Article Text

Download PDFPDF

Neuropathology of Hashimoto’s encephalopathy
  1. W Paulus1,
  2. K W Nolte1
  1. 1Institute of Neuropathology, University Hospital, Domagkstr 19, D-49149 Münster, Germany
  1. Correspondence to:
 Professor W Paulus;
  1. C P Doherty2,
  2. R Folkerth3
  1. 2Department of Neurology, Massachusetts General Hospital, Boston, Massachusetts, USA
  2. 3Brigham and Women’s Hospital, 75 Francis Street, Boston, MA 02115, USA

    Statistics from

    Doherty et al have recently reported in this journal on a case of Hashimoto’s encephalopathy clinically mimicking Creutzfeldt–Jakob disease.1 As neuropathological analysis of a brain biopsy in this case revealed spongiform change and sparse mononuclear infiltrates in perivascular spaces, the authors suggested that Hashimoto’s encephalopathy is an encephalitic process. We have previously described a necropsy case of Hashimoto’s encephalopathy showing lymphocytic vasculitis of veins and venules of the brain stem, supporting the hypothesis that vasculitis represents the morphological substrate of Hashimoto’s encephalopathy.2 Doherty et al have challenged our diagnosis and stated that the term vasculitis should be reserved for lesions with inflammation and fibrinoid necrosis of arterial vessels.

    Lymphocytic vasculitis is a generally accepted pathological subtype covered by reviews and standard textbooks.3,4 It is characterised by the presence of lymphocytes within the vessel wall and is encountered in Wegener’s disease, systemic lupus erythematosus, and Behcet’s disease, to name but a few conditions.4 That the diagnosis is more difficult than for necrotising arteritis does not, however, imply that lymphocytic vasculitis does not exist, and our case certainly belongs in this vasculitis category.

    Additional support for the vasculitic basis of Hashimoto’s encephalopathy comes from other reports on this condition where there has been angiographic demonstration of vasculitis.5 More recent experimental evidence for a vasculitic pathogenesis involves the identification of α-enolase as an autoantigen in Hashimoto’s encephalopathy,6 because anti-α-enolase antibodies are present in various autoimmune vasculitic diseases such as systemic lupus erythematosus and ANCA associated vasculitis, and because α-enolase is highly expressed in the endothelium.

    Based on the histological figure and the description provided by Doherty et al,1 we express some reservations over the presence of “spongiform” changes. True spongiform changes, usually encountered in prion diseases, are present in the neuropil between nerve cell bodies, whereas the perineuronal spaces illustrated by Doherty et al more closely resemble shrinkage artefacts which may be particularly pronounced in small biopsy specimens. In addition, we would like to point out that vasculitis is typically focal and may have been missed in this case.

    In conclusion, we feel that the available clinical, pathological, and experimental evidence suggests that vasculitis underlies a substantial proportion of cases of Hashimoto’s encephalopathy. However, additional careful necropsy studies are required to determine whether other pathologies contribute to the clinical picture.


    Authors’ reply

    We thank Drs Paulus and Nolte for their interest in our case report, which we offered as a lesson in the clinical, and to a certain extent pathological, similarities between Creutzfeldt–Jakob disease (CJD) and Hashimoto’s encephalopathy on small biopsies. Of course, in established CJD—as the correspondents rightly mention—spongiform change is easily recognised by vacuoles in the neuropil; in contrast to their contention, these are regularly accompanied by neuronal cytoplasmic vacuoles displacing surrounding neuropil (similar to those we illustrated), as described in standard textbooks1 and in clinicopathological cases.2 The difficult judgment is in the brain biopsy in early CJD, in which these changes are minimal and indeed nearly impossible to distinguish from artefact or oedema. The concurrence of astrogliosis with the spongiform appearance (no matter what the cause) in our case further lowered our threshold for submitting the tissue for definitive prion protein studies in the clinical setting of rapidly evolving cognitive decline. We hope the correspondents would agree that such would be simply good clinical practice, especially in a patient like ours without a helpful history of Hashimoto’s thyroiditis or anti-thyroid peroxidase antibodies.

    As to their assertion that we “challenged” their diagnosis of vasculitis (which we point out was followed by a question mark in the title of their own report!), we felt that the previous subarachnoid haemorrhage and apparently diffuse lymphocytic infiltration of the leptomeninges that they described suggested that the venular lymphocytic infiltrates could be secondary to the haemorrhage, or be a manifestation of a meningoencephalitis (possibly autoimmune) rather than a true primary vasculitis.3 According to the textbook they cite, the entities of Wegener’s disease, systemic lupus erythematosus, and Behcet’s disease, to use their examples, are all characterised primarily by true arteritis with necrosis, not by the lymphocytic venulitis which may accompany it.4 Drs Paulus and Nolte are correct that in our small biopsy, vasculitis may have been missed. They may also be correct that the pathogenesis of Hashimoto’s encephalopathy involves vasculitis. We stand firm in our position, however, that to make a histological diagnosis of vasculitis requires the exclusion of other factors, such as haemorrhage or tissue inflammation, in which lymphocytic diapedesis through venules is physiological. Moreover, the novel finding in our case of parenchymal inflammation does raise the prospect of an encephalitic process in Hashimoto’s encephalopathy, independent of any effect on blood vessels, whether primary or secondary.

    We thank Drs Paulus and Nolte for pointing out the interesting paper by Ochi et al,5 and look forward to future elucidation of the role of α-enolase in Hashimoto’s encephalopathy. Perhaps those of us with tissue from cases of Hashimoto’s encephalopathy will need to collaborate in that regard, to expand their studies on serum from patients with this condition.

    Finally, we want to reiterate the importance of a broad differential diagnosis in evaluating patients with rapid decline in cognition, and with vacuoles, astrogliosis, microgliosis, and parenchymal perivascular mononuclear cells on brain biopsy. That was the overarching point of our paper.


    Request Permissions

    If you wish to reuse any or all of this article please use the link below which will take you to the Copyright Clearance Center’s RightsLink service. You will be able to get a quick price and instant permission to reuse the content in many different ways.