Transcript of Technical and Biological Components of Marrow Transplantation
edited by
~.
Library of Congress Cataloging-in-Publication Data Technical and
biological components of marrow transplantation / edited by
C. Dean Buckner. p. cm. - (Cancer treatment and research; 76)
Includes bibliographical references and index. ISBN
978-1-4613-5832-9 ISBN 978-1-4615-2013-9 (eBook) DOI
10.1007/978-1-4615-2013-9 1. Bone marrow - Transplantation. 2.
Hematopoietic stern cells -
Transplantation. I Buckner, C. Dean. II. Clift, R.A. III. Series.
[DNLM: 1. Bone Marrow Transplantation. 2. Hematologic
Diseases - therapy. 3. Neoplasms - therapy. 4. Metabolie Diseases
therapy. W1 CA693 v. 761995 / WH 380 T255 1995] RD 123.5.T43 1995
617.4' 4 - dc20 DNLMIDLC for Library of Congress
Copyright © 1995 Springer Science+Business Media New York
Originally published by Kluwer Academic Publishers in 1995
Softcover reprint ofthe hardcover 1st edition 1995
95-1584 CIP
All rights reserved. No part of this publication may be reproduced,
stored in a retrieval system or transmitted in any form or by any
means, mechanical, photocopying, recording, or otherwise, without
the prior written permission of the publisher, Springer
Science+Business Media, LLC.
Printed on acidjree paper.
Cancer Treatment and Research
Muggia FM (ed): Cancer Chemotherapy: Concepts, Clinical
Investigations and Therapeutic Advances. 1988. ISBN
0-89838-381-1
Nathanson L (ed): Malignant Melanoma: Biology, Diagnosis, and
Therapy. 1988. ISBN 0-89838-384-6 Pinedo HM, Verweij J (eds):
Treatment of Soft Tissue Sarcomas. 1989. ISBN 0-89838-391-9 Hansen
HH (ed): Basic and Clinical Concepts of Lung Cancer. 1989. ISBN
0-7923-0153-6 Lepor H, RatliffTL (eds): Urologic Oncology. 1989.
ISBN 0-7923-0161-7 Benz C, Liu E (eds): Oncogenes. 1989. ISBN
0-7923-0237-0 Ozols RF (ed): Drug Resistance in Cancer Therapy.
1989. ISBN 0-7923-0244-3 Surwit EA, Alberts DS (eds): Endometrial
Cancer. 1989. ISBN 0-7923-0286-9 Champlin R (ed): Bone Marrow
Transplantation. 1990. ISBN 0-7923-0612-0 Goldenberg D (ed): Cancer
Imaging with Radiolabeled Antibodies. 1990. ISBN 0-7923-0631-7
Jacobs C (ed): Carcinomas of the Head and Neck. 1990. ISBN
0-7923-0668-6 Lippman ME, Dickson R (eds): Regulatory Mechanisms in
Breast Cancer: Advances in Cellular and
Molecular Biology of Breast Cancer. 1990. ISBN 0-7923-0868-9
Nathanson, L (ed): Malignant Melanoma: Genetics, Growth Factors,
Metastases, and Antigens. 1991.
ISBN 0-7923-0895-6 Sugarbaker, PH (ed): Management of Gastric
Cancer. 1991. ISBN 0-7923-1102-7 Pinedo HM, Verweij J, Suit HD
(eds): Soft Tissue Sarcomas: New Developments in the
Multidisciplinary
Approach to Treatment. 1991. ISBN 0-7923-1139-6 Ozols RF (ed):
Molecular and Clinical Advances in Anticancer Drug Resistance.
1991. ISBN 0-7923-1212-0 Muggia FM (ed): New Drugs, Concepts and
Results in Cancer Chemotherapy. 1991. ISBN 0-7923-1253-8 Dickson
RB, Lippman ME (eds): Genes, Oncogenes and Hormones: Advances in
Cellular and Molecular
Biology of Breast Cancer. 1992. ISBN 0-7923-1748-3 Humphrey G,
Bennett Schraffordt Koops H, Molenaar WM, Postma A (eds):
Osteosarcoma in Adolescents
and Young Adults: New Developments and Controversies. 1993. ISBN
0-7923-1905-2 Benz CC, Liu ET (eds): Oncogenes and Tumor Suppressor
Genes in Human Malignancies. 1993.
ISBN 0-7923-1960-5 Freireich EJ, Kantarjian H (eds): Leukemia:
Advances in Research and Treatment. 1993.
ISBN 0-7923-1967-2 Dana BW (ed): Malignant Lymphomas, Including
Hodgkin's Disease: Diagnosis, Management, and Special
Problems. 1993. ISBN 0-7923-2171-5 Nathanson L (ed): Current
Research and Clinical Management of Melanoma. 1993. ISBN
0-7923-2152-9 Verweij J, Pinedo HM, Suit HD (eds):
Multidisciplinary Treatment of Soft Tissue Sarcomas. 1993.
ISBN 0-7923-2183-9 Rosen ST, Kuzel TM (eds): Immunoconjugate
Therapy of Hematologic Malignancies. 1993.
ISBN 0-7923-2270-3 Sugarbaker PH (ed): Hepatobiliary Cancer. 1994.
ISBN 0-7923-2501-X Rothenberg ML (ed): Gynecologic Oncology:
Controversies and New Developments. 1994.
ISBN 0-7923-2634-2 . Dickson RB, Lippman ME (eds): Mammary
Tumorigenesis and Malignant Progression. 1994.
ISBN 0-7923-2647-4 Hansen HH (ed): Lung Cancer. Advances in Basic
and Clinical Research. 1994. ISBN 0-7923-2835-3 Goldstein FJ, Ozols
RF (eds): Anticancer Drug Resistance. Advances in Molecular and
Clinical Research.
1994. ISBN 0-7923-2836-1 Hong WK, Weber RS (eds): Head and Neck
Cancer. Basic and Clinical Aspects. 1994. ISBN 0-7923-3015-3 Thall
PF (ed): Recent Advances in Clinical Trial Design and Analysis.
1994. ISBN 0-7923-3235-0
Contents
1. Marrow Transplantation for Chronic Myeloid Leukemia. . . . . . .
. 1 REGINALD CLIFT
2. Bone Marrow Transplantation in Thalassemia. . . . . . . . . . .
. . . . . . 43 GUIDO LUCARELLI, and CLAUDIO GIARDINI
3. High-Dose Chemotherapy and Autologous Stem Cell Transplantation
for Breast Cancer. . . . . . . . . . . . . . . . . . . . . . . . .
. . 59 CHARLES WEAVER, ROBERT BIRCH, LEE SCHWARTZBERG, and WILLIAM
WEST
4. Bone Marrow Transplantation for Metabolic Diseases. . . . . . .
. . . 87 ROBERTSON PARKMAN, GAY CROOKS, DONALD KOHN, CARL LENARSKY,
and KENNETH WEINBERG
5. Cytomegalovirus Infection in Marrow Transplantation. . . . . . .
. . . 97 MICHAEL BOECKH, and RALEIGH BOWDEN
6. Marrow Transplantation from Unrelated Volunteer Donors. . . ..
137 CLAUDIO ANASETTI, EFFIE PETERSDORF, PAUL MARTIN, and JOHN
HANSEN
7. Peripheral Blood Stem Cell Transplantation ..................
169 WILLIAM BENSINGER
8. Umbilical Cord Blood Stem Cell Transplantation ............ "
195 JOHN WAGNER
v
9. In Vitro Expansion of Hematopoietic Cells for Clinical
Application .............................................. 215
STEPHEN EMERSON, BERNHARD PALSSON, MICHAEL CLARKE, SAMUEL SILVER,
PAUL ADAMS, MANFRED KOLLER, GARY VAN ZANT, SUSAN RUMMEL, R. DOUGLAS
ARMSTRONG, JAMES MALUTA, JUDITH DOUVIUE, and LESLIE PAUL
10. Recombinant Hematopoietic Growth Factors in Bone Marrow
Transplantation .......................................... , 225
JOHN NEMUNAITIS
11. Detection of Minimal Residual Disease ......................
249 JOHN GRIBBEN and LEE NADLER
12. Genetic Therapy Using Bone Marrow Transplantation ......... 271
RICHARD GILES, ELIE HANANIA, SIQING FU, and ALBERT DEISSEROTH
13 Myeloablative Radiolabeled Antibody Therapy with Autologous Bone
Marrow Transplantation for Relapsed B-Cell Lymphomas 281 OLLIE
PRESS, JANET EARY, FREDERICK APPLEBAUM, and IRWIN BERNSTEIN
14. Graft Versus Leukemia in Humans ......................... , 299
ANNA BUTTURINI and ROBERT PETER GALE
15. Interleukin-2 in Bone Marrow Transplantation. . . . . . . . . .
. . . . .. 315 UDIT VERMA, BISHAN CHARAK, CHITRA RASAGOPAL, and
AMITABHA MAZUMDER
16. Cellular Adoptive Immunotherapy after Bone Marrow
Transplantation . . . . . . . . . . . . . . . . . . . . . . . . . .
