Longer telomeres in donors aged ≥35 years reduce graft failure and non-relapse mortality after allogeneic HCT

Abstract

Abstract Background Donor age is a key determinant of recipient outcomes in allogeneic hematopoietic cell transplantation (HCT), with older donor age consistently associated with increased non-relapse mortality (NRM). However, current eligibility thresholds based on chronological age – such as prioritizing donors <35 years – may inappropriately exclude biologically fit older donors. Telomere length (TL) is a marker of cellular aging, and progressive shortening impairs adaptive immune function and reduces hematopoietic stem cell replicative capacity. While TL may have limited impact on graft efficacy in younger donors, we hypothesized that TL heterogeneity is selectively impactful in older donors, where biological and chronological aging diverge. In this multicenter cohort, we examined the association between donor TL and post-HCT outcomes in adult recipients from donors aged ≥35 years. Methods We used qPCR to measure whole blood TL in 7,373 donors ≥35 years with pre-HCT samples in the CIBMTR biorepository. Recipients underwent HCT from matched unrelated (48%), matched related (29%), mismatched unrelated (17%), or haploidentical (6%) donors for AML (59%), MDS (23%), or ALL (18%) using myeloablative (60%) or reduced intensity/non-myeloablative (40%) regimens. Graft-versus-host disease (GVHD) prophylaxis included CNI-based regimens (88%) and post-transplant cyclophosphamide (7%). We used multivariable Fine-Gray regression (Cox for OS) to evaluate associations between TL and recipient outcomes, adjusting for donor-, recipient-, and transplant-related covariates. Results Donor TL declined progressively with increased age (p<0.001) and was longer in females than males (p<0.001). In a competing risks model, longer donor TL (>40th percentile) was independently associated with lower NRM (HR 0.90, 95% CI 0.82–0.98; p=0.012). This effect was not redundant with donor age, as older age remained an independent predictor of NRM even after accounting for TL (50-59 years: HR 1.15, 95% CI 1.01-1.31, p=0.033; ≥60 years: HR 1.35, 95% CI 1.12-1.64, p=0.002, vs. 35–39). No significant difference was observed for donors aged 40–49 (HR 1.04; p=0.49). As expected, established risk factors were also significantly associated with increased NRM, including recipient age ≥40 (HR 1.43, p<0.001), Karnofsky score <90 (HR 1.15, p=0.002), higher HCT-CI (HR 1.21–1.63, p≤0.010), and female-to-male sex mismatch (HR 1.35, p<0.001). Compared to HLA-identical siblings, well-matched (HR 1.64; p<0.001) or partially matched unrelated donors (HR 2.17, p<0.001) conferred significantly higher NRM. These findings confirm expected associations while demonstrating that donor TL and chronological age contribute to NRM risk additively rather than interchangeably. To investigate potential mechanisms underlying the association between donor TL and NRM, we evaluated clinical outcomes linked to hematopoietic stem cell function and immune reconstitution. To determine whether TL impacts replicative capacity, we examined primary graft failure, which occurred in 3.4% of recipients (n=250). Longer donor TL was independently associated with reduced risk of primary graft failure (HR per SD increase: 0.88; 95% CI 0.79–0.98; p=0.02), supporting a link between replicative reserve and graft success. In contrast, TL was not associated with grade 3–4 acute GVHD (HR 0.97, 95% CI 0.82–1.14; p=0.68) or moderate-severe chronic GVHD (HR 1.06, 95% CI 0.94–1.18, p=0.34), suggesting that TL heterogeneity does not substantially influence alloimmune function. Consistent with this, donor TL was not associated with relapse risk (HR 0.97, p=0.40). Donor TL had a modest association with overall survival (HR 0.95, 95% CI 0.90–1.01; p=0.086), likely reflecting the competing impacts of relapse (42% of deaths), and multifactorial non-relapse causes including infection (15%), GVHD (11%), and organ failure (9%). Conclusions Longer TL in donors aged ≥35 was independently associated with reduced risk of graft failure and NRM among recipients of grafts, indicating that TL heterogeneity reflects biological variability among older donors that meaningfully influences transplant efficacy. The protective effect of longer TL, which is independent of established transplant-related factors, supports its potential as a biomarker of donor quality. Incorporating TL into older donor assessment could help refine selection strategies, mitigate NRM risk, and broaden the eligible donor pool beyond rigid chronological age thresholds.