The impact of interictal epileptiform activity on brain networks implicated in SUDEP in patients with epilepsy

Abstract Background: Sudden Unexpected Death in Epilepsy represents the leading cause of epilepsy-related mortality in patients with epilepsy. Previous studies have suggested that SUDEP is associated with dysfunction of neural networks involved in autonomic, cardiac, and respiratory regulation. However, the relationship between interictal epileptiform discharges (IEDs) and functional brain connectivity in patients at increased SUDEP risk remains incompletely understood. Objectives: This study aimed to investigate spike-related functional connectivity alterations in brain networks implicated in SUDEP using simultaneous EEG-fMRI analysis, and to compare connectivity patterns observed in a probable SUDEP case with those identified in patients stratified according to SUDEP risk. Materials and Methods: A cohort of 160 patients with epilepsy underwent simultaneous EEG-fMRI evaluation. Inclusion criteria were age >16 years, survival at follow-up, and significant interictal epileptiform discharges (IEDs) during EEG-fMRI acquisition. Forty patients met these criteria. Among the initial cohort, one patient was subsequently considered a probable Sudden Unexpected Death in Epilepsy case. Patients were stratified using the SUDEP-7 Inventory, SUDEP-3 Inventory, and the annual number of generalized tonic-clonic seizures (GTCS). Twenty-one patients were consistently classified as either very high-risk (vHR) or very low-risk (vLR). Regions of interest (ROIs), including the putamen, thalamus, anterior cingulate cortex, brainstem, and amygdala, were selected based on their role in autonomic and respiratory regulation. Psychophysiological interaction (PPI) analyses were performed to assess spike-related functional connectivity differences between vHR and vLR groups and to compare these patterns with those observed in the probable SUDEP case. Results: The probable SUDEP case demonstrated widespread alterations in spike-related functional connectivity, predominantly characterized by increased connectivity across several cortical regions, particularly involving the right inferior parietal lobule and right inferior frontal gyrus. Group-level analyses partially supported these findings. In the vHR > vLR comparison, increased functional connectivity was observed between the right putamen and left precuneus/DMN, suggesting abnormal cortico-subcortical synchronization in patients at elevated SUDEP risk. Conversely, the vHR < vLR contrast revealed reduced connectivity between the right putamen and left postcentral gyrus, left superior parietal lobule, and between the brainstem and right/left cerebellum. Similarly, the SUDEP case exhibited decreased connectivity between the brainstem and frontal regions, including the superior frontal gyri and the medial frontal gyri. Conclusions: These findings suggest that patients at increased SUDEP risk may exhibit both increased and decreased spike-related functional connectivity within neural networks involved in autonomic regulation and arousal mechanisms. Increased connectivity may reflect maladaptive hyper-synchronization during epileptiform activity, whereas reduced brainstem-related connectivity may indicate impaired integration of autonomic regulatory pathways. Although limited by sample size, these results support the hypothesis that large-scale functional network dysfunction contributes to SUDEP susceptibility and may represent a potential biomarker for future risk stratification studies.