April D Lake, Kevin Holsapple, Tanya McDonnell, Joseph C Arezzo, Ricardo Ramirez, Lindsay Gamelin, Mei Yu, Dylan Glatt, Ryan Dick, Xiaodong Xie, Regina Choy, Guofeng Cheng, Chin H Tay, Anne Chester, Darryl Kato, and Leigh Ann Burns-Naas
The nonclinical safety profile of GS-8873, a hepatitis B virus RNA transcript inhibitor was evaluated in rat and monkey 13-week toxicity studies with 8-week recovery phases. Vehicle or GS-8873 was dosed orally for 13 weeks at 2, 6, 20, and 60 mg/kg/day to Wistar Han rats and at 0.5, 1.5, 3, and 6 mg/kg/day to cynomolgus monkeys. In vitro and in vivo screening results from an analog discovered prior to GS-8873 informed the 13-week toxicology study designs. Neuroelectrophysiology and neurobehavioral evaluations were included at weeks 4 and 13 of the dosing and recovery phases for GS-8873. No adverse neurobehavioral effects were observed. Significant nerve conduction velocity (NCV) decreases and latency increases occurred at the high doses after 4 weeks of dosing. By week 13, dose-responsive NCV reductions and latency increases worsened across all dose groups compared with controls. Some reversal occurred 8 weeks after the last dose administered, but not to vehicle control levels. A minimal, axonal degeneration was observed in rat spinal and peripheral nerves across dose groups compared with controls. No monkey nervous system microscopic findings were observed. No-observed-adverse-effect-levels could not be determined for either species due to the neuroelectrophysiology findings and development was halted in the interest of safety. A retrospective risk assessment approach utilizing benchmark dose (BMD) modeling contributed 13-week NCV BMDL estimates (lower limits of the 95% confidence interval) in lieu of no-observed-adverse-effect-levels. The best-fitted models extrapolated NCV BMDLs for the rat caudal and monkey sural nerve at 0.3 and 0.1 mg/kg/day, respectively.