P064Evaluation of MIA FORA NGS FLEX HLA typing kit

Abstract

Aim As the HLA database expanding, traditional SBT method results in more and more ambiguities and requires increasing additional testing. We evaluated Mia Fora FLEX11 NGS typing kit with automation intending to replace SBT with NGS. Methods 108 DNA samples were selected. 92 and 16 samples were previously typed by SBT and rSSO/SSP, respectively. All DNA were prepared from blood except 4 from buccal swab and 3 from B cell line. DNA were extracted 2009-2017 with OD260/2801.51–2.05. NGS typing was performed following manufacturers automated protocol using Mia Fora FLEX11 kit. Results 313 unique alleles were tested in our validation cohort including 269 CWD alleles, 38 rare alleles, 3 null alleles and 7 new alleles. All DNA samples were successfully typed with typing results accurate to the resolution of third field according to WHO nomenclature. 11 ambiguities were identified which were all with DPB1. 7 potential new alleles were identified: 2 HLA-B, 1 DRB5, 1 DQA1 and 3 DPA1. NGS typing was 99.5% (1499/1506) concordant with previous typing at allele level. Discordance was observed with DQB1, DRB3/4/5 and DPA1 (Table 1). Allele dropout was observed in 1 out of 23 DQB1*06:02 and 2 out of 11 DRB4*01:01 in the validation cohort. Both dropouts can be predicted by allele association with DQA1 or DRB1. Sample repeat rate was 11% (18/161). Four samples failed due to software glitch with the Window 10 version but corrected when switched to the Linux Version. Most other failure was due to insufficient data which can be resolved by repeating test. The only failure that could not be resolved by repeat was DRB4*01:01 dropout. Download : Download high-res image (262KB) Download : Download full-size image Conclusions Comparing to SBT method, Mia Fora typing produced high resolution results with ambiguities only presented with DPB1 locus. Little or no additional testing was required. Mia Fora typing had a high concordance rate with previous typing results. A reasonable failure rate was observed where most could be resolved by repeating test except for some DRB4*01:01 dropout.