumatoid
arthritis and relapsing polychondritis (2,47,49–51). Sometimes a diagnosis of
“autoimmunity” is based on only the finding of a positive antinuclear antibody
or rheumatoid factor test, or detection of abnormal concentrations of particular
immunoglobulin subsets. Nevertheless, the fact that treatment directed at MDS
often ameliorates a clinical autoimmune disorder raises the possibility of a patho-
biological connection, at least in some cases.
Recent data indicated that V␥ 9V␦2 T cells, the major circulating ␥ ␦ T-cell
subset and an important component of innate immunity, are reduced in patients
with MDS who have associated autoimmune diseases (52) compared to levels
in healthy donors. These cells are not clonal in MDS, but have intrinsic defects
limited to proliferative capacity in response to IL-2 despite normal expression of
IL-2 receptor (52).
A B Cell Defect in MDS?

Although most attention has focused on the role of T cells in MDS-associated
hematopoietic failure, recent studies have implied the presence of a B-cell abnor-
mality in at least some patients with MDS (53). A CD34+ progenitor cell microar-
ray study from the United Kingdom indicated dysregulation of the expression of
a number of B-cell lineage-associated genes in patients with lower-risk MDS,
compared to patients with other forms of anemia and healthy controls (54). A
follow-up flow cytometry study then demonstrated reduced numbers of B-cell
progenitor cells in MDS patients compared to healthy controls (54), and these
findings were independently confirmed (55). One hypothesis to explain this obser-
vation is that expansion of erythroid- and neutrophil-lineage CD34+ cells may
come at the expense of CD34+ plasmacytoid dendritic cell and B-cell precursors
(56).
CLINICAL TRIALS OF IMMUNOSUPPRESSIVE AGENTS

IN PATIENTS WITH MDS
The goal of immunosuppressive therapy in MDS is to restore normal hematopoiesis
by reversing inhibitory factors suppressing normal progenitor cell maturation
(3,57). There is no question that some patients with MDS experience clinical
improvement after treatment with these agents. The greatest challenge in this
area, however, has been predicting just who these patients will be, so that patients
unlikely to benefit can be spared the potential adverse consequences of therapy
(2). Several predictive models have been proposed (58), but thus far, only younger
age at diagnosis and lower-risk disease (e.g., refractory anemia MDS subtype,
with minimal red cell transfusion needs) have been consistently linked to a higher
likelihood of response.
Antithymocyte Globulin (ATG)

Because of the striking responses to ATG treatment in patients with AA, con-
siderable similarity between hypocellular MDS and AA, and observations of
autoimmune disorders in MDS (described above), several investigators have used
ATG to treat patients with MDS. The leaders in this area have been the National
Institutes of Health group of John Barrett, Elaine Sloand, and Neil Young, who
first began a prospective trial of ATG in MDS in 1995 (2) and have now treated
more than 120 patients with this therapy. In a 2002 report (59) describing their
experience with the first 61 patients, transfusion independence had developed in
33%, attesting to the potential for benefit (Table 1). An earlier 1997 report (57) by
the NIH group inspired several other investigators to try to replicate the results.
Response rates vary considerably between the reported ATG series (Table 1),
as has the median duration of response (usually about 1 year). This variability
likely reflects not only different criteria used to assess response (less of an issue
now that standardized International Working Group criteria are available) but also
patient selection. The number of CD8+ T cells in the bone marrow of patients
who respond to ATG often decreases (60), while the flow cytometry pattern in the
marrow improves (61).
The most common ATG dose and schedule used is 40 mg/kg/day × 4 days,
with or without CSA. It is not clear whether CSA adds anything to ATG treatment
in MDS or how long CSA therapy should continue, though responses are also seen
to CSA therapy alone (see below). Prednisone is often used during and after ATG
therapy to minimize the effects of serum sickness, the most common treatment-
related adverse event from ATG. It is possible that prednisone itself may be an
active drug in some cases (62).
Several investigators have proposed patient factors that may predict response
to ATG-based therapy. For instance, the NIH group identified patient age (younger
is better), duration of red blood cell transfusion in months (shorter is better), and
HLA-DR15 status (positive is better) as factors predicting response to immunosup-
pression with ATG and CSA in their 2002 study (58). Importantly, a hypocellular
marrow did not predict response (2). HLA-DR15 was explored because of its asso-
ciation with several autoimmune disorders and AA, where it is seen at a higher
frequency than expected in the general population (63,64). In a retrospective study
of 96 ATG-treated patients at various centers, a European consortium identified
lower IPSS score and bone marrow hypocellularity as predictors of hematological
Table 1 Prospective Clinical Trials of Antithymocyte Globulin (ATG) and/or Cyclosporine A (CSA) in MDS
No. of patients
enrolled Agent used Response ratesa Factors predicting response Reference
25 ATG 40 mg/kg/day × 4 days 44% transfusion independence RA (57)

17 (including CSA 5–6 mg/kg/day 23% trilineage improvement; 82% RA (69)
8 hypoplastic) hemoglobin improvement
9 CSA 1–3 mg/kg/day 55% durable hematological N/A (84)
improvement
8 ATG + CSA 50% None (85)
61 ATG 40 mg/kg/day 34% transfusion independence; 55% Younger age, lower platelet (59)
improvement in ANC; 48% platelet count, HLA DR-15, ?trisomy
increase 8, not monosomy 7
6 CSA 0 responses N/A (86)
11 hypoplastic, CSA 1–3 mg/kg/day 73% PR in hypoplastic Hypoplastic (39)
20 RA
8 (study stopped ATG 40 mg/kg/day 0 responses N/A (87)
The Role for Immunotherapy in Myelodysplastic Syndrome
early)
31 ATG 40 mg/kg/day + CSA to 17% CR None (88)
level 200–400 mg/dL
(Continued)
445
446
Table 1 Prospective Clinical Trials of Antithymocyte Globulin (ATG) and/or Cyclosporine A (CSA) in MDS (Continued)
No. of patients
enrolled Agent used Response ratesa Factors predicting response Reference
30 ATG 1.5 vials/10 kg/day × 50% transfusion independence RA (89)

5 days
50 CSA median dose 4.6 8% PR, 53% HI-E, 35% HI-N, Only RA patients responded (90)
mg/kg/day 36% HI-P (IWG 2000)
35 Rabbit vs. equine ATG 42% response rate in each arm Younger age (68)
19 CSA 3–5 mg/kg/day 37% HI-E (major) NR (91)
20 ATG+CSA 15% CR; overall response rate 30% NR (66)
32 CSA 3–6 mg/kg/day 63% response rate NR (92)
a Non-uniform criteria were used for response assessment across trials, with some authors choosing response criteria that are more lenient than the 2006 version of
the International Working Group definitions. In some cases, patients also received erythropoietin or danazol transiently, which may have contributed to response.
Abbreviations: RA, refractory anemia; CR, complete response; PR, partial response; HI, hematological improvement (E, erythroid; P, platelet; N, neutrophil); IWG,
International Working Group response criteria; ATG, antithymocyte globulin; CSA, cyclosporin A.
Hussein and Steensma
response to ATG [IPSS Int-2/high: odds ratio (OR) 0.08, p = 0.018; BM

normo/hypercellularity: OR 0.49, p = 0.012) (65). Older patients have more
difficulty tolerating ATG than younger patients (66).
More recently, the NIH group have shown that patients with monosomy 7
are very unlikely to respond to ATG therapy, whereas those with trisomy 8 have a
67% chance of gaining benefit and the trisomy 8 clone is preferentially suppressed
(67). It is notable that this response proportion is as high as the rate of transfusion
independence with del(5q) MDS patients treated with lenalidomide.
There are currently two animal sources for ATG in the United States: horse
(equine) and rabbit. A German study of 35 patients with MDS compared horse
ATG and rabbit ATG directly (68). The investigators observed no significant
difference in response rate (42% for both) or adverse events. Factors predicting
response to ATG in that study included short interval from diagnosis to the study
therapy and a diagnosis of refractory anemia; none of the patients with RAEB or
CMML responded (68).
Prednisone and Other Corticosteroids

Corticosteroids have been used alone or in combination with other therapeutic
modalities in the management of MDS. The response rate to corticosteroids is
limited, and chronic use of these drugs is associated with significant side effects,
including an increase in infections. Additionally, increased neutrophil counts after
prednisone or dexamethasone therapy do not necessarily mean that hematopoiesis
has improved. Instead, these changes may just represent demargination or a shift
of white cells from the storage pool to the circulating pool, phenomena that are
not associated with a decrease in infections. Currently, corticosteroids are rarely
used as single agents in MDS, except to treat specific autoimmune complications
(e.g., Coombs-positive warm autoantibody hemolytic anemia). However, they
do have an important role in minimizing complications of ATG therapy (see
above.)
Cyclosporine
CSA selectively inhibits transcription of interleukin-2 and several other cytokines,
mainly in T-helper lymphocytes. It inhibits clonally expanded cytotoxic (CD8+) T
cells as well. Although CSA had been used in the transplant setting since the early
1980s, the first prospective study in MDS was reported in 1998 (69) (Table 1).
In this study, 14 of 17 patients (82%) experienced a hematological response of
some sort, with complete trilineage recovery in 23% (4/17) of patients (69). No
treatment failures occurred in the responding patients during the follow up period
of 5 to 30 months. The optimal duration of treatment is not yet defined; however,
CSA treatment interruption is associated with recurrence of cytopenias. Infectious
complications with CSA are uncommon, but blood pressure and renal function
have to be monitored closely.
TNF-α Inhibitors
TNF-␣ and its receptors are overexpressed in patients with MDS, and in vitro
studies have shown that blockade of TNF-␣ can increase hematopoietic colony
formation from marrow of MDS patients (41,70). For this reason, TNF-␣ antago-
nists have been tried in patients with MDS (42).
In the largest clinical trial performed to date in MDS patients, 37 patients
received infliximab at 1 of 2 doses: 5 or 10 mg/kg intravenously every 4 weeks
for 4 cycles (42). Response was evaluated using the International Working Group
criteria in 28 patients who completed the planned 4 cycles. Eight patients (22% of
those enrolled) showed a hematologic response, mostly at the higher infliximab
dose. However, there was no documented change in the pathological features of
the disease.
Etanercept has also been tried as a single agent in one pilot study for the
treatment of MDS in 14 cases. Etanercept was given at 25 mg subcutaneous three
times per week for 16 weeks. The study drug resulted in moderate improvement in
cytopenias in 9 of 12 evaluable patients, while cell counts declined in others (41),
similar to what was seen in a trial of etanercept in primary myelofibrosis (71). In a
follow-up study in MDS, etanercept was combined with ATG (61); adverse events
were mild and a 46% hematological improvement rate was observed.
No prospective studies of adalimumab, a newer anti-TNF-␣ agent, have
been published. Given the substantial costs of anti-TNF-␣ therapies, they are not
commonly used in MDS at present.
Sirolimus
Sirolimus inhibits T lymphocyte activation and proliferation in response to anti-
genic and cytokine stimulation, and also inhibits autoantibody production (72–74).
Sirolimus binds to FKBP-12, an intracellular protein, to form an immunosup-
pressive complex that inhibits the regulatory kinase, mTOR (mammalian target
of rapamycin). This inhibition suppresses cytokine-mediated T-cell proliferation,
halting progression of lymphocytes from the G1 to the S phase of the cell cycle.
Sirolimus is used to inhibit acute rejection of solid organ allografts and has been
shown to prolong graft survival.
Sirolimus has been shown to have anti-proliferative and direct pro-apoptotic
effects on hematopoietic and leukemia cells (72–74). One pilot study in MDS
patients showed hematological remission in 16% (3/19) of cases (75). However,
32% of the patients in this study could not complete more than 3 months of therapy
due to side effects.
Immunomodulatory Agents
Thalidomide and lenalidomide may result in their clinical effects in MDS via
immunological mechanisms, and are discussed in chapter 19.
p38 MAPK Inhibitors

As discussed above, p38 MAPK constitutive activation may be important in pro-
moting the growth inhibitory signals of TNF-␣, TGF-␤, and interferons that sup-
press hematopoiesis in MDS (44,76). MAPK activation (including p38 MAPK
and activation of downstream targets of p44/p42 MAPK) is uniformly observed in
varied morphologic subtypes of low-risk MDS (77), and correlates with enhanced
apoptosis observed in MDS hematopoietic progenitors (76). Inhibition of p38
MAPK by genetic or pharmacologic means decreases apoptosis and stimulates
in vitro hematopoiesis from primary MDS hematopoietic progenitors (76,78–80).
SCIO-469 and SD-282 are p38 MAPK inhibitors, and thus block a conver-
gence point involved in apoptosis signaling by death receptors or hematopoietic
inhibitory cytokines. These novel small molecule inhibitors occupy the ATP-
binding site on the p38 kinase (79). Unfortunately, due to limited response rates
in early-phase clinical trials, development of these agents in MDS has been
suspended.
Vaccine Approaches
PGP (primary granular proteins) such as PR1 may serve as important immunother-
apeutic targets for vaccine-based approaches in myeloid disorders, including MDS.
PR1 is a nine-amino-acid HLA-A∗ 0201-restricted peptide found in proteinase 3,
and it induces myeloid leukemia–specific cytotoxic T cell responses (81). In an
early-phase study in myeloid leukemia, a PR1 vaccine induced T-cell responses
in 22 of 37 patients and hematological responses in 16 of those patients. Some
patients with advanced myeloid leukemia also developed sustained remissions
(82). These strategies deserve further exploration, including in MDS.
The wild-type Wilms tumor gene, WT1, is overexpressed in some patients
with MDS or AML. In a phase I clinical trial of biweekly vaccination with HLA-
A∗2402-restricted WT1 peptide for these malignancies, two patients with MDS
developed severe leukocytopenia in association with a reduction in leukemic blast
cells and levels of WT1 mRNA. This occurred only after a single vaccination with
0.3 mg of WT1 peptide. Correlative studies indicated that the WT1-specific cyto-
toxic T lymphocytes elicited by WT1 vaccination eradicated the WT1-expressing
transformed stem or progenitor cells (83).
CONCLUSION
Diverse lines of evidence point to an important role for immune dysregulation
in MDS-associated hematopoietic failure. Immunotherapeutic approaches have
considerable promise, and a substantial proportion of younger patients with lower-
risk disease will benefit from therapy with ATG, with or without CSA. Better
response prediction models are needed before clinicians will feel comfortable
prescribing ATG and other agents widely in MDS. Ultimately, vaccine therapy
or other immunotherapeutic approaches may provide definitive therapy for MDS

without the toxicities currently associated with cytotoxic agents.
REFERENCES
1. Sloand EM, Rezvani K. The role of the immune system in myelodysplasia: Implica-
tions for therapy. Semin Hematol 2008; 45(1):39–48.
2. Barrett J, Sloand E, Young N. Determining which patients with myelodysplas-
tic syndrome will respond to immunosuppressive treatment. Haematologica 2006;
91(5):583–584.
3. Kerbauy DB, Deeg HJ. Apoptosis and antiapoptotic mechanisms in the progression
of myelodysplastic syndrome. Exp Hematol 2007; 35(11):1739–1746.
highlighting current controversies. Leuk Res 2003; 27(2):95–120.
5. Hinterberger W, Rowlings PA, Hinterberger-Fischer M, et al. Results of transplanting
bone marrow from genetically identical twins into patients with aplastic anemia. Ann
Intern Med 1997; 126(2):116–122.
6. Hoffman R, Zanjani ED, Lutton JD, et al. Suppression of erythroid-colony formation
by lymphocytes from patients with aplastic anemia. N Engl J Med 1977; 296(1):10–
13.
7. Nissen C, Cornu P, Gratwohl A, et al. Peripheral blood cells from patients with aplastic
anaemia in partial remission suppress growth of their own bone marrow precursors in
culture. Br J Haematol 1980; 45(2):233–243.
8. Wlodarski MW, Gondek LP, Nearman ZP, et al. Molecular strategies for detection and
quantitation of clonal cytotoxic T-cell responses in aplastic anemia and myelodys-
plastic syndrome. Blood 2006; 108(8):2632–2641.
9. Nissen C, Schbert J. Seeing the good and the bad in aplastic anemia: Is autoimmunity
in AA dysregulated or antineoplastic? Hematol J 2002; 3:169–175.
10. Maciejewski JP, O’Keefe C, Gondek L, et al. Immune-mediated bone marrow failure
syndromes of progenitor and stem cells: Molecular analysis of cytotoxic T cell clones.
Folia Histochem Cytobiol 2007; 45(1):5–14.
11. Young NS, Maciejewski J. The pathophysiology of acquired aplastic anemia. N Engl
J Med 1997; 336(19):1365–1372.
12. Mathe G, Amiel JL, Schwarzenberg L, et al. Bone marrow graft in man after condi-
tioning by antilymphocytic serum. Br Med J 1970; 2(5702):131–136.
13. Champlin RE, Feig SA, Sparkes RS, et al. Bone marrow transplantation from identical
twins in the treatment of aplastic anaemia: Implication for the pathogenesis of the
disease. Br J Haematol 1984; 56(3):455–463.
14. Speck B, Gratwohl A, Nissen C. Treatment of severe aplastic anaemia with antilym-
phocyte globulin or bone-marrow transplantation. Br Med J 1981; 282:860–863.
15. Takeda J, Miyata T, Kawagoe K, et al. Deficiency of the GPI anchor caused by a
somatic mutation of the PIG-A gene in paroxysmal nocturnal hemoglobinuria. Cell
1993; 73(4):703–711.
16. Bessler M, Mason PJ, Hillmen P, et al. Paroxysmal nocturnal haemoglobinuria (PNH)
is caused by somatic mutations in the PIG-A gene. Embo J 1994; 13(1):110–117.
17. Hillmen P, Young NS, Schubert J, et al. The complement inhibitor eculizumab in
paroxysmal nocturnal hemoglobinuria. N Engl J Med 2006; 355(12):1233–1243.
18. Wlodarski M, Schade A, Maciejewski J. T-large granular lymphocyte leukemia: Cur-

rent molecular concepts. Hematology 2006; 11(4):245–256.
19. Saunthararajah Y, Molldrem JL, Rivera M, et al. Coincident myelodysplastic syn-
drome and T-cell large granular lymphocytic disease: Clinical and pathophysiological
features. Br J Haematol 2001; 112(1):195–200.
20. Plasilova M, Risitano A, Maciejewski JP. Application of the molecular analysis of
the T-cell receptor repertoire in the study of immune-mediated hematologic diseases.
Hematology 2003; 8(3):173–181.
21. Wang H, Chuhjo T, Yasue S, et al. Clinical significance of a minor population of
paroxysmal nocturnal hemoglobinuria-type cells in bone marrow failure syndrome.
Blood 2002; 100(12):3897–3902.
22. Slaughter DP, Southwick HW, Smejkal W. Field cancerization in oral stratified squa-
mous epithelium; clinical implications of multicentric origin. Cancer 1953; 6(5):963–
968.
23. Braakhuis BJ, Tabor MP, Kummer JA, et al. A genetic explanation of Slaughter’s
concept of field cancerization: Evidence and clinical implications. Cancer Res 2003;
63(8):1727–1730.
24. Okamoto M, Shichishima T, Noji H, et al. High frequency of several PIG-A mutations
in patients with aplastic anemia and myelodysplastic syndrome. Leukemia 2006;
20(4):627–634.
25. Tichelli A, Gratwohl A, Wursch A, et al. Late haematological complications in severe
aplastic anaemia. Br J Haematol 1988; 69:413–418.
26. Yue G, Hao S, Fadare O, et al. Hypocellularity in myelodysplastic syndrome is an
independent factor which predicts a favorable outcome. Leuk Res 2008; 32(4):553–
558.
Hematol 2000; 37(1):15–29.
28. Matsui WH, Brodsky RA, Smith BD, et al. Quantitative analysis of bone marrow CD34
cells in aplastic anemia and hypoplastic myelodysplastic syndromes. Leukemia 2006;
20(3):458–462.
29. Brodsky RA, Jones RJ. Aplastic anaemia. Lancet 2005; 365(9471):1647–1656.
30. Orazi A, Albitar M, Heerema NA, et al. Hypoplastic myelodysplastic syndromes can
be distinguished from acquired aplastic anemia by CD34 and PCNA immunostaining
of bone marrow biopsy specimens. Am J Clin Pathol 1997; 107(3):268–274.
31. Otawa M, Kawanishi Y, Iwase O, et al. Comparative multi-color flow cytometric
analysis of cell surface antigens in bone marrow hematopoietic progenitors between
refractory anemia and aplastic anemia. Leuk Res 2000; 24(4):359–366.
32. Elghetany MT, Hudnall SD, Gardner FH. Peripheral blood picture in primary hypocel-
lular refractory anemia and idiopathic acquired aplastic anemia: An additional tool
for differential diagnosis. Haematologica 1997; 82(1):21–24.
33. Maciejewski JP, Selleri C. Evolution of clonal cytogenetic abnormalities in aplastic
anemia. Leuk Lymphoma 2004; 45(3):433–440.
34. Parker JE, Mufti GJ. The role of apoptosis in the pathogenesis of the myelodysplastic
syndromes. Int J Hematol 2001; 73(4):416–428.
35. Epperson DE, Nakamura R, Saunthararajah Y, et al. Oligoclonal T cell expansion in
myelodysplastic syndrome: Evidence for an autoimmune process. Leuk Res 2001;
25(12):1075–1083.
36. Maciejewski JP, Hibbs JR, Anderson S, et al. Bone marrow and peripheral blood
lymphocyte phenotype in patients with bone marrow failure. Exp Hematol 1994;
22(11):1102–1110.
37. Kitagawa M, Saito I, Kuwata T, et al. Overexpression of tumor necrosis factor (TNF)-
alpha and interferon (IFN)-gamma by bone marrow cells from patients with myelodys-
plastic syndromes. Leukemia 1997; 11(12):2049–2054.
38. Gersuk GM, Beckham C, Loken MR, et al. A role for tumour necrosis factor-alpha,
Fas and Fas-Ligand in marrow failure associated with myelodysplastic syndrome. Br
J Haematol 1998; 103(1):176–188.
39. Selleri C, Maciejewski JP, Catalano L, et al. Effects of cyclosporine on hematopoietic
and immune functions in patients with hypoplastic myelodysplasia: In vitro and
in vivo studies. Cancer 2002; 95(9):1911–1922.
40. Lindberg EH. Strategies for biology and molecular-based treatment of myelodysplas-
tic syndromes. Curr Drug Targets 2005; 6(6):713–725.
41. Deeg HJ, Gotlib J, Beckham C, et al. Soluble TNF receptor fusion protein (etaner-
cept) for the treatment of myelodysplastic syndrome: A pilot study. Leukemia 2002;
16(2):162–164.
42. Raza A, Candoni A, Khan U, et al. Remicade as TNF suppressor in patients with
myelodysplastic syndromes. Leuk Lymphoma 2004; 45(10):2099–2104.
43. Platanias LC. Map kinase signaling pathways and hematologic malignancies. Blood
2003; 101(12):4667–4679.
44. Zhou L, Opalinska J, Verma A. p38 MAP kinase regulates stem cell apoptosis in
human hematopoietic failure. Cell Cycle 2007; 6(5):534–537.
45. Giafis N, Katsoulidis E, Sassano A, et al. Role of the p38 mitogen-activated protein
kinase pathway in the generation of arsenic trioxide-dependent cellular responses.
Cancer Res 2006; 66(13):6763–6771.
46. Copplestone JA, Mufti GJ, Hamblin TJ, et al. Immunological abnormalities in
myelodysplastic syndromes. II. Coexistent lymphoid or plasma cell neoplasms: A
report of 20 cases unrelated to chemotherapy. Br J Haematol 1986; 63(1):149–159.
47. Enright H, Jacob HS, Vercellotti G, et al. Paraneoplastic autoimmune phenomena in
patients with myelodysplastic syndromes: Response to immunosuppressive therapy.
Br J Haematol 1995; 91(2):403–408.
48. Castro M, Conn DL, Su WP, et al. Rheumatic manifestations in myelodysplastic
syndromes. J Rheumatol 1991 May;18(5):721–727.
49. Billstrom R, Johansson H, Johansson B, et al. Immune-mediated complications in
patients with myelodysplastic syndromes—Clinical and cytogenetic features. Eur J
Haematol 1995; 55(1):42–48.
50. Marisavljevic D, Kraguljac N, Rolovic Z. Immunologic abnormalities in myelodys-
plastic syndromes: Clinical features and characteristics of the lymphoid population.
Med Oncol 2006; 23(3):385–392.
51. Saif MW, Hopkins JL, Gore SD. Autoimmune phenomena in patients with myelodys-
plastic syndromes and chronic myelomonocytic leukemia. Leuk Lymphoma 2002;
(11):2083–2092.
52. Kiladjian JJ, Visentin G, Viey E, et al. Activation of cytotoxic T-cell receptor gam-
madelta T lymphocytes in response to specific stimulation in myelodysplastic syn-
dromes. Haematologica 2008; 93(3):381–319.
53. Vyas P, Sternberg A. Characterization of the hemopoietic defect in early stages of the
myelodysplastic syndromes. Advances in enzyme regulation 2006; 46:98–112.
in early (low-risk) myelodysplastic syndrome. Blood 2005; 106(9):2982–2891.
55. Maftoun-Banankhah S, Maleki A, Karandikar NJ, et al. Multiparameter flow cytomet-
ric analysis reveals low percentage of bone marrow hematogones in myelodysplastic
syndromes. Am J Clin Pathol 2008; 129(2):300–308.
56. Matarraz S, Lopez A, Barrena S, et al. The immunophenotype of different immature,
myeloid and B-cell lineage-committed CD34(+) hematopoietic cells allows discrimi-
nation between normal/reactive and myelodysplastic syndrome precursors. Leukemia
2008 Jun; 22(6):1175–83.
57. Molldrem JJ, Caples M, Mavroudis D, et al. Antithymocyte globulin for patients with
myelodysplastic syndrome. Br J Haematol 1997; 99(3):699–705.
58. Saunthararajah Y, Nakamura R, Wesley R, et al. A simple method to predict response
to immunosuppressive therapy in patients with myelodysplastic syndrome. Blood
2003; 102(8):3025–3027.
59. Molldrem JJ, Leifer E, Bahceci E, et al. Antithymocyte globulin for treatment of the
bone marrow failure associated with myelodysplastic syndromes. Ann Intern Med
2002; 137(3):156–163.
60. Kochenderfer JN, Kobayashi S, Wieder ED,et al. Loss of T-lymphocyte clonal domi-
nance in patients with myelodysplastic syndrome responsive to immunosuppression.
Blood 2002; 100(10):3639–3645.
61. Deeg HJ, Jiang PY, Holmberg LA, et al. Hematologic responses of patients with
MDS to antithymocyte globulin plus etanercept correlate with improved flow scores
of marrow cells. Leuk Res 2004; 28(11):1177–1180.
62. Bagby GC, Jr., Gabourel JD, Linman JW. Glucocorticoid therapy in the preleukemic
syndrome (hemopoietic dysplasia): Identification of responsive patients using in-vitro
techniques. Ann Intern Med 1980; 92(1):55–58.
sented in myelodysplastic syndrome and aplastic anemia and predicts a response
to immunosuppression in myelodysplastic syndrome. Blood 2002; 100(5):1570–
1574.
64. Sugimori C, Yamazaki H, Feng X, et al. Roles of DRB1 ∗1501 and DRB1 ∗1502 in
the pathogenesis of aplastic anemia. Exp Hematol 2007; 35(1):13–20.
65. Lim ZY, Killick S, Germing U, et al. Low IPSS score and bone marrow hypocellularity
in MDS patients predict hematological responses to antithymocyte globulin. Leukemia
2007; 21(7):1436–1441.
66. Broliden PA, Dahl IM, Hast R, et al. Antithymocyte globulin and cyclosporine A
as combination therapy for low-risk non-sideroblastic myelodysplastic syndromes.
Haematologica 2006; 91(5):667–670.
67. Sloand EM, Mainwaring L, Fuhrer M, et al. Preferential suppression of trisomy
8 compared with normal hematopoietic cell growth by autologous lymphocytes in
patients with trisomy 8 myelodysplastic syndrome. Blood 2005; 106(3):841–851.
68. Stadler M, Germing U, Kliche KO, et al. A prospective, randomised, phase II
study of horse antithymocyte globulin vs rabbit antithymocyte globulin as immune-
modulating therapy in patients with low-risk myelodysplastic syndromes. Leukemia
2004; 18(3):460–465.
69. Jonasova A, Neuwirtova R, Cermak J, et al. Cyclosporin A therapy in hypoplastic
MDS patients and certain refractory anaemias without hypoplastic bone marrow. Br
J Haematol 1998; 100(2):304–309.
70. Sekeres MA, List A. Immunomodulation in myelodysplastic syndromes. Best Pract

Res Clin Haematol 2006; 19(4):757–767.
71. Steensma DP, Mesa RA, Li CY, et al. Etanercept, a soluble tumor necrosis fac-
tor receptor, palliates constitutional symptoms in patients with myelofibrosis with
myeloid metaplasia: Results of a pilot study. Blood 2002; 99(6):2252–2254.
72. Hultsch T, Martin R, Hohman RJ. The effect of the immunophilin ligands rapamycin
and FK506 on proliferation of mast cells and other hematopoietic cell lines. Mol Biol
Cell 1992; 3(9):981–987.
73. Hosoi H, Dilling MB, Shikata T, et al. Rapamycin causes poorly reversible inhibition
of mTOR and induces p53-independent apoptosis in human rhabdomyosarcoma cells.
Cancer Res 1999; 59(4):886–894.
74. Honma Y, Ishii Y, Kasukabe T, et al. Induction of differentiation of human myeloid
leukemia cells by novel synthetic neurotrophic pyrimidine derivatives. Exp Hematol
2001; 29(2):194–201.
75. Platzbecker U, Haase M, Herbst R, et al. Activity of sirolimus in patients
with myelodysplastic syndrome—Results of a pilot study. Br J Haematol 2005;
128(5):625–630.
76. Navas TA, Mohindru M, Estes M, et al. Inhibition of overactivated p38
MAPK can restore hematopoiesis in myelodysplastic syndrome progenitors. Blood
2006;108(13):4170–4177.
77. Steensma DP, Mesa RA, Reeder TL, et al. Effects of the MEK inhibitor CI-1040 (PD
184352) on progenitor growth from normal and myelodysplastic marrow. Haemato-
logica 2003; 88(9):1072–1074.
78. Salituro FG, Germann UA, Wilson KP, et al. Inhibitors of p38 MAP kinase: Therapeu-
tic intervention in cytokine-mediated diseases. Curr Med Chem 1999; 6(9):807–823.
79. Nikas SN, Drosos AA. SCIO-469 Scios Inc. Curr Opin Investig Drugs 2004;
5(11):1205–1212.
80. Hideshima T, Podar K, Chauhan D, Ishitsuka K, Mitsiades C, Tai YT, et al. p38
MAPK inhibition enhances PS-341 (bortezomib)-induced cytotoxicity against multi-
ple myeloma cells. Oncogene 2004; 23(54):8766–8776.
81. Barrett J, Rezvani K. Neutrophil granule proteins as targets of leukemia-specific
immune responses. Curr Opin Hematol 2006; 13(1):15–20.
82. Qazilbash M, Wieder E, Rios R. Vaccination with the PR1 leukemia-associated anti-
gen can induce complete remission in patients with myeloid leukemia. Blood 2004;
104(Suppl.1 A):259.
83. Kawakami M, Oka Y, Tsuboi A, et al. Clinical and immunologic responses to very
low-dose vaccination with WT1 peptide (5 microg/body) in a patient with chronic
myelomonocytic leukemia. Int J Hematol 2007; 85(5):426–429.
84. Catalano L, Selleri C, Califano C, et al. Prolonged response to cyclosporin-A in
hypoplastic refractory anemia and correlation with in vitro studies. Haematologica
2000; 85(2):133–138.
85. Asano Y, Maeda M, Uchida N, et al. Immunosuppressive therapy for patients with
refractory anemia. Ann Hematol 2001; 80(11):634–638.
86. Atoyebi W, Bywater L, Rawlings L, et al. Treatment of myelodysplasia with oral
cyclosporin. Clinical and laboratory haematology 2002; 24(4):211–214.
87. Steensma DP, Dispenzieri A, Moore SB, et al. Antithymocyte globulin has limited
efficacy and substantial toxicity in unselected anemic patients with myelodysplastic
syndrome. Blood 2003; 101(6):2156–2158.
88. Yazji S, Giles FJ, Tsimberidou AM, et al. Antithymocyte globulin (ATG)-based
therapy in patients with myelodysplastic syndromes. Leukemia 2003; 17(11):2101–
2106.
89. Killick SB, Mufti G, Cavenagh JD, et al. A pilot study of antithymocyte globulin
(ATG) in the treatment of patients with ‘low-risk’ myelodysplasia. Br J Haematol
2003; 120(4):679–684.
90. Shimamoto T, Tohyama K, Okamoto T, et al. Cyclosporin A therapy for patients
with myelodysplastic syndrome: Multicenter pilot studies in Japan. Leuk Res 2003;
27(9):783–788.
91. Dixit A, Chatterjee T, Mishra P, et al. Cyclosporin A in myelodysplastic syndrome:
A preliminary report. Ann Hematol 2005; 84(9):565–568.
92. Chen S, Jiang B, Da W, et al. Treatment of myelodysplastic syndrome with cyclosporin
A. Int J Hematol 2007; 85(1):11–17.
19
Lenalidomide Therapy in MDS
Rami Komrokji and Alan F. List

