Gragert · The New England journal of medicine 2014 · population-based genetic modeling study · n=?

HLA match likelihoods for hematopoietic stem-cell grafts in the U.S. registry.

Cited 1107 times in the scientific literature.

Level 4 - case-series / case-control

Population-based genetic modeling and observational registry analysis

PubMed 25054717 · doi:10.1056/NEJMsa1311707 · record verified 2026-08-30

What was done

Population-based genetic models were constructed using HLA data from the National Marrow Donor Program (NMDP) Be the Match Registry to estimate the likelihood of finding a suitable adult donor or banked umbilical cord-blood unit across 21 U.S. racial and ethnic groups. The models incorporated the degree of HLA matching, adult donor availability to donate, and cord-blood unit cell dose.

What was found

The likelihood of identifying an optimal adult donor (high-resolution match at HLA-A, HLA-B, HLA-C, and HLA-DRB1) ranged from 75% in whites of European descent to 16% in blacks of South or Central American descent, with other racial/ethnic groups intermediate. While optimal cord-blood units were rarely identified, cord-blood units mismatched at one or two HLA loci were identified for nearly all patients younger than 20 years of age and for more than 80% of patients aged 20 years or older regardless of racial or ethnic background.

Why it matters

This study defines the extent of racial and ethnic disparities in finding matched adult stem-cell donors in the United States, highlighting where adult registry recruitment falls short and demonstrating that cord-blood units provide a viable alternative graft source across diverse populations.

Limits

The findings are derived from mathematical and genetic models rather than real-world prospective match outcomes. Registry sample sizes, confidence bounds, and clinical transplantation outcomes for the identified donor matches were not reported in the abstract.

Cited by