Abstracts

Rationale: RNS SYSTEM (r) System (NeuroPace, Inc.) was FDA approved in 2013 for the treatment of medically refractory epilepsy. In the multi-center randomized clinical trial, RNS system was shown to reduce seizure frequency as compared to sham stimulation; however not every patient responded to this treatment modality. In this study we evaluated the surgical outcomes of all patients who underwent RNS system placement at our center and sought to evaluate potential predictors of surgical response over time. Methods: We evaluated patients that underwent RNS system placement at our center from 2005-2018 who were at least one year post surgery. Data collected pertained to patient demographic information, seizure duration, seizure type, epilepsy classification (temporal vs extratemporal), previous or concomitant epilepsy surgery at time of or prior to RNS system implant, electrographic onset of seizures during their intracranial EEG monitoring and placement of RNS system electrodes. Seizure frequency data was obtained for focal aware, focal impaired and bilateral tonic clonic seizures averaged over a three month time period pre-surgery, post RNS system implant, post initial RNS system stimulation and most current. Surgical outcome was classified using the ILAE system. Responders were considered to have greater than 50% seizure reduction. We compared patient’s electrographic onset location based on intracranial EEG to the location of their implanted RNS system and categorized the location as same if they matched or different if the locations differed. This was completed by reviewing the intracranial electrode map at onset to post implant imaging by the neurosurgeons (DDS and EY). We also evaluated the total number of detections and long episodes daily as recorded by the RNS system for 7 days prior to stimulation and 7 days post stimulation. Exclusions included evaluating number of all detections and long episodes on day of stimulation along with patients that underwent detection setting changes at time of initial stimulation. Results: A total of 25 patients were identified who had RNS system implanted during the designated time frame but only 20 patients had greater than one year post surgical outcomes. There were 10 patients with extratemporal lobe epilepsy and 10 with temporal, either bitemporal or medial with neocortical involvement. Duration of epilepsy ranged from 3-37 years with a mean duration of 20 years of epilepsy. A total of 13 patients maintained their detection settings at time of stimulation initiation. Stimulation parameters ranged from 0.03uC-6.1uC with a mean of 1.0uC. 30% (6/20) of patients were seizure free for greater than a two year sustained period with majority (4/6) achieving seizure freedom in less than two years from RNS system implant date. Only one of these patients underwent resective epilepsy surgery at time of RNS system implant. Responder rate was 65% (13/20) with mean reduction rate of seizures at greater than 87%. RNS system electrode placement matching region of seizure onset seen on intracranial EEG(6/9) predicted response to RNS system stimulation with greater than 50% seizure reduction (p=0.011) for patients with extratemporal lobe epilepsy. Electrode placement for medial temporal lobe patients did not appear to impact outcome or response rates (p=1.0). All patients who experienced any reduction in seizure frequency after initiation of RNS system stimulation had a reduction in the total number of detections within 7 days post stimulation (range 4.7%-82.9%, mean 26.7%, N = 9). Those with greater than 20% reduction in the total number of detections(5/13) post stimulation were responders to RNS system stimulation with greater than 50% seizure reduction (p=0.04). Reduction or increase in the total number of long episodes did not have an association with outcome or response rate. Conclusions: In our cohort of patients, 30 percent of patients were seizure free for >2 years after treatment with RNS system, greater than what is reported in the literature. For extratemporal lobe patients, having the RNS system electrode placement match the region of electrographic onset seen on intracranial EEG was associated with better outcome (p=0.01). Our findings on this small subset of patients are suggestive that response to RNS system can be predicted within the first week post initial stimulation and that those with greater than 20% reduction in detection rate will be a responder to RNS system stimulation (ie, will have >50% seizure reduction). Additional studies with a larger subset of patients will be needed to validate these findings. Funding: No funding