Predicting success of vagus nerve stimulation (VNS) from interictal EEG

Abstract Purpose Vagus nerve stimulation (VNS) has shown to be an effective treatment for drug resistant epilepsy in numerous patients, however, not in all. It is still not possible to predict which patients will profit from VNS. In this pilot study, we explore predictive interictal EEG features for seizure reduction after VNS. Methods 19 Patients with medically refractory epilepsy and an implanted VNS system were included. Interictal EEG registrations, recorded before implantation, were retrospectively analysed. A quantative symmetry measure, the pair wise derived brain symmetry index (pdBSI), was tested to predict VNS outcome. Reduction in seizure frequency was used to define the responders. Results 10 Patients did respond to VNS, of whom 7 patients had a seizure reduction of at least 50% in a follow-up period of 2 years. On average, we find higher pdBSI values for delta, theta, alpha and beta bands for non-responders than for responders. The average pdBSI of the theta and alpha bands could significantly discriminate between responders and non-responders. Conclusion In this study, quantifying EEG symmetry using the pdBSI shows promising results in predicting the reduction of seizure frequency after VNS treatment.