Authors :
Presenting Author: Venethia Danthine, MD, PhD student – UCLouvain
Lise Cottin, PhD student – Université Libre de Bruxelles (ULB); Enrique Ignacio Germany Morrison, PhD – UClouvain; Jean Delbeke, MD, PhD – UClouvain; Giulia Liberati, PhD – UClouvain; Antoine Nonclercq, PhD – Université Libre de Bruxelles (ULB); Riëm El Tahry, MD, PhD – UClouvain
Rationale: Previous studies have shown that Vagus Nerve Stimulation (VNS) induces acute electroencephalogram (EEG) desynchronization. Laryngeal Motor Evoked Potentials (LMEPs) are markers of efferent A-fiber activation and can be recorded using two surface electrodes on the ventral surface neck. Currently, no reliable biomarker for optimizing VNS treatment exists. In addition, there is no information available on how (i) vagus nerve recuperates after surgery, (ii) VNS-induced EEG desynchronization evolves postoperatively, and (iii) LMEPS change after the implantation. This study aims at assessing longitudinally efferent and afferent markers of VNS action.
Methods: Adult patients with drug-resistant epilepsy, candidates to VNS implantation, were prospectively recruited for the current study. EEG was recorded one to two months before surgery (V1). LMEP and EEG recording were acquired two weeks after surgery (V2-LMEPs recording); as well as one month (V3), three months (V4) and six months (V5) after VNS implantation. For the LMEPs recordings, trains of 7s were delivered at increasing current, up to the routine stimulation intensity +0.25 mA allowing to build dose response curves. EEG recordings were performed using a 64-channel cap. For the EEG part, clinical parameters of stimulation (intensity, pulse width, frequency) were kept stable and 180s of EEG with eyes open (EO) and 180s of eyes closed (EC) were recorded. Weighted Phase Lag Index (wPLI – a connectivity metric) was computed in each frequency band (delta, theta, alpha, beta and broadband) during VNS ON and VNS OFF epochs. Using paired t-tests, these values were compared between conditions (VNS OFF/VNS ON) and across visits.
Results: So far, six patients were included (three females, three males) and completed V1, V2 and V3; five patients finished V4 and three patients completed the full study (V5). LMEPs were recorded in patient one (V2,V4,V5), patient three (V4,V5) and patient six (V2,V3,V4) while no LMEPs were recorded in other patients yet (patient two, four, and five). Based on these available data, intensity threshold leading to LMEPs induction showed to decrease over time in one patient (patient one) (V2-0,375mA; V4/V5-treshold 0,25mA) while no change was observed in the two others (patient three, six) (threshold 0,375mA). In 6/6 patients, a significant higher whole brain wPLI was found in V3-V4 compared to V1 during VNS ON in the delta band (p=0,021) and in the theta band (p=0,049) respectively, in EO and EC conditions, while no significant effect was found in the other bands or during the VNS OFF condition. However, a trend can be observed between V1-V3/V4 (p= 0,097) and V4-V5 (p=0,097) in delta band in EC conditions (Fig.1).
Conclusions: LMEPs may depend not only on the intensity of stimulation but also on the recovery of the nerve. VNS may (i) increase wPLI in the initial months after surgery and (ii) decrease wPLI after six months in delta and theta bands. These preliminary results may reflect modulatory effects of VNS over time.
Funding: This work was supported by Fond de la Recherche Scientifique (F.R.S.-FNRS) with a FRIA grant.