Abstracts

Rationale: Canine epilepsy is a naturally occurring neurological disorder that shares features with human epilepsy. Dogs with epilepsy present a valuable pre-clinical model for implantable human devices. Electrical brain stimulation (EBS) is an established therapy for drug-resistant epilepsy, yet seizure-free outcomes are rare, and optimizing stimulation parameters can be a lengthy process. Stereotactic EEG (sEEG) monitoring provides an opportunity to trial multiple stimulation parameters and evaluate EBS effects on electrophysiological biomarkers, such as interictal epileptiform discharges (IES). However, EBS artifacts superimposed on intracranial EEG (iEEG) traces hinder biomarker analysis on the stimulation channels.

Methods: The Cadence Neuroscience workstation is a stimulating and recording bedside device for trial EBS, featuring artifact rejection (AR) capable of removing EBS artifacts from stimulating channels, thus enabling biomarker analysis even during stimulation. A female dog with generalized convulsive seizures and interictal generalized epileptiform activity was implanted with 4 leads (16 total contacts) targeting the bilateral dorsal-medial nucleus of the thalamus (DM), centromedian nucleus of the thalamus (CM), and pulvinar nucleus of the thalamus (Pul). Trial EBS was utilized intraoperatively to test different EBS paradigms. Implanted electrodes were connected to the novel implantable neural sensing and recording (INSR) device.

Results: Intraoperative recording and electrical stimulation revealed generalized bilateral interictal epileptiform spikes. A 5-second burst of high-frequency CM EBS (100 Hz, 90 us, 1.5 mA) modulated the iEEG with suppressed IES activity (baseline: 30.5 ± 0.50; modulation: 1.5 ± 1.5; recovered: 23.67 ± 4.19 spikes per 10 seconds). iEEG and IES returned to baseline after 30 seconds of no EBS. Additionally, INSR demonstrated the capability to longitudinally monitor iEEG and deliver therapeutic EBS. The data showed variable brain excitability dependent on behavioral state, with evoked response potentials and IES rates varying with sleep state (non-REM: 59.36 ± 56.83 vs. Awake: 1.21 ± 3.78 spikes per 10 minutes, p < 0.001). Both the Cadence Neuroscience workstation and INSR were effective in rejecting stimulation artifacts. The workstation artifact rejection was demonstrated also in 13 human subjects wundergoing sEEG monitoring.