Abstracts

Rationale: MRI analyses in temporal lobe epilepsy (TLE) have focused on cortical morphology by means of volumetry, voxel-based morphometry, and thickness measurements. As these metrics reflect a complex combination of biological events, substrates of cortical anomalies remain incompletely understood. Novel imaging contrasts provide metrics of cortical microstructure in clinically-feasible acquisition times. Particularly, quantitative T1 mapping offers a window into cortical myelin. Given the link between myelin and MRI contrast, earlier studies have suggested that myelin alterations may contribute to measurable atrophy. Due to the role of myelin in axonal conductivity, alterations may also reflect anomalies in inter-regional connectivity. Here, we utilized quantitative T1 mapping to probe intracortical myelin in TLE in-vivo. In addition to assessing the topography of myelin changes, systematic vertical profiling assessed alterations at different intracortical depths. Surface-wide association analyses furthermore established the relation between T1 changes, markers of atrophy, functional connectivity, and baseline myeoloarchitecture. Methods: We studied 24 drug-resistant TLE patients and age-and sex-matched 22 healthy individuals using multi-modal 3T MRI. For quantitative T1 mapping, we acquired a MP2RAGE sequence, which combines two images at different inversion times to provide intrinsic bias field cancellation. We also obtained a T1-weighted MPRAGE and 2D EPI-BOLD resting-state fMRI scan. T1-maps and fMRI time series were linearly co-registered to T1-weighted images, and registered to MNI152 space. To examine intracortical T1 times, we positioned three surfaces between the inner (GM-WM) and outer (GM-CSF) cortical surfaces at 25%, 50%, and 75% cortical thickness to sample the axis perpendicular to the cortical ribbon. Resting-state data were also co-registered to surface-models. Vertex-wise Student's t-tests mapped alterations in intracortical T1 between patients and controls. Multi-surface profiling assessed the cortical depth at which effect sizes were most marked. Surface-wide analyses assessed correlations between TLE-related T1 increases, cortical thinning, baseline myelin, and hippocampal functional connectivity. Results: Surface-based analysis showed T1 increases in TLE compared to controls in ipsilateral parahippocampal, temporopolar, lateral temporal, and bilateral orbitofrontal regions (P < 0.007; Fig. 1A). Increases were more consistently observed towards the pial surface, and less marked close to the GM/WM interface (Fig. 1B). Compared to controls, patients presented with widespread cortical thinning in bilateral fronto-parietal, and temporal regions that did not resemble patterns of T1 increases. Repeating the comparisons in T1 times after values were corrected for corresponding thickness measures revealed virtually identical findings. Conversely, T1 increases in TLE were highly correlated to baseline myelination and hippocampal functional connectivity (r>