H. Lee Moffitt Cancer and Research Institute, Tampa,
Florida, U.S.A.
Aristoteles Giagounidis
St. Johannes Hospital, Medizinische Klinik II, An der Abtei 7–11,
Duisburg, Germany
INTRODUCTION
The last decade, generally, witnessed major advancements in understanding the
pathobiology of myelodysplastic syndromes (MDS) and the 5q− syndrome in par-
ticular (see chaps. 3 and 12) (1–6). For the first time, three medications for various
forms of MDS were developed and approved by the Food and Drug Administra-
tion (FDA): 5-azacytidine (azacitidine), 5-aza-2 -deoxycytidine (decitabine), and
lenalidomide.
The demethylating agents (azacitidine and decitabine) and lenalidomide
represent different strategies, targeting MDS from different pathobiological angles
(see chap. 17). Lenalidomide, now known as immunomodulator drugs (IMiDs),
is a second-generation drug in class; thalidomide is the lead compound of the first
generation. In this chapter, we will discuss in details the role of IMiDs in treatment
of MDS.
457
458 Komrokji et al.
Figure 1 Deletion 5q in MDS. Source: Adapted from Ref. 3.
THE EVOLVING STORY OF IMiDs

The concept of IMiDs as a drug class with unique properties was evolved after
a long-term use of the parent compound thalidomide, and more recently, its ana-
logue lenalidomide. These agents earned this classification due to variety of bio-
logical effects they exert, including cytokine modification and immune-mediated
responses (7,8).
Thalidomide was originally developed from the glutamic acid derivative
␣-phthaloylisoglutamine in the 1950s in Germany as anti-epileptic medication,
but it lacked efficacy for this purpose (9). However, it was found to be sedating,
and the structure was similar to barbiturates (but it lacked many of the toxicities
of barbiturates, including overdose potential); this led to its marketing as sedative
and sleep aid. Thalidomide was approved in Germany in 1957 and in several other
countries, including Canada and Australia, shortly thereafter (8). Thalidomide was
found to have antiemetic properties in pregnancy, which led to an increased use in
the first trimester. By the early 1960, reports suggesting its teratogenic effects were
published (10). The drug was withdrawn from the market in 1961, but, tragically,
only after 5000 to 12,000 deformed born babies were reported in 46 countries
(11,12). Fortunately, the FDA did not approve the drug in the United States due to
neurotoxicity concerns (13).
Lenalidomide Therapy in MDS 459
In 1965, the first report of thalidomide activity in leprosy was published (14).
This was discovered because the drug was being used as a sedative for patients
with erythema nodusum leprosum (ENL). The drug was found to also reduce
symptoms and lesions associated with ENL. In 1998, the FDA approved the drug
in the United States for ENL (8).
Antitumor activity of thalidomide in multiple myeloma was first reported
in 1999 (15). Thalidomide was subsequently studied in many different types of
cancer, including MDS. In an effort to improve efficacy and reduce toxicity,
thalidomide analogues were developed using the backbone of thalidomide struc-
ture. 4-Amino analogues were developed in the mid-1990s, and lenalidomide
(CC-5013) and pomalidomide (CC-4047) were identified as the most promising
next-generation IMiDs.
Lenalidomide was tested in MDS and multiple myeloma. In 2005, the FDA
approved lenalidomide for use in transfusion dependent IPSS low or intermediate-
1 MDS patients with 5(q) deletion alone or with other cytogenetic abnormalities.
In the United States, in 2006, lenalidomide was also approved for use in relapsed
multiple myeloma, and thalidomide was also approved for use in newly diagnosed
multiple myeloma patients. Lenalidomide is currently being tested in other hema-
tological malignancies (e.g., lymphoproliferative disorders) with early promising
results. Third generation IMiDs are being developed (16).
PHARMACOKINETICS OF IMiDs
Thalidomide
Thalidomide [␣-(N-phthalimido)glutarimide] is a synthetic glutamic acid deriva-
tive. The empirical formula is C13 H10 N2 O4 (17). Thalidomide is formulated as
a racemic mixture of two active enantiomers, S(−) and R(+) [Fig. 2 (A)]. Ini-
tially, the S(−) isoform was thought to be associated with the teratogenic effects
and the R(+) isoform was responsible for the sedative properties. Purification
of the R(+) enantiomer, despite its relatively lower potency, was attempted in
order to improve drug safety, but was eventually not found to be technically
feasible secondary to the rapid interconversion of isomers under physiologic con-
ditions (7,16). Furthermore, eventually both forms were found to be teratogenic
(18,19).
Thalidomide is not soluble in water, so it is exclusively an oral agent. The
absorption through GI tract is slow with peak plasma concentration of 2.9 to 5.7
hours after a dose is administered. The mean plasma protein binding is 55% to
66%. The drug is metabolized by nonenzymatic hydrolysis to different metabolites.
Thalidomide and its metabolites are eliminated in the urine, with a half-life of 5
to 7 hours. The pharmacokinetic properties of thalidomide in the presence of
renal or hepatic dysfunction are largely unknown. Dosing with thalidomide varies
according to the disease and patient tolerance of adverse effects, typical doses
range from 50 to 800 mg daily (17).
460 Komrokji et al.
A O O O O
NH HN
N O O N
H H
O O
R-(+)-Thalidomide S-(−)-Thalidomide
B Thalidomide
O O
NH
N O
4 O
Phthaloyl ring
CC-5013 CC-4047
O O O O
NH NH
N O N O
NH2 NH2 O
Figure 2 Structure of (A) Thalidomide and (B) IMiDs lenalidomide and pomalidomide.
Source: Adapted from Ref. 16.
Lenalidomide
The second-generation IMiDs, lenalidomide and CC-4047, were developed by
modifying the thalidomide backbone. Both are 4-amino-gultaramide derivatives
of thalidomide in which an amino group was added to the fourth carbon of the
phthaloyl ring of the parent compound, and one of the carbonyls (C = O) of the
4-amino–substituted phthaloyl ring has been removed in lenalidomide [Fig. 2(B)]
(7,16).
The chemical name for lenalidomide is 3-(4-amino-1-oxo-1,3-dihydro-2H-
isoindol-2-yl)piperidine-2,6-dione (20). It also exists as a racemic mixture of R
and S forms. Lenalidomide is rapidly absorbed following oral administration with
maximum plasma concentrations occurring between 0.625 and 1.5 hours post-
dose. Co-administration with food does not alter the absorption but does reduce
the maximal plasma concentration by 36%. The drug is 30% bound to plasma
protein, and the exact metabolism is not well known. Two-thirds of lenalidomide
is excreted unchanged in the urine with a half-life of 3 hours (20).
The current approved initial dose of lenalidomide in MDS is 10 mg oral
daily. The effects of age, gender, and race on pharmacokinetics are not known.
Table 1 Recommended Dose Adjustments for Patients with Impaired Renal Function
Renal impairment Dose
Mild (80 ⬎ CLcr ≥ 50 mL/min) 10 mg (full dose) every 24 hr

Moderate (30 ≤ CLcr ⬍ 50 mL/min) 5 mg every 24 hr
Severe (CLcr ⬍ 30 mL/min, not requiring dialysis) 5 mg every 48 hr
End-stage renal disease (CLcr ⬍ 30 mL/min, requiring 5 mg 3 times a week after each
dialysis) dialysis
Recommendations based on a pharmacokinetic study by Chen et al. (57).

Abbreviation: CLcr , creatinine clearance.
Patients with renal impairment were excluded from clinical trials with lenalido-
mide. Dose adjustment for renal insufficiency has been suggested based on recent
pharmacokinetic data (Table 1).
PHARMACODYNAMICS OF IMiDs
The IMiDs exert a variety of biological effects including cytokine modification, T
cell activation, antiangiogenic, and antiapoptotic effects (Fig. 3); which of these,
Thal/IMids
Inhibition of proliferation Thal/IMids
induction of cell cycle Inhibition of
MM cells arrest/apoptosis adhesion:
VCAM ↓ Bone marrow
Thal/IMids
Selection ↓ Stromal cells
Cellular
interaction
T cell activation
Bone marrow Stimulation of and proliferation
Inhibition of Blood vessels production
expression Thal/IMids
IL-2 ↑
IVEGG ↑ IFN-γ ↑
IL-6 ↑
TNF-α ↑ Inhibition of Activation
anglogenesis Th-1 cells
Thal/IMids
NK cells
Release of
cytotoxic
mediators
MM cells lysis
Figure 3 Variety of biological effects of IMiDs in multiple myeloma. Source: Adapted

from Ref. 8.
462 Komrokji et al.
if any, are responsible for the clinical effects of thalidomide and lenalidomide in
MDS is unknown.
Cytokine Modification
Thalidomide and its analogues modulate cytokine production, with inhibitory
effects in inflammation and stimulatory immune–mediated effects. The specific
effect is based on the cell type and the triggering stimuli (21). Originally, Thalido-
mide was shown to inhibit TNF-␣ production from monocytes (22). The inhibitory
effect is via degradation of TNF-␣ mRNA (8,23). Thalidomide can also inhibit
TNF-␣ from other cells (24). The effect of thalidomide on TNF-␣ is dependent
on the stimuli where, as mentioned, it inhibits TNF-␣ production from monocytes
due to microbial stimuli or lipopolysaccharide stimuli; however, through T cell
co-stimulation, it can increase the cytokine production (8,16,21,25).
The second-generation IMiDs are more potent TNF-␣ modulators. Lenalido-
mide and pomalidomide (CC-4047) are 100 to 50,000 times more potent TNF-␣
inhibitors from monocytes in response to lipopolysaccharide trigger (26,27).
Lenalidomide and pomalidomide (CC-4047) increased TNF-␣ production
more than thalidomide through T cell co-stimulation from both CD4 and CD8
cells (28).
IMiDs have similar modulatory effects on other cytokines including inter-
leukin 12 (IL-12), IL-6, IL-1␤, IL-6, and GMCSF (8). For instance, IL-12 produc-
tion is suppressed by IMiDs in lipopolysaccharide-stimulated monocytes, while
it is potentiated through T cell co-stimulation. This may play a role in the future
using this class of drugs as adjunct to vaccine therapy (8,29,30).
Immune Effects
T cells are activated by two steps. The first occurs when antigen presenting cells
(APC) present major histocomptability complex (MHC)-bound peptides to the T
cell receptor (TCR). A second signal via the interaction of B7 molecule on APC
and CD28 on T cell is needed. Activated T cells release cytokines enhancing
further T cell stimulation and proliferation. IMiDs facilitate activation of T cells
in the absence of a second signal and further enhance the response in the presence
of two signals. Lenalidomide and pomalidomide (CC-4047) are more potent co-
stimulators (16).
Thalidomide and its analogues co-stimulate CD8+ cytotoxic cells. Lenalido-
mide increased cell lysis via virus specific CD8+ cells for influenza and CMV
antigens (31). IMiDs co-stimulate CD4+ helper cells and alter the (milieu) bal-
ance between Th1 and Th2 cells (32). The effects of IMiDs were demonstrated
in vivo when pomalidomide (CC-4047) was shown to enhance anti-tumor activ-
ity of cancer vaccine in murine model. The enhancement was mainly Th-1 cell
mediated (33). Lenalidomide also enhances NK cell and T cell activity resulting
in lysis of multiple myeloma cells (16,34).
Antiangiogeneic Activity
The anti-angiogeneic activity of IMiDs were first recognized in an attempt to
explain the limb-bud malformations observed with thalidomide (35).
IMiDs decrease vascular endothelial growth factor (VEGF) and beta fibrob-
last growth factor (bFGF) (35,36). IMiDs inhibited new microvessel growth in rat
aorta model; lenalidomide was a more potent antiangiogeneic drug (37). They also
demonstrated antiangiogeneic activity evident by decreased microvessel density in
rat lymphoma model (38). In addition, lenalidomide decreased bFGF endothelial
cell migration via attenuating AKT phosphorylation (39). The above-mentioned
studies demonstrate antiangigogeneic activity of IMiDs in vitro and in vivo with
more potency observed with second generation.
Alteration of Microenvironment and Stroma

IMiDs-modify surface-cell adhesion molecules. Thalidomide was shown to reduce
VCAM-1, ICAM-1, E-selectin, and L-selectin adhesion molecules (40). IMiDs
interfere with adhesion of myeloma cells with the bone marrow stromal cells
decreasing VEGF and IL-6 (16,36).
Antitumor Activity
IMiDs have direct antitumor activity independent from immunomodulatory
effects. In myeloma cell lines and patient-derived myeloma cells, IMiDs demon-
strated inhibitory effects in doxorubicin, melphalan, and dexamethasone resistant
cells; they also demonstrated additive effects to dexamethasone. IMiDs inhib-
ited DNA synthesis and induced G1 growth arrest. IMiDs induced apoptosis of
myeloma cells (41). IMiDs activate caspase-8, enhance FAS-induced apoptosis
and increased TNF-related apoptosis inducing ligand (TRAIL). In addition, they
downregulate nuclear factor (NF)-␬ B activity (42). Lenalidomide demonstrated
antiproliferative activity against a 5q mutant cell line (MUTZ-1) (43).
ROLE OF THALIDOMIDE IN TREATMENT OF MDS

Thalidomide was the first IMiDs to be tested in MDS (though the term IMiD was
not widely used at the time clinical trials were conducted). Several clinical trials
examined the role of thalidomide in treatment of MDS (44–50) (Table 2). Overall,
one can notice that thalidomide had modest activity in treating MDS patients,
with substantial adverse events. In many cases, this is because studies attempted
to escalate to high-dose thalidomide, and due to the prolonged time required for a
response, large number of patients discontinued treatment because of side effects
before a response could be seen.
The largest thalidomide MDS study reported was by Azra Raza and col-
leagues (45), a protocol that enrolled 83 patients. The median age was 67 years—
77 patients had de novo MDS, while 6 patients had secondary MDS. Thirty-six
464 Komrokji et al.
Table 2 Major Studies of Thalidomide in MDS

No. of Dose Response rate Discontinued
Study patients (mg/day) (%) (ITT) treatment (%)
(45) 83 100–400 3 (19) 38

(46) 73 200–1000 25 (11) 56
(47) 53 200–800 59 (19) 64
(48) 40 100–300 32 (20) 37
(49) 34 100–500 65 (56) 15
(50) 13 100–300 62 (38) 38
Abbreviation: ITT, intention to treat analysis.

Source: Adapted from Ref. 44
patients had refractory anemia (RA), 13 refractory anemia with ringed sideroblasts
(RARS), 24 with refractory anemia with excess blasts (RAEB), 6 patients with
refractory anemia with excess blasts in transformation (RAEB-t), and 4 patients
with chronic myelomonocytic leukemia (CMML). Sixty-three patients were RBC-
transfusion dependent.
Fifty-one patients completed 12 weeks of therapy, while 32 patients dis-
continued treatment before 12 weeks. Only few patients were able to take the
400-mg dose, while majority were taking 150- to 200-mg dose. Sixteen patients
showed hematological improvement according to the International Working Group
(IWG) criteria (15 in erythroid series and 1 with platelet response). Among the 15
patients with erythroid response, 11 had major hematological improvement, while
4 patients had minor response. By intention to treat analysis, the overall response
rate was 19% (16 out of 83), while the overall response rate was 31% (16 out
of 51) in patients who continued therapy at least for 12 weeks. No cytogenetic
responses were seen. The median time to response was 12 weeks and median
duration of response was 306 days. Responders had lower percentage of blasts,
shorter duration of platelet transfusion, and higher pretreatment platelet count.
The major side effects included fatigue in 79% of patients, constipation in 71%,
shortness of breath in 54%, and fluid retention in 54%.
The second largest study of thalidomide in MDS was conducted by the
North Central Cancer Group (46). Seventy-three patients were accrued to this
study. Patients were divided into two groups, 43 patients in favorable group (IPSS
scores 0 or 1) and 30 patients in unfavorable group (IPSS 1.5–3.5). Five patients
were not evaluable. Median age was 73 and 71 years, respectively. Out of 39
patients in favorable group, 34 were RBC-transfusion dependent, and 19 out of 29
patients in the unfavorable group were RBC-transfusion dependent. The maximum
planned thalidomide dose was 1000 mg.
The median number of cycles received was 3, and the median dose of
thalidomide was 300 mg. Two patients in the favorable group had minor hema-
tological response and four patients in the unfavorable group had a response
(1 partial remission, 2 major hematological, and 1 minor hematological responses).

More than 50% of the patients discontinued treatment because of side effects or
other reasons. Around 62% of all the patients had grade 3 or 4 nonhematologic
side effects. The most frequent were fatigue (24%), infection (19%), neuropathy
(13%), shortness of breath (8%), and constipation (7%).
Cytogenetic responses on thalidomide therapy were reported by Strupp and
colleagues. They observed three cytogenetic responses out of 16 MDS patients
with chromosomal abnormalities (51). Two of the three patients were poor-risk
karyotypes (RA, 45, XY, −7; and RAEB-t, complex karyotype), but most impor-
tantly, two of these patients had involvement of the long arm of chromosome
5—one as a single abnormality (RA, 46, XX, del(5)(q22q33), and one as part of a
complex karyotypic abnormality [RAEB-t, 46, XY, del(2)(p13?4), del(5)(q13q33),
add(17)(p11), del(20)(q11)], indicating that IMiDs could have a special role in
del(5q) myeloid neoplasias.
Thalidomide use was a proof of concept that IMiDs may have a potential
role in treating patients with MDS. But the nature of the disease requiring long-
term treatment limited the use of thalidomide, since potential activity is offset
by significant side effects, especially with longer use. Nevertheless, thalidomide
paved the road for lenalidomide studies in MDS.
CLINICAL TRIALS OF LENALIDOMIDE IN MDS

Three consecutive clinical trials were done using lenalidomide for treatment of
MDS patients: MDS-001, MDS-002, and MDS-003 (52–54) (Table 3) summarizes
these clinical studies.
MDS-001
The MDS-001 study was the first clinical study to address the role of lenalido-
mide in treatment of MDS (52). The study was phase I/II single institution study.
Alan List and his colleagues reported the results of this study in the New England
Journal of Medicine in 2005. Eligibility criteria included established diagnosis of
de novo MDS according to FAB criteria of more than 3 month with symptomatic
anemia (defined as hemoglobin less than 10 g/dL) or transfusion dependent anemia
(defined as need for 4 units of RBC within 8 weeks before enrollment). Patients
must have no response to treatment with erythropoietin (EPO) or had an endoge-
nous serum level of more than 500 mU/mL. Patients with neutropenia less than
500/mm3 or platelets less than 10,000 (mm3 were excluded). Patients received 1
of 3 oral lenalidomide dosing schedules: 25 mg daily, 10 mg daily, and 10 mg per
day administered for 21 days of every 28-day cycle. Modified IWG criteria were
used to evaluate response.
Forty-three patients were enrolled on this study. The median age was 72
years; 25 patients (58%) were males. Out of the 43 patients, 20 (47%) had RA,
13 (30%) RARS, 8 (19%) RAEB, 1 (2%) RAEB-t, and 1 (2%) CMML. Accord-
ing to the IPSS risk classification 22 (51%) patients were low risk, 16 (37%)
Table 3 Lenalidomide in MDS Clinical Studies 466
MDS-001 MDS-002 MDS-003
No. of Patients 43 214 148

Eligibility De Novo MDS by FAB classification De Novo MDS De Novo MDS by FAB classification
criteria Symptomatic anemia (Hgb No 5q31 deletion 5q31 deletion that was either isolated or
⬍10 g/dL) or Low or int-1 IPSS risk with additional cytogenetic abnormalities
RBC transfusion dependence (4 units Transfusion dependent anemia (2 units Low or int-1 IPSS risk
within 8 wk) RBC/8 wk) Transfusion dependent anemia (2 units
No response to Epo or Serum EPO Platelets ⬎50,000/mm3 and ANC RBC/8 wk)
level ⬎500 ⬎500/mm3 Platelets ⬎50,000/mm3 and ANC
Adequate renal and hepatic function ⬎500/mm3
Treatment Lenalidomide 25 mg po daily, 10 mg Lenalidomide 10 mg po daily po for 21 Lenalidomide 10 mg po daily po for 21
po daily or 10 mg po for 21 days, days, each cycle repeated every 28 days, each cycle repeated every 28 days
each cycle repeated every 28 days days then amended to 10 mg po daily then amended to 10 mg po daily
Baseline Median age = 72 yr Median age = 72 yr Median age = 71 yr
characteristics FAB classification—no. (%) FAB classification—no. (%) FAB classification—no. (%)
RA 20 (47) RA 47 (22) RA 77 (52)
RARS 13 (30) RARS 86 (40) RARS 18 (12)
RAEB 8 (19) RAEB 24 (11) RAEB 30 (20)
RAEB-t 1 (2) RAEB-t 5 (2) CMML 3 (2)
CMML 1 (2) CMML 20 (9) AML 1 (1)
IPSS risk category—no. (%) AML 1 (1) aCML 3 (2)
Low 22 (51) inadequate specimen 31 (14) Inadequate specimen 16 (11)
Intermediate-1 16 (37) IPSS risk category—no. (%) IPSS risk category—no. (%)
Intermediate-2 4 (9) Low 92 (43) Low 55 (37)
High 1 (2) Intermediate-1 76 (36) Intermediate-1 65 (44)
Karyotype—no. (%) Intermediate-2 or high 8 (5) Intermediate-2 or high 8 (5)
Komrokji et al.
Normal: 23 (53) Unclassified 38 (18) Unclassified 20 (14)
Abnormal: 20 (47) Karyotype—no. (%) Karyotype—no. (%)
Patients with del(5)(q31.1) 12 (28) Normal: 167 (76) Isolated 5q 110 (74)
Abnormal: 47 (24) 5q+ additional abn 37 (25)
Response Overall erythroid response 24 (56%) Overall erythroid response 93 (43%) Overall erythroid response 112 (76%)
Major 21 (49%) Transfusion independence (TI) 56 TI 99 (67%)
Minor 3 (7%) (26%) ≥50% Transfusion reduction 13 (9%)
Erythroid response by karyotype ≥50% Transfusion reduction 37 TI frequency by karyotype complexity
Del(5)(q31.1) only 10 (83%) (17%) Isolated del(5q) 79 (72%)
Normal karyotype 13 (57%) TI by IPSS risk Del(5q) + 1 additional 12 (48%)
Cytogenetic responses Low 31 (34%) Complex (⬎30) 8 (67%)
Lenalidomide Therapy in MDS
Overall cytogenetic response 11 (55%) Intermediate-1 23 (30%) Overall cytogenetic response 62/85
Overall cytogenetic response in High risk 0 (73%)
del(5)(q31.1) 10 (83%) TI by Karyotype Complete response 38/85 (45%)
Complete response 9 (75%) Normal 42 (26%) Cytogenetic response by karyotype
Abnormal 13 (28%) complexity (85 evaluable pts)
Del(5q) (n = 64) 49 (77%)
Del(5q) + 1 (n = 15) 10 (67%)
Complex karyotype (n = 6) 3 (50%)
Time to response Median time to response 9–11.5 wk Median time to response 4.8 wk Median time to response 4.6 wk
Median time to cytogenetic response 8 wk Median Hgb increase 3.2 g/dL Median Hgb increase 5.4 g/dL
Median duration of response ⬎48 wk Median duration of response 41 wk Median duration of response 115 wk
Major adverse Any grade ⬎10% Grade 3 or 4 Grade 3 or 4
events [no. of Neutropenia 28 (65) Neutropenia 25% Neutropenia 81 (55)
patients (%)] Thrombocytopenia 32 (74) Thrombocytopenia 20% Thrombocytopenia 65 (44)
Pruritus 12 (28) Rash 4% Anemia 10 (7)
Diarrhea 9 (21) Rash 9 (6)
Urticaria 6 (14)
467
468 Komrokji et al.
intermediate-1, 4 (9%) intermediate-2, and 1 (2%) high risk. Thirty-two patients

(74%) were RBC transfusion dependent, 12 (28%) were neutropenic, and 10 (23%)
were thrombocytopenic. Twenty-three patients (53%) had normal karyotype and
20 (47%) had abnormal karyotype. Twelve patients had deletion 5q. All had either
not responded to prior treatment with EPO (77%) or had a low probability of EPO
response based on high transfusion burden and serum EPO concentration (23%).
Overall, 30% had not responded to thalidomide therapy.
The overall response rate by intention to treat analysis was 56% (24 patients
out of 43). Twenty-one patients (49%) achieved major erythroid response. Out of
transfusion dependent patients 63% (20 of 32 patients) achieved transfusion inde-
pendence. Median time to response ranged from 9 to 11.5 weeks in dose-dependent
fashion. The observed erythroid responses were durable, with the median duration
of major response not reached after a median follow-up of 81 weeks. Cytogenetics
correlated significantly with the hematologic response where 83% of patients with
a deletion of 5q31.1 had a response, as compared with 57% of those with a normal
karyotype and 12% of those with other cytogenetic abnormalities (p = 0.007). The
FAB category, IPSS, age, duration of disease, and number of previous treatments
had no significant correlation with response. The median time to a response was
shorter in patients with a deletion of 5q31.1. Remarkably, for cytogenetic responses
defined as 50% or more, reduction in abnormal metaphases were seen. Among
20 patients with abnormal karyotype, 11 patients achieved cytogenetic response.
Out of 12 patients with deletion 5q31.1 abnormality, 10 (83%) had a cytogenetic
response and 9 (75%) had a complete cytogenetic response. All patients with
a cytogenetic response also had a hematological response. The median time to
cytogenetic response was 8 weeks.
Correlative studies were also performed as part of MDS-001 (55). The
marrow plasma concentration of six cytokines, including TNF-␣, significantly
decreased in erythroid responders after 16 weeks of lenalidomide treatment. In the
erythroid responders, patients’ erythroid burst forming units (BFU-E) and multi-
lineage myeloid progenitors (CFU-GEMM) improved significantly compared to
nonerythroid responders (Fig. 4). The microvessel density measured by immuno-
histochemical staining was significantly reduced in patients with major erythroid
response, proliferation and apoptotic indices were not significantly changed. In
patients with del5(q) reduction of microvessel density and increase in apoptotic
indices was significant compared to non- 5q patients. The reduction of proliferation
was limited to non-5(q) patients.
MDS 003
This study was the confirmation of the striking activity of lenalidomide in patients
with 5q31 deletion (54). A total of 148 patients were enrolled on this study. Eligi-
bility criteria included confirmed diagnosis of MDS by FAB criteria with low- or
intermediate-risk IPSS, presence of 5q31 deletion either isolated or with additional
cytogenetic abnormalities, transfusion dependent anemia (more than 2 units RBC
in last 8 weeks). Patients with neutrophil count less than 500/mm3 , platelets less
350
BFUE Responders
Mean colony number × 105 MNC
300
BFUE Nonresponders
CFUGEMM Responders
250
CFUGEMM Nonresponders
200
150
100
50
Baseline Treated
Figure 4 Change in bone marrow progenitor recovery after lenalidomide. Mean BFU-E
and CFU-GEMM increased significantly in erythroid responders. Adapted from Ref. 55.
than 50,000/mm3 , proliferative CMML (WBC ⬎12,000/mm3 ), therapy-related

MDS, or known sensitivity to thalidomide were excluded. Treatment consisted
of 10 mg lenalidomide oral for 21 days every 28-day cycle but then amended to
10 mg po daily given shorter time to response observed on the pilot study.
The median age of patients was 71 years. More female patients, 97 (66%),
were enrolled. According to FAB classification, 77 patients (52%) were RA, 18
(12%) RARS, 30 (20%) RAEB, 3 (2%) CMML, 1 (1%) AML, 3 (2%) atypical
CML, and 16 (11%) had inadequate specimen. Majority of patients were low-
risk IPSS 55 (37%) and low intermediate-1 65 (44%), only 8 (5%) patients were
intermediate-2 or high risk, and 20(14%) were unclassified. An isolated 5q dele-
tion was found in 110 patients (74%), only 40 patients (27%) met the criteria
for the diagnosis of the 5q− syndrome. Thirty-seven patients (25%) had one or
more cytogenetic abnormalities in addition to del(5q). The median red blood cell
transfused in the 8 weeks prior to study was 6 units. Majority of patients were
treated previously with erythropoietin 108 (73%), and 58 (39%) had received
chemotherapy previously.
The overall response rate was 76% (112 patients). Ninety-nine patients
(67%) achieved transfusion independence and 13 patients (9%) had more than
50% reduction in transfusion. The median time to response was 4.6 weeks. The
median hemoglobin (Hgb) increased from baseline of 7.8 to 13.4 g/dL (5.4 g/dL
470 Komrokji et al.
median increase). After longer follow up approaching 4 years, the median time
of response was 115 weeks with 53% of the 99 responding patients having an
ongoing response (56). In multivariable analysis, thrombocytopenia and number
of RBC units transfused were the only factors correlated with response. Age, sex,
IPSS, FAB-type, and cytogenetics did not correlate with response.
Among the evaluable patients for cytogenetics, 64 patients had isolated 5q
deletion, 15 patients had del(5q) and one additional chromosomal abnormality,
and 6 patients had complex karyotype. Forty-nine patients (77%) with isolated
5q abnormality achieved cytogenetic response including 29 patients (45%) with
complete cytogenetic response. In patients with 5q deletion and one additional
abnormality, 10 patients (67%) had cytogenetic response with 6 (40%) complete
cytogenetic response. The cytogenetic response was 50% in patients with com-
plex karyotype. Although the sample size was small, no difference in cytogenetic
response was observed according to the complexity of cytogenetics. The cytoge-
netic responses were correlated with hematological responses.
Assessment of bone marrow by central review revealed that 36% of evaluable
patients had resolution of dysplastic features, 74% of patients with myeloblasts,
more than 5% had decrease in myeloblasts to below 5%, and 64% of patients had
reduction in ring sideroblasts. Sixteen patients had progression to more advanced
FAB-type or AML.
Based on those two discussed clinical studies, the FDA approved use of
lenalidomide for transfusion dependent IPSS low/int-1 risk MDS patients with a
5q deletion, with or without additional cytogenetic abnormalities.
MDS-002
The MDS-002 clinical trial examined the role of lenalidomide in MDS patients
lacking a deletion of chromosome 5q (53). A total of 214 patients were enrolled
on this protocol. Inclusion criteria included low- or int-1–risk MDS, transfusion
dependent anemia (more than 2 units in last 8 weeks), absence of 5q deletion,
neutrophil count more than 500/mm3 and platelets count more than 50,000/mm3 .
The median age of patients was 72 years, and 65% were males. According
to the FAB classification, 47 patients (22%) were RA, 86 (40%) RARS, 24 (11%)
RAEB, 20 (9%) CMML, 5 (2%) RAEB-t, 1 (1%) AML, and 31 (14%) had
inadequate specimen to classify subtype. The MDS subtypes were also classified
according to the WHO classification. Only 5% of the patients were classified
as RA, 16% as RCMD, 3% RARS, and 40% RARS-MD. Most cases of RAEB
(89%) were classified as RAEB-1 and all of CMML cases had ⬍10% blasts (i.e.,
CMML-1). Based on IPSS 92 (43%) were low risk, 76 (36%) int-1, 8 (4)% int-2
or high risk, and 38 (18%) unclassified. The median baseline RBC transfusion
was 4 units.
The overall response rate to lenalidomide was 43% according to IWG
2000 criteria, where 56 (26%) patients achieved transfusion independence and 37
(17%) patients had ≥50% reduction in transfusions. The median hemoglobin was
3.2 g/dL. The median time to response was 4.8 weeks and the median duration
of response was 41 weeks. Major neutrophil or platelets response was seen

in six patients. Baseline transfusion of less than 4 units, platelets more than
150,000/mm3 , and shorter duration of MDS diagnosis were the only significant
independent factors for the response in multivariable analysis. According to the
new IWG 2006 response criteria, 33% achieved hematological improvement either
by transfusion response or Hgb response.
Among 47 patients with cytogenetic abnormalities, 9 (19%) were cytoge-
netic responders (4 complete, 5 partial responses). The responding karyotypes
included trisomy 8 (n = 3), −Y (n = 3), deletion 11q (n = 2), and deletion 17p
(n = 1). All complete cytogenetic responses were associated with hematological
response, and 3 of 5 partial cytogenetic responders also attained RBC-transfusion
independence. Bone marrow central review revealed that 2 out 27 patients with
myeloblasts more than 5% had ⬎50% reduction in myeloblasts percentage. Ten
(5%) patients experienced disease progression.
The response rates observed in this study were considerably lower than in
patients with 5q31 deletion. However, lenalidomide induced clinical meaningful
responses in some patients who had limited treatment options. These clinical stud-
ies also suggest that there may be different mechanisms of action of lenalidomide
in 5q31 deletion MDS, where eradication or suppression of the clone seems to
be the major effect, compared to a more modest differentiation effect observed in
non-del(5q) patients.
CLINICAL SAFETY OF LENALIDOMIDE

