A randomized, blinded, placebo-controlled trial of divalproex sodium prophylaxis in adults with newly diagnosed brain tumors.

Abstract

Seizures occur after the diagnosis of brain tumors in up to 40% of patients. Prophylactic anticonvulsants are widely advocated despite a lack of convincing evidence of their efficacy in preventing first seizures. We conducted a randomized, double-blind, placebo-controlled study comparing the incidence of first seizures in divalproex sodium- and placebo-treated patients with newly diagnosed brain tumors. Patients who had not previously had a seizure were randomized within 14 days of diagnosis of their brain tumor to receive either divalproex sodium or placebo. All patients had at least one supratentorial brain lesion, a Karnofsky Performance Score (KPS) > or = 50%, and no previous anticonvulsant use or other brain disease. Compliance and adequacy of dosing were assessed by pill counts and monthly blood levels. Seventy-four of 75 consecutive eligible patients were entered in this study. Median follow-up was 7 months. The drug and placebo groups did not differ significantly in age, sex, KPS, primary tumor type, number or location of brain lesions, frequency of brain surgery, or pretreatment EEG. Thirteen of 37 patients (35%) receiving divalproex sodium and 9 of 37 patients (24%) on placebo had seizures. The odds ratio for a seizure in the divalproex sodium arm relative to the placebo arm was 1.7 (95% CI 0.6 to 4.6; p = 0.3). The hypothesis that anticonvulsant prophylaxis provides a reduction in the frequency of first seizure as small as 30% was rejected (p = 0.05). Anticonvulsant prophylaxis with divalproex sodium is not indicated for patients with brain tumors who have not had seizures.