Abstracts

Rationale: Carisbamate is a novel neuromodulating agent in development for the adjunctive treatment of epilepsy. Its efficacy and tolerability were studied in one Phase 2 and two Phase 3 multicenter, randomized, double-blind placebo-controlled trials in partial onset seizures (POS). An integrated PK/PD model was developed to characterize the exposure-response relationship over the entire dose range studied. Methods: A population pharmacokinetic model that described the pharmacokinetics of carisbamate following oral administration was developed using data from 2 Phase 3 studies and 3 Phase 2 studies in the patient population and 10 Phase 1 studies in healthy subjects. In total 15695 concentrations were collected in 1427 individuals. Using NONMEM®, a structural model was identified that described the pharmacokinetics of carisbamate. Covariate model building was used to identify influential factors that explained carisbamate’s pharmacokinetic behavior and accounted for variability. The individual estimates of the model parameters (ka, CL/F, V/F) of patients were used to simulate patient-specific plasma concentration time profiles at steady state of the target dose. From these profiles, steady state trough concentrations were derived which were used as input for the PK/PD-model. In the PK/PD-analysis the relationship between the primary efficacy variable (percent reduction of POS from baseline) and the steady state trough concentration was investigated. Data of three clinical trials were used (Phase 2 study:100, 300, 800, and 1600mg/d, N=533 and two identical Phase 3 trials: 200 and 400 mg/d, total N=1116) Results: The pharmacokinetics was adequately described by a one-comparment model with first-order absorption. The most important factor that affected carisbamate pharmacokinetics was the co-administration of enzyme inducing anti-epileptic drugs (e.g. carbamazepine phenytoin and phenobarbital). Co-administration of these drugs reduced carisbamate exposure by about 33-36%, confirming earlier findings in drug-drug interaction Phase 1 studies. In all three studies separately, a statistically significant positive relationship between seizure rate reduction and steady state trough concentrations was found. When all three studies were pooled, the PK/PD-relationship could be adequately described by a single Emax-model, including only a different placebo-response between Phase 2 and Phase 3. Simulations predicted a clinically meaningful increase of the placebo-substracted seizure rate reduction over the dose range of 400 mg/day to 1600 mg/day. Conclusions: Data of three randomized, double-blind placebo-controlled trials were adequately described by a single Emax PK/PD-model, allowing only for a different placebo response between Phase 2 and Phase 3. The predicted dose-response relationship confirmed the effectiveness of carisbamate as adjunctive treatment of partial onset seizures at doses of more than 400 mg/d.