Carbamazepine versus valproate monotherapy for epilepsy: a meta-analysis

PURPOSE: To provide an overview of the evidence comparing carbamazepine (CBZ) and valproate (VPA) monotherapy for epilepsy, investigating whether existing data support the current practice of preferring CBZ for partial-onset and VPA for generalized-onset seizures. METHODS: We performed meta-analysis of randomized controlled trials by using individual patient data. Our strategy included searches of (a) Medline, 1966-2000; (b) The Cochrane Library 2000, issue 4; and (c) the pharmaceutical industry. Outcome measures were time to discontinuation of allocated treatment, time to 12-month remission, and time to first seizure after randomization. Results are expressed as hazard ratios (HRs; 95% CI), where HR>1 indicates that an event is more likely with VPA. A test for an interaction between treatment and seizure type (partial vs. generalized onset) also was undertaken. RESULTS: Data were available for 1,265 patients from five trials. Overall results (HR, 95% CI) were Time to treatment discontinuation, 0.97 (0.79-1.18); 12-Month remission, 0.87 (0.74-1.02); and First seizure, 1.09 (0.96-1.25), suggesting no overall difference for these outcomes. The test for an interaction between Treatment and Seizure type was significant for time to first seizure, but for no other outcome. The age distribution of adults classified as having generalized seizures indicated that significant numbers of patients may have had their seizures misclassified. CONCLUSIONS: We found some evidence to support the preference of CBZ for partial-onset seizures, but no evidence to support the preference of VPA for generalized-onset seizures. Confidence intervals are too wide to infer equivalence. Misclassification of patients may have confounded our results and has important implications for future trials.