Abstracts

Rationale: The contribution of long-range network properties to ictogenesis remains poorly understood. Seizures induced by electrical stimulation mapping (ESM) have a defined site of initiation in which functional connectivity of remote seizure foci may be evaluated. We analyzed ESM-induced neuronal recruitment with ictal high gamma activity (HGA) as a biomarker of synchronized neuronal population firing to characterize epileptic networks and evaluate the predictive value of recruitment patterns on identification of the spontaneous seizure onset zone (SOZ) and surgical outcome.

Methods: ESM-induced electroclinical seizures with typical semiology in patients implanted with subdural/depth electrodes at three epilepsy centers were retrospectively studied. Ictal HGA was visually analyzed after filtering (FIR, 80-150Hz, order 500) using previously validated methods, defined by onset within 1s after stimulation artifact and repetitive bursting above pre-stimulation baseline for ≥10s. Epileptic networks were defined as the aggregate of seizure-inducing ESM contacts and recruited contacts per patient. To assess if increased functional connectivity determined recruitment patterns, pre-stimulation interictal broadband (1-30 Hz) EEG cross-correlation was measured between ESM and recruited contact pairs, and compared to ESM and non-recruited contact pairs at equivalent Euclidean distances. Outcomes at ≥6mo were assessed using the Engel scale. Fisher Exact, Mann-Whitney, and Wilcoxon signed-rank tests were used for categorical, continuous, and paired variables respectively, two-sided, α=0.05.

Results: There were 33 patients included (21 female, mean age 31yr [range 10-57]). Spontaneous SOZs were neocortical in 24/33 and mesial temporal in 9/33. ESM-induced ictal HGA occurred in 30/33 (91%) of cases, and extended beyond stimulated contacts in 27/33 (82%). The median network size was 5 contacts (range 2-41). Epileptic network contacts overlapped a greater proportion of the spontaneous SOZ compared to stimulated contacts alone (mean 51% vs 31% of SOZ, p=0.047). Recruited and equidistant non-recruited contacts showed similar interictal cross-correlation values with seizure-inducing ESM sites (n=10 patients, mean 0.29 [range 0.06-0.65] vs 0.30 [0.07-0.70], p=0.49). Resection of all SOZ contacts was associated with Class I outcome (OR 13.81, p=0.041), with a trend seen for resection of all ESM-recruited contacts (OR 8.43, p=0.058). There was no association between outcome and number of network contacts (p=0.50).

Conclusions: ESM-induced ictal HGA reveals neuronal recruitment commonly occurring beyond stimulation sites, indicating mechanisms for long-range emergence of ictal activity, often including the SOZ. Identification of recruited sites thus improves prediction of the spontaneous SOZ. Stimulated and recruited sites did not show enhanced interictal functional connectivity; further investigation into mechanisms underlying seizure susceptibility remote from seizure initiation is warranted. Clinical use of ESM as an independent epileptic biomarker should be evaluated with respect to recruited brain areas rather than the stimulated area alone.

Funding: Please list any funding that was received in support of this abstract.: AES RTFC #704541.