作者
Sirisha Nouduri,Arianna Damiani,Jiahao Chen,Elvira Pirondini,Jorge A. González-Martínez
摘要
INTRODUCTION: Drug-resistant epilepsy (DRE) affects nearly 30% of individuals with epilepsy and remains a major clinical challenge, especially when surgical resection is not feasible. Neuromodulation offers an alternative, but its effectiveness depends on identifying optimal stimulation targets. the pulvinar nucleus (PUL), with its widespread cortical connectivity, is a promising yet understudied target—particularly in posterior quadrant epilepsy. However, the functional roles of its subnuclei during seizures remain poorly defined. We hypothesize that distinct PUL subregions exhibit dynamic interactions with epileptogenic cortex during ictogenesis, and that mapping these interactions through SEEG can guide more effective, targeted neuromodulation. METHODS: We studied a cohort of 11 patients with drug-resistant epilepsy involving the posterior quadrant who underwent clinically indicated SEEG. We retrospectively included patients with: 1) one SEEG contact in the medial PUL and one in the lateral PUL and 2) at least one relevant recording (≥2 seizures, evoked potentials, HDFT, or thalamic stimulation). Patients were excluded for refusal to participate or major SEEG complications (e.g., severe hemorrhage). Electrodes targeted the medial or lateral PUL, based on individual hypotheses about seizure networks. to assess the thalamocortical relationship, we computed h2 correlation coefficients between PUL and cortical contacts within the epileptogenic zone (EZ). Analyses focused on seizure onset and propagation periods to evaluate dynamic functional connectivity during ictogenesis. RESULTS: In the thalamocortical-relevant frequency band (4–8 Hz), we observed distinct patterns of functional connectivity between pulvinar subregions and cortical EZ based on the anatomical location of the seizure focus. Patients with mesial temporal (n=5) and lateral temporal (n=3) EZs exhibited significantly stronger preictal and ictal connectivity with the lateral PUL compared to the medial PUL. This lateralized thalamocortical coupling suggests that the lateral PUL may play a preferential role in the initiation and early propagation of temporal lobe seizures. Notably, this connectivity asymmetry diminished over the course of the seizure, with both PUL subregions showing similar levels of connectivity by seizure termination, potentially reflecting a network-wide synchronization state as the ictal activity resolves. Conversely, in patients with occipital lobe EZs (n=3), the pattern was reversed. These individuals demonstrated greater functional connectivity between the medial PUL and the cortical EZ at seizure onset, suggesting a region-specific PUL-cortical coupling that aligns with the medial PUL’s known anatomical projections to visual and posterior associative cortices. As in the temporal lobe cases, connectivity to both PUL subregions converged by seizure termination, indicating a dynamic reorganization of thalamocortical interactions throughout the ictal evolution. CONCLUSIONS: Our findings demonstrate that distinct PUL subregions exhibit preferential and dynamic functional connectivity with specific cortical EZ during seizure onset and propagation. These results support the hypothesis that the PUL is not a homogenous structure but instead contains functionally distinct subnuclei that differentially participate in ictogenesis, particularly in posterior quadrant epilepsy. By mapping these thalamocortical interactions in vivo using SEEG, we provide foundational evidence that can inform the selection of patient-specific targets for neuromodulation. This work addresses a critical gap in our understanding of thalamic contributions to seizure networks and lays the groundwork for future studies aimed at optimizing DBS strategies through more precise, connectivity-guided targeting of PUL subregions in drug-resistant epilepsy.