Abstracts

Rationale: Hippocampus is commonly involved in seizure generation in temporo-perisylvian epilepsy and participates in seizure propagation in neocortical epilepsy. SEEG method allows studying the role of the hippocampus in epileptogenic network, however, significance of various SEEG ictal patterns as biomarkers of hippocampal epileptogenecity remains unclear. Change in FLAIR signal is a known feature of hippocampal sclerosis, but its implication in non-sclerotic hippocampus still has to be evaluated. In this study, we aimed to characterize SEEG ictal patterns and MRI signal changes that predict hippocampal epileptogenicity Methods: Subjects were selected from a consecutive cohort of patients who underwent SEEG with hippocampal exploration (n=243). Only patients who, became seizure free following surgical resection and had recorded ictal patterns involving the hippocampus were included (n=43). The hippocampus was identified as epileptogenic and resected in 29 patients (mesial and mesio-lateral temporal group 1) and was spared in 14 patients (neocortical group 2). All surgical specimens were microscopically reviewed and hippocampal cell loss classified according to the ILAE classification 2013. Eleven patients from group 1 had histopathologically confirmed HS ILAE Type 1 (group 1A) and 18 patients had no evidence for segmental cell loss (no-HS; group 1B). Electrographic patterns were determined by visual inspection and characterized with power spectral density (PSD) signal analyses. Changes in FLAIR signal in the hippocampus were assessed visually and automatically on a voxel basis (MATLAB SPM12). 5 patients were excluded from MRI post-processing because of big structural lesion (n=4) and motion artifacts (n=1). Exact Fisher test were used for statistical analysis. Results: On SEEG, ictal pattern initiated from rhythmical spikes or brief bursts of spikes with transition to low voltage fast activity (LVF) in gamma/beta range was statistically more frequent in group 1 compare to group 2, 23 (83%) vs 2 (14%), p=0,001, respectively. Difference was not significant for LVF without initial spikes, 4 (14%) vs. 5(36%). Onset from rhythmical spikes with duration 5 seconds and longer was found exclusively in epileptogenic hippocampus, and was more frequently in group 1A compare to group 1B, 9(80%) vs. 3 (17%). Abnormal FLAIR signal in the hippocampus upon visual analysis was identified in all patients from group 1A. FLAIR signal was statistically more frequent in group 1B compare to group 2, 9 (58%) vs. 1 (17%) p=0,02. MRI post-processing revealed hippocampal FLAIR signal change in 8 (89%) patients from group 1A, in 4 (24%) patients from group 1B, and in 1 (6%) from group 2. Conclusions: Hippocampal epileptogenicity is common in hippocampal sclerosis but also confirmed when the hippocampus is histopathologically not sclerotic (no-HS, ILAE classification 2013). The most specific SEEG biomarkers of the epileptic hippocampus were spikes with transition to LVF. Changes in FLAIR signals correlate with the sclerotic and non-sclerotic epileptogenic hippocampus. Funding: No funding.