High density probes reveal medullary seizure and rapid medullary shutdown in a model of fatal apnea in seizure.

objectiveSudden unexpected death in epilepsy (SUDEP) is suggested to be a cardiorespiratory collapse that occurs shortly after a seizure. The impacts of seizure on medullary respiratory control remain poorly understood. Prior work in rats suggests that reflexive apneas are highly fatal during seizure but well tolerated otherwise. These reflexes share network connectivity in the medulla, particularly the caudal solitary nucleus (NTS) and ventral respiratory column (VRC), and possibly other intermediate structures. We sought to observe the activity in these regions in fatal ictal apneas.

methodsWe collected data from urethane anesthetized long evans rats. To record neural activity we used either 125 µm silver wire in the caudal NTS or a Neuropixel 1.0 probe along a dorsoventral trajectory that spanned the caudal NTS to the VRC. We additionally recorded cardiorespiratory activity via several methods. We induced a reflexive apnea - the diving reflex - by nasal irrigation of cold water for several seconds, which produces a period of apnea, then gasping, and then a gradual return to eupnea. We repeated several trials while the animal was healthy and subsequently induced continuous seizure activity with kainate and repeated the reflexes, which are ultimately fatal during seizure.

resultsSeizure activity confounds many established methods of analyzing high-density single unit data such as provided by Neuropixels probes, and so our analyses focus on averaging responses over larger anatomical regions (120 µm) covering small populations of neurons. Seizure produces broad increases in neuronal activity across the medullary tract, which by itself is not dangerous. Ictal reflexive apneas were broadly more inhibitory (producing a reduction in firing rate) than they were preictally, and fatal ictal responses resulted in a very rapid shutdown of all medullary activity. We only rarely observed ictal central apneas (apneas with no apparent stimuli), but when we did they were apparently safe, always survived, and produced no significant change in network activity (neither increase nor decrease).

conclusionsThese data support the theory that central apnea events in seizure are relatively safe as we observed they produce little change in the medullary tract network, while stimuli-induced-reflexive-apneas are dangerous because they produce profound quieting across respiratory centers. Our data suggest that seizure spreads to this medullary tract at approximately the same rate and intensity as forebrain, as previously described in this model. These data are supportive of SUDEP mechanisms involving brainstem inhibition as a primary cause, such as spreading depolarization waves. These findings likely extend beyond nasal irrigation to any sensory reflexive apnea caused by airway irritation of any kind, and may bear relevance to similar deaths seen in infants.