The Current Status of Treatment for Acute Myeloid Leukemia
This article reviews current data on acute myeloid leukemia (AML) therapy. Several treatment phases, including induction, postremission, second-remission and supportive therapy, are considered.
INDUCTION THERAPY FOR ACUTE MYELOID LEUKEMIA
In 1990, the Australian Leukemia Study Group conducted one of the most interesting recent investigations of induction therapy for AML.1 The study, chaired by James Bishop, MD, assessed a standard regimen of cytarabine and daunorubicin and the same regimen combined with etoposide. Complete response rates were similar in the two treatment groups, but the triple-drug regimen yielded statistically significantly better disease-free and overall survival among patients <55 years of age. Statistically significant differences in outcome for this age group have persisted (Fig. 1); however, no significant outcome differences were observed among older patients.
The Bishop study is important for several reasons. First, it clearly demonstrates that a manipulation during induction can improve remission duration and survival. Second, it suggests that, as increasingly effective treatments become available, cure may be achieved with remission induction alone, without postremission therapy. Postremission regimens are ineffective in other hematologic malignancies, such as advanced Hodgkin's disease and diffuse large-cell lymphoma, for which substantially efficacious induction therapy has been developed. Third, the study demonstrates the inability to achieve meaningful therapeutic responses in elderly patients, which are observed in younger patients. Potential treatments for elderly patients should still be investigated, but the study of an intervention in this population alone may not reveal even a modest improvement, presumably because of the myriad comorbidities that afflict older patients.
Arlin et al2 compared a standard regimen of cytarabine and daunorubicin with a similar regimen in which daunorubicin was replaced by mitoxantrone. No significant differences in outcome were observed between the two therapeutic approaches, suggesting similar efficacy for daunorubicin and mitoxantrone. To date, this is the only published investigation comparing daunorubicin and mitoxantrone for induction therapy in AML patients; no comparisons of idarubicin and mitoxantrone in this setting have been published. However, the Eastern Cooperative Oncology Group (ECOG) has recently initiated a study (E3993) in which elderly AML patients are randomly assigned to receive induction therapy with cytarabine plus daunorubicin, mitoxantrone or idarubicin.
The combination of cytarabine, mitoxantrone and etoposide has been popular, especially in Europe, as induction therapy for patients with relapsed or refractory AML.3 I have not been as impressed with this combination as have other investigators, and no prospective, comparative trials show this regimen to be superior to other regimens in the same setting. This situation is reminiscent of the acceptance of high-dose cytarabine plus amsacrine based on anecdotal data as superior therapy for patients with acute leukemia4; in that case, a prospective study conducted by the Southwest Oncology Group (SWOG) demonstrated no advantage to the combination therapy compared to high-dose cytarabine alone.5
Random-assignment trials have shown cytarabine plus idarubicin to be superior to cytarabine plus daunorubicin in the treatment of patients with AML. For example, Berman et al,6 at the Memorial Sloan Kettering Cancer Center, conducted a study of AML patients <60 years of age and reported a statistically significantly greater complete response rate, remission duration and survival among those treated with cytarabine plus idarubicin compared with those treated with cytarabine plus daunorubicin. Statistically significant differences in survival continue to be apparent, as demonstrated in Fig. 2. In a multicenter study of similar design,7 the superiority of cytarabine plus idarubicin to cytarabine plus daunorubicin in improved outcome was evident among patients <60 years of age but not among older patients. Statistically significant outcome differences in favor of cytarabine plus idarubicin have been maintained among patients <=50 years of age (Fig. 3).
In an investigation by the Southeast Cancer Study Group (SECSG),8 complete response rate was statistically significantly greater with cytarabine plus idarubicin than with cytarabine plus daunorubicin. The postremission protocol incorporated a test of late intensification with cytarabine plus idarubicin or cytarabine plus daunorubicin, depending on the patient's initial induction regimen. Although a recent update demonstrated no difference between induction regimens with respect to disease-free or overall survival, remission duration was statistically significantly longer among patients who received late intensification than among those who did not. Compared to any other patient group, patients who were given cytarabine plus idarubicin during induction and late intensification phases had statistically significantly longer disease-free and overall survival (Fig. 4).
