High-Dose Consolidation Chemotherapy V Bone Marrow Transplantation as Postremission Therapy in Adult Acute Myeloid Leukemia
As many as 60% to 80% of adults with acute myeloid leukemia (AML) achieve complete remission with induction chemotherapy.1-4 However, most patients ultimately relapse, primarily because of the proliferation of leukemia cells surviving the induction regimen. Prognostic factors related to outcome include age, leukocyte count at presentation, history of preceding hematologic disturbance and karyotype.5-9
Postremission strategies attempting to eradicate subclinical disease in patients with a high probability of recurrence include consolidation and maintenance chemotherapy, allogeneic bone marrow transplantation (BMT) and autologous BMT with or without myeloablative conditioning.10-16. Conventional postremission chemotherapy is expected to yield a median remission duration of 12 to 18 months and a 20% to 30% likelihood of leukemia-free survival 5 years after complete remission is achieved.17,18 High-dose cytarabine-based consolidation chemotherapy has produced the most encouraging results among patients receiving non-myeloablative postremission chemotherapy.19-23
This article: (1) describes the long-term outcome for adults with AML treated with high-dose, cytarabine-based consolidation chemotherapy; (2) compares the efficacy of this form of postremission treatment to that of allogeneic BMT; and (3) briefly discusses other potential strategies in postremission treatment, particularly autologous BMT.
LONG-TERM ANALYSIS OF HIGH-DOSE CONSOLIDATION CHEMOTHERAPY
In a study conducted over 10 years at UCLA, 19 leukemia-free survival was investigated among patients receiving high-dose cytarabine-anthracycline consolidation chemotherapy. A total of 227 patients with new diagnoses of AML were enrolled in two sequential studies dating from 1982 to 1991. Median follow-up was 57.6 months (range: 8.4 to >109 months) for patients who achieved complete remission and who were eligible for consolidation chemotherapy.
The standard procedure for histologic diagnosis of AML was employed.24 De novo AML was considered to be present if no evidence of a hematologic abnormality was documented more than 2 months before the leukemia diagnosis. Preleukemia, therefore, was defined as ineffective hematopoiesis identified more than 2 months before AML diagnosis and characterized by unexplained peripheral blood cytopenia or abnormal bone marrow. The FAB system was used for morphologic classification, and cytogenetic analysis was classified as abnormal, but favorable, if there were abnormalities of 16q, t(8;21) or t(15;17). Abnormal and unfavorable karyotype consisted of trisomy 8, abnormalities of chromosomes 5, 7 or 11 or multiple or complex translocations.18
The 227 patients enrolled were treated with induction chemotherapy containing daunorubicin and one of several schedules of cytarabine.5,19 Of these patients, 151 achieved complete remission. Twenty-eight patients " 45 years of age had histocompatible sibling donors and were assigned to allogeneic BMT. The remaining 123 were eligible to receive high-dose, cytarabine-based consolidation, which was administered in two or three planned courses, beginning a median of 1 month postremission.
A total of 108 patients received the first cycle of high-dose consolidation with cytarabine 2 to 3 g/m" administered as a 2-hour IV infusion every 12 hours for 4 days plus daunorubicin (45 mg/m"/d IV bolus ◊ 3 days) or mitoxantrone (2 mg/m"/d IV bolus ◊ 3 days). The remaining 15 patients did not receive the first cycle of consolidation because of infection, excessive toxicity from induction, patient refusal or early relapse.
Of the 108 patients, 79 went on to receive a second cycle of consolidation, consisting of either cytarabine (200 mg/m"/d continuous IV ◊ 7 days) and daunorubicin (45 mg/m"/d IV ◊ 3 days) or etoposide (200 mg/m"/d IV over 1 hour ◊ 5 days) and mitoxantrone (10 mg/m"/d IV bolus ◊ 3 days). No further treatment was administered to the remaining 29 patients because of infection, excessive toxicity, patient refusal or early relapse, all of which exemplify one of the most limiting features of the consolidation regimen, ie, the patientís difficulty proceeding through the planned course of therapy.
After 1987, a total of 23 patients received a third cycle of consolidation chemotherapy with high-dose cytarabine (2 g/m" IV every 12 hours ◊ 4 days) and daunorubicin (45 mg/m"/d IV ◊ 3 days). Nine elderly patients entering the study after 1988 received consolidation chemotherapy at a reduced regimen because of concerns about the increased risk of dose-related toxicity in older persons. After completion of consolidation chemotherapy, no maintenance therapy or other antileukemia treatment was administered until relapse.
