Similar 5-Year Survival after Peripheral Blood Autotransplants (AutoPB) Versus HLA Matched Sibling Myeloablative Transplants (AlloBMT) for Acute Myeloid Leukemia (AML) in First Complete Remission (CR1).

Optimum post-remission strategies for AML CR1 remain uncertain. Previous studies comparing autotransplants with alloBMT favored the latter because of less relapse despite treatment-related mortality (TRM) rates of 20–30%. In those studies, the autotransplants were exclusively with bone marrow (BM) and reported TRM of 12–20%. We were interested in determining the effect of lower TRM with AutoPB shown in many series to be 5–10%. We compared TRM, relapse, leukemia-free survival (LFS), and overall survival (OS) of 1133 patients 19–60 years of age in CR1 who underwent HLA-identical myeloablative BM or peripheral blood (PB) allogeneic transplants or AutoPB for AML reported to the CIBMTR from 1995–2004. For alloBM, alloPB and autoPB, median ages were 36, 40 and 44 yrs; cytogenetics were missing/unknown in 26%, 12% and 17%; proportion with poor risk cytogenetics was 6%, 9% and 7%; and median follow-up of survivors was 82 (6–149), 60 (4–144), and 62 (1–139) mo., respectively. Outcomes were evaluated in univariate and by multivariate analyses using the pseudo-value technique.

Results: 5 year transplant outcomes by univariate analysis are as follows:

All analyses showed that risks associated with alloBM and AlloPB were identical, hence relative risk (RR) was modeled for any allotransplant (alloBMT) vs autoPB. Multivariate analysis of alloBMT vs autoPB cohorts revealed that age >40 yrs was associated with increased RR for TRM, LFS and OS at 5yrs:1.39 (1.05–1.84;n=518;p=0.019);1.25(1.04– 1.51;n=521;p=0.015);1.28(1.05–1.56;n=521;p= 0.012), respectively. For relapse at 5 yrs, the presence of extramedullary disease was a poor prognostic factor: [RR: 1.60 (1.13– 2.27); for no n=1036; yes n=88]. Cytogenetics risk category was not a significant covariate for TRM (p=0.84), relapse (p=0.13), LFS (p=0.27) or OS (p=0.062) at 5 yrs.

Conclusions: AutoPB may provide acceptable alternative post-remission therapy for AML CR1 in the absence of a matched sibling donor, and its role in this regard should be reinvestigated prospectively.