Manuscripts

Objectives: No reliable biomarkers exist for predicting and assessing vagus nerve stimulation (VNS) response. While VNS induces acute electroencephalography (EEG) desynchronization after implantation, longitudinal evaluations of EEG synchronization changes are lacking. This study constitutes the first prospective investigation evaluating EEG synchronization before and after VNS device implantation and correlating it with the clinical response to VNS.
Materials and Methods: High-density EEG recordings were obtained from 12 adults with drug-resistant epilepsy before and after VNS device implantation (one, three, and six months). EEG resting state (180 seconds), with eyes open and eyes closed (EC), was recorded in VNS ON and OFF conditions. The global weighted phase lag index (wPLI) was computed as an EEG phasesynchronization measure and correlated with the VNS response using various assessment methods, including binary classification (>50% or <50% seizure frequency reduction), percentage of seizure reduction, and the newly developed Clinical-Research Response Scale (CRRS).
Results: We observed a progressive decrease of wPLI in the delta band during the EC VNS OFF condition, which correlated with the VNS response over time, particularly when assessed using the new CRRS compared with other assessment methods. Additionally, a higher preimplant global wPLI predicted a better outcome of VNS, as did an early magnet response.
Conclusions: Overall, VNS may positively influence specific brain states, with a time-dependent evolution of EEG synchronization reflecting therapeutic efficacy. Preimplantation EEG synchronization and an early magnet response may predict VNS response. Moreover, the CRRS could constitute a more sensitive method for characterizing VNS response compared with traditional assessment methods.

There are currently no established biomarkers for predicting the therapeutic effectiveness of Vagus Nerve Stimulation (VNS). Given that neural desynchronization is a pivotal mechanism underlying VNS action, EEG synchronization measures could potentially serve as predictive biomarkers of VNS response. Notably, an increased brain synchronization in delta band has been observed during sleep–potentially due to an activation of thalamocortical circuitry, and interictal epileptiform discharges are more frequently observed during sleep. Therefore, investigation of EEG synchronization metrics during sleep could provide a valuable insight into the excitatory-inhibitory balance in a pro-epileptogenic state, that could be pathological in patients exhibiting a poor response to VNS. A 19-channel-standard EEG system was used to collect data from 38 individuals with Drug-Resistant Epilepsy (DRE) who were candidates for VNS implantation. An EEG synchronization metric–the Weighted Phase Lag Index (wPLI)—was extracted before VNS implantation and compared between sleep and wakefulness, and between responders (R) and non-responders (NR). In the delta band, a higher wPLI was found during wakefulness compared to sleep in NR only. However, in this band, no synchronization difference in any state was found between R and NR. During sleep and within the alpha band, a negative correlation was found between wPLI and the percentage of seizure reduction after VNS implantation. Overall, our results suggest that patients exhibiting a poor VNS efficacy may present a more pathological thalamocortical circuitry before VNS implantation. EEG synchronization measures could provide interesting insights into the prerequisites for responding to VNS, in order to avoid unnecessary implantations in patients showing a poor therapeutic efficacy.