Graduate Teaching Assistant University of Texas at Arlington Arlington, Texas
Rationale: Epilepsy is increasingly seen as a disorder of both structural and functional networks. An advantage of studying functional networks is that they show abnormal connectivity even in the absence of lesions. Epilepsy connectivity studies have been limited so far to intracranial electroencephalography (icEEG) data filtered at lower frequencies (< 70 Hz) ignoring the emerging role of high frequency oscillations as epilepsy biomarkers. Here, we assess the feasibility of using functional connectivity metrics that contain frequency components over the 0–250 Hz range (i.e. spikes, ripples, and fast ripples) in interictal icEEG. We use these metrics as candidate biomarkers to discriminate the seizure onset zone (SOZ) from surrounding areas and test if these can be used as predictors of outcome. Methods: We examined 25 children with medically refractory epilepsy (MRE) who underwent surgery. We dichotomized surgical outcome into seizure-free (14 patients; Engel 1) and non-seizure-free (11 patients; Engel≥2). On icEEG recordings, we identified: (i) segments with interictal epileptiform discharges (IEDs); (ii) resting-state activity; and (iii) segments with IEDs on ripples and/or fast ripples. For each patient, we obtained three connectivity matrices [Amplitude Envelope Correlation (AEC), Correlation (CORR), and Phase Lock Value (PLV)] that consisted of connectivity measures between all icEEG electrodes pairwise (Figure 1). Using the Wilcoxon signed-rank test, we tested differences in mean connectivity between the icEEG electrodes inside and outside the SOZ, which was previously defined by the epileptologists. We also compared connectivity values between resected and non-resected areas for seizure-free (good outcome) and non-seizure-free (poor outcome) patients. Results: We observed decreased mean CORR for the IEDs segments inside (0.2156±0.081) compared to outside (0.24944±0.0732) the SOZ for all patients (Figure 2a; p=0.023). No significant changes were observed for the other two connectivity metrics (i.e. AEC and PLV). We also observed a significant decrease in the mean CORR values for the IED segments recorded form electrodes located in resected (0.2433±0.0743) areas compared to non-resected areas (0.2760±0.06) for patients with good outcome (p=0.0479; Figure 2b); such a difference was not observed for patients with poor outcome (p=0.320; Figure 2c). Finally, we observed a decrease in mean CORR for the IEDs on ripples for electrodes inside resection (0.2705±0.068) compared to outside resection (0.3125±0.1088) only for patients with good outcome (p=0.0195; Fig. 2d). Conclusions: Functional connectivity measures of interictal icEEG segments with ripples overlapping on IEDs are promising biomarkers for predicting surgical outcome in children with MRE. Such a biomarker can help discriminate the epileptogenic zone (EZ) from less epileptogenic areas without having to wait for a seizure to occur and guide the EZ resection, thus improving the surgical outcome of patients.
Funding: Please list any funding that was received in support of this abstract.: Cook Children's Health Foundation