Antidepressants, thiazide diuretics and alcoholism: central pontine myelinolysis waiting to happen?

A 50-year-old man, with a known weekly alcohol intake of almost 100 units, presented to his GP feeling drowsy. Routine blood tests revealed a profound hyponatraemia of 100 mmol/litre, and he was referred to hospital for investigations. On admission he was drowsy, but had a Glasgow Coma Scale score of 15, an abbreviated mental test score of 10/10 and an entirely normal neurological examination. Liver function tests were consistent with excessive alcohol consumption. He had been prescribed bendroflumethazide and clomipramine (a tricyclic antidepressant) by his GP. In view of their potential tocause hyponatraemia these were subsequently withheld and he was fluid restricted to 1.5 litres per day. His sodium level rose gradually and on day 5 of admission it had increased to 122 mmol/litre. He was discharged on day 8 having been advised to stop drinking.

Less than a week after discharge he represented to his GP with symptoms of excessive sweating, shaking, constipation, disorientation and dysarthria. He was febrile at 38.6°C , and had a blood pressure of 180/90 mmHg. Electrolytes were normal (Na+ 135 mmol/litre). Although he denied consuming any alcohol since his initial admission, his symptoms were consistent with acute alcohol withdrawal and he was treated accordingly. He failed to improve on two attempts at reducing doses of chlordiazepoxide. His dysarthria progressed until his speech was difficult to understand and his swallow became unsafe. He continued to spike temperatures throughout the admission and although a catheter specimen of urine was positive for enterococcus, treatment with a prolonged course of antibiotics had not resolved his fever.

A computed tomography scan of his head and a lumbar puncture were performed. Other than cortical atrophy in keeping with alcohol abuse, neither investigation provided any answers. It was considered that his symptoms could be explained by autonomic dysfunction. In view of his corrected hyponatraemia, central pontine myelinolysis was suggested and a magnetic resonance imaging head scan was performed (Figure 1). This confirmed the presence of central hyperintensity on the water sensitive sequences within the central pons, in keeping with an osmotic demyelination syndrome. Additionally these changes were identified within the caudate nuclei, possibly accounting for the patient's inability to fully control voluntary movement. The ongoing temperatures without sepsis suggested hypothalamic involvement.

Following diagnosis this patient failed to improve. Treatment became supportive, and he died 7 weeks later.