Abstracts

Rationale: While resective epilepsy surgery offers the best chance of seizure freedom for medically intractable epilepsy, patients with multifocal and wide network epilepsy are often better candidates for neurostimulation treatment including RNS device. However, there are knowledge gaps understanding RNS efficacy based on seizure onset zone, MRI or PET findings , and prior SEEG mapping.

Methods: This is a retrospective study that included patients with medically refractory epilepsy who were implanted with RNS system for a minimum of one year between July 2015 and September 2023. Patients were stratified based on seizure-onset zone ( temporal vs.extra-temporal), presence or absence of a structural lesion on MRI, presence or absence of hypometabolism on FDG-PET, and prior SEEG implantation. The main goals of this study are First, to report overall seizure response rates among among different patient groups. Second, to examine the effect of RNS on the number of disabling seizures , severe seizures , convulsions , and falls per month within one year of RNS implantation compared to prior implantation. Third, to investigate the impact of RNS on generalized tonic-clonic seizures at 3 months, 6 months , and 12 months post RNS implantation compared to 3 months prior.

Results: A total of 25 patients met the inclusion criteria. 76% had temporal lobe epilepsy and 20% had Extra-temporal. 48%had prior surgical intervention. 60% had lesional MRI and 40%had non-lesional MRI. PET scans were positive in 60% with temporal location in 40% and extra-temporal in 20% of the patients. 20% had non lesional PET. Overall seizure control was 96% . 84% had SEEG prior RNS implantation. More than half of the patients had no change in their AED medications. 20% had reduction while 16% had an increase. RNS was  effective in reducing seizures in both temporal epilepsy (95%) and extra-temporal epilepsy (100%). PET lesional patients showed a better seizure control rates (93%) compared to the non lesional group ( 83%) ( 95 % CI 0.14, 0.048, p< 0.05). Regardless to MRI findings, RNS was effective in the lesional and non-lesional group. There was a significant reduction in the number of disabling seizures (95% CI -0.05, -0.11, p< 0.05), severe seizures (95% CI -0.02,-1.04, p< 0.05) , and convulsions ( 95% CI -0.03,-0.98 , p< 0.05) per month within one year of RNS implantation. While the number falls decreased per month ( mean difference -1.7), but it was not statistically significant (95% CI -0.16, 1,4 , p > 0.05). There was significant reduction of the number of GTCs across all time points  post RNS implantation compared to 3 months prior . The mean number of seizures significantly declined from 1.27  three months before RNS to 0.21 one year after RNS.

Conclusions: Overall seizure response rate in our study surpasses what has been reported in other studies. Although limited by small sample size, subpopulation analysis showed similar results regardless to seizure onset zone and MRI findings. However, RNS response rates were better in PET lesional group. Moreover, in our cohort RNS was effective in reducing the number of generalized tonic-clonic seizures.

Funding: none