Functional Connectivity Alterations Based on Hypometabolic Region May Predict Clinical Prognosis of Temporal Lobe Epilepsy: A Simultaneous 18F-FDG PET/fMRI Study.

Abstract

Simple SummaryAccurate localization of the epileptogenic zone and understanding the related whole-brain functional alteration patterns are critical for the prediction of postsurgical outcomes in patients with temporal lobe epilepsy (TLE). Previous studies suggested that functional connectivity (FC) alterations based on structurally abnormal lesion on MRI could indicate the clinical prognosis. However, hypometabolic regions localized by 18F-FDG PET precedes the appearance of structural abnormalities in lesion localization. By using hybrid 18F-FDG PET/fMRI, we aim to obtain spatially and temporally matched structural, functional, and metabolic information of TLE patients, localize their hypometabolic regions, compare the differences of FC alteration patterns based on hypometabolic region and structural lesion, respectively, and further explore their relationships with disease progression and postsurgical outcome. The results showed that hypometabolic region could be used for detecting a more extensive long-range functional alteration pattern, which correlated with disease progression (epilepsy duration) and surgical outcome (seizure-free or not in follow-up). In this way, we conclude that FC alterations based on hypometabolic region constructed by 18F-FDG PET/fMRI may provide additional value in the prediction of clinical prognosis in TLE patients.(1) Background: Accurate localization of the epileptogenic zone and understanding the related functional connectivity (FC) alterations are critical for the prediction of clinical prognosis in patients with temporal lobe epilepsy (TLE). We aim to localize the hypometabolic region in TLE patients, compare the differences in FC alterations based on hypometabolic region and structural lesion, respectively, and explore their relationships with clinical prognosis. (2) Methods: Thirty-two TLE patients and 26 controls were recruited. Patients underwent 18F-FDG PET/MR scan, surgical treatment, and a 2–3-year follow-up. Visual assessment and voxel-wise analyses were performed to identify hypometabolic regions. ROI-based FC analyses were performed. Relationships between clinical prognosis and FC values were performed by using Pearson correlation analyses and receiver operating characteristic (ROC) analysis. (3) Results: Hypometabolic regions in TLE patients were found in the ipsilateral hippocampus, parahippocampal gyrus, and temporal lobe (p < 0.001). Functional alterations based on hypometabolic regions showed a more extensive whole-brain FC reduction. FC values of these regions negatively correlated with epilepsy duration (p < 0.05), and the ROC curve of them showed significant accuracy in predicting postsurgical outcome. (4) Conclusions: In TLE patients, FC related with hypometabolic region obtained by PET/fMRI may provide value in the prediction of disease progression and seizure-free outcome.