Long-Term Follow-up of Autologous Hematopoietic Stem Cell Transplantation (AHSCT) for Acute Myeloid Leukemia (AML): A Survey of 3567 Patients From the Acute Leukemia Working Party of the EBMT.

Abstract 3112

AHSCT is an option to consolidate remission in a subgroup of patients with AML. Although a significant decrease of number of such procedures has been reported after year 2000, this treatment modality continues to be used by some centers. The major disadvantage of AHSCT is leukemia relapse which occurs mainly during the first 2 years. Recently, a survey of long-term disease-free survivors after AHSCT was published by North American institutions showing mortality rates similar to that of the general population [Majhail NS et al., BMT 2011]. The goal of the current study was to characterize risk factors for outcome of patients who survived at least 2 years after AHSCT without AML relapse.

We analyzed 3567 patients with de novo AML treated with AHSCT in first (n=3087, 86%) or second (n=480, 14%) complete remission (CR) between 1990 and 2008 who remained alive with no signs of disease recurrence at least 2 years thereafter. The analysis was based on data from the EBMT registry. Median age was 45 y (range, 18–78); gender male-52%, female-48%; FAB M0–2%, M1–18%, M2–28%, M3–10%, M4–24%, M5–15%, M6-2%, M7-1%. Among patients with known karyotype, 26% had favorable, 71% intermediate and 3% unfavorable cytogenetic features. Stem cell source was bone marrow (BM) (n=1153, 32%) and peripheral blood (PB) (n=2414, 68%) with a significant shift towards PB after year 2000 (11% before and 69% after). The conditioning regimen was based on chemotherapy in 78% and TBI in 22% of cases, and graft purging was performed in 6% of cases. G-CSF was used to promote engraftment in the immediate post-transplant period in 23% of patients. Median AHSCT year was 2000 and overall follow-up was 6.9 years (range, 2.0–21.5).

The probability of leukemia-free survival (LFS) for the whole study population at 5 and 10 years after AHSCT was 86% and 76%, the relapse incidence (RI) 11% and 16%, and the non-relapse mortality (NRM) 3% and 8%, respectively (all calculated from the day of AHSCT). In univariate analysis, older age at the time of AHSCT, FAB M067 vs. M1–5 vs. M3, adverse cytogenetics, use of PB as stem cell source, unpurged graft and transplantation performed after year 2000 were associated with a worse outcome. CR status, type of conditioning and G-CSF use did not affect the long-term results (see Table 1). Factors that independently affected unfavorable outcome were older age, use of PB and M067 FAB for LFS and RI, whereas older age and M067 FAB for NRM (see Table 2).

In summary, our analysis indicates a relatively favorable outcome for patients surviving disease-free 2 years after AHSCT. However, several patient-, disease- and procedure-related factors may influence the overall results. At present, the role of AHSCT needs to be revisited in the context of minimal residual disease assessment at the time of stem cell collection, as well as the value of post-transplant maintenance (e.g. hypomethylating agents).