OA3-AM24-TU-15 - Real-World Feasibility of a Universal Blood Donor Typing Array for Large-Scale Genotyping of Red Cell, Platelet, and Leukocyte Antigens in an International Multi-Ethnic Donor Cohort

Tuesday, October 22, 2024

10:30 AM - 11:30 AM

Location: 362

CE: Zero

Presenting Author(s)

William J. Lane, MD, PhD, A(ACHI)

Tissue Typing Lab Director
Brigham and Women's Hosptial, Massachusetts, United States

Disclosure(s): CareDx: Consultant/Advisory Board (Ongoing); One Lambda: Consultant/Advisory Board (Ongoing); Thermo Fisher Scientific Inc: Consultant/Advisory Board (Ongoing), Royalties (Ongoing)

PDF

Disclosure(s):

William J. Lane, MD, PhD, A(ACHI): CareDx: Consultant/Advisory Board (Ongoing); One Lambda: Consultant/Advisory Board (Ongoing); Thermo Fisher Scientific Inc: Consultant/Advisory Board (Ongoing), Royalties (Ongoing)

Background/Case Studies: The Blood transfusion Genomics Consortium (BGC) was formed to develop an all-inclusive DNA array-based genotyping solution for human erythrocyte (HEA), platelet (HPA), and leukocyte (HLA) antigens. We have previously published the results of 7,474 samples, mostly of European ancestry, with genotyping performed in research labs. Here, we present four multi-ethnic cohorts, including the real-world feasibility of array genotyping at blood centers.

Study

Design/Methods:
DNA samples were collected from seven international blood centers. Ancestry was inferred using array ethnicity probes. HEA and HPA were determined using bloodTyper and HLA using HLA*IMP:02. Results were compared to the donor test-of-record. Existing PCR-based genotyping assays and gene sequencing were used for discordance resolution. The first three cohorts were used to improve the assay through iterative improvement cycles using 1,925 African samples, 6,953 multi-ethnic samples, and 367 rare blood group samples. The fourth multi-ethnic cohort of 6,946 samples was used as a real-world feasibility study with genotyping performed at three blood centers.

Results/Findings: The first three datasets led to many improvements. The best array probes for each small nucleotide variation (SNV) were identified, including evaluation for inter-gene cross-reactivity due to homologous genes. The interpretive software was updated with array-based copy number analysis for additional structural variants (SV), especially in the Rh and MNS blood group systems. SV detection also allowed for special probe clustering rules to genotype SNVs co-existing with SVs. The fourth dataset comprised 6,775 multi-ethnic samples: 65% European, 13% African, 9% Ad mixed American, 8% South Asian, 3% East Asian, and 2% Other. The HEA concordance at the three centers compared to the test-of-record was 99.89%, 99.90%, and 99.88%. For HPA, concordance was 99.64%, 99.57%, and 99.65%. For a subset of 767 samples with previous HLA results, concordance was 99.7%, 98.7%, and 99.7% for class I A, B, and C and 96.9%, 99.9%, 98.9% for class II DPB1, DQB1, and DRB1. The inter-lab reproducibility was 99.93%. For the 181 HEA and HPA discordances, 54% were due to the previous test-of-record being wrong, and another 38% should be fixable with software updates.

Conclusions: We present data that DNA arrays can produce highly accurate HEA, HPA, and HLA genotypes at scale and in ethnically diverse samples. Importantly, all genotyping in this study was conducted at accredited blood service molecular laboratories, demonstrating that this technology can be implemented for routine donor typing, paving the way for better-matched transfusions in the near future.