Survival after nonmyeloablative versus myeloablative allotransplantation for AML/MDS.

Abstract

6563 Background: Results from the few retrospective studies comparing nonmyeloablative (NMAT) vs conventional myeloablative (CMAT) allotransplant for the treatment of AML/MDS are mixed. Methods: 143 consecutive patients with AML (n=104) or MDS (n=39) underwent HLA-matched allogeneic peripheral hematopoietic cell (PHC, n=119) or bone marrow (BM, n=24) transplantation from a matched related (MRD, n=84) or unrelated (MUD, n=59) donor between January 2000 and June 2010. Patients were treated with CMAT (n=80) or NMAT cyclophosphamide/fludarabine (n=63) conditioning. Among patients with AML, 87 (90.6%) had intermediate/poor risk cytogenetics; among patients with MDS, 32 (84.2%) had intermediate/high IPSS. Patients in frank relapse at transplant were not included. Follow-up was until death or > 120 days. Kaplan-Meier, log-rank and Cox-regression methods were used to retrospectively compare disease/progression free survival (DFS/PFS) and overall survival (OS). Results: The distribution of gender, cytogenetic risk (for AML) and donor source (MRD/MUD) were similar between NMAT and CMAT groups. The NMAT group was older (median age 58 vs. 45 years; p<0.001), more likely to have received PHC (98.4% vs. 71.3%; p<0.001) and have MDS (42.9% vs. 15%; p<0.001). Among patients with AML, the CMAT group was more likely to be in 1st remission or untested relapse (69.1% vs. 33.3%, p<0.001); the NMAT group was more likely to be in 2nd or 3rd remission or have uncategorizable remission status owing to concurrent MDS (n=8) or second hematologic condition (n=2). Among patients with MDS, the NMAT group had higher IPSS (median 1.0 vs. 0.5; p=0.004). Nineteen patients (32.8%) in the NMAT group experienced relapse/progression compared to 24 (32.4%) after CMAT; median DFS/PFS were 426 days and 428 days, respectively (p=0.822). The median OS was 597 and 421 days for NMAT and CMAT groups, respectively (p=0.943). After adjusting for age, graft source (PHC/BM), donor source (MRD/MUD) and disease (MDS/AML), DFS/PFS [hazard ratio 0.90 (p=0.777)] and OS [hazard ratio 0.76 (p=0.401)] were not different between groups. Conclusions: NMAT and CMAT conditioning appear to provide equivelent disease control and survival in patients with high risk, low burden AML/MDS.