Better outcomes with haploidentical than matched sibling donors for adult AML patients allo-transplanted in CR with detectable MRD: A study from the global committee and the ALWP of the EBMT.

Abstract Background Disease relapse is a major challenge for acute myeloid leukemia patients after allogeneic hematopoietic cell transplantation (allo-HCT). The risk of post-transplant relapse significantly increases in patients who are transplanted in complete remission with detectable measurable disease (MRD positive) as opposed to MRD negative. Haploidentical donor (HAPLO) grafts may exert a higher graft-versus-leukemia effect than human leukocyte antigen (HLA)-matched donors. In a European Society for Blood and Marrow Transplantation (EBMT) global multi-center, registry-based analysis, we compared the outcomes following allo-HCT from either a HAPLO, matched sibling donor (MSD), or a 10/10 HLA-matched unrelated donor (MUD). Methods Data from 3385 adult AML patients receiving a first allo-HCT in first or second complete remission (CR1/CR2) from 269 EBMT centers between 2015 and 2022 were analyzed. 2765 patients were transplanted in CR1 and 620 in CR2. All had detectable measurable residual disease by molecular biology (MRD positive). Transplants from HLA-mismatched UD (<10/10), umbilical cord blood, or grafts with ex-vivo manipulation were excluded. HAPLOs were restricted to recipients with post-transplant cyclophosphamide (PTCy) as graft-versus-host disease (GVHD) prophylaxis. Multivariable Cox regression models were used to evaluate the impact of donor types on outcomes. Results Results from 3385 HCTs were analyzed comprising 609 (18%) HAPLO, 1057 (31%) MSD and 1719 (51%) MUD. For the entire study population, the median age was 55 years (range: 18-83). In the HAPLO group, the proportion of patients transplanted in CR2 was higher (22%) than in the MSD (15%) or MUD (18%) groups (p< 0.001). Peripheral blood was the major graft source (91%) and reduced-intensity conditioning (RIC) was usedfor approximately half (49%) of the study population. Primary disease relapse was the major cause of death in each of the three cohorts. On multivariable analysis, compared with HAPLO HCT (as reference group), MSD HCT (hazard ratio [HR]=1.55, 95% CI 1.26-1.91; p<0.001) was associated with a significantly higher risk of disease relapse (HR=1.55, 95% CI 1.26-1.91; p<0.001)and lower non-relapse mortality (NRM) (HR=0.63, 95% CI 0.46-0.86; p=0.004) leading to an inferior LFS (HR=1.19, 95% CI 1.01-1.41; p=0.048) and a lower GVHD/relapse free survival (GRFS) (HR=1.22, 95% CI 1.05-1.41; p=0.01). Interestingly, in reference to HAPLO, although MUD HCT was associated with a higher incidence of disease relapse (HR=1.26, 95% CI 1.03-1.54; p=0.03), there was no significant differences concerning LFS, GRFS and OS. We also found, as expected, that HCT at CR2 was associated with a significantly higher incidence of relapse (HR=1.5, 95% CI 1.27-1.78; p<0.001) compared with at CR1 and therefore significantly poorer survival outcomes (OS, LFS and GRFS). Myeloablative conditioning regimen was associated with a significantly lower relapse incidence (HR=0.75, 95% CI 0.64-0.87; p<0.001) and significantly superior survival outcomes (OS, LFS and GRFS) compared with reduced intensity conditioning. In addition, patients in better physical condition at transplant (Karnofsky Performance Status >= 90) had a lower NRM (HR=0.79, 95% CI 0.62-0.99; p=0.048) and a better OS compared with those in poorer physical condition. Finally, older patient age (by 10-year intervals) was associated with higher NRM (HR=1.23, 95% CI 1.12-1.36; p<0.001) and poorer OS (HR=1.10, 95% CI 1.04-1.17; p<0.001). Conclusions For AML patients transplanted in MRD positive complete remission, HAPLO HCTs were associated with better LFS and GRFS than MSD HCTs, primarily due to better relapse prevention.