Although lenalidomide yielded impressive erythroid responses in both del(5q)
and non-del(5q) MDS patients, it is not a totally benign drug. The clinical trials
have identified several safety measures that should be taken in consideration while
using this compound. These relate to the pharmacokinetics of the drug (due to its
predominantly renal excretion), to its myelosuppressive effect (predominantly in
del(5q) karyotypes), and its ability to induce nonhematologic adverse events. If
certain precautions are being taken, lenalidomide can be administered in a safe
and effective way.
Although clinical trials have shown that the overall adverse event burden is
not significantly higher in older patients, patients over the age of 65 experience
more serious adverse events than younger patients reflecting their overall reduced
physical function (20). Because most of the MDS patients eligible for lenalidomide
therapy will be ⬎65 years of age, they should be followed closely at least during
the first 8 weeks of lenalidomide administration to prevent complications due to
the initial phase of lenalidomide-induced cytopenias (see later).
Lenalidomide is renally excreted, and therefore, the risk of toxicity is
expected to be greater in patients with known renal impairment. In those patients,
it is prudent to adjust the dosage to their creatinine clearance (Table 1) (57). Also,
given that those patients are more likely to suffer from comorbidities affecting
renal function and may be taking concomitant medications reducing renal blood
472 Komrokji et al.
flow (e.g., nonsteroidal anti-inflammatory drugs, angiotensin-converting enzyme

inhibitors, etc.), close monitoring of renal function is advised during lenalidomide
treatment.
In patients on concomitant digoxin therapy, lenalidomide may increase
digoxin levels and periodical monitoring of digoxin levels are advisable. Lenalido-
mide does not interact with the cytochrome P450 system and can be co-adminis-
tered safely with other drugs metabolized by this route (20).
Other than its hematological effects, lenalidomide is usually well tolerated.
In MDS, the most common adverse events are neutropenia and thrombocytopenia,
rash, diarrhea, muscle cramps, and pruritus. Rare, but important, adverse events
include hypothyroidism and hypogonadism.
Neutropenia and Thrombocytopenia in Del(5q)

Neutropenia and thrombocytopenia are the most common adverse events that
have been observed in clinical trials (52–54). In del(5q) MDS patients with low-
or intermediate-1–risk disease, 55% experienced neutropenia of WHO grade III
or IV severity, and 44% suffered from thrombocytopenia of WHO grade III or
IV (54). These hematological adverse events tend to occur early in the treatment
course; 62% happen within the first 8 weeks of lenalidomide administration (54).
To reduce the frequency, severity, and consequences of these events, an
expert panel has recently published safety recommendations on the use of lenalido-
mide in MDS patients (58). These recommendations include a specific algorithm
for patients experiencing neutropenia (Fig. 5). It should be born in mind that
patients with del(5q) MDS often suffer from mild-to-moderate neutropenia at
diagnosis (5). This contributes to the fact that in those patients, WHO grade III or
IV neutropenia can occur abruptly even after a short time of exposure to lenalido-
mide, emphasizing the fact that all patients receiving lenalidomide should have
their full blood count checked every week for a minimum of 8 weeks.
Thrombocytopenia occurs also frequently in patients with del(5q) MDS
(44% WHO grade III and IV in lenalidomide MDS-003 trial). However, severe
bleeding has not been recorded in clinical trials as these protocols determined a
limit of 25,000/␮L to 50,000/␮L platelets at which treatment with lenalidomide
had to be temporarily interrupted. If these safety limits are observed, serious
bleeding will almost certainly be avoided. This, however, raises the question
of how to treat patients with significant thrombocytopenia at presentation. Such
thrombocytopenia, although atypical for the 5q− syndrome, may occur in patients
with additional chromosomal abnormalities or an increase in bone marrow blasts.
To date, no recommendation can be made for the treatment of patients with baseline
thrombocytopenia of ⬍50,000/␮L, as there is not enough evidence from clinical
studies that therapy with lenalidomide in this setting is safe and tolerable and,
whether long-term platelet transfusion dependence might be a consequence.
Eighty-four percent of patients in the del(5q) patient cohort needed dose
reductions for adverse events. An analysis of dose adjustment frequencies showed
Neutropenia
Neutrophils < 1000/µL Neutrophils < 500/µL
If possible, introduce G-CSF to

Interrupt lenalidomide therapy
prevent worsening of neutropenia
Monitor Neutrophils > 750/µL
Reintroduce/continue
lenalidomide at one lower-dose
level, if possible with G-CSF
support
Figure 5 Recommendations for the management of neutropenia in MDS patients treated

with lenalidomide. Source: Adapted from Ref. 58.
that dose reductions occurred less often in the syncopated regimen (21 days/28)
than in the continuous regimen (28 days/28). The frequencies were 91% and
67%, respectively, and reached statistical significance (p ⬍ 0.05). Given that there
was no difference in efficacy between the syncopated and the continuous dosing
schedule, it seems prudent to advise a syncopated dosing outside clinical trials.
The median time to first dose reduction was 22 days, emphasizing again the
necessity for a frequent outpatient visit schedule during the first 8 weeks. One in
five patients taking lenalidomide for del(5q) MDS will have to discontinue the
drug for intolerable adverse effects.
Neutropenia and Thrombocytopenia in Non-Del(5q)

In MDS patients with non-del(5q) karyotypes, the most common and the most
severe adverse events were also neutropenia (27%) and thrombocytopenia (22%),
but only about half as frequent than in del(5q) karyotypes (53). Also, there was
no significant difference between a continuous dosing schedule (28 days/28) and
a syncopated one (21 days/28). Fifty-five percent of patients required dose adjust-
ments with a median of 7 weeks to the first dose reduction. Given that four deaths
474 Komrokji et al.
Table 4 Nonhematological Adverse Events in Del(5q) and Non-Del(5q) Karyotypes of

MDS Patients with Low- or Intermediate-1–Risk Myelodysplastic Syndromes
Non-del(5q) MDS Del(5q) MDS

Adverse event All grades Grades III/IV All grades Grades III/IV
Pruritus 21 1 32 3
Rash 24 2 28 6
Diarrhea 16 1 24 3
Fatigue 18 4 12 3
Hypothyroidism NA NA 7 0
Numbers represent percentages.
are attributable to severe infections (one septic shock, one respiratory failure,
two pneumonias), and that 43% of grade III and IV hematologic adverse events
occurred during the first 8 weeks of therapy, the same precautions are advisable for
the treatment of non-del(5q) MDS karyotypes as outlined in the del(5q) paragraph.
Non-Hematological Adverse Events

Non-hematological adverse events are less pronounced and less common than
hematological adverse events, both in del(5q) and non-del(5q) MDS patients
(Table 4).
Unlike thalidomide, lenalidomide does not lead to peripheral neuropathy
or somnolence. Dry skin, rash, and pruritis occur regularly, and itching of the
scalp is a typical adverse event in lenalidomide-treated patients. However, these
adverse events are often self-limited and require only rarely any pharmacological
intervention. Unselective histamines are of some use, and in more severe cases, a
short course of low-dose corticosteroids (up to 14 days of 10 mg of prednisone)
can be helpful. Diarrhea may be more difficult to tackle, and in chronic cases, tem-
porary interruption of lenalidomide might be necessary. Some anecdotal reports
exist where the addition of lactase to the diet improved diarrhea in patients with
unknown lactose intolerance, as lenalidomide capsules contain small amounts of
lactose. Loperamide, pipaverium bromide, uzara root extract, and tincture opii
have been used to alleviate diarrhea in some patients, with varying efficacy. Mus-
cle cramps have occurred in some patients and are best treated with quinine
sulphate, where available. Magnesium supplementation has rarely been useful in
this setting. Hypothyroidism has been reported in about 7% of patients and is
almost exclusively of autoimmune nature. Patients should have their TSH levels
checked every other month to prevent clinically significant hypothyroidism. If
TSH levels rise significantly, hormone replacement therapy is indicated. Patients
losing their response to lenalidomide during treatment may have developed either
hypothyroidism or hypogonadism, and should be screened for both conditions.
Grade 3 or 4 venous thromboembolism (VTE) was observed in 3% of

patients in del(5q) karyotypes and 1% in non-del(5q) disease. Patients with
antecedents of VTE should be carefully monitored. It is probably prudent to
use low–molecular-weight heparin in such patients to prevent recurrent throm-
bosis. Although aspirin is effective in the prevention of thrombosis in multiple
myeloma, aspirin cannot be recommended alongside lenalidomide as lenalido-
mide frequently leads to grade 3 or 4 thrombocytopenia in patients with MDS with
and without del(5q). Since the incidence of VTE in patients with lenalidomide-
treated MDS is generally low (52), prophylaxis in patients without antecedents
of VTE is not recommended. If VTE does occur, it should be treated according
to standard protocols. Treatment with lenalidomide should be interrupted until
stable anticoagulation is achieved and then carefully reintroduced. Once VTE has
occurred during lenalidomide therapy, patients should remain on anticoagulation
therapy as lenalidomide treatment continues.
In summary, lenalidomide can be safely administered to both del(5q) and
non-del(5q) MDS patients, if some basic precautions are being taken. Patients
should be monitored closely for hematological adverse events for the first 8 weeks,
and G-CSF should be used if neutropenia of ⬍1000/␮L occurs. Platelet counts
down to 25,000/L may be tolerated, but lower levels will require treatment inter-
ruption. Nonhematological adverse events are frequent but seldom serious. Regu-
lar thyroid-stimulating hormone (TSH) monitoring is required. Finally, although
lenalidomide has not been shown to be teratogenic in the most sensitive animal
model for thalidomide-induced birth defects, every precaution must be taken to
prevent pregnancies during lenalidomide treatment. This is independent of the fact
whether the patient is male or female and includes two independent measures of
birth control, the prohibition to donate blood during lenalidomide therapy, and, of
course, the interdiction to share the medication with others.
NEW INSIGHTS IN THE MECHANISM OF ACTION OF LENALIDIOMIDE

The complementary action of lenalidomide on both the malignant clone and the
surrounding microenvironment distinguishes it from more selective therapeutics
such as the hematopoietic cytokines. Lenalidomide has direct anti-tumor effects
that are amplified by its inhibitory effects on the supportive stroma and its inherent
ability to augment tumor selective T-cell and NK cell immune responses (59,60).
In the MDS-001 trial, emergence of multiple lymphoid aggregates composed
of B- and T-lymphocytes was noted in 36% of marrow trephine biopsies after
lenalidomide treatment that was closely linked to erythroid response. Whether this
finding represents an immune response against the ineffective clone is unknown,
but is currently under investigation.
The concordance between cytogenetic and erythroid response in patients
with del(5q) MDS suggests that lenalidomide restores red blood cell produc-
tion in a karyotype dependent manner by eliminating ineffective myelodysplastic
clones. This is supported by the observation that dysplastic cytological features
476 Komrokji et al.
persist in non-del(5q) responders, indicating that clonal suppression is selective

for del(5q) and complemented by promotion of effective erythropoiesis in suscep-
tible non-del(5q) MDS progenitors (53,54). Companion biomarker studies per-
formed in the MDS-001 trial offer additional insight into the agent’s contrasting
karyotype-dependent mechanisms of action (55). A direct cytotoxic effect leading
to suppression of the deletion 5q MDS clone is supported by the marked rise in
apoptosis in major erythroid responders with del(5q) compared to non-del(5q)
responders (207% vs. 48%, respectively). In contrast, restoration of effective ery-
thropoiesis in the non-del(5q) patients was associated with proliferation arrest as
evidenced by a marked reduction in Ki67 labeling (−76%) compared to del(5q)
responders (−9%). In addition, lenalidomide significantly suppressed elabora-
tion of an array of inflammatory cytokines in bone marrow plasma, including
TNF-␣, IL-1␤, IFN-␣, IFN-␥ , SDF-1␤, and IL-2, in erythroid responders by 16
weeks of treatment. Consistent with lenalidomide’s action to suppress the del(5q)
malignant clone, reduction in medullary microvessel density was greatest in major
erythroid responders with the chromosome 5q deletion. These findings provide
added support for the duality of drug action in MDS, characterized by promotion
of erythropoiesis in non-del(5q) versus direct cytotoxicity in del(5q) MDS.
Although a single cellular target has not been identified, it accounts for
lenalidomide’s actions in MDS; recent investigations indicate that lenalidomide
inhibits either directly or indirectly, the CD45 protein tyrosine phosphatase, which
serves as a negative regulator the EPO-receptor signal (61). Myelodysplastic ery-
throid progenitors display an attenuated response to EPO stimulation that is asso-
ciated with impaired EPO-receptor activation of its primary transcriptional target,
Signal Transduction and Activator of Transcription (STAT)-5 (62). In non-del(5q)
erythroid cell lines and MDS erythroid precursors, lenalidomide enhances the
EPO-receptor/STAT5 signal by relieving CD45 repression of ligand-dependent
Jak2 and Lyn kinase mediated STAT5 activation (61). Moreover, these data sug-
gest that CD45 activity may be aberrantly upregulated in MDS erythroid progen-
itors and contribute to the recognized impairment in EPO-receptor signaling. In
CD34+ selected cells from normal marrow donors, treatment with lenalidomide
or its analogue, pomalidomide, induces the expansion of immature progenitors,
and in particular, erythroid bursts (59,63). Both IMiDs delay erythroid maturation
in vitro as measured by glycophoryn-A induction, while increasing the generation
of immature erythroids that are erythropoietin responsive with coincident induc-
tion of hemoglobin transcription, with potent induction of hemoglobin-F (63).
Taken together, these data indicate that lenalidomide and possibly other members
of the IMiD family have direct effects on non-del(5q) erythroid progenitors that
serve to promote erythropoiesis by enhancing erythropoietin receptor signaling
and consequent transcriptional response.
The genetic loss of heterozygosity in the critical deleted region (CDR) of
del(5q), which encompasses a 1.5 Mb region between 5q31 and 5q33, has pro-
vided a long-standing rationale to identify a select gene or cluster that accounts
for the pathogenesis of the “5q− syndrome” and the possible contrasting effects
of lenalidomide in this karyotypically distinct MDS subset. Investigation of the

transcriptional profile of del(5q) clones has identified a number of candidate
genes whose expression is significantly reduced and consistent with monoal-
lelic loss, including the tumor suppressor gene SPARC, RPS14—a component
of the 40 S ribosomal subunit, ␣-catenin [CTNNA1, and early growth response
gene-1 (EGR-1)] (64–66). Using a functional RNA interference screen to eval-
uate the phenotype arising from knock down of each of the 41 genes in the
del(5q) CDR, Benjamin Ebert and colleagues recently reported that it is the
RSP14 gene encoding a ribosomal protein component of the 40 S subunit, which
yields a block in erythroid differentiation with increased apoptosis while preserv-
ing megakaryocyte differentiation (see chaps. 3 and 12) (67). More importantly,
expression of lentivirus transduced RPS14 cDNA in primary bone marrow cells
from patients with 5q− syndrome was sufficient to rescue erythropoiesis, impli-
cating the RPS14 gene as a lead candidate in the pathogenesis of the del(5q)
hematologic phenotype. In clonogenic assays, lenalidomide inhibits the growth of
del(5q) erythroblasts without discernable effects on normal erythroid progenitors
(68).
Analysis of drug-induced changes in gene expression has shown that
lenalidomide upregulates expression of several genes, including induction of the
haplodeficient SPARC gene expression in del(5q) erythroblasts. Although the
SPARC gene product has antiproliferative and antiangiogenic effects, its precise
role, if any, in mediating the apoptotic response to lenalidomide in del(5q) MDS
remains unclear. Recent studies have implicated inhibition of two cell cycle regula-
tory phosphatases, that is, Cdc25 C and PP2 A-C␣ that are haplodeficient in del(5q)
MDS, as key targets underlying the karyotype dependent cytotoxicity of lenalido-
mide (69). Indeed, selective suppression of gene expression by shRNA promotes
selective sensitivity to lenalidomide-induced apoptosis in both a cell-line model
and primary MDS bone marrow cells with a normal karyotype. These findings
suggest that haploinsufficiency for critical drugable gene products may offer a
potential strategy for therapeutic exploitation in other types of malignancies, and
provide opportunities to optimize treatments for MDS.
CLINICAL USE OF LENALIDOMIDE IN MDS

Use of lenalidomide in MDS patients should be addressed in context of other avail-
able treatment options. The choice of treatment should be dictated by evidence-
based medicine, cost-effectiveness considerations, and awareness of potential side
effects in different patients with variable comorbidities. In every day practice,
other logistics including insurance coverage and social support may affect treat-
ment decisions. Algorithms should be used as broad-line guidelines for treatment
but yet individualized for each patient.
The first step in treatment decision is risk stratification. Here, the 1997
IPSS is currently most commonly utilized (70). In patients with lower risk disease
(low-risk and intermediate-1–risk IPSS categories), the principle goal of therapy
478 Komrokji et al.
is transfusion independence and offering patients better quality of life (QOL).

Patients with cytopenia(s) can be divided into three groups: (1) deletion 5q with
or without additional cytogenetic abnormalities, where lenalidomide will be the
treatment of choice (54); (2) patients with thrombocytopenia and/or neutropenia,
where demethylating agents could be the first choice option; and (3) patients with
symptomatic anemia without del(5q).
In the third group, the probability of response to erythroid stimulating agents
(ESA) dictates first choice of therapy (71). If the serum erythropoietin level is
≤500 mU/mL, a 6- to 8-week trial of an ESA with or without G-CSF is reasonable.
If the trial is successful, the patient can be continued on ESA. For patients with
serum erythropoietin level ⬎500 mU/mL or ESA failure/progression, options for
treatment include demethylating agents or lenalidomide. A small subset of young
patients with low-risk disease may benefit from underutilized immunosuppressant
therapy with ATG and cyclosporine (see chap. 19) (72). Although one may argue
that responses to lenalidomide were less robust in non-del(5q) deletion patients,
some clinically meaningful responses were seen in patients with limited treatment
options (53). The fact that lenalidomide is an oral medication with different side
effect profile may make it an appealing treatment option for certain patients.
Allogeneic stem cell transplant for candidate patients with low-risk group is an
option best used at time of disease progression, based on a 2004 Markovian model
analysis (see chap. 22) (73).
In patients with higher risk disease (intermediate-2 and high-risk IPSS cat-
egories), the outcomes are poor, and the goals of the treatment include delaying
disease progression and improving overall survival. Allogeneic stem cell transplant
is the treatment of choice for this group, if feasible; however, it is limited to small
percentage of patients because most are ineligible due to age and comorbidities.
For the majority of patients who are not a candidate for transplant, the first choice
treatment is a demethylating agent (74–76). Appropriate use of lenalidomide in
these patients has yet to be determined; further clinical trials are needed. In a
recent study presented in abstract form, 46 patients with intermediate-2 or high-
risk MDS patients with del5(q) were treated with lenalidomide 10 mg orally for 21
days every 28 days (77). The median age was 68 years and majority of patients had
5q with additional cytogenetic abnormalities. The overall hematological response
rate was 26%, 5 (12%) patients achieved CR and 2 (5%) patients had marrow CR.
Transfusion independence was achieved in 24% and overall cytogenetic response
was 19%. Isolated 5q deletion and platelets count ⬎100,000/mm3 were associ-
ated with better probability of CR or marrow CR. Treatment was associated with
significant hematological toxicity and 80% of patients required hospitalization.
The main challenge with many of the new medications for MDS is their very
high cost. Once efficacy and safety are established in clinical trials, then a new
medication has to be cost-effective to become widely adapted. A recent decision
analysis study evaluated cost-effectiveness of lenalidomide in patients with 5q
deletion (78). The study compared lenalidomide use to best supportive care (BSC)
utilizing growth factors and transfusion. The time horizon was 1 year and the
perspective was from health care payers in the United States. The endpoints of
the study were transfusion independence and QOL gains. Costs include drug cost,
transfusion costs, outpatient services, and costs related to complications. With
the assumptions of the model, the total annual cost for lenalidomide group was
estimated at $63,385, compared to $54,940 in the BSC group. The incremental
cost per transfusion-free year gained was $16,066 and incremental cost of QALY
gained was $35,050. Those results are within the $50,000 or less for QALY-
gained range that is considered by some analysts to be an important component
of cost-effective treatment.
FUTURE DIRECTIONS
The development of IMiDs in MDS has been an informative example of both
serendipity and rational translational research, where observations made in the
clinic are studied in the laboratory, and then the knowledge gained is brought back
into the clinic. Use of IMiDs has shed light on important biological aspects of
MDS; those findings, in turn, have guided our use of IMiDs. As we learn more
about the biology of MDS and gain new insights in the mechanism of action
of IMiDs, we will see further fine tuning of their role and better identification of
patients who will benefit from their use. Combination therapies are currently being
explored, such as lenalidomide with an ESA or lenalidomide with demethylating
agents.
REFERENCES
1. Bennett J, Komrokji R, Kouides P. The myelodysplastic syndromes. In: Abeloff
MD, Armitage JO, Niederhuber JE, eds. Clincial Oncology. New York: Churchill
Livingstone, 2004:2849–2881.
2. Bennett JM, Komrokji RS. The myelodysplastic syndromes: Diagnosis, molecular
biology and risk assessment. Hematology 2005; 10(Suppl. 1):258–269.
3. List AF. Evolving applications of lenalidomide in the management of anemia in
myelodysplastic syndromes. Cancer Control 2006; 13(Suppl.):12–16.
4. Giagounidis AA, Germing U, Wainscoat JS, et al. The 5q– syndrome. Hematology
2004; 9:271–277.
5. Giagounidis AA, Germing U, Haase S, et al. Clinical, morphological, cytogenetic, and
prognostic features of patients with myelodysplastic syndromes and del(5q) including
band q31. Leukemia 2004; 18:113–119.
6. Giagounidis AA, Germing U, Strupp C, et al. Prognosis of patients with del(5q) MDS
and complex karyotype and the possible role of lenalidomide in this patient subgroup.
Ann Hematol 2005; 84:569–571.
7. Melchert M, List A. The thalidomide saga. Int J Biochem Cell Biol 2007; 39:1489–
1499.
8. Teo SK. Properties of thalidomide and its analogues: Implications for anticancer
therapy. AAPS J. 2005; 7:E14–E19.
9. Stephens T, Brynner R, eds. Dark Remedy: The Impact of Thalidomide and Its Revival
as Vital Medicine. Cambridge, MA: Perseus Publishing, 2001.
480 Komrokji et al.
10. Lenz W, Knapp K. Thalidomide embryopathy. Arch Environ Health 1962; 5:100–105.
11. McBride WG. Thalidomide and congenital abnormalities. J Am Med Assoc 1961;
2:1358.
12. Mellin GW, Katzenstein M. The saga of thalidomide. Neuropathy to embryopathy,
with case reports of congenital anomalies. N Engl J Med 1962; 267:1238–1244
(conclusion).
13. Kelsey FO. Thalidomide update: Regulatory aspects. Teratology 1988; 38:221–
226.
14. Sheskin J. Thalidomide in the treatment of lepra reactions. Clin Pharmacol Ther 1965;
6:303–306.
15. Singhal S, Mehta J, Desikan R, et al. Antitumor activity of thalidomide in refractory
multiple myeloma. N Engl J Med 1999; 341:1565–1571.
16. Bartlett JB, Dredge K, Dalgleish AG. The evolution of thalidomide and its IMiD
derivatives as anticancer agents. Nat Rev Cancer 2004; 4:314–322.
17. Thalidomide Package Insert, Celgene Corporation, 2007.
18. Eriksson T, Bjorkman S, Roth B, et al. Intravenous formulations of the enantiomers of
thalidomide: Pharmacokinetic and initial pharmacodynamic characterization in man.
J Pharm Pharmacol 2000; 52:807–817.
19. Fabro S, Smith RL, Williams RT. Toxicity and teratogenicity of optical isomers of
thalidomide. Nature 1967; 215:296.
20. Celgene Corporation. Revlimid R
(Lenalidomide) 5 mg, 10 mg, 15 mg and 25 mg
Capsules Prescribing Information USA. Available from: http://www.revlimid.
com/pdf/REVLIMID PI.pdf. Accessed December 2007.
21. Crane E, List A. Immunomodulatory drugs. Cancer Invest 2005; 23:625–634.
22. Sampaio EP, Sarno EN, Galilly R, et al. Thalidomide selectively inhibits tumor
necrosis factor alpha production by stimulated human monocytes. J Exp Med 1991;
173:699–703.
23. Moreira AL, Sampaio EP, Zmuidzinas A,et al. Thalidomide exerts its inhibitory action
on tumor necrosis factor alpha by enhancing mRNA degradation. J Exp Med 1993;
177:1675–1680.
24. Deng L, Ding W, Granstein RD. Thalidomide inhibits tumor necrosis factor-alpha
production and antigen presentation by langerhans cells. J Invest Dermatol 2003;
121:1060–1065.
25. Wolkenstein P, Latarjet J, Roujeau JC, et al. Randomised comparison of thalidomide
versus placebo in toxic epidermal necrolysis. Lancet 1998; 352:1586–1589.
26. Muller GW, Corral LG, Shire MG, et al. Structural modifications of thalidomide
produce analogs with enhanced tumor necrosis factor inhibitory activity. J Med Chem
1996; 39:3238–3240.
27. Muller GW, Chen R, Huang SY, et al. Amino-substituted thalidomide analogs:
Potent inhibitors of TNF-alpha production. Bioorg Med Chem Lett 1999; 9:1625–
1630.
28. Marriott JB, Clarke IA, Dredge K, et al. Thalidomide and its analogues have distinct
and opposing effects on TNF-alpha and TNFR2 during co-stimulation of both CD4(+)
and CD8(+) T cells. Clin Exp Immunol 2002; 130:75–84.
29. Corral LG, Haslett PA, Muller GW, et al. Differential cytokine modulation and T cell
activation by two distinct classes of thalidomide analogues that are potent inhibitors
of TNF-alpha. J Immunol 1999; 163:380–386.
30. Haslett PA, Klausner JD, Makonkawkeyoon S, et al. Thalidomide stimulates T cell
responses and interleukin 12 production in HIV-infected patients. AIDS Res Hum
Retroviruses 1999; 15:1169–1179.
31. Haslett PA, Hanekom WA, Muller G, et al. Thalidomide and a thalidomide analogue
drug costimulate virus-specific CD8+ T cells in vitro. J Infect Dis 2003; 187:946–
955.
32. McHugh SM, Rifkin IR, Deighton J, et al. The immunosuppressive drug thalidomide
induces T helper cell type 2 (Th2) and concomitantly inhibits Th1 cytokine production
in mitogen- and antigen-stimulated human peripheral blood mononuclear cell cultures.
Clin Exp Immunol 1995; 99:160–167.
33. Dredge K, Marriott JB, Todryk SM, et al. Protective antitumor immunity induced
by a costimulatory thalidomide analog in conjunction with whole tumor cell vac-
cination is mediated by increased Th1-type immunity. J Immunol 2002; 168:4914–
4919.
34. Davies FE, Raje N, Hideshima T, et al. Thalidomide and immunomodulatory deriva-
tives augment natural killer cell cytotoxicity in multiple myeloma. Blood 2001;
98:210–216.
35. D’Amato RJ, Loughnan MS, Flynn E, et al. Thalidomide is an inhibitor of angiogen-
esis. Proc Natl Acad Sci U S A 1994; 91:4082–4085.
36. Gupta D, Treon SP, Shima Y, et al. Adherence of multiple myeloma cells to bone
marrow stromal cells upregulates vascular endothelial growth factor secretion: Ther-
apeutic applications. Leukemia 2001; 15:1950–1961.
37. Dredge K, Marriott JB, Macdonald CD, et al. Novel thalidomide analogues display
anti-angiogenic activity independently of immunomodulatory effects. Br J Cancer
2002; 87:1166–1172.
38. Lentzsch S, LeBlanc R, Podar K, et al. Immunomodulatory analogs of thalidomide
inhibit growth of Hs sultan cells and angiogenesis in vivo. Leukemia 2003; 17:41–44.
39. Dredge K, Horsfall R, Robinson SP, et al. Orally administered lenalidomide (CC-
5013) is anti-angiogenic in vivo and inhibits endothelial cell migration and Akt
phosphorylation in vitro. Microvasc Res 2005; 69:56–63.
40. Geitz H, Handt S, Zwingenberger K. Thalidomide selectively modulates the density
of cell surface molecules involved in the adhesion cascade. Immunopharmacology
1996; 31:213–221.
41. Hideshima T, Chauhan D, Shima Y, et al. Thalidomide and its analogs overcome drug
resistance of human multiple myeloma cells to conventional therapy. Blood 2000;
96:2943–2950.
42. Mitsiades N, Mitsiades CS, Poulaki V, et al. Apoptotic signaling induced by
immunomodulatory thalidomide analogs in human multiple myeloma cells: Ther-
apeutic implications. Blood 2002; 99:4525–4530.
43. Gandhi AK, Naziruddin S, Verhelle D, et al. Anti-proliferative activity of CC-5013 in
5q– myelodysplastic syndrome (MDS) and acute lymphocytic leukemia (ALL) cell
lines. J Clin Oncol (Meeting Abstracts) 2004; 22:6618.
44. Musto P. Thalidomide therapy for myelodysplastic syndromes: Current status and
future perspectives. Leuk Res 2004; 28:325–332.
45. Raza A, Meyer P, Dutt D, et al. Thalidomide produces transfusion independence in
long-standing refractory anemias of patients with myelodysplastic syndromes. Blood
2001; 98:958–965.
482 Komrokji et al.
46. Moreno-Aspitia A, Colon-Otero G, Hoering A, et al. Thalidomide therapy in adult

