HLA typing by NGS in the clinical lab: our one and a half years experience

Aim We present our experience using the Holotype™ Kit by Omixon for next generation sequencing (NGS) of HLA genes in a clinical setting for one and a half years, starting December 2013. Methods Samples were prepared using the X4 Holotype™ kit by Omixon for up to 7 loci (full length HLA-A, B, C, DQA1 and DQB1, intron 1 to intron 4 of DRB1, and intron 1–3’ UTR of DPB1) and were sequenced in both an individual and pooled amplicon approach using 2 × 251 bp reads on the Illumina MiSeq. Samples were also characterized in parallel with a second method (Sanger SBT, SSO, or SSP) to confirm the genotyping results by NGS and identify possible causes of discrepancies, if any. Two analysis programs were used to genotype the NGS data, Target by Omixon (v 1.7/1.8) and NGSengine by GenDX (v 1.3/1.6). Results A total of 1046 samples have been sequenced: 497 (47%) related to bone marrow transplant cases, 250 (24%) related to solid organ transplantation, and 299 (29%) samples being either proficiency, outreach or disease association studies. Approximately 975 typings were performed for HLA-A, B, C, DRB1 and DQB1, and 600 typings for DQA1 and DPB1. On average, one sequencing run was performed per week with 12 samples on the MiSeq for a total of 83 runs utilizing both nano and full flow cells. At best, the turnaround time from amplification to analysis completion was 4 days, but averages 10 days due to the laboratory workflow and a single sequencing run per week. Concordance was 99.7% between NGS and secondary typing methods (SBT/SSP/SSOP). Discordance was due to incorrect or incomplete (ambiguous) SBT/SSP typing, allele imbalance, NGS genotyping software errors, incorrect reporting of results and new alleles. Two genotyping programs were used for analysis to safeguard against systematic analysis errors and agree for 94.8% of the best match allele calls. Ambiguous results are most common at the DPB1 locus due to a lack of phasing between exons 2 and 3 or the unsequenced exon 1 (17% of alleles) and the DRB1 locus due to not sequencing exon 1 (3.8% of alleles). No ambiguities were detected among the other loci. Conclusions We have genotyped over 1000 samples using NGS, with a very high concordance rate with other HLA typing methods and low ambiguity rates for DRB1 and DPB1 as described above. This technology is therefore very appropriate for HLA typing in a clinical setting. D. Monos: Grant/Research Support; Company/Organization; Omixon. 7. Other (Identify); Company/Organization