Abstract
In this study, we sought to identify specific individual high-intensity or reduced-intensity conditioning regimens with the best relapse-free survival (RFS) rather than the global high- vs reduced-intensity regimen comparison. Patients (median age, 58 years) with acute myeloid leukemia (AML; n = 1258), who were in first or subsequent remission, or with MDS (n = 951) who had refractory anemia with unilineage or multilineage dysplasia, 5q- syndrome, or refractory anemia with excess blasts received nonirradiation-containing regimens and were transplanted between 2009 and 2014 in the United States. Three-year RFS with high-intensity busulfan/cyclophosphamide (Bu4/Cy; 44%) was comparable to conditioning with high-intensity fludarabine/busulfan (Flu/Bu4; 44%), reduced-intensity fludarabine/melphalan (Flu/Mel; 52%; P = .53), and Flu/Mel + anti-thymocyte globulin (ATG; 44%; P = .38). RFS was lower with reduced-intensity Flu/Bu2 + ATG (31%; P = .0006). RFS was also lower with high-intensity Flu/Bu4 + ATG (38%; P = .05) and reduced-intensity Flu/Bu2 (38%; P = .02), although the difference did not reach the level of significance set for these analysis. RFS with Flu/Mel was superior to RFS with Flu/Bu2 (P = .01) and Flu/Bu2 + ATG (P = .0006). The 3-year incidence of relapse was 22% with Flu/Mel compared with 46% with Flu/Bu2 and 56% with Flu/Bu2 + ATG. With only a modest reduction in nonrelapse mortality with the Flu/Bu2 regimens, the higher relapse incidence resulted in lower RFS. The data support optimal RFS with Bu4/Cy, Flu/Bu4, and Flu/Mel regimens for AML in remission or MDS. The low relapse rate with reduced-intensity Flu/Mel resulted in RFS comparable to that after the higher-intensity regimens.