Abstracts

Demonstrate acute safety of responsive stimulation in the seizure onset zone and provide evidence that responsive stimulation can affect epileptiform activity.
Patients undergoing intracranial seizure monitoring were connected to an external Responsive Neurostimulator (eRNS) system. The eRNS is capable of customizable seizure detection, closed-loop delivery of automated responsive stimulation (RS), and intracranial EEG activity storage. Once each subject had all the seizures needed for surgical planning, the eRNS was programmed to deliver RS. The rate of adverse device effects (ADE) was monitored to evaluate safety. Efficacy of RS was assessed by investigator review of the electrographic activity in stored EEGs immediately pre- and post-RS, and by the seizure rate.
Safety was evident in all 34 subjects enrolled to date (mean age 33, 65% female). Of the 3379 RS bursts delivered, no ADEs were reported. One instance of a brief, non-propagated afterdischarge occurred, but on no occasions did RS elicit seizure activity. Two subjects indicated awareness of RS (tingling/visual flash), but neither reported discomfort. The RS settings were adjusted and the phenomena promptly subsided. Detection algorithms were quickly tuned to achieve 100% sensitivity. The mean study duration for the 27 subjects receiving RS was 31 hours (range 6-117). An average of 4 stimulations/hour/subject was delivered with a mean total stimulation time of 17 seconds/subject. A positive electrographic effect (EE) was evident in 11 of the RS subjects (41%). In 4 of the RS subjects, the EE was indeterminate due to short closed-loop operating time. In 6 (22%) of the RS subjects, there was no evidence of EE. In several subjects, early results did not show a positive EE, but subsequent adjustments to the RS settings produced evidence of efficacy. The opportunity to assess a range of RS settings in most subjects was limited due to surgical schedule constraints. In a subject with multiple daily seizures prior to RS, increasing detection sensitivity to include stimulation of subclinical epileptiform activity produced complete seizure cessation over a 3-day period. Initial results stimulating hippocampal onset seizures after the seizures were well developed did not show evidence of efficacy. Qualitative EEG assessments indicate that early stimulation of epileptiform activity is important.
The eRNS successfully detected typical seizure onsets and subclinical epileptiform activity. RS was well tolerated and performed safely. Ongoing review of stored EEGs was important, resulting in parameter adjustments that produced positive results in some subjects who did not initially respond. Optimal responsive stimulation parameters are still being investigated. Preliminary data suggests that responsively stimulating epileptiform activity early may improve electrographic and clinical response.
[Supported by: NeuroPace, Mt. View, CA]