Article Text

Guillain-Barré, Fisher, and Bickerstaff syndromes: nature versus well established ideas
  1. 415 Morris Street, Suite 401, Charleston, West Virginia 25301, USA
    1. Department of Neurology, Dokkyo University School of Medicine, Kitakobayashi 880, Mibu, Shimotsuga, Tochigi 321–0293, Japan
    1. Dr N Yuki yuki{at}

    Statistics from

    As the author of the first ever report documenting central nervous system involvement in Miller Fisher syndrome,1 I beg permission to discuss certain aspects of the work of Yukiet al on Guillain-Barré syndrome and Bickerstaff's brainstem encephalitis, an example of which was published in this Journal.2Twenty three years ago, I perused the literature in search of a diagnosis for a child I had seen in Iran. Eyeing incessantly a few certain guide posts in the neurological heavens (Babinski's response, internuclear ophthalmplegia, spasticity) I sailed the lore diligently, arriving at two inescapable conclusions: (a) that my patient with a midbrain lesion demonstrated by CT was an instance of Miller Fisher syndrome, hitherto a variant of Guillain-Barré syndrome; (b) that Bickerstaff's brainstem encephalitis and Miller Fisher syndrome are one and the same entity. With that perusal behind me, I thought it entirely sufficient to include Bickerstaff's work in my references and ended the article by the following statement: It is evident that our finding, if confirmed, raises questions regarding pathophysiology of Landry-Guillain-Barré syndrome, of which Fisher syndrome is a limited form. That pivotal observation has since been amply substantiated (see below for an incomplete list of references).

    In 1982 the article became the subject of an editorial in theArchives of Neurology in opposition to my position. It was acknowledged parenthetically by Bickerstaff himself who eventually embraced my second conclusion even though he sided with the splitters, considering the entity of a separate disease. The arrival of MRI turned the tide entirely in favour of the lumpers soon after 3, confirming my impression of a need for a reappraisal indicated above. Electrophysiological studies pointed to the same conclusion—that is, Miller Fisher syndrom often, if not always, is superimposed on a Guillain-Barré syndrome background sometimes so mild requiring complex techniques to disclose its presence.4 Frequent shuttling of Yukiet al between Guillain-Barré syndrome, Bickerstaff's brainstem encephalitis, and Miller Fisher syndrome when reporting “overlapping” hybrid cases means the same thing as well. Many others have made similar observations describing an ascending or descending march of the pathology along the brainstem, displaying corresponding neurological features.5 The report of Odakaet al concerning ballism6 in a case of Guillain-Barré syndrome/Miller Fisher syndrome is a novel observation, and another warning to those remaining sceptics who, while ignoring other guide posts, consider Miller Fisher syndrome strictly a peripheral nervous system disease. But when Yuki et al draw any distinction between Fisher and Bickerstaff syndromes they are displaying a lack of historical perspective and paucity of clinical perspicuity which I am rectifying here, as this is an important area of neurology where the stakes are high and simple logic must prevail. It is this blurring of traditional border between certain maladies affecting the peripheral and central nervous system that is the thrust of what has followed our ground breaking observation, including the work of Yuki et al (whatever its eventual immunological import may be).

    The role of MRI in the ensuing events which sometimes resembled an eponymous war deserves a comment. Whereas Ropper and others mistakenly relied on the absence of a lesion in conventional MRI to refute the role of a central nervous system lesion in Miller Fisher syndrome such instances are more cogently explained as follows: (a) the existence of a pathological difference between a signal producing inflammation and mere presence of a viral infection in the central nervous system; for example, in a recent report 30% of clinically proved enterovirus 71 related rhombencephalitides had negative conventional MRI7; (b) the subject of “normal appearing” white matter when in fact it is not normal may one day loom large here as it has in the case of multiple sclerosis.

    Thus our novel observation of 1979 alleviated Fisher's “certain reluctance to upset well established ideas concerning the disease (Guillian-Barré syndrome)” and removed the stigma of “oddity” and “aberration” from the syndrome described by the two luminaries, Miller Fisher and Bickerstaff, who did not know of each others' contributions; nor could they have known of the fact that they were describing the same clinical entity. And for two good reasons—that is, sharing the above mentioned reluctance to upset the well established ideas concerning Guillain-Barré syndrome (as admitted by Fisher himself) and preceding the wonderful age of computerised neuroimaging by a quarter of a century. It was left to the serendipitous observation in a 7 year old Persian girl, and its aftermath, for the facts to be gleaned—as depicted here.


