Something new with Hashimoto’s encephalopathy

TOPIC: Hashimoto's encephalopathy

Title: Anti-thyroperoxidase antibodies from patients with Hashimoto's encephalopathy bind to cerebellar astrocytes.

Authors: Blanchin S, Coffin C, Viader F, Ruf J, Carayon P et al.

Reference: Journal of Neuroimmunology 192: 13-20, 2007



Hashimoto's encephalopathy is a rare and controversial neurological disorder, which is responsive to treatment with corticosteroids and associated with autoimmune thyroid disease. It is currently unclear whether this condition is a direct association between thyroid autoimmunity and a presumptive autoimmune process within the central nervous system, an indirect association or an artefact arising from the frequent occurrence of thyroid autoimmunity in health individuals.


In order to test the first of these possibilities, the authors in the present study have attempted to assess the reactivity of thyroid peroxidase autoantibodies from patients with Hashimoto's encephalopathy with central nervous system tissue.


The authors obtained sera in cerebrospinal fluid (CSF) from 10 patients with presumptive Hashimoto's encephalopathy. Control sera were obtained from patients admitted with other neurological conditions, including meningitis, intractable headache, peripheral neuropathy and multiple sclerosis. Two of these coincidentally had thyroid peroxidase antibodies. Further sera were obtained from patients with straightforward Hashimoto's thyroiditis without encephalopathy. These sera were characterised in detail, and in particular TPO antibodies were sought in the CSF. The main experiments concerned the binding of TPO antibodies to primate central nervous system tissues. Cerebrum, cerebellum and thyroid from rhesus monkeys, as well as cultured human astrocytes were used in standard immunofluorescence assays.


The clinical characteristics of the patients were typical of those with this syndrome. Manifestations included impaired consciousness and a series of other neurological complications but no dementia or behavioural disturbance. EEG analysis on the other hand was abnormal in only 3 of the 10 patients. The patients were variably responsive to steroids. All patients had high levels of TPO antibodies (>50IU/ml) and all those analysed had TPO antibodies in CSF. TPO antibodies declined in most but not all patients after corticosteroid treatment. Indirect immunofluorescence assays were performed with all of the encephalopathy samples, 12 of the Hashimoto's thyroiditis samples and 4 healthy donor samples. The 10 sera from the encephalopathy patients but none of the Hashimoto's thyroiditis sera or the healthy sera bound to structures in primate cerebellar tissues. After depletion of the sera on purified human TPO by affinity chromatography, the serum did not bind significantly to either thyroid cells or cerebellar cells, excluding a role for thyroglobulin antibodies in the binding. Four monoclonal antibodies against TPO which were tested bound to primate cerebellar cells. No sera bound to cerebrum. Further double immunofluorescence assays revealed that the sera appeared to bind to normal human primary astrocyte cultures (GFAP+).


These results indicate that TPO autoreactivity may have a specific role in Hashimoto's encephalopathy.


The association between Hashimoto's thyroiditis and a specific form of encephalopathy has been hotly disputed, and this paper may have been overlooked by those who concentrate on the endocrine literature. Although there is circumstantial evidence to support a role for thyroid autoimmunity in encephalopathy, detailed reviews of patient cohorts have failed to establish that this is a cause and effect relationship. The present results provide some convincing evidence that TPO antibodies bind to central nervous system structures. The depletion experiments are particularly important in this regard, although of course it remains possible that another cross-reactive autoantibody species, directed against a second autoantigen besides TPO, could be responsible for the encephalopathy. In this regard it is noteworthy that the patients with Hashimoto's thyroiditis, who had no clinical evidence of encephalopathy, had TPO antibodies which did not bind to cerebellar structures. This clearly suggests that there is an epitopic difference in TPO autoreactivity which determines susceptibility to encephalopathy, although of course the numbers sampled so far are small.

There are parallels to the hypothesis being put forward in this paper. Several neurological disorders such as stiff man syndrome and Lambert-Eaton myasthenic syndrome have been strongly associated with autoantibodies against glutamic acid decarboxylase and calcium channel autoantibodies. Simple experiments could easily be attempted using the autoantibodies from patients with encephalopathy to see if these can induce encephalopathic changes in animal models, and of course it would be fascinating to compare the epitopic recognition of TPO by antibodies from patients with encephalopathy to compare these with straightforward thyroiditis. In addition it would be important to assess the activity of these sera against human rather than primate cerebellum and to understand why encephalopathy can arise from reactivity to only the cerebellar part of the central nervous system. There is clearly much more to be done in this rare but intriguing and clinically important condition.

Summary and commentary prepared by Anthony Weetman (Related to Chapter 8 of TDM)