Postictal Modulation of Breathing by the Central Amygdala (CeA) Can Induce Sudden Unexpected Death in Epilepsy (SUDEP).

Rationale SUDEP is the most common cause of death in refractory epilepsy patients. The MORTality in Epilepsy Monitoring Unit Study (MORTEMUS) reported monitored SUDEP cases, and fatal apnea preceded terminal asystole in all patients. The mechanisms involved in seizure‐induced apnea are not known, but evidence suggests the amygdala may be a critical node in the pathway by which seizures spread from the forebrain to the brainstem respiratory network. Depth electrode recordings from the amygdala of epilepsy patients showed that prolonged apneas occur when seizures spread into the CeA. Apneas also occur when the CeA is electrically stimulated in awake patients. Here we examined the role of the amygdala in peri‐ictal apnea and its involvement in SUDEP in two epilepsy models: Scn1a R1407X/+ (Dravet Syndrome; DS) mice and DBA/1 mice. Methods Baseline breathing, hypercapnic ventilatory response (HCVR) and hypoxic ventilatory response (HVR) were assessed using whole‐body plethysmography in DS (n=7) and DBA/1 (n=8) mice before and after lesioning the CeA. Under surgical anesthesia (1–2% isoflurane), electrolytic lesions were induced in the CeA, after which mice were instrumented with EEG, ECG and EMG electrodes. After 5 days of recovery, mice were tethered to a preamplifier and placed in a plethysmography chamber. Seizures were induced in DS mice (n=22) by increasing body temperature 0.5°C per minute using a heat lamp and in DBA/1 mice (n=13) with loud white noise. In a separate cohort of DS mice (n=4), the CeA was stimulated under light anesthesia (0.5–1% isoflurane) by injecting current with stereotactically guided monopolar electrodes while measuring breathing using head‐out plethysmography. Electrode placement was verified post hoc using histology. In a third cohort of DS mice (n=12), CeA lesions or sham lesions were made and mice were video‐monitored for 100 days. A subset of these mice had spontaneous seizures and SUDEP. Results CeA lesions did not affect baseline breathing, HCVR or HVR, and did not prevent seizures in DS or DBA/1 mice. Apnea and death occurred in control (sham‐lesioned) DS mice following heat‐induced seizures in 54.4% of cases (n=11). In contrast, only 9.1% of DS mice with CeA lesions showed respiratory arrest and death after heat‐induced seizures (n=11; p=0.016). Stimulation of the rostral CeA produced apnea that was dependent on the frequency and amplitude of stimulation, without causing seizures. At 50 Hz and 500 μA, apnea was induced for 1–4 minutes without causing death (n=4) or reversal of anesthetic immobility, suggesting lack of an arousal response. Kaplan‐Meier survival curves showed that mice with CeA lesions had higher survival (64.5%) than sham lesions (28.3%) (p=0.015). Conclusion Our data suggest that the CeA can induce apnea but, is not directly involved in baseline respiratory rhythm generation or the ventilatory response to hypercapnia or hypoxia. CeA lesions reduce the risk of apnea during seizures. Therefore, the CeA is a key component of the neural pathway through which seizures spread from the forebrain to respiratory centers in the brainstem to elicit ventilatory arrest after seizures. Support or Funding Information NIH/NINDS U01‐NS0904143 SUDEP Research Alliance