Abstracts

Rationale: High frequency oscillations (HFO, ripple: 80-250 Hz, fast ripple (FR): 250-500 Hz) are EEG biomarkers in epilepsy. Previous studies in human patients have focused on analysis of interictal intervals and suggest that HFOs are a good indicator of epileptogenicity. In addition, HFOs have been linked to ictogenesis. However, their role has not been studied extensively so far. We hypothesized a relationship between ictal HFOs and seizure severity. Methods: We recorded intracranial EEG data from patients with drug-resistant temporal lobe epilepsy with simultaneous video-monitoring. Seizures were classified depending on their secondary clinical generalization. HFOs and their duration were marked in the EEG for a 30s preictal and postictal period as well as during the ictal period each 30s before and after secondary generalization. We included all recording channels located in the seizure onset zone (SOZ) and selected 5 remote contacts for comparison. HFO activity was identified visually by two reviewers independently. Durations of HFOs per minute were calculated to account for varying length of HFOs during the different periods. Durations of ripples and FRs/min were statistically compared for the different periods as well as different seizure severity. Results: 5 complex partial and 6 secondary generalized seizures from 4 patients were analyzed. 36 channels were in the SOZ, 55 channels were outside. Both ripple and FR duration/min increased during the course of seizures and decreased afterwards. It can also be shown that the duration/min of ripple/FR is significantly higher over the SOZ in the preictal (p=0.004/p=0.001) and ictal (p < 0.001/p < 0.001) period (Figure 1). Besides, ripple duration/min during the preictal (p=0.011), ictal (p=0.050) and postictal (p=0.019) intervals of non-generalizing seizures was significantly lower than in secondary generalizing ones (Figure 2). Conclusions: The present results show that the duration of HFOs per minute increases during the ictal period and shows a postictal depression. As expected HFOs were predominant over the SOZ areas during all phases of the seizure. Ripples were significantly more prominent in seizures that secondarily generalized, even prior to the EEG and clinical generalization. This supports our hypothesis that HFOs reflect seizure severity, which indicates an involvement in ictogenesis. In addition, they might be useful for estimating the risk of secondary generalization. Analysis of more seizures from further patients is needed to secure the results. Funding: No funding