patients with myelodysplastic syndrome. A north central cancer treatment group phase
II trial. Cancer 2006; 107:767–772.
47. Bouscary D, Legros L, Tulliez M, et al. A non-randomised dose-escalating phase
II study of thalidomide for the treatment of patients with low-risk myelodysplastic
syndromes: The thal-SMD-2000 trial of the groupe francais des myelodysplasies. Br
J Haematol 2005; 131:609–618.
48. Musto P, Falcone A, Sanpaolo G, et al. Thalidomide abolishes transfusion-dependence
in selected patients with myelodysplastic syndromes. Haematologica 2002; 87:884–
886.
49. Strupp C, Germing U, Aivado M, et al. Thalidomide for the treatment of patients with
myelodysplastic syndromes. Leukemia 2002; 16:1–6.
50. Bortolheiro TC, Fonseca LG, Peres ALM,et al. Thalidomide therapy for MDS, one
year follow-up. Leuk Res 2002; 27(Suppl. 1, addendum), [abstract].
51. Strupp C, Hildebrandt B, Germing U, et al. Cytogenetic response to thalidomide
treatment in three patients with myelodysplastic syndrome. Leukemia 2003; 17:1200–
1202.
N Engl J Med 2005; 352:549–557.
53. Raza A, Reeves JA, Feldman EJ, et al. Phase 2 study of lenalidomide in transfusion-
dependent, low-risk, and intermediate-1 risk myelodysplastic syndromes with kary-
otypes other than deletion 5q. Blood 2008; 111:86–93.
55. List AF, Baker AF, Green S, et al. Lenalidomide: Targeted anemia therapy for
myelodysplastic syndromes. Cancer Control 2006; 13(Suppl.):4–11.
56. List AF, Dewald GW, Bennett JM, et al. Long-term clinical benefit of lenalidomide
(Revlimid) treatment in patients with myelodysplastic syndrome and chromosome
deletion 5q. ASH Annu Meet Abstr 2006; 108:251.
57. Chen N, Lau H, Kong L, et al. Pharmacokinetics of lenalidomide in subjects with
various degrees of renal impairment and in subjects on hemodialysis. J Clin Pharmacol
2007; 47:1466–1475.
58. Giagounidis AAN, Fenaux, P, Mufti G, et al. Practical recommendations on the use of
lenalidomide in the management of myelodysplastic syndromes. Ann Hematol 2008
May; 87(5):345–352.
59. Verhelle D, Corral LG, Wong K, et al. Lenalidomide and CC-4047 inhibit the prolif-
eration of malignant B cells while expanding normal CD34+ progenitor cells. Cancer
Res 2007; 67:746–755.
60. Hideshima T, Anderson KC. Molecular mechanisms of novel therapeutic approaches
for multiple myeloma. Nat Rev Cancer 2002; 2:927–937.
61. List AF, Estes M, Williams A, et al. Lenalidomide (CC-5013; Revlimid(R)) pro-
motes erythropoiesis in myelodysplastic syndromes (MDS) by CD45 protein tyrosine
phosphatase (PTP) inhibition. ASH Annu Meet Abstr 2006; 108:1360.
62. Hoefsloot LH, van Amelsvoort MP, Broeders LC, et al. Erythropoietin-induced acti-
vation of STAT5 is impaired in the myelodysplastic syndrome. Blood 1997; 89:1690–
1700.
63. Moutouh-de Parseval LA, Verhelle D, Glezer E, et al. Pomalidomide and lenalidomide
regulate erythropoiesis and fetal hemoglobin production in human CD34+ cells. J
Clin Invest 2008; 118:248–258.
in patients with the 5q– syndrome. Br J Haematol 2007; 139:578–589.
65. Liu TX, Becker MW, Jelinek J, et al. Chromosome 5q deletion and epigenetic suppres-
sion of the gene encoding alpha-catenin (CTNNA1) in myeloid cell transformation.
Nat Med 2007; 13:78–83.
110:719–726.
67. Ebert BL, Pretz J, Bosco J, et al. Identification of RPS14 as the 5q– syndrome gene
by RNA interference screen. ASH Annu Meet Abstr 2007; 110:1.
5q– syndrome patients. Proc Natl Acad Sci U S A 2007; 104:11406–11411.
69. Wei S, Rocha K, Williams A, et al. Gene dosage of the cell cycle regulatory phos-
phatases Cdc25 C and PP2 A determines sensitivity to lenalidomide in del(5q) MDS.
71. Hellstrom-Lindberg E, Gulbrandsen N, Lindberg G, et al. A validated decision model
for treating the anaemia of myelodysplastic syndromes with erythropoietin + granu-
locyte colony-stimulating factor: Significant effects on quality of life. Br J Haematol
2003; 120:1037–1046.
72. Molldrem JJ, Leifer E, Bahceci E, et al. Antithymocyte globulin for treatment of the
bone marrow failure associated with myelodysplastic syndromes. Ann Intern Med
2002; 137:156–163.
risk myelodysplasia is associated with improved outcome. Blood 2004; 104:579–585.
leukemia group B. J Clin Oncol 2002; 20:2429–2440.
myelodysplastic syndromes. Results of a phase III randomized study. Cancer 2006;
106:1794–1803.
76. Lyons RM, Cosgriff T, Modi S, et al. Results of the initial treatment phase of a
study of three alternative dosing schedules of azacitidine (Vidaza(R)) in patients with
myelodysplastic syndromes (MDS). ASH Annu Meet Abstr 2007; 110:819.
77. Burcheri S, Prebet T, Beyne-Rauzy O, et al. Lenalidomide (LEN) in INT 2 and high
risk MDS with del 5q. Interim results of a phase II trial by the GFM. ASH Annu Meet
Abstr 2007; 110:820.
78. Goss TF, Szende A, Schaefer C, et al. Cost effectiveness of lenalidomide in the
treatment of transfusion-dependent myelodysplastic syndromes in the United States.
Cancer Control 2006; 13(Suppl.):17–25.
20
DNA Methyltransferase Inhibitor
Therapy in the Treatment of
Steven D. Gore
Sidney Kimmel Comprehensive Cancer Center at Johns Hopkins, Baltimore,
Maryland, U.S.A.
INTRODUCTION
In the past decade, the science of epigenetics has seen explosive growth. Epige-
netics refer to heritable changes in transcription patterns which are not due to
structural defects in DNA such as mutations and deletions. An increasing number
of important epigenetic alterations have been identified in human cells. The best
characterized include the methylation of cytosine residues in CpG dinucleotide
sequences. Methylation of such cytosines in CpG-rich regions of gene promot-
ers, referred to as CpG islands, predicts for transcriptional silencing of the gene.
Changes in chromatin conformation effect the transcriptional silencing; chromatin
conformation depends importantly on post-translational modifications of lysine
residues in histone proteins (Fig. 1). The acetylation status of the lysine residues
derives from local action of histone acetyltransferases and histone deacetylases
(HDACs). Methylated promoters recruit HDACs as the part of transcriptional
inhibitory complexes targeted by specific methyl–binding proteins. A wide vari-
ety of other epigenetic modifications contribute to this highly complex system,
including histone lysine methylation, phosphorylation, ubiquitination, and sumoy-
lation [reviewed in (1,2)].
While epigenetic gene silencing mediates physiological transcriptional con-
trol such as the silencing of alleles in imprinted genes and during X-chromosome
485
486 Gore
Figure 1 Relationship between DNA methylation and chromatin compaction. In this

simplified conceptual illustration, transcriptionally active open chromatin (euchromatin)
features lysine acetylation on the histone tails. Cytosine methylation (beige beads) recruits
methyl-binding proteins (green) which in turn recruit transcriptional corepressors includ-
ing HDACs (red beads). The resulting lysine deacetylation leads in part to compaction of
the chromatin and transcriptional silencing. Blue cylinders represent nucleosomes. Source:
Courtesy of Keith Robertson, Department of Molecular Biology and Biochemistry, Uni-
versity of Florida College of Medicine, Gainesville, Florida, U.S.A.
inactivation, promoter CpG dinucleotides appear to be protected from methyla-

tion in most somatic cells. Cancer cells demonstrate markedly abnormal epigenetic
marks; epigenetic silencing may represent a more frequent pathway to gene silenc-
ing than mutation or deletion (1,3). Because these epigenetic marks can poten-
tially be reversed, and because many of these marks are cancer-specific, epigenetic
DNA Methyltransferase Inhibitor Therapy in the TMS 487
gene silencing has become an attractive therapeutic target in experimental cancer

therapeutics.
DEVELOPMENT OF AZACYTOSINE NUCLEOSIDE ANALOGUES

The azacytosine nucleoside analogues were synthesized in the 1960s and were
studied as high-dose antimetabolites for the treatment of acute leukemias (4,5).
5AC failed to receive FDA approval, and further development of this nucleoside
fell under the aegis of the Cancer Therapy Evaluation Program (CTEP) of the
US National Cancer Institute. In vitro, these nucleosides are incorporated into
DNA in lieu of cytosine residues, whereupon they form irreversible adducts with
DNMT, the enzyme required for transmitting the pattern of cytosine methylation
from parent to daughter strand of DNA. The functional depletion of DNMT by the
azanucleosides leads to gradual reversal of methylation over several cell divisions
(6–8).
Administration of lower doses of 5AC as putative differentiation therapy to
patients with MDS was modeled after trials demonstrating efficacy in the induction
of fetal hemoglobin expression in patients with sickle cell anemia (9,10). While
these early trials in hemoglobinopathies included a variety of dose schedules, no
formal pharmacodynamically oriented dose finding studies were performed. Not
commonly appreciated, these early investigators demonstrated significant activity
of 5AC when administered orally along with the cytidine deaminase inhibitor
tetrahydrouridine (10).
Many assume that DNMT inhibitors represent the result of rational drug
development, and that epigenetic abnormalities must underpin the pathophysiol-
ogy of MDS. While the latter remains a strong possibility, the development of these
nucleosides in MDS resulted from empiric development as putative “differenti-
ation therapy.” The successful development of 5AC in MDS must be attributed
to Lewis Silverman, who championed this drug through early Phase trials at
Mt. Sinai Medical Center in New York, and subsequently in the Cancer and
Leukemia Group B (CALGB) (11–14). Dr. Silverman and colleagues developed
the first robust criteria for hematologic responses in MDS and demonstrated clin-
ical activity of intravenously and subsequently subcutaneous administration of
5AC. They described clinical behavior of this compound, which applies equally
well to decitabine—administration of the azacytosine nucleosides often worsens
cytopenias early in the treatment course, and a median of three cycles of adminis-
tration must be administered before clinical responses are noted. Silverman led the
seminal CALGB9221 study in which patients with MDS and chronic myelomono-
cytic leukemia (CMMoL) received 5AC in the currently labeled dose schedule, or
were observed with best supportive care; patients progressing on best supportive
care crossed over to 5AC (13). Reanalysis of these data using the more rigorous
International Working Group (IWG2000) clinical response criteria (15) demon-
strated an overall hematologic response rate of 47% including 10% complete
responses. Importantly, analysis of the impact of 5AC on time to acute myeloid
488 Gore
leukemia (AML) (defined by 30% blasts) or death revealed an approximately

1-year prolongation in the group initially assigned to 5AC compared to the control
arm (including patients crossed over to 5AC). A companion quality-of-life study
showed improvement in several quality-of-life parameters in patients receiving
5AC (16). The CALGB9221 data served as the basis for the FDA approval of
5-azacitidine.
The cross-over design of the CALGB9221 study precluded a rigorous anal-
ysis of the impact of 5AC on patients’ survival. However, a multicenter trial
in which patients with high-risk MDS (IPSS intermediate-2 or high), dysplastic
CMMoL, and AML with trilineage dysplasia (who formerly met FAB criteria for
RAEB-t) with survival as the primary endpoint has now been completed. In this
trial, physicians preselected one of three control treatments as standard therapy
for each individual patient prior to randomization. The control treatment could
be supportive care only, low-dose cytarabine, or cytarabine–anthracycline AML
induction therapy. Following selection of the control arm, patients were random-
ized for receiving that control treatment or 5AC. Preliminary data were presented
at the 2007 meeting of the American Society of Hematology (17). This analysis
indicates an improvement in median survival from 15 to 24 months and an increase
in 2-year survival from 26% to 51%.
Similar response data in Phase II trials of DAC culminated in a Phase III trial
randomizing patients with IPSS int-1, int-2 and high-risk MDS to treatment with
DAC or observation (18). Among 89 patients treated with DAC, 8 achieved CR, 9
PR, and 13 hematologic improvement. Time to AML or death was not statistically
significantly different between the two arms, but the two Kaplan–Meier curves
separated between 100 and 300 days, and was significantly different in the subset of
patients with IPSS int-2 or high-risk disease. Median overall survival did not differ
between the decitabine-treated patients and supportive care arm (14.0 months vs.
14.9 months, respectively).
Comparison of the DAC registration trial with CALGB9221 is not straight-
forward. The CALGB trial was performed prior to development of IPSS scores,
and cytogenetic analysis was not required. Among 39 patients in whom IPSS
could be assigned, 41% had int-2 or high-risk disease; 69% of patients in the
DAC study fell into these categories. The two trials differed in other important
aspects. The maximum number of cycles of therapy planned in the DAC trial was
two cycles past-CR, six cycles if no PR, and eight cycles if no CR. In fact, study
physicians administered a median number of three cycles of therapy. In contrast,
in the CALGB trial of 5AC, patients who achieved CR received three additional
cycles, while all other patients were treated until progression. Given the small
number of complete responders, this design essentially administered maintenance
therapy to most of the patients receiving 5AC. The CALGB investigators admin-
istered a median number of nine cycles of therapy. A similar number of cycles
was reportedly administered in the survival trial (17).
Examination of the deaths on study provides insight into why less cycles
of therapy were administered on the DAC study. A total of 10 out of 89 patients
treated with DAC died during the first two cycles, compared with only one death in
99 patients treated with 5AC on the CALGB trial. This suggests that the toxicity of
the regimen studied on the DAC trial, which is the FDA-approved dose schedule,
likely exceeds that of the approved schedule of 5AC. However, it must be noted
that a similar percentage of patients in the control arm of the DAC study died while
on-study. In addition to induction deaths, treatment toxicity may have accounted
for early exit from the treatment arm on the DAC study, sabotaging the time to
AML-or-death endpoint. The “maintenance” administered in the 5AC trial may
account for the apparent greater median duration of response (15 months vs.
10 months).
The European Organization for Research and Treatment of Cancer (EORTC)
has completed a Phase III trial of DAC versus supportive care in high-risk MDS
using the same 3-day dose schedule as the United States registration trial. Overall
survival is a primary endpoint of this trial. Data presentation is expected within
2008.
As with 5AC, the currently approved dose schedule of DAC was not arrived
at through careful dose finding. Both drugs have subsequently been studied in
alternative dose schedules. 5AC was administered at doses ranging from 25 to
75 mg/m2 /day for 5 to 14 days, in sequence with the HDAC inhibitor sodium
phenylbutyrate. Complete responses developed in patients receiving 50 mg/m2 /day
for 10 days as well as 25 mg/m2 /day for 14 days (19). An ongoing community
practice–based trial randomizes patients between a variation on the labeled indi-
cation (75 mg/m2 /day × 7 days, interrupted by a 2-day weekend break) and two
alternative dose schedules (75 mg/m2 /day × 5 days, and 50 mg/m2 /day × 10 days,
interrupted by a 2-day weekend break) (20).
MD Anderson investigators performed a Phase I study of prolonged dosing
of low-dose DAC (21). Doses ranged from 5 to 20 mg/m2 /day for 10 days (with a
weekend break). The investigators selected 15 mg/m2 /day as the most promising
schedule. However, rather than expanding that study, the investigators proceeded
with a randomized Phase II trial examining three 5-day intravenous schedules
administering a total dose of 100 mg/m2 /day. They selected 20 mg/m2 /day intra-
venous administration as the most promising schedule (22). A multicenter Phase
II trial has been completed using that dose schedule (23).
Clinical investigation of oral formulations of both DNMT inhibitors has
recently begun. The availability of effective oral formulations of azacytosine
analogues would likely increase the application of these important therapies to
appropriate patients, many of whom decline frequent clinic visits for parenteral
injections of these drugs on an indefinite basis.
Upon peer-review of the two survival studies, data will allow determination
of the survival benefit in high-risk MDS patients of the administration of the two
FDA-indicated dose schedules of azacitidine and DAC. Extrapolation of survival
benefit from these trials to any of the alternative dose schedules would require
leaps of faith and will not be scientifically valid. Thus, physicians purporting to
practice evidence-based medicine will administer the approved schedule of those
490 Gore
nucleosides which have demonstrated survival benefit if improving survival is the

goal of the particular patient with high-risk MDS.
Two small studies have examined the potential use of azacytosine nucleo-
side analogues in patients relapsed following allogeneic stem cell transplantation.
Various doses of DAC were administered to 14 patients with acute leukemia and
patients with blast crisis CML after failure of allogeneic SCT in a Phase I trial.
Donor stem cells were reinfused following 5 days of DAC. Eight patients demon-
strated hematologic responses (24). 5AC has also been used to induce remission
in a patient with AML relapsed after allogeneic stem cell transplantation (25).
AZACYTOSINE NUCLEOSIDES AND MDS EPIGENETICS

The efficacy of DNMTi for the treatment of MDS led investigators to search for
evidence of DNA methylation in this group of hematologic malignancies. Epige-
netic abnormalities in hematologic malignancies have been recently reviewed (26).
In AML, a variety of important cell regulatory genes are frequently methylated
and silenced including p15INK4b (cyclin-dependent kinase inhibitor), CDH-1 (E-
cadherin, adhesion molecule), SOCS-1 (signaling molecule), p73 (p53 analogue),
DAP kinase 1 (apoptosis), HIC1 (p53-like), RAR␤2 (retinoid signaling), CRBP1
(retinoid signaling).
The methylome of MDS has been less-well characterized. Due to its fre-
quency of methylation in AML, p15 has been extensively investigated and likely
represents the most frequently methylated gene in MDS studied to date (27,28).
SOCS-1 methylation has been reported in 31% to 47% of MDS patients (29).
RASSF1 was methylated in 9% of cases studied (30). In a series of 13 MDS
patients, CD34+ cells harbored frequent methylation of p15, p16, p73, RAR␤,
DAP kinase, and WT-1 (31). Recently, a putative tumor suppressor gene, RIL,
which maps to chromosome 5q31.1, was found to be methylated in 36% of MDS
samples (32); the potential importance of this finding derives from the frequency
of deletion of this band in MDS. Preliminary presentation of a genomics-based
methylation assay (HELP) suggested that hundreds of genes may be methylated
in MDS compared to normal CD34+ cells, and greater than the number of genes
methylated in de novo AML (33).
Gene methylation has been associated with a decreased likelihood of
response to cytarabine/anthracycline-based induction chemotherapy in patients
with high-risk MDS and MDS-related AML. Interestingly, only 2 out of 15 patients
with CDH-1 methylation achieved remission, compared with 58% of the patients
who did not demonstrate CDH-1 methylation. None of the six patients methylated
at p15, CDH-1, and HIC achieved remission (34).
A variety of studies have demonstrated that treatment with 5AC and DAC
leads to reversal of DNA methylation. Reversal of the usual heavy methylation
of noncoding elements of DNA (so-called LINE elements) has been used as
a marker of reversal of global DNA methylation. Several studies indicate that
treatment with DAC or 5AC leads to transient decrease in LINE methylation. The
decrease in LINE methylation ranges from 3% to 20%, peaks between 7 and 14

days following the initiation of therapy, mostly recovers prior to retreatment, and
does not correlate with clinical response. The significance of LINE methylation
reversal appears limited to providing evidence that the nucleoside impacts its
putative target, DNMT (35–37).
Investigations of gene-specific methylation reversal have focused on the
frequently methylated p15 and CDH-1 genes. Early studies suggested that DAC
treatment induced reversal of p15 methylation, but this did not correlate with clin-
ical response (21,38). Reversal of p15 or CDH-1 methylation during the first cycle
of therapy was highly correlated with response in a small Phase I study sequencing
5-azacitidine with sodium phenylbutyrate (19). p15 baseline expression appeared
lower in clinical responders to DAC than in nonresponders; in addition, expression
of this gene significantly increased during the first cycle of therapy in responding
patients compared with nonresponders (22). The highest mean level of expression
in this study remained quite low, and the biological significance of such expression
is unclear. Thus, while it is now established that treatment with DNMT inhibitors
induces methylation reversal in tumor cells early in treatment, the essential role
of such methylation reversal in the induction of clinical response has not yet been
conclusively established.
In two clinical studies, administration of 5AC and DAC led to hyperacety-
lation of histones in peripheral blood and bone marrow cells (19,39). The mech-
anism behind this surprising finding has not been explained. If methylation is
reversed, and transcriptional corepression complexes are released, histone acetyl-
transferases would re-establish acetylation of histones; however, global hyper-
acetylation would not necessarily follow. Recent evidence links chromatin mod-
ifications with double-strand breaks in DNA [reviewed in (40)]. In vitro, DAC
induces expression of the unmethylated p21WAF1/CIP1 gene. p21 expression was
induced in cells in which DNMT activity was almost completely knocked-out, but
was dependent on the presence of wild-type p53. Expression was also abrogated
by an inhibitor of ATM phosphorylation. Treatment with very low doses of DAC
induced expression of the phosphorylated variant histone ␥ -H2AX, closely asso-
ciated with double-stranded breaks in DNA (ED50 16 nM) (41). 5AC can also
induce expression of ␥ -H2AX (42).
The induction of γ-H2AX by DAC required ATM activation (42). Other
DNA repair proteins are also activated following DAC treatment. DNMT1-
deficient cells demonstrate a blunted response in these proteins. While these
data lend further support to a potential role for DNA damage and repair in
response to azacytosine analogues, most of the studies have been performed
at supra-pharmacologic concentrations (10 ␮M) making their clinical relevance
unclear (42).
These data suggest that azacytosine nucleosides may be DNA-damaging
agents and that DNA damage responses need to be considered in assessing the
molecular effects of these drugs. These compounds may have other unexplored
molecular mechanisms in addition to their DNMT inhibitory effects.
492 Gore
STRATEGIES TO IMPROVE THE EFFICACY OF DNMT INHIBITORS

Despite the efficacy of DNMT inhibitors for the treatment of MDS, the overall
response rates remain inadequate, and disease eventually recurs or progresses in
essentially all patients. Because methylated promoters silence gene expression
through chromatin remodeling which requires HDAC activity, combinations of
DNMT and HDAC inhibitors seem logical. In vitro, the sequential application of
DNMT inhibitors followed by HDAC inhibitors leads to synergistic expression
of a wide variety of methylated genes in a broad range of cancers (43). In vitro,
the sequence of application appears critical to achieve synergy; cells must be
pretreated with the DNMT inhibitor. Lest this combination approach appear to
represent rational “epigenetic targeting,” sequential DNMT and HDAC inhibitor
combinations also synergistically induce apoptosis, mandating the examination of
the impact of the combinations on DNA damage and repair as well as epigenetic
gene silencing (41).
Phase I studies combining 5-azacitidine with sodium phenylbutyrate
(sequential) (19) and DAC with valproic acid [sequential (39) and concomitant
(36)] have been published. These studies show that the combinations are tolerable
and active. Recent and current studies combine 5AC with the benzamide SNDX-
275 (formerly MS-275) (44), 5AC and vorinostat, 5AC and the benzamide MGCD
0103 (45), DAC and belinostat, and DAC and romidepsin. Randomized trials will
be required to assess the marginal benefit of the addition of the HDAC inhibitor
to the DNMT inhibitors.
CURRENT ROLE OF AZACYTOSINE ANALOGUES IN THE

TREATMENT OF MDS
Apart from the unique activity of lenalidomide in low-risk MDS patients with
deletions of chromosome 5q31.1, the azacytosine nucleosides represent the most
active single agents for the treatment of MDS. The drugs are highly active in
both lower and higher risk MDS. Red cell transfusion independence can be
achieved in approximately 50% of transfusion-dependent patients. 5AC admin-
istration has been associated with improved quality-of-life measurements (16).
Thus, these drugs represent important options for hematologically compro-
mised low-risk MDS patients who are not appropriate for or who are failing to
respond to erythropoietic stimulating agents, lenalidomide, or immunosuppressive
drugs.
If peer review and publication of the 5AC survival trial validate the
findings presented in abstract form (17), 5AC will become the standard-of-care
for high-risk MDS patients, including those AML-TLD patients formerly
classified as RAEB-t. The doubling of 2-year survival for this high-risk group
of patients must be considered a major achievement. Important unanswered
questions which this trial will hopefully address include whether patients with
stable disease but without objective hematologic response benefit from ongoing
therapy with azacytosine nucelosides. The outcomes of the EORTC DAC survival
trial will determine whether the current FDA-indicated schedule of DAC has
a comparable impact on survival of high-risk patients. No current or planned
studies address the impact of alternative dosing schedules of either drug on patient
survival.
Allogeneic stem cell transplantation remains the only therapy with curative
potential and remains the gold-standard treatment for MDS patients. No appropri-
ate patient should be denied access to allogeneic stem cell transplant in order to be
treated with azacytosine nucleosides. However, the cure rate of allogeneic stem cell
transplant in patients with IPSS int-2 and high-risk MDS does not exceed 40% (46).
The ability of the azacytosine nucleoside analogues to “down-stage” MDS patients
prior to allogeneic stem cell transplant requires investigation. While these drugs
will no doubt reduce blast percentages in many patients, whether this translates
into improved transplant outcomes requires formal testing. One might suspect that
patients treated with azanucleosides might be at lower risk for preparative regimen
toxicity than patients treated with cytarabine-based induction therapy; however,
the relative impact of several months of azanucleoside therapy on opportunistic
infections prior to transplant must also be considered.
Hematologic response to azanucleosides may be a test of disease responsive-
ness which could predict for disease-free outcome following transplant, similar
to the chemosensitivity of relapsed non-Hodgkins lymphoma. The use of aza-
cytosine nucleosides prior to transplant raises other important questions. How
should azacytosine nonresponders be handled regarding subsequent transplant?
Should they receive induction chemotherapy prior to transplant? Should they be
transplanted directly? Should they be excluded from transplant? Ideally, transplant
consortiums would study the integration of azanucleosides into transplant strate-
gies in a streamlined cooperative fashion to answer these questions rapidly and
decisively.
If the 5AC survival data appear robust, the community will need to consider
whether high-risk patients should be offered treatment with 5AC prior to eligibility
for clinical trials. 5AC does not represent a curative option for patients. Phase
II window-of-opportunity trials may remain very appropriate to offer high-risk
patients prior to therapy with 5AC. However, one must consider that in CALGB
9221, time to AML-or-death was greater in patients randomized to receive 5AC
initially, compared to the observation arm, half of whom received 5AC later. This
could suggest that delay of 5AC treatment could compromise survival if the Phase
II window drug were not active. It is the opinion of this author that validation
of the 5AC survival data would put into question the appropriateness of offering
trials of drugs in early-stage development to high-risk MDS patients who are
azanucleoside-naı̈ve. The ongoing development of combination therapies which
aim to improve the outcomes of azacytosine nucleosides must remain a major
priority for the MDS community; whenever possible, such combinations should
be studied in trials randomizing against the azanucleoside alone to obtain early
signal that the combination continues to be viable for development.
494 Gore
REFERENCES
1. Jones PA, Baylin SB. The fundamental role of epigenetic events in cancer. Nat Rev
Genet 2002; 3(6):415–428.
2. Jenuwein T, Allis CD. Translating the histone code. Science 2001; 293(5532):1074–
1080.
3. Esteller M. The necessity of a human epigenome project. Carcinogenesis 2006;
27(6):1121–1125.
4. Von Hoff DD, Slavik M, Muggia FM. 5-Azacytidine. A new anticancer drug with
effectiveness in acute myelogenous leukemia. Ann Intern Med 1976; 85(2):237–
245.
5. Saiki JH, McCredie KB, Vietti TJ, et al. 5-Azacytidine in acute leukemia. Cancer
1978; 42(5):2111–2114.
6. Jones PA, Taylor SM. Cellular differentiation, cytidine analogs and DNA methylation.
Cell 1980; 20(1):85–93.
7. Jones PA, Taylor SM. Hemimethylated duplex DNAs prepared from 5-azacytidine-
treated cells. Nucleic Acids Res 1981; 9(12):2933–2947.
8. Taylor SM, Jones PA. Mechanism of action of eukaryotic DNA methyltransferase.
Use of 5-azacytosine-containing DNA. J Mol Biol 1982; 162(3):679–692.
9. Charache S, Dover G, Smith K, et al. Treatment of sickle cell anemia with 5-
azacytidine results in increased fetal hemoglobin production and is associated with
nonrandom hypomethylation of DNA around the ␶ -␦-␤-globin gene complex. Proc
Natl Acad Sci USA 1983; 80:4842–4846.
10. Dover G, Charache S, Boyer SH, et al. 5-Azacytidine increased HbF production and
reduces anemia in sickle cell disease. Dose–response analysis of subcutaneous and
oral dosing regimens. Blood 1985; 66(527):532.
11. Silverman L, Holland JF, Demakos EP, et al. Azacitidine in myelodysplastic syn-
dromes: CALGB studies 8421 and 8921. Ann Hematol 1994; 68:A12.
12. Silverman LR, Holland JF, Weinberg RS, et al. Effects of treatment with 5-azacytidine
on the in vivo and in vitro hematopoiesis in patients with myelodysplastic syndromes.
Leukemia 1993; 7(Suppl. 1):21–29.
leukemia group B. J Clin Oncol 2002; 20(10):2429–2440.
14. Silverman LR, McKenzie DR, Peterson BL,et al. Further analysis of trials with
azacitidine in patients with myelodysplastic syndrome: Studies 8421, 8921, and
9221 by the Cancer and Leukemia Group B. J Clin Oncol 2006; 24(24):3895–
3903.
15. Cheson BD, Bennett JM, Kantarjian H, et al. Report of an International Working
Group to standardize response criteria for myelodysplastic syndromes. Blood 2000;
96(12):3671–3674.
16. Kornblith AB, Herndon JE, Silverman LR, et al. Impact of azacytidine on the quality
of life of patients with myelodysplastic syndrome treated in a randomized Phase
III trial: A Cancer and Leukemia Group B study. J Clin Oncol 2002; 20(10):2441–
2452.
17. Fenaux P, Mufti GJ, Santini V, et al. Azacitidine (AZA) treatment prolongs overall
(CCR): Results of the AZA-001 Phase III study. Blood 2007; 110(Suppl. 1).
myelodysplastic syndromes: Results of a Phase III randomized study. Cancer 2006;
106(8):1794–1803.
19. Gore SD, Baylin S, Sugar E, et al. Combined DNA methyltransferase and histone
deacetylase inhibition in the treatment of myeloid neoplasms. Cancer Res 2006;
66(12):6361–6369.
20. Lyons RM, Cosgriff T, Modi S, et al. Results of the initial treatment phase of a
study of three alternative dosing schedules of azacitidine (Vidaza) in patients with
myelodysplastic syndromes (MDS). Blood 2007; 110(Suppl. 1).
21. Issa JP, Garcia-Manero G, Giles FJ, et al. Phase 1 study of low-dose prolonged
exposure schedules of the hypomethylating agent 5-aza-2 -deoxycytidine (decitabine)
in hematopoietic malignancies. Blood 2004; 103(5):1635–1640.
22. Kantarjian H, Oki Y, Garcia-Manero G, et al. Results of a randomized study of three
schedules of low-dose decitabine in higher risk myelodysplastic syndrome and chronic
myelomonocytic leukemia. Blood 2006.
23. Steensma DP, Baer MR, Slack JL, et al. Preliminary results of a Phase II study of
decitabine administered daily for 5 days every 4 weeks to adults with myelodysplastic
syndrome (MDS). Blood 2007; 110(Suppl. 1).
24. Ravandi F, Kantarjian H, Cohen A, et al. Decitabine with allogeneic peripheral blood
stem cell transplantation in the therapy of leukemia relapse following a prior trans-
plant: Results of a Phase I study. Bone Marrow Transplant 2001; 27(12):1221–1225.
25. Graef T, Kuendgen A, Fenk R, et al. Successful treatment of relapsed AML after
allogeneic stem cell transplantation with azacitidine. Leuk Res 2007; 31(2):257–
259.
26. Galm O, Herman JG, Baylin SB. The fundamental role of epigenetics in hematopoietic
malignancies. Blood Rev 2006; 20(1):1–13.
27. Quesnel B, Guillerm G, Vereecque R, et al. Methylation of the p15(INK4b) gene
in myelodysplastic syndromes is frequent and acquired during disease progression.
Blood 1998; 91:2985–2990.
28. Uchida H, Zhang J, Nimer SD. AML1 A and AML1B can transactivate the human
IL-3 promoter. J Immunol 1997; 158(5):2251–2258.
29. Brakensiek K, Langer F, Schlegelberger B, et al. Hypermethylation of the suppressor
of cytokine signalling-1 (SOCS-1) in myelodysplastic syndrome. Br J Haematol 2005;
130(2):209–217.
30. Johan MF, Bowen DT, Frew ME, et al. Aberrant methylation of the negative regulators
RASSFIA, SHP-1 and SOCS-1 in myelodysplastic syndromes and acute myeloid
leukaemia. Br J Haematol 2005; 129(1):60–65.
31. Hopfer O, Komor M, Koehler IS, et al. DNA methylation profiling of myelodysplastic
syndrome hematopoietic progenitor cells during in vitro lineage-specific differentia-
tion. Exp Hematol 2007; 35(5):712–723.
32. Boumber YA, Kondo Y, Chen X, et al. RIL, a LIM gene on 5q31, is silenced by
methylation in cancer and sensitizes cancer cells to apoptosis. Cancer Res 2007;
67(5):1997–2005.
33. Figueroa ME, Fandy T, McConnell MJ, et al. Myelodysplastic syndrome (MDS) dis-
plays profound and functionally significant epigenetic deregulation compared to acute
myeloid leukemia (AML) and normal bone marrow cells. Blood 2007; 110(Suppl. 1).
34. Grovdal M, Khan R, Aggerholm A, et al. Negative effect of DNA hypermethylation
on the outcome of intensive chemotherapy in older patients with high-risk myelodys-
496 Gore
plastic syndromes and acute myeloid leukemia following myelodysplastic syndrome.

