We investigated human leukocyte antigen (HLA) class I and class II antigens in 56 Japanese patients with subacute thyroiditis (SAT) who visited our out-patient clinic between 1988 and 1990. We found SAT to be associated with not only HLA-B35 (40 patients; P < 0.000001; relative risk, 18.02), but also with HLA-B67 antigens (9 patients; P < 0.00001; relative risk, 11.20). No heterozygotes of HLA-B35 or HLA-B67 were found in any of the 56 patients with SAT. Either HLA-B35 or HLA-B67 antigen is found in 87% of patients with SAT. When season of onset and clinical course of SAT were compared in the 49 patients with HLA-B35-positive SAT (B35-SAT) and HLA-B67-positive SAT (B67-SAT), we were able to identify certain characteristics: 1) B67-SAT often followed the course from transient thyrotoxicosis to a hypothyroid phase to a euthyroid phase [6 of 9 B67-SAT (67%) vs. 10 of 40 B35-SAT (25%); P < 0.05]; and 2) B67-SAT occurred mostly during the summer or autumn and at a higher rate than did B35-SAR [8 of 9 B67-SAT (89%) vs. 17 of 40 B35-SAT (43%)], whereas B35-SAT occurred throughout the year. We conclude that there are at least two types of SAT that can be classified by association with either HLA-B35 or HLA-B67 antigens.

TRH tests were performed in 206 clinically and biochemically euthyroid relatives of patients with Graves' disease. In 117 of the 206, T3 suppression tests were performed. Results revealed that 56 of the 206 (27.1%) showed abnormal responses to TRH. Twenty-nine of these (14.1%) revealed absent or decreased responses, and 27 (13.1%) revealed augmented responses to TRH. Eight of the 117 (6.8%) were T3 nonsuppressible. These eight subjects consisted of 4 subjects out of 17 hyperesponders and 4 subjects out of 90 normal responders. The majority of suppressible subjects (86 among 109) demonstrated normal responses to TRH. Sixty-nine of the 206 subjects were followed for 6 months to 5 yr to observe changes in their thyroid functions. Among all 69 subjects 3 became clinically thyrotoxic 12, 12, and 18 months after their initial visit, respectively, and 2 became clinically hypothyroid 2 yr after their initial visit. Since 69 subjects were clinically and biochemically euthyroid and had no goiter or exophthalmos at their initial visit, the incidence of thyrotoxicosis or hypothyroidism in these subjects could be considered to be remarkably high. It is of interest that the 3 thyrotoxic patients were TRH hyporesponders at their first visit. One patient was T3 suppressible; T3 suppression tests were not performed in the other 2 patients at their initial visit. There was no abnormality in the first TRH test in 2 relatives who became hypothyroid. It is suggested that 1) among euthyroid relatives with a family history of Graves' disease, there are many with abnormalities in TRH responsiveness and T3 suppressibility, 2) nonsuppressible subjects are more likely to be TRH hyporesponders and vice versa, 3) hyperthyroidism or hypothyroidism occurs frequently in euthyroid relatives with a family history of Graves' disease, and 4) thyrotoxicosis occurs frequently in TRH-hyporesponders, and hypothyroidism occurs in the others.