Convergence of spatially dispersed atrophy into functional networks in temporal lobe epilepsy: glial-immune signature associated with circuit dysfunction and treatment response.

Temporal lobe epilepsy (TLE) exhibits spatially heterogeneous cortical atrophy, yet how these distributed patterns drive unified pathophysiology remains unknown. Here, we test the hypothesis that locally discrete atrophic regions coalesce into functionally coherent networks associated with unified molecular mechanisms. Using large-scale normative modeling of patient-specific regional atrophy and functional connectivity mapping, we demonstrate that marked variability in atrophic loci coalesce into remarkable functional convergence within temporal-limbic circuitry in TLE patients. This convergence was selectively enriched in glial-immune gene pathways, with transcriptomic analysis revealing astrocytic markers as key molecular substrates. Cross-sectional disease-duration analyses showed functional coupling shifts toward unimodal cortices with disease chronicity, while fronto-parietal network coupling strength predicted anti-seizure medication responsiveness, suggesting a network-based stratification biomarker. Our findings suggest that spatially dispersed cortical atrophy in TLE converges on a shared temporal-limbic functional network. The spatial association of this network with glial-immune transcriptomic signature, together with its associations with disease duration and anti-seizure medication response, provides a multiscale framework for understanding the network-level pathology of TLE.