Of the 108 patients who received the first consolidation course, 5 developed neurotoxicity and 9 died from treatment-related complications. Treatment-related deaths also occurred in 2 of the 79 patients undergoing a second cycle and 2 of the 23 given a third course. One patient died from cardiomyopathy 2 years after completing consolidation treatment. No difference in the incidence of consolidation-related deaths was observed among the various age groups, although some of the elderly patients had received a dose-reduced regimen.
When our original study was published in December of 1992, 40 (33%) of the 123 patients eligible for consolidation were alive, with 28 having a continued remission duration of 12 to > 110 months. Median remission duration for all eligible patients was 12.8 months (range: 1.3 to > 110 months) and the actuarial 5-year, leukemia-free survival was 26 ± 8% (Fig. 1). Median survival from remission was 24 months and actuarial survival 5 years from remission was 33 ± 10%. Actuarial risk of relapse 5 years postremission was 69 ± 10%.
Age was a significant predictor of both disease-free and overall survival from remission. Among patients " 45 years of age, actuarial 5-year, disease-free survival was 35 ± 13%, which was statistically significantly better than survival for those > 45 years of age (P = .03; Fig. 2). Actuarial survival for the younger patients was much better because they were eligible for salvage therapy with BMT. Disease-free survival was also statistically significantly better for women than for men (P = .0054; Fig. 3).
Analysis of other pretreatment characteristics showed that cytogenetic abnormalities, specific AML subtypes, LDH, marrow cellularity and blood counts were not significant predictors of outcome. History of antecedent hematologic disturbance also was not related to leukemia-free survival, perhaps because of the small number of patients involved.
We concluded from the 1992 study that intensive consolidation was efficacious as a form of postremission chemotherapy for patients with AML, the effect being most pronounced among younger persons. In fact, when data were compared to earlier findings on conventional-dose consolidation chemotherapy, a statistically significant improvement in leukemia-free survival was evident among patients receiving high-dose, cytarabine-based consolidation, although only in the younger group. We also concluded that regimen-related toxicity and infection had a significant adverse effect on the patient's ability to complete the prescribed treatment course, suggesting that further dose intensification is most likely to improve results in younger patients.
CONSOLIDATION CHEMOTHERAPY VERSUS ALLOGENEIC BMT
"An allogeneic bone marrow transplant from an HLA-identical sibling produces a low rate of leukemia relapse and high overall survival.24- 26 This improved antileukemia effect, however, is accompanied by a greater risk of treatment-related mortality, with approximately 30% of patients dying within the first 6 months after transplantation from graft-versus-host disease, veno-occlusive disease, interstitial pneumonia or other complications. 26 Furthermore, adverse predictors of leukemia-free survival observed among patients receiving conventional chemotherapy also seem to apply to those undergoing BMT.
In a 1992 study, we assessed 28 patients 16 to 45 years of age, who were assigned to undergo a closely HLA-matched BMT for AML in first remission versus an age-matched group of 54 patients treated in the same period with high-dose cytarabine-based consolidation chemotherapy.21 Of the 28 patients in the transplantation group, 25 had a genotypic-matched, allogeneic bone marrow transplant; the median interval from remission to transplantation was 44 days (range: 10 to 154 days). One patient each received a syngeneic marrow graft and a marrow graft matched for 5 of 6 HLA loci. Donor refusal for the remaining patient led to treatment with chemotherapy; nonetheless, the patient's outcome was included in the transplantation group, based on intent-to-treat analysis.
In 8 of the 28 patients, donor bone marrow was treated with Leu-2b monoclonal antibody to selectively deplete CD8-positive cells.27 Two patients underwent pan-T-cell depletion with anti-CD2 monoclonal antibody. Patients assigned to T-cell depletion and those receiving non-T-cell-depleted grafts were similar with respect to pretreatment characteristics. Preparative conditioning for BMT consisted of 1125 cGy total body irradiation given in five fractions over 2.5 days, cyclophosphamide 60 mg/kg/d for 2 days and cytarabine 500 mg/m" every 12 hours for four doses (9 patients) or cytarabine 2 g/m" every 12 hours for four doses (1 patient). Other preparative conditioning regimens containing total body irradiation were used; no patient received pure chemotherapy conditioning. Prophylaxis for graft-versus-host disease consisted of cyclosporine or methotrexate.