. . . . . . . . . . . . . . . .. 337 STAN RIDDELL and PHILIP
GREENBERG
Index ....................................................... ,
371
Contributing Authors
AD AMS, Paul T., Department of Internal Medicine, University of
Michigan 48105, 3105/Box 0368 Taubman Street, Ann Arbor,
MI48109
ANASETTI, Claudio, Fred Hutchinson Cancer Research Center,
Director, Unrelated Donor Transplant Program, 1124 Columbia Street,
Mailstop E611, Seattle, WA 98104
APPELBAUM, Frederick R., Fred Hutchinson Cancer Research Center,
1124 Columbia Street, M-l27, Seattle, WA 98104
ARMSTRONG, R. Douglas, Aastrom Biosciences, Inc., Domino Farms,
Lobby L, Ann Arbor, MI 48105
BENSINGER, William I., Fred Hutchinson Cancer Research Center, 1124
Columbia Street, Mailstop E100, Seattle, W A 98104
BERNSTEIN, Irwin D., Fred Hutchinson Cancer Research Center, 1124
Columbia Street, Cl-169, Seattle, WA 98104
BIRCH, Robert, Response Technologies, 1775 Moriah Woods Boulevard,
Memphis, TN 38117
BOECKH, Michael, Fred Hutchinson Cancer Research Center, 1124
Columbia Street, Mailstop AC142, Seattle, WA 98104
BOWDEN, Rowley, Fred Hutchinson Cancer Research Center, 1124
Columbia Street, Mailstop AC142, Seattle, WA 98104
BUTTURINI, Anna, Salick Healthcare, Inc., 8201 Beverly Boulevard,
Los Angeles, CA 90048
CHARAK, Bishan S., Georgetown University School of Medicine,
Department of Medical Oncology, 3800 Reservoir Road NW, Washington,
DC 20007-2197
CLARKE, Michael F., Department of Hematology-Oncology, 102
Observatory Street, Ann Arbor, MI 48109
CLIFT, Reginald, Fred Hutchinson Cancer Research Center, 1124
Columbia Street, Mailstop E100, Seattle, WA 98104
CROOKS, Gay, Instructor of Pediatrics, Children's Hospital, Los
Angeles, 4650 Sunset Boulevard, Los Angeles, CA 90027
DEISSEROTH, Albert, Department of Hematology, University of Texas,
MD Anderson Cancer Center, 1515 Holcombe Boulevard #24, Houston, TX
77030-4009
VB
DOUVILLE, Judith, Aastrom Biosciences, Inc., Domino Farms, Lobby L,
Ann Arbor, MI 48105
EARY, Janet F., Department of Nuclear Medicine, University of
Washington, 1959 NE Pacific Street, RC-70, Seattle, WA 98195
EMERSON, Stephen Chief, Division of Hematology/Oncology, University
of Pennsylvania, School of Medicine, 3400 Spruce Street,
Philadelphia, PA 19104-4283
FU, Siqing, Department of Hematology, University of Texas, MD
Anderson Cancer Center, 1515 Holcombe Boulevard #24, Houston, TX
77030- 4009
GALE, Robert Peter, Salick Healthcare, Inc., 8201 Beverly
Boulevard, Los Angeles, CA 90048
GIARDINI, Claudio, Department of Hematology, Hospital of Pesaro,
6110 Pesaro, ITALY
GILES, Richard, Department of Hematology, University of Texas, MD
Anderson Cancer Center, 1515 Holcombe Boulevard #24, Houston, TX
77030-4009
GREENBERG, Philip, Fred Hutchinson Cancer Research Center,
Director, Unrelated Donor Transplant Program, 1124 Columbia Street,
Mailstop AC100, Seattle, WA 98104
GRIBBEN, John, Tumor Immunology Division, Dana-Farber Cancer
Institution, 44 Binney Street, Boston, MA 02115
HANANIA, Elie G., Department of Hematology, University of Texas, MD
Anderson Cancer Center, 1515 Holcombe Boulevard #24, Houston, TX
77030-4009
HANSEN, John A., Fred Hutchinson Cancer Research Center, 1124
Columbia Street, M-718, Seattle, W A 98104
KOHN, Donald B., Associate Professor of Clinical Pediatrics and
Micro biology, Children's Hospital, Los Angeles, 4650 Sunset
Boulevard, Los Angeles, CA 90027
KOLLER, Manfred R., Aastrom Biosciences, Inc., Domino Farms, Lobby
L, Ann Arbor, MI 48105
LENARSKY, Carl, Associate Professor of Clinical Pediatrics,
Children's Hospital, Los Angeles, 4650 Sunset Boulevard, Los
Angeles, CA 90027
LUCARELLI, Guido, Department of Hematology, Hospital of Pesaro,
6110 Pesaro, ITALY
MALUTA, James, Aastrom Biosciences, Inc., Domino Farms, Lobby L,
Ann Arbor, MI 48105
MARTIN, Paul J., Fred Hutchinson Cancer Research Center, 1124
Columbia Street, M-718, Seattle, WA 98104
MAZUMDER, Amitabha, Georgetown University School of Medicine,
Department of Medical Oncology, 3800 Reservoir Road NW, Washington,
DC 20007-2197
NADLER, Lee, Tumor Immunology Division, Dana-Farber Cancer Institu
tion, 44 Binney Street, Boston, MA 02115
Vlll
NEMUNAITIS, John, Director of Clinical Research, Texas Oncology,
P.A., Director of Cytokine Research, Baylor University Medical
Center, PA Research #400, 3320 Live Oak, Dallas, TX 75204
PALSSON, Bernard 0., University of Michigan, Department of Chemical
Engineering, 2300 Hayward Street, Ann Arbor, MI48109-2136
PARKMAN, Robertson, Children's Hospital of Los Angeles, Department
of Immunology MS62, 4650 Sunset Boulevard, Los Angeles, CA
90027
PAUL, Leslie, Aastrom Biosciences, Inc., Domino Farms, Lobby L, Ann
Arbor, MI 48105
PETERSDORF, Effie W., Fred Hutchinson Cancer Research Center, 1124
Columbia Street, M-718, Seattle, W A 98104
PRESS, Ollie, Assistant Professor of Medicine, University of
Washington Medical Center, Mailstop ED-08, 1959 NE Pacific,
Seattle, WA 98111
RAJAGOPAL, Chitra, Georgetown University School of Medicine,
Depart ment of Medical Oncology, 3800 Reservoir Road NW,
Washington, DC 20007-2197
RIDDELL, Stan, Fred Hutchinson Cancer Research Center, Director,
Unrelated Donor Transplant Program, 1124 Columbia Street, Mailstop
AC100, Seattle, WA 98104
RUMMEL, Susan, Aastrom Biosciences, Inc., Domino Farms, Lobby L,
Ann Arbor, MI48105
SCHWARTZBERG, Lee S., Response Technologies, 1775 Moriah Woods
Boulevard, Memphis, TN 38117
SILVER, Samuel M., Department of Hematology-Oncology, 102 Observa
tory Street, Ann Arbor, MI 48109
VAN ZANT, Gary, Aastrom Biosciences, Inc., Domino Farms, Lobby L,
Ann Arbor, MI 48105
VERMA, Udit, Georgetown University School of Medicine, Department
of Medical Oncology, 3800 Reservoir Road NW, Washington, DC 20007-
2197
WAGNER, John, Department of Pediatrics, University of Minnesota,
Box 366 UMHC, 420 Delaware Street, SE, Minneapolis, MN 55455
WEAVER, Charles, Response Technologies, 1775 Moriah Woods
Boulevard, Memphis, TN 38117
WEINBERG, Kenneth, Associate Professor of Pediatrics, Children's
Hospital, Los Angeles, 4650 Sunset Boulevard, Los Angeles, CA
90027
WEST, William H., Response Technologies, 1775 Moriah Woods
Boulevard, Memphis, TN 38117
ix
Preface
This is not a textbook and it is not intended to be a work of
reference. We hope it is a book that can be read from cover to
cover by physicians and scientists involved with, or interested in,
bone marrow transplantation. The objective is to present up-to-date
information and recent citations. For the most part, the
contributions are directed at scientific and technologic advances
designed to extend and improve the clinical application of treat
ment usually described as bone marrow transplantation.
Two chapters deal with the treatment of chronic myeloid leukemia
(CML) and thalassemia, which are spectacularly successful
applications of allogeneic marrow transplantation that have now
become conventional therapy. These therapies are still being fine
tuned, particularly with a view to increasing the number of
patients who can avail themselves of this treatment. The use of
volunteer unrelated donors is clearly an option favored by the pace
of disease progression in these diseases, and it is already widely
used for CML. Autologous marrow transplantation is an option that
will be studied for both diseases. In the case of CML, the rapidly
increasing understanding of the molecular biology of the underlying
genetic flaw will add special oppor tunities to studies of in vivo
or in vitro purging. In the case of thalassemia, autologous
transplantation will provide the vehicle for introducing the
genetic revisions needed for cure.
The term bone marrow transplantation is not always an accurate des
cription of the field we cover. Increasingly it is applied to the
reinfusion of autologous hematopoietic progenitors, either as part
of a strategy of ex vivo marrow protection or as a vehicle for
introducing genetic change. Indeed, it is likely that even
allogeneic marrow transplantation soon will be routinely
accomplished by the transfer of peripheral blood stem cells rather
than bone marrow. However, bone marrow transplantation has a nice
old-fashioned ring to it, and the phrase will probably continue in
use to describe any manipulation that involves the reconstitution
of the hematopoietic system.
The development of this field was made possible by advances in
supportive care, including platelet transfusions and powerful
antibiotics, and these advances have continued to the point where
allogeneic transplants can now be performed with very little
morbidity and mortality in patients who do not
Xl
have a big legacy of organ damage from intensive prior therapy.
Improved management of cytomegalovirus infection should have a
dramatic impact on survival after allogeneic transplantation. The
chapters dealing with cytokines and progenitor expansion indicate
that the time is not far distant when marrow transplantation can be
contemplated as an outpatient procedure. This will undoubtedly have
an influence on the timing of transplantation. We may soon be able
to define circumstances in which early transplantation for leukemia
will be less dangerous and more effective than initial remission
induction without the support of early marrow reconstitution from a
trans plant. The genetic implantation of resistance factors into
stem cells could enable a survival advantage over resident
hematopoiesis for modified, rein fused stem cells, permitting the
exploitation of selection pressures favoring the modified cell
population over an extended time scale. This may encour age the
development of chronic therapy removing some of the cataclysmic
associations that history has bestowed on marrow
transplantation.
Enthusiastic investigators of cord blood stem cell technology
clearly con template global depositories of stem cells collected
at the time of birth and, perhaps, outliving the unwitting donors:
A brave new world indeed.
