For certain we know that the autoantibodies can stimulate the thyroid and cause overactivity in Graves' disease, and can in select circumstances inhibit thyroid function and cause hypothyroidism in neonates and some adults. Whether anti-TG or anti-McAg/TPO antibodies are primary cytotoxic agents in AITD remains an unsettled issue. TG antibodies are probably not normally cytotoxic, but TPO antibodies can certainly mediate complement dependent thyroid cell cytotoxicity and ADCC. However, the frequently reproduced natural experiment of transplacental antibody passage from a mother with AITD to her fetus, without evidence of thyroid damage, clearly shows that antibodies alone are not destructive to the thyroid.

Cell-mediated immunity is thought to be important in thyroid cell destruction, and T cells have been shown to be reactive to TECs. T cell lines or clones have been shown to react to TECs (140 - 142), but the nature of the antigen recognized is unknown. One CD8 ⁺T cell clone in man has been shown to be cytotoxic specifically to autologous TECs (152), suggesting that cell-mediated TEC destruction is an important process, and similar activity has been reported in CD8 ⁺T cell lines and clones derived from mice with experimental autoimmune thyroiditis (207). A second type of T cell-mediated cytotoxicity is that mediated by gd TCR-bearing T cells and specific recognition of TECs by such cells has been reported in Graves’ disease, but the exact autoantigen involved is unknown (208). In animals it is clearly shown that there can be a marked dissociation between the extent of histologic thyroiditis and the levels of antibodies, again suggesting that T cells rather than antibodies mediate cell destruction. However, it must be admitted that the hard evidence for direct T cell-mediated cytotoxicity in thyroid autoimmunity in man is meagre at present.

There are 3 mechanisms by which T cells might mediate TEC destruction and evidence for all 3 operating in AITD has accrued. Firstly, cell lysis might be effected via T cell-derived perforin, which leads to pore formation in target cell surfaces, and certainly the thyroid lymphocytic infiltrate contains perforin-expressing T cells in AITD (209). Secondly, T cells expressing Fas ligand, especially the CD8 ⁺subset, can induce apoptosis in TECs expressing Fas (210). Fas is induced by IL-1b on TECs, whereas TSH-R stimulation inhibits Fas expression (211), and this may lead to the involvement of this pathway in Hashimoto’s thyroiditis but not Graves’ disease, as TSI would act like TSH in the latter to diminish Fas expression (and other regulatory molecules) (209). It has been suggested that T cells may not be necessary, as Hashimoto TEC may express Fas ligand, and autocrine/paracrine interaction with Fas may lead to TEC death (212). The mechanisms for this are unclear and as yet there is no consensus on the role this may have in AITD. The picture is complicated by the upregulation of molecules which protect against apoptosis such as Bcl-2. The pattern of expression of this molecule is different in Graves’ and Hashimoto’s diseases, suggesting that TECs are protected in the former and more sensitive to destruction in the latter (213). Whether these differences depend on cytokines, genetics or other factors is at present unknown (214). Finally, T cell-derived cytokines can injure the TECs directly, leading to functional impairment (164, 165), and by triggering other phlogistic pathways such as nitric oxide synthesis (215).