Central pontine myelinolysis
Central pontine myelinolysis was first described in four patients in whom autopsy studies revealed a single, sharply outlined focus of myelin destruction in the rostral part of the pons, indiscriminately involving all descending and ascending fibre pathways, with the exception of the ventrolateral tract, but sparing nerve fibres (Adams et al. 1959). The disorder is not rare and much has since been learned concerning the aetiology, pathology, and clinical manifestations. Although the original descriptive term ‘central pontine myelinolysis’ remains valid, demyelination is often observed outside the pons (Wright et al. 1979). A high proportion of the cases first described were metabolically disturbed, often as a result of alcohol abuse, but the condition is now known to occur in association with Wernicke's encephalopathy, liver cirrhosis not due to alcohol, as part of Wilson's disease, after hepatic transplantation, and in non-cirrhotic liver disease. Central pontine myelinolysis is also seen as a complication of leukaemia, uraemia and haemodialysis, hyperemesis gravidarum or other causes of prolonged vomiting, following diuretic therapy, and in children. In several of these situations, the diagnosis can be confused with Wernicke's encephalopathy, but an important difference is that patients with central pontine myelinolysis are often hyponatraemic before the onset of neurological symptoms. This observation led to the suggestion that the pons is unusually susceptible to changes in electrolyte balance, but it is now believed that central pontine myelinolysis is caused by overzealous correction of a low serum sodium (Sterns et al. 1986). A prospective study of electrolyte correction in patients with hyponatraemia demonstrated that pontine damage correlates both with the degree of hyponatraemia and speed with which it is corrected—starting levels of less than 110 mmol/l or rates of correction above 2 mmol/l/h increasing the risk of pontine damage (Brunner et al. 1990). Experimental studies have shown that rapid changes in sodium can better be tolerated when hyponatraemia has arisen acutely than in chronic electrolyte imbalance. Central pontine myelinolysis is occasionally seen as a result of sustained or rapidly corrected hypernatraemia.
Clinically, the process begins with damage to pathways placed centrally within the pons and then spreads centrifugally. The fully evolved clinical picture is of flaccid paralysis with facial and bulbar weakness, disordered eye movements, profound imbalance, and alterations in consciousness. The recent literature highlights movement disorder and other extrapyramidal manifestations of central pontine and extrapontine myelinolysis (Seiser et al. 1998). Some manifestations can be attributed to the poor medical condition of affected individuals. Since central pontine myelinolysis tends to occur following therapeutic intervention and correction of the serum sodium, other features of hyponatraemia, such as fits, tend not to be associated. In a series observed in hospitals in New York and Oxford, Ellis (1995) reported only one case of central pontine myelinolysis (with a striatal syndrome) amongst 184 patients presenting with hyponatraemia and managed in a variety of ways, including rapid correction of serum sodium in some instances. Newell and Kleinschmidt-DeMasters (1996) carried out an autopsy-based epidemiological survey of the prevalence and features of central pontine myelinolysis over a period in which management of hyponatraemia was changing, but observed no change in the incidence. Of 15 clinically undiagnosed cases, 5 of 6 with active lesions were associated with overzealous correction of the serum sodium; these correlations could not be made for comparison in those with remote episodes of central pontine myelinolysis.
The prognosis for clinical recovery is largely determined by the underlying metabolic disorder and the extent to which that can be managed; but with stabilization of the serum sodium, management of the bulbar failure and time, the prognosis for neurological recovery is good and the condition does not relapse spontaneously. Menger and Jorg (1999) reviewed retrospectively the outcome in 44 patients and concluded that death in two, or poor recovery in 10, with residual deficits not affecting independence in a further 11, hinged on the general medical complications of the more acute illness or its precipitating events, rather than the severity of pontine myelinolysis; they did not identify any radiological or electrophysiological predictors of poor outcome.
The clinical features of brainstem disease occurring in a patient with hyponatraemia are sufficiently distinctive not to cause diagnostic difficulties, but uncertainty may arise if only a small reduction in serum sodium has occurred or the fall has not been documented, in which case brain imaging may be informative. The acute changes can be recognized by computed tomography or magnetic resonance imaging and persistent abnormal signals attributable to astrocytic proliferation and gliosis can be detected long after the clinical features have resolved. Ultrastructural studies support the view that the condition is due to a metabolic process which is directed primarily at oligodendroglia.