Soon more patients will receive marrow transplants as part of the
treat ment of solid tumors and metabolic diseases than for the
treatment of hematologic disease. The contributions in this volume
explicitly describe some of these applications and contain hints of
other exciting possibilities. Marrow transplantation is clearly
here to stay.
xii
Reginald A. Clift
Introduction
Chronic myeloid leukemia (CML) is a relatively common disease
mainly afflicting patients older than 40 years. It was the first
malignant disease shown to be associated with a change in
chromosomal pattern [1,2], and the molecular biology of CML has
been intensively investigated [3-6]. It is one of a group of
leukemias known to arise because of a translocation that
repositions part of the c-ABL proto-oncogene situated on chromosome
9 to a position adjacent to the breakpoint cluster region (BCR) on
chromosome 22 [3,4]. This translocation usually produces a
distinctively malformed chromosome 22 (referred to as the Ph
chromosome) and always creates a length of corrupted genetic
information known as the BCR-ABL rear rangement. The resulting
fusion gene directs the synthesis of chimeric proteins (p2IOBCR-ABL
or pI90BCR-ABL) with readily detectable in vitro tyrosine kinase
activity [5]. It is thought that these proteins arise as a result
of different breakpoint locations in the BCR region.
The p190 protein is found in approximately 50% of patients with Ph
positive acute lymphocytic leukemia (ALL), whereas more than 90% of
patients with Ph-positive CML have the p210 protein. The
introduction of the base sequence associated with the BCR-ABL
rearrangement into the hematopoietic cells of mice can produce a
disease with most of the charac teristics of human CML [6].
Studies of female patients with CML who were heterozygous for
polymorphic markers on the X chromosome demonstrated that the
population of leukemia cells in CML is monoclonal. This suggests,
but does not prove, that the disease resulted from a single event
in a primitive myeloid precursor. Clonal studies in certain cell
popUlations in humans suggest there may be a stage of evolving
monoclonality that precedes the detection of BCR-ABL transcripts
[7-9]. However, an emerging body of data from molecular and other
studies suggests that the development of the BCR-ABL rearrangement
in hematopoietic stem cells is the determining event for the
development of CML [10-12].
The genetic abnormality in CML occurs in multipotential, and
probably the most primitive multipotential, hematopoietic cells,
and the Ph chromo-
C. Dean Buckner (ed.), TECHNICAL AND BIOLOGICAL COMPONENTS OF
MARROW TRANSPLANTATION. Copyright © 1995. Kluwer Academic
Publishers, Boston. All rights reserved.
some can usually be demonstrated in granulocyte and red cell
precursors and megakaryocytes but not in lymphocytes. Patients who
relapse after treatment by bone marrow transplantation (BMT)
usually have cells of host origin in granulocyte, red cell, and
megakaryocyte lines. CML can be cured by BMT [13,14] and is a
particularly interesting disease for the study of many aspects of
this form of treatment. Because of the unique cytogenetic pattern
in malignant cells, the disease is easily detected with great
sensitivity, and this facilitates both early transplantation and
the recognition of very low levels of residual malignancy
[13-15].
When first diagnosed, the disease is usually distinguished by
increased proliferation of normally maturing granulocytes and very
mild symptom atology, often limited to the consequences of
splenomegaly. This stage of the disease is referred to as the
chronic phase (CP), which usually persists for a period of years.
Eventually the character of the disease changes with transformation
into a stage known as blast phase (BP), characterized by disorderly
maturation and increased symptomatology. This phase resembles acute
leukemia, usually with myeloid characteristics but sometimes is
clearly lymphoid in nature. Frequently, transformation to BP is
anticipated by the development of an accelerated phase (AP) with
increased symptomatology and hematologic and cytogenetic changes
[16-18]. A very small number of reports describe extremely
prolonged survival in CP [19,20], sometimes for more than 20 years,
but for the overwhelming majority of patients CML is fatal within 5
years unless treated by BMT or, perhaps, with interferon
(IFN).
A 1924 study of 166 cases of CML suggested that the median survival
from diagnosis was about 3 years and that this was not improved by
the only treatments available at the time (radium or X-irradiation)
[21]. Several studies with large numbers of patients have looked
for patient characteristics present at diagnosis that predict for
survival. Tura et al. examined the prognostic value of nine
clinical and hematologic features recorded at diagnosis in 255
Italian patients and found that six characteristics could be used
to classify these patients into three prognostic categories. The
prog nostic value of this classification was then confirmed in a
further series of 153 patients [22]. Sokal et al. used a Cox model
to examine a 'training' population of 361 patients (including many
of the patients in the Tura study) and devised an algorithm that
was then applied to a 'test' population of 317 patients. This
approach permitted the classification of 'good risk' newly
diagnosed patients (i.e., patients not in BP) into three roughly
equal groups with median survivals of 2-3, 3-4, and 5-6 years [23].
Spleen size, per centage of circulating blasts, platelet count,
and age were the only features with unequivocal prognostic
significance.
Three forms of treatment are currently used for patients with CML.
These are palliation with chemotherapy, treatment with IFN, and
BMT. Patients who have suitable marrow donors may be treated with
BMT, which offers a high probability of cure at the cost of some
early morbidity and
2
mortality. Treatment with IFN can reduce the size of the leukemic
clone (as identified by cytogenetics) in 40-50% of patients and
eliminate the clone [as determined by cytogenetics and occasionally
by polymerase chain reaction (PCR) determination of BCR-ABL
transcripts] in 5-15%. Randomized studies indicate that populations
of patients treated with IFN have some prolongation of survival
compared with patients treated with hydroxyurea or busulfan. No
randomized studies have been conducted comparing survival of
patients who have been transplanted with that .of patients
receiving palliative chemotherapy. Treatment with IFN is associated
with substantial continuing cost, discomfort, and disability.
Treatment by BMT is expensive and involves much disability and
discomfort, which is transient for most patients but permanent for
a small proportion.
Because CML is a disease for which there is a reasonable prospect
of quite prolonged and comfortable survival for patients treated
with palliation, the selection of intervention strategies for
treatment with IFN or BMT is important and difficult. This chapter
deals with the use of BMT for the treatment of CML and pays
particular attention to the issue of timing. It is, however,
impossible to discuss this topic without considering alternative
approaches.
Diagnosis
The most common hematologic abnormalities at diagnosis are marked
granulocytosis and thrombocytosis. These abnormalities may exist
for years without symptoms and, in societies with modern health
care practices, they may be detected unexpectedly in the course of
routine medical examinations. The most common presenting symptom is
early satiety and abdominal discomfort related to the most common
presenting sign, which is sple nomegaly. Diagnosis is based on the
demonstration of Ph chromosomes in marrow metaphases. Keating et
al. [24] demonstrated that about 25% of colonies from cultured
marrow of patients with Ph-positive CML were Ph positive but
BCR-ABL negative by PCR. The reason for this is not known, and the
implications of this finding for the use of BCR-ABL in diagnosis
and in monitoring patients for relapse are unclear. Because
cytogenetic exami nation of the marrow or peripheral blood is
essential to staging and Ph chromosome-positive patients are nearly
always positive with PCR testing for the BCR-ABL rearrangement,
usually it is not necessary to perform molecular analyses as a
routine diagnostic procedure. Occasionally, patients with
apparently typical CML will lack Ph chromosomes and BCR-ABL
positivity can be detected in marrow or blood by Southern blot or
PCR techniques. In such instances, the translocation is complex and
hidden, but the disease behaves in all respects like Ph-positive
CML. Patients with myeloproliferative disorders in which the
BCR-ABL rearrangement cannot
3
be detected usually have a disease other than classical CML, and
this chapter does not deal with the treatment of such
conditions.
Staging
Patients in CP have stable disease with only minor symptomatology,
no extramedullary disease, and with granulocyte and platelet counts
easily controlled by palliative chemotherapy (see below). The
definition of AP has been contentious [25,26] but requires at least
one of the following findings: 1. The persistent presence of 10-30%
myeloblasts in marrow or peripheral
blood 2. Major perturbations of white blood cell count (>50 x
109/L), platelet
count «100 or >1000 x 109/L), and hematocrit «25%) uncontrolled
by chemotherapy with busulfan (BU), hydroxyurea (HU), or IFN
3. Progressive splenomegaly 4. Extramedullary tumor 5. The presence
of any nonconstitutional cytogenetic abnormality in addition
to a single Ph chromosome 6. Persistent unexplained fever or bone
pain Blast phase is associated with more than 30% myeloblasts in
marrow or blood.
Palliation
The use of chemotherapy in doses intended to control the
hematologic manifestations and symptoms of CML in CP has usually
been referred to as conventional therapy. However, the widespread
acceptance of BMT and IFN as nonexperimental therapy means that
these therapies must also be considered conventional, and, given
the results, treatments other than these are best described as
palliative.
Early attempts at palliation used either total body or splenic
irradiation or the isotope 32p. This relieved the discomfort
associated with hypersplenism and decreased very high granulocyte
and platelet counts, but studies sug gested that the treatment did
not produce significant prolongation of survival [21]. Several
drugs have been shown to control the hematologic and clinical
manifestations of the disease, but still there has been no major
prolongation of survival, and the continued presence of the
abnormal leukemic clone is signalled by the persistence of
metaphases containing Ph chromosomes [7].
Busulfan is a drug with activity against the most primitive myeloid
stem cells and was the first drug demonstrated to have a major
impact on the quality of life for patients in CPo In low doses, it
is effective in reducing platelet and granulocyte counts, and in
reducing spleen size for patients in CPo Unfortunately it does not
delay the development of AP and BP, and it
4
probably does not increase the duration of survival [27].
Interestingly there have been several reports of patients receiving
'overdoses' of busulfan with cure of the CML but with subsequent
death from aplastic anemia [28-30]. This demonstrates that the drug
can eliminate the disease clone without producing lethal changes in
organs other than the bone marrow, but clearly the patients lacked
normal precursors, probably due to the busulfan treat ment, and
there was no healthy marrow available to replace the diseased
cells. Busulfan in doses used for the control of CP has been
relatively nontoxic, but there has been concern about the
development of pulmonary complications in patients treated for a
long time with the drug [31-34].
