The Complexity and Challenges of Developing Effective Epilepsy Therapies: Lessons to be Learned from Comparing Cenobamate, Carisbamate, and Padsevonil.

Most of the about 30 clinically approved antiseizure medications (ASMs) have been discovered using animal seizure models. Thus, such models are commonly considered highly predictive of clinical antiseizure efficacy. However, this has recently been questioned by the apparent dichotomy in clinical antiseizure efficacy between the novel ASMs padsevonil and the alkyl-carbamate cenobamate, both of which were highly potent in a battery of acute and chronic rodent models. In this review, we analyze and compare the preclinical and clinical data for cenobamate and padsevonil, along with their proposed mechanisms of action, to identify differences between the two drugs that might have predicted the contrasting clinical outcome, as well as another alkyl-carbamate, carisbamate, which was highly effective in preclinical models but failed in clinical trials. While padsevonil was more potent than cenobamate and carisbamate in suppressing seizures in most acute and chronic animal models, an important difference was observed in the amygdala kindling model of focal epilepsy. Cenobamate suppressed both focal and secondary generalized convulsive seizures, whereas padsevonil was less effective against focal versus generalized seizures, and carisbamate was only partially effective in amygdala kindled rats. Padsevonil was developed in a target-based, rational drug design program that combined high affinity for synaptic vesicle glycoprotein 2 (SV2) with moderate affinity at the benzodiazepine binding site of the GABA