Exploratory EEG studies for the assessment of neurological liabilities in conscious dogs in early drug development.

INTRODUCTION: Convulsions in toxicology studies can be the first indication of seizure liability. Drug levels during convulsions are not usually evaluated. This, and exposure variability after oral administration, complicates estimation of safety margins. The electroencephalogram (EEG) enables symptoms to be attributed to seizures and to collect samples during epileptiform activity without clinical convulsion. We evaluated an EEG-study design for optimized detection of neurological symptoms. Additionally, we assessed whether EEG- based anticonvulsive treatment is feasible, to prevent progression to convulsions and if dogs have higher sensitivity towards neurological symptoms than non-human-primates. METHODS: Three compounds that previously were tested in non-human-primates were selected to evaluate the dog EEG-study design. Two substances were administered in escalating intravenous doses; the third was given as single oral dose. Per compound, one male and one female telemetered dog were evaluated; males also had cerebrospinal-fluid-ports. Drug levels, video-EEG and clinical symptoms were evaluated and compared to previous studies. RESULTS: While similar neurological symptoms were induced, intravenous administration reduced experimental time compared to standard toxicology studies. EEG analysis could link animal behavior to seizures but did not allow convulsion prevention. This was due to artefacts and the short latency between onset of epileptiform EEG activity and clinical convulsions. Free plasma concentrations during convulsions were comparable between dogs and non-human-primates. DISCUSSION: The findings suggest that infusion studies provide a possibility to investigate neurological adverse effects in few animals in a short time period. For candidates with a high risk for seizures, such studies can guide dose selection for longer regulatory studies and improve safety margin definition.