Abstracts

Rationale: Identifying the seizure onset zone (SOZ) is crucial for surgical interventions for patients with epilepsy. Pathological activity in epileptic patients may extend further than the SOZ, making its identification a challenge in the field. There is evidence that EEG biomarkers such as interictal spikes and high frequency oscillations can be helpful in identifying the SOZ. However, it is hard to determine biomarker specificity, given that the electrical dynamics of pathological activities can differ from one region to the next. Recently, phase amplitude coupling and power spectral density have been found to be informative for the SOZ. We investigated whether there is a difference in the power of different frequency bands between SOZs and non-SOZs, and whether these differences are region dependent.

Methods: Five-minute interictal periods of awake (n=99) and asleep (n=98) data from patients (n=33) with medically refractory focal epilepsy who underwent presurgical evaluation with intracranial electrodes were analyzed. Patients were classified into four groups based on their SOZ: hippocampus, lateral temporal, frontoparietal, and occipital lobes. We computed spectral power in six frequency bands (delta, theta, alpha, beta, low gamma, and high gamma) in seven brain regions (hippocampus, lateral temporal, frontoparietal, cingulate, amygdala, occipital, and insula) during the selected interictal periods for each patient, before averaging over the interictal data of patients in each SOZ group. Within each SOZ group, the average power was compared for all region pairs within each frequency band, using Kruskal-Wallis tests to identify significance (p=0.05).

Results: The SOZ of each patient was identified as either hippocampus (n=19), lateral temporal (n=7), frontoparietal (n=5), or occipital (n=2) lobe. Seizures originating in the hippocampus had significantly larger power in the hippocampus compared to other regions (non-SOZ) across all bands (p < 0.05). Otherwise, there was no correlation between SOZ and the region with the highest power. Within each SOZ group, we found a consistent distribution of average interictal power across regions. Interestingly, interictal activity exhibited relatively large power readings in non-SOZs as well as SOZs. The non-SOZs that exhibited higher power were in structural proximity to the SOZ (e.g. in the lateral temporal onset group, occipital regions exhibited high power compared to other regions, while exhibiting lower power in the frontoparietal onset group). We also found that the cingulate and amygdala had the smallest powers across all bands for all SOZ groups, while the hippocampus generally had the largest power across bands.

Conclusions: Our findings demonstrate that the relationship between the region of seizure onset and spectral power magnitude is complex. Some regions tend to exhibit higher power compared to others, even at baseline, and this variability makes the identification of SOZ even more challenging. In addition, since the average interictal power in a given region can exhibit large power readings in non-SOZs, we speculate that epileptogenesis may extend further beyond the SOZ than initially thought.

Funding: Please list any funding that was received in support of this abstract.: NIH/NINDS K24-NS088568, R01- 2NS062092.