Abstracts

Rationale: The RNS® System provides responsive stimulation to 1 or 2 seizure foci and is an effective adjunctive therapy to reduce seizure frequency in patients ≥ 18 years of age with medically refractory partial onset epilepsy (≥ 2 medications) having frequent and disabling seizures. Data from a controlled clinical trial were assessed to determine whether responsive stimulation was associated with cognitive morbidity and whether stimulation delivered to mesial temporal lobe structures (MTL) affected memory and language differently than stimulation delivered to other brain regions (non-MTL). Methods: Subjects were participants in the RNS® System Pivotal Clinical Investigation who completed either the Boston Naming Test (BNT) or Rey Auditory Verbal Learning (AVLT) test (learning and delayed recall) at baseline, and at 1 and/or 2 years post implantation. These cognitive measures were selected from a larger neuropsychological data set to manage Type I error rate since language and verbal memory are the greatest risks associated with anterior temporal lobectomy, particularly resection of the language-dominant mesial temporal lobe. Subjects who had both MTL and non-MTL seizure onsets were excluded. Results: The sample size for each analysis varied depending on the test and the number of subjects with available data for each of the comparative time points. The sample size of MTL subjects ranged from 67-78, and the sample size of non-MTL subjects ranged from 62-69. To maximize the sample size for each pair-wise comparison, independent 2-way AVOVAs (location x follow-up interval) were calculated separately for each variable at 1 and 2 years of follow-up. Across the 6 analyses (3 cognitive measures; 2 follow-up intervals), there were no statistical location by follow-up interval interactions. A significant time effect (practice) was observed at 1 year for AVLT verbal learning (p=.02) but was not present at 2 years (p=.47). There was a trend for time (practice) for AVLT delayed recall (p=.09) at 1 year (p=.58). Significant practice effects were present for BNT at both 1 year (p=.00008) and at 2 years (p=.0002). No suggestion of cognitive performance decline were detected for any measure or follow-up interval. Conclusions: These analyses indicate that not only is there no significant risk to naming and verbal memory with responsive cortical stimulation at 1 and 2 years, there is no suggestion of a differential risk to these important cognitive functions based upon whether stimulation is applied to either MTL or non-MTL structures. The primary goal of treatment with the RNS® System is to reduce seizures in medically refractory patients who are not candidates for surgical resection due to seizure onset location, risk of significant cognitive decline, or having had a prior failed surgical intervention. These data provide reassurance that long-term treatment is not associated with additional cognitive decline of verbal memory or naming.