Authors :
Guy Gurevitch, MA – Tel Aviv Sourasky Medical Center
Noam Fallik, MD – Tel Aviv Sourasky Medical Center
Segev Gabay, MD – Tel Aviv Sourasky Medical Center
Ido Strauss, MD, PhD – Tel Aviv Sourasky Medical Center
Presenting Author: Firas Fahoum, MD, MSc – Tel Aviv Sourasky Medical Center
Rationale:
Vagus Nerve Stimulation (VNS) is an approved palliative treatment option for drug-resistant
focal epilepsy (DRE). In this patient population, VNS results in 30-50% long-term seizure reduction, yet a substantial number of DRE patients do not respond to VNS. To date, no biomarkers or clinical features are associated with a favorable response to VNS. Previous studies show that the neuromodulatory effect of VNS is related to functional connectivity (FC) changes between different cortical and subcortical brain hubs. We hypothesized that specific pre-implantation resting-state functional MRI (RS-fMRI) FC features are related to clinical response to VNS.
Methods:
We included all consecutive adult drug-resistant focal epilepsy patients who underwent pre-surgical 3 Tesla fMRI scans and eventually underwent VNS implantation. The frequency of disabling seizures was quantified from clinical charts before implantation and a year after reaching therapeutic current. Responders were defined by a 50% reduction in disabling seizures. A region-based FC (ROI-ROI FC) analysis was applied to RS-fMRIs from the patient cohort seeking group differences between responders and non-responders. RS-fMRIs of 73 healthy volunteers were similarly analyzed and used as a control group. Outcome measures were compared between groups and evaluated within the responder group against a continuous measure of seizure reduction.
Results:
We included 55 patients: 19 responders and 36 non-responders to VNS (35% and 65% respectively). ROI-to-ROI FC analysis revealed significantly increased connectivity in responder compared to non-responder groups, between regions in the primary visual cortex and frontal cortex regions including primary and associative motor areas (corrected for multiple comparisons using spatial pairwise clustering, pFDR< 0.05). Average visual-motor FC was extracted for all patients and controls, and a one-way ANOVA found a significant group effect (p< 0.001). Post-hoc paired comparisons showed increased connectivity in responders compared to both non-responders and controls (p< 0.001), but no differences between controls and non-responders (p=0.58). Finally, visual-motor FC within the responder group was significantly correlated with the seizure frequency reduction in this patient group (r=-0.68, p=0.003), showing that higher connectivity resulted in a greater reduction in post-VNS seizure frequency.