Fourteen patients assigned to BMT were alive 153 to 2197 days postremission. Actuarial disease-free survival at 5 years was 48 ± 21% (median duration of disease-free survival: 363 days). Actuarial survival was 45% ± 24%. Five patients relapsed: two who had received a conventional transplant, two who had received a T-cell-depleted transplant and one who was given chemotherapy despite assignment to the transplantation group. Actuarial risk of relapse for the transplantation group was 32% ± 26% at 5 years &8212; statistically significantly lower than the risk observed among the group receiving consolidation chemotherapy (60 ±14%). Nonetheless, a high incidence of treatment-related complications, including graft-versus-host disease and interstitial pneumonia, led to a statistically significantly higher risk of treatment-related mortality and lower survival in patients who underwent BMT (P =.07;Fig. 4).
Thus, allogeneic BMT for AML in first remission did not offer any advantage with regard to overall survival compared to intensive postremission consolidation chemotherapy. Furthermore, the relatively high 5-year, leukemia-free survival among young patients assigned to consolidation chemotherapy suggests that this less toxic approach is the preferred form of postremission treatment. Whether autologous transplantation provides improved results has not been elucidated as yet. Our results confirm those of most other studies comparing allogeneic transplantation with consolidation chemotherapy for adults with AML in first remission. That is, a trend toward improved disease-free survival in the transplantation group does not translate into improved overall survival.
OTHER POTENTIAL POSTREMISSION STRATEGIES
When pursuing dose intensification, it is reasonable to consider autologous bone marrow or stem-cell transplantation.28-31 Most long-term studies of autologous BMT show a leukemia-free survival of 30% to 50%, although some trials demonstrate an even higher disease-free survival Ė as high as 60%.28,30-32 This may reflect differences in the biologic characteristics of patients studied rather than differences in the type of conditioning regimen or transplant. Leukemia relapse is a major cause of mortality in patients undergoing autologous transplantation. Strategies that may be used to improve postremission autologous transplantation include monoclonal antibody treatment or ex vivo expansion.<33,34 Post-transplant immunologic manipulation with IL-2 or another type of immunostimulant should be considered.
At UCLA, we are currently performing autologous peripheral blood progenitor cell transplantation for patients <= 70 years of age with AML in first remission.35 All patients with new diagnoses of leukemia patients receive an identical induction regimen of continuously infused cytarabine and idarubicin. Those who achieve remission subsequently receive high-dose cytarabine-based consolidation chemotherapy with mitoxantrone and G-CSF. Peripheral blood progenitor cells are collected when leukocyte count reaches 1000/mm≥. After collection of 6 ◊108 total mononuclear cells/kg, patients are admitted for 1125 cGy total body irradiation in five doses and cyclophosphamide 60 mg/kg/d for 2 days, followed by reinfusion of peripheral blood progenitor cells. After reinfusion, no growth factors are given and patients are followed until hematologic recovery.
This postremission regimen has been quite tolerable. Because peripheral blood progenitor cells are used as a form of hematopoietic support, time to hematologic recovery is approximately 14 days, making the regimen tolerable even for elderly patients and those who have been heavily pretreated. So far, there have been no treatment-related deaths and treatment-related morbidity has been confined to stomatitis and gastrointestinal toxicity. Whether this type of myeloablative treatment with peripheral blood progenitor transplantation provides any long-term benefit will not be known until further follow-up analysis has been completed. We are also performing cytogenetic analysis on the progenitor cell product and using fluorescent in situ hybridization to detect minimal residual disease in the concentrated infused stem-cell product.
Once we have defined a fairly nontoxic approach to deliver intensive chemoradiotherapy and a fairly purified form of autologous progenitor cell support, transplantation would be studied best using a random-assignment design with currently available consolidation chemotherapy as the comparative treatment.36 Because patient and disease-related factors have a significant impact on treatment-related complications, as well as on leukemia-free and overall survival, only a random-assignment study could identify any incremental benefit of autologous transplantation. Ideally, the study would include a heterogeneous group with an advanced median age, thus being more representative of adult patients with AML.
Gary Schiller, MD is Assistant Professor of Medicine, Division of Hematology/Oncology, Dept. of Medicine, UCLA School of Medicine, Los Angeles, California.