Abstracts

Rationale: Invasive monitoring with stereo electroencephalography (SEEG) is an important diagnostic tool for further characterization of the epileptogenic zone (EZ) in children with intractable epilepsy. However, invasive monitoring can be costly, necessitates additional hospitalization for the patient, and places them at greater risk for seizures during monitoring. The goal of this study is to examine the pre-admission variables that predict the duration of SEEG monitoring in the pre-surgical evaluation of pediatric patients. Methods: Duration of monitoring and onset to first clinical seizure in children admitted for SEEG at Cincinnati Children's Hospital Medical Center between December, 2014 and April, 2019 was evaluated retrospectively. The pre-admission variables were gender, prior epilepsy surgical resection, duration (days) of scalp EEG monitoring during Phase 1, number of seizure days on scalp EEG during Phase 1, onset to first seizure (hours) during Phase 1, ratio of seizure days during Phase 1 to total days during Phase 1, and use of steroids longer than 24 hours after initial SEEG implantation. Monitoring duration was further dichotomized prior to analysis to <= 7 days and > 7 days to plot a survival probability (KM) curve for the cohort as a function of onset to seizure during SEEG. Multiple regression analysis was used to assess for predictors of duration and onset to seizure in SEEG. Results: The study included 51 patients (24 females, mean age 10.8 ± 5.9 years). The mean duration of SEEG monitoring was 7.8 ± 6.6 days with a range of 2-28 days. Only the number of seizure days on scalp EEG during Phase 1 was predictive for the entire cohort (p = 0.031, regression coefficient = -0.15). 45 patients had a clinical seizure during SEEG monitoring with mean onset to first clinical seizure 35.9 ± 7.6 hours. Within this sub-group, the duration of Phase 1 monitoring (p = 0.042, regression coefficient = -0.50) and the onset to seizure during Phase 1 (p = 0.042, regression coefficient = 0.02) were both predictors to the onset of first seizure during SEEG. 3 patients did not have any seizures during SEEG monitoring and 2 patients had only electrographic seizures. In our KM analysis, patients in our cohort who did not have a seizure in the first 80 hours of SEEG monitoring would go on to have SEEG monitoring durations of 7 days or more (n=6). Conclusions: The number of seizure days during Phase 1 scalp EEG monitoring may be predictive of the duration of SEEG monitoring in children undergoing pre-surgical evaluation for intractable epilepsy, with a greater number of seizure days correlated with a shorter SEEG duration. Within our subgroup of patients who had clinical seizures on SEEG, the duration of scalp EEG monitoring during Phase 1 and onset of first seizure during Phase 1 were both predictors to the onset of first electroclinical seizure in SEEG. Future studies are needed to determine the applicability of findings to all types of invasive pre-surgical monitoring and what additional clinical factors may contribute. Funding: No funding