While Hashimoto’s thyroiditis (HT) is well understood and a cornerstone for understanding autoimmune thyroid disease, Hashimoto’s encephalopathy (HE) remains a challenge to clinicians not so much for its effective management, but because of the potential for unraveling an important aspects of autoimmune encephalitis, and related childhood and adult cognitive and neuropsychiatric disorders.
Epidemiology: Hashimoto encephalopathy (HE) is a rare disorder and there are no epidemiologic studies of its population incidence or prevalence. The diagnosis, as originally described by Brain and colleagues in 1966 rested on the presence of Hashimoto’s thyroiditis (HT) and a high serum concentration of anti-thyroid antibodies, but then as now, it is unknown whether anti-thyroid antibodies or concomitant thyroid dysfunction contributes to the pathogenesis of HE. In 2002, Rowland and coworkers characterized the clinic-pathologic findings of all literature cases of HE, beginning with the patient described by Brain and colleagues through 2002, adding their own patient. The authors searched Medline database with the terms “Hashimoto”, “autoimmune thyroid disease” and “encephalopathy,” identifying 105 patients, 85 of whom, met inclusion criteria of encephalopathy (clouding of consciousness with reduced wakefulness, attention, or cognitive function); absence of CSF evidence of bacterial or viral infection and positive test result for high titers of circulating thyroid peroxidase (TPO) or anti-thyroglobulin (Tg) antibodies. Patients with stroke-like or postictal focal signs, seizures, dementia, or psychiatric symptoms were included if consciousness was depressed. Of four patients studied histologically, one patient examined at autopsy showed congestion of cerebral vessels without focal infarction on macroscopic examination, and another postmortem examined case showed lymphocytic infiltration of brainstem veins consistent with vasculitis. Two patients underwent brain biopsy also showing lymphocytic infiltration of the walls of arterioles and veins, or small perivascular cuffs of lymphocytic cells.
Clinical Presentation: In the series of Rowland and coworkers the mean age at onset of symptoms of HE was 44 years (range, 9-78 years); 19 of whom were boys or girls, age 18 years or younger. Among the adults, there were 53 women and 13 men. In addition to encephalopathy as required, stroke-like signs presented in 23 (27%) case, seizure in 56 (66%), myoclonus in 32 (38%), and visual hallucination or paranoid delusion in 31 (36%). The course was relapsing and remitting in 51 (60%) of cases.
Laboratory Investigation: In the series of Rowland and coworkers, both Tg and microsomal or TPO antibodies were together in 60 (71%) cases with one antibody of the two normal in 20 (24%) cases. There was no relationship between the neurologic symptoms and signs and the type or serum concentration of anti-thyroid antibodies. Altogether, 30 (35%) cases were sub-clinically hypothyroid, 19 (22%) were euthryoid, and 17 (20%) were overtly hypothyroid. Fourteen (16%) cases had an elevated ESR or ANA, and three had a concomitant connective tissue disease (psoriatic arthritis, Sjogren’s syndrome, and sarcoidosis). An elevated cerebrospinal fluid (CSF) protein level was noted in 66 (78%) patients, with abnormal findings in neuroimaging in 40/82 (49%) or electroencephalography (EEG) in 80/82 (98%) patients. A goiter was detected in 24/39 (62%) patients, with a serum microsomal antibodies in 55/58 (95%), TPO antibodies in 26/26 (100%), and Tg antibodies in 45/62 (73%) patients.
Immunopathogenic Mechanisms: Unlike the close relation between anti-thyroid antibodies and thyroiditis in HT, neither high titers of anti-thyroid antibodies nor presence of subclinical or overt hypothyroidism accounted for the observed encephalopathy in the series of Rowland and coworkers. The neurological findings in patients that were euthyroid were similar to those with subclinical or overt hypothyroidism. High serum anti-thyroid antibody titers have been associated with neuropathy and myopathy, bipolar disease and dementia. Their increased prevalence in healthy individuals and patients with subclinical hypothyroidism makes it difficult to consider a high titer a defining feature of any disorder other than autoimmune thyroid disease (AITD). Given the lack of a well-defined pathophysiologic link between anti-thyroid antibodies or HT and encephalopathy, HE may be a misleading term, the disorder being one of a larger group of autoimmune encephalopathies.
