Therapy response prediction of focal cortex stimulation based on clinical parameters: a multicentre, non-interventional study protocol.

Abstract

Introduction: A novel focal cortex stimulation (FCS) device has recently received approval in Europe for patients with focal drug-resistant epilepsy (DRE). After 6 months of stimulation, 17 of 32 patients achieved ≥50% reduction in seizure frequency compared with their prestimulation baseline (responders). Currently, there is no established method for predicting FCS treatment response prior to implantation.

Methods and analysis: This is an ongoing combined retrospective-prospective non-interventional multicentre study. Clinical data of up to 100 patients treated with FCS are collected across 20 collaborating epilepsy centres in four European countries. The key outcome parameters, seizure frequency and severity, are measured along with metrics on cognition, mood and quality of life, both pre-electrode and postelectrode implantation. The data are complemented by demographics, medical history and information on antiseizure medication and FCS treatment parameters during the stimulation period. In addition to clinical data, MRI and electroencephalography registrations are used to gain insights into spatial and electrophysiological aspects of FCS. Multivariate statistical and machine learning analyses are employed to identify key predictive biomarkers associated with patient outcomes (responders vs non-responders). The primary goal is to improve counselling for DRE patients by identifying promising candidates for FCS treatment.

Ethics and dissemination: This study has received approval from the ethics committee of the University of Freiburg, Germany (23-1540 S1; 23-1183_1-S1-retro). The same approval is applicable for all participating centres in Germany as part of a multicentre study. Ghent University Hospital, Belgium, has received approval for participation in the retrospective arm from their local ethics committee (ONZ-2024-0168). The final approvals for the participating Swiss and Austrian sites are still pending. The results will be made available to the public through peer-reviewed journals and conference presentations.