Allogeneic stem cell transplantation with post-transplantation cyclophosphamide in patients with acute myeloid leukemia achieving first complete remission after one versus two inductions: A study from the ALWP/EBMT

Abstract

Abstract Background: Allogeneic transplantation (HSCT) is a curative therapeutic option for high-risk AML. The results of HSCT are in correlation with the disease status at transplant. Undergoing HSCT while in CR1 is an important prognostic factor. Moreover, HSCT outcomes in patients (pts) receiving conventional graft-versus (vs.)-host disease (GVHD) prophylaxis also correlate with the number of induction courses needed to achieve CR1 (Loke J, Cancer 2024). However, whether it holds for HSCT with post-transplant cyclophosphamide (PTCy) is unknown. Methods:The study aim was to compare transplantation outcomes in AML pts achieving CR1 after one vs. two inductions. HSCTs were performed between 2012-2022, and all pts received PTCy as GVHD prophylaxis. Statistical tests included a multivariate analysis (MVA) adjusting for potential confounding factors using a Cox proportional-hazards regression model for main outcomes. Results: 677 pts (de novo-542, secondary-135) were included, comprising 518 (77%) with 1 induction and 159 with 2 inductions. Median follow-up was 2.2 (range, 2-2.6) years, and the median year of transplant was 2020 (interquartile range, 2017-2021) with no significant difference between the two groups (p=0.14). Median pt age was 56.5 (interquartile range, 43.4-64.1) years, and 58.8% were male, comparable between both groups (p=0.08 and p=0.93, respectively). The cytogenetic risk was categorized as intermediate (69.1% vs. 69.7%), adverse (25.8 % vs.27%), and favorable (5.1% vs. 3.3%), in the 1 and 2 induction groups, respectively (p=0.65; data missing for 33 pts). Donors were haploidentical 65.6% vs. 69.2%; HLA matched sibling 9.3% vs. 11.3%; matched (10/10) unrelated 14.3% vs. 10.1%, and mismatched (9/10) unrelated 10.8 vs. 9.4%, respectively (p=0.46). PB grafts were the main stem cell source in 74.7 % and 71.7% of the pts, respectively (p=0.45). Myeloablative conditioning (MAC) was used in 46.3% vs. 53.5%, respectively (p=0.12). Cumulative incidence of day 30 absolute neutrophil count (>0.5 x 109/L) was 90.4% vs.91.1percentage and of day 60 platelet count (>20,000/L), 87.3% vs. 86.5% for pts receiving 1 vs. 2 inductions to achieve CR1, respectively. Day 180 incidence of acute (a) GVHD II-IV was 29.7% vs. 31.2 % and of III-IV it was 9.1% vs. 9.8%, respectively. The 1-year (y) total and extensive chronic (c) GVHD was 29.8% vs. 34.2%and11.3% vs. 16.1 %, respectively, with no significant difference between pts receiving 1 vs. 2 inductions to achieve CR1. These data were confirmed by MVA. Higher age per 10-y and year of transplantation per 5-y were poor prognostic factors for total cGVHD, with a hazard ratio (HR) of 1.13 (95% CI, 1-1.28, p=0.045) and HR=0.71 (95% CI 0.53-0.96, p=0.024), respectively. The 2-y relapse incidence (RI) was higher in pts receiving 2 inductions compared to those receiving 1 induction to achieve CR1, 27% vs. 19.3%, HR = 1.75 (95% CI, 1.21-2.53, p=0.003). The 2-y overall survival (OS) 69% vs. 62.8%, GVHD-free, relapse-free survival (GRFS) 50.4% vs. 46,3%, leukemia-free survival (LFS) 65% vs. 61.1% and non-relapse mortality (NRM) 15.7% vs. 11.9%, did not differ significantly between pts receiving 1 vs. 2 inductions to achieve CR1, respectively. These data were confirmed by MVA. The main cause of death was the original disease, 48.7% vs. 70.9% of the deaths, respectively. Adverse risk cytogenetics predicts higher RI, aGVHD II-IV and III-IV, and lower LFS, OS, and GRFS in the MVA. MAC was associated with lower RI and superior LFS, OS, and GRFS. A haploidentical donor was a poor prognostic factor for NRM (in comparison to a sibling donor). Other prognostic factors for NRM were higher age per 10 y, female donor to male pt, and CMV seronegativity of both pt and donor compared to other serostatus combinations. Female donor to male pt, intensity of the conditioning, and CMV seronegativity of both pt and donor, compared to other serostatus combinations, were prognostic factors for neutrophil engraftment. Conclusions: In this registry-based retrospective analysis of HSCT for AML in CR1, achieved after one vs. two inductions, we observed similar transplantation outcomes other than for pts with a higher incidence of relapse achieving CR1 after 2 inductions. These data may indicate that in pts receiving PTCy as GVHD prophylaxis, the number of induction courses to attain CR loses its predictive value, which we previously observed in non-PTCy transplants, suggesting the uniqueness of PTCy biology.