PubMed1992
OBJECTIVES: To re-evaluate the associations of HLA types with Japanese patients having hyperthyroid Graves' disease, HLA types and clinical findings were correlated. DESIGN: Four independent clinical findings (ophthalmopathy, family history, age at onset and size of goitre) and two autoantibody titres, thyrotrophin binding inhibitor immunoglobulin (TBII) and anti-thyroid microsmall antibody (anti-M), were analysed. PATIENTS: Eighty-eight Japanese patients with hyperthyroid Graves' disease and 186 control subjects were assessed. MEASUREMENT: Serological HLA typing was performed on 73 antigens in HLA-A, -B, -C, -DR and -DQ loci. HLA-D and -DP (29 antigens) were determined by the restricted fragment length polymorphism (RFLP) methods. TBII and anti-M were measured by commercially available kits. RESULTS: Patients with potent antibody titres had HLA antigens commonly seen among all the patients with Graves' disease. Interestingly, however, HLA-B35 and -Cw11 were found to relate with negative and/or weak TBII, and HLA-B7 and absence of HLA-Aw19 with negative anti-M. Significant associations were observed between HLA-DRw8 and large goitre and absence of ophthalmopathy, and between HLA-DQw4 and a negative family history of diffuse goitre (corrected P less than 0.05). Several other antigens were also found to be significant. Among these antigens, four pairs of MHC classes I and II were found to relate to the clinical findings independently. HLA-DQw4 and negative -A31 pair was closely related to ophthalmopathy, negative family history and late onset of disease. The HLA-B5 and -Dw12 pair was associated with ophthalmopathy, positive family history and early onset of disease. The HLA-A11 and negative -DPw2 pair was associated with ophthalmopathy, negative family history and early onset of disease. The HLA-Bw46 and -DRw8 pair did not increase in frequency above that seen with HLA-DRw8 alone. These four antigen groups (HLA-DRw8, HLA-DQw4 and negative-A31, HLA-B5 and -Dw12, and HLA-A11 and negative -DPw2) were observed in the majority (68%) of patients with Graves' disease and at a significantly higher incidence than in the control group (P less than 0.05). CONCLUSION: There are four subpopulations of Japanese patients with hyperthyroid Graves' disease. This is one of the reasons why the association of HLA types in Japanese patients is rather weak when they are studied as one group.