Haploidentical Mismatched Allogeneic Versus Autologous Hematopoietic STEM CELL Transplantation in Adult Patients with ACUTE Myeloid Leukemia (AML) in First Complete Remission (CR1): A Pair-Matched Analysis From the Acute Leukemia Working Party of EBMT.

Abstract 3093

In the absence of a matched family donor, adult patients with AML in CR1 are nowadays usually offered, as a first choice, an allogeneic transplant from an alternative donor (haplomismatched donor or cord blood cells), but there has been thus far no demonstration that outcome when using an alternative donor would be superior to outcome following an autologous stem cell transplantation (ASCT).

Between January 2000 and December 2010, 113 AML patients transplanted in CR1 with an haplo-mismatched family donor and 4444 patients autografted were reported to the EBMT registry. For the purpose of this analysis, we performed a matched pair analysis (113 haplo/226 autos) using age and interval from diagnosis to transplant as matching factors. The median follow up was 18 months (range, 1–128). Patients allografted were transplanted more recently (median year of transplant: 2008 vs 2004, p<10e-4), and they received more frequently total body irradiation in the pretransplant regimen (58% vs 17%). Engraftment following an haplo- mismatched transplant occurred in 93% of cases vs 99% after ASCT and recovery of ANC>500/mm3 was slightly delayed (15 days vs 13 days, p<10e-4). The outcome at two years, following an haplo mismatched transplant versus an ASCT was: LFS 44±5% vs 54±4% (p=0.008); OS 44±6 % vs 68±4% (p<10e-4); Non Relapse Mortality (NRM) 37±5% vs 2±2 ( p<10e-4) and Relapse Incidence (RI) 19±4% vs 43±4% (p<10e-4). Following an haplo mismatched transplantation the incidence of chronic extensive GVHD was 24+/− 6%. By multivariate analysis adjusted on year of transplantation, source of stem cells and TBI, the RI was no longer statistically different between the 2 groups, NRM was higher following haplo (HR=11.8; p<0.0001) and LFS was better following ASCT (HR=0.52; p=0.004).

Though this study did not account for cytogenetics and molecular markers (data collection is ongoing), these results suggest that outcome following ASCT in CR1 may be better compared to outcome following a haplo-mismatched transplantation. Recent developments in both transplant approaches may change the results in the near future and may justify a randomized prospective study.