Another drug that controls symptoms and counts in patients in CP is
hydroxyurea, which does not eliminate marrow cell precursors in
otherwise tolerated doses but is effective in reducing granulocyte
and platelet levels. Whereas busulfan has activity against the most
primitive myeloid precursors, this is almost certainly not true of
hydroxyurea. In a very large randomized study the median survival
of patients treated with hydroxyurea was signifi cantly longer
than that of patients treated with busulfan [35]. There have been
no reports of irreversible marrow aplasia in patients treated with
hydroxyurea, and the drug cannot eradicate the leukemic clone.
However, hydroxyurea has emerged as the drug of choice for
controlling the manifes tations of CML in CP because of the
relative freedom from side effects, including a lesser adverse
effect on subsequent BMT [14].
During the past decade there has been some improvement in the
survival of patients treated with palliation [36]. This has
probably been a consequence of earlier diagnosis and the use of
hydroxyurea for palliation instead of busulfan.
Interferon
In 1983, Talpaz and his colleagues at the M.D. Anderson Hospital
reported that the administration of partially purified human a-IFN
produced a cytoreductive effect and hematologic remissions in
patients with CML [37]. Encouraged by these results they studied
the use of recombinant IFN, and in 1986 they reported that its use
in 17 patients with CML in CP resulted in hematologic remissions in
14 patients and cytogenetic improvements in 6 [38]. It was shown
that cytogenetic remissions from CML in CP induced by IFN resulted
in polyclonal myelopoiesis [39,40]. In 1991, a study of IFN in 96
consecutive patients treated less than 1 year from diagnosis
revealed a complete hematologic response in 73 %, a partial
cytogenetic response in 19%, and a complete cytogenetic response in
7% [41]. This was the first study to show sustained complete
cytogenetic responses in a subset of patients with CML after any
form of treatment other than BMT.
Stimulated by these findings, several large cooperative group
studies of this form of treatment have been conducted in an attempt
to determine
5
whether treatment with IFN is beneficial in terms of prolonging
survival. Ozer et al. reported a multi-institution study of 107
patients with previously untreated CML in CP given 5 x 106 IU/m2
subcutaneously daily [42]. Most patients had initial toxicity with
ftulike symptoms, but this usually resolved after several weeks of
therapy. However, severe chronic fatigue occurred in 26%, grade 3
hepatotoxicity in 12%, and neurologic symptoms in 10% of patients.
Sixty-three patients (59%) had some form of hematologic remission,
which was complete in 22%. Cytogenetic responses were observed in
40% of patients with cytogenetic follow-up, but 27 of the 107
patients had no follow-up because of failure to achieve hematologic
response or disease progression. Analyses of the effect of
cytogenetic response upon survival using time-dependent covariate
and landmark techniques failed to provide statistically significant
evidence of survival benefit from cytogenetic response.
The Italian Cooperative Group on CML conducted a study in which all
previously untreated or minimally treated patients with CML between
1986 and 1988 were randomly assigned to treatment with either IFN
(218 patients) or palliative chemotherapy (104 patients) [43]. The
dose of IFN was increased from 3 x 106 IU/day to 9 X 106 IU/day at
1 month. One patient in each arm died from therapy-related
complications and toxicity was greatest in the IFN arm. The time to
progression from CP to AP or BP was significantly longer in the IFN
group than in the chemotherapy palliation group (median >72
months vs. 45 months; p = 0.002). Treatment with IFN is unpleasant
and expensive, and the authors concluded that the optimum
circumstances for obtaining a good result needed investigation.
Thus far it appears that the dose of IFN must be large, that the
patient should be in CP, and that treatment should be initiated
early in the disease. Older patients (>60 years) tolerate the
treatment less well than younger ones [36].
The superiority of treatment with IFN over palliative chemotherapy
is undoubted but, although these results are very exciting, there
is very little experience with the discontinuation of expensive and
toxic treatment, and there is as yet no evidence to indicate that
patients with CML can be cured with this form of therapy.
Marrow transplantation
Identical twins
In 1979 Fefer et al. reported experience in transplanting four
patients in CP from identical twin donors after treatment with a
busulfan derivative (dimethylbusulfan), cyclophosphamide, and a
single exposure of 920 cGy of total body irradiation (TBI). The
leukemic clone was successfully eliminated in all patients [44].
These studies were extended and in 1982 a report described the
total experience of transplants for CML from identical twins (22
patients, 12 of them in CP) [45]. Figure 1 presents the
probabilities of
6
0.2
0 0 2 4 6 8 10 12 14 16 18
YEARS
Figure 1. The probabilities of survival and relapse for 12 patients
with CML in CP transplanted from syngeneic donors before May 1981
after a regimen of CY, dimethylbusuIfan, and TBI and first reported
in 1982 [45]; survival and events updated as of April 1994.
survival and relapse for the 12 patients transplanted in CP,
updated as of April 1994. It is clear from this figure that
syngeneic transplantation with an effective conditioning regimen
has a high probability of curing patients with CML in CP and that
an allogeneic effect is not essential for cure. This topic and
subsequent experience in transplantation from identical twins is
discussed in more detail in the section dealing with the biology of
cure.
HLA-identical related donors
Early experience with BMT using donors selected by
histocompatibility typing was limited to patients with advanced
disease. Initial results of 14 such transplants for advanced CML
were reported in 1978 [46]. All patients died and only one survived
for more than 1 year.
Chronic phase. Encouraged by the demonstration that prolonged
disease free survival could be obtained in patients transplanted
in CP from identical twins, the first marrow transplants from
HLA-identical siblings for CML in CP took place in Seattle in 1979,
and the first 10 such transplants were reported in 1982 [47].
Figure 2 depicts the probabilities of survival and relapse for
these patients updated through March 1994 and presents com pelling
evidence that this form of therapy has the potential to cure CML.
Four of these patients died, all within 100 days of transplant
[three from interstitial pneumonia (IP) and one from acute
graft-versus-host disease
7
YEARS
10 12 14
Figure 2. The probabilities of survival and relapse for the first
10 patients transplanted in Seattle for CML in CP from
HLA-identical siblings and reported in 1982 [47]. Survival and
events updated as of April 1994.
(GVHD)]. Only one patient relapsed (4.3 years after transplant),
and that patient was promptly treated with a second transplant from
the same donor and survives 4.2 years after the second and 8.7
years after the first transplant.
These results and those of others [48,49] led to an increased use
of transplantation for patients with CML, and in 1986 the Seattle
team published its experience with 167 patients with CML
transplanted through 1983 from HLA-identical siblings [15]. This
report revealed many of the opportunities and problems provided by
this form of therapy, which have since been amply confirmed by many
investigators. The probabilities of survival and relapse for these
patients updated through March 1994 are presented in Figures 3 and
4. It is clear from Figure 3 that phase at the time of transplant
is an important determinant of post-transplant survival. Thirty-one
of the 67 patients transplanted in CP through 1983 were alive and
disease free between 9% and 14.2 years after transplant, and the
latest relapse among this group of patients was at 5.3 years after
transplant. Transplantation during CP gave by far the best results,
whereas in this analysis there was not a lot of difference between
the results of transplantation during AP or BP. Sur prisingly, the
survival of 12 patients transplanted during remission after being
in BP was as good as for the patients transplanted in CP, and none
of these patients relapsed.
Figure 4 illustrates the problem of describing post-transplant
relapse. Eighteen CP patients had a reappearance of Ph-positive
metaphases in the marrow after transplantation, but in six patients
this reappearance was transient, with Ph-positive metaphases
subsequently becoming undetectable. One of the patients with
transient cytogenetic relapse and 12 other patients
8
YEARS
10 12 14 18
Figure 3. The probabilities of survival for 167 patients
transplanted in Seattle from HLA identical siblings for CML
through 1983 and reported in 1986 (47). Survival and events updated
as of April 1994.
RELAPSE 1
0.8 BLAST
0.8 ACCELERATED
YEARS
Figure 4. The probabilities of relapse for 167 patients
transplanted in Seattle from HLA identical siblings for CML
through 1983 and reported in 1986 [47). Survival and events updated
as of April 1994.
developed clinical relapse. All patients who developed clinical
relapse have died, and all five patients with transient cytogenetic
relapse are alive with no evidence of leukemia. The cytogenetic
marker associated with CML provides a sensitivity for detecting
'relapse' not available in most other transplant
9
situations, and we still do not know how best to utilize this
sensitivity to reduce the incidence of clinical relapse.
Another unanticipated finding was that the interval from diagnosis
to transplant influenced the outcome of BMT for patients
transplanted in CP and of patients transplanted in AP. This
observation has now been confirmed in many studies, is important to
the design of treatment strategies, and is discussed in more detail
later.
For cytoreduction nearly all the CP patients reported in the 1986
paper received a regimen of cyclophosphamide 120mg/kg followed by
2.0Gy of TBI on each of six successive days. These patients,
together with patients in first remission of acute myeloid
leukemia, were entered into studies of prophylaxis against GVHD,
first comparing 100 days of weekly intravenous methotrexate (MTX)
with 6 months of treatment with cyclosporine (CSP) [50], and then
comparing the CSP regimen to the same regimen with four doses of
MTX (MTX-CSP) [51]. These randomized trials clearly demon strated
that the MTX-CSP regimen was superior in reducing the incidence of
acute GVHD and in improving survival for patients transplanted in
CP of CML. Building on this experience, studies were designed
seeking cytoreduc tive regimens with a lower probability of
post-transplant relapse. It was demonstrated that an increase in
TBI dose from 12.0 Gy in six exposures to 15.75 Gy in seven
exposures was effective in reducing the incidence of relapse but
did not improve survival or disease-free survival due to an
increase in nonrelapse mortality [52].
In 1987, Tutschka reported the use of a conditioning regimen
consisting of busulfan (BU; 16mg/kg administered over 4 days),
followed by 60mg/kg cyclophosphamide (CY) on each of 2 successive
days [53]. This regimen (BU-CY) had low toxicity, was effective in
facilitating allogeneic engraft ment, appeared to be particularly
effective in the treatment of patients with myeloid malignancy, and
was used increasingly in the treatment of patients with CML
[54,55]. In 1988 a randomized study was initiated comparing this
regimen with the CY + 12.0 Gy TBI regimen in patients receiving
marrow transplants from HLA-identical related donors for the
treatment of CML in CP [56]. All patients received MTX + CSP for
GVHD prophylaxis. There was no significant difference between the
CY-TBI and the BU-CY groups in the 3 year probabilities of survival
(0.80 for both), in relapse (0.13 for both), in event-free survival
(CY-TBI 0.68, BU-CY 0.71), in speed of engraftment, or in incidence
of veno-occlusive disease of the liver. The 4 year probabilities of
survival and event-free survival for patients transplanted within 1
year of diagnosis were 0.86 and 0.72, respectively, for each group.
