Authors from the University Children’s Hospital in Heidelberg estimate the incidence of fatal valproate (VPA) hepatotoxicity in West Germany at around 1 in 5000 and find it hard to justify the use of VPA as a drug of first choice for children with generalised epilepsies. Analyses of data on 16 cases (15 between 1980 and 1986) and 75 additional published cases (a total of 91 cases) showed that no single high-risk age group could be defined; only 2 of the 16 German cases (12%) and 26% of the 91 cases reviewed were under 3 years of age. Fatalities were more frequent in young children on polytherapy, but 14 (15%) followed monotherapy with VPA and 10 (71%) of these were in patients older than 3 years. [1, 2]

COMMENT: I agree with.the authors that valproate, a drug with known hepatotoxicity and potentially fatal side-effects, should not be used as a first-line therapy in children with epilepsy. Also, its use in the treatment of febrile convulsions seems unacceptable. Despite the less worrisome estimates of fatalities from the US (1 in 10,000; 1 in 7,000 for polytherapy and 1 in 37,000 for monotherapy)1 and from England (1 in 20,000)2, careful monitoring of valproate therapy should be mandatory. Fatalities for VPA polytherapy in children < 3 yr was 1/500. [3, 4, 5]