Graves' disease (GD) and Hashimoto's thyroiditis (HT) represent the commonest forms of autoimmune thyroid disease (AITD) each presenting with distinct clinical features. Progress has been made in determining association of HLA class II DRB1, DQB1 and DQA1 loci with GD demonstrating a predisposing effect for DR3 (DRB1*03-DQB1*02-DQA1*05) and a protective effect for DR7 (DRB1*07-DQB1*02-DQA1*02). Small data sets have hindered progress in determining HLA class II associations with HT. The aim of this study was to investigate DRB1-DQB1-DQA1 in the largest UK Caucasian HT case control cohort to date comprising 640 HT patients and 621 controls. A strong association between HT and DR4 (DRB1*04-DQB1*03-DQA1*03) was detected (P ¼ 6.79 Â 10 À7 , OR ¼ 1.98 (95% CI ¼ 1.51-2.59)); however, only borderline association of DR3 was found (P ¼ 0.050). Protective effects were also detected for DR13 (DRB1*13-DQB1*06-DQA1*01) (P ¼ 0.001, OR ¼ 0.61 (95% CI ¼ 0.45-0.83)) and DR7 (P ¼ 0.013, OR ¼ 0.70 (95% CI ¼ 0.53-0.93)). Analysis of our unique cohort of subjects with well characterized AITD has demonstrated clear differences in association within the HLA class II region between HT and GD. Although HT and GD share a number of common genetic markers this study supports the suggestion that differences in HLA class II genotype may, in part, contribute to the different immunopathological processes and clinical presentation of these related diseases.

Summary The identification of genes placing individuals at an increased risk for the development of autoimmune thyroid disease (AITD) has been a slow process. However, over the last 20 years or so real progress has been made with the mapping of novel loci, via a number of different approaches. First, through the use of traditional immunological methods, Human Leucocyte Antigen (HLA)/Major Histocompatibility Complex (MHC) was the first gene region to be associated with AITD and consistent replications have been reported. Second, the CTLA‐4 gene region on 2q33 was the first non‐MHC replicated locus to be primarily identified using the candidate gene method. Third, family‐based linkage studies led to the mapping of a new type 1 diabetes locus, the PTPN22 gene, which has subsequently been independently replicated as a susceptibility gene for Graves’ disease (GD). Fourth, despite many unsuccessful attempts at implicating the TSHR gene as a susceptibility locus for GD, a recent approach of ‘tagging’ all the common variation within the gene has led to its identification as the first GD specific locus. Moreover, the use of tag single nucleotide polymorphisms (SNPs) has also been used to implicate the recently identified type 1 diabetes locus, CD25 as a susceptibility gene for GD. Finally, large scale, ongoing genome‐wide association studies in multiple autoimmune diseases (AID) states, including AITD seem likely to lead to the identification of additional MHC and non‐MHC susceptibility loci.

No abstract