Significantly more patients in the CY-TBI group experienced major
creatinine elevations. There was significantly more acute GVHD in
the CY-TBI group. Fever days, positive blood cultures,
hospitalizations, and inpatient hospital days were significantly
more common in the CY-TBI group than in the BU-CY group.
In a major study of veno-occlusive disease (VOD) by McDonald et
al.
10
[57] the incidence of severe VOD in 45 patients transplanted for
CML in CP with TBI-containing regimens was 4%. Biggs et al. [55]
have reported the results of allogeneic marrow transplantation
after treatment with BU-CY in 115 patients with CML (62 in CP).
Patients in CP transplanted within 1 year of diagnosis had a 4 year
survival of 70%, and the authors concluded that the survival
statistics and transplant-related mortality were similar to those
seen in patients conditioned with regimens containing CY-TBI. The
inci dence of VOD in patients transplanted in CP was 6.6%. Essell
et al. [58] reported that in patients receiving MTX plus CSP for
GVHD prophylaxis, hepatotoxicity (particularly VOD) was
significantly higher for patients conditioned with BU-CY than for
those conditioned with CY-TBI. The study did not allocate treatment
by randomization, and it involved patients with several different
types and stages of leukemia. In the studies reporting the use of
BU-CY in patients with CML in CP, there is no consistent evidence
of an increase in hepatotoxicity compared with that seen after CY
TBI, whereas there is consistent evidence of an increase in VOD in
patients with advanced CML or other hematopoietic malignancies
receiving BU-CY. One of the reasons for this difference may be the
much greater exposure to pre transplant chemotherapy experienced by
such patients.
For the 101 patients transplanted within 1 year of diagnosis, the
4-year probability of survival with either regimen was 0.86. In a
recent report, a regimen consisting of VP16 and TBI produced
results in patients in CP similar to those seen with the BU-CY and
CY-TBI regimens [59]. Thus, there are now three regimens that
produce excellent and similar results in terms of survival and
disease-free survival. The number of patients required for
randomized studies aimed at improving this survival would be very
large, and it will be difficult to devise a practicable study of
regimens aimed at improving survival. The BU-CY regimen offers
opportunities for studying protocols that might reduce the
toxicity, cost, and inconvenience of BMT in this setting.
A long follow-up will be required to determine whether the known
late effects of CY-TBI (which have been reported to include growth
retardation and the development of cataracts and second
malignancies [60,61]) also occur in patients treated with BU-CY.
The problem of post-transplant relapse remains both complex and
challenging [13]. The 3-year probability of persistent cytogenetic
relapse with the three regimens was between 0.10 and 0.20. The
testing of conditioning regimens for improved antileukemic effect
will be very difficult. It may be more rewarding to study the
effect of the treatment of, or prophylaxis against, clinical
relapse in patients identified after transplantation as being at
high risk for this event [62]. For this purpose we need a better
understanding of the nature and definition of post transplant
relapse, and this problem is discussed in more detail later.
From 1983 through 1993, 327 patients in CP were transplanted in
Seattle from HLA-identical siblings after either CY-TBI or BU-CY
with MTX-CSP for prophylaxis against acute GVHD. Cox multivariate
analyses were
11
performed examining the influence of pre transplant variables upon
survival and post-transplant relapse. The characteristics examined
were patient and donor age; the four permutations of patient and
donor gender; the interval from diagnosis to transplant by day as a
continuous variable and categorized as less than 1 year, between 1
and 2 years, and more than 2 years; cyto megalovirus (CMV)
seropositivity of patient and donor; and the patient's spleen size
at diagnosis and transplant. In the analysis of the impact of these
variables upon survival, patient age less than 35 years versus
patient age greater than 35 and less than 51 years, transplantation
within 1 year of diagnosis, and female gender of both patient and
donor were independently associated with survival, and all were
beneficial. When patient age greater than 50 years was compared
with patient age between 35 and 50 years, there was no significant
difference either univariately or in the multivariate analysis.
When time-dependent covariates representing the development of
acute GVHD grade 2 or worse, the development of acute GVHD grade 3
or 4 (severe acute GVHD), and the development of clinically
extensive chronic GVHD were entered into the model, age less than
35 years ceased to be independently influential (suggesting that
the adverse impact of increasing age may be associated with acute
GVHD) , and both severe acute and chronic GVHD were independently
adversely influential. In the analysis with relapse as the
endpoint, only female donor gender was independently influential,
and this was beneficial irrespective of patient gender. None of the
variables representing acute or chronic GVHD was influential in
either the univariate or multivariate analysis. These results are
presented in Table 1. Figures 5-7 present the influence of age,
interval from diagnosis to transplant, and patient and donor gender
on the Kaplan-Meier statistics for survival and relapse.
Of the 327 patients, 49 developed persistent relapse and 7 of these
received second transplants from the same donor. The Kaplan-Meier
proba bilities of survival were 0.68 at 7 years after the first
transplant and 0.65 at 5 years from relapse for these 49 patients
(Figure 8). Only one of the survivors had received a second
transplant, but many of the patients had received other therapy,
including IFN and infusions of donor lymphocytes. This relatively
prolonged survival of patients who have relapsed after BMT is
surprising and has been reported by others [63].
It is particularly important to have an understanding of the
influence of age on outcome because the median age at diagnosis of
CML is relatively high. Reports from individual referral centers
usually indicate a median age at diagnosis of less than 50 years.
However, the National Cancer Institute Report of Surveillance,
Epidemiology, and End Results for the United States lists the
median age at death of patients with CML as 65.8 years, indicating
that the median age at diagnosis for patients not selected by
referral is close to 60 years [64]. Because of this age structure
and because increasing age is believed to exert a powerful adverse
influence on the outcome of allogeneic BMT, most patients with
newly diagnosed CML
12
0.8
0.4
0.2
8 10
Figure 5. The probabilities of survival by age about the median for
327 patients transplanted in Seattle from HLA-identical donors for
CML in CP between 1983 and 1994 using CY-TBI or BU-CY with MTX +
CSP as prophylaxis against acute GVHD.
SURVIVAL 1
0.8
0.4
0.2
8 10
Figure 6. The probabilities of survival by the interval from
diagnosis to transplant for 327 patients transplanted in Seattle
from HLA-identical donors for CML in CP between 1983 and 1994 using
CY-TBI or BU-CY with MTX + CSP as prophylaxis against acute
GVHD.
13
0.8
0.4
o~~~--~----~------~----~------~--
8 10
Figure 7. The influence of patient and donor gender on the
probabilities of survival and relapse for 327 patients transplanted
in Seattle from HLA-identical donors for CML in CP between 1983 and
1994 using CY-TBI or BU-CY with MTX + CSP as prophylaxis against
acute GVHD.
Table 1. Cox multivariate analyses of factors influencing outcome
in 327 patients transplanted in CP with currently used
regimensa
95% confidence Covariate p value Risk ratio interval
Mortality Patient and donor both female 0.035 0.45 0.22-0.94 Less
than 1 yr diagnosis to transplant 0.0004 0.43 0.27-0.69 Acute GVHD
grade 3 or 4 0.0010 2.52 1.45-4.36 Clinical extensive chronic GVHD
0.0077 2.15 1.22-3.77
Relapse Donor gender female 0.0003 0.376 0.22-0.64
a CY 120 mg/kg + six daily exposures each of 2.0 Gy TBI or BU 16
mg/kg + CY 120 mg/kg. All GVHD prophylaxis was with MTX-CSP.
are never offered the option of BMT, even if they have suitable
donors. However, the Seattle experience using current regimens
suggests that for patients in CP the subsequent deterioration of
survival expectations asso ciated with increased age is very small
over the age of 35, and patients over the age of 50 with newly
diagnosed CML in chronic phase can derive substantial benefit from
transplantation from HLA-identical related donors. Through 1993, 47
patients 50 years of age or older (17 were aged 56-60 years) have
been transplanted in CP using one of the two current
regimens,
14
SURVIVAL
0.2
o+-~--~~-.--,--.--,-.--,--,-~
o 1 2 3 4 5 6 7 8 9 10 11
YEAR
Figure 8. The probabilities of survival from first transplant and
from post-transplant relapse for 49 patients who developed
persistent cytogenetic relapse after transplantation in Seattle
from HLA-identical related donors for CML in CPo Seven of these
patients received second transplants, and one of these survives and
is marked with a * on the survival curves.
SURVIVAL 1
0.4
0.2
6 8
Figure 9. The probabilities of survival for 47 patients older than
50 years transplanted in Seattle from HLA-identical siblings for
CML in CP through 1993.
and the survival of these patien"s is presented in Figure 9. Twelve
of these patients died (including four of those over 55 years of
age). Five of the deaths occurred within the first 100 days
post-transplant (two among the patients older than 55 years), and
one death was due to leukemic relapse.
15
There has been one death (on day 472) among seven patients older
than 55 years transplanted less than 1 year after diagnosis. There
is obviously a strong case for BMT in older patients with CML in
CP.
Accelerated phase. The accelerated phase of CML is transitional
between CP and BP, and the category is less well defined than
either of the other phases. Some characteristics used to define
phase, such as the proportion of blasts and promyelocytes in marrow
and peripheral blood, can be evaluated readily, while others, such
as bone pain, fever, and response to chemo therapy, are defined
less objectively. There is no firm agreement on the cytogenetic
characteristics that indicate a worse prognosis for a patient
otherwise in CP [65,66]. The Seattle group has accepted the
presence of any chromosomal abnormalities additional to a single Ph
chromosome as an indication of AP. All the characteristics that are
used to define AP have been demonstrated to be prognostic for the
survival of patients receiving conventional therapy [25,26].
