Abstracts

Rationale:
Responsive neurostimulation (RNS) is an FDA approved therapy for patients with refractory mesial temporal lobe epilepsy, where stimulation is delivered in response to recorded electrographic activity in order to disrupt seizures. Although most patients receiving RNS therapy experience a decrease in seizure frequency, clinical outcomes are variable across patients and few patients receive long-term seizure freedom. It is currently unknown which structures or networks within the temporal lobe should be modulated to achieve seizure freedom. We hope to improve the efficacy of neurostimulation for patients with refractory epilepsy by identifying predictors of seizure reduction and implementing novel stimulation approaches that lead to seizure arrest. The objective of this study was to determine if stimulation location in the mesial temporal lobe is predictive of seizure reduction.
Method:
Retrospective data was collected for a cohort of patients receiving bilateral mesial temporal RNS® (n=13) including pre-operative structural MRI and diffusion-weighted imaging, post-operative CT, stimulation settings, and clinical outcomes. Volumes of tissue activated (VTA) were estimated using computational models that integrated patient-specific imaging and stimulation parameters.1 To compare across patients, active contact locations and VTA were mapped to MNI space and the VTA from both hemispheres were mapped to a single hemisphere (n=26 VTA). To determine whether stimulation location was associated with seizure reduction, we generated a stimulation atlas of the average patient-reported seizure reduction from their last clinical follow-up compared to baseline at each voxel.
Results:
The active electrodes were in the amygdala (11/104), hippocampus (57/104), parahippocampal gyrus (13/104), and adjacent white matter (23/104) (Figure 1A). Figure 1C shows maximal overlap of the VTA across subjects occurred in the inferior amygdala and the long axis of the hippocampus. Initial spatial trends in seizure reduction based on the VTA suggest that activation of the amygdala and anterior hippocampus correspond to increased seizure reduction (Figure 1C). The median patient-reported seizure reduction for this cohort was 78% (Figure 1B).
Conclusion:
Overall, these initial results suggest there may be a link between activation of the amygdala/anterior hippocampus and seizure reduction. This finding is consistent with regions implicated in seizure freedom following laser ablation.2 Despite these promising results, we will also incorporate electrophysiological recordings and network analyses in order to develop a clearer picture of potential predictors of seizure reduction. A better understanding of the local and network effects of RNS in the mesial temporal lobe will hopefully guide novel patient-specific programming strategies that will lead to long-term seizure freedom. 1. Butson. Neuroimage. 2007; 34: 661-670. 2. Wu. Epilepsia. 2019; 60: 1171–1183.
Funding:
:NSF Graduate Research Fellowship 1747505; “Center for Integrative Biomedical Computing” NIH P41 GM103545