Abstracts

Rationale: Electrical modulation of the thalamus using deep brain stimulation (DBS) is an emerging treatment for patients with refractory epilepsy, including Lennox-Gastaut Syndrome (LGS). However, the optimal thalamic targets are uncertain. Furthermore, functional organization of thalamic nuclei, and their cortical projections, is poorly understood in LGS. Here, we used task-free fMRI to study functional connectivity (FC) between the thalamus and cortical networks in LGS patients and healthy controls. We hypothesized that patients would show abnormal FC in specific thalamocortical circuits. Methods: For each of 20 LGS patients (mean age 28.5 yrs; 10 female) and 26 age-matched controls (mean age 27.6 yrs; 10 female), 210 fMRI volumes were acquired using a 3T scanner. fMRI data were warped to a common template, and signal changes attributable to motion and physiological noise were removed. 10 cortical networks were determined from the fMRI data of all subjects using group independent components analysis (Fig. 1). Thalamocortical FC was computed using the Pearson correlation between the time-series of each thalamic voxel, and the the first Eigen time-series across voxels within each cortical network. For each cortical network, we studied a) thalamic regions with significant positive FC, computed separately for patients and controls; and b) thalamic regions with significant between-group differences in FC. Significance (p < 0.05, family-wise error corrected) was assessed using nonparametric permutation tests. Anatomical locations of significant thalamic FC were identified by quantifying spatial similarity (using Dice's coefficient [DC]) with an MRI atlas generated from thalamic histology (NeuroImage, 2010; 49(3); 2053-62). Results: Fig. 1 displays each cortical network and its territory of significant thalamic FC in patients and controls. In both groups, posterior thalamic regions showed FC with networks in primary/sensory cortex (visual, auditory, and sensorimotor networks), while the anterior, medial and dorsal thalamus showed FC with networks in limbic/association areas (including default-mode [DMN], anterior salience, and executive control [ECN] networks). Four networks (left/right ECN, ventral/dorsal DMN) showed significantly stronger thalamocortical FC in patients relative to controls (Fig. 2). Quantification using DC indicated that these FC increases maximally involved the mediodorsal (Md) and adjacent ventrolateral (VL) nuclei. Conclusions: Posterior thalamic nuclei, connecting to primary/sensory cortical networks, showed similar FC in LGS and healthy controls. However, FC was significantly enhanced in LGS between the Md/VL thalamus and the DMN/ECN, thalamocortical circuits that normally support key cognitive processes. Our previous EEG-fMRI studies show that the DMN and ECN are recruited during epileptic activity in LGS (Epilepsia, 2014; 55(8); 1245-54). Taken together, the present results suggest that the Md/VL thalamus may be a candidate target for modulating epileptic activity in these networks, potentially via thalamic DBS. Funding: National Health & Medical Research Council (NHMRC) project grant #628725. Aaron Warren is supported by an Australian Postgraduate Award.