Abstracts

Rationale: Padsevonil (PSL; UCB0942) is an antiepileptic drug candidate that binds with high affinity to the three isoforms of synaptic vesicle 2 proteins (SV2A, B, C) and with ~100-fold lower affinity to the benzodiazepine (BZD) site of the GABAA receptor. Its anticonvulsant properties were characterized in a range of mouse models representing focal and generalized seizures. The protective effect of PSL in the 6Hz model was compared directly to that of the selective SV2A ligands levetiracetam (LEV) and brivaracetam (BRV), and the BZD receptor agonist diazepam (DZP), administered alone or in combination (LEV/DZP and BRV/DZP). Methods: Seizures were induced acoustically (audiogenic seizures), electrically (maximal electroshock [MES] and 6Hz models), or chemically by administration of pentylenetetrazol (PTZ; 89 mg/kg sc), bicuculline (3 mg/kg sc), pilocarpine (373 mg/kg ip) or 11-deoxycortisol (1.0–1.2 mmol/kg iv). PSL was administered (10 ml/kg ip) 15 (audiogenic) or 30 min (all others) before testing. The adverse effects of PSL on motor function were assessed in mice using the rotarod test. Experiments consisted of several groups of 10 mice; one group received vehicle and the others received different doses of PSL. For the comparative 6Hz study, PSL was tested at a dose of 0.17 mg/kg, resulting in approximately 2% occupancy at the BZD receptor and 35% occupancy at SV2A. DZP, LEV, and BRV were tested at corresponding doses to provide similar in vivo target occupancies (0.017, 1.83, and 0.42 mg/kg, respectively). Results: PSL showed marked potency in the 6Hz, audiogenic, and pilocarpine models (Table). It was also potent in the PTZ and 11-deoxycortisol models. In the latter model, many AEDs including BRV, phenytoin, carbamazepine, and valproate were ineffective, while LEV and DZP showed partial protection at the highest doses only (Kaminski R, et al. Neuropharmacol 2011;60:1098). PSL showed low potency against bicuculline-induced seizures and MES. A dose estimated to induce an impairment of motor performance in 50% of the animals (TD50) was equal to 11.8 mg/kg. Consequently, PSL displays a high margin between doses that induce side effects and protect against seizures in the 6Hz, audiogenic, and pilocarpine models (~60-70 fold difference). In the comparative 6Hz study, PSL protected a significantly greater proportion of mice than the LEV/DZP and BRV/DZP combinations (p Conclusions: Owing to its dual SV2 and BZD binding properties, PSL’s antiseizure profile differs considerably from selective SV2A ligands LEV and BRV. PSL showed substantially higher potency across seizure models and was active in several models where LEV or BRV are ineffective. Importantly, PSL offered significantly greater protection against seizures than LEV or BRV in combination with DZP. These findings suggest that the unique SV2A binding mode (Wood et al., AES 2017), the additional interaction with SV2B and SV2C, and the partial agonism at the BZD receptor would confer enhanced antiseizure properties beyond combination of compounds targeting SV2A and BZD. Funding: UCB Pharma-sponsored.