SCIENTIFIC COMMENTARY The other BSE

Brainstem encephalitis has long proved complex and cyptic, but the waters are beginning to clear. Six decades ago, Edwin Bickerstaff (1921–2008), writing from Birmingham with Professor Philip Cloake (1890–1969), described three cases of ‘mesencephalitis and rhombencephalitis’ (Bickerstaff and Cloak, 1951). Each individual presented with a brainstem oculomotor disturbance, other cranial neuropathies and ataxia (with variable other signs); two had a hypercellular cerebrospinal fluid. The patients recovered well; and though lamenting the lack of pathological characterization, the authors considered ‘their clinical features and course were so remarkably similar that it seemed highly probable that the same or a closely related pathological process was involved’. Six years later, Bickerstaff published eight cases of what he now designated ‘brainstem encephalitis’ (Bickerstaff, 1957); others later came to attach Bickerstaff’s name to the disorder. The waters were, however, already becoming rather muddied: only one case had come to autopsy, and no brainstem inflammation was apparent (Bickerstaff, 1957). What is more, four of these eight cases of ophthalmoplegia and ataxia had areflexia (one also with extensor plantar responses); and 1 year earlier Charles Miller Fisher had described three cases of ‘an unusual variant of acute idiopathic polyneuritis (syndrome of ophthalmoplegia, ataxia and areflexia)’ (Fisher, 1956), which of course also came to bear his name. There followed a long period of uncertainty as to whether Fisher and Bickerstaff syndromes were one and the same, and of central or peripheral origin; only in the last few years has (mainly) immunological characterization begun to resolve these matters. This brief neuro-history offers a backdrop to last year’s fascinating description in Brain of a new brainstem inflammatory syndrome. Sean Pittock, Jan Debruyn, Brian Weinshenker and Mark Keegan and colleagues, from the Mayo Clinic (USA) and from Ghent University Hospital (Belgium), beautifully and succinctly (if perhaps not euphoniously) described ‘CLIPPERS’ (Pittock et al., 2010)—Chronic Lymphocytic Inflammation with Pontine Perivascular Enhancement Responsive to Steroids—as a distinct form of idiopathic brainstem encephalitis centred on the pons. All eight patients with CLIPPERS had a brainstem oculomotor disturbance and ataxia; but what was striking and may (speculatively) have inspired the initial identification of these cases was a remarkable pattern of MRI change, ‘symmetric curvilinear gadolinium enhancement peppering the pons’, changes extending in all cases to other parts of the brainstem, principally the medulla or the midbrain, and occasionally beyond to the cerebellum or spinal cord. One patient had a CSF pleocytosis, three had cyto-albuminic dissociation, and three had oligoclonal bands specific to the cerebrospinal fluid. Importantly, four patients underwent brain biopsy, and the tissue again showed a similar picture—hindbrain lymphocytic infiltration, largely but not wholly perivascular in distribution, accompanied by a moderate number of histiocytes and activated microglia. No demyelination was seen; no vasculitic, granulomatous or lymphomatous changes; and no neuronal or astrocytic changes are mentioned. All patients improved after corticosteroid treatment, some proving steroid dependent. Reports of similar cases have quickly followed—four are printed online as part of the current issue of Brain—and already have begun to extend the phenotype. Taeib et al. (2011) describe a patient with comparable imaging changes and steroid responsiveness, with childhood onset, and persisting over decades—with, notably, normal MRI for the first 9 years. Duprez and Sindic (2011) present a patient with little steroid response, and propose that new imaging criteria obviate the need for biopsy no spinal fluid results are reported and there is no neuropathology. In their reply, Keegan et al. (2011) demur here, consider the imaging picture dissimilar and suggest that the individual reported likely had an alternative disorder. List et al. (2011) report a single more typical case, with negative anti-ganglioside antibody testing; spinal fluid examination showed transient oligoclonal bands, but no cellular reaction. These authors also did not pursue biopsy, partly ‘given the typical clinical and radiological features of CLIPPERS’. Keegan et al. (2011) are alert to the risk that their description should encourage any reluctance to biopsy, notwithstanding the delicate site of the disorder. Making CLIPPERS a ‘disease,’ diagnosed only on imaging, will hardly advance our understanding of brainstem inflammatory disorders. Cerebral and brainstem biopsies are increasingly seen as safer than often previously considered, and informative (Rachinger et al., 2009; Rajshekhar and Moorthy, 2010; Rice et al., 2011). Moreover, as Jones et al. (2011) illustrate, these magnetic resonance findings are not pathognomic: low-grade glioma may masquerade as CLIPPERS. Lastly, Kastrup et al. (2011) report three further cases. All had extra-pontine involvement clinically and radiologically. Interestingly, two exhibited a marked elevation of serum IgE. None had a cerebrospinal fluid pleocytosis or oligoclonal bands. All three patients had normal peripheral nerve conduction studies, and all had brain biopsies. These showed a similar pattern of mainly but not exclusively angiocentric infiltration by lymphocytes, with doi:10.1093/brain/awr181 Brain 2011: 134; 2194–2196 | 2194