Studies of factors predictive of outome of transplantation have
identified phase as the most influential disease-related variable
and have shown that survival is worse for patients transplanted in
AP than for those transplanted during CP, with increased
probabilities of relapse and of nonrelapse death [15,55,67].
However, it does not follow that the indicators used to categorize
patients as being in AP have influence on the outcome of
transplantation.
The early experience of transplantation in AP was discussed
earlier, and all subsequent reports have demonstrated a worse
outcome than achieved after transplantation in CP [55,68,69]' Both
the relapse rates and the nonrelapse mortality were higher, but it
cannot be determined whether this is a consequence of disease
phase, because the patients transplanted in AP were subjected to
more aggressive cytoreductive regimens. In a recent Seattle
analysis of 58 patients with Ph-positive CML in AP who received
transplants of unmodified marrow from genotypically HLA-identical
siblings [70], the 4-year probabilities of survival and event-free
survival for the entire group of patients were 0.49 and 0.43, and
the 4-year actuarial probability of relapse censoring for other
causes of death was 0.12 (Figure 10), which is not different than
the relapse probability for patients transplanted in CP. The 4-year
probability of survival for patients aged 37 years or less was 0.66
compared with 0.35 for older patients (Figure 11; p = 0.01). The
4-year probability of survival for patients categorized as in AP
because of factors other than cytogenetic abnormalities was 0.34
compared with 0.66 for patients whose only reason for
categorization as AP was the presence of cytogenetic abnormalities
other than a single Ph chromosome in marrow metaphases (Figure 12;
p < 0.001). The 4-year probability of survival for patients
transplanted in AP less than 1 year from diagnosis of CML was 0.61
compared with 0.39 for patients who had delayed transplantation for
more than 1 year (Figure 13; p = 0.03). The 4-year probability of
survival for the
16
8 7 8
Figure ZO. The probabilities of survival, event-free survival, and
relapse of 58 patients transplanted in Seattle for CML in AP.
SURVIVAL 1
0.8
0.4
04---~--~--~--~--~--~--~--~
8 7 8
Figure 11. The influence of patient age on the probability of
survival of 58 patients transplanted in Seattle for CML in
AP.
17
0.6
0.2
o~--~----~--~--~----~--~--~----~--~
YEARS
6 7 8 I
Figure 12. The influence of being categorized as AP solely because
of cytogenetic abnormalities on the survival of 58 patients
transplanted in Seattle for CML in AP.
SURVIVAL 1
0.2
6 7 8
Figure 13. The influence of the interval from diagnosis to
transplant on the survival of 58 patients transplanted in Seattle
for CML in AP.
18
16 patients categorized as AP because of chromosomal abnormalities
and transplanted less than 1 year from diagnosis was 0.74.
In the Cox model with survival as an endpoint, the interval from
diagnosis to transplant, age 35 years or less at the time of
transplant and categorization as AP on the basis of cytogenetic
abnormalities were the only significant variables in the initial
univariate analysis. During the stepwise multivariate analysis, the
interval from diagnosis to transplant ceased to be significant when
the variable representing categorization as AP on the basis of cy
togenetic abnormalities was entered. After completion of the
stepwise multivariate analysis, patient and donor gender and CMV
serology, spleen status at diagnosis and at the time of transplant,
peripheral blood white blood cell (WBC) count at the time of
diagnosis, previous chemotherapy, the interval from diagnosis to
transplant, regimen, and acute or chronic GVHD were not
significantly independently associated with survival or nonrelapse
mortality. Age 37 years or less at the time of transplant and
classification as in AP solely on the basis of cytogenetic
abnormalities emerged as factors independently significantly
associated with improved survival and reduced nonrelapse mortality.
The probabilities, relative risks, and confidence levels for the
instantaneous relative risks are described in Table 2. Sample size
considerations undermined confident assessment of the relative
influence of different chromosomal abnormalities.
The low probability of relapse observed in these patients together
with the fact that relapse can now be treated with IFN [71,72], or
with infusions of donor lymphocytes [73J (see later), suggests that
more aggressive preparative regimens should not be used in view of
the risk of increasing the incidence of nonrelapse mortality. It is
possible that the nonrelapse mortality associated with less
aggressive regimens would be significantly lower than that of the
regimens commonly used for patients transplanted in AP. Currently
in Seattle, patients in AP are transplanted with the same regimens
used for patients in CP. This should permit an assessment of the
association with survival after transplantation of chromosomal
abnormalities and of the phase categorization.
The finding that age is a significant determinant of outcome in
patients transplanted for the treatment of CML in AP is in
accordance with experience in all allogeneic marrow transplant
situations. The median age at
Table 2. Cox multivariate analyses of factors influencing mor
tality of 58 patients transplanted in AP
Relative Confidence Variable p value risk limits
Age 37 years or less 0.02 0.32 0.12-0.85 Classified as AP because
of 0.003 0.30 0.13-0.67
cytogenetics only
19
time of transplant was 37 years, and the 4-year probability of
survival for patients over 37 years was 0.35. Decisions affecting
the timing of trans plantation for patients with newly diagnosed
CML will take into this account (see later).
Blast phase. Patients transplanted after transformation to blast
phase have had a very poor post-transplant survival in all
published studies [55,74,75]. This is a result of a very high
post-transplant relapse rate and also of a high nonrelapse
mortality. The Seattle team had transplanted 100 patients in BP
before 1993 after a variety of conditioning regimens. The
event-free survival probabilities at 100 days, 1 year, and 3 years
were 0.43, 0.18, and 0.11, and the probability of relapse at 2
years was 0.73. Despite this disappointing result, it is important
to recognize that there are 10 survivors in continuous remission
between 2 and 16 years, with 8 patients more than 8 years after
transplant. Clearly, a small but significant proportion of patients
transplanted in BP can be cured, and because combination
chemotherapy is ineffective in producing prolonged survival in such
patients they should be offered trans plantation if they have
suitable donors.
A small proportion of patients with CML in BP achieve hematologic
remission when treated with combination chemotherapy [76,77]. These
remissions are usually of very short duration but sometimes endure
long enough to permit BMT while in remission. Figure 14 presents
the survival and relapse probabilities for 28 patients in remission
after BP transplanted through 1992 in Seattle. It is surprising
that these patients have a survival
PROBABILITY 1
0 0 2 4 8 8 10 12 14 18
YEARS
Figure 14. The probabilities of survival and relapse for 28
patients transplanted in Seattle through 1992 while in remission
from BP of CML.
20
probaility of 0.41 and a relapse probability of only 0.18. It is
important to note that only 1 of the 9 patients older than 35 is
alive and relapse free after 7 have died and 2 have relapsed. Ten
of the 19 patients 35 years of age or younger are relapse-free
survivors after 9 deaths and 1 relapse. Eleven of the 16 deaths
occurred within 100 days of transplant and were due to causes other
than relapse. It is likely that the intensive chemotherapy received
in the course of remission induction had rendered these patients
particularly susceptible to transplant-related complications in a
situation analogous to that of patients transplanted in first
remission of acute myeloid leukemia, and this susceptibility may be
more severe in older patients. We do not know how many patients
were treated with combination chemotherapy in BP in order to obtain
this group of patients, but clearly they represent a highly
selected population and it is not possible to construct treatment
strategies based on these data.
Effect of splenomegaly. Most patients have splenomegaly at
diagnosis and many at transplant. Sometimes the spleen size at the
time of transplant is so large that it could interfere with the
supportive care of the patient. Moreover, in such circumstances the
spleen may represent a large tumor mass that might influence the
probability of post-transplant relapse. There are reports
indicating that splenomegaly is associated with delayed engraft
ment in patients undergoing BMT for CML [78], and that splenectomy
resulted in earlier granulocyte and platelet engraftment and
reduced platelet transfusion requirements [79], but there was no
effect on survival or the probability of relapse [80]. Reports from
the European Group for Bone Marrow Transplantation showed that
neither routine splenectomy nor routine splenic irradiation
improved survival or relapse probabilities, and both were
associated with some adverse effects, including an increase in
acute and chronic GVHD and infection [80,81]. None of these studies
addressed the issue of benefit for patients with massive
splenomegaly, and this must be considered on a case-by-case basis.
Certainly some patients present for BMT with a degree of
splenomegaly that will jeopardize a successful transplant, and the
desirability of splenic irradiation or splenec tomy will depend on
individual factors such as patient age and size and the urgency of
the need to transplant.
Mismatched related donors
Most patients do not have genotypically HLA-identical siblings, but
a very small proportion of patients will have parents or children
with whom they are genotypically HLA identical for one haplotype
and have the same HLA antigens on the other. The results of
transplantation from such donors are similar to those obtained by
using HLA-identical siblings as donors. A slightly more common
situation is when the patient has a sibling or other relative who
is genotypically HLA identical for one haplotype and has some
21
similarity less than phenotypic HLA identity for the other. It has
been shown that the success of BMT in this setting is related to
the degree of mismatching for the non identical haplotype.
Mismatching for one antigen is associated with an small increase in
the probability of rejecting the graft and a moderate increase in
the incidence of grade 2 or worse acute GVHD. However, the
prospects of success in this situation are still good enough to
contemplate transplants while the patient is in CP. Thus through
1993 the Seattle team has performed 66 transplants for patients in
CP from donors with whom they are genotypically HLA identical for
one haplotype and one antigen mismatched on the other. The results
are presented in Figure 15 and show a very low probability of
post-transplant relapse (0.03) and an event free survival of 0.55
at 4 years with 18 survivors from 4 years to more than 10 years
after transplant. Thirty-five of these patients were transplanted
less than 1 year after diagnosis, and the probability of survival
at day 230 for these patients is 0.63 with 22 survivors on a
plateau to 11 years and no relapses. During the same period 25
transplants were performed in Seattle from one antigen - mismatched
family members for patients in AP with 4 year probabilities of
survival and relapse of 0.39 and 0.34 respectively. Because the
prospects for prolonged survival without marrow transplant for
patients in AP are very poor, transplants from family members
genotypically identical for one HLA haplotype and mismatched for
two antigens of the other were undertaken in 18 patients, and 2 of
these patients, aged 7 and 43 years at the time of transplant,
survive disease-free at 2.8 and 3.0 years after
PROBABILITY 1
YEARS
8 9 10 11
Figure 15. The probabilities of survival and relapse for 66
patients transplanted for CML in CP from donors with whom they are
genotypically HLA identical for one haplotype and one antigen
mismatched on the other.
