ANTIBODIES PRODUCING COMPLEMENT-MEDIATED THYROID CYTOTOXICITY IN PATIENTS WITH ATROPHIC OR GOITROUS AUTOIMMUNE-THYROIDITIS

Thyroid-cytotoxic antibodies (thyroid-cytotoxic Abs) have been described in patients with autoimmune thyroiditis, but their role in the development of hypothyroidism remains to be clarified. In this study, we evaluated the pathogenetic role of thyroid-cytotoxic Abs in 20 patients with atrophic thyroiditis (idiopathic myxedema; AT) and 94 patients with goitrous Hashimoto's thyroiditis (HT). Among patients with HT, 27 were euthyroid (HT-E), 27 had subclinical hypothyroidism (HT-SH), and 40 had overt hypothyroidism (HT-H). Seventeen normal subjects and 8 patients with nonthyroidal illnesses were used as controls (C). To detect thyroid-cytotoxic Abs, human thyroid cells expressing thyroid peroxidase (TPO) were labeled with Cr-51,and challenged with the immunoglobulin G (IgG) fraction of serum plus rabbit complement. The cytotoxic effect of IgGs was calculated as the percent specific lysis (% SL), taking into account the lytic effect of complement alone and the maximal lysis produced by a detergent. Most C-IgGs decreased the cytotoxic effect of complement (median % SL, -3.3). IgGs from hypothyroid patients with thyroiditis had a greater cytotoxic effect than C-IgGs, either as a whole group (P < 0.001) or when subdivided according to clinical diagnosis: HT-SH (median % SL, 4.8; P < 0.005), HT-H (%SL, 2.2; P < 0.0001), or AT (%SL, 0.9; P < 0.01). Among patients with HT, the lytic activity of IgGs from patients with subclinical and overt hypothyroidism was higher than that of IgGs from euthyroid patients (P < 0.05). The results of IgGs from euthyroid patients with HT (median % SL, -0.9) did not significantly differ from those of C-IgGs. By taking a cut-off over the upper range of % SL produced by C-IgGs (>2), the prevalence of thyroid-cytotoxic Abs was 30% in AT, 59% in HT-SH, and 55% in HT-H. However, 37% of euthyroid patients with HT also had thyroid-cytotoxic Abs. No IgG containing TPO antibodies (TPOAb) at low titer (<40(2)) was cytotoxic. However, the levels of thyroid-cytotoxic Abs did not correlate with TPOAb titers, and preabsorption with TPO only partially abolished the lytic effect of some HT-IgG. These findings suggest that TPO is a target of thyroid-cytotoxic Abs, but other thyroid antigens are also involved in the cytotoxic reaction. In conclusion, 1) we developed a radiometric method to detect antibodies producing thyroid-specific complement-mediated cytotoxicity; 2) thyroid-cytotoxic Abs were detected with a greater prevalence and at higher levels in patients with Hashimoto's thyroiditis who were hypothyroid; however, some euthyroid patients also had cytotoxic antibodies; and 3) thyroid-cytotoxic Abs appear to contribute as a second line mechanism to thyroid destruction, because they probably require preceding disruptive event of follicular structure to exert their action.