Stereo-electroencephalography Performance in Bilateral Independent/Unclear Scalp Seizures

Scalp electroencephalography (EEG) may reveal bilateral independent or unclear (BI/U) ictal onset patterns in patients with focal drug-resistant epilepsy, presenting a challenge to surgical decision-making. The utility of stereo-electroencephalography (SEEG) in this subgroup, particularly the probability of delineating a single seizure onset zone (SOZ) that would permit curative resection, remains poorly understood. This study examined whether BI/U scalp EEG findings could predict SEEG outcomes in this population.

We conducted a retrospective cohort study of consecutive patients with focal drug-resistant epilepsy and BI/U ictal onset on scalp EEG who underwent SEEG evaluation at the London Health Sciences Centre (Ontario, Canada) between January 2012 and December 2024. All patients had undergone non-invasive and invasive presurgical assessments. Surgical outcomes were determined using the Engel classification following at least one year of postoperative follow-up. A blinded decision validation sub-study was also performed. Blinded to actual outcomes, the team made decisions regarding SEEG and surgical interventions when patients found to have a single SEEG SOZ were presented. Responses were stratified to inform the added diagnostic value of SEEG.

Of 255 SEEG cases screened, 84 patients (33%) met inclusion criteria. The cohort was 56% female, with a median seizure onset age of 12 years (IQR 6–20); 65.5% had temporal lobe epilepsy (TLE). A single SOZ was identified in 14.3% of cases (TLE: 14.5%, extratemporal: 13.8%). Patients with a single SEEG SOZ were found to have shorter recording durations (mean of 11 vs 15.79 days in those with multifocal SEEG SOZs; p=0.009). Curative focal resections were performed in 12% ( n= 10), with long-term Engel I outcomes achieved in one patient (1.2%). Palliative resections occurred in 26% ( n= 22), with Engel I outcomes in 7% ( n= 6). In 50% of the blinded cases, the epilepsy surgery team reported that they would not have recommended SEEG based on phase I data.

These findings suggest that patients with BI/U scalp EEG SOZs may be associated with a low likelihood of identifying a single SEEG SOZ and curative outcome. Using BI/U scalp EEG ictal onset as a predictor in preoperative decision-making will assist in refining SEEG candidate selection in this large subgroup.