Abstracts

To improve treatment of human epilepsy, animal models are developed to determine the effectiveness of anticonvulsants. The kainate model of temporal lobe epilepsy has been investigated extensively and produces rats which are chronically epileptic. To fully understand the results and statistical significance of drug studies, it is necessary to understand the expected variation in seizure number and duration over a fixed period. For this study we consider 11 two month old Sprague-Dawley rats implanted with EEG radio-telemetry units. The rats were treated with varying doses of kainate and monitored using a video-EEG system for up to 5 months. EEG seizures were identified using routines written in Visual Basic 6.0 (DClamp software) and expert confirmation. We have determined that there are three distinct phases after kainate-treatment: (1) a latent phase which extends up to 5 months in which there are sporadic, clustered seizures, (2) a transitional phase, in which the number of seizures increases dramatically, and (3) a stable epileptic phase in which the rats have frequent seizures. We determined the daily and 5 day average variability in each of these phases of both the number of seizures and the time spent seizing. We expressed this as the mean fractional change in these parameters, defined as the average change in the seizure parameter (either number of seizures or total length of time seizing) from one period to the next, divided by the mean value of the parameter during the phase. The average daily and 5 day changes during phase 1 were 0.96 and 0.64 for the number of seizures and 1.10 and 0.67 for the time spent seizing. Much of the reason for the large variability in daily seizure occurrence was due to seizure clustering. During both the transition phase and the stable phase, the variability decreased dramatically. For the transition phase the average daily mean fractional change in number of seizures was 0.32 for the stable epileptic phase, the average daily mean fractional change was 0.23. Consideration of differences in variance of seizure frequency are important in the design of experiments using the kainate model of temporal lobe epilepsy. From these results it is evident that studies performed while the rats are in the [ldquo]latent phase.[rdquo] For example, investigations of epileptogenesis will require either larger numbers of animals or longer observation times to achieve statistical significance in the face of higher variance in seizure frequency. Studies performed during the two later phases will more easily yield significant results. (Supported by the NIH.)