High Accuracy and Significant Savings Using Tag-SNP Genotyping to Determine HLA-B*27 Status

To the Editor:

Axial spondyloarthritis (axSpA) is characterized by inflammatory back pain, spinal ankylosis, peripheral arthritis, and extraarticular manifestations¹. HLA-B*27 status is particularly helpful in diagnosis because the combination of severe sacroiliitis on magnetic resonance imaging with HLA-B*27 positivity is an excellent predictor of disease outcome². However, because of inherent challenges associated with multiallelic genetic testing, clinical tests evaluate the entire HLA-B locus using commercially available kits. Consequently, the clinical value of the HLA-B locus test as a tool for primary care physicians is constrained, primarily because of cost. A large genome-wide association study demonstrated that a tag-single-nucleotide polymorphism (SNP) can be used to determine HLA-B*27 status in whites³. The objective of our study was to assess the analytical validity and economic effect of using a HLA-B*27 tag-SNP assay compared with an HLA-B locus test in a clinical genetics laboratory. The ethics committee from Memorial University approved the study. Patient consent was obtained.

The study cohort consisted of 1000 consecutive patient samples received by the clinical genetics laboratory at Eastern Health to determine the HLA-B*27 genotyping status. Institutional ethics was approved from the Health Research Ethics Board at Memorial University (Health Research Ethics Authority reference #15.051). Genomic DNA was extracted from whole blood using an automated MagnaPure Compact according to manufacturer’s protocols (Roche Molecular Systems). The concentration of extracted DNA was determined using a NanoDrop 2000 (ThermoFisher Scientific). Targeted analysis of the HLA-B locus was performed using a commercially available kit (LABType SSO HLA-B locus kit) on a Luminex 100/200 platform as per manufacturer’s instructions (One Lambda, ThermoFisher Scientific). Genotyping of rs4349859 (NM_001289154.1: c.-2921G>A) and rs116488202 (NC_000006.12:g.31377139C>T) was performed using a TaqMan SNP genotyping assay (Assay ID number C_28023949_10) and a custom-designed TaqMan SNP genotyping assay (Assay ID number AHCS8KA), respectively, as per the manufacturer’s instructions (Applied Biosystems). Genotypes were assigned using the endpoint genotyping analysis software (Applied Biosystems). HLA-B*27 positivity was inferred by the presence of the “T” allele and “A” allele for the rs116488202 and rs4349859 SNP, respectively. Statistical analysis was performed using Prism 7 (GraphPad).

The analytical validation study cohort, which consisted of 572 men and 428 women, contained 172 HLA-B*27–positive and 828 HLA-B*27– negative patients. Successful genotypes were generated for 99.9% of samples compared with using the HLA-B locus test. When the results of each assay were compared with the HLA-B locus test, each tag-SNP assay demonstrated an analytical sensitivity, specificity, and accuracy of 97.6%, 99.9%, and 99.6%, respectively, with a false-negative rate of 2.4% and a false-positive rate of 0.1% (Table 1). The HLA-B*27 SNP, rs4349859 and rs116488202, tag the major white AS-associated suballeles with 98% sensitivity and 99% specificity^3,4,5,6, findings consistent with the results of our study. Given that the rs116488202 assay outperformed the rs4349859 assay (i.e., more discrete separation of genotype clusters; data not shown), the former was selected for clinical validation and implementation. That the HLA-B*27 rs116488202 tag-SNP assay demonstrated an analytical sensitivity, specificity, accuracy, and precision of 97.6%, 99.9%, 99.6%, and 100%, respectively, strongly suggests that this tag-SNP assay represents a very good genetic screening marker of HLA-B*27 positivity.

Replacing the HLA-B locus kit with the HLA-B*27 tag-SNP assay represents about a 14-fold cost reduction for healthcare authorities ($64.91 for the locus kit vs $4.61 for the tag-SNP assay, based on 20 samples; Table 2; all costs in Canadian dollars). For a clinical laboratory receiving about 500 HLA-B*27 test requests annually, replacing the HLA-B locus test with a HLA-B*27 tag-SNP test would equate to about $30,150 in cost savings.

HLA-B*27 status is of considerable interest to the healthcare system as part of a first-tier screening test for individuals with inflammatory back pain because it is an excellent predictor of future axSpA disease outcome. SNP genotyping to determine HLA-B*27 status in the Newfoundland (NL), Canada, population, which is primarily white, represents a highly sensitive, specific, accurate, precise, cost-effective, and time-efficient clinical screening method for patients with axSpA. Although genetic results from the NL population are generalizable to others⁷, uncertainty exists regarding how well the tag-SNP assay will perform in more ethnically diverse populations given that the rs116488202 SNP displays the most robust association with the HLA-B*27 allele in whites.

To our knowledge, ours is the first study to assess the analytical validity and economic effect of using a HLA-B*27 tag-SNP assay compared with a HLA-B locus test in a clinical genetics laboratory. That our study used 1000 consecutive specimens that arrived at the clinical genetics laboratory for HLA-B*27 genotyping provides an accurate assessment of the analytical validity and economic effect of using a HLA-B*27 tag-SNP assay. Based on our findings, the HLA-B*27 tag-SNP assay will be used clinically as a first-tier screening test for patients with inflammatory axial pain in the NL population. The ordering of the HLA-B locus assay will be limited to rheumatologists for those patients with a negative tag-SNP test result where a strong clinical suspicion remains that the patient might be HLA-B*27– positive. Adoption of such a cost-effective screening approach is recommended to other healthcare authorities striving to offer equal services at reduced costs.