Young (35 years) matched sibling donors for allogeneic transplantation with post-transplantation cyclophosphamide in patients with secondary acute myeloid leukemia in first complete remission: A study from the ALWP/EBMT

Abstract Background: Allogeneic hematopoietic stem cell transplantation (HSCT) is a curative therapeutic option for secondary acute myeloid leukemia (sAML), a distinct type of high-risk AML, typically associated with inferior outcomes. Human leukocyte antigen disparity and donor age are two key factors affecting HSCT outcomes. Whether a young haploidentical donor is preferred over an older sibling as a potential donor for HSCT with post-transplant cyclophosphamide (PTCy) in sAML is unknown. Methods:The study aimed to assess whether a young (<35 years [y]; median 27.7 [interquartile range: IQR 23.7, 31.3]) haploidentical donor (yHaplo) is preferential to an older (> 35 y; median 57.9 [IQR 50.0, 61.5]) matched sibling donor (oMSD) in sAML patients (pts) undergoing HSCT in first complete remission with PTCy, between 2012 and 2022. Statistical tests included a multivariable analysis. Results: 406 pts were included (yHaplo=306, oMSD=100). The median follow-up was 2.4 y (IQR, 2-23.9), and the median year of transplant was 2019 (IQR, 2017-2021), with no significant difference between the two donor groups. Median pt age was 58.9 y (IQR, 52.9-63.4), and 61.6% were male, comparable between the groups. The antecedent hematological disease was myelodysplastic/myeloproliferative disorder in both groups, 86.3% vs. 82%, respectively (p=0.92). The cytogenetic risk was categorized as intermediate (65.2% vs. 61.4%), adverse (33.6% vs. 37.5%), and favorable (1.2% vs. 1.1%) in the yHaplo and oMSD groups, respectively (p = 0.79) (data missing for 71 pts). Karnofsky performance status, the frequency of pt cytomegalovirus seropositivity, and female donor to male pt combination did not differ between the two groups. Median time from diagnosis to transplant was significantly longer for yHaplo vs. oMSD, being 5.1 vs. 4.1 months, respectively (p<0.001). Bone marrow was the stem cell source in 19% of the yHaplo vs. 5% of the oMSD grafts (p<0.001), following myeloablative conditioning (MAC) in 56.4% vs. 51% (p=0.35). PTCy was most frequently combined with cyclosporine A /mycophenolate mofetil-based immunosuppression as graft-versus-host disease (GvHD) prophylaxis, 62.8% vs.58 %, respectively. Thiotepa/busulfan was the most frequent conditioning in yHaplo pts, while it was busulfan/fludarabine in the oMSD group. The cumulative incidence of day +30 absolute neutrophil count (>0.5 x 109/L) was inferior in yHaplo compared to oMSD 83.9% vs. 93.8%, hazard ratio (HR) 0.6 (95% CI, 0.43-0.82, p=0.002). Day +60 platelet count (>20,000) was 78.2% vs. 90.4%, respectively. Day +180 incidence of acute (a) GvHD II-IV was 24.5% vs.25.2 % and of III-IV it was 8.3% vs. 11.6%, respectively. The 2-y total and extensive (ext) chronic (c) GvHD were 26.2% vs. 36.5% and5.4% vs. 20.2 %, respectively. In the multivariable analysis, ext cGvHD was significantly lower in the yHaplo compared to the oMSD groups, with an HR 0.38 (95% CI, 0.18-0.78, p=0.008), while aGvHD II-IV and III-IV and total cGvHD did not differ significantly. The 2-y non-relapse mortality (NRM) was higher in HSCTs from yHaplo compared to those from oMSD, being 24.6% vs. 19% (HR 1.86; 95% CI, 1.07-3.2, p=0.028). The main cause of death was sAML, accounting for 40.2% vs. 66.7% of deaths, respectively. The 2-y overall survival (OS:59.6% vs. 49.4%), GvHD-free, relapse-free survival (GRFS: 45% vs. 33.5%), leukemia-free survival (LFS: 53.6% vs. 45.4%), and relapse incidence (RI: 21.8% vs. 35.7%) did not differ significantly between the two groups. A lower Karnofsky score (<90) was a poor prognostic factor for neutrophil engraftment, NRM, and ext cGvHD. Higher age (per 10-y) was a poor prognostic factor for neutrophil engraftment and GRFS. A peripheral blood graft was a poor prognostic factor for aGvHD II-IV, total cGvHD, NRM, and GRFS. Adverse-risk cytogenetics was a poor prognostic factor for RI, LFS, OS, and GRFS. MAC was a poor prognostic factor for aGvHD II-IV. Conclusions: In this registry-based retrospective analysis of HSCT with PTCy for sAML in first complete remission, comparing yHaplo to oMSD, we observed similar transplantation outcomes, but a higher incidence of NRM and slower neutrophil engraftment, while a lower incidence of ext cGvHD was observed in the yHaplo compared to the oMSD grafts.