Long-term efficacy and predictors of vagus nerve stimulation in drug-resistant epilepsy: a multicenter cohort study.

Background: At present, a number of indicators have been analyzed for the relationship with the efficacy of vagus nerve stimulation (VNS) in drug-resistant epilepsy (DRE) patients, but there is still no definite predictor of efficacy. This study is to assess the long-term effectiveness and predictors of VNS in DRE patients.

Methods: We analyzed DRE patients monitored for over a year post-surgery (2016-2019) to evaluate VNS outcomes. Logistic regression was used to identify efficacy predictors.

Results: Out of 162 DRE patients with VNS, 99 were followed for over 12 months, 80 for over 24 months, and 70 for over 36 months. At 12 months, 33 (33.4%) showed effectiveness, including 7 (7.1%) who were seizure-free. At 24 months, 32 (40.0%) were effective, including 12 (15.0%) who were seizure-free. At 36 months, 36 (51.4%) were effective, including 11 (15.7%) who were seizure-free. After 5 years, 27 (55.1%) were effective, including 8 (16.3%) who were seizure-free. Multivariate regression analysis identified structural etiology as a predictive factor for the effective VNS treatment (P = 0.039, OR = 0.35 [0.13-0.95]).

Conclusions: VNS effectively controls seizures, with effectiveness and seizure-free rates improving over time. Patients with structural factors are at higher risk of ineffective VNS, suggesting epilepsy etiology may predict VNS outcomes.