Abstracts

Seizure Frequency Correlate to Hippocampal Subfield Volume of Tissue Activation in Responsive Neurostimulation

Abstract number : 2.315
Submission category : 9. Surgery / 9A. Adult
Year : 2023
Submission ID : 930
Source : www.aesnet.org
Presentation date : 12/3/2023 12:00:00 AM
Published date :

Authors :
Presenting Author: Lesley Kaye, MD – University of Colorado School of Medicine

Trey Jouard, MS – Neurology – University of Colorado School of Medicine; John Thompson, PhD – Neurosurgery – University of Colorado School of Medicine

Rationale: Treatment with responsive neurostimulation (RNS) results in variable efficacy, with some patients considered “non-responders," indicating less than 50% seizure reduction, while other patients are “super-responders,” referring to a >90% seizure reduction. The cause of this disparity in response is not well understood. It is thus imperative to develop better modelling of stimulation parameter sets to improve neuromodulation efficacy and to determine which focal and network brain targets best respond to neuromodulation.

Methods: We retrospectively reviewed and processed brain imaging that is collected as standard clinical care and correlated volume of tissue activation (VTA) to clinical outcomes and long term electrocorticography (ECog). A University of Colorado IRB protocol is currently approved for use of this clinical data. Subjects were identified from the University of Colorado Comprehensive Epilepsy Program RNS database. Patients with at least one hippocampal depth electrode and greater than one year of ECog data without significant gaps were included. Patients with extratemporal RNS electrodes were excluded. Lead-DBS toolbox software was used to reconstruct lead placement from pre- and post-operative MRI and CT imaging and simulate a VTA area for each electrode contact. We then wrote a program to visualize and calculate the percent overlap of each VTA area with hippocampal subfields. CoBrALab Merged Atlas was utilized for hippocampal subfield identification. Seizure rates, based on both retrospective chart review and recorded RNS ECog data will be correlated to the percent overlap of each region.

Results: A total of 10 patients were identified, five each in super-responder and non-responder categories. For subject 1 (a non-responder), VTA and % subfield overlap were calculated for 2.5 µC/cm² and 5.1 µC/cm²charge density (CD) in a cathodal monopolar stimulation pathway. We investigated two epochs: A and B. Duration of programming for each epoch was 100 days and 105 days respectively. Seizure diary reported six seizures per month at lower CD and daily seizures at higher CD. Figure 1 shows fusion of hippocampal depth lead reconstruction to MRI and hippocampal segmentation with electrode contact VTA overlap. Final analysis of seizure frequency correlate to VTA % overlap of hippocampal subfield for all ten subjects is ongoing.

Conclusions: Utilizing an open access software program (Lead-DBS) to apply VTA modelling to RNS hippocampal depth electrodes may help elucidate ideal targets within the hippocampus and improve outcomes for patients who are non-responders to hippocampal responsive neurostimulation.

Funding: Intradepartmental Grant, University of Colorado Dept of Neurology

Surgery