Prospective Tractography-guided Nucleus Targeting for Thalamic RNS in Drug-resistant Epilepsy

Introduction: Thalamic responsive neuromodulation (RNS) is an increasingly used treatment for patients with drug-resistant epilepsy (DRE). Optimal therapeutic stimulation relies on proper selection of the thalamic target, which is currently done based on seizure semiology. We recently presented a retrospective study showing greater seizure onset zone (SOZ) to volume of tissue activated (VTA) correlated to better seizure outcomes. Here, we present a cohort of patients wherein tractography was used during the sEEG mapping phase to identify thalamic connections to the seizure onset zone (SOZ). This structural connectivity was used to inform selection of the thalamic nucleus for neuromodulation.

Methods: 5 patients received pre-SEEG DTI MRI imaging. After implantation of sEEG, seizures were recorded and used to identify SOZ. The same DTI imaging was used to conduct individualized probabilistic tractography to define the SOZ network and its thalamic involvement. Each patient’s probabilistic map was used to define the area of highest connectivity of the thalamus with the SOZ. This area was then incorporated into the multidisciplinary conference to decide optimal thalamic nucleus for RNS electrode implantation. Outcomes are recorded after minimum one year follow up.

Results: Patients with minimum one-year follow up (n=4) were responders (>50% reduction). Remaining clinical outcomes, results of the correlation analysis between connectivity, stimulation modeling of the RNS electrode, and seizure outcomes will be presented along with illustrative cases of patient-specific targeting using tractography for RNS placement.

Conclusion : We demonstrate the potential value of thalamic seizure network mapping to perform personalized selection of thalamic RNS targets, as opposed to typical semiology-based selection. Incorporating tractography into standard RNS workflow may lead to consistent therapeutic response in patients. The identification of these networks prior to surgery could help identify specific corticothalamic connections and their anatomical trajectories to optimize the implantation of the leads for better and more predictable seizure control.