Ochi and colleagues provided a link between HT autoimmunity and the central nervous system (CNS). They developed a human brain proteome map using two-dimensional electrophoresis and applied it to the immunoscreening of brain proteins that reacted with serum anti-thyroid antibodies in HE patients, identifying the novel antigen -enolase (ENO1), encoded on 1p36.23, as a candidate autoantigen in HE, considering it a useful marker for HE-related pathology and corticosteroid sensitivity. Kishitani and coworkers extended the findings of Ochi and colleagues noting anti-NH2-terminal of a-enolase (NAE) antibodies in 24% of HE patient sera and limbic abnormalities on MRI demonstrating abnormal signal in unilateral or bilateral medial temporal lobes, and diffuse slow wave activity with epileptogenic discharges. These findings suggested that limbic encephalitis (LE) associated with anti-NH2-terminal of a-enolase antibodies could be a possible manifestation of HE in some cases. Graus and colleagues proposed that HE be regarded as a probable autoimmune encephalitis (AE) after excluding all other well-characterized syndromes of AE, and other syndromes accompanied by well-defined autoantibodies. The authors noted that -enolase antibodies, like anti-thyroid antibodies, were present in up to 68% of patients with HE and anti-thyroid antibodies, making them useful in the serological diagnosis of HE. However, they were of limited reliability as a biomarker of the brain disease since they were detectable in healthy individuals and in patients with other autoimmune disorders.
In support of an antibody-mediated autoimmune mechanism in HE, it is notable that patients with meningoencephalitis and corticosteroid-responsive encephalopathy may present with circulating anti-thyroid antibodies, rheumatoid factor, ANA, Sjogren’s (SSA, SSB), and cardiolipin antibodies like those with AITD and HE. In five such patients, one had high serum anti-thyroid antibody levels compatible with HE (anti-thyroid levels were not mentioned in the other four) and brain tissue biopsy revealed perivascular lymphocytic infiltrates without vessel wall invasion similar to the case described by Rowland and colleagues. At present, it is unclear whether anti-thyroid antibodies represent an immune epiphenomenon in a subset of patients with encephalopathy or are truly associated with pathogenic mechanisms of the disorder. In turn, the significance of classifying encephalopathies under the term HE will be determined in the future once the relevance of the role of anti-thyroid antibodies is demonstrated or dismissed by more detailed experimental and immunopathological studies.
A vasculitic etiopathogenesis of HE is likely in some cases based upon the tendency for systemic autoimmunity and vasculitis in cases of HT, and like those with AAV, the shared genetic predisposition to autoimmune disease in HT is in establishing and maintaining immunologic non-responsiveness or tolerance to self-antigens through interactions of the CTLA4 gene. According to Rowland and colleagues, one subgroup of patients with HE have a vasculitic clinical presentation of multiple stroke-like episodes. The available histopathology in HE also supports an inflammatory vasculopathy, so noted in one postmortem case that showed lymphocytic infiltration of brainstem veins in a case studied in brain tissue biopsy as a form of isolated angiitis of the CNS due to lymphocytic infiltration of the walls of arterioles and veins. Small perivascular cuffs of lymphocytic cells were noted in the case of Rowland and colleagues. It is noteworthy that patients with HE and circulating -enolase antibodies are at risk for heightened autoimmune activity, and the tendency for systemic and invasive autoimmune disorders including AAV.
Treatment: The significance of corticosteroid sensitivity in HE is widely accepted as a criterion for the diagnosis. However, as Rowland and colleagues suggested, it would be unwise to define any condition by response to any particular therapy especially if not replacing a specific deficit or directing it at a particular target. Patients with HE improve in association with, but not necessarily due to corticosteroid therapy. Moreover, those that respond to corticosteroids have no distinguishing clinical characteristics nor receive treatment in other fashions for a meaningful comparison.
1. Brain L, Jellinek EH, Ball K. Hashimoto’s disease and encephalopathy. Lancet 1966; 2:512-514.
2. Chong JY, Rowland LP, Utiger RD. Hashimoto encephalopathy. Syndrome or myth? Archives of Neurology 2003; 60:164-171.
3. Ochi H, Horiuchi I, Araki N, et al. Proteomic analysis of human brain identifies -enolase as a novel autoantigen in Hashimoto’s encephalopathy. FEBS Lett 2002; 528:197-202.
4. Kishitani T, Matsunaga A, Masamichi I, et al. Limbic encephalitis associated with anti-NH2-terminal of a-enolase antibodies. A clinical subtype of Hashimoto encephalopathy. Medicine 2017; 96:10:e6181.
5. Graus F, Titulaer MJ, Balu R, et al. A clinical approach to diagnosis of autoimmune encephalitis. Lancet Neurology 2016; 15:391-404.