Matched Sibling Donors Are Still ‘Gold Standard’ in PTCy Allo-HCT Era

Matched sibling donors (MSD) and matched unrelated donors (MUD) continue to yield the best 2-year survival outcomes following allogeneic hematopoietic cell transplantation (HCT), while mismatched unrelated donors (MMUD) and haploidentical donors perform comparably to each other and all outperform umbilical cord blood (UCB), according to a large retrospective registry analysis presented at the European Blood and Marrow Transplantation Society Annual Meeting.¹

The analysis of 13,336 patients across 166 US centers examined outcomes through 2 years posttransplant in the contemporary era of posttransplant cyclophosphamide (PTCy)-based graft-vs-host disease (GVHD) prophylaxis. PTCy has increasingly replaced calcineurin inhibitor (CNI)–based regimens and has expanded access to allogeneic HCT by enabling safer use of mismatched donors, but questions have remained about whether outcome differences between donor types would persist or narrow beyond 1 year of follow-up.

“Matched sibling donor remains the gold standard with better [OS] and less [GVHD] and nonrelapse mortality relative to all the alternative donor types,” Steven M. Devine, MD, of the National Marrow Donor Program and the Center for International Blood and Marrow Transplant Research (CIBMTR) said in his oral presentation.

Key Survival Findings

In multivariable analyses adjusted for patient age, sex, race, Karnofsky performance status, HCT comorbidity index, donor gender, graft source, donor/recipient CMV serostatus, disease risk, conditioning, time from diagnosis to HCT, and year of HCT, MSD recipients had the highest 2-year adjusted overall survival (OS) at 70%, followed by MUD at 68% (HR, 1.14; P =.071), MMUD at 65% (HR, 1.29; P =.002), haploidentical at 62% (HR, 1.41; P <.001), and UCB at 57% (HR, 2.45; P <.001).

For GVHD-free, relapse-free survival (GRFS), defined as survival without grade 3 to 4 acute GVHD, systemic therapy-requiring chronic GVHD, or relapse, MSD again led at 46%, with MUD (45%; HR, 0.99; P =.839) and MMUD (45%; HR, 1.09; P =.147) performing comparably. Haploidentical was modestly lower at 41% (HR, 1.13; P = .011), and UCB notably worse at 35% (HR, 1.86; P <.001).

All alternative donor groups showed lower relapse rates compared with MSD. However, this benefit was offset by significantly higher nonrelapse mortality (NRM). NRM hazard ratios compared with MSD were 1.67 for MUD, 2.30 for MMUD, 2.80 for haploidentical, and 6.75 for UCB (all P <.001). Relapse-free survival (RFS) was similar across MUD, MMUD, and haploidentical compared with MSD, while UCB was significantly worse (HR, 1.63; P <.001).

GVHD Outcomes

Rates of grade 3 to 4 acute GVHD were significantly higher across all alternative donor types vs MSD: HR 1.55 for MUD (P =.003), 2.06 for MMUD, 1.96 for haploidentical, and 7.62 for UCB (all P <.001). Notably, moderate-to-severe chronic GVHD did not differ significantly between any donor type and MSD.

Study Cohort

The cohort included adults aged 18 years or older with acute lymphoblastic leukemia, acute myeloid leukemia, or myelodysplastic syndrome who received first allogeneic HCT in the US between 2017 and 2023. All patients were treated with PTCy-based GVHD prophylaxis without antithymocyte globulin, except UCB recipients, who received CNI-based prophylaxis per American Society for Transplantation and Cellular Therapy guidelines. Peripheral blood was the most common allograft source (84%). Median follow-up for survivors was 35 months. All patients in the MMUD group had 7/8 matched donors.

This study extends a prior CIBMTR analysis covering 2017 to 2022 that showed persistent but narrowing 1-year outcome gaps between matched and mismatched donor types in the PTCy era.²

Clinical Implications

The authors noted one important caveat: MSD recipients in the current US dataset are younger than in prior eras of GVHD prophylaxis, with a median age of 56 years, which may partly explain their superior outcomes.¹ Despite this, the findings support a clear donor selection hierarchy.

“Donor selection practices, at least in the United States, should still prioritize matched sibling donors and matched donor related donors when available, and favor MMUD and [haploidentical] over unmanipulated cord blood,” Devine said in his conclusion.

The reasons for the difference in outcomes by donor source remain under investigation. “I can't say that HLA doesn't matter. I still think it does,” Devine said. “Maybe it matters less with PTCy, and…you have the outcomes with the cord blood. Most of it was driven by worse acute GVHD, and [NRM]; probably also immune recovery and issues related to poor immune recovery, causing higher risk of infections.”

Devine also emphasized enrollment on prospective clinical trials and called for future evaluation of longer-term outcomes beyond 2 years.

REFERENCES

1. Devine SM, Litvinovich I, Astha V, et al. Do 2 year patient outcomes differ by donor type using contemporary graft vs host disease prophylaxis post-transplant? A CIBMTR retrospective analysis. Presented at: 2026 EBMT Annual Meeting; March 22-25, 2026; Madrid, Spain. OS03-08.

2. Devine SM, Astha V, Kuxhausen M, et al. Do patient outcomes differ by donor type if post-transplant cyclophosphamide is used for GVHD prophylaxis? A CIBMTR retrospective analysis. Presented at: 2025 EBMT Annual Meeting; March 30-April 2, 2025; Florence, Italy. O157.