Abstracts

Rationale: Chronic epilepsy, when unsuccessfully managed, causes significant neurological changes. In addition to the seizure focus, primary epileptogenic cortex may have a noxious effect on more distal brain regions. Evidence suggests that structural alterations occur along neural pathways remote from the seizure focus. Research on functional connectivity in mTLE has shown distinct changes in neural networks, suggesting possible neuropathophysiological underpinnings to this disorder which can only be detected using network-based analysis methods. However, these methods are prone to bias, limiting their generalizability and applicability. Here we use meta-analytic connectivity modeling to develop robust models independent of task and ROI selection for future connectivity analyses. Methods: We used PubMed to compile a database of papers relating to the structural deficits in mTLE as determined by voxel-based morphometry (VBM). We included papers that met the following criteria: whole-brain VBM analyses implemented,coordinates reported in standard space,studies involving adult, pre-lobotomy patients, and papers written in English. Additional papers were added that were identified in a review by Keller and colleagues(2008). A total of 14 papers were initially identified. Activation Likelihood Estimation (ALE) Voxel-Based Morphometry Meta-Analysis. We used ALE to determine distal brain regions that exhibited structural deficits in mTLE patients. After identifying the total number of coordinates reported in each of the 14 papers, we excluded those papers reporting 35 or less coordinates. Six papers met this criterion to be included in the analysis. The focus of the main structural deficit cluster within the hippocampus was identified (24,-36,0). ALE Functional Neuroimaging Meta-Analysis. To examine the functional connectivity of the cluster center from the VBM meta-analysis, we drew a 10mm cube around the coordinate within the hippocampus and seeded the BrainMap database to search for all studies that reported activation within the ROI boundary. Whole-brain coordinates of activations were downloaded. The search produced 42 papers reporting activation within the hippocampal ROI. Taken together, these data propose a model of temporal lobe epilepsy that accounts for both functional and structural deficits. Results: Consistent structural deficits in the hippocampus, insula, fusiform gyrus, anterior cingulate, parahippocampus, thalamus and mid-occipital gyrus were identified. Functional connectivity was demonstrated in bilateral parahippocampal gyri, superior temporal gyri, precentral gyri, insula, and putamen. Other regions of functional coherence included the thalamus, red nucleus, substantia nigra, right amygdala, anterior cingulate, and lingual gyrus. Conclusions: The present study used ALE meta-analysis driven connectivity methods to develop a model of temporal lobe epilepsy inclusive of both structural and functional observations. The model holds promise in identifying previously undetected network deficits by capitalizing on the robustness of MACM.