Clin Cancer Res 2007; 13(23):7107–7112.
35. Yang AS, Doshi KD, Choi SW, et al. DNA methylation changes after 5-aza-2 -
deoxycytidine therapy in patients with leukemia. Cancer Res 2006; 66(10):5495–
5503.
36. Garcia-Manero G, Kantarjian HM, Sanchez-Gonzalez B, et al. Phase I/II study of the
combination of 5-aza-2 -deoxycytidine with valproic acid in patients with leukemia.
Blood 2006; 108:3271–3279.
37. Soriano AO, Yang H, Faderl S, et al. Safety and clinical activity of the combination
of 5-azacytidine, valproic acid, and all-trans retinoic acid in acute myeloid leukemia
and myelodysplastic syndrome. Blood 2007; 110(7):2302–2308.
38. Daskalakis M, Nguyen TT, Nguyen C, et al. Demethylation of a hypermethy-
lated P15/INK4B gene in patients with myelodysplastic syndrome by 5-aza-2 -
deoxycytidine (decitabine) treatment. Blood 2002; 100(8):2957–2964.
39. Blum W, Klisovic RB, Hackanson B, et al. Phase I study of decitabine alone or
in combination with valproic acid in acute myeloid leukemia. J Clin Oncol 2007;
25(25):3884–3891.
40. Karagiannis TC, El Osta A. Chromatin modifications and DNA double-strand breaks:
The current state of play. Leukemia 2007; 21(2):195–200.
41. Jiemjit A, Fandy TE, Carraway H, et al. p21(WAF1/CIP1) induction by 5-azacytosine
nucleosides requires DNA damage. Oncogene 2008; 27:3615–3623.
42. Palii SS, Van Emburgh BO, Sankpal UT, et al. DNA methylation inhibitor 5-aza-
2 -deoxycytidine induces reversible genome-wide DNA damage that is distinctly
influenced by DNA methyltransferases 1 and 3B. Mol Cell Biol 2008; 28(2):752–
771.
43. Cameron EE, Bachman KE, Myohanen S, et al. Synergy of demethylation and histone
deacetylase inhibition in the re-expression of genes silenced in cancer. Nat Genet 1999;
21:103–107.
44. Gore SD, Jiemjit A, Silverman LB, et al. Combined methyltransferase/histone
deacetylase inhibition with 5-azacitidine and MS-275 in patients with MDS, CMMoL
and AML: Clinical response, histone acetylation and DNA damage. Blood 2006;
108(Suppl. 1).
45. Garcia-Manero G, Yang AS, Klimek V, et al. Phase I/II study of MGCD0103, an oral
isotype-selective histone deacetylase inhibitor, in combination with 5-azacitidine in
higher-risk myelodysplastic syndrome and acute myelogenous leukemia. Blood 2007;
110(Suppl. 1).
lated donors in patients with myelodysplastic syndrome. Blood 2002; 100(4):1201–
1207.
21
Intensive Chemotherapy and
Stem Cell Transplantation in
Guillermo F. Sanz
Department of Hematology, Hospital Universitario La Fe, Valencia, Spain
Theo de Witte
Department of Hematology, Radboud University Hospital, Nijmegen,
The Netherlands
INTRODUCTION
The myelodysplastic syndromes (MDS) include a heterogeneous group of clonal
disorders of the hematopoietic stem cell, with widely varying natural histories and
risks for disease-related complications, including progression to acute myeloid
leukemia (AML) or death from the effects of peripheral blood cytopenias (1). The
incidence of MDS increases markedly with age (see chap. 2), and since the median
age of patients is above 70 years, many patients with MDS have important comor-
bid conditions at presentation that complicate therapeutic decisions, especially
those involving more intense and riskier therapies (2,3). When MDS occurs in
younger patients, the disease is often preceded by treatment with radiotherapy or
certain chemotherapeutic agents, such as alkylating substances and topoisomerase
II inhibitors (see chap. 8) (4–6). Such an exposure history marks patients’ disease
as high risk, and furthermore, other residual nonhematological effects from those
prior treatments may also limit the ability to successfully deliver intensive therapy
for MDS.
The diagnosis and subclassification of MDS into morphological subtypes
is carried out according to the proposals of the French-American-British (FAB)
497
498 Sanz and Witte
cooperative group (7) and, more recently, the World Health Organization (WHO)
(8) classification systems (see chaps. 1 and 9). Even within a given FAB or
WHO subtype, however, in some cases, the disease has an indolent and relatively
stable course, whereas in other instances there is a rapid increase in the severity
of cytopenias or progression to AML (9). The International Prognostic Scoring
System (IPSS) (10) has been shown to be a useful tool for further evaluating
prognosis in MDS (see chaps. 1 and 16), and it is widely used by clinicians for
risk assessment and treatment planning.
The IPSS is based on the percentage of bone marrow blasts, the number
of cytopenias, and cytogenetics, and the scoring system distinguishes four patient
risk groups with clearly different median survival and AML risk (designated as
low, intermediate-1, intermediate-2, and high risk). However, in clinical practice,
only two general risk groups are usually considered: lower risk (i.e., IPSS low
and intermediate-1–risk groups) versus higher risk (i.e., IPSS intermediate-2 and
high-risk groups) MDS (11). For higher risk patients with MDS, the median time
to AML evolution is between 0.2 and 1.1 years, while the median survival is only
0.4 to 1.2 years (10). While the main therapeutic goals for lower risk MDS include
symptom management, amelioration of cytopenias, and improvement of quality
of life; for high-risk MDS, the objectives should be, whenever possible, to attempt
to alter the natural history of the disease, increase survival, and cure.
At present, the only treatment modality with demonstrated curative potential
is allogeneic stem cell transplantation (SCT). But despite recent advances in the
field, this procedure is applicable to a minority of MDS patients, due to lack of
donor availability and patient criteria (advanced age and presence of comorbidi-
ties). The mortality rate continues to be high, due to relapse and treatment-related
mortality (12). High-intensity AML-type chemotherapy, with or without autolo-
gous stem cell transplantation in patients who achieved complete remission (CR),
is associated with not only restoration of normal autologous hematopoiesis in
many patients with MDS, but also carries substantial morbidity and mortality.
Thus, this kind of therapy is usually reserved for relatively young and fit MDS
patients with higher risk disease. In some studies, the long-term outcome with
autologous SCT has been comparable to that achieved with allogeneic SCT (13).
Hypomethylating agents, such as azacitidine (14,15) and decitabine (see
chap. 20) (16), have recently shown to prolong overall survival (OS) and time to
progression to AML or death as compared to best supportive care in high-risk
MDS, but it is unclear whether there are any cures with these drugs. Lenalidomide
(see chaps. 12 and 19) is able to induce transfusion independence and cytoge-
netic remission in a high proportion of patients with transfusion-dependent lower
risk MDS with the deletion 5q chromosomal abnormality (17), but the effect of
lenalidomide on long-term outcomes is uncertain.
In this chapter, we review the results and the main prognostic factors for
outcome after intensive therapies—defined here as AML-type myelosuppressive
chemotherapy and SCT—for patients with MDS, aiming to identify those MDS
patients who can benefit most from these approaches.
Stem Cell Transplantation in Myelodysplastic Syndromes 499
INTENSIVE AML-TYPE CHEMOTHERAPY IN MDS

High-risk MDS are clonal disorders that frequently overlap with AML. In fact,
there is a blurred and arbitrary border of 20% blasts (WHO) or 30% blasts (FAB)
used by hematopathologists to separate AML from MDS (8,9). In addition to
the known secondary AML cases who had a hematological disorder prior to
developing AML, a sizeable proportion of older patients classified as “de novo”
AML have probably evolved from prior high-risk MDS, but lack documented
antecedent MDS. In these cases, the presence of multilineage dysplasia is frequent
and is associated with a bad prognosis (18).
In addition to morphological similarities between MDS and AML, there is
also a genetic overlap, demonstrated best by the numerous shared chromosomal
abnormalities. While 40% to 50% of de novo AML are accompanied by cyto-
genetic abnormalities (19,20), this proportion increases to 60% to 70% in MDS
(21,22). With the exception of core-binding factor (CBF) abnormalities, such as
the t(15;17), t(8;21), or inv(16) that qualify for a diagnosis of AML by definition
(9), MDS cases may exhibit the same cytogenetic abnormalities as AML cases.
Some abnormalities, such as complex karyotypes, nonisolated del(5q), monosomy
5, del(7q), monosomy 7, t(6;9), del(17p), or 11q23 rearrangements, are frequent
in MDS and in de novo AML and confer a poor prognosis in both diseases (19–
22). Other abnormalities, like del(20q), trisomy 8, trisomy 21, or loss of the Y
chromosome, have a more benign outlook. In addition, some molecular markers
universally recognized in cases of AML, like overexpression of WT1, FLT3-ITD,
and N-RAS mutations, are not unusual in patients with MDS, where they fre-
quently predict progression to AML (23–25).
The clinical course in many high-risk MDS cases is indistinguishable from
AML, including a short-life expectancy. Finally, the outcome after intensive AML-
type chemotherapy in high-risk MDS is not different from that observed in patients
with de novo AML treated with identical regimens, when the prognostic impact
of age and cytogenetics are accounted for (26–28).
All the above arguments are the bases for the use of intensive AML-type
chemotherapy in high-risk MDS. In fact, several studies have demonstrated a
beneficial impact of attaining CR on OS after AML-type chemotherapy in MDS
patients (29,30). The usefulness of intensive AML-type chemotherapy in high-
risk MDS is more pronounced in younger patients, as shown in a recent series
including 232 patients under age 50 published by Andrea Kuendgen in Düsseldorf
and her colleagues (31). In contrast to the results in higher risk patients, in the
same study, younger patients belonging to the IPSS low- and intermediate-1–risk
groups did not clearly benefit from intensive chemotherapy and allogeneic SCT as
compared to supportive care only (31). In keeping with these results, aggressive
treatment approaches should rarely be recommended to younger MDS patients
belonging to the IPSS low- and intermediate-1–risk groups, but they are indicated
for many patients with higher risk disease. Possible exceptions to this rule are
lower risk MDS with high-risk features not captured by the IPSS, such as marked
500 Sanz and Witte
myelofibrosis, high-risk chromsomal abnormalities like t(3;21), or life-threatening

neutropenia and thrombocytopenia.
Results of Intensive AML-Type Chemotherapy

The main objective of intensive chemotherapy is to eradicate or suppress the
myelodysplastic clone, restore polyclonal haematopoiesis, and thereby induce
long-term CR. The first reports that demonstrated success in achieving CR in
MDS by using intensive chemotherapy appeared in the early 1980s (32,33) and
now have been extensively confirmed in many other series (34–51). The reported
CR rate in high-risk MDS after the classical AML–type remission induction
chemotherapy “7+3”, combination of cytarabine and one of the anthracyclines
(daunorubicin or idarubicin), has ranged from 48% to 60%, with a median of
55% (34–40). Remission induction chemotherapy in patients with MDS is not
apparently as effective as in de novo AML, in which remission induction rates
overall exceed 65% to 70%. This difference likely reflects the older age of MDS
patients and the more frequent presence of poor-risk cytogenetics, rather than a
specific unfavorable effect due to MDS per se. Early death rates with “7+3” are
also generally higher in MDS patients (10–20%) than in de novo AML patients
(5–10%), probably again because MDS patients are generally older.
To improve CR rate and diminish treatment-related morbidity and mortality,
some trials have reduced the intensity of induction remission chemotherapy in
older patients with MDS. In two studies (37,39), this approach has appeared to
be effective in lowering the early death rate (5% and 8%, respectively), but was
unsuccessful in increasing the CR rate (60% and 48%, respectively) due to a high
rate of resistance disease (35% and 44%, respectively). It should be stressed that the
main cause of induction failure in MDS is leukemic resistance (25–45%), higher
than in de novo AML. As in de novo AML, the addition of a third chemotherapeutic
agent, like etoposide, has not clearly improved the CR rate in MDS (36,38).
Long-term outcome using only chemotherapy consolidation with an anthra-
cycline and cytarabine, with or without maintenance, is very poor, mainly due to a
very high relapse risk (higher than 75% in most series) (34–40). The estimated CR
duration in patients who attain CR is only 5 to 8 months (38,40), the median OS
reported is only 12 months (range, 4–15 months) (34–40), and the estimated prob-
ability of OS at 5 years is lower than 10% in most reports (34,35,38–40). These
results are disappointing and are comparable to the OS in high-risk MDS patients
who receive only supportive care (10). In one series that included younger patients
(median age of 48 years) and the use of autologous or allogeneic transplantation
as post-remission therapy, OS reached 26% at 5 years (36).
Given the poor outcome with classical AML-type chemotherapy regimens,
newer regimens including topotecan, a topoisomerase I inhibitor; fludarabine, a
purine analog; or the immunoconjugate gemtuzumab ozogamycin (GO) have been
investigated in clinical trials (41–49). Despite a lower remission induction mortal-
ity rate (7–8%) with the combination of topotecan and cytarabine with or without
idarubicin or even cyclophosphamide (35,49), the CR rate (median, 52%; range,

33–61%), CR duration (8–12 months), and OS (8% at 5 years) were not substan-
tially improved compared to the classical regimens (35,42,46,48,49). In contrast,
several reports with the “FLAG-Ida” regimen (fludarabine, cytarabine, idarubicin,
and G-CSF) have been more encouraging (50,51). The Spanish PETHEMA coop-
erative group has recently reported the results of remission induction chemotherapy
with fludarabine, idarubicin, cytarabine, and G-CSF in 103 patients with a median
age of 62 years with either high-risk MDS (66 patients) or AML secondary to
MDS (40 patients) (50). Sixty-six patients (64%) attained CR, 25 patients (24%)
had resistant disease, the induction death rate was 12%, and median disease-free
survival (DFS) was 11 months.
However, the benefit of adding fludarabine remains unclear. In one study
(35), the outcome of 94 high-risk MDS patients who were treated with fludarabine
and cytarabine was retrospectively compared to that of patients receiving regimens
containing topotecan or idarubicin and cytarabine. The treatment resistance rate
was lower when fludarabine was added (23%), but the induction death rate was
higher (23%), resulting in a similar CR rate (54%). Another retrospective study
compared remission induction outcomes of 29 P-glycoprotein–negative patients
treated with mitoxantrone, high-dose cytarabine, and fludarabine with that of 32
patients with similar characteristics who were treated with the same combination
but without fludarabine (45). There were no significant differences in terms of CR,
survival or DFS between both treatment arms. In another study (52), the addition
of fludarabine to cytarabine and G-CSF in 91 MDS and 43 AML patients did not
clearly improve OS (39% FLAG, 24% AG) and DFS (23% FLAG, 16% AG).
It is unclear whether there is a role of GO in induction remission in MDS;
results to date have been discouraging. In a single institution study, GO alone
or combined with interleukin-11 was tested in 51 elderly patients with high-risk
MDS and AML (47). The CR rate (22%) and median OS (4 months) observed
were poor and clearly inferior to those achieved in 31 similar patients treated in
the same institution with idarubicin and cytarabine. Preliminary results with the
combination of GO and cytarabine (with or without idarubicin or topotecan) are
also available (41,43,44). The CR rate in these series has ranged from 18% to
52%, and the death rate during induction was 11% to 36%.
In summary, we can affirm that the newer intensive chemotherapy regimens
do not substantially improve the short- and long-term outcomes obtained with the
classical “7+3” regimen of idarubicin and cytarabine. Table 1 summarizes the
results observed in several series of intensive chemotherapy for high-risk MDS,
and Figure 1 shows OS and DFS in one of the largest published series.
Use of Hematopoietic Growth Factors Along with Chemotherapy

The addition of granulocyte-colony stimulating factor (G-CSF) or granulocyte
macrophage–colony stimulating factor (GM-CSF) to intensive chemotherapy in
AML and MDS has been tested, with a goal of improving the CR rate by two
502
Table 1 Results of Intensive Chemotherapy in Several Large Series of Patients with MDS
Median CR Induction
No. of patients/ age rate mortality Resistance
References disease (years) Chemotherapy regimen (%) rate (%) rate (%) Long-term outcome
(34) 160/AML or MDS 67 Cytarabine and an anthracycline 59 10 31 Median OS: 10 mo

PI: CT 5 yr OS: 8%
(36) 184/high-risk MDS or 47 Cytarabine, idarubicin, and 54 15 31 Median OS: 13 mo
sAML etoposide 5 yr OS: 26%
PI: CT ± SCT
77/high-risk MDS Topotecan–cytarabine 56 6 38
PI: CT
67/high-risk MDS Topotecan–cytarabine, and 52 16 31
cyclophosphamide
PI: CT
(35) 96/high-risk MDS 63 Fludarabine–cytarabine 54 23 23 5 yr OS: 8%
PI: CT
270/high-risk MDS Cytarabine and idarubicin 55 17 28
(± fludarabine)
PI: CT
Sanz and Witte
(49) 59/MDS 64 Topotecan and cytarabine 61 7 Median OS: 14 mo
PI: CT Median CRD: 12 mo
27/CMML 44 Median OS: 9 mo
Median CRD: 8 mo
(53) 112/RAEB-t or 58 Idarubicin, cytarabine, etoposide, 62 12 26 5 yr OS: 12%.
sAML and G-CSF 5 yr CRD: 16%
PI: CT
(54) 93/RAEB-t (25) and 72 TAD 2 + 7 with or without 43 Median OS: 10 mo (B)
sAML (68) addition GM-CSF
PI: CT
(50) 103/high-risk MDS or 62 Fludarabine, idarubicin, 64 12 24 3 yr CRD: 15%
sAML cytarabine, and G-CSF
PI: CT ± SCT
Stem Cell Transplantation in Myelodysplastic Syndromes
Abbreviations: CR, complete remission; AML, acute myeloblastic leukemia; MDS, myelodysplastic syndrome; PI, post-induction therapy; CT, chemotherapy; OS,
overall survival; mo, months; CMML, chronic myelomonocytic leukemia; CRD, complete remission duration; RAEB-t, refractory anemia with excess of blasts in
transformation; sAML, AML after MDS; TAD, thioguanine and cytarabine and daunorubicin; GM-CSF, granulocyte-macrophage–colony stimulating factor; G-CSF,
granulocyte-colony stimulating factor; SCT: stem cell transplantation.
503
504 Sanz and Witte
Figure 1 Overall survival and complete remission duration in 510 patients with MDS
treated with intensive chemotherapy. Source: Adapted from Ref. 35.
different mechanisms of action. The first goal of using hematopoietic growth factor
was to shorten the duration of the neutropenic phase and, thus reduce the incidence
and severity of infections and the length of the hospitalization period. If feasible,
this could be particularly important in high-risk MDS, since this population is
more vulnerable to infectious complications than MDS as a whole. Secondly,
hematopoietic growth factors could increase the susceptibility of leukemic cells
to chemotherapy drugs by driving leukemic cells into cell cycle, enhancing the
efficacy of S-phase specific drugs such as cytarabine.
In a randomized trial performed in 722 patients with newly diagnosed AML
with a median age of 68 years, the CR rate was significantly higher in patients
who received G-CSF during chemotherapy (58% vs. 49%), but no significant
differences were observed in terms of OS (55). Patients who received G-CSF
after chemotherapy had a shorter time to neutrophil recovery (median, 20 days vs.
25 days) and a shorter hospitalization (mean, 27 days vs. 30 days). Several studies
have also analyzed the use of G-CSF during and/or after induction chemotherapy
in high-risk MDS (50,52,53,56,–60). The median CR rate reported in those studies
was 65% (range, 51–74%), which compares well with the CR rate achieved with
the classical idarubicin and cytarabine combination without G-CSF. The reported
remission induction treatment-associated mortality rate in high-risk MDS, when
adding G-CSF, was 9% to 12% (50,53,56). Further, two of the three randomized
clinical trials comparing chemotherapy alone versus chemotherapy with G-CSF

found an advantage in the CR rate for the G-CSF arm (57,58,60). However, the
addition of G-CSF to induction or consolidation chemotherapy or both did not
improved CR duration or OS in any of those studies.
In contrast to G-CSF, the addition of GM-CSF to intensive chemotherapy has
failed to show any benefit and, indeed, might be detrimental (54,61). A randomized
controlled trial comparing GM-CSF with placebo after induction chemotherapy
with sequential mitoxantrone and cytarabine was prematurely closed due to high
induction mortality in the GM-CSF arm (44%) (61). And another randomized
study that evaluated the addition of GM-CSF to induction chemotherapy with
thioguanine, cytarabine, and idarubicin showed no clinical benefit in terms of CR
rate (43% in both arms), early death rate, and OS (10 months in both arms) and
carried an increased risk for side effects (54).
Reverting Multidrug Resistance (MDR) Phenotype

The lower CR rate after intensive chemotherapy in MDS compared to de novo
AML could be due in part to a higher incidence of the expression of MDR
proteins by leukemic blasts, as already demonstrated for elderly patients with
AML compared to younger patients with AML (62). Modulation of drug resistance
by using agents capable of reverting the MDR phenotype has been attempted
numerous times in AML without benefit, and in MDS the trial findings have been
contradictory and trial sizes relatively small (63,64–66).
In 1998, Wattel and colleagues in France reported the results of a ran-
domized trial that evaluated the possible effect of the addition of quinine, an
MDR phenotype-reverting drug, to induction chemotherapy with mitoxantrone
and high-dose cytarabine in patients aged less than 66 years with high-risk MDS
(63). Among 42 patients who expressed P-glycoprotein, those who received qui-
nine had a significant improvement in CR rate (52% vs. 18%, p = 0.02) and OS
(13 months vs. 8 months; p = 0.01). In contrast, the addition of quinine did not
influence outcomes in 49 patients who did not show P-glycoprotein expression
(63). In a randomized multicenter clinical trial involving 55 patients more than 60
years old with AML secondary to MDS, the addition of cyclosporine to induction
chemotherapy with idarubicin, cytarabine, and etoposide resulted in improvement
in CR rate (52% vs. 27%; p = 0.01) and leukemia-free survival (median, 12 months
vs. 7 months; p = 0.03), and in lower treatment failure rate (48% vs. 73%; p ⬍
0.0001), without increasing drug toxicity and treatment-related mortality (64).
However, two recent large randomized trials with second- and third-
generation MDR modulators, valspodar (65) and zosuquidar (66), have failed
to demonstrate any benefit for patients with high-risk MDS and AML in terms of
CR or DFS rates. Thus, concern about the likely hazard of exposing all patients to
potentially toxic agents, without clear confirmatory data of their beneficial effect
in larger series, have prevented most clinicians dealing with patients with MDS
from incorporating modulators of drug resistance into their clinical practice.
506 Sanz and Witte
Quality of Life After Intensive Chemotherapy

To date, no definite data assessing quality of life (QOL) in patients with MDS
undergoing intensive AML-type chemotherapy are available. Nonetheless, remain-
ing lifetime spent in hospital could be used as a potential surrogate for QOL,
because QOL decreases markedly during hospitalization but rebounds after dis-
charge (67). Patients with MDS receiving intensive chemotherapy spend 79% of
their remaining lifetime in hospital, whereas this proportion of life-in-hospital
is only 16% for patients treated with decitabine and 14% for those receiving
only supportive care (68,69). Thus, QOL is likely to be inferior in MDS patients
treated with intensive AML–type chemotherapy compared to other less intensive
approaches.
Autologous SCT as Post-Remission Therapy

In view of the high risk of relapse after chemotherapy alone, autologous SCT
has been applied as intensification therapy after successful conventional induc-
tion chemotherapy for patients younger than 60 to 65 years of age, who lack
an appropriate allogeneic donor. In 1995, it was shown that polyclonal primitive
hematopoietic progenitors can be detected in mobilized peripheral blood from
patients with high-risk MDS after intensive chemotherapy (70). The two prerequi-
sites for performing an autologous SCT are achievement of CR following induction
chemotherapy, and harvest of an adequate amount of autologous nonclonal pro-
genitor cells. A successful autologous harvest is feasible in only 40% to 65% of
the patients entering CR (Theo de Witte, unpublished data) (70), which might
reflect the low number of residual normal stem cells or the bone marrow stroma
damage caused by proapoptotic cytokines produced by the MDS clone (71).
The largest series of autologous SCT for high-risk MDS and AML evolving
from MDS (sAML) have been reported by the European Group for Blood and
Marrow Transplantation (EBMT) (13,36,72–74). The first report of the EBMT,
which included 39 patients with MDS or sAML and 21 patients with t-MDS,
showed 2-year survival, DFS, and relapse rates of 39%, 34%, and 64%, respectively
(72). A later EBMT study reported favorable results in 65 patients with t-MDS
who received an autologous graft. The probabilities of 5-year OS and DFS were
35% and 32%, respectively. The cumulative incidences of relapse- and transplant-
related mortality (TRM) were 58% and 22%, respectively (73). The most recently
published EBMT retrospective study included 280 patients with MDS and sAML
transplanted in first CR (74). Survival, DFS, TRM, and relapse risk at 3 years were
41%, 28%, 17%, and 62%, respectively. Interestingly, in this study relapse after
autologous HCT remained constant over time, suggesting that the cure rate would
be lower after a longer follow-up.
This pattern of relapse has been also observed in two other prospective
studies (50,75,76). The Groupe Français des Myélodysplasies reported in 1999 a
Figure 2 Overall survival (A) and disease free survival (B) after autologous SCT in 53
patients with high-risk MDS or AML evolving from MDS. Source: Adapted from Ref. 76.
study on 24 patients who received an autologous SCT. Median DFS and OS were
29 and 33 months, respectivley, from the autograft, and 50% were still in CR after
8 to 55 months (75). In an updated report from this cooperative group including
53 patients and with a median follow-up of 6 years, early TRM was 9%, relapse
rate was 75%, median DFS and OS were 8 and 17 months, respectively, and DFS
and OS at 4 years were 15% and 19%, respectively (Fig. 2) (76). Similarly, in a
study of 22 patients who received an autologous HSC by the Spanish PETHEMA
cooperative group, the relapse rate at 3 years was 69% and no plateau was observed
(Guillermo Sanz, unpublished data).
The relative place of autologous SCT among the other intensive post-
remission alternatives is still disputed. A recent retrospective joint study by the MD
508 Sanz and Witte
Anderson Cancer Center (MDACC) and European Organization for Research and
Treatment of Cancer (EORTC) compared the outcome after intensive chemother-
apy alone (215 patients from MDACC) or after induction chemotherapy followed
when possible by autologous or allogeneic SCT (180 patients from the EORTC,
65 of whom received a transplant) in high-risk MDS younger than 60 years of
age (77). Although DFS was higher in the EORTC cohort (29% vs. 17%, p =
0.02), OS was not clearly different in the two parallel cohorts. Another study has
also been unable to demonstrate any advantage of autologous SCT over intensive
post-remission chemotherapy (50).
On the other hand, two EBMT reports have shown comparable DFS and OS
after allogeneic and autologous SCT for high-risk MDS and sAML (13,36). In
these studies, the lower relapse risk after allogeneic SCT was offset by the higher
TRM of this modality. The multicenter study of the EORTC, EBMT, the Swiss
Group for Clinical Cancer Research (SAKK), and the Italian Group for Adult
Hematologic Diseases (GIMEMA) compared the results of 100 patients who had
entered CR after remission induction chemotherapy and who were candidates for
allogeneic and autologous SCT, depending on the availability of an HLA-identical
sibling (12). The 4-year DFS rates in the group of patients with or without a donor
(31% and 27%) were not clearly different. However, in a more recent study
by the same groups, the 4-year DFS rate of the patients with a donor (46%) was
significantly better than in the group without a donor (Theo de Witte, unpublished).
The significantly higher DFS in patients with a donor compared to the DFS
of this group in the previous study may reflect the improvement in the results of
allogeneic SCT observed in recent years. In fact, nonrelapse mortality (NRM) in
the group with a donor was only 14% compared with 27% in the previous study.
Subgroup analysis of the last study showed that the advantage of the presence
of an HLA-identical sibling donor was only apparent in the patient group with
intermediate and high-risk cytogenetics (Theo de Witte, unpublished). Allogeneic
SCT from matched unrelated donors (MUD) has also recently demonstrated to be
superior to autologous SCT in patients with high-risk MDS or sAML (74). In this
EBMT study, TRM was lower after autologous SCT but RR was lower, and DFS
and OS higher after allogeneic SCT from MUD.
Prognostic Factors for Response to Intensive Chemotherapy

and Autologous SCT
It is crucial to be able to identifiy the 10% to 30% of high-risk MDS patients
who may achieve long-term DFS after AML-type remission induction chemother-
apy followed by further chemotherapy or autologous SCT. The most powerful
indicators of outcome after intensive chemotherapy alone are age, cytogenetics,
performance status, and presence of comorbidities. Older age has been associ-
ated in several studies with a lower CR rate, shorter CR duration, and poorer OS
(35,53,78). The presence of an abnormal karyotype (79,80), especially when
chromosomal aberrations are complex (35,78) or categorized as high-risk as
defined by the IPSS (i.e., complex or chromosome 7 abnormalities) (34,50) por-

trays a poor prognosis, attributable to low rates of CR and a high risk of early
relapse.
The study by Hagop Kantarjian and colleages at MDACC (35) highlighted
the importance of combining age and cytogenetics to predict an acceptable out-
come after intensive chemotherapy in high-risk MDS. In this series, the OS of the
entire cohort of 510 patients was only 8%, but among 82 patients younger than
65 years of age with a normal karyotype the CR rate was 67%, 33% remained in
remission at 5 years, and OS was 27% at 5 years. Further, in a recent report in
MDS patients older than 60 years (34), the CR rate and median OS in patients with
high-risk karyotype according to the IPSS were only 34% and 4 months, respec-
tively. In a study including high-risk patients with MDS or AML evolving from
MDS who were treated with the FLAG-Ida regimen, (50) poor-risk cytogenetics
according to the IPSS classification was the only independent factor associated
with relapse risk or death. In this report, the median OS for patients pertaining to
this category was 7 months and no patient was predicted to be alive at 3 years.
One important question to be addressed is whether the adverse prognostic
effect of advanced age is due to age per se or instead due to the poor performance
status or comorbid conditions which are very common in elderly MDS (3). A poor
Eastern Cooperative Oncology Group (ECOG) performance status (i.e., greater
than 2) (78), the presence of abnormal organ functions (78), and a greater Charlson
comorbidity index (81) or SCT-specific comorbidity index (82) have recently been
found to independently predict a poorer outcome in MDS patients treated with
AML-type chemotherapy. In one series (82), patients with a SCT comorbidity
index score greater than 2 had an early death rate of 29% and median OS was
only 19 weeks. Thus, assessment of comorbidity or frailty using accepted tools
is clearly important for selecting elderly MDS patients for intensive treatment
approaches.
As expected, the main prognostic factors after autologous SCT are not dif-
ferent from those observed in patients treated with intensive chemotherapy alone.
Younger age and normal karyotype or IPSS good-risk cytogenetics have been
constantly associated with a better outcome in different reports (13,36,72,76,77).
Male gender had a negative impact in two reports from the same cooperative group
(75,76), but this association has not been observed by others. The possible effect
of comorbidity on outcome after autologous SCT has not been addressed.
From all the studies on prognostic factors after intensive chemotherapy, it is
clear that MDS patients who are more likely to benefit from intensive chemother-
apy and autologous SCT are those who are younger, fitted, and do not have
poor-risk chromosomal abnormalities.
ALLOGENEIC SCT
Allogeneic SCT is universally considered the treatment of choice for young MDS
patients with an available histocompatible donor, since this modality is able to
510 Sanz and Witte
cure a substantial fraction of these patients. Nonetheless, as discussed in the

previous section, the superiority of allogeneic SCT in comparison to other intensive
approaches still has not been definitely proven. Allogeneic SCT from an HLA-
identical sibling after a standard myeloablative conditioning is applicable in less
than 10% of MDS patients due to the lack of availability of a donor (only 30–35%
have a suitably matched donor), age (a minority are less then 55–60 years), and
performance status/comorbidites. In the last two decades, the use of other sources
of stem cells [e.g., volunteer MUD or from mismatched umbilical cord blood
(UCB) units] and reduced intensity conditioning regimens have greatly expanded
the proportion of patients with potential access to allogeneic SCT. In this section,
we will review the results and prognostic factors of the different modalities of
allogeneic SCT already available, and critically discuss several questions still
unanswered.
Allogeneic SCT from HLA-Identical Siblings

One of the largest series on allogeneic SCT from an HLA-identical sibling, includ-
ing 452 patients with MDS, was reported in 2002 by the International Bone Mar-
row Transplantation Registry (IBMTR) (83). The DFS, relapse risk, and TRM at
3 years were 40%, 23%, and 37%, respectively. It is highly likely that the results
with this transplant modality have improved in recent years thanks to a reduction
in TRM, as reflected by surveys from the IBMTR and EBMT. In an EBMT study,
the 3-year survival and DFS rates were better in patients who underwent SCT after
1989, and this was chiefly due to a decrease in TRM (84). Table 2 summarizes the
main results of several large series of allogeneic SCT from HLA-identical sibling
donors.
Prognostic Factors
The most important prognostic factors after allogeneic SCT from an HLA-identical
sibling are age, stage of disease at transplant (i.e., FAB subtype, or proportion of
Table 2 Summary of the Results of Several Large Studies of Allogeneic SCT from
HLA-Identical Sibling Donors in MDS
References No. of patients DFS TRM Relapse risk
(85) 45 56% at 3 yr NA 16% at 3 yr

