Brivaracetam-induced elevation of carbamazepine epoxide levels: A post-hoc analysis from the clinical development...

To assess the association, if any, between brivaracetam (BRV)-induced elevated carbamazepine-10,11-epoxide (CBZ-E) and toxicity and efficacy in patients with epilepsy. Data were pooled from three double-blind, placebo-controlled, Phase III studies of adjunctive BRV in adults with uncontrolled focal seizures (N01252/NCT00490035, N01253/NCT00464269, N01358/NCT01261325). Treatment-emergent adverse events (TEAEs) of interest (ataxia, diplopia, dizziness, nystagmus, somnolence, accidental overdose or poisoning, and toxicity), discontinuations due to TEAEs, and serious TEAEs (SAEs) were assessed in subgroups who did/did not receive carbamazepine (CBZ) at study entry (CBZ+ and CBZ-). Logistic regression analysis evaluated CBZ-E/CBZ plasma concentrations and TEAEs. SAEs suggestive of CBZ-E toxicity were summarized from the BRV safety database up to a cut-off of October 1, 2014. Percent reduction in focal seizure frequency over placebo was assessed in subgroups of CBZ-E/CBZ ratios. Data from 1558 patients were included in the pooled safety population. Of these, concomitant CBZ was received by 184/459 (40.1%) placebo-treated and 315/803 (39.2%) BRV-treated patients (≥50 mg/day). In BRV-treated patients, study completion rates were similar in the CBZ+ (92.7%) and CBZ- (88.7%) groups; incidence of TEAEs of interest was similar (CBZ+ 24.4%; CBZ- 24.2%), and did not appear affected by CBZ dosage; SAEs and discontinuations due to TEAEs were CBZ+ 1.6%; CBZ- 3.9% and 2.9%; 9.2%, respectively. Likelihood of TEAEs of interest decreased with increasing CBZ-E/CBZ ratio for BRV-treated patients: odds ratio 0.88 (95% confidence intervals 0.74, 1.03; p = 0.112). In the safety database, five SAEs suggestive of CBZ-E toxicity were identified. Efficacy outcomes did not appear to have a consistent pattern across CBZ-E/CBZ ratio subgroups. This post-hoc analysis does not support an association between CBZ-E levels and TEAEs potentially associated with CBZ-E toxicity, or with increases in efficacy. Overall, current evidence does not suggest that BRV dose adjustment is required with concomitant CBZ.