To our knowledge, only leucopenia and thrombocytopenia have been described as possible haematological side effects of lamotrigine.1^–3 We report a case of agranulocytosis associated with lamotrigine treatment.

A white girl aged 11 years 10 months who had left renal agenesis, an imperforate anus (corrected postnatally), neuronal heterotopia at the left insula, and seizures had been free of seizures for nearly two years while taking carbamazepine. She had positive serological results for hepatitis B and C. During the anticonvulsive treatment she developed a clinically significant increase in serum transaminase concentrations, raising suspicion of chronic hepatitis C (results of a polymerase chain reaction assay were positive on two occasions and negative on one). As she was clinically well, normally developed, and free of seizures, the anticonvulsive treatment was discontinued, which resulted in serum transaminase concentrations becoming normal.

Twelve months later she was admitted because she had had three focal seizures with secondary generalisation. Anticonvulsive treatment was reinstituted. We opted for lamotrigine treatment alone as this drug does not induce liver enzymes so would not mask the increase in serum transaminase concentrations from progression of hepatitis. Instead of the recommended initial dose of 25 mg/day, she was given 50 mg/day (1.5 mg/kg/day).

Two weeks later her parents observed a cutaneous rash. Lamotrigine was discontinued immediately. After two days she presented to our service still with a maculopapular rash, with confluent papules on the face, slight pruritus, and some abdominal discomfort. A haemogram showed leucopenia (leucocyte count 2.2×10⁹/l, proportion of neutrophils 0.34). Liver enzyme activities were slightly raised (alanine aminotransferase 34 U/l (normal range 5-17), aspartate aminotransferase 33 U/l (2-23)). After 3 days she developed agranulocytosis (leucocyte count 3.1×10⁹/l, proportion of lymphocytes 0.92 and of monocytes 0.08). She had no clinical problems, and the rash disappeared on the fifth day.

One week after the second haemogram her leucocyte count improved (4.9×10⁹/l, proportion of neutrophils 0.56), increasing during the following week (6.5×10⁹/l, proportion of neutrophils 0.50). Liver enzyme concentrations remained slightly raised.