Epileptogenicity in tuberous sclerosis complex: A stereoelectroencephalographic study

Abstract

Objective

In tuberous sclerosis complex (TSC)‐associated drug‐resistant epilepsy, the optimal invasive electroencephalographic (EEG) and operative approach remains unclear. We examined the role of stereo‐EEG in TSC and used stereo‐EEG data to investigate tuber and surrounding cortex epileptogenicity.

Methods

We analyzed 18 patients with TSC who underwent stereo‐EEG (seven adults). One hundred ten seizures were analyzed with the epileptogenicity index (EI). In 13 patients with adequate tuber sampling, five anatomical regions of interest (ROIs) were defined: dominant tuber (tuber with highest median EI), perituber cortex, secondary tuber (tuber with second highest median EI), nearby cortex (normal‐appearing cortex in the same lobe as dominant tuber), and distant cortex (in other lobes). At the seizure level, epileptogenicity of ROIs was examined by comparing the highest EI recorded within each anatomical region. At the patient level, epileptogenic zone (EZ) organization was separated into focal tuber (EZ confined to dominant tuber) and complex (all other patterns).

Results

The most epileptogenic ROI was the dominant tuber, with higher EI than perituber cortex, secondary tuber, nearby cortex, and distant cortex (P < .001). A focal tuber EZ organization was identified in seven patients. This group had 80% Engel IA postsurgical outcome and distinct dominant tuber characteristics: continuous interictal discharges (IEDs; 100%), fluid‐attenuated inversion recovery (FLAIR) hypointense center (86%), center‐to‐rim EI gradient, and stimulation‐induced seizures (71%). In contrast, six patients had a complex EZ organization, characterized by nearby cortex as the most epileptogenic region and 40% Engel IA outcome. At the intratuber level, the combination of FLAIR hypointense center, continuous IEDs, and stimulation‐induced seizures offered 98% specificity for a focal tuber EZ organization.

Significance

Tubers with focal EZ organization have a striking similarity to type II focal cortical dysplasia. The presence of distinct EZ organizations has significant implications for EZ hypothesis generation, invasive EEG approach, and resection strategy.

0