Second Allogeneic Stem Cell Transplantation (allo-SCT) Using Reduced-Intensity Conditioning (RIC) for Acute Myeloid Leukemia (AML) Patients Who Relapsed Following a First RIC Allo-SCT: a Survey From the Acute Leukemia Working Party of EBMT.

Abstract 3447

Second allogeneic stem cell transplants for hematological malignancies are associated with a high incidence of non-relapse mortality (NRM) due to the cumulative incidence of toxicity. On the other hand, the RIC approach represents an attempt to harness the immune graft-versus-leukemia (GVL) effect while attempting to control or overcome toxicity. The aim of this analysis was to assess the outcomes (overall survival: OS, leukemia-free survival: LFS, NRM, and relapse incidence: RI) in a cohort of 103 AML patients (57 males) who have received a second RIC allo-SCT (RIC2) as a salvage therapy after relapse following a first RIC allo-SCT (RIC1), and who were reported to the EBMT registry with adequate data.

In this series, the median age at time of RIC2 was 54 (range, 20–69) years. The median time from RIC1 to relapse was 184 (range, 28–2377) days. The median time from RIC1 to RIC2 was 307 (range, 44–3165) days, and the median time from relapse to RIC2 was 84 (range, 7–582) days. For the RIC2 transplant, 58 patients (56%) received an allogeneic graft from an HLA-identical sibling donor, while 45 patients (44%) received an HLA-matched or mismatched graft. The stem cell source was PBSCs in the majority of cases (89%).

At time of RIC2 transplant, 82 patients (80%) were in relapsed or refractory disease. Only one patient was in first complete remission (CR), while 20 patients (19%) were in CR2 or CR3. Various RIC regimens as per EBMT definition were used for preparation prior to the second allo-SCT. After RIC2, 89 patients (86%) engrafted and the median time to ANC>500/μL was 15 (range, 1–40) days. The incidence of grade 3–4 acute GVHD after RIC2 was 14%. With a median follow-up of 11 (range, 2–99) months after RIC2, 22 patients were still alive. At 2 years, the rates of OS, LFS, NRM and RI were 19+/−4%, 13+/−4%, 27+/−4%, and 60+/−4% respectively. In multivariable analysis, no factors were associated with a higher risk of NRM after RIC2. However, multivariate analysis showed that the interval from first transplant (RIC1) to relapse (>6 months) was the strongest predictive factor significantly associated with improved LFS and a lower RI after RIC2 (P=0.004, HR=1.91, 95%CI, 1.23–2.95; and P=0.0008, HR=0.41, 95%CI, 0.25–0.69, respectively). When comparing patients progressing before or beyond 6 months after the first RIC allo-SCT, LFS and RI rates were 6+/−3% vs. 21+/−6% (P=0.002) and 71+/−7% vs. 48+/−8% (P=0.001), respectively.

In all, these results suggest that the initial use of a RIC regimen prior to allo-SCT can reduce overall transplant-related toxicity, thus making a second RIC transplant feasible, especially if compared to historical results achieved in the setting of standard myeloablative conditioning second allo-SCT. Results achieved after a second salvage RIC allo-SCT compare favorably with those achieved when testing new drugs in phase I trials performed in this setting of relapsed AML (e.g. HDAC inhibitors etc.). Of note, AML patients who relapse beyond 6 months after a first RIC allo-SCT are potential candidates for a second RIC allo-SCT with a relatively acceptable rate of NRM.