Abstracts

Rationale: Scalp electroencephalography (scEEG) remains a cardinal tool in the pre surgical evaluation of epilepsy. Scalp EEG, along with other investigative modalities, guides intracranial electrode placement, but it is seldom used simultaneously with intracranial EEG (IEEG). ScEEG ictal and interictal discharge localization, amplitude and morphology reflect intracranial generating source area size and the vector of the discharge’s dipole, as previously demonstrated in temporal lobe epilepsy. We hypothesize that a focal seizure would typically appear on scEEG with a slight delay, but would maintain onset, morphological and propagation patterns similar to those recorded in IEEG. Lack of ictal concordance between scEEG and IEEG in the presence of adequate scalp electrode coverage, may indicate that the seizure source recorded in IEEG is not the true onset zone, leading potentially to poor localization of seizure onset and rendering the patient a suboptimal candidate for surgical resection. Our goal is to explore the relationship between the onset time, onset pattern, and evolution of seizures with simultaneous scEEG and IEEG in patients whose seizures were localized to a single focus and underwent a surgical resection in comparison to patients who did not have resective surgery.Methods: A retrospective analysis of records of patients who underwent intracranial monitoring as part of a presurgical evaluation in UPMC, Pittsburgh, PA during 2012-2014 was performed. Patients who underwent simultaneous scEEG and IEEG recording were subsequently selected and six patients were included in the preliminary analysis. For each patient, we analyzed the following in the first 5 seizures, 24 hours post implantation: seizure onset time, seizure onset pattern and location of seizure onset and evolution in the first 15 seconds in the scEEG and IEEG. Examples of concordant seizure patterns are low amplitude fast activity in IEEG with attenuation of paroxysmal fast activity on scalp EEG and rhythmic spikes in IEEG and rhythmic spikes or rhythmic delta on scEEG. We then compared these parameters between patients whose seizures were localized to a single focus and underwent surgical excision and patients whom did not undergo resective surgery.Results: The mean delay in seizure onset between scEEG and IEEG was 17.8 seconds in patients whose seizures were localized to a single focus and underwent resective surgery, compared to 5.5 seconds in patients who did not undergo surgery. Most of the patients who underwent resective surgery had a concordant ictal onset location and ictal patterns as opposed to patients who did not have resective surgery.Conclusions: Absence of concordance between scEEG and IEEG in terms of seizure onset time, seizure pattern and seizure onset and evolution in the first 15 seconds in a patient with adequate scEEG coverage is associated with poor localization of ictal onset zone and inability to undergo resective surgery. This mismatch may occur when the seizure onset zone is not adequately covered by IEEG electrodes.