Haploidentical Donor versus Mismatched Unrelated Donor in Reduced-Intensity Conditioning Stem Cell Transplantation: A...

The optimal donor for allogeneic stem cell transplantation (HSCT) remains debated, particularly when neither a matched related nor an unrelated donor is available, a scenario that is likely to increase in the near future. This study compares the outcomes of haploidentical HSCT (HAPLO group) versus mismatched unrelated donor (MMUD) HSCT performed with reduced-intensity conditioning (allo-RIC) protocols and various graft-versus-host disease (GVHD) prophylaxis strategies. The primary endpoint was to compare GVHD-free and relapse-free survival (GRFS) between the HAPLO and MMUD groups. Outcomes for MMUD transplantations using post-transplantation cyclophosphamide (PTCy) as GVHD prophylaxis (MMUD-PTCy group) were analyzed separately from those using alternative GVHD prophylaxis regimens (MMUD-OTHERS group). Secondary endpoints included acute and chronic GVHD, disease-free survival (DFS), overall survival (OS), nonrelapse mortality (NRM), relapse, organ toxicities, and hospitalization burden. Patients undergoing their first allo-RIC between January 2012 and March 2022 at 12 GETH-TC/EBMT centers were included. HAPLO transplantations used PTCy-based GVHD prophylaxis. MMUD transplantations received either PTCy-based regimens (MMUD-PTCy) or alternative strategies, mainly sirolimus plus tacrolimus (MMUD-OTHERS). A total of 330 HAPLO, 49 MMUD-PTCy, and 76 MMUD-OTHERS HSCTs were analyzed. Two-year GRFS was comparable across the 3 groups: 47% for HAPLO, 52% for MMUD-PTCy, and 43% for MMUD-OTHERS (P = .8). The predominant cause of GRFS failure differed by group: NRM was more common after HAPLO, while relapse was the main contributor to GRFS failure in MMUD transplantations with or without PTCy. No significant differences among the 3 groups were observed in 2-year OS (59% for HAPLO, 64% for MMUD-PTCy, and 64% for MMUD-OTHERS; P = .7) or 2-year DFS (52%, 57%, and 53%, respectively; P = .9). NRM and relapse also were similar across the 3 groups. PTCy-based prophylaxis significantly impacted neutrophil and platelet engraftment times. At 6 months post-HSCT, the incidence of grade II-IV or III-IV acute GVHD did not differ among the groups. The use of antithymocyte globulin in the MMUD-OTHERS group may have contributed to the comparable rates of acute GVHD across the 2 groups; however, patients receiving PTCy-based GVHD prophylaxis had a lower incidence of moderate/severe cGVHD, regardless of donor type: 12.2% (95% confidence interval [CI], 8.8% to 16.2%) for HAPLO, 11% (95% CI, 3.9% to 22.2%) for MMUD-PTCy, and 21.7% (95% CI, 12.8% to 31.7%) for MMUD-OTHERS. Donor selection strategies varied across centers. Although haploidentical donors are often chosen for their ready availability, unrelated donor searches are efficient even in urgent settings, with MMUDs associated with the greatest likelihood of identifying a suitable donor. MMUD and haploidentical transplantations using PTCy achieve comparable GRFS, OS, relapse, and NRM in patients without a fully matched donor. Moreover, MMUD HSCT without PTCy carried a higher incidence of moderate-to-severe cGVHD and should be avoided when a PTCy-based approach is feasible.