22
transplant. For patients transplanted in BP, there are 2 survivors
(at 1 and 12 years) from 17 transplants from one antigen-mismatched
related donors, and none of 28 patients transplanted from two
antigen - mismatched donors survive.
In summary, most patients in CP and a small proportion of patients
with advanced disease benefitted from one antigen-mismatched
transplants but only 2 of 46 patients (one aged 7 years) with
advanced disease survived after transplant from two
antigen-mismatched donors. Partly matched related donors are rare,
and consequently there has been great interest in extending
allogeneic BMT for CML by using unrelated donors [82-86].
Unrelated donors
The topic of BMT from unrelated donors is discussed in detail by
Anasetti et al. in Chapter 6. Mackinnon et al. [87] described a
series of 17 patients with CML in CP transplanted from unrelated
donors selected by the Anthony Nolan Centre in England. The marrow
was T-cell depleted to reduce the incidence of GVHD, but five
patients died during the first 100 days and nine died within the
first year from causes other other than relapse (although two of
them had relapsed). A total of five patients relapsed. McGlave et
al. [84] reported a series of 196 patients transplanted in 21
centers (115 during CP) with marrow from unrelated donors furnished
by the NMDP. The 2-year probability of disease-free survival for
patients transplanted in CP less than 1 year from diagnosis was
0.45.
Through 1993 the Seattle team had transplanted more than 300
patients with CML from unrelated donors, and this experience,
together with the overall experience of the use of unrelated
donors, is described by Anasetti in Chapter 6. Through September
1991 the Seattle team had transplanted 105 patients in CP, AP, or
BP from HLA-matched unrelated donors. The survival probabilities
for these patients are presented in Figure 16, and the 4-year
probabilities of survival for 67 patients transplanted in CP and 29
patients transplanted in AP were 0.51 and 0.37, respectively.
Twenty-one patients were transplanted in CP less than 1 year from
diagnosis, and the 4 year probability of survival for these
patients was 0.57. Nine patients were transplanted in BP, and they
all died within 21f2 years of transplant. For patients transplanted
in CP or AP, CMV IP was the most frequent cause of death,
accounting for 6 of 33 deaths in patients in CP and 5 of 19 deaths
for patients transplanted in AP. One patient each died after
relapse in patients transplanted in CP or AP, whereas 5 of 8 deaths
in BP patients occurred after relapse.
Drobyski et al. [86] have reported on the use of T-cell-depleted
marrow from matched and mismatched unrelated donors. Two of 28
recipients of mismatched marrow rejected their grafts, whereas all
20 recipients of matched marrow achieved engraftment. The incidence
of acute GVHD was relatively low (grade II or worse 39%), and there
were four relapses in
23
0.4
0.2
8 8
Figure 16. The probabilities of survival for 105 patients
transplanted in Seattle for CML from matched unrelated
donors.
patients with advanced disease at the time of transplantation. The
probability of survival at 2 years was 0.52.
The problem of relapse
Definition of relapse
Special problems and opportunities are created by the very great
sensitivity of techniques currently available for detecting
molecular and cytogenetic signs of persistent or recurring CML.
Four different types of relapse can be recognized. Clinical relapse
is the reappearance of clinical signs or symptoms of the original
disease. This has usually been accompanied by hematologic relapse,
which is the reappearance of characteristic changes in hematologic
values, although cases have been reported in which relapse was
limited to the development of a chloroma without other signs of
relapse. Transient hematologic or clinical relapse has not been
reported and, once developed, such relapses tend to produce
progressive disease, although the rate of disease progression may
be very slow [63]. The presence in marrow or peripheral blood of
metaphases containing Ph chromosomes is referred to as cytogenetic
relapse, and the sensitivity of this method of relapse detection
depends on the number of metaphases examined. Subsequent
examinations may fail to detect Ph chromosomes, even if no
therapeutic intervention has been made, and this is known as
transient relapse. Transient relapse may be
24
the result of sampling probabilities or of a real decline in the
tumor burden consequent on biologic phenomena. Very rare reports
have described cytogenetic relapse without molecular evidence of
the BCR-ABL rearrange ment, but the reasons for this are unknown
and cytogenetic relapse is usually accompanied by molecular
relapse.
The most sensitive technique for detecting the presence of the
BCR-ABL rearrangement uses PCR. This technique permits the
detection of one CML cell in a population of 106 cells [88] and
frequently reveals the presence of BCR-ABL transcripts in marrow
transplant patients with no other evidence of relapse. The
technique is very susceptible to technical error, but agree ment
has emerged from many investigators that post-transplant PCR posi
tivity is not uncommon, particularly soon after transplant [89-91],
and especially in recipients of T-cell-depleted marrow [92]. Like
cytogenetic relapse, PCR relapse is frequently transient. The
logical expectation is that the incidence of cytogenetic,
hematologic, and clinical relapse will be higher in patients who
already show PCR positivity, but this expectation has not yet been
confirmed in clinical studies and lacks dimension. Moreover, some
studies were unable to confirm a strong correlation between PCR
positivity and relapse [93-95]. A modified PCR technique has been
reported to be quantitative in nature, and it has been suggested
that an increase in the number of detectable transcripts presages
imminent cytogenetic relapse and can be used to identify patients
who might benefit from pre-emptive therapy [62]. Further study of
pre cytogenetic relapse should permit the early institu tion of
measures to forestall the progression to cytogenetic and
hematologic relapse.
Treatment of relapse
Frequently the pace of disease progression is very slow after
post-transplant relapse, particularly for patients who have only
early cytogenetic relapse [63,96]. Some patients will have stable
disease over many years with no increase in the proportion of
Ph-positive metaphases, and they may not require an early second
attempt at transplantation. Methods of treatment other than second
transplant are available, and it has been shown that the longer the
interval between the first and second transplant, the greater the
chance of a successful outcome.
Two methods that have been used widely for the treatment of post
transplant relapse are treatment with IFN with or without the
infusion of lymphocytes from the marrow donor. Treatment with IFN
alone is effective in producing both clinical and cytogenetic
remission in patients who have relapsed after transplantation, and
complete remissions are more frequent when treatment is initiated
at an early stage of relapse [63,71,97]. The same considerations
apply to this form of treatment as to the use of IFN in the
management of patients with CML who have not had marrow
transplants, namely, success requires high doses of IFN, the
treatment is toxic and
25
expensive, and it is not known whether successful treatment can be
discontinued without relapse. Kolb et al. produced hematologic and
cyto genetic remission by treatment with IFN accompanied by the
infusion of donor buffy coat cells in three patients who relapsed
after transplantation [73]. This form of treatment has a high
success rate for patients in early relapse, with most patients
achieving hematologic remission, many achieving cytogenetic
remission, and some becoming negative to PCR testing for BCR-ABL
[73,98,99]. Most patients have reactivation of acute GVHD, and some
patients have developed fatal GVHD. Another serious complication of
this treatment is the development of marrow aplasia, presumably in
patients whose hematopoiesis became entirely of host origin when
they lost the myeloid component of their grafts in the same process
that produced relapse. Lymphocyte transfusions without IFN have
also been demonstrated to be effective in producing remission
[100].
As mentioned earlier, successful second transplants have been
reported. For second transplants, chemotherapy only is used when
the first transplant was with a TBI-containing regimen, and
TBI-containing regimens are used when the first transplant regimen
consisted of chemotherapy only. In cases where immune tolerance of
host tissues persists, the regimen is not con strained by the need
to overcome the possiblity of graft rejection. In Seattle through
1990, 12 patients who relapsed after transplants from identical
twins received second transplants. Two of these patients relapsed a
second time and died, and four remain alive and disease free
between 6 and 15 years after the first and 4 and 14 years after the
second transplant.
Cullis et al. [101] reported 16 patients who received second
transplants from the same donors for relapse after transplantation
with T -depleted marrow from HLA-identical siblings. Eight patients
were alive disease free a median of 424 days after the second
transplant (range 158-1789 days). Five of these patients had been
conditioned for second transplant with a regimen of BU only. In
Seattle 30 patients who had received transplants from HLA
identical siblings received second transplants through 1989. Twenty
of these patients relapsed after the second transplant and died,
and four patients remain alive and disease-free between 6 and 15
years after the first, and 4 and 14 years after the second
transplant. Fourteen of these patients were in CP when they
received the first transplant, and one of these died from
rejection, five died after a second relapse, and three each died
from VOD or infection. Two of these patients survive disease-free 6
and 9 years after the second, and 7 and 11 years after the first
transplants. Clearly, second transplants for patients transplanted
for CML who relapse are possible but rarely successful.
The timing of transplantation
Since 1983 the Seattle transplant team has recommended
transplantation as soon as possible after diagnosis.
Transplantation before the disease
26
accelerates is beneficial because the results during the chronic
phase are much better than in accelerated or blast phase and
because in the Seattle experience delay has an adverse effect on
survival, even when transplants are performed in chronic phase
[15,102].