(83) 452 40% at 3 yr 37% at 3 yr 23% at 3 yr
(86) 131 34% at 5 yr 44% at 5 yr 39% at 5 yr
(87) 234 50% at 2 yr 42% at 2 yr 13% at 2 yr
(88) 71 32% at 7 yr 39% at 7 yr 48% at 7 yr
(84) 885 36% at 3 yr NA 36% at 3 yr
(89) 621 27% at 3 yr 32% at 3 yr 27% at 3 yr
Abbreviations: DFS, disease-free survival; TRM, transplant-related mortality; NA, not available.
blasts in bone marrow), IPSS cytogenetic risk category, presence of comorbidi-

ties, and time from diagnosis to transplant (83,84,88,90–93). Advanced age, the
presence of comorbidities, and longer disease duration increase TRM.
In a retrospective EBMT study, TRM for patients younger than 20 years,
between 20 and 40 years, and older than 40 years was 30%, 43%, and 50%,
respectively (84). The negative impact on TRM of disease duration longer than
12 months before transplant is present even in patients with lower-risk disease
(91), and may reflect the number of transfusions and subsequent iron overload.
A recent study from the Dana-Farber Cancer Center in Boston showed a strong
inverse correlation between transplantation serum ferritin levels before transplant
and OS in patients with MDS and AML. In patients with MDS, the inferior
OS was attributable to a significant increase in TRM (94). Further, transfusion
dependence at transplant has resulted in significantly higher TRM and reduced
OS in MDS patients without excess of blasts (95). In this study, the recent WHO-
based Prognostic Scoring System (WPSS) had a relevant value in post-transplant
outcome.
The stage of disease at transplant and cytogenetics have a strong association
with relapse risk and DFS. The EBMT reported a 5-year actuarial relapse rate
of 44% in 35 patients with refractory anemia with excess of blasts (RAEB) and
52% in 28 patients with RAEB in transformation (86). Long-term DFS can be
achieved in more than 50% of the patients without excess of blasts—refractory
anemia (RA) and refractory anemia with ringed sideroblasts (RARS)—whereas it
is lower than 20% in those with greater than 20% of blasts (83,84,86,88,90–92).
Relapse rate may be greater than 80% in patients with poor prognosis cytogenetic
abnormalities according to the IPSS, but is lower than 20% in those with good
prognosis chromosomal abnormalities or a normal karyotype (88,92).
Chronic Myelomonocytic Leukemia and Therapy-Related MDS
Data on allogeneic SCT for patients with chronic myelomonocytic leukemia
(CMML; see chap. 13) are limited. In an analysis of 50 CMML patients reported
to the EBMT registry, the estimated 2-year DFS rate was 18%, with a relapse risk
of 42% (96). The published results of allogeneic SCT for patients with t-MDS are
also scarce. A report from Ibrahim Yakoub-Agha and colleagues in Lille, France,
involved 70 patients with t-MDS and t-AML (97). Of these patients, 34% were in
CR at the time of SCT. The 2-year DFS, relapse, and TRM rates were 28%, 42%,
and 49%, respectively, and only 5 of 46 patients with active disease at transplant
were long-term survivors. A large study from Seattle reported on 99 patients (47
t-MDS, 52 t-AML), of whom 65 received marrow from a family member, and 34
received marrow from an unrelated donor. The probabilities of survival, relapse,
and NRM were 13%, 47%, and 78%, respectively (98).
Unsolved Issues
Several relevant questions regarding allogeneic SCT from an HLA-identical sib-
ling remain unclear. First, the best time to proceed to transplant is debatable. An
512 Sanz and Witte
analysis by Cutler and colleagues, using a Markov-type decision model and the
databases of the IBMTR and the IPSS, concluded that OS was maximized by
transplanting upfront in patients with IPSS intermediate-2 or high-risk groups,
and delaying the transplant until disease progression in patients with low or
intermediate-1 IPSS risk groups (99). The authors hypothesized that the opti-
mal timing of SCT in the later cohort was at the time of development of a new
cytogenetic abnormality, the appearance of a clinically important cytopenia, or an
increase in the percentage of marrow blasts.
However, the Cutler et al. algorithm may not be applicable to particular
cases. A delay does not seem convenient for an intermediate-1–risk patient with
transfusion dependent anemia, multilineage dysplasia, and a poor-risk karyotype,
or for a low-risk patient with life-threatening neutropenia or thrombocytopenia
(100). Furthermore, this analysis did not take into account the negative impact on
transplantation outcome that may have longer disease duration or the appearance
of comorbidities.
Although excellent results have been reported with targeted-dose oral busul-
fan (administered to maintain blood levels at 800–900 ng/mL) and cyclophos-
phamide (85), the best conditioning regimen for allogeneic transplantation is not
established. An analysis performed by the IBMTR showed that, despite being
associated with an increased relapse risk, T-cell depletion did not influence out-
come (83). However, an earlier single-center study showed a 73% DFS rate at
2 years after SCT with T-cell–depleted grafts from HLA-identical siblings using
elutriation compared with a 39% DFS rate at 2 years for patients who received a
full graft (101).
The use of AML-type chemotherapy before transplant to reduce tumor bur-
den is controversial (102,103). This strategy could simply serve to select patients
with a higher chance of cure. Patients who attain CR will have a lower risk of
relapse whereas those who do not will have a higher TRM (100). Further, many
patients may die or develop severe organ dysfunction that preclude the transplant.
The EBMT has launched a prospective randomized study to address the possible
benefit of remission induction chemotherapy prior to transplant.
Finally, the best source of stem cells, mobilized peripheral blood or bone
marrow, is uncertain. Engraftment was faster and chronic graft-versus-host disease
(GVHD) more frequent with mobilized peripheral blood than with bone marrow
in a retrospective EBMT study that included 234 patients (87). Actuarial DFS was
better and TRM lower with mobilized peripheral blood, except in patients with
refractory anemia or unfavorable karyotype in whom those outcomes were similar.
Reduced-Intensity Conditioning Regimens
The principal aim of reduced-intensity conditioning (RIC) in MDS is to minimize
the toxicity associated with conventional myeloablative regimens and to harness
the graft-versus-MDS effect of the infused donor lymphocytes. Table 3 summarizes
the main outcomes of several series of RIC for allogeneic SCT from HLA-identical
siblings in MDS. Initial reports have showed an encouraging lower TRM rate
Table 3 Summary of the Results of Several Retrospective Studies of Reduced Intensity

Conditioning in MDS
(104) 24 61% at 1 yr 5% at 1 yr –
(105) 37 38% at 3 yr 27% at 3 yr 32% at 3 yr
(106) 37 66% at 1 yr 5% at 1 yr 28% at 1 yr
(107) 23 39% at 2 yr 31% at 2 yr 17% at 2 yr
(89) 215 33% at 3 yr 22% at 3 yr 45% at 3 yr
Abbreviations: DFS, disease-free survival; TRM, transplant-related mortality.
compared with conventional myeloablative conditioning (104–107). In a study by

Martino and colleagues enrolling 37 patients (median age 57 years) with MDS
and AML, TRM rate was only 5%. This group used a regimen of fludarabine
and busulfan (106). The 1-year progression-free survival was 66% and disease
progression was clearly lower in patients who developed GVHD than in those
without GVHD (13% vs. 58%). These results support the notion that a graft-
versus-MDS/AML response is critical in reducing the risk for relapse after RIC
SCT (106).
Kröger in Hamburg and colleagues reported on 37 MDS patients who were
ineligible for conventionally conditioned SCT (105). The RIC used consisted
of fludarabine, a reduced dosage of busulfan, and antithymocyte globulin. The
overall TRM rate was 27%, with a significantly higher rate of mortality in those
with poor-risk cytogenetics (75% vs. 20%) or with an HLA-MUD (45% vs. 12%).
In total, 32% of patients relapsed. The actuarial DFS rate at 3 years was 38%, with
a median follow-up of 20 months.
A group from King’s College Hospital, London, reported more favorable
results following conditioning with fludarabine, busulfan, and alemtuzumab in 62
MDS patients (104). The 1-year DFS rates were 61% and 59% in patients with
matched sibling donors (n = 24) and unrelated donors (n = 38), respectively.
These favorable results might be explained by the low-estimated 1-year TRM rate
[5% for recipients of sibling donors and 21% for recipients of volunteer unrelated
donors (VUD)], the relatively high number of patients with less than 5% marrow
blasts at the time of SCT (greater than 75%), the high number of patients who
received donor lymphocyte infusions (67% of sibling recipients and 26% of VUD
recipients), and the relatively short period of follow-up. No long-term survival
was observed in patients with progressive disease (104).
The EBMT has recently published the largest series to date of RIC for MDS
patients (89). This retrospective study analyzed the outcomes of 836 MDS patients
who underwent allogeneic SCT from an HLA-identical sibling donor. Outcomes
were compared according to two types of conditioning: RIC in 215 patients,
and standard myeloablative conditioning (SMC) in 621 patients. In multivariate
analyses, the 3-year relapse rate was significantly higher after RIC (45% vs. 27%,
514 Sanz and Witte
Figure 3 Cumulative incidence at 36 months of non-relapse mortality (NRM) and relapse

(REL) after standard myeloablative conditioning (standard) or reduced-intensity condition-
ing (RIC) regimens in patients with MDS undergoing transplants from an HLA-identical
sibling. Source: Adapted From Ref. 89.
p = 0.001), the 3-year NRM was lower after RIC (22% vs. 32%, p = 0.015), and
the 3-year probabilities of DFS (39% with SMC vs. 33% with RIC), and OS (45%
vs. 41%, respectively) were similar in both groups. Figure 3 shows the cumulative
incidence of NRM and relapse rate at 3 years according to the type of conditioning
in this study. The lower 3-year NRM after RIC is encouraging, since patients in
the RIC group were older (median age, 56 years in the RIC and 45 years in the
SMC group; p ⬍ 0.0001) and had more adverse prognostic factors at transplant.
Similar results were observed in subanalyses of different age groups and stage of
the disease at transplant (89).
A recent study from Seattle, adjusting for comorbidity and stage of disease
at transplant has also reported similar outcomes after SMC (n = 452) and RIC
(n = 125) in patients with AML (n = 391) or MDS (n = 186) who underwent
an allogeneic SCT (93). All these data clearly suggest that RIC is a valuable
alternative for MDS patients with advanced age or comorbidities who would not
be good candidates for SMC.
Allogeneic SCT from Volunteer Unrelated Donors

Several reports have demonstrated that transplantation from matched VUDs is a
feasible and potentially curative strategy for MDS patients who lack an HLA-
identical sibling. Estimated overall DFS probabilities in published studies range
from 28% to 59% (85,108,109). However, all available data show that, despite the
young median age of the patients in all published series, NRM remains the most
important problem and patients’ age is a critical factor for this specific endpoint.
A major focus of future research should be the reduction of NRM, with the aim
of being able to offer this type of transplantation to more patients who are in need
of it, most of whom are elderly. In this sense, evaluation of the usefulness of RIC
is clearly required. Preliminary results with RIC for allogeneic SCT from VUDs
are already available.
Clinical Results
The largest series of allogeneic SCT from VUDs in MDS was reported by the
National Marrow Donor Program (NMDP) (109). The study included 510 patients,
with a median age of 38 years. The study was seriously limited by the wide
inclusion period and high heterogeneity in transplantation-related procedures,
such as conditioning regimen, GVHD prophylaxis, and use of T-cell depletion.
However, it offers a clear overview of the possible role of VUD SCT in the
treatment of MDS. At 2 years, the probability of DFS was 29%, the cumulative
incidence of relapse was 14%, and the cumulative incidence of TRM was 54%.
Data from other series with fewer patients are also consistent with a high
TRM and an acceptable incidence of relapse. In an early study on the results of
allogeneic SCT from VUDs in 52 MDS patients, the 2-year DFS, relapse risk, and
NRM rates were 38%, 28%, and 48%, respectively (110). In the EBMT experience,
among 118 patients who received a transplantation from a VUD, the DFS, relapse
risk, and TRM rates were 28%, 35%, and 58%, respectively (Table 4) (108). A
more recent series by H. Joachim Deeg and colleagues in Seattle described 64
MDS patients, and reported a 3-year relapse-free survival rate of 59% (85). The
conditioning regimen in this study consisted of oral busulfan targeted to plasma
concentrations of 800 to 900 ng/mL and cyclophosphamide. The 3-year cumulative
incidence of relapse was 11% and the NRM was 30% (13% at day 100). These
results are encouraging, particularly given that the patients in this study had a
median age of 46 years, greater than in other reports. The results were clearly
inferior in 11 recipients of mismatched grafts, in whom 3-year DFS and NRM
rates were 27% and 52%, respectively (85). As discussed earlier in this chapter,
Table 4 Summary of the Results of Some Large Studies of Allogeneic SCT from
Volunteer Unrelated Donors in MDS
(110) 45 38% at 2 yr NA 28% at 2 yr

(108) 118 28% at 2 yr 58% at 2 yr 35% at 2 yr
(109) 510 29% at 2 yr 54% at 2 yr 14% at 5 yr
(84) 198 25% at 3 yr 58% at 3 yr 41% at 3 yr
(85) 64 59% at 3 yr 30% at 3 yr 11% at 3 yr
Abbreviations: DFS, disease-free survival; TRM, transplant-related mortality; NA, not available.
516 Sanz and Witte
this conditioning regimen appears to be superior to other regimens in reducing

NRM in MDS patients, but these data require confirmation.
Prognostic Factors
The disease stage at SCT, whether measured according to the FAB classification
subtype or as the proportion of blasts in bone marrow, is probably the most
important variable influencing the relapse rate after SCT from VUD (85,108–
110). In the NMDP series, patients with RAEB-t and AML secondary to MDS
had a relative risk for graft failure of 4.92 and 7.67, respectively, as compared with
the remaining patients (109). In the other three studies, the probability of relapse
increased from 3–13% for patients with RA or RARS to 29–49% for patients with
RAEB-t or AML (98,108,110). Other factors clearly related to the relapse risk are
IPSS score (85), IPSS cytogenetic risk category (85), etiology of the MDS (85),
and the development of acute GVHD after SCT (108,109). Patients with a low
IPSS score, good-risk IPSS cytogenetics, de novo MDS, and those developing
acute GVHD had a lower risk for relapse. This last finding suggests that the VUD
transplant modality has a significant graft-versus-MDS effect.
In contrast to the report by the Fred Hutchinson Cancer Research Center
(85), recipients’ age has shown a strong impact on TRM both in studies from
registries (108,109) and in an earlier study from the Hutchinson Center (110). In
the NMDP series, patients aged ≥20 years had a significantly higher NRM than
younger patients (109). In the EBMT experience, NRM at 2 years was 40% for
patients aged ⬍18 years, 61% for patients aged 18 to 35 years, and 81% for those
aged ⬎35 years (108). Other factors that have been reported to influence NRM
include time elapsed from diagnosis to SCT (108–110), degree of HLA dispar-
ity between donor and recipient (85,109), donor age (109) and the presence of
acute GVHD (85,109), neutropenia at SCT (110), and recipient cytomegalovirus
(CMV) serology (109,110). A longer interval from diagnosis to SCT, HLA mis-
matching, increasing age of the donor, development of grades II–IV acute GVHD,
and the presence of neutropenia at SCT results in higher NRM. Recipient CMV
serology has yielded conflicting results. In the NMDP experience, CMV seroneg-
ativity was a favorable factor for TRM (89); in the earlier Hutchinson Center
study, however, CMV seronegativity was unexpectedly associated with a higher
NRM (110).
The status of the disease at transplantation, the time from diagnosis to
transplantation, and MDS etiology have shown close relationships with DFS after
VUD transplantation for MDS. Patients with FAB subtypes without excess of
blasts (85,109,110), or belonging to a lower-risk IPSS category (85), experienced
higher DFS rates. Similarly, early transplantation (i.e., less than 9–12 months
from diagnosis to SCT) resulted in longer DFS (108–110). Patients with MDS
secondary to chemo/radiotherapy have poorer DFS rates than patients with de
novo MDS (85). There also appears to be an improvement in DFS rates in patients
who have been transplanted in recent years (109). Other characteristics that have
been associated with better DFS rates are younger recipients’ age (108,109),
greater cell dose (109), CMV seronegativity (109), and grades 0–I acute GVHD
(85,109).
Umbilical Cord Blood Transplants

Umbilical cord blood (UCB) transplantation from unrelated donors has emerged
as a clear alternative for children and adults with acute leukemia who lack an
HLA-matched donor, particularly for those who require urgent SCT (111). The
main advantages of UCB are the shorter time from the start of the search to trans-
plantation, and the lower degree of HLA matching required for transplantation.
The major drawback is the lower content of hematopoietic progenitor cells in a
typical UCB unit as compared with a graft from an adult donor. Experience with
UCB transplantations from unrelated donors for MDS patients is still very scarce.
Jun Ooi and colleagues in Japan reported the outcomes of 13 patients with
advanced and transformed MDS (112). Despite the very advanced disease status
at transplantation, 10 out of 13 patients were alive, and in CR with an estimated
76% DFS rate at 2 years. The Eurocord Cooperative Group has reported a pre-
liminary experience with UCB transplantation in 50 MDS patients (113). With
a median follow-up of 21 months, 2-year probabilities of relapse, DFS, and OS
were 32%, 29%, and 34%, respectively. These preliminary data suggest that UCB
transplantation could expand the access to transplant of MDS patients without a
matched donor.
CONCLUSIONS
Allogeneic SCT is the treatment of choice for the majority of young patients
with MDS who have a histocompatible donor (either a sibling or an unrelated
volunteer donor). The use of RIC instead of SMC regimens may be valuable for
patients aged 50 to 65 years or with comorbidities. High-risk MDS patients without
a matched donor should be considered for intensive chemotherapy. Outcomes
with autologous SCT or further intensive chemotherapy after achievement of CR
with intensive chemotherapy appear comparable. Intensive approaches benefit
especially young patients, those with good performance status, and those with
good prognosis cytogenetics.
Patients of very advanced age with substantial comorbidities, particularly if
they have a poor-risk karyotype, are not good candidates for intensive therapy, and
should instead be considered for alternative low-intensity approaches or newer
therapeutic agents in the context of well-designed clinical trials. Figure 4 offers
a potential algorithm for the selection of treatment in high-risk MDS patients.
Intensive approaches, despite their substantial morbidity and mortality rates, are
currently the only available therapies that offer MDS patients a chance of cure.
518
Age <65 years Age >65 years

No significant comorbidities Significant comorbidities
Available donor
(HLA-identical sibling or matched VUD or UCB)
Yes No
Good-risk cytogenetics Poor-risk cytogenetics
Intensive chemotherapy
Allogeneic SCT
with or without Azacitidine/decitabine
(consider RIC if age >50 years or comorbidities;
autologous SCT in CR1 or
consider intensive chemotherapy before
or clinical trial
transplant if bone marrow blasts >10%)
clinical trial
Figure 4 Suggested algorithm for the treatment of patients with high-risk MDS or in selected cases of
lower risk MDS (i.e., life-threatening neutropenia or thrombocitopenia, bone marrow fibrosis, or poor-risk
cytogenetics). Abbreviations: VUD, volunteer unrelated donor; UCB, umbilical cord blood; SCT, stem cell
transplantation; RIC, reduced-intensity conditioning; CR1, first complete remission.
Sanz and Witte
REFERENCES
1. San Miguel JF, Sanz GF, Vallespı́ T, et al. Myelodysplastic syndromes. Crit Rev
Oncol Hematol 1996; 23:57–93.
2. Germing U, Strupp C, Kuendgen A, et al. No increase in age-specific incidence of
3. Ma X, Does M, Raza A, et al. Myelodysplastic syndromes: Incidence and survival
in the United States. Cancer 2007; 109:1536–1543.
4. Pedersen-Bjergaard J, Specht L, Larsen SO, et al. Risk of therapy-related leukaemia
and preleukaemia after Hodgkin’s disease. Relation to age, cumulative dose of
alkylating agents, and time from chemotherapy. Lancet 1987; 2:83–88.
5. Rubin CM, Arthur DC, Woods WG, et al. Therapy-related myelodysplastic syn-
drome and acute myeloid leukemia in children: Correlation between chromosomal
abnormalities and prior therapy. Blood 1991; 78:2982–2988.
6. Pedersen-Bjergaard J, Philip P, Larsen SO, et al. Therapy-related myelodysplasia
and acute myeloid leukemia. Cytogenetic characteristics of 115 consecutive cases
and risk in seven cohorts of patients treated intensively for malignant diseases in the
Copenhagen series. Leukemia 1993; 7:1975–1986.
8. Brunning RD, Head D, Bennett JM, et al. Myelodysplastic syndromes: Introduction.
In: Jaffe ES, Harris NL, Stein H, Vardiman JW, eds. Tumours of Haematopoietic
and Lymphoid Tissues. Lyon, France: IARC Press, 2001:63–67.
9. Sanz GF, Sanz MA, Vallespı́ T, et al. Two regression models and a scoring system
for predicting survival and planning treatment in myelodysplastic syndromes. A
multivariate analysis of prognostic factors in 370 patients. Blood 1989; 74:395–
403.
11. National Comprehensive Cancer Network (NCCN). Myelodysplastic syndromes.
NCCN clinical practice guidelines in Oncology. V.2.2008. www.nccn.org (accessed
December 1, 2007).
12. Barrett AJ, Savani BN. Allogeneic stem cell transplantation for myelodysplastic
syndrome. Semin Hematol 2008; 45:49–59.
13. Oosterveld M, Suciu S, Verhoef G,et al. The presence of an HLA-identical sibling
donor has no impact on outcome of patients with high-risk MDS or secondary AML
(sAML) treated with intensive chemotherapy followed by transplantation: Results of
a prospective study of the EORTC, EBMT, SAKK and GIMEMA Leukemia Groups
(EORTC study 06921). Leukemia 2003; 17:859–868.
azacitidine in patients with the myelodysplastic syndrome: A study of the cancer
and leukemia group B. J Clin Oncol 2002; 20:2429–2440.
15. Fenaux P, Mufti GJ, Santini V,et al. Azacitidine (AZA) treatment prolongs overall
(CCR): Results of the AZA-001 phase III study. Blood 2007; 110:817 (abstract).
myelodysplastic syndromes: Results of a phase III randomized study. Cancer 2006;
106:1794–1803.
520 Sanz and Witte
17. List A, Dewald G, Benett J, et al. Lenalidomide in myelodysplastic syndrome with

5q deletion. N Engl J Med 2006; 355:1456–1465.
18. Arber DA, Stein AS, Carter NH, 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; 119:672–680.
19. Grimwade D, Walker H, Oliver F, et al. The importance of diagnostic cytogenetics
on outcome in AML: Analysis of 1612 patients entered into the MRC AML 10 trial.
The Medical Research Council Adult and Children’s Leukaemia Working Parties.
Blood 1998; 92:2322–2333.
20. Byrd JC, Mrózek K, Dodge RK, 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; 100:4325–4336.
21. Fenaux P. Chromosome and molecular abnormalities in myelodysplastic syndromes.
Int J Hematol 2001; 73:429–437.
22. Vallespi T, Imbert M, Mecucci C, et al. Diagnosis, classification, and cytogenetics
23. Cilloni D, Gottardi E, Messa F, et al. Significant correlation between the degree of
WT1 expression and the International Prognostic Scoring System Score in patients
with myelodysplastic syndromes. J Clin Oncol 2003; 21:1988–1995.
24. Shih LY, Lin TL, Wang PN, et al. Internal tandem duplication of fms-like tyrosine
kinase 3 is associated with poor outcome in patients with myelodysplastic syndrome.
Cancer 2004; 101:989–998.
892.
26. De Witte T, Muus P, De Pauw B, et al. Intensive antileukemic treatment of patients
younger than 65 years with myelodysplastic syndromes and secondary acute myel-
ogenous leukemia. Cancer 1990; 66:831–833.
27. Estey E, Thall P, Beran M, et al. Effect of diagnosis (refractory anemia with excess
blasts, refractory anemia with excess blasts in transformation, or acute myeloid
leukemia [AML] on outcome of AML-type chemotherapy. Blood 1997; 90:2969–
2973.
28. Bernstein SH, Brunetto VL, Davey FR, et al. Acute myeloid leukemia-type
chemotherapy for newly diagnosed patients without antecedent cytopenias hav-
ing myelodysplastic syndrome as defined by French-American-British criteria: A
Cancer and Leukemia Group B study. J Clin Oncol 1996; 14:2486–2493.
29. Beran M, Shen Y, Kantarjian H, et al. High-dose chemotherapy in high-risk
myelodysplastic syndrome covariate-adjusted comparison of five regimens. Can-
cer 2001; 92:1999–2015.
30. Hast R, Hellström-Lindberg E, Ohm L, et al. No benefit from adding GM-CSF to
induction chemotherapy in transforming myelodysplastic syndromes: Better out-
come in patients with less proliferative disease. Leukemia 2003; 17:1827–1833.
32. Mertelsmann R, Tzvi Thaler H, To L, et al. Morphological classification, response
to therapy, and survival in 263 adult patients with acute nonlymphoblastic leukemia.
Blood 1980; 56:773–783.
33. Armitage JO, Dick FR, Needleman SW, et al. Effect of chemotherapy for the dys-
myelopoietic syndrome. Cancer Treat Rep 1981; 65:601–603.
34. Knipp S, Hildebrand B, Kundgen A, et al. Intensive chemotherapy is not recom-
mended for patients aged ⬎60 years who have myelodysplastic syndromes or acute
myeloid leukemia with high-risk karyotypes. Cancer 2007; 110:345–352.
35. Kantarjian H, Beran M, Cortes J, et al. Long-term follow-up results of the combi-
nation of topotecan and cytarabine and other intensive chemotherapy regimens in
myelodysplastic syndrome. Cancer 2006; 106:1099–1109.
36. de Witte T, Suciu S, Verhoef G, et al. Intensive chemotherapy followed by allogeneic
or autologous stem cell transplantation for patients with myelodysplastic syndromes
(MDSs) and acute myeloid leukemia following MDS. Blood 2001; 98:2326–2331.
37. Okamoto T, Kanamaru A, Shimazaki C, et al. Combination chemotherapy with
risk factor-adjusted dose attenuation for high-risk myelodysplastic syndrome and
resulting leukemia in the multicenter study of the Japan Adult Leukemia Study
Group (JALSG): Results of an interim analysis. Int J Hematol 2000; 72:200–205.
38. Latagliata R, Breccia M, Pulsoni A, et al. Acute myeloblastic leukemia secondary
to myelodysplasia (MDS-AML): A comparison of remission induction with three
drugs versus standard two-drugs induction. Leuk Lymphoma 2000; 36:539–541.
39. Invernizzi R, Pecci A, Rossi G, et al. Idarubicin and cytosine arabinoside in the
induction and maintenance therapy of high-risk myelodysplastic syndromes. Haema-
tologica 1997; 82:660–663.
40. Ruutu T, Hänninen A, Järventie G, et al. Intensive chemotherapy of poor prognosis
myelodysplastic syndromes (MDS) and acute myeloid leukemia following MDS
with idarubicin and cytarabine. Leuk Res 1997; 21:133–138.
41. Langston AA, McMillan S, Lonial S, et al. A phase I trial of ara-C, topotecan,
and gemtuzumab ozogamicin (Mylotarg) R for advanced MDS and secondary or
relapsed AML. Blood 2004; 104:1814 (abstract).
42. Kim Y, Cheong JW, Kim JS, et al. Extended study for topotecan, idarubicin, and
intermediate-dose cytarabine combination chemotherapy in patients with refractory
or relapsed acute myelogenous leukemia and high-risk myelodysplastic syndrome.
Blood 2005; 106:4609 (abstract).
43. Murali S, Winton EF, McMillan S, et al. Tolerability and anti-leukemic activity
of ara-c, topotecan, and gemtuzumab ozogamicin [ATGO] in advanced MDS and
AML: Interim analysis of phase I/II data. Blood 2006; 108:1960 (Abstr).
44. Kang P, Seiter K, Liu D, et al. Phase II study of gemtuzumab ozogamycin (GO)
and cytarabine (ARA-C) in patients with high risk MDS and AML. Blood 2004;
104:4723 (abstract).
45. Prébet T, Ducastelle S, Debotton S, et al. A phase II study of intensive chemotherapy
with fludarabine, cytarabine, and mitoxantrone in P glycoprotein-negative high-risk
myelodysplastic syndromes. Hematol J 2004; 5:209–215.
46. Weihrauch MR, Staib P, Seiberlich B, et al. Phase I/II clinical study of topotecan
and cytarabine in patients with myelodysplastic syndrome, chronic myelomonocytic
leukemia and acute myeloid leukemia. Leuk Lymphoma 2004; 45:699–704.
47. Estey EH, Thall PF, Giles FJ, et al. Gemtuzumab ozogamicin with or without
interleukin 11 in patients 65 years of age or older with untreated acute myeloid
leukemia and high-risk myelodysplastic syndrome: Comparison with idarubicin
plus continuous-infusion, high-dose cytosine arabinoside. Blood 2002; 99:4343–
4349.
522 Sanz and Witte
48. Lee ST, Jang JH, Suh HC, et al. Idarubicin, cytarabine, and topotecan in patients with
refractory or relapsed acute myelogenous leukemia and high-risk myelodysplastic
syndrome. Am J Hematol 2001; 68:237–245.
49. Beran M, Estey E, O’Brien S, et al. Topotecan and cytarabine is an active combination
regimen in myelodysplastic syndromes and chronic myelomonocytic leukemia. J
Clin Oncol 1999; 17:2819–2830.
50. Sanz GF, Mena-Duran AV, Ribera JM, et al. Autologous stem cell transplantation
after FLAG-IDA chemotherapy for high-risk myelodysplastic syndromes (MDS) and
acute myeloid leukemias secondary to MDS (sAML) does not improve outcome: A
PETHEMA experience in 103 patients. Blood 2005; 106:793 (abstract).
51. Parker JE, Pagliuca A, Mijovic A, et al. Fludarabine, cytarabine, G-CSF and idaru-
bicin (FLAG-IDA) for the treatment of poor-risk myelodysplastic syndromes and
acute myeloid leukaemia. Br J Haematol 1997; 99:939–944.
52. Ossenkoppele GJ, Graveland WJ, Sonneveld P, et al. The value of fludarabine in
addition to ARA-C and G-CSF in the treatment of patients with high-risk myelodys-
plastic syndromes and AML in elderly patients. Blood 2004; 103:2908–2913.
53. Hofmann WK, Heil G, Zander C, et al. Intensive chemotherapy with idarubicin,
cytarabine, etoposide, and G-CSF priming in patients with advanced myelodysplastic
syndrome and high-risk acute myeloid leukemia. Ann Hematol 2004; 83:498–503.
54. Hast R, Hellström-Lindberg E, Ohm L, et al. No benefit from adding GM-CSF to
induction chemotherapy in transforming myelodysplastic syndromes: Better out-
come in patients with less proliferative disease. Leukemia 2003; 17:1827–1833.
55. Amadori S, Suciu S, Jehn U, et al. Use of glycosylated recombinant human G-CSF
(lenograstim) during and/or after induction chemotherapy in patients 61 years of
age and older with acute myeloid leukemia: Final results of AML-13, a randomized
phase-3 study. Blood 2005; 106:27–34.
56. Ferrara F, Leoni F, Pinto A, et al. Fludarabine, cytarabine, and granulocyte-colony
stimulating factor for the treatment of high risk myelodysplastic syndromes. Cancer
1999; 86:2006–2013.
57. Virchis A, Koh M, Rankin P, et al. Fludarabine, cytosine arabinoside, granulocyte-
colony stimulating factor with or without idarubicin in the treatment of high risk
acute leukaemia or myelodysplastic syndromes. Br J Haematol 2004; 124:26–32.
58. Ossenkoppele GJ, Van Der Holt B, Verhoef GE, et al. A randomized study of
granulocyte colony-stimulating factor applied during and after chemotherapy in
patients with poor risk myelodysplastic syndromes: A report from the HOVON
Cooperative Group. Leukemia 1999; 13:1207–1213.
59. Estey EH, Thall PF, Pierce S, et al. Randomized phase II study of fludarabine +
cytosine arabinoside + idarubicin ± all-trans retinoic acid ± granulocyte colony-
stimulating factor in poor prognosis newly diagnosed acute myeloid leukemia and
myelodysplastic syndrome. Blood 1999; 93:2478–2484.
60. Bernasconi C, Alessandrino EP, Bernasconi P, et al. Randomized clinical study
comparing aggressive chemotherapy with or without G-CSF support for high-risk
myelodysplastic syndromes or secondary acute myeloid leukaemia evolving from
MDS. Br J Haematol 1998; 102:678–683.
61. Verbeek W, Wörmann B, Koch P, et al. Results of a randomized double-blind placebo-
controlled trial evaluating sequential high-dose cytosine arabinoside/mitoxantrone
chemotherapy with or without granulocyte/macrophage-colony-stimulating factor in
high-risk myelodysplastic syndromes. J Cancer Res Clin Oncol 1999; 125:369–374.
62. Leith CP, Kopecky KJ, Chen IM, et al. Frequency and clinical significance of the
expression of the multidrug resistance proteins MDR1/P-glycoprotein, MRP1, and
LRP in acute myeloid leukemia: A Southwest Oncology Group Study. Blood 1999;
94:1086–1099.
63. Wattel E, Solary E, Hecquet B, et al. Quinine improves the results of intensive
chemotherapy in myelodysplastic syndromes expressing P glycoprotein: Results of
a randomized study. Br J Haematol 1998; 102:1015–1024.
64. Matsouka P, Pagoni M, Zikos P, et al. Addition of cyclosporin-A to chemotherapy
in secondary (post-MDS) AML in the elderly. A multicenter randomized trial of
the Leukemia Working Group of the Hellenic Society of Hematology. Ann Hematol
2006; 85:250–256.
65. Greenberg PL, Lee SJ, Advani R, et al. Mitoxantrone, etoposide, and cytarabine with
or without valspodar in patients with relapsed or refractory acute myeloid leukemia
and high-risk myelodysplastic syndrome: A phase III trial (E2995). J Clin Oncol
2004; 22:1078–1086.
66. Cripe LD, Li X, Litzow M, et al. A randomized, placebo-controlled, double blind
trial of the MDR modulator, zosuquidar, during conventional induction and post-
remission therapy for pts ⬎ 60 years of age with newly diagnosed acute myeloid
leukemia (AML) or high-risk myelodysplastic syndrome (HR-MDS): ECOG 3999.
Blood 2006; 108:423 (abstract).
67. Sekeres MA, Stone RM, Zahrieh D, et al. Decision-making and quality of life in
older adults with acute myeloid leukemia or advanced myelodysplastic syndrome.
Leukemia 2004; 18:809–816.
68. Pitako JA, Haas PS, Van Den Bosch J, et al. Quantification of outpatient management
and hospitalization of patients with high-risk myelodysplastic syndrome treated with
low-dose decitabine. Ann Hematol 2005; 84(Suppl. 1):25–31.
69. Deschler B, de Witte T, Mertelsmann R, et al. Treatment decision-making for older
patients with high-risk myelodysplastic syndrome or acute myeloid leukemia: Prob-
lems and approaches. Haematologica 2006; 91:1513–1522.
70. Delforge M, Demuynck H, Vandenberghe P, et al. Polyclonal primitive hematopoietic
progenitors can be detected in mobilized peripheral blood from patients with high-
risk myelodysplastic syndromes. Blood 1995; 86:3660–3667.
71. Demuynck H, Delforge M, Verhoef GE, et al. Feasibility of peripheral blood pro-
genitor cell harvest and transplantation in patients with poor-risk myelodysplastic
syndromes. Br J Haematol 1996; 92:351–359.
72. De Witte T, Van Biezen A, Hermans J, et al. Autologous bone marrow trans-
plantation for patients with myelodysplastic syndrome (MDS) or acute myeloid
leukemia following MDS. Chronic and Acute Leukemia Working Parties of the
European Group for Blood and Marrow Transplantation. Blood 1997; 90:3853–
3857.
73. Kroger N, Brand R, van Biezen A, et al. Autologous stem cell transplantation
for therapy-related acute myeloid leukemia and myelodysplastic syndrome. Bone
74. Al-Ali HK, Brand R, van Biezen A, et al. A retrospective comparison of autol-
ogous and unrelated donor hematopoietic cell transplantation in myelodysplastic
syndrome and secondary acute myeloid leukemia: A report on behalf of the Chronic
Leukemia Working Party of the European Group for Blood and Marrow Transplan-
tation (EBMT). Leukemia 2007; 21:1945–1951.
524 Sanz and Witte
75. Wattel E, Solary E, Leleu X, et al. A prospective study of autologous bone marrow or
peripheral blood stem cell transplantation after intensive chemotherapy in myelodys-
plastic syndromes. Groupe Français des Myélodysplasies. Group Ouest-Est d’étude
des Leucémies aiguës myéloı̈des. Leukemia 1999; 13:524–529.
76. Ducastelle S, Adès L, Gardin C, et al. Long-term follow-up of autologous stem
cell transplantation after intensive chemotherapy in patients with myelodysplastic
syndrome or secondary acute myeloid leukemia. Haematologica 2006; 91:373–376.
77. Oosterveld M, Muus P, Suciu S, et al. Chemotherapy only compared to chemotherapy
followed by transplantation in high risk myelodysplastic syndrome and secondary
acute myeloid leukemia; two parallel studies adjusted for various prognostic factors.
Leukemia 2002; 16:1615–1621.
78. Kantarjian H, O’brien S, Cortes J, et al. Results of intensive chemotherapy in
998 patients age 65 years or older with acute myeloid leukemia or high-risk
myelodysplastic syndrome: Predictive prognostic models for outcome. Cancer 2006;
106:1090–1098.
79. Fenaux P, Morel P, Rose C, et al. Prognostic factors in adult de novo myelodysplastic
syndromes treated by intensive chemotherapy. Br J Haematol 1991; 77:497–501.
80. de Witte T, Suciu S, Peetermans M, et al. Intensive chemotherapy for poor prognosis
myelodysplasia (MDS) and secondary acute myeloid leukemia (sAML) following
MDS of more than 6 months duration. A pilot study by the Leukemia Cooperative
Group of the European Organisation for Research and Treatment in Cancer (EORTC-
LCG). Leukemia 1995; 9:1805–1811.
81. Etienne A, Esterni B, Charbonnier A, et al. Comorbidity is an independent predictor
of complete remission in elderly patients receiving induction chemotherapy for acute
myeloid leukemia. Cancer 2007; 109:1376–1383.
82. Giles FJ, Borthakur G, Ravandi F, et al. The haematopoietic cell transplantation
comorbidity index score is predictive of early death and survival in patients over 60
years of age receiving induction therapy for acute myeloid leukaemia. Br J Haematol
2007; 136:624–627.
83. Sierra J, Pérez WS, Rozman C, et al. Bone marrow transplantation from HLA-
identical siblings as treatment for myelodysplasia. Blood 2002; 100:1997–2004.
84. de Witte T, Hermans J, Vossen J, et al. Haematopoietic stem cell transplantation for
patients with myelodysplastic syndromes and secondary acute myeloid leukaemias:
A report on behalf of the Chronic Leukaemia Working Party of the European Group
for Blood and Marrow Transplantation (EBMT). Br J Haematol 2000; 110:620–
630.
lated donors in patients with myelodysplastic syndrome. Blood 2002; 100:1201–
1207.
86. Runde V, de Witte T, Arnold R, et al. Bone marrow transplantation from HLA-
identical siblings as first-line treatment in patients with myelodysplastic syndromes:
Early transplantation is associated with improved outcome. Chronic Leukemia Work-
ing Party of the European Group for Blood and Marrow Transplantation. Bone
87. Guardiola P, Runde V, Bacigalupo A, et al. Retrospective comparison of
bone marrow and granulocyte colony-stimulating factor-mobilized peripheral
blood progenitor cells for allogeneic stem cell transplantation using HLA
identical sibling donors in myelodysplastic syndromes. Blood 2002; 99:4370–
4378.
88. Sutton L, Chastang C, Ribaud P, et al. Factors influencing outcome in de novo
myelodysplastic syndromes treated by allogeneic bone marrow transplantation: A
long-term study of 71 patients Societe Francaise de Greffe de Moelle. Blood 1996;
88:358–365.
89. Martino R, Iacobelli S, Brand R, et al. Retrospective comparison of reduced-intensity
conditioning and conventional high-dose conditioning for allogeneic hematopoietic
stem cell transplantation using HLA-identical sibling donors in myelodysplastic
syndromes. Blood 2006; 108:836–846.
90. Appelbaum FR, Barrall J, Storb R, et al. Bone marrow transplantation for patients
with myelodysplasia. Pretreatment variables and outcome. Ann Intern Med 1990;
112:590–597.
for refractory anemia: A comparison of two preparative regimens and analysis of
prognostic factors. Blood 1996; 87:51–58.
92. Nevill TJ, Fung HC, Shepherd JD, et al. Cytogenetic abnormalities in primary
myelodysplastic syndrome are highly predictive of outcome after allogeneic bone
93. Sorror ML, Sandmaier BM, Storer BE, et al. Comorbidity and disease status based
risk stratification of outcomes among patients with acute myeloid leukemia or
myelodysplasia receiving allogeneic hematopoietic cell transplantation. J Clin Oncol
2007; 25:4246–4254.
94. Armand P, Kim HT, Cutler CS, et al. Prognostic impact of elevated pretransplantation
2007; 109:4586–4588.
95. Alessandrino EP, Della Porta MG, Bacigalupo A, et al. WHO classification and
WPSS predict post-transplant outcome in patients with myelodysplastic syndrome:
A study from the GITMO (Gruppo Italiano Trapianto di Midollo Osseo). Blood 2008
(Epub ahead of print).
96. Kroger N, Zabelina T, Guardiola P, et al. Allogeneic stem cell transplantation
of adult chronic myelomonocytic leukaemia. A report on behalf of the Chronic
Leukaemia Working Party of the European Group for Blood and Marrow Transplan-
tation (EBMT). Br J Haematol 2002; 118:67–73.
97. Yakoub-Agha I, de la Salmoniere P, Ribaud P, et al. Allogeneic bone marrow
marrow transplantation. J Clin Oncol 2000; 18:963–971.
98. Witherspoon RP, Deeg HJ. Allogeneic bone marrow transplantation for secondary
leukemia or myelodysplasia. Haematologica 1999; 84:1085–1087.
585.
100. Deeg HJ. Optimization of transplant regimens for patients with myelodysplastic
syndrome (MDS). Hematol Am Soc Hematol Educ Program 2005:167–173.
526 Sanz and Witte
101. Mattijssen V, Schattenberg A, Schaap N, et al. Outcome of allogeneic bone mar-

