Poster Board I-221
Currently, most treatment algorithms reserve the use of allogeneic hematopoietic stem cell transplantation (HCT) in patients with acute myeloid leukemia (AML) in first complete remission (CR) to patients with a matched related donor (MRD) and intermediate/high-risk disease. However, the role of HCT from a matched or mismatched unrelated donor (MUD/MMUD) in patients with AML remains to be defined. We retrospectively analyzed a cohort of 219 consecutive adult patients (98 female, 121 male) with AML who received HCT from 2000-2009 at our institution. The patients were transplanted after either myeloablative (MAC, n=139) or dose-reduced-conditioning regimens (RIC, n=80). Median age of patients was 50 years (range, 18-76). 77 patients were transplanted from MRD, 80 patients from MUD and 62 patients from MMUD (one antigen mismatch (MM)=31; two antigen MM=2; one allel MM=24; two allel MM=3, one antigen/one allel MM= 2). In all but six patients receiving MMUD grafts, ATG was included in the conditioning. Age, risk profile and pretreatment were evenly distributed among the three cohorts of patients. At time of HCT 22 (MRD), 18 (MUD) and 28 (MMUD) patients were not in CR. Current overall survival is 40 of 77 (52%) in patients transplanted from MRD, 48 of 80 (60%) from MUD and 34 of 62 (55%) from MMUD with a median follow-up of 1309 (range, 98-3173), 796 (range, 87-3075) and 648 (range, 111-1973) days of alive patients, respectively. Kaplan-Meier-estimated 3-year overall survival (OS) was similar with 54% after MRD-, 56% after MUD- and 46% after MMUD-HCT (p=0.4554). In patients transplanted in CR, 3-year estimated OS was also comparable (64% MRD vs. 58% MUD vs. 55% MMUD, p=0.6614). However, in patients transplanted in partial remission (PR) we observed a trend for a better survival in patients receiving a MUD graft (30% MRD vs. 46% MUD vs. 39% MMUD, p=0.1707). In the patients receiving MAC we observed a better OS compared to RIC with an estimated 3-year OS of 58% vs. 38% (p=0.1047) mainly due to a lower incidence of relapse. In the subgroup of patients receiving MRD-HCT this survival benefit was significant (61% vs. 21%, p= 0.0327) while there was only a trend for MUD- or MMUD-HCT (60% vs 45%, p=0.5702 and 49% vs. 43%, p= 0.7566, respectively). There was no significant difference in the incidence of acute GvHD >II with 25% (MRD), 35% (MUD) and 34% (MMUD) or chronic GvHD with 43% (MRD), 46% (MUD) and 34% (MMUD), respectively. A significant better survival of patients with limited cGvHD vs. extensive or without cGvHD (estimated 3-year OS 73% vs. 34% vs. 47%, p=0.0001) was observed. This advantage was present in all subgroups with a significant better survival in the group with MRD (86% vs. 38% p= 0.0034), a trend in MUD (67% vs. 55% p= 0.0564) and MMUD (59% vs. 55%, p= 0.3111). No significant influence on survival or GVHD of the degree and loci of HLA-mismatch could be detected. In conclusion in our cohort of patients, HCT from MUD or MMUD in AML resulted in a similar outcome compared to MRD. In patients with PR at time of HCT, the use of MUD and occurrence of limited cGVHD may lead to improved survival due to an enhanced graft-versus-leukemia-effect.
Off Label Use: some chemotherapeutical agents in the conditioning are off-label-use.
Asterisk with author names denotes non-ASH members.