Abstracts

Rationale: Epilepsy is neurological disease characterized by recurring spontaneous seizures caused by hypersynchronous neuronal firing. It was previously shown that epileptogenic areas exhibit abnormal anatomical and functional brain connectivity of clinical and prognostic value. Our aim is to evidence the modifications of the effective connectivity in the epileptic brain using cortico-cortical evoked potentials (CCEPs). Methods: We performed a post-hoc analysis on a cohort of 41 epileptic patients undergoing presurgical evaluation for drug-resistant epilepsy using stereo-EEG method. Single pulse electrical stimulation (SPES) has been systematically performed for each pair of adjacent electrodes as part of the routine investigation, while CCEPs were recorded on the remaining electrodes. The CCEPs, quantified by the RMS of the recorded signal in the 10-110ms interval post-stimulation pulse, allowed us to assess the effective connectivity between stimulation and recording locations. Effective connections between any pair of structures were averaged and normalized at the individual level. Being a post-hoc analysis, we were able to label each stimulation location as being part of the epileptogenic zone (EZ) or not (NEZ), and we created two subsets of effective connections at the group level, based on epileptogenicity. An epileptogenicity modulation index (EMI) was defined for a pair of structures (A,B) as  Rez-Rnez  / Rez+Rnez ( where the effective connection strength if represented by the CCEP response R recorded in structure B when A was stimulated. EMI was computed for all pairs of sublobar temporal structures implanted with intracranial electrodes Results: Overall, EZ structures exhibit significant changes in effective connectivity patterns, showing both enhanced and decreased connectivity. The antero-mesial structures (temporal pole, amygdala, and entorhinal cortex) exhibit an increase in effective connectivity when part of EZ, while the more posterior and lateral structures (hippocampus, fusiform, inferior and superior temporal gyrus) tend to be more disconnected. The highest absolute values of the EMI were observed in the strongest physiological connections (ex. Amygdala-hippocampus, amygdala – anterior cingulate, middle temporal gyrus – lingual gyrus).  Conclusions: Our results suggest that the inclusion of a brain structure in the epileptogenic zone may be decided in the future based on the connectivity patterns it exhibits and the comparison to an atlas of effective connectivity computed solely on NEZ structures. This could be of prognostic significance  for both postsurgical seizure freedom rate as well as for the functional outcome. However, given the inherent problem of sparse brain sampling with intracranial electrodes, such atlas may require a multi-center effort.  Funding: -