row transplantation with lymphocyte-depleted marrow grafts in adult patients with
myelodysplastic syndromes. Bone Marrow Transplant 1997; 19:791–794.
102. Nakai K, Kanda Y, Fukuhara S, et al. Value of chemotherapy before allogeneic
hematopoietic stem cell transplantation from an HLA-identical sibling donor for
myelodysplastic syndrome. Leukemia 2005; 19:396–401.
103. Anderson JE, Gooley TA, Schoch G, et al. Stem cell transplantation for secondary
acute myeloid leukemia: Evaluation of transplantation as initial therapy or following
induction chemotherapy. Blood 1997; 89:2578–2585.
104. Ho AY, Pagliuca A, Kenyon M, et al. Reduced-intensity allogeneic hematopoietic
stem cell transplantation for myelodysplastic syndrome and acute myeloid leukemia
with multilineage dysplasia using fludarabine, busulphan, and alemtuzumab (FBC)
conditioning. Blood 2004; 104:1616–1623.
105. Kröger N, Bornhäuser M, Ehninger G, et al. Allogeneic stem cell transplantation
after a fludarabine/busulfan-based reduced-intensity conditioning in patients with
myelodysplastic syndrome or secondary acute myeloid leukemia. Ann Hematol
2003; 82:336–342.
106. Martino R, Caballero MD, Simón JA, et al. Evidence for a graft-versus-leukemia
effect after allogeneic peripheral blood stem cell transplantation with reduced-
intensity conditioning in acute myelogenous leukemia and myelodysplastic syn-
dromes. Blood 2002; 100:2243–2245.
107. Parker JE, Shafi T, Pagliuca A, et al. Allogeneic stem cell transplantation in the
myelodysplastic syndromes: Interim results of outcome following reduced-intensity
conditioning compared with standard preparative regimens. Br J Haematol 2002;
119:144–154.
108. Arnold R, de Witte T, van Biezen A, et al. Unrelated bone marrow transplantation in
patients with myelodysplastic syndromes and secondary acute myeloid leukemia: An
EBMT survey. European Blood and Marrow Transplantation Group. Bone Marrow
Transplant 1998; 21:1213–1216.
109. Castro-Malaspina H, Harris RE, Gajewski J, et al. Unrelated donor marrow trans-
plantation for myelodysplastic syndromes: Outcome analysis in 510 transplants
facilitated by the National Marrow Donor Program. Blood 2002; 99:1943–1951.
110. Anderson JE, Anasetti C, Appelbaum FR, et al. Unrelated donor marrow transplan-
tation for myelodysplasia (MDS) and MDS-related acute myeloid leukaemia. Br J
Haematol 1996; 93:59–67.
111. Sanz MA. Cord-blood transplantation in patients with leukemia—A real alternative
for adults. N Engl J Med 2004; 351:2328–2330.
112. Ooi J, Iseki T, Takahashi S, et al. Unrelated cord blood transplantation for adult
patients with advanced myelodysplastic syndrome. Blood 2003; 101:4711–4713.
113. Sanz GF, Rocha V, Asano S, et al. Outcomes after umbilical-cord blood transplan-
tation from unrelated donors (UCBT) in patients with myelodysplastic syndromes
(MDS). An Eurocord study. Leuk Res 2005; 29(Suppl. 1):S20 (abstract).
Index
Page number with “t” and “f” indicate table and from an HLA-identical sibling, 510–514
figure. umbilical cord blood (UCB) transplantation,
517
ABCB7 gene, 160 from volunteer unrelated donors, 514–517
B cell defect in MDS, 443 (This should be under AMG 531, 407, 428
B cell) Aminoacyl-tRNA biosynthesis, 274
Abnormal localization of immature myeloid 3-(4-Amino-1-oxo-1,3-dihydro-2Hisoindol-2-
precursors (ALIP), 10, 199, 359–360 yl) piperidine-2,6-dione,
Abnormal skin pigmentation, 37 460
5-AC. See azacitidine ? Don’t know what this AML–MLL cell lines, 275
isAC133+ cells, 94 AML1 (RUNX1/CBFA2) gene, 179
ACPL, 96 AML-specific translocations in children, 321
Acquired idiopathic sideroblastic anemia AML-type chemotherapy, 352
(AISA), 138 Androgens, 407
Acquired marrow failure states, 440–441 Anemia, associated with MDS, 153
Acquired sideroblastic anemias, 159t of copper deficiency, 131
ACT2 gene, 96 therapeutic approaches
Activin A, 275 effect of the combination of G-CSF and
Acute leukemia, 55 EPO, 423–426
Acute lymphoblastic leukemia (ALL), 55, 256, erythropoiesis stimulating agents, 421-423,
268 426–427
Acute megakaryoblastic leukemia (AMKL), 36 iron chelation therapy, 418–421
Acute myelogenous leukemia. See Acute red cell transfusion therapy, 416–418
myeloid leukemia (AML) transfusion-related iron overload, 420–421
Acute myeloid leukemia (AML), 2, 33, 50–51, Anemia, WHO definition, 153
58, 63, 88, 110, 197, 312, 372–373 Angiogenesis inhibitors, 120
Acute panmyelosis with fibrosis (APMF), Annexin V, 112, 119
211–212 Anthracycline plus cytarabine, 185, 400
Adalimumab, 448 Anthracyclines, 40
Adenosine triphosphate (ATP), 111, 148 Antigen presenting cells (APC), 462
Adrenocorticotropin, 164 Antihemoglobin, 201
Adriamycin, 39, 182 Antithymocyte globulin (ATG), 228, 253, 406,
Affymetrix GeneChipTM technology, 87 439, 442, 444–447, 445t–446t
Affymetrix HG-U95Av2, 88, 91 Apaf-1, 111
Affymetrix HG U133 Plus 2.0 array platform, Aplastic anemia (AA), 37, 60, 164, 209, 240,
89, 93 312, 321, 439
Age-related damages, to mtDNA, 136 bone marrow histology of, 233
ALAS2 protein, 159 diagnose, 228
ALA synthase gene (ALAS2), 19, 159 Apoptosis in MDS, 140
Alkylator-induced t-AML, 180 cell proliferation versus, 117
Allogeneic stem cell transplantation (SCT), 242, and disease progression, 117–118
352, 397–400, 402 epigenetic modifications, 118
527
528 Index
Apoptosis in MDS, (Continued) B94 gene, 92

extrinsic pathway, 111 Basophilic stippling, 197
in hematopoietic system, 108–110 B-cell precursors, 252, 257
intrinsic pathway, 110–112 Bcl-2 family of proteins, 108
issues with BCL2 gene, 69t, 180
clinical implications, 119 BCNU, 39
in clonal cells, 119–120 BCR/ABL fusion gene, 213–214, 286, 292, 304
in relation to the stages of cell BCR/ABL-negative chronic myeloid disorders,
differentiation, 118–119 286
parameters BCR/ABL-positive CML, 305
caspase-3 activity, 112, 114t Benzene, 177
mitochondrial changes, 117 Beta fibroblast growth factor (bFGF), 463
NF-Kappa B activation, 112–115, 115t BFU-E progenitors, 240
p38MAPK (mitogen-activated protein BIRC5 gene, 93
kinase) activation, 115–116 Blast proportion, and MDS, 196–197
pro-apoptotic and anti-apoptotic proteins, Bleomycin, 39
116–117 B-lymphocyte development genes, 90
TNF-␣ levels, 112, 113t BM hypoplasia, 325
TRAIL, 112, 113t BMI1 gene, 92, 93–94, 274
upregulation of Fas or Fas-L, 112, 113t Bone marrow biopsy, 199–202, 228
therapies for Bone marrow cellularity, methods to evaluate
higher risk patients with decreased comparison of histology with MRI, 234–235
apoptosis,120–121 histological and cytological evaluation,
lower risk patients with increased 228–233
apoptosis, 120 magnetic resonance imaging, 233–234
ARG1 gene, 96 Bone marrow dysplasia, 358
ARPC2 gene, 75 Bone marrow hypocellularity, 444
ARPP21 gene, 90 Bone marrow morphology in MDS, 198f,
Arsenic trioxide, 402 358–361
ATRX gene, 96, 97f Bone marrow stroma, 185
Atypical chronic myeloid leukemia (aCML), Borderline dysplasia, 202
213–214, 285 BRCA2 gene, 92
clinical features, 305 Breast radiotherapy, 182
definition, 304 5-Bromodeoxyuridine, 249
genetics, 306
laboratory and pathologic findings, Caenorhabditis elegans, 108, 109f
305–306 CALCA gene, 319
prognosis, 306 CALGB 9221 trial, 488
Auer rods, 177 CAPZA2 gene, 75
Autoimmune manifestations in MDS, Caspase-9, 111, 161, 426
442–443 Caspase-3 activity in MDS, 112, 114t
Autoimmune mechanisms in MDS, 142 Caspases, 108 [Shouldn’s caspase-9 and
Autoimmune myelofibrosis, 208 caspase-3 be subsets of this]
Autophagy, 128–129 CBFB gene, 179
Autoreactive T cells, 439 CCAAT/enhancer-binding protein-alpha
Azacytidine or azacitidine, 325, 400, 402, 498 transcription factor gene (CEPBA), 69t,
Azacytosine nucleoside analogues, 487–490 73
effect on epigenetic abnormalities, 490–491 CCDC6 protein, 62
for the treatment of MDS, 492–493 CCR2 gene, 92
Azurophilic granules, 196 CD4, 202
Index 529
CD7, 253 Ced-4, 108

CD8+cytotoxic T cells, 439, 440f, 441462 Ced-9, 108
CD10, 259 Cell proliferation in MDS, 117
CD11b, 250, 257, 259 CFU-GEMM, 468
CD11c, 202, 376 CGH microarrays
CD13, 202, 376 contemporary studies, 100–101
CD14, 92 early studies, 98–100
CD15, 376 studies with del(5q), 101–102
CD16, 250, 376 Chemotherapy, 242, 400–401
CD18, 96 Childhood MDS, 35–37, 36t
CD19, 257 bone marrow features, 320
CD24, 90 clinical and laboratory features, 319–320
CD31, 201 cytogenetics, 320–321
CD33, 259 diagnostic categories of, 313–314, 313t
CD36, 257 differences with adult, 312t
CD34+ cells, 74, 96, 118–120, 132–133, 197, differential diagnosis, 321–324
209, 238, 249, 255, 262, 373–375, 426, epidemiology
441, 443, 490 acquired aplastic anemia, 317–318
deficiency of RPS19 in, 276 age difference, 315
measured by flow cytometry, 202 familial, 318
microarray studies, 89–92 incidence, 315
as predictors of acute leukemic inherited bone marrow failure, 317
transformation, 200 male/female distribution, 316
in the 5q− syndrome, 271 subtype distribution, 315–316
CD34+ basophils, 257 therapy-related, 318
CD34+ myeloblasts, 257, 262 flow cytometry immunophenotyping, 321
CD34+ plasmacytoid dendritic cell, 443 historical background of the childhood MDS
CD34+CD38− HSC compartment, 269 classification, 312–313
CD34+CD19+ pro-B cells, 269 pathophysiology, 319
CD34+CD38−Thy1+ cells, 93, 269 primary and secondary, 314
CD41b, 92 prognosis and natural course, 324–325
CD45, 202, 249–250, 254–257 treatment
CD55, 439 AML-type chemotherapy, 326
CD56, 202, 259 hematopoietic allogeneic stem cell
CD59, 439 transplantation, 326–327
CD68 (macrosialin), 201 relapse following HSCT, 327
CD61 (platelet glycoprotein IIIa), 75, 201 secondary MDS, 327–328
CD64, 259 transplant-related mortality (TRM), 326
CD71, 257 versus childhood AML, 323t
CD105, 201 Chlorambucil, 39, 179
CD117 (c-Kit), 201, 256–257, 262, 376 Chloramphenicol toxicity, 131, 159
CD123, 287 Chromatin clumping, 197
CD163 (macrophage-associated antigen), 201 Chromosomal abnormalities in MDS, 2, 58-61
CDC25C gene, 477 Chromosomal analysis of MDS, 366–372
CDH-1 methylation, 490–491 Chromosomal instability and mitochondrial
CDKN1A/p21 gene, 65 dysfunction, 140–141
CDKN2B (p15INK4B )gene, 71t, 73, 319 Chromosome 7 or del(7q), loss of, 60
CDR of the 5q− syndrome, 270–271 Chromosome 8, gain of, 59
CD45/SSC plots, 256 Chromosome 17 (17p−) syndrome,
Ced-3, 108 60–61
530 Index
Chromosome 20, del(20q), deletion of, 58 Clonal cytogenetic abnormalities, 202

Chronic idiopathic myelofibrosis, 212 Clonality in MDS, 51
Chronic lymphocytic leukemia (CLL), 33, 136, CMML. See chronic myelomonocytic leukemia.
179 Commonly deleted region (CDR), 270
Chronic myelogenous leukemia (CML), 55, 256 Commonly deleted segments (CDS), 64
Chronic myelomonocytic leukemia (CMML), Congenital thrombocytopenia, 317
27, 56, 196, 212–213, 285–286, 312, Continuous erythropoietin receptor activator
363, 372, 377–378, 464 (CERA), 427
absolute lymphocyte count in, 290 Coombs-positive warm autoantibody hemolytic
blood and marrow blasts numbers in, 287 anemia, 447
CD34 expression, 287 Copper deficiency, 207
clinical features, 286–287 Core-binding factor (CBF) complex, 72
definition, 286 C3ORF27 gene, 63
differential diagnosis, 292–293 CORO1C gene, 75
dysmegakaryopoietic and dyserythropoietic Corticosteroids, 447
features, 287 COX2 gene, 96
genetics, 287–289 CpG dinucleotides, 486
immature granulocytic precursors, 290 CRBP1, 490
karyotypes, 290 CREBBP gene, 62, 62f
laboratory and pathologic findings, 287 CSF1R/FMS gene, 69t
prognosis CSNK1A1 gene, 75, 93, 271, 274
CMML-1 versus CMML-2, 290 CTNNA1 gene, 65, 278–279
myeloproliferative versus myelodysplastic CUTL1 gene, 66
type, 289–290 Cyclophosphamide, 39, 176, 179, 182
prognostic scoring systems for, 290–292, Cyclosporine A (CSA), 228, 439, 442,
291t 445t–446t, 447
serum lactate dehydrogenase (LDH) levels, CYP3A4 gene, 174
290 CYP3A5 gene, 39
Chronic myeloproliferative disorders (CMPD), Cystathionine-␤-synthetase, 331
233, 285 Cytarabine, 185, 242, 331
Cisplatin, 39 Cytochrome c, 111, 129, 139, 160–161, 426
Class I and Class II mutations, 74 Cytochrome P450 system, 39, 472
Classical anthracycline–Ara C chemotherapy, Cytogenetics of MDS
400, 402 abnormalities, 10
Classification of MDS cases of trilineage dyplasia, 240
associated with an isolated del(5q) in predicting patient outcomes, 187–189
chromosomal abnormality, 206 studies of the interstitial 5q deletion, 270
FAB, 5–8, 27, 177, 312, 349–350, 361–366, analysis of rare recurring translocations
376 abnormalities of 3q, 63
refractory anemia with excess of blasts beta chain of the platelet-derived growth
(RAEB), 206 factor receptor, 62–63
refractory anemia with ring sideroblasts 11q23 abnormalities, 61
(RARS), 205 t(11;16), 61–62
refractory cytopenia with multilineage chromosomal abnormalities, 54f
dysplasia (RCMD), 205–206 chromosome 17 (17p−) syndrome, 60–61
refractory cytopenia with unilineage complex karyotypes, 61
dysplasia, 203 deletion of chromosome 20, del(20q), 58
unclassifiable, 206–207 gain of chromosome 8, 59
WHO, 196, 203, 204t isolated del(5q) (5q− syndrome), 59
Clonal chromosome abnormalities, 177 loss of chromosome 5 or del(5q), 58–59
Index 531
loss of chromosome 7 or del(7q), 60 flow cytometry, 202

loss of the Y chromosome (−Y), 57–58 morphology, 196
clonality analysis, 51 Diamond–Blackfan anemia, 66, 276, 317
identification of abnormalities in the Dietary iron, 163
karyotype, 54f, 56–57 DiGuglielmo syndrome, 2
prognosis and clinical correlations, 55–56 Dihydroorotate dehydrogenase (DHODH), 141
recurring cytogenetic abnormalities, 51–55, DKC1 gene, 37–38
52t–53t DLK1 gene, 88, 89
significance, 50 D-loop of mtDNA, 136
versus fluorescence in situ hybridization DNA methyltransferase inhibitors, 325
(FISH), 55 DNMT inhibitors, 487–489. See also
Cytomegalovirus infection, 208 azaciditine, decitabine.
Cytoplasm, hypogranularity of, 197 efficacy improvement strategies, 492
Cytosine methylation, 118 Down syndrome (DS), 36, 314, 328–331
myeloblasts, 331
DAPK gene, 118 Down syndrome megakaryoblastic leukemia, 36
DAP kinase 1, 490 DPH5 gene, 93, 274
Darbepoetin alfa, 405, 426. See also DRAQ5, 249
Recombinant Erythropoietin. Drug-induced sideroblastic anemia, 159t
Death-inducing signal complex (DISC), 111 Dyserythropoiesis (Dys-E), 28, 197, 228
Decitabine, 325, 498. See also DNMT inhibitors. Dysgranulopoiesis (Dys-G), 34, 60, 177, 197,
DEFA4 gene, 92 209, 228, 287, 305, 358
Deferasirox, 421 Dyshematopoietic changes MDS-like
Deferoxamine, 420 (non-clonal), 207t
Del(5q), 59, 64–66, 206, 364, 472-3. See also Dyskeratosis congenita, 37, 317
5q-. Dysmegakaryocytopoieis (Dys-Meg), 177, 199,
Del(11q), 367 209, 228, 358
Del(12p), 367 Dysmorphic syndrome, 37
Del(13q), 367 Dysplastic-type CMML, 488
Del(20q), 367 Dysplastic megakaryocytes, 199
Del(20q), molecular analysis of, 67 Dysplastic neutrophils, 178f, 214, 441
Denaturing HPLC (dHPLC) for mutation
detection, 135 Early growth response 1 gene (EGR1), 65, 477
Deoxythymidine kinase (sTK), 356 EBF gene, 90
Deoxyuridine monophosphate (dUMP) EGR1 gene, 65, 278
synthesis, 39 EKLF expression, 180
Dexamethasone, 179 ELA2 gene, 38, 90
DHRS8 gene, 90 Endonuclease G, 139
Diagnostic evaluation of MDS Environmental factors predisposing to MDS,
blasts, 196–197 40–41
bone marrowbiopsy, 199–202 Eosinophilia, 288
cytogenetics, 202 Epidemiology of MDS
difficulties in Bournemouth group study results, 32–33
differentiation from non-neoplastic childhood, 35–37
disorders, 207–209 clinical features of Japanese patients, 34
with fibrosis (MDS-F), 210–212 differences between European and Asian
hypocellular MDS (h-MDS), 209–210 populations, 34
dyserythopoiesis, 197 Düsseldorf group study results, 30–32, 31t
dysgranulopoiesis, 197 estimation problems, 27–28, 32–33
dysmegakaryopoiesis, 199 European Population Studies, 29–31
532 Index
Eosinophilia, (Continued) aspects, 279–280

French study results, 30–31, 31t bone marrow morphology study, 269
German patients, 34 candidate genes for, 274–278
incidence estimated by individual studies, 31t cell of origin, 269
incidence in relation to other hematological clinical, laboratory, and prognostic features,
malignancies, 33 268–269
increasing rate of incidence, 31–32 cytogenetic studies, 270
Japanese study results, 31, 31t genes mapping to the commonly deleted
LRF study results, 29–31, 31t region of the, 272t–273t
male to female ratio, 30 genetic stability, 270
at older ages, 30 hematological indicators for, 267–268
SEER database results, 28–29 molecular mapping of the CDR, 270–271
Swedish study results, 30–31, 31t relation with myeloid malignancies with the
true incidence of, 27–28 del(5q), 278–279
various countries, 33–34 survival rate, 268
younger adults, 34–35 treatment, 279
Epipodophyllotoxins, 40 FAB classification, see French-American-British
Epigenetics. See DNMT inhibitors, azacitidine, classification.
decitabine. Familial MDS, 38, 318
EPO, see Erythropoietin and Recombinant Fanconi anemia, 37, 60, 312, 314, 317
Erythropoietin. Farnesyl-transferase inhibitors, 120, 402
EPO gene, 67 Fas, see Fas-L/Fas pathway
Erythema nodusum leprosum (ENL), 459 Fas-associated death domain protein (FADD),
Erythroblast cultures, analysis, 94 111
Erythroblast mitochondria, impaired, 138–139 Fas-L/Fas pathway, 112, 113t, 140, 238, 442
Erythroblasts, 257 FBXL13 gene, 67
Erythrocyte macrocytosis, 208 Fc-gamma receptor IIIb, 250
Erythrocytic stimulating agents (ESAs), 405 Ferritin level in MDS, 55
Erythroid burst–forming units (BFU-E), 275, Ferrochelatase, 138
468 Fetal hemoglobin (HbF), 320
Erythroid hyperplasia, 197, 240 Field cancerization theory, 441
Erythroid-specific ALA synthase gene, 157 FKBP-12, 448
Erythron iron turnover, 154 FLAG-Ida regimen, 509
Erythropoiesis Flavocytochrome b558, 139
disturbed, 1 FLICE-like inhibitory protein (FLIP), 115
ineffective, 154–156 Flow cytometric scoring systems (FCSS), 202,
Erythropoietin, 253, 405 251
Etanercept, 120, 442, 448 analysis of dysplasia compliments
Ethanol-induced sideroblastic anemia, 159t morphologic assessment, 252
Etoposide, 39 approaches
European Working Group on MDS in childhood, analysis of maturing myeloid cells,
322, 324–327 257–259
EVI1 (Ecotropic Virus Insertion site −1) gene, analysis of monocytes, 257
63, 182. See also MDS-EVI1. analysis of multiple phenotypic
Extramedullary myeloid tumor, 320 abnormalities, 259–262
Extrinsic pathway of apoptosis in MDS, 111 analysis of myeloblasts, 254–257
analysis of the erythroid compartment, 259
5q− syndrome, 32, 93–94, 101–102, 206, 216, gating strategies, 254
313, 364, 367, 369, 414, 469, 472 reagent combinations, 254
analysis in hematopoietic stem cells, 271–274 diagnosis, 251–252
Index 533
historical perspectives CGH and SNP microarrays

antigenic combinations, 248 contemporary studies, 100–101
DNA content measurement, 248–249 early studies, 98–100
multicolor flow cytometric analysis, studies with Del(5q), 101–102
250–251 microarray application, 87–89
multiple antigen studies, 249–250 AC133+ hematopoietic stem cells (HSC),
single antigen studies, 249 transcriptome of, 88
utility of CD45 and right-angle light scatter, ATMDS microarray results, 96–98, 97f
250 in CD34+ cells, 89–90
prognosis, 253 effect of chromosomal abnormalities,
Flow cytometry immunophenotyping, 321 90–93
FLT3 gene, 67, 69t, 73, 74, 90, 180 karyotype-related, 90–91
FLT3-ITD mutations, 67, 72 lenalidomide in, 94
FLT3-length mutations (FLT3-LM), 118 studies in mature cells, 94–98
Fludarabine, 179 studies in 5q− syndrome, 93–94
Fludarabine–Ara C combinations, 400 Genomic instability and mitochondrial
Fluorescent in situ hybridization (FISH), 55, 239 dysfunction, 140–141
FMS gene, 319 Globin genes, 96
FMS-like tyrosine kinase 3 (FLT3) gene, 67 GLRA1 gene, 65, 270
Folic acid deficiency, 207 Glutathione S-transferase (GST), 39
FOS gene, 96 Glycophorin A or C, 201
Four-color flow cytometry, 202 Glycosylphosphatidylinositol (GPI), 439
FPC (Fetal hemoglobin, Platelets, Cytogenetics) GP1BA gene, 75
scoring system for childhood MDS, 36 GPI-linked proteins, 439
Free-radical theory of ageing, 131–133 Granulocyte-colony stimulating factor (G-CSF),
French-American-British (FAB) Cooperative 38, 161–162, 182, 253, 317–318
Group classification system, 5–8, 27, Granulocytes, 250
177, 312, 349–350, 361–366, 376 Granulopoiesis, 213
Granulopoietic precursors, 199
GATA1 gene, 180, 329 Gravin/AKAP12 gene, 90
Gating strategies, 254 GSTT1/GSTM1 genotypes, 40
G-banded metaphase cytogenetics, 2, 240 GSTT1 null genotype, 40
GCSFRG gene, 69t GTPase RABEP1 (Rabaptin 1), 62
GDF15, 155–156
Genetic factors of MDS ␥ -H2AX, 491
alterations in gene function, 67–73 HbF-containing erythroblasts, 441
congenital, 37–38 H4/D10S170 gene, 289
emerging technologies, 76–77 Heavy-metal intoxication, 207
familial clustering, 38 Hematologic malignancies, 201
genes altered, 69t–71t in children, 315t
genetic pathways, 74–76 in older persons, 158
molecular analysis of the −5/del(5q), 64–66 Hematologic neoplasms, 38
molecular analysis of the −7/del(7q), 66–67 Hematologic response of patients with MDS,
molecular analysis of the del(20q), 67 236f
molecular models for chromosome Hematopoiesis
abnormalities, 63–64 and hematopoietic failure in MDS, 441–442
polymorphisms, 38–40 steady state, 109
Genetic pathways, leading to MDS, 76f Hematopoietic allogeneic stem cell
Genetic stability in 5q− syndrome, 270 transplantation, 326–327
Genomic approaches in MDS Hematopoietic neoplasms, 208
534 Index
Hematopoietic progenitor cells, 51, 162, 442 IFI56 gene, 96

