Similar outcome after allogeneic stem cell transplantation with a modified FLAMSA conditioning protocol substituting 4 Gy TBI with treosulfan in an elderly population with high-risk AML

The fludarabine, amsacrine, and cytarabine (FLAMSA)-reduced-intensity conditioning (RIC) protocol has been described to be effective in patients with high-risk and refractory acute myeloic leukemia (AML) undergoing allogeneic hematopoietic stem cell transplantation (aSCT). To increase safety and tolerability of the conditioning, we previously reported the feasibility to substitute the TBI component by treosulfan in elderly AML patients. We now present long-term follow-up data on patients treated with FLAMSA/treosulfan compared to the original FLAMSA/4Gy TBI protocol. We retrospectively analyzed 130 consecutive patients with high-risk or relapsed AML after aSCT following FLAMSA conditioning at our center. Fifty-eight patients were treated with FLAMSA/treosulfan due to age and/or comorbidities. Seventy-two patients were treated with FLAMSA/TBI. Median age of patients treated with FLAMSA/treosulfan was 60 years compared to 46 years in those treated with FLAMSA/TBI. The cumulative incidence of a non-relapse mortality at 4 years was 28% in FLAMSA/treosulfan patients as compared to 13% in FLAMSA/TBI. Cumulative incidence of relapse was higher in patients treated with FLAMSA/TBI (46 vs. 32%). This difference was even more prominent for patients treated in blast persistence prior to transplant (relapse incidence 70% for TBI vs. 35% for treosulfan). The overall and relapse-free survival rates at 4 years were 47 and 41%, respectively, for patients treated with FLAMSA/TBI as compared to 43 and 40% in patients treated with FLAMSA/treosulfan. These data indicate an anti-leukemic activity by FLAMSA/treosulfan especially in patients with a blast persistence prior to transplant. Older age was an independent factor for a higher non-relapse mortality. Translating FLAMSA/treosulfan to younger patients, a lower non-relapse mortality, and an improved anti-leukemic activity might add up to improved overall survival. Randomized studies are required to demonstrate an improved efficacy of treosulfan- versus TBI-based FLAMSA conditioning.