Functional Segregation of Epileptogenicity within the Human Amygdala

Objective: In temporal lobe epilepsy (TLE), the amygdala in the epileptogenic network is underestimated compared to other regions such as the hippocampus. Recent advances in anatomical neuroimaging and stereoelectroencephalography (SEEG) signal analyses could help better understand the involvement of the different amygdala nuclei in the genesis of temporal lobe seizures. Methods: We retrospectively included 51 patients suffering from TLE who underwent SEEG over the past 5 years. The Virtual Epileptic Patient atlas with an integrated amygdala atlas was used to automatically localize SEEG contacts within the brain regions, including 9 amygdala nuclei. The Epileptogenicity Index (EI) and Connectivity Epileptogenicity Index (cEI) were computed on ictal SEEG recordings. We used a beta mixed model to evaluate the relative effects of amygdala nuclei, TLE subtypes, and lateralization of the epileptogenic zone on the epileptogenicity. We used the Wilcoxon rank sum test to study the associations between epileptogenicity level of distinct amygdala nuclei and ictal semiology (sensory, affective, cognitive, motor, and autonomic). Results: We observed higher epileptogenicity within the basolateral (BL) nucleus compared to other nuclei of the basolateral complex (lateral (LA), accessory basal (BM), and paralaminar (PL) nuclei) across all TLE subtypes. Regarding semiology, BL was more epileptogenic in patients with sensory phenomena and LA in patients with autonomic phenomena, while PL was less epileptogenic in patients with cognitive phenomena. Interpretation: Our findings disentangle the different epileptogenicity of amygdala nuclei in temporal lobe seizures. The observed epileptogenicity variance across amygdala nuclei can be explained by underlying neuronal and cytoarchitectural substrates. ANN NEUROL 20269999:n/a-n/a.