Abstract 4129

In patients with cytogenetically normal AML, the mutational status of FLT3, NPM1 and CEBPA are associated with the outcome (Schlenk, NEJM 2008). In that study, the benefit of ASCT was limited to the subgroup of pts with FLT3-ITD or the genotype consisting of WT NPM1 and CEBPA without FLT3-ITD (triple neg). In these pts, ASCT provided a better RFS not translating into a better overall survival. In that study, pts were under the age of 60 and were transplanted with an HLA matched related donor after a myeloablative conditioning. In an effort to further explore the role of ASCT in AML with FLT3-ITD or triple neg, we undertook a single centre retrospective analysis of de-novo AML with a cytogenetically intermediate-risk profile (Döhner, Blood 2010) treated at 1st CR with a RIC ASCT or conventional consolidation chemotherapy in the absence of a suitable donor.


All pts age 18 up to 65 diagnosed with AML in our center between January 2001 and December 2010 were reviewed. Secondary AML and APL were excluded as were AML with favorable or unfavorable karyotypes (according to Döhner). Pts who never reached CR were also excluded. Furthermore, pts not genotypically defined at diagnosis with available frozen leukemic cells were retrospectively analyzed and only AML with FLT3-ITD or triple neg were included in the study. To avoid biases in favor of the donor group, pts excluded from ASCT because of a poor performance status were excluded as were pts deceased before the median time between CR1 and ASCT in the donor group. As a consequence, the only reason for not performing ASCT was the absence of an appropriate donor. The aim of our study was to compare RIC ASCT to conventional chemotherapy as the post-CR1 therapy.


67 pts were included (30 treated with conventional chemotherapy, the “no donor” group and 37 treated with ASCT, the “donor” group). Both groups (donor vs no donor) were comparable with respect to med age at dg: 57 y (31–64) vs 54 y (19–63), WBC at dg, sex ratio, proportion of normal/abnormal karyotypes: 28/9 vs 23/7, proportion of FLT3-ITD/triple negative genotypes: 10/27 vs 14/16, median time between dg and CR1: 52 d (29–230) vs 45 d (32–75), and number of lines (n=1/ n=2/ n=3) to reach CR1: 23/12/2 vs 21/9/0. The med time between CR1 and ASCT was 114 days (24–295). Conditioning were fludarabine+busulfan+ATG (n=20), fludarabine+cyclophosphamide+TBI2Gy (n=3), fludarabine+TBI2Gy (n=11), fludarabine+treosulfan+ATG (n=3). The source of stem cells were PB (n=33), BM (n=1), or cord blood (n=3). Donors were matched-related or -unrelated, in 51% and 30% of patients, respectively. Med F.U after CR1 was 28 months (6 to 112) and 54 months (6–83) in the donor and no donor groups, respectively. In the donor group, 10 patients relapsed at a med time of 8 months (4–39) after CR1. In the no donor group, 19 patients relapsed at a med time of 8 months (1–44) after CR1. In the donor vs no donor groups, the 3-years relapse rate were 29% ±8% vs 65% ± 9%, p=0.007. The 3-years NRM were 25% ± 10% vs 6 % ± 6%, p=0.02. At the last follow-up, 18 patients have died in the donor group from the following causes: disease (n=9), infections (n=7), GvHD (n=1), suicide (n=1). Fifteen patients have died in the no donor group from disease (n=14) or infections (n=1). The 3-years OS were 51% ± 9% vs 41% ± 10%, p=0.9.


in pts with intermediate-risk de-novo AML and FLT3-ITD genotype or WT NPM1 and CEBPA without FLT3-ITD, a RIC ASCT as post-remission therapy improves the PFS as compared to conventional chemotherapy, demonstrating a potent graft-versus-leukemia effect in these pts with AML at a high-risk of relapse. Efforts remain to be done to decrease RIC ASCT associated NRM.


No relevant conflicts of interest to declare.

Author notes


Asterisk with author names denotes non-ASH members.