Stability of the Epileptogenic Index across seizures and the effects of sleep and withdrawal of anti-seizure medications.

Abstract

Objective: Epileptogenic Index (EI) is a promising index to evaluate the epileptogenicity of the brain tissue and is useful in planning epilepsy surgery. We aimed to evaluate whether EI maps are consistent across seizures within patients and whether EI-derived measures are affected by the vigilance state and anti-seizure medication (ASM) dose.

Methods: We analyzed 127 stereoelectroencephalography (SEEG) seizures from 19 patients. EI was calculated in all seizures. Within-patient EI consistency was assessed using pairwise Spearman correlations between seizure-specific EI maps. To assess the effects of vigilance and ASM on EI consistency, we calculated a contact-level p_weighted EI score that combines the frequency of suprathreshold EI involvement and mean EI magnitude, separately across vigilance states and in a subset of patients with different ASM states.

Results: Mean seizure-to-seizure EI correlations ranged from 0.11 to 0.81, indicating substantial variability in within-patient EI consistency. Seven patients showed high consistency (≥0.6) and four showed low consistency (≤0.3). At the contact level, no differences in p_weighted EI were found between wakefulness and sleep. In contrast, p_weighted EI was significantly higher during reduced ASM compared with full ASM doses.

Significance: EI maps are not uniformly stable across seizures. ASM reduction may broaden or intensify EI-defined epileptogenic involvement, suggesting that ASM status should be considered when interpreting EI during SEEG evaluation. These findings are hypothesis-generating and require prospective validation to determine their clinical utility for surgical planning.

Plain language summary: The Epileptogenic Index (EI) measures the involvement of a recorded brain area during a specific seizure. This study shows that EI patterns can vary significantly between different seizures in the same patient. While sleep states do not change these patterns, reducing a patient's anti-seizure medications increases EI values, making seizures appear to spread more broadly. These medication-induced changes must be carefully considered to accurately map the epileptic network for surgical decisions.