Abstracts

Rationale: Video electroencephalography (VEEG) is an essential component of presurgical evaluation and adequate localization of seizure onset using EEG is important for seizure free outcome. Multiple recorded seizures may be necessary to confidently understand the epileptogenic zone. Duration of EEG monitoring may depend on multiple variables, including patient age, seizure frequency, number of pre-op seizure semiologies, patient response to provocative stimuli, and imaging results, though little data exists to predict duration of monitoring or number of seizures required for localization. With increased constraints on healthcare costs, it is important to limit duration of stay and to make an accurate a priori prediction of monitoring days needed to complete a presurgical evaluation. It is essential to understand factors which influence the number of seizures and duration of VEEG monitoring necessary for accurate localization, after which time the likelihood of more widespread or multifocal epileptogenic zones is negligent. Methods: Patients undergoing phase I presurgical evaluations were prospectively enrolled between 1/2014 and 1/2016. Seizures were analyzed sequentially during the VEEG admission by the treating epileptologist and included if the seizure provided additional localizing clinical or neurophysiologic information. The total number of seizures and duration of monitoring needed for the epileptologist to confidently localize seizure onset was compared to multiple pre-admission patient characteristics (i.e. presence of lesion on MRI, seizure frequency, age, number of pre-op semiologies, and temporal vs extratemporal localization) to determine predictors. Results: Eighty three patients were investigated, 10 (12%) were undergoing evaluation after a prior failed surgery. Average age at evaluation was 9.72 y (0.5-18) with an epilepsy duration of 5.13 y (0-17). Forty three (52%) had a single seizure semiology, 24 (29%) had 2, and 16 (19%) had 3 or more. We recorded an average of 17.74 (1-469) seizures over mean monitoring duration of 78 hours (6-214). Patients had an average 3.17 seizures which provided new information (1-19) over a mean 33 hours of monitoring. Forty eight percent required only 1 seizure for localization, 31% required 2-4, and 21% required 5 or more seizures. Patients with a single pre-op semiology required fewer seizures for localization (p=.018), though other pre-op variables did not significantly impact seizure number. Patients with daily seizures (p=.033) and those undergoing a repeat evaluation (p=.003) required shorter durations of monitoring. Of those treated with surgery (n=32), 63% were seizure free at 6 months. Conclusions: The majority of patients require < 4 seizures to adequately localize seizure onset and for those with a single semiology at presentation, a single seizure is often adequate for localization. Patients with daily seizures and those undergoing repeat evaluations after failed surgery require shorter duration of phase I VEEG monitoring to localize seizure onset with most completed within 48 hours. Funding: None