Old mismatched donors are associated with inferior outcomes after allogeneic hematopoietic stem cell transplantation in myelodysplastic syndrome, but not in Acute Myeloid Leukemia

Abstract Introduction: Allogeneic hematopoietic stem cell transplantation (alloHCT) is the only potentially curative option for high-risk myelodysplastic syndrome (MDS) and acute myeloid leukemia (AML). While donor selection has traditionally focused on HLA compatibility, improvements in GVHD prophylaxis have expanded the use of mismatched donors. Donor age is increasingly recognized as a prognostic factor, with older donors associated with diminished stem cell quality and potentially worse outcomes. However, the clinical implications of donor age—especially in relation to donor type—remain incompletely understood. AML and MDS are often analyzed together as prototypical myeloid malignancies, yet their biological and risk profiles differ substantially, warranting disease-specific evaluation of donor-related factors such as age. Methods:We retrospectively analyzed adult AML and MDS patients who underwent alloHCT with busulfan/fludarabine conditioning and T-cell depletion using antithymocyte globulin (ATG), post-transplant cyclophosphamide (PTCy), or both, at Seoul National University Hospital between Jan 2016 and Jun 2024. Patients who received second transplantation were excluded. Donors were classified by age (<35 vs. ≥35 years) and donor type (matched: matched related donor/matched unrelated donor vs. mismatched: mismatched unrelated donor/haploidentical donor). Survival and transplant-related outcomes were compared, with subgroup analyses according to disease type. Results: A total of 320 patients were included; 131 (40.9%) received grafts from young donors and 189 (59.1%) from old donors. Young donor recipients were slightly younger (median age, 57 vs. 58 years; P=0.002), had fewer CMV-positive donors (80.9% vs. 99.5%; P<0.001), and received higher CD34⁺ cell doses (median, 5.06 vs. 4.87×10⁶/kg; P=0.009). Mismatched donors were more common among young donors (51.1% vs. 38.1%; P=0.028). Overall, donor age alone was not associated with significant differences in survival (4-year OS: 55.7% vs. 48.0%, P=0.150; RFS: 41.0% vs. 38.3%, P=0.280; GRFS: 31.6% vs. 31.8%, P=0.310). However, stratification by both donor age and type revealed significant outcome differences. The 4-year OS was 58.9% for matched donors, 47.4% for young mismatched donors, and 34.6% for old mismatched donors (P=0.007). Multivariate analysis confirmed the association between old mismatched donors and inferior OS (HR 1.49; 95% CI, 1.01–2.21; P=0.046). Additionally, disease type was a significant factor, with AML being an adverse indicator of OS compared to MDS (HR 2.21; 95% CI, 1.47–3.31; P<0.001). Given its significant impact on survival, disease type was used to stratify subgroup analyses. Subgroup analysis showed that the adverse impact of donor age was driven primarily by the MDS cohort (N=105). In MDS, young donor recipients had significantly superior OS (66.2% vs. 39.7%; P=0.018). Importantly, outcomes were comparable between matched donors and young mismatched donors, whereas old mismatched donors had markedly worse survival. The 4-year OS was 64.2% for matched donors, 51.4% for young mismatched donors, and 18.6% for old mismatched donors (P=0.002). Multivariate analysis confirmed the independent negative impact of old mismatched donors on OS (HR 2.66; 95% CI, 1.38–5.15; P=0.004), RFS (HR 2.45; 95% CI, 1.27–4.74; P=0.008), and GRFS (HR 2.13; 95% CI, 1.16-3.90; P=0.015). In contrast, in the AML cohort (N=215), donor age and type were not significantly associated with survival, whereas disease risk and disease status were the primary prognostic determinants. Conclusions:Donor age and donor type significantly affect alloHCT outcomes in MDS but not in AML. In MDS, matched and young mismatched donors yielded comparable outcomes, whereas old mismatched donors were associated with markedly inferior survival. These findings underscore the importance of exercising particular caution when considering old mismatched donors for MDS patients, especially when alternative options exist. Future efforts to refine donor selection should incorporate biologic measures of graft quality in addition to chronological age.