Abstract
BACKGROUND AND OBJECTIVES: Peri-ictal apnea is common in persons with epilepsy (PWE) and may contribute to sudden unexpected death in epilepsy (SUDEP). It is not yet known whether brainstem respiratory centers in PWE respond differently to voluntary apnea. We addressed this issue using breath-holding (BH) functional MRI (fMRI). METHODS: Adult PWE were recruited from a single outpatient clinic with the aim of achieving a comparable proportion of patients with and without generalized or focal-to-bilateral tonic-clonic seizures. Age-matched and sex-matched healthy controls were recruited through public advertisements. Exclusion criteria included physiologic and medical conditions known to alter respiration. All participants underwent fMRI while performing voluntary inspiratory and expiratory BH tasks. Respiratory rate, oxygen saturation, and end-tidal O(2) and CO(2) were recorded. Functional data were analyzed using a standard general linear model, as well as seed-to-voxel and ROI-to-ROI connectivity analyses targeting predefined brainstem regions of interest. RESULTS: The study included 31 PWE (mean age 34.0 ± 11.6 years, 51.6% female) and 21 controls (mean age 32.8 ± 9.9, 52.4% female). At the group level, PWE had significantly lower brainstem activation than controls during both expiratory BH (Cohen d = 1.43, 95% CI 0.81-2.05, p = 0.005) and inspiratory BH (Cohen d = 1.31, 95% CI 0.70-1.92, p = 0.006). Decreased activations in PWE were observed in regions corresponding to the cuneiform nucleus during expiratory BH and the median raphe nucleus during the inspiratory BH. BH-triggered fMRI changes at the individual level showed a significant brainstem activation during expiratory BH in 61% of PWE vs 90% of controls. When comparing each participant with the control group, 35% of PWE demonstrated a significantly decreased brainstem activation, vs none of the controls. Intrinsic connectivity during self-paced breathing and BH showed reduced brainstem-cortical connectivity in PWE. DISCUSSION: A significant proportion of PWE seem to suffer from interictal dysfunction of brainstem regions involved in respiratory control. These abnormalities could be detected at the individual level using a simple BH fMRI paradigm, suggesting potential for translation into a clinical biomarker. Future studies are needed to confirm these findings in larger populations and investigate their relation to SUDEP.