The Importance of Impedance: A Case of Responsive Neurostimulation Failure Characterized by Concurrent Stereoelectroencephalography.

Summary: Epilepsy neuromodulation treatment failure is a significant challenge, with multiple possible causes. The responsive neurostimulation (RNS) system delivers stimulation from a single current source, and the relative flow of the electrical current through each stimulating contact is inversely proportional to the relative impedance of each contact. Current shunting through low-impedance contacts (i.e., intraventricular contacts) can divert therapy away from the intended targets and may be a cause of treatment failure. We present a case of a patient with bitemporal epilepsy and bitemporal encephaloceles, with poor response to bilateral mesial temporal RNS, who completed stereotactic EEG (sEEG) monitoring to investigate the possible causes of treatment failure. The sEEG was safely completed without damaging the RNS device. The sEEG recorded independent bitemporal interictal epileptiform discharges and seizures, which did not arise from sampled encephalocele regions. The sEEG-recorded RNS stimulation artifact was reduced in the left mesial temporal region relative to the right, which suggested potential current shunting through the right-sided contacts. Impedance measurements confirmed several low-impedance contacts from the right lead, with associated intraventricular position on imaging. At last follow up, 161 days after replacement of the right lead, the patient experienced an additional 58% reduction in seizure burden. Effective therapy delivery by single-current-source neurostimulation systems, such as RNS, critically depends on relative electrode impedances. Current shunting through low-impedance contacts is an underappreciated potential cause of treatment failure. Routine impedance assessments and individualized stimulation programs are recommended to avoid unintended current diversion. Concurrent sEEG monitoring and active RNS are feasible and can characterize stimulation effects.