Authors :
Masud Seyal, University of California, Davis; fady Girgis, University of California, Davis; Kiarash Shahlaie, University of California, Davis; Katherine Park, University of California, Davis
Rationale: Sudden unexpected death in epilepsy (SUDEP) is the most common cause of death in epilepsy patients. Respiratory dysfunction is implicated as a critical factor in SUDEP pathophysiology. In the MORTEMUS study, patients dying of SUDEP in epilepsy monitoring units had postictal tachypnea, followed by apnea starting 25-180 sec postictally, then terminal asystole (Ryvlin P et al 2014).
Human studies show that electrical stimulation of the amygdala results in apnea (Dlouhy BJ et al. 2015; Lacuey N et al. 2017; Nobis WP et al. 2018) indicating that the amygdala has a role in respiration control. Amygdala stimulation resulted in immediate onset of respiratory dysfunction occurring only during nose breathing (Nobis WP et al. 2018). In a single patient, spontaneous seizures resulted in apnea occurring within one sec of seizure spread to the amygdala (Dlouhy BJ et al 2015). In homozygous Cacna1a mice, cortical seizures induced spreading depolarization in the brainstem within 8-24 sec coinciding with apnea, followed by cardiac arrest and death (Loonen ICM et al. 2019).
To determine whether spontaneous seizure onsets occurring in the amygdala or if seizure spread to the amygdala was necessary and sufficient to cause apnea, we investigated the temporal relationship between apnea onset and initial seizure involvement within the amygdala in patients with implanted depth electrodes.
Methods: Patients had pharmacoresistant focal epilepsy. Intracranial electrodes had been placed as determined during multidisciplinary presurgical assessment. Patients were selected if depth electrodes had been implanted in the amygdala. Implantation technique involved stereotactic targeting using a frame. Postoperative CT scan was fused with the pre-operative MRI to confirm electrode location. All patients in this EMU have synchronized recordings of nasal airflow, abdominal excursions and digital pulse oximetry. A subset of patients was selected for analysis if either a usable nasal airflow signal, respiration-related abdominal excursion signal or both were available at seizure onset. Apneas were defined as cessation of either the nasal airflow or abdominal excursion signal for 5 sec or longer in proximity to seizure onset. Hypopneas were defined as a greater than 50% drop in the amplitude of these signals. Results: Data from 16 patients (45 seizures) were analyzed. There was no apnea with 4 seizures (2 patients) that involved the amygdala. With 6 seizures (3 patients) apnea/hypopnea onset preceded amygdala seizure involvement by 3 to 55 sec. With 7 seizures apnea/hypopnea occurred within 2 sec of seizures involving the amygdala. With the remaining 25 seizures, apnea started >2 sec after seizure involvement of the amygdala (range 3-158 sec) (Figure 1). There was a significant difference between apnea onset and amygdala seizure onset (p<0.001). Following seizure onset, the mean amygdala seizure onset was 39.2 (range 0-1222) sec and mean apnea onset was 60.2 (range -11-1167) sec. Conclusions: The wide range of apnea/hypopnea onset times relative to initial amygdala seizure involvement, amygdala seizure involvement without apnea, and apneas preceding amygdala seizures suggest that amygdala involvement may not be crucial to inducing apneas with spontaneous seizures and may require involvement of other brain regions. Spreading depolarization affecting brainstem respiratory centers rather than seizure involvement restricted to the amygdala may be of greater relevance in SUDEP pathophysiology. Funding: No funding