Stassi et al.1 have proposed that in thyroid autoimmunity thyrocyte survival depends on the differential effect of TH1 and TH2 cytokines. Their conclusion is based on the demonstration that the TH2 cytokines IL-4 and IL-10, which are abundant in glands from patients affected by Graves' disease (GD) but not Hashimoto's thyroiditis (HT), are able to up-regulate expression of FLIP and of the anti-apoptotic gene that encodes Bcl-xl on thyrocytes. In situ, both FLIP and Bcl-xl are up-regulated in the thyroids of patients with GD but not HT. The up-regulation of these proteins would enable GD thyrocytes to become resistant to Fas-mediated apoptosis. The TH1 cytokine IFN-γ, which is detected in HT thyroids, does not exert an effect on Bcl-xl expression on HT thyrocytes. However, it can stimulate expression of caspase-8 and caspase-3 on these cells, which leads to apoptosis.

There are two points we wish to address: both relate to our difficulty in reconciling these results1 with previous findings in the literature. First, in previous reports, no differential pattern of TH1 and TH2 cytokine expression has been shown in different thyroid autoimmune glands2,3,4,5. Although it has been suggested that a functionally, rather than morphologically, dominant subset of T cells is critical for the pathogenesis of either destructive or stimulating thyroid autoimmunity6, clear documentation of a restricted TH1- or TH2-dependent profile has in general, remained elusive.

In particular, we have studied the expression of IL-10 in various thyroid conditions by reverse-transcribed polymerase chain reaction (RT-PCR), in situ hybridization and immunohistochemistry5 and found the following. (i) By RT-PCR high amounts of IL-10 were seen in 3/6 GD samples and in 1/2 HT samples. In 3/6 GD samples the amount of IL-10 found was similar to that found in the controls. (ii) By in situ hybridization IL-10 expression was observed within the most intense infiltrates (not within the thyrocytes) not only of GD but also of HT glands. (iii) High amounts of IL-10 expression were associated with earlier stages of disease rather than with the severity of the lymphocytic infiltration. (iv) The highest amounts of IL-10 were present in 5/6 glands from autoimmune patients with the highest titers of circulating TPO antibodies, thus suggesting a role for this cytokine in B cell stimulation rather than in protecting against autoimmunity.

The second issue concerns the detection of “normal” amounts of Bcl-xl expression on thyrocytes from thyroids affected by HT. Because it has been found that expression of another anti-apoptotic gene, which encodes Bcl-2, is decreased on thyrocytes taken from patients with this autoimmune-destructive condition7,8 and Bcl-xl is more easily induced than Bcl-2, at least on lymphocytes9, we would have expected Stassi et al. to give an interpretation of their data in the light of these reports.

As autoimmune thyroid disease represents a broad spectrum of conditions with an extensive clinical overlap it is critical that one remains cautious when, from the results obtained in a small cohort of patients, hypotheses are formulated.

See Response to 'TH1 and TH2 cytokine control of thyrocyte survival in thyroid autoimmunity' by Giorgio Stassi and the Control of target cell survival in thyroid autoimmunity by T helper cytokines via regulation of apoptotic proteins by Giorgio Stassi