Abstracts

Rationale: Encephaloceles are protrusions of brain parenchyma through cranial defects or natural foramina, often found in the anterior temporal lobe. Although they may be associated with up to 14% of refractory epilepsy cases, their causal role and relationship to seizure networks remains unclear. Most investigations involved small cohorts or case studies. Here, we characterized encephaloceles, including associated demographic factors, number of encephaloceles per patient, location and size, and their precise relationship to seizure onset zones (SOZs) using gold-standard stereoelectroencephalography (SEEG).

Methods: We retrospectively studied all epileptic patients with suspected encephalocele(s) on imaging who underwent intracranial investigation at our institution. Encephaloceles were manually segmented using a high-definition T2 MRI to localize and determine their volume. We used FreeSurfer to automatically segment and parcellate preoperative T1 MRI into anatomical regions, and objectively determine cortical regions involving encephalocele(s) and SOZ(s). Postoperative CT was used to manually label intracranial recording contacts. Imaging spaces were coregistered and normalized to MNI space using the Advanced Normalization Tools (ANTs) toolbox. A neurologist identified contacts involved in the first 1 second (s), next 3 s, and next 7 s of the seizure to define SOZ and early spread. Finally, relevant demographic information was collected through chart review.

Results: Our cohort comprises 21 patients with 90% (19) females, 76% (16) Blacks, 95% (20) with additional stigmata of IIH, a mean body mass index (BMI) of 39.9, a mean age at epilepsy onset of 29.1 years, and a mean duration of epilepsy of 9.6 years. A total of 58 encephaloceles were identified, with on average 2.8 encephaloceles per patient (median 2, range 1-7, bilateral in 11 patients), of average volume 0.33ml (range 0.05-1.31 ml). The temporal pole was the region most often involved with encephaloceles, but other involved regions included the fusiform, inferior temporal, entorhinal, and orbito-frontal cortices. 75% of encephaloceles were adequately sampled by SEEG (defined by electrodes placed < 1 cm from encephalocele). Overall, 3-dimensional visualization of SOZs, defined by SEEG contacts showing activity in the first second of an event, indicated that encephalocele locations were highly associated with SOZs which themselves varied with respect to extent of involved temporal lobe structures, some of which included the hippocampus. Notably, even small (< 6 mm in diameter) encephaloceles co-localized with SOZs.

Conclusions: We observed encephalocele associations with female sex, Black demographic, obesity, imaging stigmata of intracranial hypertension, and adult-onset epilepsy. Encephaloceles co-localized with SOZs, but the extent of SOZ was variable, possibly suggesting a value of SEEG in mapping SOZ to inform subsequent intervention.

Funding: T32EB025816