Multivariate analysis of HLA loci in conjunction with a thyrotropin receptor codon 52 polymorphism in conferring risk of Graves' disease

Recent reports have suggested that the HLA alleles DRB3*0101 or DQA1*0501 confer greater susceptibility to Graves' disease than does the DR3 allele. We have reported previously that a non-HLA-linked allele, a polymorphism in codon 52 of the human thyrotropin receptor gene, is highly associated with Graves' disease in females. To determine which of these four susceptibility alleles confers greater independent risk for the development of Graves' disease, we analyzed the alleles in 134 North American Caucasian females who have Graves' disease (n = 69) or are normal controls (n = 65) in a logistic regression model. While we found each of these alleles to be associated with Graves' disease when analyzed independently (corrected p < 0.01 for each of 4 alleles tested), only DR3 (p = 0.0001) and the thyrotropin receptor polymorphism (p = 0.0060) maintained a statistically significant independent association when assessed in conjunction with each of the other alleles in a logistic model. We conclude that DR3 confers the greatest susceptibility to Graves' disease (odds ratio = 7.6) of the alleles within the HLA locus, and that any association between DRB3*0101 or DQAl*0501 and Graves' disease may be a result of the tight linkage disequilibrium between these alleles and DR3. In addition, we found the non-HLA-linked thyrotropin receptor codon 52 polymorphism to confer significant independent risk of Graves' disease (odds ratio = 9.0). Further, because 6 of 6 individuals who possessed both DR3 and the thyrotropin receptor polymorphism had Graves' disease, while no individual in the normal control group possessed both alleles, study of a larger population to assess the potential synergism between these 2 alleles is warranted.