Abstract
The value of stimulation-induced seizures for multimodal determination of the epileptogenic zone in preoperative epilepsy diagnostics has not yet been sufficiently investigated. Patients with focal pharmacorefractory epilepsy who underwent invasive electroencephalography with cortical 50 Hz stimulation at the Epilepsy Center Erlangen between 2018 and 2023, had at least one stimulation-induced seizure, underwent resective epilepsy surgery, and had a postoperative follow-up ≥ 1 year were analyzed. 20 patients were included, 11 (55.0%) with temporal, 7 (35.0%) with frontal and 2 (10.0%) with parietal lobe epilepsy. 12 patients (60.0%) had a good Engel outcome (Engel 1A). Associated with a good vs. poor (Engel 1B-4) surgical outcome were not only the percentage of resected electrode contacts of the spontaneous seizure onset zone, SOZ (p = 0.005), but also the stimulation SOZ (p = 0.022), as well as stimulation-induced seizure with a typical seizure semiology (p = 0.033), the electrodes inducing a stimulation-induced seizure (p = 0.014), electrodes with an identical seizure onset pattern (p = 0.035), and the occurrence of low voltage fast seizure onset pattern, LVFA (p = 0.015). ROC analyses showed that the AUC for the predictors of the spontaneous SOZ were greatest for the stimulation SOZ (AUC 0.876) and stimulation-induced seizures with LVFA (0.860). Analysis of combined predictors showed higher odds of predicting SOZ for combinations including LVFA. Electroclinical stimulation seizures have prognostic value in determining the epileptogenic zone. Characteristics such as the seizure onset zone, seizure pattern and stimulation seizure semiology predict seizure freedom in case of resection of electrode contacts. Electrodes should be resected where both stimulation seizures have been induced or the seizure pattern has been localized and low voltage fast seizure pattern has occurred.