Severe hepatorenal failure in a child receiving carbamazepine and erythromycin.

Sir: Severe hepatorenal failure possibly related to carbamazepine (CBZ) therapy has been reported in two children who had been receiving CBZ for several years [6] and in an old man who had begun treatment a few days before [1]. The man and one of the children died. The child at autopsy showed hepatic necrosis, focal hyaline necrosis of muscle, and renal tubular necrosis with myoglobinuric nephrosis. We have recently seen an 8-year-old boy who for 1 year had been receiving CBZ 500 mg daily for treatment of partial seizures. Neurological and general examinations had always been normal. Steady-state plasma drug levels had been about 7 gg/ ml. Three weeks before admission he was given erythromycin (EM) for an upper respiratory infection; 12h later he became drowsy with ataxic gait. Withdrawal of EM was followed by immediate recovery. Three days before admission EM was reintroduced associated with paracetamol for upper respiratory infection with fever; 24 h later the child began to show fatigue, vomiting, anorexia and oliguria. Later, he became drowsy; all drugs were stopped. At admission, his examination was normal. Blood examinations revealed: GOT 2090 IU/1 (0-37), GPT 4110 IU/1 (0-40), GGT 109 units/l, urea nitrogen 260mg/dl; creatinine 6.2 mg/dl. The small amount of urine which was produced showed erythrocytes > 50/MF but not proteinuria nor casts; urinary osmolarity 320mosmol/1; urinary sodium concentration 43 mmol/1. Other extensive examinations, including antigen and antibodies for hepatitis A, B, C and antibodies for cytomegalovirus, Epstein-Barr and herpes simplex virus, autoantibodies and immunocomplexes, revealed normal findings. CBZ level was determined 2 days after withdrawal and was 9.6 gg/ml. The patient was treated by dialysis. The hepatic and renal function improved progressively until recovery within 20 days. Renal biopsy performed 15 days after admission showed a picture of a late stage of acute renal necrosis. In this case the relatively high CBZ concentration 2 days after its discontinuation suggests a relationship between the symptoms of hepatorenal failure and high CBZ level. A sudden increase of CBZ level after EM introduction had probably occurred. The effects of macrolides in increasing CBZ levels are well known [2]. The laboratory findings including the renal biopsy and the fact that the hepatorenal failure occurred I year after the beginning of the CBZ therapy are against the hypothesis of a hypersensivity reaction. In our case hepatorenal failure was probably a dose-related result of CBZ therapy. This is in contrast with the clinical and laboratory findings observed in