    Yuki replies:

    Derakhshan wrote “when Yuki et aldraw any distinction between Fisher and Bickerstaff syndromes they are displaying a lack of historical perspective”, but we had described in detail the historical background in theintroduction of our original paper,1-1 which he did not cite. In 1951, Bickerstaff and Cloake1-2 described three patients who had ophthalmoplegia and ataxia as a grave syndrome they called “mesencephalitis and rhombencephalitis”. They speculated that the aetiology of Bickerstaff's brainstem encephalitis is similar to that of Guillain-Barré syndrome because prodromal upper respiratory infection, areflexia, and CSF albumino-cytological dissociation were detected. In 1956, Fisher1-3 described three patients who presented with acute ophthalmoplegia, ataxia of the cerebellar type, and areflexia as a variant of Guillain-Barré syndrome. One of the three patients showed mild drowsiness. In 1957, Bickerstaff1-4 added five more patients to the original study and named the condition “brain-stem encephalitis”. In 1982, Al-Din et al 1-5 (Bickerstaff being one of the authors) described 18 patients, and considered Bickerstaff's brainstem encephalitis to be a distinct clinical entity, not a variant of Guillain-Barré syndrome on the basis of radiological (three patients) and pathological (one patient) changes in the brainstem. In 1983, Ropper,1-6 however, criticised their report. He considered that six of the 18 cases reported by Bickerstaff's group were typical Miller Fisher syndrome, and that the other 12 represented obscure brainstem lesions without peripheral polyneuropathy. In 1987, Al-Din1-7 described two cases in which there was an altered state of consciousness and motor nerve dysfunction, in addition to the triad of Miller Fisher. He presented a “spectrum hypothesis”, in which Guillain-Barré syndrome and the syndrome of ophthalmoplegia, ataxia, and areflexia are at opposite ends of a broad spectrum, although clinically and pathologically distinct. He regarded his reported cases as being in the middle of that spectrum.

    Patients showing drowsiness, brisk reflexes, extensor plantar responses, and hemisensory disturbance, or central nervous system pleocytosis usually are considered to have Bickerstaff's brainstem encephalitis rather than Miller Fisher syndrome. Opinions differ as to whether these two conditions are distinct or related because the aetiology had yet to be established. While studying serum samples from patients with Miller Fisher syndrome, our attention was brought back to a patient with Bickerstaff's brainstem encephalitis previously treated in our hospital. In addition to acute ophthalmoplegia and cerebellar ataxia, this patient (patient 2 in Yuki et al 1-1) had abnormal plantar responses and became comatose. These neurological symptoms, however, disappeared 2 months after onset. At that time, we thought that anti-GQ1b antibody testing could be used to differentiate between Bickerstaff's brainstem encephalitis and Miller Fisher syndrome. Unexpectedly, the patient had the anti-GQ1b IgG antibody. Therefore, we investigated whether two other patients with Bickerstaff's brainstem encephalitis had this antibody. All three patients had high anti-GQ1b IgG antibody titres, which decreased with their clinical improvement.1-1 As prior infection is frequent in Bickerstaff's brainstem encephalitis, as in Miller Fisher syndrome and Guillain-Barré syndrome, an autoimmune mechanism was suggested in the pathogenesis of Bickerstaff's brainstem encephalitis. The finding that Bickerstaff's brainstem encephalitis and Miller Fisher syndrome share common autoantibodies suggested that the autoimmune mechanism is common to both and that they are not distinct, but closely related conditions. Bickerstaff and Cloake1-2 speculated that the aetiology of Bickerstaff's brainstem encephalitis is similar to that of Guillain-Barré syndrome, but the nosological relation between the two had yet to be clarified because of the lack of pathological findings in overlapping cases. We therefore reported necropsy findings for a patient in whom overlapping Bickerstaff's brainstem encephalitis and Guillain-Barré syndrome had been diagnosed clinically.1-8

    Effective therapy for Bickerstaff's brainstem encephalitis has yet to be established. As shown, Bickerstaff's brainstem encephalitis and Guillain-Barré syndrome are closely related; therefore, steroids should not be used to treat these disorders. Instead, the established treatments—plasmapheresis and intravenous immunoglobulins (IVIg)— should be used. Recent studies1-9 1-10 provide theoretical evidence that the removal of anti-GQ1b antibodies is reasonable and beneficial. Some patients with Bickerstaff's brainstem encephalitis respond favourably to plasmapheresis and IVIg.1-11 1-12 We recommend that no steroids, rather IVIg (or plasmapheresis), be used to treat Bickerstaff's brainstem encephalitis. Controlled clinical trials are needed to establish the efficacy of these procedures as therapy for this disease.


    1. 1-1.
    2. 1-2.
    3. 1-3.
    4. 1-4.
    5. 1-5.
    6. 1-6.
    7. 1-7.
    8. 1-8.
    9. 1-9.
    10. 1-10.
    11. 1-11.
    12. 1-12.
    View Abstract

    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.