Authors :
Presenting Author: Emily Liao, BE – University of Minnesota
Derek Doss, PhD – Vanderbilt University
Ghassan Makhoul, BS – Vanderbilt University Medical Center
Addison Cavender, BS – Vanderbilt University Medical Center
Anas Reda, MS – Vanderbilt University Medical Center
Bruno Hidalgo, BS – Vanderbilt University Medical Center
Kate Wang, BS – Vanderbilt University Medical Center
Lucas Sainburg, BS – Vanderbilt University
Graham Johnson, MD, PhD – Mayo Clinic
Angela Crudele, MD – Vanderbilt University Medical Center
Sarah Bick, MD – Vanderbilt University Medical Center
Victoria Morgan, PhD – Vanderbilt University Medical Center
Dario Englot, MD PhD – Vanderbilt University Medical Center
Rationale:
The Extended Network Inhibition Hypothesis (eNIH) suggests that in patients with mesial temporal lobe epilepsy (mTLE), persistent seizure propagation from mesial structures to the subcortex and basal forebrain yields lasting arousal network damage. This damage then likely contributes to deficits in vigilance and cognition interictally. The thalamus and ascending reticular activating system (ARAS) are among the regions affected by this pathophysiology. Prior imaging studies of mTLE patients have shown decreased structural and functional connectivity in thalamocortical and ARAS-to-cortical connections. We hypothesize subcortical-to-subcortical connectivity may also contribute to arousal network dysfunction, specifically among the thalamus, ARAS, hippocampus, and amygdala.
Methods:
We collected resting-state fMRI (TR=2s,3x3x4mm3) from mTLE patients (n=33) and controls (n=64). DWI (b=1000s/mm2, HARDI, 2.5mm isotropic) was obtained for a subset (patients n=17, controls n=45). For parcellation, we combined atlases in patient specific space (Desikan-Killiany, ScLimbic, Harvard AAN, and HSSM thalamus). Functional connectivity (FC) and streamline weights were computed between regions of interest (ROIs). Connectivity values were age corrected, and patients were z-scored against controls. To align seizure onset zone laterality, connectivity values were left-right flipped in patients with right mTLE and an equal proportion of controls. Node strengths were computed by summing connectivity with all other ROIs and dividing by the number of connected regions. Patient and control differences were tested with two-sample t-test for Fisher z-transformed FC and Wilcoxon rank-sum for streamline weights. P-values were adjusted for Holm-Bonferroni correction.
Results:
In functional analysis, patients had significantly reduced mean subcortical connectivity strength in the ipsilateral anterior thalamus, centrolateral (CL) thalamus and caudate (p< 0.05 two-sample t-test). Contralaterally, they had reductions in the connections of the centromedian (CM) and ventral thalamus (p< 0.05, Fig. 1A). Edgewise differences were significant between ipsilateral anterior and contralateral ventral thalamus, ipsilateral CL and ipsilateral parabrachial complex (PBC), and contralateral CM and PBC (p< 0.05, Fig. 1B). In structural analysis, reduced strength was seen in the ipsilateral hippocampus and contralateral ventral thalamus (p< 0.05 Wilcoxon rank-sum, Fig. 1A).