Finally, the Italian group GIMEMA studied cytarabine plus idarubicin versus cytarabine plus daunorubicin as induction therapy in elderly patients.9 They found no differences in response rate, remission duration or overall survival between the two treatment groups, as might be expected, based on the results obtained for elderly patients by Bishop et al1 and in the multicenter study described previously.7 A statistically significantly greater number of patients, however, achieved a complete response with one course of cytarabine plus idarubicin than with one course of cytarabine plus daunorubicin in the GIMEMA study. This observation was also made in the Berman6 and multicenter trials,7 but not in the SECSG study.8
In conclusion, these investigations demonstrate the superiority of idarubicin over daunorubicin in combination with cytarabine as induction therapy for AML, especially in patients <60 years of age. This clinical observation is supported by laboratory investigations that indicate more favorable phar- macokinetics and less interaction with P-glycoprotein for idarubicin than for daunorubicin.10-13 The presence of P-glycoprotein on the surface of leukemic cells from patients with de novo AML has repeatedly been shown to impair response to cytarabine plus daunorubicin.14-16
INDUCTION THERAPY FOR ACUTE PROMYELOCYTIC LEUKEMIA
Investigators generally agree that all-trans-retinoic acid in acute promyelocytic leukemia (APL) is limited by its clinical pharmacokinetics. In particular, trans-retinoic acid induces its own catabolism, presumably by a cytochrome P-450 mechanism.17 Edward Schwartz, PhD, and I have shown that fluconazole, a P-450 inhibitor, blocks the in-vitro catabolism of all-trans-retinoic acid; we also recently demonstrated this inhibition in a patient with APL. Whether fluconazole inhibition of all-trans-retinoic acid catabolism translates into more effective antileukemic activity, however, remains to be studied. At present, optimal APL treatment appears to be initial therapy with all-trans-retinoic acid, followed by administration of an anthracycline plus cytarabine, as first demonstrated by French investigators.18 In their pilot study, this approach yielded a 96% complete response rate and a substantial fraction of durable remissions.18 Although daunorubicin was used by the French investigators, unpublished data from the Memorial Sloan Kettering Cancer Center strongly suggest that idarubicin is significantly more active than daunorubicin in APL, as it generally is in AML (Fig. 5).
POSTREMISSION THERAPY FOR ACUTE MYELOID LEUKEMIA
Considerable debate continues about the most effective postremission therapy for AML patients. Published studies suggest similar long-term survival among patients undergoing bone marrow transplantation and among those receiving chemotherapy alone; however, none of these trials offers an analysis of outcome according to major prognostic factors. For this reason, the ECOG initiated a multicenter study (EST 3489) of the relative merits of chemotherapy alone and autologous and allogeneic bone marrow transplantation in patients who have been characterized by immunophenotype and karyotype. Their results may provide a basis for selecting optimal AML treatment according to specific patient characteristics.
Broad conclusions can be drawn regarding the comparative efficacy of various chemotherapy regimens during postremission. In my opinion, no treatment has emerged as clearly superior. Fig. 6 illustrates the results for several regimens that have recently attracted interest:
Although the data in Fig. 6 were not obtained from prospective, random-assignment trials and, thus, are not suitable for statistical analysis, we can conclude that the various regimens examined in the five studies yielded similar outcomes. These studies suggest that several therapeutic alternatives currently available have the potential to produce optimal long-term results. These options include: (1) improve induction therapy (Bishop and Berman curves); (2) provide short-term, intensive postremission therapy (CALGB and Schiller curves); and (3) provide long-term, less intensive, postremission therapy (Wiernik curve). Options 1 and 3 are available to patients of all ages; while option 2 is too toxic for patients beyond middle age. No evidence from other studies indicates that combining alternatives improves outcome.
SECOND-REMISSION THERAPY FOR ACUTE MYELOID LEUKEMIA
Bone marrow transplantation provides long-term, disease-free survival among patients in second remission.23 Because this procedure is not appropriate or available to a large number of patients for various reasons (eg, early relapse, inaccessibility of transplant facility, no HLA-compatible donor), however, other potentially beneficial forms of treatment must be assessed.24 In an ECOG study, patients receiving low-dose ara-C had a statistically significantly longer duration of second remission than those without treatment, indicating that chemotherapy during second remission can be beneficial.25 Biological response modifiers, alone or in conjunction with other modalities, have recently been examined among patients with AML in second remission. Results from two small studies of interleukin-2 administered alone26 or in conjunction with autologous bone marrow transplantation27 were provocative enough to justify prospective, multicenter trials, which will soon be underway.
SUPPORTIVE CARE FOR ACUTE MYELOID LEUKEMIA
Two recent placebo-controlled studies28,29 examined whether GM-CSF administered soon after the completion of induction therapy in elderly AML patients would clinically significantly improve hematologic recovery. The ECOG and CALGB used yeast- and E. coli-derived GM-CSF, respectively. No evidence of a detrimental effect from GM- CSF was apparent in either study—an important finding because of concern that such a growth factor might stimulate the leukemic clone. In the ECOG study, patients receiving GM-CSF had a significantly more rapid recovery of peripheral granulocyte count and significantly fewer infections, compared to patients receiving placebo. In the CALGB study, however, outcomes with the growth factor and placebo were similar. These discrepant results could be explained by the different sources of recombinant growth factor, doses of chemotherapy and age groups in the two studies. Further investigation is necessary to resolve the role, if any, of GM-CSF in AML.