Related and Unrelated Donor Match Presence for Haematopoietic Stem Cell Transplantation in a Portuguese HLA Referral Cohort

Identifying a suitable HLA-matched donor is essential for allogeneic haematopoietic stem cell transplantation (HSCT), but match presence depends on donor source, family workup, matching thresholds, and donor-dataset representation. We retrospectively analysed fully anonymised records from a Portuguese referral cohort (2021-2026), comprising 924 patients with complete six-locus HLA data, evaluated relatives, and Portuguese Bone Marrow Donor Registry records. HLA-A, -B, -C, -DRB1, -DQB1 and -DPB1 assignments were harmonised at high resolution, retaining P groups without conversion to specific two-field alleles. Related donor match presence was evaluated at 12/12, ≥ 11/12, ≥ 10/12 and probable haploidentical levels, defined by allele sharing at all six loci without confirmed haplotypic phase. Unrelated donor thresholds were 8/8, ≥ 7/8, 10/10, ≥ 9/10, 12/12 and ≥ 11/12. Population-background descriptors, HLA frequency patterns, and exploratory prediction models were also examined. Among all patients, 185 (20.0%) had a 12/12 related donor and 596 (64.5%) a probable haploidentical relative; among 653 patients with an evaluated relative, these proportions were 28.3% and 91.3%. Unrelated donor match presence was 13.1% at 8/8, 48.5% at ≥ 7/8, 12.8% at 10/10, 36.5% at ≥ 9/10, 2.9% at 12/12 and 12.4% at ≥ 11/12. In a same-pool analysis of 23,004 HLA-DPB1-complete donors, match presence decreased from 10.5% (10/10) to 2.9% (12/12), and from 31.8% (≥ 9/10) to 12.4% (≥ 11/12). Strict 12/12 identity was descriptive, not a clinical donor-selection target. Donor match potential is best characterised through structured search profiles integrating family workup, HLA thresholds, HLA-DPB1, and dataset representation. These findings describe genetic match presence within analysed datasets, rather than clinical donor availability.