Long-term outcome of postremission chemotherapy for adults with acute myeloid leukemia using different dose-intensities

Abstract

The long-term results of postremission chemotherapy for 122 consecutive, unselected adults (15-65 years) with acute myeloid leukemia (AML) were assessed in two sequential prospective studies involving an identical 3/7-type induction regimen, and in those achieving remission, another course for early consolidation using 1 day of daunorubicin instead of three. Forty-one patients reaching C.R. during the first study period, were treated with an intensive ablative maintenance ("IM") program for a period of 9 months. They were randomized to either 6 cycles of induction-type regimen or to 6 cycles of an alternating-type regimen consisting of high-dose (HD)-Ara C/AMSA or 5-azacytidine/AMSA every 6 weeks. There was no difference in disease-free survival (DFS) or survival. Results are compared with 27 patients reaching C.R. on the subsequent protocol where IM was replaced by intensive, short-term consolidation ("IC") using 1 cycle of intermediate-dose Ara C plus AMSA and 1 cycle of HD-AraC/AMSA. Fifteen patients received both courses of IC as scheduled, 12 refused the second cycle. There was no significant difference in DFS or survival. Seventeen out of the 122 patients refused either IM or IC following early consolidation ("refusals"). They received no further treatment and served as control. Fourteen percent of all patients underwent autologous or allogeneic bone marrow transplantation (BMT) at different stages of their disease, equally distributed amongst the IM and IC-group. Median DFS was 3.3 months in the refusal group, 12.4 months in the IM-group, and 18.4 months in the IC-group when censored for BMT (p = 0.01) with 6%, 12%, and 40% in C.C.R. at 50 months. Accordingly, median survival was 5.4, 20 and 47 months (p = 0.001) with 6%, 15%, and 45% of patients alive at 5 years. There was a definite trend (p = 0.14) for a higher proportion of long-term survivors in the IM-group when BMT was performed (not censored), while long-term survival was identical in the IC-group whether BMT was considered for analyses (not censored) or not (censored). Median follow-up for both studies is 5.6 years, the longest, 10 years. In conclusion, progressive increments in the intensity of postremission therapy yields in a graded, significant improvement of remission duration and survival.