Abstracts

Rationale: Rolandic epilepsy (RE) is the most common focal developmental epilepsy, characterized by a transient period of seizure susceptibility during childhood. The pathophysiology behind the seizures and symptom resolution in this disease is not known, but recent studies have identified focal disruption of thalamocortical circuits to the Rolandic cortex. Children with RE have an abnormal developmental trajectory of thalamocortical white matter structural connectivity to the Rolandic cortex and transiently decreased sleep spindles rate in the Rolandic cortex during the active phase of disease. Because seizures are due to abnormal synchronization between brain regions, we hypothesized the thalamocortical functional connectivity to the Rolandic cortex would also follow an aberrant developmental trajectory in Rolandic epilepsy compared to controls.

Methods: We recruited 28 children with RE including 18 with active disease (defined as seizure within 12 months, 5F, Ages 10.9±2.9) and 10 in disease resolution (seizure free for > 12 months, 2F, Ages 13.6±2.1), and 23 controls (10F, Ages 11.2±2.5). Subjects underwent structural and functional MRI in a 3T Magnetom Prisma Simens MRI scanner. The thalamus and the inferior Rolandic cortex were defined as regions of interest (ROIs) in each subject using Freesurfer labels. Resting-state functional connectivity MRI (rsfcMRI) preprocessing used FSFAST and included slice time correction for multiple interleaved acquisition, motion correction to the middle volume, spatial smoothing using a 5mm FWHM Gaussian filter, and removal of artifacts from head motion, CSF, white matter, and global signals using regression. Functional signals from each voxel were averaged within each ROI. Signals were filtered from 0.01-0.08Hz. Volumes were identified as having motion artifact if the mean global signal was >3 STD from the mean of temporal change, if there was framewise displacement >0.5mm from the reference volume, or rotation or translational values > 1mm between contiguous volumes. Volumes with motion artifact were then ignored and the rsfcMRI was computed as the cross-correlation of the final time-series signal from the thalamus and inferior Rolandic ROI for each hemisphere for each subject (Figure 1A-F). The developmental trajectory of inferior Rolandic thalamocortical functional connectivity was evaluated for each group (active, resolution, control) using mixed effect generalized linear models.

Results: Control children (Green, p=0.003) and children with active RE (Red, p=0.006), but not those with resolved RE (Blue, p=0.31) have increased thalamocortical functional connectivity to the inferior Rolandic cortex with age (Figure 2). Children with resolved RE have decreased thalamocortical functional connectivity to the Rolandic cortex with age compared to controls (p=0.01).

Conclusions: Children with active RE demonstrate increased Rolandic thalamocortical functional connectivity with age, similar to controls. Epilepsy resolution coincides with disruption of this relationship, suggesting a potential compensatory role of decreased Rolandic thalamocortical functional connectivity to facilitate symptom resolution.

Funding: Please list any funding that was received in support of this abstract.: NIH NINDS R01NS115868.