Valproic acid and thrombocytopenia: cross-sectional study.

OBJECTIVE To investigate the relationship between platelet count and serum valproic acid level, age, duration of valproic acid therapy, and polytherapy, and to determine the clinical significance of thrombocytopenia associated with high-dosage valproic acid therapy. DESIGN Cross-sectional study. SETTING Residential unit for neurologically impaired children and paediatric out-patient clinic, Hong Kong. PATIENTS Ninety-six neurologically impaired children who were treated with valproic acid between 1 July 1991 to 3 June 1999. The comparison group consisted of 48 children receiving antiepileptic drugs other than valproic acid. INTERVENTION Low- or high-dosage valproic acid, using the threshold value of 40 mg/kg/d. MAIN OUTCOME MEASURES Platelet count and liver function, duration of valproic acid treatment, dosage, and trough serum valproic acid concentration. RESULTS Seventeen (17.7%) patients in the treatment group developed thrombocytopenia, compared with two (4.2%) in the comparison group (P<0.05). The platelet count was negatively correlated to serum valproic acid level and age, and positively correlated to polytherapy. The duration of valproic acid treatment was not a confounding factor in the age-related decrease in platelet count. Children with a trough level of >450 micromol/L or a daily dose of >40 mg/kg were more likely to develop thrombocytopenia. Thrombocytopenia was mild in most cases. CONCLUSIONS A trough valproic acid level of >450 micromol/L or a daily dose of >40 mg/kg should alert the clinician to the risk of developing thrombocytopenia. The risk is further increased for older children. The platelet count should be monitored for patients receiving a high concentration of valproic acid who are also receiving drugs that would affect homeostasis, or who are undergoing surgical procedures.