(M-050) Exposure-Response Analysis of Cenobamate

Janice Laramy, PharmD, PhD – Associate Director, Clinical Pharmacology, SK Life Science, Inc.; Vijay Vashi, PhD – Executive Director and Head, Clinical Pharmacology, SK Life Science, Inc.

Objectives: Cenobamate is a small molecule approved and marketed for focal seizures (partial-onset seizures, POS) in several countries under the brand name of Xcopri® or Ontozry®. The aim of this analysis was to develop an exposure-response (ER) model to describe the relationship between cenobamate exposure and efficacy based on data obtained in adult POS patients. This model was subsequently used to simulate seizure reduction that supported the selected (approved) dosing regimen.

Methods: Data were pooled from two efficacy clinical studies, YKP3089C013 and YKP3089C017. These studies included data from patients with POS treated with cenobamate at maintenance doses ranging from 100-400 mg/day. A previously developed population pharmacokinetic model (Jonsson et al, 2024) was used to derive the estimated daily cenobamate exposure. A population pharmacodynamic model was optimized and implemented, using nonlinear mixed-effects methods while treating the seizure data as count data. Different distributions were considered to describe the count data and different models to characterize the relationship between daily cenobamate exposure and seizure frequency. The final model was qualified by standard numerical goodness-of-fit checks, including visual predictive check. The analyses were conducted using NONMEM (version 7.4) and R (version 3.5.2).

Results: The analysis dataset included 104,037 daily records including 59,645 seizures among 665 subjects, over a cenobamate dose range from 100-400 mg/day. The seizure count data were best described using a generalized Poisson distribution, and a power relationship between area under the concentration-time curves (AUCs) and seizure responses, in order to describe the ER relationship. Inter-individual variability was estimated for overall seizure frequency parameter lambda (90.4% coefficient of variation [CV]), dispersion parameter delta (354% CV), and for the AUC-dependent cenobamate response (83.5% CV). Goodness-of-fit and simulation-based diagnostics indicated that the final model described the data well. The model was used to simulate the expected seizure count and fraction of subjects with >50% reduction in seizure frequency, as responses at doses up to 400 mg/day, while assuming a patient population similar to that enrolled in the two efficacy clinical studies. The dose-response curve was nonlinear in the simulated 100-400 mg/day dose range, with a simulated fraction of subjects with more than 50% reduction in seizure frequency of 0.42 at 100 mg/day, 0.52 at 200 mg/day, and 0.60 at 400 mg/day. The expected proportion of seizure-free subjects was 0.13 at 100 mg/day during the maintenance phase, 0.20 at 200 mg/day, and 0.28 at 400 mg/day, respectively.

Conclusion: These results support the cenobamate maintenance dosing regimen of 100-400 mg/day in adult patients with POS.
Supported by: SK Life Science, Inc.

Citations: Jonsson F, et al. [Manuscript under preparation].