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