Hematopoietic stem cells (HSC), 129, 133 IFIT1 gene, 74
age-related deterioration of, 134 IFITM1 gene, 74
effects of mtDNA mutations, 134 Ifosfamide, 39
for t-AML, 186–187 IL-7 receptor, 96
and therapy-related leukemia, 183 Imatinib mesylate, 56, 274
Heme synthesis, 137–139 Immunohistochemistry, 199
Hepatic fibrosis, 164 Immunomodulatory drugs (IMiDs)
Hepatosplenomegaly, 60 anti-angiogeneic activity of, 463
Hepcidin, 154–155 antitumor activity, 463
Hereditary sideroblastic anemias, 159t biological effects of, 461f
Herpesvirus 6 (HHV6), 321 concept of, 458–459
Heterogenetic gene expression patterns in MDS, immune effects of, 462
96 pharmacodynamics of, 461–463
Heteroplasmic mtDNA mutations, 135 pharmacokinetics of, 459–461
H-ferritin, 156 for renal impairment, 461t
Hierarchical clustering, 88, 91f–92f Immunomodulatory agents, 448
HIST1 gene, 90 Immunophenotyping
Histone acetyltransferases, 485 protocols, 51
Histone deacetylases (HDACs) inhibitors, 120, Immunophenotyping, 374–376
402, 485 Immunosuppressive therapy (IST), 325, 439
HIV infection, 208, 321 clinical trials
HLA-DR15 (DR2), 228, 325, 444 antithymocyte globulin (ATG), 444–447
HMSH2 and hMLH1 genes, 176 cyclosporine (CSA), 447
Hodgkin disease, 357 immunomodulatory agents, 448
HOX genes, 62, 92 p38 MAPK inhibitors, 449
HSPA9B, 130 prednisone and other corticosteroids, 447
Hydroquinone, detoxification of, 39 sirolimus, 448
Hyperfibrotic MDS, 360 TNF-␣ inhibitors, 448
Hyperplastic panmyelosis, 2 vaccines, 449
Hypocellular acute myeloid leukemia for MDS, 236–238
(Hypo-AML) mechanism of hematologic remission by,
clinical and hematologic features of, 241–242 238–239
response to chemotherapy, 242 patient profile, 228, 229t–232t
Hypocellular MDS (h-MDS), 209–210, 209f Impaired GTPase activities, and MDS, 372
clinical and hematological features of, 235 Incidence of MDS. See Epidemiology of MDS
immunosuppressive therapy in, 236-238, 240 Ineffective hematopoiesis, 1
severe AA as an extreme case of, 239–240 Infliximab, 120, 442, 448
Hypolobular megakaryocytes, 269 Inhibitor of apoptosis (IAP) family of proteins,
Hypomethylating agents, 401–402, 406 110f
Hypoproliferative erythropoiesis, 415 In situ end labeling (ISEL) as technique for
Hypoxia, 155 detection of apoptosis, 119
Intensive AML-type chemotherapy
ICSBP/IRF8 gene, 74 autologous SCT as, 506–508
Idiopathic cytopenia of undetermined background, 499
significance (ICUS), 217–218, 252 bases for the use of, 499
Idiopathic refractory sideroblastic anemia, 154, prognostic factors, 508–509
see also refractory anemia with ring quality of life post, 506
sideroblasts results, 500–501
IEX1 gene, 88 reverting of MDR phenotypes, 505
Index 535
use of hematopoietic growth factors, KIAA0001 gene, 96

501–505 Ki-67+ cells, 441
Interferon-␥ , 442 KIT gene, 70t
Interleukin 11 (IL-11), 407 mutations, 118, 186
Internal tandem duplications (ITDs), 67 Klinefelter syndrome, 316
International Classification of Disease for Knudson two-hit model, 63, 271
Oncology 27, 28 Kostmann syndrome, 38
International MDS Risk Assessment Workshop KRAS gene, 60
(IMRAW), 10–13
International Prognostic Scoring System (IPSS), Lactate dehydrogenase (LDH), 377
8–13, 9t, 11t–12t, 16–19, 51, 59, 196, Large granular lymphocyte disorders, 439–440
235–236, 253, 324, 352, 361, 379–380, Lenalidomide (CC-5013), 56, 94, 95f, 253, 275,
379t–380t, 394, 464 279, 405, 407, 448, 457, 459
Intrinsic pathway of apoptosis in MDS, 110–112 absorption of, 460
Investigational agents, 402 clinical safety of, 471–475
Inv(3)/t(3;3) abnormalities, 63 clinical studies, 466t–467t
Iodine-131 tositumomab, 179 cytokine modification, 462
IPSS, see International Prognostic Scoring insights to the mechanism of, 475–477
System. non-hematological adverse events, 474–475
Iron chelation therapy, 165, 406 patients with del(5q) karyotypes, 472–473
Iron-dependent oxidative damages, 138 patients with non-del(5q) karyotypes,
Iron-laden mitochondria, 128 473–474
due to ineffective erythropoiesis, 154–156 pharmacokinetics, 460–461
ferritin sideroblasts, 156–157 treatment of MDS, 465–471
mitochondrial iron ring sideroblasts, 157 clinical use, 477–479
pathogenesis of, 137f Leukocyte common antigen, 249
secondary iron overload, 163–165 Leukocytosis, 60
sideroblastic anemia, 157–162 Leukopenia, 177
thrombocytosis, 162–163 L-ferritin, 156
Iron-loading anemias, 154–155 LH2 gene, 89
Iron overload Li-Fraumeni syndrome, 176
negative effects, 164 LINE methylation, 490–491
progressive, 154 Liver cirrhosis, 406
Isochromosome 17q MDS/MPN, 216, 367 L3MBTL expression, 67
Isoniazid, 159 Loss of chromosome 5 or del(5q), 58–59, see
also 5q-.
JAK2 gene, 69t, 73,162, 215, 288 Loss of heterozygosity (LOH), 98
JAK2 V617F mutation, 15, 162–163, 180, 182, Loss of the Y chromosome (−Y), 57–58
205, 286, 292, 300–301, 303, 365 Low-dose Ara C, 400
JAK2 V617F in MPD, 301 LRCC17 gene, 66
JAK2 V617F in RARS-T, 304 Lymphocytosis, 199
Juvenile myelomonocytic leukemia (JMML), Lymphoid follicles, 199
36, 60, 286, 314–315. See also childhood Lysozyme, 201
MDS.
Macrocytic anemia, 134, 205, 320
Kaplan–Meier survival curves, 362–364, 363f, Magnetic resonance imaging (MRI), 165
488 Marrow failure, 109
Karyotypes, complex, 61. Cf. individual Marrow fibrosis, 210–212, 361
karyotypes. MDM2 gene, 70t, 176
Karyotypic abnormalities in RARS-T, 300 MDR1 gene, 70t
536 Index
MDS1/EVI1 genes, 63 Mitochondrial ferritin (mitoferrin, MtF), 138,

MDS with fibrosis (MDS-F), 210–212 139, 161
Mean corpuscular volume (MCV), 177 Mitochondrial respiratory chain (RS), 129
Megakaryoblasts, 196 dysfunction, 137f
Megakaryocyte growth and differentiation factor Mitogen-activated protein kinase (MAPK)
(MGDF), 428 family, 442
Megakaryocyte hypoplasia, 199 Mitomycin C, 39
Megakaryocytes, 205, 214, 269, 305 Mitosis, 133, 141
premature death of, 139–140 Mitoxantrone, 39–40, 179
Megakaryocytopoiesis, 211, 428 MLL (Mixed Lineage Leukemia) gene or
Megaloblastic erythropoiesis, 320 protein, 61, 62f, 70t, 72, 118, 179, 180
MEGF1 gene, 271 MME gene, 90
Melphalan, 39, 400 Moderate-to-marked poikilocytosis, 298
Metabolic disorders, 321 Molecular mapping in 5q− syndrome,
Methionine, 39 270–271
Methylene tetrahydrofolate reductase (MTHFR) Molecular mutations in MDS, 68t
enzyme, 39 Monoblasts, 196
Methylome of MDS, 490 Monocytes, 196
Methyltransferease inhibitor therapy, 325 Monocytosis, 287
Mfrn−/− hematopoietic cells, 139 Monosomy 5, 177
␤2-Microglobulin levels in MDS, 55 Monosomy 21, 367
Micromegakaryocytes, 178f, 199 Monosomy 7 syndrome, 36, 60, 92–93, 177,
Mild-to-marked reticulin fibrosis, 177 239, 312, 320, 400
Miller–Dieker syndrome as another “MDS”, 37 Morphologic dysplasia, 196
Mitochondrial biogenesis, 131 MPD/MDS overlap syndrome, 13
Mitochondrial changes in MDS, 117, 143f MPD-type megakaryocyte morphology, 300
apoptosis related, 140 MPL gene, 70t. See also thrombopoietin
autoimmune mechanisms, 142 receptor gene.
and copper deficiency, 131 MtDNA mutations, 132–133
evidence of mitochondrial dysfunction, 129 MTHFR polymorphisms, 39
in the elderly, 131 Mucosal leucoplakia, 37
in hematopoietic cells, 144f Multicolor flow cytometric analysis,
possible causes, 129–131 250–251
and free-radical theory of ageing, 131–133 Multilineage dysplasia, 158, 164, 252
genomic instability, 140–141 Multilobular megakaryocytes, 269
granulocyte dysfunction, 139 Multiparameter flow cytometry, 250
impact on bone marrow functions, 133–136 Myeloablative therapy, 325
impaired ATP synthesis, 141 Myeloblasts, 196, 199, 254–257
impaired erythroid maturation, 138–139 Myelodysplastic/myeloproliferative neoplasms
light and electron microscopic findings, (MDS/MPN), 60
127–129 associated with isochromosome 17q, 216
megaloblastic changes, 141–142 atypical chronic myeloid leukemia (aCML),
mtDNA mutations, 133–136 213–214
pathogenesis of the sideroblastic phenotype, chronic myelomonocytic leukemia (CMML),
137–138 212–213
premature death of megakaryocytes (MK), “early” MDS and ICUS, 217–218
139–140 5q−/JAK2 positive atypical
RARS-T, 143 myeloproliferative disease, 215–216
Mitochondrial DNA (mtDNA), mutations of, refractory anemia with ringed sideroblasts and
129 thrombocytosis (RARS-T), 215
Index 537
therapy-related MDS (t-MDS) and t-AML, pathobiology, 329–330

216–217 treatment, 330–331
unclassifiable, 214 Myeloid maturation, 257–259
Myelodysplastic/myeloproliferative syndrome, Myeloid neoplasia, 314
73 Myeloid neoplasms, deletions of 5q and 7q in,
Myelodysplastic syndromes (MDS), 129. See 58–59, 58f
also Diagnostic evaluation of MDS; Myeloma, 33
Epidemiology of MDS; Prognostic Myeloperoxidase (MPO), 139, 201
parameters in MDS; Therapeutic strategy Myeloproliferative disease, 135
in MDS MYL6 gene., 75
abnormalities associated with, 2
autoimmune manifestations in, 442–443 NADH dehydrogenase, 142
classification NAD(P)H:quinone oxidoreductase (NQO1), 39,
as cancer, 19 174
French-American-British (FAB) Nail dystrophy, 37
Cooperative Group, 5–8, 27 NAIP gene, 93
International Classification of Disease for ␣-naphthyl butyrate esterase, 196, 287
Oncology (ICD-O-3), 28 National Surgical Adjuvant Breast and Bowel
International Classification of Diseases Project, 182
(ICD-9), 27 NDUFV1 gene, 89
MD Anderson CMML classification Neoplastic myeloblasts, 248
system, 13 Neurofibromatosis type 1, 37, 60, 177
2008 WHO Revision, 15–16 Neutropenia, 1, 406, 440, 472–474
World Health Organization (WHO) management of, 427
Classification, 13–15, 14t, 28, 50t Neutrophil nuclei, hypersegmentation of, 197
WPSS, 18t, 19 Neutrophils, 94
clinical overlap with AA, 239–240 morphologic abnormalities in, 197
connotation, 1 NF1 gene, 37, 70t, 177
cytogenetic risk groups defined by the NF-Kappa B activation in MDS, 112–115, 115t
German-Austrian Consortium, 17t Nicotinamide adenine dinucleotide, see NADH
diagnostic criteria for, 3t–4t or NAD(P)H.
early developments, 2–5 Non-del(5q) karyotypes, 473–474
history of classification and treatment of, 3t–4t Non-Hodgkin lymphomas (NHL), 33, 94, 357
International Prognostic Scoring System Non-Hodgkin lymphomas (NHL) cell lines,
(IPSS), 8–13, 9t, 11t–12t, 16–19, 51 275
as a marrow failure syndrome, 438–441 Noonan syndrome, 37
mechanisms of ineffective hematopoiesis in, NRAS, 16, 60, 71t, 118, 287, 319, 372. See also
414–415 RAS.
minimal diagnostic criteria for, 322t ␣-(N-phthalimido)glutarimide/C13H10N2O4,
peripheral blood cytopenias in, 2 459
risk stratification, 12t NPM1 gene, 70t, 72–73
Myelofibrosis, 360 NQO1, 174
Myelofibrotic myeloid neoplasms, 211 Nuclear hypolobation, 197
Myeloid diseases, associated with abnormalities NUP98 gene, 179
of 3q, 63
Myeloid dysplasia, 134 O(6)-guanine alkylating agents, 176
Myeloid leukemia in Down syndrome, 316 Oligomycin-sensitive ATPase, 129
Clinical and laboratory features, 330 Oligonucleotides, 87
cytogenetic abnormalities, 330 8-oxoG DNA glycosylase (hOGG1), DNA
epidemiology, 329 repair gene, 38
538 Index
P15 gene, 118 Prognostic parameters in MDS, 353t. See also

P53, 201, 319; see also TP53 IPSS and WPSS.
P73, 490 bone marrow morphology, 358–361
Pancytopenia, 228 chromosomal analysis, 366–372
Paraneoplastic myelodysplasia, 208 clinical parameters, 353–355
Parenchymal iron loading, 154 FAB and WHO classifications, 361–366
Paroxysmal nocturnal hemoglobinuria, 37, 228, immunophenotyping, 374–376
439 laboratory features, 355–358
PCNA+ cells, 209 molecular genetic findings, 372–374
PDGFRA, 214 prerequisites for using clinical and biological
PDGFRB, 56, 62, 213 variables as risk factors, 349–353
Pearson syndrome, 128, 137, 160, 205, 322 scoring systems, 376–383
Pediatric sideroblastic anemia, 35. See also Programmed cell death, 108
Childhood MDS. Promonocytes, 196
PF4 gene, 92 Promyelocytes, 199
P190 fusion protein, 213 Protein ubiquitination, 274
PF4V1 gene, 75 Protoporphyrin IX, 138
Phosphatidylinositol 3-kinase (PI3 Prussian blue staining, 156
K)/AKT/mammalian target of rapamycin Pseudoaplastic anemia, 2
(mTOR), 117 Pseudo-Pelger-Huët anomalies, 214, 305
Photolithography, 87 Pseudo-Pelger-Hüet hypolobulation, 60
PIASx-␤ gene, 89 PTPN11 gene, 60, 71t, 182
PIASy gene, 89 Pure sideroblastic anemia (PSA), 362. See also
PIGA gene, 439. See also GPI. refractory anemia with ring sideroblasts
PIK3C2B gene, 180 (RARS).
p53-induced gene, 274 PUS1 gene, 160
PLAB gene, 92 P21WAF1 /CIP1 gene, 491
Plasma cell proliferative disorder (PCPD), Pyrimidines, 141–142
268
Plasmacytosis, 199 RA. See Refractory anemia (RA)
Platelet-derived growth factor receptor beta. See RAB20 gene, 96
PDGFRB. RAD51 gene, 39
P38 MAPK inhibitors, 115–116, 442, 449 RAEB. See Refractory anemia with excess
PML/RARA gene, 179 blasts (RAEB)
PNMA2 gene, 89 RAEB in transformation (RAEB-t), 8, 13, 35,
Poikilocytosis, 197 117, 127–128, 312–313, 322, 324, 326,
Polg mouse models, 134 363, 372–373, 465, 492
Polymorphisms in MDS, 38–40 RAI3 gene, 88
Pomalidomide (CC-4047), 459, 462 RAR␤2, 490
Posttransfusion purpura, 418 RARS. See Refractory anemia with ringed
PP2 A-C␣, 477 sideroblasts (RARS)
PPBP gene, 75 RARS-T. See Refractory anemia with ringed
PPIC gene, 275 sideroblasts associated with marked
PRAME gene, 92 thrombocytosis (RARS-T)
Prednisone, 444 RASA4 gene, 67
Preleukemia, as older term for MDS, 5, 312 RAS pathway and mutations, 60, 289
Preleukemic anemia, as older term for MDS, 2 RASSF1A gene, 118, 490
Pro-apoptotic versus anti-apoptotic molecules in RA (WHO) morphological subtype, see
MDS, 116–117 Refractory anemia.
Procarbazine, 176 RBM22 gene, 75, 93, 271, 274
Index 539
RCMD. See Refractory cytopenia with Refractory normoblastic anemia, as older term
multilineage dysplasia (RCMD) for MDS, 2
RCMD with ring sideroblasts (RCMD-RS), 205 Refractory thrombocytopenia (RT), 203
RCUD. See Refractory cytopenia with R(+) enantiomer, 459
unilineage dysplasia (RCUD) Residual megakaryopoiesis, 209
Reactive oxygen species (ROS) production in Reticulin fibrosis, 201, 209, 211, 305–306
MDS, 132, 355 Retinoblastoma gene product (Rb), 130
Recombinant erythropoietin (EPO), 416, Revlimid, see Lenalidomide
421–423 Rheumatoid arthritis, 321
in combination with G-CSF, 423–426 Ribosomal RNA (rRNA), 276
predictive algorithm for, 425t RIL gene, 490
Red blood cell (RBC) anisocytosis, 197 Ringed sideroblasts with thrombocytosis, see
Red blood cell transfusions and hepcidin, 155 RARS-T
Red cell alloimmunization, 418 Rituximab, 179
Red cell transfusion therapy, 416–418 RNA interference (RNAi) technology, 277
Red–fluorescing DNA probes, 249 RPL22L1 (EAP) gene, 63
Refractory anemia (RA), 28, 35, 89, 127, RPN1 (ribophorin 1), 63
162–164, 199, 203, 414, 464 RPS14 gene, 59, 66, 75, 93, 180, 206, 271, 274,
Refractory anemia with excess blasts in 276–278, 477
transformation (RAEB-t), 464 RPS19 gene or protein, 66, 276, 277
Refractory anemia with excess blasts (RAEB), RPS25 gene, 66
8, 29, 90, 127, 203, 206, 312, 372–373, RUNX1/AML1, 63–64,71t, 72
414, 464
with a hypocellular marrow, 227 S-adenosyl-methionine, 331
Refractory anemia with ringed sideroblasts SCIO-469, 449
associated with marked thrombocytosis SCT-specific comorbidity index, 509
(RARS-T), 13, 162–163, 215, 286, 365 SD-282, 449
characteristics of patients, 294t–297t, 298 Serum ferritin, 165
clinical features, 293–298 Severe congenital neutropenia (SCN), 317
definition, 293 S-G2 M-phases of the cell cycle, 248
differential diagnosis, 302–304 Short TI inversion recovery (STIR) imaging, 233
genetics Shwachman–Diamond syndrome, 37, 317
JAK2 V617F mutation in, 300–301 Sideroblastic anemias, 32, 157–162
karyotypic abnormalities, 300 Single nucleotide polymorphism (SNP) arrays,
other aberrations, 301 51, 173
laboratory and pathologic findings, 298–300 detection of chromosomal microlesions and
megakaryocyte morphology in, 298 uniparental disomy (UPD), 240–241
prognosis, 301–302 detection of clonal hematopoiesis, 241
Refractory anemia with ringed sideroblasts Sirolimus, 448
(RARS), 29, 90, 112, 128, 157–158, 199, Smac/DIABLO proteins, 112, 139
203, 205, 312, 362, 414, 464 Small-cell lung carcinoma, 256
Refractory cytopenias, 36, 164, 321 Smoldering acute leukemia, 2
Refractory cytopenia with multilineage SNP array analysis, 76–77
dysplasia (RCMD), 15, 29, 199, 203, contemporary studies, 100–101
205–206, 228 examples of graphs, 99f
Refractory cytopenia with unilineage dysplasia studies with del(5q), 101–102
(RCUD), 203 SOCS1 gene, 118, 490
Refractory megaloblastic anemia, as older term SPAG6 gene, 93, 274
for MDS, 2 SPARC gene, 75, 93, 271, 274–275, 477
Refractory neutropenia (RN), 203 S-phase cells, 117, 130
540 Index
S-phase of cell cycle, 248 recommended strategies, 402–404

SPI1/PU.1 gene, 64 intensive AML-type chemotherapy
Splenomegaly, 212, 287 autologous SCT as, 506–508
SRPK2 gene, 66 background, 499
STAT5 transcription factor, 426 bases for the use of, 499
Stem cell transplant (SCT), 251 prognostic factors, 508–509
STIP1, 88 quality of life post, 506
Superconducting quantum interference device results, 500–501
(SQUID) magnetometer, 421 reverting of MDR phenotypes, 505
Survivin, siRNA-mediated downregulation of, use of hematopoietic growth factors,
238 501–505
Syngeneic hematopoietic stem cell international working group (IWG) response
transplantion, 439 criteria, 395t–399t
IPSS-based strategy, 394
TAL1 gene, 180 iron chelation therapy, 406
T-cell clones in MDS, 238 for lower risk MDS, 407–409, 408f
T-cell mediated myelosuppression, 209 of anemia, 404–405
T cell receptor (TCR), 462 antithymocyte globulin (ATG), 406
TdT (terminal deoxynucleotidyl transferase), erythropoiesis stimulating agents (ESAs),
202 405
TEL(ETV6) gene, 62, 289 hypomethylating agents, 406
Telomerase, 71t lenalidomide, 405
Teniposide, 39 thalidomide, 405
TERC mutations, 319 for neutropenia, 406
Thalassemia, 96, 154, 418 response criteria, 394–397
Thalidomide, 405, 448 for thrombocytopenic lower risk MDS, 407
activity in leprosy, 459 Therapy-related MDS or AML , 37, 51, 61, 64,
alteration of microenvironment and stroma, 216–217, 318, 354
463 after autologous hematopoietic cell
antiemetic properties, 458 transplantation (HCT), 183
antitumor activity of, 459, 463 after breast cancer, 182
background, 458 cytogenetic abnormalities and, 187–189
clinical studies, 464t decision tree for the management of, 189f
cytogenetic responses with, 465 factors influencing, 175t, 183–185
cytokine modification, 462 genetic pathways, 181f
pharmacokinetics, 459 and cooperating mutations, 180–182
TNF-␣ production modification, 462 risk factors, 173–177, 184f
in treatment of MDS, 463–465 subtypes of, 177–179
THBS1 gene, 75 treatment, 185
Therapeutic strategies in MDS with balanced chromosomal
allogeneic SCT rearrangements, 185–186
from an HLA-identical sibling, 510–514 hematopoietic cell transplantation (HCT),
umbilical cord blood (UCB) 186–187
transplantation, 517 recommendations, 189–190
from volunteer unrelated donors, 514–517 Thrombocytopenia, 1, 60, 199, 287, 406–407,
for higher risk MDS, 403f, 518f 472–474
allogeneic SCT, 397–400 management of, 428–429
chemotherapy, 400–401 Thrombocytosis, 162–163, 205, 293
hypomethylating agents, 401–402 Thrombopoietin receptor gene (MPL), 162
investigational agents, 402 T lymphocytes, 439. See also CD8, CD4.
Index 541
TNF-␣ inhibitors, 442, 448 Vaccines, 449

TNF-␣ levels in MDS, 112, 113t Vascular endothelial growth factor (VEGF),
Topoisomerase II inhibitors, 40, 61, 178, 182, 463
186 VEGF receptors, 200
Topotecan–Ara C combination, 400 Vinca alkyloids, 39
TPBG gene, 275 Viral disorders, 208
TP53 gene, 16, 60-61, 64, 72t, 73, 74, 176, 177 Visceral leishmaniasis, 321
201, 319 Vitamin B12, 207
TRAIL (TNF-related apoptosis inducing ligand) Deficiency of, 321
receptor, 111, 463 Von Willibrand Factor (VWF), 201
Transferrin receptor (TfR), 425 VP16–Ara C combination therapy, 400
Transforming growth factor beta (TGF-␤), 115 V␤ profiles, 238
Transfusion-associated graft-versus-host VSIG4 gene, 275
disease, 418 V␥ 9V␦2 T cells, 443
Transient abnormal myelopoiesis, 328–329
Translocations, rare and recurring WASPIP gene, 75
abnormalities of 3q, 63 WHO Classification. See World Health
beta chain of the platelet-derived growth Organization Classification of MDS.
factor receptor (PDGFRB), 62–63 WHO Classification-Based Prognostic Scoring
11q23 abnormalities, 61 System (WPSS), 18t, 19, 55, 59, 196,
t(11;16), 61–62 253, 351, 380, 382t
Tretinoin, 185 WIG-1 gene, 93, 274
Trilineage dysplasia, 178f, 240, 242 Wnt/␤-catenin pathway, 75, 274
TRIP11, 63 World Health Organization Classification of
Trisomy 6, 367 MDS. 13–15, 14t, 28, 50t, 196, 203,
Trisomy 8, 92, 238, 316 204t
Trisomy 21, 367 Wright–Giemsa stained cell populations,
t(11;16) translocations, 61–62 248
Tumorigenesis, 92 WT1 gene, 72t, 319, 327, 449
Tumor necrosis factor-␣, see TNF-␣ WT1-specific cytotoxic T lymphocytes, 449
TUNEL (method of measuring apoptosis), 112,
118 X-linked sideroblastic anemia (XLSA), 154,
Turner syndrome, 316 157, 159t
Type III receptor tyrosine kinase (c-kit receptor), XLSA female heterozygotes, 159
257 XRCC3 gene, 39
Tyrosine kinase domain (TKD), 67
gene Younger adults, MDS in, 34–35
Yttrium-90 ibritumomab tiuxetan (ZevalinTM),
Unclassifiable (MDS/MPDU), 306–307 179
Unilineage dysplasia, 252
Uniparental disomy (UPD), 98, 100, 102, ZNF183 gene, 96
240–241 ZNF261 gene, 92
Hematology Second
about the book… Edition
Written by a team of leading authorities in pathogenesis, diagnostic techniques, and
Myelodysplastic
clinical management strategies in myelodysplastic syndromes (MDS), this text provides a
concise, easy-to-follow review of the advances in the science, classification, diagnosis, and
Pathobiology and Clinical Management

management of these challenging conditions.
An ideal source for hematologists, oncologists, and researchers, this Second Edition
Syndromes
features:
• a new eight-page color insert
• 200 color and black-and-white illustrations
• reworked content, organized into three sections: MDS epidemiology and biology;
MDS diagnosis, classification, and prognosis; and MDS therapy
Pathobiology and
• 21 thoroughly updated chapters, reflecting a shift in topical focus as dictated by
the evolution of the field
New topics in Myelodysplastic Syndromes include:
• del(5q) and 5q-syndrome, including new biological insights such as RPS14
down-regulation
• MDS-MPN overlap syndromes including CMML
Clinical Management
Second Edition
• the “reborn” DNA methyltransferase inhibiting nucleoside analogs, azacitidine
and decitabine
• the potential importance of iron overload in MDS
• global genomics approaches, such as gene expression arrays, array-based comparative
genomic hybridization, and single nucleotide polymorphism arrays
• the latest classification revisions, including the 2008 changes to the WHO classification
• the role of mitochondrial ferritin accumulation and other mitochondrial metabolic Edited by
anomalies in MDS
• other newly defined disease-associated alterations, such as abnormalities of
B lymphocyte populations
David P. Steensma
• three recently FDA-approved drugs for MDS: azacitidine, lenalidomide, and decitabine
about the editor...
DAVID P. STEENSMA is a Consultant and Associate Professor of Medicine and Oncology at
Mayo Clinic, Rochester, Minnesota, USA. Originally from the New York City suburbs, he
received his M.D. from the Pritzker School of Medicine at the University of Chicago,
Chicago, Illinois, USA. Dr. Steensma completed his clinical training in internal medicine,
hematology and medical oncology at Mayo Clinic, and research training at the Weatherall
Institute of Molecular Medicine in Oxford, England. His laboratory efforts focus on the
molecular genetics of MDS, and he is also currently conducting several clinical trials aimed
at improving outcomes and quality of life for patients with myelodysplastic syndromes
(MDS) and various forms of anemia. He is a member of the North Central Cancer Treatment
Group, Mayo Clinic Cancer Research Consortium, and the Eastern Cooperative Oncology
Group Leukemia Committee. Steensma
Printed in the United States of America H7439

Myelodysplastic Syndromes (2009) PDF

Uploaded by

Copyright:

Available Formats

Myelodysplastic Syndromes (2009) PDF

Uploaded by

Document Information

Original Title

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

Myelodysplastic Syndromes (2009) PDF

Uploaded by

Copyright:

Available Formats

Hematology Second

about the book… Edition

Pathobiology and Clinical Management

1. Chronic Lymphocytic Leukemia: Scientific Advances and Clinical Develop-

Library of Congress Cataloging-in-Publication Data

The landscape of the myelodysplastic syndromes (MDS) has changed consid-

by a section on diagnosis, classification, and prognosis, and concluding with a sec-

1. The Myelodysplastic Syndromes: History and Classification 1

2. Epidemiology of the Myelodysplastic Syndromes 27

3. The Cytogenetics and Molecular Biology of the Myelodysplastic

4. Genomic Approaches in MDS 87

5. The Role of Apoptosis in MDS 107

6. The Role of Mitochondria in MDS 127

7. Defects in Iron Metabolism and Iron Overload in MDS 153

8. Therapy-Related Myelodysplastic Syndrome and Myeloid Leukemia 173

9. Diagnosis of Myelodysplastic Syndromes: Criteria and Challenges 195

10. Hypocellular Myelodysplastic Syndrome: Relationship to Aplastic Anemia

11. Diagnostic and Prognostic Utility of Flow Cytometry in MDS 247

12. Molecular Pathogenesis of the 5q− Syndrome 267

13. Chronic Myelomonocytic Leukemia and Myelodysplastic Syndrome/

14. MDS in Children 311

15. Prognostic Factors in the Assessment of Patients with Myelodysplastic

17. Management of Cytopenias in MDS 413

18. Immune Dysregulation and the Role for Immunotherapy in

19. Lenalidomide Therapy in MDS 457

20. DNA Methyltransferase Inhibitor Therapy in the Treatment of

21. Intensive Chemotherapy and Stem Cell Transplantation in Myelodysplastic

Lionel Ades Hôpital Avicenne (Assistance Publique-Hôpitaux de Paris),

Henrik Hasle Department of Pediatrics, Aarhus University Hospital Skejby,

Masao Tomonaga Department of Hematology, Atomic Bomb Disease

INTRODUCTION: WHAT ARE “MYELODYSPLASTIC SYNDROMES”?

Table 1 Selected Important Events in the History of Classification and Treatment of

Year Event Source

1926 Giovanni Di Guglielmo in Naples describes “acute erythremia,” a (79)

Table 1 Selected Important Events in the History of Classification and Treatment of

Year Event Source

One of the first thorough literature reviews of “preleukemia” was published

EVOLVING CRITERIA FOR MDS DIAGNOSIS AND RISK ASSESSMENT

surrogate marker for monocyte/macrophage activity). Finally, the FAB group’s

Table 2 The 1982 French-American-British (FAB) Cooperative Group Revised

Limitations of the 1982 FAB Classification

Table 3 Important Cell Morphological Features to Assess in Suspected MDS Cases

Assessment of Prognosis: Steps Towards the International Prognostic

1. In multivariate regression analyses, a poor prognosis was apparent for older

Dutch Japanese German Lille IPSS

Year of 1985 1987 1989 1992 1993 1994 1996 1997

Hemoglobin abnormal Low white ≤100 × Presence of Hemoglobin cytopenias

Only series with >100 MDS patients are included.

2. The presence of a specific morphological pattern in the bone marrow, abnormal

Category score (sum all 3 for overall IPSS score)

Marrow blasts ⬍5 5–10 – 11–20 21–30b

blasts, number of cytopenias, cytogenetic subgroup (good, intermediate, and

Table 6 Risk Stratification of International Prognostic Scoring System (IPSS)

Low risk 0 5.7 11.8 4.8 9.4

The World Health Organization (WHO) Classifications of MDS

Table 7 2001 World Health Organization (WHO) Classification of Neoplastic Diseases

Categories relevant to the myelodysplastic Approximate proportion of

if clonal karyotypic findings include inv(16), t(8;21) or t(15;17).

The 2008 WHO Revision