GRAVES' DISEASE AND HLA: CLINICAL AND EPIDEMIOLOGIC ASSOCIATIONS

Correlation between clinical and epidemiological features of Graves' disease and HLA were sought in 175 patients, eighty-three of whom were typed for HLA DR antigens. The relative risks (x) conferred by HLA-B8 and DR3 were 3.1 and 5–7 respectively ( P <0.0005 uncorrected). The disease occurred at an earlier age in HLA-DR3 positive patients compared to negative patients ( P <0.005 in females> and <30-years-old). Eighty-four out of ninety-two patients could specify the season of onset of hyperthyroidism; an excess of HLA-B8 positive patients in the summer and a lack of these in the spring and autumn was found. No associations between B8 and exophthalmos and/or soft tissue eye changes were observed. However, significant associations between exophthalmos and either exophthalmos and/or soft tissue changes were found with DR3 (x=3.6 and 3.8 respectively). HLA-DR3 positive patients were found to be more resistant to radioiodine therapy than patients negative for these antigens. No heterogeneity in the distribution of HLA antigens was found when the following indices were examined: sex, goitre size, severity of disease, pretibial myxoedema, antecedent psychological disorder, consumption of oral contraceptives, family history of Graves' disease or thyroid antibody titre. HLA does not appear to distinguish subvariants of Graves' disease, rather it influences the susceptibility to disease and its persistence once it becomes clinically manifest.