Abstracts

Rationale: Neurostimulation devices can reduce seizure frequency in drug-resistant epilepsy. Responsive neurostimulator (RNS) treatment provides closed-loop seizure detection and subsequent electrical brain stimulation with four-contact electrodes placed directly onto or into the brain. While electrical stimulation may trigger symptoms when performing cortical mapping, this is not acceptable when treating abnormal epileptiform activity and seizures with RNS. Initial studies documented that successful RNS treatment is associated with several hundred stimulations per day, but not with stimulation-triggered symptoms (STS) other than rare photopsia.We identified 3 patients with RNS STS at our institution who required adjustment of stimulation settings and electrode position. Here we aim to document the patient experience, location of electrodes, and the successful modification of stimulation settings and montages to eliminate STS that do not serve a therapeutic purpose. We also provide a practical guide to avoid and address STS.  Methods: This is a retrospective case series. Data were extracted from chart review, including age, age of epilepsy onset, results of intracranial monitoring and cortical mapping, RNS electrode placement, and changes made to RNS stimulation settings. Results: Case 1: A 50-year-old (yo) right-handed (RH) woman with seizures since age 5 months and dysplastic cortex in the left frontal lobe. Stereo-EEG monitoring (SEEG) and cortical mapping revealed close proximity of seizures and hand motor function. RNS depth electrodes were placed along three trajectories associated with ictal patterns. In the operating room (OR), under anesthesia RNS stimulation was performed along all electrodes at 6 mA (charge density of 3 μC/cm2), and only one lead was associated with right hand movement. RNS was connected to the other two electrodes. Increasing current in clinic has not led to any motor symptoms. (Figure 1A)Case 2: A 35-yo left-handed woman with seizures since age 19 years and nonlesional imaging. SEEG showed ictal patterns in the dominant right posterior temporal region. Stimulation caused a warm sensation over both eyebrows and nausea, related to a strip electrode over the right temporo-parietal convexity. Symptoms were suppressed by changing stimulation montage from current flow between electrodes to current flow between each electrode and generator, reducing stimulation frequency (SF) from 200 Hz to 100 Hz and pulse width (PW) from 160 μs to 120 μs. (Figure 1B)Case 3: A 20-yo RH man with seizures since age 4 months and a dysplastic lesion in the left frontal lobe. SEEG at an outside institution led to a left frontal resection and placement of 6 RNS depths electrodes in the left frontal and parietal lobes. Initial settings did not produce any symptoms, but increasing current to >1mA, caused right thigh twitching with stimulation. Movements were abolished by identifying the contacts involved and removing two contacts from the stimulation montage, reducing SF to 100 Hz and PW to 120 μs.Table 1 provides a practical approach to modifying RNS settings and eliminating STS. Conclusions: Placement of RNS electrodes should take into account that STS can occur when placement is close to eloquent cortex. At electrode placement, stimulation under anesthesia can provide information on whether motor symptoms are triggered, and electrodes in other positions can be used, if more than two were placed. STS encountered when increasing current in clinic can be eliminated by identifying involved electrodes, reducing SF and PW, and removing contacts from montage, if needed. Funding: No funding