Authors :
Presenting Author: Gabriel Reyes, BS – Baylor College of Medicine
Liam Ferreria, BS – Baylor College of Medicine; Zulfi Haneef, MD – Baylor College of Medicine; Adel Hassan, BS – Baylor College of Medicine; John Myers, PhD – Baylor College of Medicine; Sameer Sheth, MD, PhD – Baylor College of Medicine
Rationale:
Epileptiform discharges are associated with increased neuronal firing at specific phases of narrowband gamma oscillations, attributed to the activity of inhibitory interneurons and the amplification of gamma oscillations within the ictal core. Disruptions in theta-gamma coupling have been associated with brain pathology. Deep brain stimulation (DBS) at 5 Hz within the medial septum may reduce spontaneous recurrent seizures, underscoring the significance of theta rhythms in seizure inhibition. Optimizing neuromodulation approaches, refining device programming, and identifying neurophysiological biomarkers are vital for enhancing therapeutic outcomes and predicting responses in patients.
Methods:
We analyzed data from twelve patients with drug-resistant epilepsy with bilateral hippocampal Responsive Neurostimulation System (RNS) implantations at a single site. The data collected included preoperative clinical information, monthly seizure counts, electrocorticography (ECoG) data, and RNS stimulation parameters. Multiple linear regression was employed to investigate predictive factors for monthly seizure count, considering average monthly power spectral density, RNS charge density, and time in therapy. ECoG data consisted of scheduled 90-second recordings from four distinct RNS channels located in the left-anterior hippocampus (LAH), left-posterior hippocampus (LPH), right-anterior hippocampus (RAH), and right-posterior hippocampus (RPH). Monthly assessments of oscillatory power at each hippocampal electrode were derived by averaging the power spectral densities (PSDs) of the scheduled recordings for theta (4-7 Hz) and gamma (31-100 Hz).
Results:
The average seizure reduction for patients was 49.96% relative to their pre-implant baselines. We found that monthly seizures were associated with statistical interactions between gamma and theta oscillations in most regions of the hippocampus, indicating that higher amplitudes of both oscillations were associated with increased monthly seizures
(p < 0.001
). Lower gamma power coupled with higher theta power was correlated with reduced monthly seizure seizures
. Neither RNS charge density nor the duration of RNS therapy independently influenced monthly seizure counts
(p > 0.05
). Instead, the
interaction between charge density and time significantly reduced monthly seizure count across patients. Preliminary mediation analyses suggest that the seizure-reducing effects of RNS therapy in this patient cohort were mediated by the effects of charge density on theta/gamma interactions (
p <