Session Type: Abstract Submissions (ACR)
Background/Purpose : The purpose of this study is to examine associations of HLA class I and class II alleles with AS in different patients populations of whites of European ancestry (EA) and African-American (AA) ethnicities, as well as HLA-B locus associations in Han Chinese (HC).
Methods: HLA-A, -B, -C, -DRB1, -DQB1 and -DPB1 alleles were examined by DNA typing in unrelated patients from the Prospective Study of Outcomes in Ankylosing Spondylitis cohort, the North American Spondylitis Consortium and Australo-Anglo-American Spondyloarthritis Consortium. For the HLA-B locus analyses, an additional 578 British and Australian AS patients from the Australo-Anglo-American Spondylitis Consortium and 360 HC AS patients were also analyzed. Included therefore in the study were 1829 EA, 62 AA patients and 360 HC who met modified New York criteria for AS. Controls were North American white and African American as well as Han Chinese subjects without history of rheumatic disease. To remove an associated effect of HLA-B*27 due to linkage disequilibrium, analyses were also conducted on HLA-B*27 non-carriers only. Statistical analysis was done by construction of 2×2 tables and testing the proportion of alleles in cases vs. controls with Fisher’s exact test. Other analyses included permutation-based omnibus testing and “relative predispositional effects” (RPE) analysis
Results: HLA-B27 occurred in 88.7% of EA, 61% of AA, and 93% of HC patients compared to 7.5%, 2% and 7.6%, respectively, of ethnicity matched controls. HLA-B*07 was negatively associated with AS in all three ethnic groups (6.2 % versus 14.9% in EA, p 3.655 x10-26 , 3.2% versus 14% in AA(p=0.02), and 1% versus 12% in HC (p=0.0018). Among EA AS patients, HLA-B27 noncarriers showed positive associations with HLA-B*38 (OR=2.94, p =0.0008) and DRB1*04:04 (OR= 3.02 p = 0.0065) and negative associations with HLA-B*07 and HLA-DRB1*03, HLA-DRB1*15:01 and their respective linked alleles DQB1*02:01 and DQB1*06. Additional associations with HLA-B*14(OR = 1.74, p < 0.001) and HLA-B*40 (OR = 1.32, p = 0.02) were observed via RPE analysis, which excludes the HLA-B*27 alleles. No associations were seen with HLA-DPB1 alleles or with HLA-A*02 (the latter seen in a much larger study where HLA alleles were imputed but not directly genotyped). Among AA patients, positive associations were seen in HLA-B*27 (OR = 75.11, p <0.0001), HLA-B*40 (OR = 8.33, p = 0.01) and HLA-DRB1*13:02 (OR = 2.43, p = 0.02). No statistically significant associations were seen in HLA-DQB1, HLA-DPB1 alleles. In the HC, no association was seen with B*40:01 (B60) although an association was seen by a covariate via logistic regression analysis (p=0.02, OR=2.3) and by RPE analysis (p=0.01, OR=1.56). Negative associations were also demonstrated with HLA-B*13, B*15, B*46 and B*51.
Conclusion: This is the largest directly genotyped HLA study to date. These data, analyzing the largest number of AS patients in three patient populations examined to date, suggest other HLA alleles to be operative in AS predisposition in addition to HLA-B*27. The shared association of certain alleles in all three groups suggests a direct role in AS pathogenesis
M. H. Weisman,
M. M. Ward,
L. S. Gensler,
M. A. Brown,
J. D. Reveille,
« Back to 2014 ACR/ARHP Annual Meeting
ACR Meeting Abstracts - https://acrabstracts.org/abstract/shared-hla-class-i-and-ii-alleles-in-susceptibility-to-ankylosing-spondylitis-among-three-ethnic-groups/