Abstracts

Rationale: The spike-wave complex is a typical EEG pattern of interictal epileptic discharge. The spike includes high-frequency oscillations (HFOs) which are linked to epileptic excitation. The post-spike slow wave (PSS) relates to inhibitory activity. The spike-wave complex in patients with focal cortical dysplasia (FCD) type II is well known to be very intrinsic, continuous and focal in the epileptogenic zone. According to this evidence, we hypothesized that power reduction of PSS, relative to spike-related HFOs in the same spike-wave complex, would contribute to seizure initiation in patients with FCD type II. We explored the spatio-temporal power relationship between spike-related HFOs and PSS in patients with FCD type II. Methods: We retrospectively analyzed intracranial EEG data in pediatric patients with FCD type II, who underwent extraoperative video EEG. In each patient, we selected 10 spike-wave complexes including HFOs with 80 - 200 Hz from the electrode in the seizure onset zone (SOZ). We performed spike peak-locked averaging based on the selected spike-wave complex during both 30-sec interictal (> 1 hour apart from seizures) and 30-sec preictal periods. We calculated the power of spike-related HFOs and PSS during both periods for three areas: 1) inside SOZ, 2) inside resection area (RA) but outside SOZ (RA-SOZ), and 3) outside RA. Across all 10 patients, we performed correlation (Spearman's coefficient) and simple linear regression analyses comparing HFO and PSS power in the same spike wave complex within each area during both interictal and preictal periods. Results: Ten patients (6 males and 4 females, age 3 - 14 years) were evaluated. Across all patients, a total of 807 electrodes were analyzed during each period (Figure 1). During interictal periods, there were significant positive correlations between HFO and PSS power in all areas (inside SOZ, r = 0.568, p < 0.0001; RA-SOZ, r = 0.700, p < 0.00001; outside RA, r = 0.320, p < 0.00001). During preictal periods, the correlation became weaker inside SOZ (r = 0.149, p = 0.345) and remained unchanged both RA-SOZ (r = 0.704, p < 0.00001) and outside RA (r = 0.346, p < 0.0001). From interictal to preictal period, the slope (ΔPSS power / ΔHFO power) of simple regression line decreased inside SOZ (0.349 to 0.051) but increased in RA-SOZ (0.534 to 0.734) and outside RA (0.267 to 0.435). Conclusions: Preictal reduction of PSS power, relative to spike-related HFOs inside SOZ, is associated with seizure initiation, which will contribute to future studies focusing on seizure prediction. Also, the correlation coefficients and the slope of the regression lines were so specific each of inside and outside resection areas. These parameters could be new biomarkers to differentiate between epileptogenic and non-epileptogenic areas.