Single-center retrospective analysis of responsive neurostimulation outcomes

Responsive neurostimulation (RNS) is a promising treatment for drug-resistant epilepsy (DRE). RNS has been shown to be safe and effective in reducing seizure frequency in patients with DRE.

We performed a retrospective cohort study of patients with DRE implanted with the RNS system at our institution between 2017 and 2024. The patients had at least 6 months of clinical follow-up and two RNS programming sessions postimplantation. We collected data on patient demographics and postimplantation seizure frequency. Efficacy was assessed as the median percent change in postimplantation seizure frequency compared to baseline, as well as responder rate defined as >50% reduction in seizures. A one-way analysis of variance test was used to assess differences in mean postoperative seizure frequency across all lead placement groups, and Fisher's exact test was used to analyze differences in the proportion of patients achieving a postimplantation Engel classification of ≤2B.

Thirty patients were implanted with the RNS system at our institution between 2017 and 2024. Seventy-three percent of patients were male (n = 22), and the median age at RNS implantation was 36.5 years (interquartile range: 25-46). The mean seizure frequency reduction was 71.4 (±20.9), with a proportion of patients achieving a >50% seizure frequency reduction (responder rate) at 70% postimplantation, four patients had Engel 1A/1B outcome, 17 patients had Engel 2B outcome, and nine patients had achieved Engel 3A outcome. None of the patients had Engel 4 outcome at the last follow-up visit. Interestingly, the use of cenobamate had a statistically significant relationship with achieving a postoperative Engel 2B or better classification (P = 0.03) as well as prior vagal nerve stimulation placement (P < 0.001).

RNS system is a safe and effective treatment for DRE. The responder rate in our cohort is consistent with the seizure outcomes reported in the literature. Additional studies need to characterize the differences between RNS responders and nonresponders.