The Seattle team does not report its experience to the
International Bone Marrow Transplant Registry (IBMTR). An analysis
of outcome for patients with CML in CP reported to the IBMTR [14]
showed an improvement in survival and a lessened incidence of
relapse when patients were transplanted within 1 year of diagnosis
compared with later. The negative effect of delay upon survival in
this analysis was not a consequence of the increased risk of
relapse because post-transplant relapse did not have an early
effect on survival (i.e., patients who relapse may survive for a
long time after relapse). Instead, the poorer survival seen with
increased delay was solely the result of an increase in mortality
from causes other than relapse. This suggests that the advantage of
early transplant likely will become even greater as the impact of
the increased relapse rate upon survival becomes apparent. One can
devise explanations for an increased risk of post transplant
relapse in patients who have had CML longer before being
transplanted, but we do not know why patients who remain in chronic
phase without obvious physical, hematologic, or cytogenetic change
have an increasing risk of dying of the complications of BMT. No
single cause of death accounts for the difference. Since the
diagnosis of CML in CP is frequently made fortuitously, it seems
likely that the deterioration in survival prospects is associated
with making the diagnosis rather than the inception of disease, and
this would implicate medical attention as a possible cause. BU was
the standard treatment for CML until a few years ago, when most
physicians began to use HU instead. Consequently, until recently,
patients with a long interval between diagnosis and transplantation
were much more likely to have been treated with BU than with HU.
This has made it difficult to examine separately the effect of
delayed transplantation and the effect of pretreatment with BU.
However, the IBMTR study [14] shows that palliative treatment with
BU has an adverse effect on the outcome of subsequent BMT and that
delay was detrimental, even in patients who did not receive BU. It
may well be that treatment with HU is also detrimental, which could
only be recognized if there were a com parative series of patients
who had received no treatment before transplant.
The hazards associated with delay in BMT for patients with newly
diagnosed CML can be evaluated only in a setting where a cohort of
patients receiving transplants is followed from the time of
diagnosis. It would be extremely difficult to design a protocol
that provided a broad spectrum of delay for patients with donors.
It is reasonable to ask whether the effect of delay upon survival
after transplantation simply reflects a relationship between the
timing of transplantation and survival from the date of diagnosis.
Figure 17 presents the Kaplan-Meier survival curves for all
patients transplanted in chronic phase from matched sibling donors
after a
27
SURVIVAL
L.
YEARS FROM DIAGNOSIS
Figure 17. The probabilities of survival for patients transplanted
less than 2 years after diagnosis for CML in CP after CY-TBI. A
describes survival from the date of transplant, and B describes
survival from the date of diagnosis.
CY + 12.0Gy TBl. Prophylaxis against acute GVHD was provided by MTX
+ CSP. The cases are stratified on the basis of being transplanted
less than or more than 2 years after diagnosis. Figure 17 A
describes survival from the date of transplant and Figure 16B
describes survival from the date of diagnosis. The log-rank p
values are 0.0001 for Figure 16A and 0.05 for Figure 17B.
A population of patients with newly diagnosed CML may not have a
uniform susceptibility to the influence of delay on post-transplant
survival. We have transplanted eight patients in CP 8 or more years
after diagnosis and, with a follow-up of 2-10 years, three of these
patients have died (on days 90, 147, and 175), and there are five
disease-free survivors between 2 and 10 years after transplant.
Thus, the effect of delay may be different in different groups of
patients.
Of course, patients who delay transplantation for 2 years will be
at hazard for transformation into AP or BP during the period of
delay, and we have no way of estimating the attrition that this
will cause, or the modification of this hazard by the benefits of
transplantation during AP or BP. In this respect, age influences
the relative risks associated with delay. Figure 18 shows the
probabilities of survival for patients transplanted less than 1
year from diagnosis and younger than 35 or older than 35 for
patients while in CP (Figure 18A) or in AP (Figure 18B). The
adverse effect of age is greater for transplants during AP than
during CP, so that older patients gain more by transplantation in
CP than younger patients and are therefore placed in greater
jeopardy by delaying transplantation.
The demonstration that treatment with IFN can produce complete
cytogenetic (and even molecular) remission in a small proportion of
patients has provided an additional rationale for delay, adding to
the difficulties in counseling patients [41]. There is a report
that the outcome of transplanta tion was not adversely affected by
prior treatment with IFN in a study that
28
SURVIVAL SURVIVAL
... ... ACCELERATED PHASE (N-12)
YEARS
Figure 18. The probabilities of survival for patients transplanted
less than 1 year after diagnosis for CML in CP or AP. A describes
the survival of patients 35 years of age or less. B describes the
survival of patients older than 35 years.
involved only 15 patients transplanted within 1 year of diagnosis.
The authors concluded that the sample size was too small to derive
any definite conclusions on whether delaying transplantation for a
trial of IFN has any effect on transplant outcome. Controlled
trials and further study involving large numbers of patients will
be needed to examine this question, but given the results of early
transplantation from HLA-identical siblings, it is difficult to
design a randomized study of the problem. Clearly patients 55 years
of age or less, and probably to the age of 60, with HLA-identical
siblings or one antigen-mismatched related donors, should be
transplanted as soon as possible after diagnosis. For patients
between 60 and 65 years, no useful data are yet available for
either transplantation or IFN therapy, but if they have donors they
should receive transplants at the first sign of disease
progression. For other patients, treatment with IFN should be
started and a search should be initiated for unrelated donors. If a
matched unrelated donor is found and the patient has not achieved a
complete cytogenetic response with IFN therapy, it seems reasonable
that BMT should be performed as soon as possible. Patients who have
achieved complete cytogenetic responses to IFN probably should not
be subjected to unrelated donor transplants until they have disease
progression.
Autologous transplantation
The earliest studies of autologous bone BMT for the treatment of
CML did not attempt cure but were aimed at the restoration of CP in
patients whose disease had evolved to AP or BP [103,104]. Marrow
from patients in CP was harvested and cryopreserved without any
attempt at purging. When patients entered BP, they were conditioned
with CY and TBI, and the stored
29
marrow was reinfused. In these studies, some patients did not have
marrow repopulation, and in others there was a high incidence of
infection associated with poor lymphoid engraftment, generating the
suspicion that stored marrow from patients with CML might be of
poor quality. Moreover, when engraftment was achieved, the duration
of a second CP was disappointingly short. The number of stem cells
in the peripheral blood (PBSC) is much increased over normal in
patients with CML, and Goldman and his colleagues at the
Hammersmith Hospital in London conducted a program of research
using stem cells collected from the peripheral blood of patients in
CP [105]. The results of using this approach in 51 patients were
summarized in 1984 [106] and suggested that the numbers of
collectable PBSC permitted rapid marrow recovery. Forty-eight
patients transplanted in transformation were restored to a second
CP, but recrudescence of BP occurred between 8 and 40 weeks after
transplantation. Of interest in this report, three patients had a
proportion of Ph-negative marrow metaphases after the transplant,
but these patients also relapsed into BP. Other studies reported
similar findings [107,108]. The lack of substantial benefit from
autologous trans plantation for advanced CML may have been in part
a consequence of failure to eliminate the transformed disease from
the patient, and attention has turned to autologous BMT during CP,
at which time it is easier to achieve this. Obviously this will not
be a rewarding endeavor unless the stem cells can be treated in
some way to eliminate or reduce the leukemic component.
All attempts at achieving this rely on the assumption that patients
in CP have a population of Ph-negative stem cells, even if they
cannot be detected readily by routine cytogenetic examinations. The
cytogenetic responses of patients treated with IFN suggest that
this is so for many patients in CP, and the observation that
patients treated with IFN soon after diagnosis are more likely to
develop some Ph-negative hematopoiesis than those treated later
suggests that the proportion of Ph-negative hematopoietic
precursors decreases with time from diagnosis. Laboratory studies
tend to confirm this [109], although the relationship between the
absolute number of Ph-negative (or BCR-ABL negative) cells and
phase has not been fully explored. Most purging protocols either
select patients who already have 'useful' pro portions of
Ph-negative cells or attempt to increase the proportion of Ph
negative stem cells in the patient's marrow or blood, collect the
stem cells, and further treat them to reduce the proportion of
Ph-positive stem cells. Some protocols add early post-transplant
treatment to increase the competitive capability of Ph-negative
hematopoiesis.
Barnett et aI. [110,111] reported studies in which patients were
selected on the basis of containing Ph-negative long-term culture
initiating cells in the marrow. The marrow was cultured for 10 days
to reduce the proportion of Ph-positive cells, and the patients
were then treated with intensive therapy and had the marrow
cultures reinfused. Of 87 patients evaluated, 36 were considered
eligible and 22 were transplanted. Thirteen patients achieved
30
complete hematologic and cytogenetic remISSIon, but only one of
these remained in remission at last report [112]. Nine of the
relapsed patients were treated with IFN, and four returned to
complete remission. A major disadvantage of this approach has been
the small proportion of patients eligible for study because they
lacked a sufficient proportion of Ph-negative cells in the marrow,
and intensive chemotherapy has been used in an attempt at
increasing this proportion. Some success has been achieved in this
endeavor, but it is not clear whether intensive chemotherapy
increases the absolute number of Ph-negative stem cells in the
marrow by removing inhibitory effects from the leukemic stem cells
or whether such treatment simply increases the proportion of
Ph-negative stem cells by selective destruction of Ph-positive stem
cells.
Complete cytogenetic conversion to Ph negativity in a small
proportion of patients treated with IFN encouraged the hope that
this would provide an opportunity to collect normal hematopoietic
stem cells from these patients. Unfortunately patients undergoing
treatment with interferon frequently have severe suppression of
granulo-erythropoietic precursors, which persists for long periods
after discontinuation of the IFN therapy [113], and this has
frustrated the use of this approach. However, Simonssen et al.
[114] per formed autologous transplants in 18 patients using
marrow harvested after prolonged treatment with IFN and HU followed
by intensive chemotherapy. There was one early death from
interstitial pneumonia and seven patients have relapsed. Nine
patients are Ph negative between 1 and 32 months after transplant.
Carella et al. [115,116] demonstrated that stem cells collected by
leukapheresis after simulation with granulocyte-colony-stimulating
factor (G-CSF) when recovering from treatment with a combination of
idarubicin, ara-C, and etoposide were completely Ph negative in 9
of 15 patients treated in CP and in 3 of 10 treated in AP. Nine
patients were transplanted with these collections at a time when
90-100% of metaphases from the patient's marrow were Ph positive,
and two patients died after failing to achieve engraftment. At the
time of the report, 5 of the 7 survivors were completely Ph
negative between 2 and 18 months after transplant. This approach
has stimulated much interest and is currently being studied by many
groups. One problem has been t