Abstracts

This abstract has been invited to present during the Neuroimaging platform session

Rationale: Drug-refractory temporal lobe epilepsy (TLE) is often treated by resection of the epileptic focus. Patients with TLE have alterations in fMRI activity in their epileptic foci as well as in distant connected regions.¹ However, the distant effects of the resection of a large region on brain activity in TLE is not well understood. Here we sought to find distant regions with postsurgical changes in brain activity in TLE using resting-state fMRI (rsfMRI). We then hypothesized that these functional changes were mediated by a structural disconnection to the resected region and tested this hypothesis in a subset of TLE patients who had selective amygdalohippocampectomies (SAH) and therefore a very specific resection.

Methods: Twenty-six right TLE patients who underwent surgery (18 SAH, 8 temporal lobectomies) took part in this study. Subjects underwent a T1-weighted scan (1x1x1 mm³), two 10 min rsfMRI scans (TR = 2 s, 3x3x4 mm³), and diffusion-weighted imaging (DWI, 2.5x2.5x2.5 mm³, 92 directions, b = 1600 s/mm²) on a 3T MRI before and after surgery (26.2 ± 16.2 mos. postsurgery). A total of 111 regions were segmented from the T1-weighted scan._x000D_  _x000D_ FMRI activity was calculated as the voxel-wise amplitude of low frequency fluctuations by summing the frequency spectra of the preprocessed fMRI data across the 0.0083-0.1 Hz band, and normalizing across the grey matter.¹ Postsurgical change in fMRI activity was tested region-wise, followed by voxel-wise tests within the thalami using paired t-tests with family-wise error correction._x000D_  _x000D_ The thalamic cluster from the fMRI analysis was seeded for tractography in patients who underwent SAH. Presurgical structural connectivity (SC, weighted streamline count) and fractional anisotropy (FA) were computed between the thalamic cluster and the resected region (defined as right anterior hippocampus) to estimate structural disconnection caused by surgery. Right-tailed Spearman correlations were run between postsurgical change in fMRI activity in the thalamic cluster and the two estimates of structural disconnection.

Results: FMRI activity was increased postsurgery in the right thalamus, fusiform gyrus, and transverse temporal gyrus and decreased postsurgery in the right opercular inferior frontal and the left supramarginal gyri (p_FWE < 0.05). Voxel-wise analysis in the thalamus showed that the anterior, mediodorsal, and pulvinar nuclei of the right thalamus had increased fMRI activity postsurgery (p_FWE < 0.05, Figure 1A). Postsurgical increase in fMRI activity in the thalamic cluster was positively related with both presurgical SC (r_s = 0.41, p = 0.051, Figure 2A) and FA (r_s = 0.66, p = 0.0024, Figure 2B) from the thalamic cluster to the resected region.

Conclusions: We found that (1) TLE surgery leads to functional changes distant from the resected region, (2) these changes are present in thalamic nuclei that have been implicated in TLE, and (3) these functional changes are related to structural connectivity to the resected region prior to surgery. This structure-function relationship may provide insight into widespread surgical and stimulation effects in TLE._x000D_ _x000D_ Reference:_x000D_ 1. Sainburg et al., Cereb Cortex. 2022

Funding: NIH T32 EB021937, R01 NS075270, R01 NS108445, R01 NS110130, R00 NS097618