P09-S Searching for neurophysiological markers of VNS response predictors: PLI quantification during VNS stimulation

Background Neuronal mechanisms of Vagus nerve stimulation (VNS) efficacity in seizure control are not clearly understood. In animal models as in patients, VNS seems to produce desynchronization of interictal electroencephalography (EEG). We aim to confirm the role of EEG desynchronization as possible marker of VNS efficacity. Material and methods We prospectively selected all EEG recordings of patients with drug resistant epilepsy treated by chronic VNS, recorded in our epileptology center from Jan 2017 to May 2018. EEG signals were acquired during a 30 min eyes-closed resting state period, using a 21 channel EEG system. Phase lag index (PLI), was used to calculate the functional connectivity strength on the first 16 artifact-free epochs for each condition (ON et OFF). Analyses were performed over a broad-filtered signal (0,5–70 Hz) and besides overs different sub-bands filtered signals. Patients were assigned to two different groups, responders (R: reduction > 50% in seizure frequency) and non-responders (NR) Results Eighteen patients were included. Lower synchronization values were highlighted in R patients, compared to NR, in broadband signal and in sub-bands filtered signals. These differences were significantly demonstrated for alpha (p = 0,03, Mann–Whitney) and beta (p = 0,01, Mann–Whitney) bands. Lower synchronization values were as well lower in condition ON compared to condition OFF. Conclusions These results confirm that the beneficial effect induced by VNS in seizure control is related to global interictal EEG desynchronization and that functional connectivity parameters could be used in clinical practice as potential predictor of VNS efficacity.