Abstract:INTRODUCTION: Approximately 30% of epilepsy patients are refractory to pharmacotherapy. Patients with drug-resistant epilepsy (DRE) with seizure foci located bilaterally or in the eloquent cortex may be ineligible for resective surgery. For such cases, responsive neurostimulation (RNS) is a viable treatment strategy. METHODS: Retrospective analysis of 23 patients treated with thalamic RNS of the centromedian (CM) and anterior nuclei of the thalamus (ANT) from 2015-2022. Clinical features and seizure outcomes were recorded. RNS electrodes were used to model the electrical field using the parameters at last follow up. Patient-specific probabilistic tractography was used to define the network of the seizure onset zone (SOZ) and its thalamic involvement. Each patient’s probabilistic map was used to define the area of highest connectivity of the thalamus with the seizure network. Each of these patient-specific thalamic areas demonstrating highest connectivity with the seizure network were correlated with electrode position, stimulation modeling, and seizure outcomes. RESULTS: Seizure network from neocortical areas were highly connected to the CM and from limbic cortical areas to the ANT, MD, and parafascicularis nuclei. Results of the correlation analysis between connectivity, stimulation modelling of the RNS electrode, and seizure outcomes will be presented along with illustrative cases of patient-specific targeting using tractography for RNS placement. CONCLUSIONS: Probabilistic tractography is useful to define the networks associated with improved seizure control in thalamic RNS. The identification of these networks prior to surgery could help to identify specific corticothalamic connections and their anatomical trajectories to optimize the implantation of the leads for better and more predictable seizure control.

483 Network Analysis of the Thalamic RNS in Patients With Drug Resistant Epilepsy: Insights to Improve Surgical Targeting