Abstracts

Rationale: We have previously demonstrated that preemptive low frequency sine wave stimulation (LFSWS) decreased the incidence of amygdala and hippocampal kindled seizures (Goodman et al., 2005). The purpose of this study was to use a closed-loop responsive neurostimulation system to detect epileptiform activity and to deliver stimulation similar to LFSWS in a spontaneous, pharmacoresistant seizure model: the tetanus toxin model of epilepsy.Methods: Tetanus toxin (50ng, 0.5μl) was injected into the motor cortex of male Sprague-Dawley rats (275-350g, n=8). A bipolar, Teflon-coated stainless steel electrode was lowered to the same location. Two stainless steel screw electrodes were placed in the skull on either side of the injection site and a ground screw was placed behind Lambda. Gold-plated pins connected to the electrodes were placed in a socket and fixed to the skull with dental cement. A modified version of the NeuroPace® RNS™ responsive neurostimulation system for epilepsy was used for event detection, electrical stimulation, and data collection. Each rat was fitted with a vest and a neurostimulator (NS) was attached to the vest and connected to the electrodes. The vest and NS were worn for at least 4h/day. Baseline EEG activity was collected from screw electrodes and microelectrodes and stored in the NS. Event detectors were configured to detect epileptiform activity from each rat. Event detection counts were stored in hourly histograms by the NS. Stimulation consisted of a continuous 1-Hz biphasic square wave, at 100μA per phase. Subjects were divided into two groups. In Group I, stimulation was delivered for 2s by the NS each time epileptiform activity was detected. In Group II stimulation was delivered for 60s, once/h, for 4h each day. In both groups, baseline detection counts without stimulation were collected for one week, followed by one week of stimulation, then one week of baseline.Results: Tetanus toxin-induced epileptiform activity consisted of wide-band bursting activity in the beta and gamma frequency range, as well as frequent spiking, occasionally resembling epilepsia partialis continua. Behaviorally, each rat exhibited facial fasciculations correlated with EEG burst activity. Some test stimulations of 30s acutely decreased epileptiform activity for up to 5min. Group I received an average of 166s/day of stimulation, while Group II received 240s/day. Only the rats in Group I who received the most stimulation showed evidence of altered detection rates in response to stimulation, suggestive of a response threshold. One rat with an average of 175s/day of stimulation showed decreased detections (p<0.001), while another rat with 359s/day of stimulation showed increased detections (p<0.01).Conclusions: This study demonstrates that closed-loop responsive neurostimulation can be used in freely moving animals to evaluate the efficacy of new electrical stimulation paradigms in experimental models of epilepsy. Support for this research was provided by Neuropace, Inc. and ATP Cooperative Agreement No. 70NANB3H3044