Abstracts

Rationale: In the context of assessing candidacy for epilepsy surgery, direct cortical electrical stimulation (DCES) via intracranial EEG electrodes remains the most powerful method to map eloquent cortex. However, the procedure is invasive and carries a risk of patient discomfort and of stimulation induced seizures. Here, we present the results of a large number of such stimulations, centred on the medial frontal wall, and ask if the range of elicited responses is broad enough to potentially map different functional zones within this region and also establish the safety profile of this technique within our centre. Methods: We reviewed 147 consecutive admissions to the National Hospital for Neurology and Neurosurgery, London, UK for prolonged intracranial EEG recording to establish surgical candidacy, in the context of drug-resistant focal epilepsy. From these, we selected only patients who underwent electrical stimulation of electrodes within the medial frontal wall and audited the stimulation locations, currents, response types and side-effects. Electrodes were surgically implanted using a frameless technique and locations verified post-operatively with CT imaging. Stimulations were monopolar (to a distant reference) or bipolar (with a neighbouring electrode) delivered as trains of 50Hz biphasic square-wave pulses with pulse width of 500 microseconds and an amplitude ranging from 0.5mA to 7mA. Results: A total of 46 patients had electrodes implanted to the medial frontal wall, and 38 of those (29M, 9F) underwent DCES. The mean age (+/-SD) was 32.4 (10.3) years, mean age of epilepsy onset was 11.8 (9.2) years. The total number of sites stimulated was 537, but stimulation at 8 sites resulted in seizures (1.5% per site) whilst stimulation at 52 sites resulted in after-discharges. Therefore, functional data is reported for the remaining 477 stimulation sites, whose locations are shown in Figure 1. We assessed for positive and motor symptoms and used a reading tasks to screen for speech difficulties. A clinical response was seen in 252 electrode positions (52%). Positive motor responses were the most frequent response (n=141, 31.2%), followed by somatosensory responses (n=43, 9.5%), speech disturbance (n=38, 8.4%) and negative motor responses (n=30, 6.6%). A mean (+/-SD) stimulation current of 2.35 (1.12) mA was required to elicit a positive motor response, 2.81(1.35) mA to elicit a negative motor response, 2.03 (0.77) mA to elicit a somatosensory response and 2.79 (1.02) mA to affect speech. Conclusions: DCES of the medial frontal wall carried a risk of seizure of 1.5% per site stimulated. A range or positive and negative motor, sensory and speech effects, which may be useful in demarcating different functional zones within the medial frontal wall, are elicited at low stimulation intensities of under 3mA with positive motor and sensory responses at the lowest intensities. Funding: This work was undertaken at UCLH/UCL who receive a proportion of funding from the Department of Health's NIHR Biomedical Research Centres funding scheme. AJ was funded by a Guarantors of Brain fellowship.