(934) Joint analysis of EEG and MRI in Rasmussen’s Encephalitis questions the concept of a unihemispheric pathology

Rasmussen’s encephalitis (RE) can be considered an archetype of an unihemispheric pathology. To put this paradigm to the test, we systematically quantify Electroencephalography (EEG) characteristics of RE patients and relate them to disease onset, disease duration and patterns of cortical atrophy in MRI in the largest dataset of RE patients reported.

Methods:

Seventy-two patients (41 females, 39 left focus) were retrospectively included. All available EEG data (573 datasets, >3000 total hours) were evaluated by an experienced rater according to a fixed scheme (cf. Kane et al., 2017). Interictal epileptiform patterns (sharp waves or spikes) and seizure onset were regarded as epileptiform activity. A sample of 100 datasets was re-evaluated by a second rater (κ=0.40). In 12 patients, EEG data were available before and after hemispherotomy. Patterns of cortical atrophy were evaluated in a subset of 52 patients (35 females, 29 left focus) with at least one isotropic T1-weighted MRI scan available. Spatial normalization and cortical reconstruction was performed using Freesurfer (https://surfer.nmr.mgh.harvard.edu). All images were reoriented leaving us with an ipsilesional and a contralesional hemisphere. Differences in cortical thickness between patients with contralesional epileptiform activity and patients who never presented with contralesional epileptiform activity were assessed using unpaired, two-sided t-tests.

Results:

An unspecific interictal focus (theta or delta slow activity) was found in 48.2% of all EEG timepoints in the ipsilesional and in 14.2% in the contralesional hemisphere. Interictal epileptiform patterns were found in 39.3% in the ipsilesional, in 12.9% in the contralesional and in 14.3% in both hemispheres simultaneously. Seizure onset was found in 25.9% in the ipsilesional, in 3.3% in the contralesional and in 3.7% in both hemispheres. In a mixed effects logistic regression model, contralesional epileptiform activity was predicted by age at disease onset (OR=0.89, p=0.002) but independent of disease duration (OR=0.99, p=0.65). Considering a subset of 12 patients who underwent hemispherotomy, contralesional epileptiform activity remained after surgery in 5 cases (41.7%). Cortical atrophy analysis revealed reduced cortical thickness in superior frontal, anterior cingulate and insular cortices in the contralesional hemisphere in patients with contralesional epileptiform activity as compared to patients without contralesional epileptiform activity (FDR-corrected p< 0.05).

Conclusions:

Based on a large, longitudinal study cohort, our results prove that contralesional epileptiform EEG activity in RE is not uncommon and may persist after hemispherotomy, suggesting epileptiform EEG activity to originate from the contralesional hemisphere. Further analyses show that contralesional epileptiform EEG activity is more likely to occur in patients with disease onset at a younger age and is associated with typical atrophy patterns located solely in the contralesional hemisphere. Altogether, our findings question the concept of RE as a unilateral pathology.

Funding: Please list any funding that was received in support of this abstract.: BONFOR Research Commission of the Medical Faculty of the University of Bonn, 'BonnNI' structured MD program of the Medical Faculty of the University of Bonn funded by the Else Kröner-Fresenius-Foundation, 'Verein zur Förderung der Epilepsieforschung'