Post-Transplant Cyclophosphamide Versus Antithymocyte Globulin in Haploidentical Transplantation for Relapsed/Refractory Acute Myeloid Leukemia: A Study from the Acute Leukemia Working Party of the EBMT

Background. Despite greater use of post-transplant cyclophosphamide (PTCy) in Europe and the United States in unmanipulated haploidentical transplantation (Haplo-HSCT), efficacy of antithymocyte globulin (ATG)-based platforms in effectively preventing graft-versus-host disease (GVHD) has also been reported. Better transplant outcomes with PTCy compared to ATG have been reported previously by our group in patients undergoing Haplo-HSCT for acute myeloid leukemia (AML) in complete remission, while comparative analyses focusing on primary induction failure (PIF) or relapsed/refractory (Rel/Ref) AML are lacking. Methods. The current study aimed to compare transplant outcomes of patients with PIF or Rel/Ref AML undergoing their first Haplo-HSCT with either PTCy or ATG as GVHD prophylaxis during the period 2012-2022. Both bone marrow and peripheral blood were included as stem cell sources. Propensity score matching was used to reduce or eliminate confounding effects. Each patient receiving ATG was matched with three patients receiving PTCy using the nearest neighbor method. Variables included in the propensity score model were: stem cell source, age of patient at Haplo-HSCT, year of transplant, female donor to male recipient, interval from diagnosis to transplant, disease status at transplant, cytogenetic risk. Univariable Cox analysis was then performed on pair-matched data. Before matching, a total 977 patients (ATG, n=82; PTCy, n=895) were identified. Herein we report the results after pair-matching. Results. After pair-matching, 234 and 78 patients receiving PTCy or ATG, respectively, were included. Median age at Haplo-HSCT was 52 (range 37-61) for the PTCy and 51 (range 20-72) years for the ATG group. A higher proportion of patients with a Karnofsky score<90 was observed in the ATG group (66% [n=45] compared to 44% [n=97] for PTCy). Distribution of disease status according to PTCy or ATG was as follows: PIF in 45% (n=105) and 44% (n=34), Rel/Ref in 48% (n=113) and 43% (n=34). Most patients were transplanted for intermediate-risk cytogenetic AML (72% [n=121] for PTCy and 67% [n=37] for ATG) while adverse karyotype was registered in 21% (n=35) and 27% (n=15) of cases in PTCy and ATG groups, respectively. Peripheral blood was the preferred stem cell source in both groups (69% [n=161] and 73% [n=57] for PTCy and ATG, respectively). A female donor to male recipient was recorded for 18% (n=41) and 19% (n=15) of PTCy and ATG groups, respectively. Thiotepa-busulfan-fludarabine was the preferred conditioning regimen for both groups (41% [n=97] for PTCy and 27% [n=21] for ATG). More than half of patients received a myeloablative conditioning regimen (57% [n=134] and 54% [n=42] for PTCy and ATG groups, respectively). Neutrophil engraftment at 30 days occurred more frequently in the ATG group (93.3% [95% CI 84.2-97.3] versus 87.9% [95% CI 82.8-91.6] for PTCy, hazard ratio [HR] 1.57 [standard error 1.16-2.14], p=0.004). No differences were observed between the 2 groups for most transplant outcomes, except for a lower risk of extensive chronic GVHD in patients receiving ATG (3% [95% CI 0.5-9.3] versus 16.5% [95% CI 11.7-21.9] for PTCy (HR 0.16, p=0.012), and for higher neutrophil recovery at 30 days for ATG (93.3% [95% CI 84.2-97.3]) versus PTCy (87.9% [95% CI 82.8-91.6]; HR=1.57; p=0.004). With a median follow-up of 4 years, survival outcomes at 3 years were distributed as follows for PTCy and ATG groups, respectively: overall survival 29.4% (95% CI 23.2-35.9) versus 25.1% (95% CI 14.8-36.6), p=0.773; leukemia-free survival 25.6% (95% CI 19.8-31.8) versus 22.8% (95% CI 13.2-33.9), p=0.824; relapse incidence 47.1% (95% CI 40.2-53.6) versus 45.9 (95% CI 33.4-57.5), p=0.909; non-relapse mortality 27.3% (21.5-33.5) versus 31.4% (95% CI 20.6-42.7), p=0.630; 180-day grade II-IV acute GVHD 26.1% (95% CI 20.4-32.1) versus 22.8% (95% CI 13.7-33.2), p=0.511; grade III-IV acute GVHD 11% (95% CI 7.3-15.6) versus 7.1% (95% CI 2.6-14.6), p=0.290; chronic GVHD of all grades 25.1% (95% CI 19.4-31.1) versus 22% (95% CI 12.9-32.6), p=0.391. Conclusions. Both ATG and PTCy represent valid alternatives for GVHD prophylaxis in Haplo-HSCT for PIF or Rel/Ref AML. However, use of ATG may result in lower incidence of extensive chronic GVHD, that remains a major cause of quality-of-life impairment in transplanted patients with high morbidity and mortality.