Abstracts

Rationale: Thalamus is known to have participated in the epilepsy network. Cortical fast activity usually precedes occurrence of clinical semiology. Whereas the association between thalamic fast activity and seizure behaviors is yet unknown. An in-depth understanding of properties of thalamic ictal fast activity and seizure behaviors can assist with better description of ictal epilepsy network.

Methods: We prospectively included 52 patients with focal epilepsy who underwent stereotactic implantation of electrode targeting at thalamus. We summarized characteristics of ictal thalamic fast activity using intracranial electroencephalography (EEG) reading, power analysis, and time-frequency analysis. We further divided patients into focal- and non-focal onset patterns based on whether the seizure-onset zone was confined to a single cortex and whether the seizure-onset pattern of EEG was consistent across different seizures. To ensure reliability of the study, only patients with focal-onset patterns were included in the statistical analysis. We analyzed the association between ictal thalamic fast activity and clinical semiology, and further explored clinical influencing factors by univariate analysis and multivariate regression analysis.

Results: Forty-eight patients (92.3%) had both non-specific thalamic nuclei and specific nuclei implanted along the implanted electrode trajectory and the rest 4 patients had only thalamic specific nuclei implanted. Among the 27 patients out of the total 52 patients presenting with a focal-onset pattern, 20 (74.1%) expressed low-voltage fast activity (LVFA) at seizure onset. In most seizures, cortical fast activity consistently preceded clinical symptoms. However, there was no significant correlation between occurrence of thalamic fast activity and different seizure-onset patterns or seizure origin, regardless of whether the cortex expressed LVFA or not. With the help of simultaneous intracranial and scalp EEG recordings, we found a correlation between ictal thalamic fast activity and arousal provoked by seizures during sleep. Multivariate regression analysis revealed found that presence of seizure during sleep was an independent predictor for ictal thalamic fast activity [p = 0.004; OR 42.000 (95% CI 3.198-551.565)], after excluding influences caused by epilepsy types, pathological findings, lesional finding on magnetic resonance imaging, and surgical outcomes.

Conclusions: For the first time, this study explored associations between cortico-thalamic fast activity and seizure behaviors in epilepsy, enhancing understandings of ictal epilepsy network. Thalamic fast activity is closely correlated with arousals during sleep, suggesting thalamus may be involved in the occurrence of seizure behaviors during sleep, which is an important step during ictogenesis. We need to look again at the role of thalamus in brain network. Relationships between epileptic seizures and sleep-awake network should be taken into consideration in the future when analyzing thalamic electrophysiological recordings.

Funding: National Natural Science Foundation of China (grant number, 82171437), Natural Science Foundation of
Zhejiang Province (grant number, LD24H090003).