Abstracts

Rationale: Frequency, clinical features, and mechanisms of seizure related asystole are poorly understood. This information may promote understanding and prevention of sudden death in epilepsy (SUDEP).Methods: Case series of seizure related asystole encountered during video-monitoring in an adult inpatient epilepsy unit over a 2 year period.Results: Seizure related asystole occurred in 5/504 patients (1%). Age range was 28-60 years, epilepsy duration 1-23 years. 12 lead EKG was normal in 2 and showed sinus bradycardia in 3. Echocardiography and interictal cardiac telemetry (range 3-12 days) were normal in all cases. Only one patient had a risk factor for cardiac disease. Twelve seizures were recorded, 6 with asystole, 3 with significant bradycardia (nadir 20-36 beats per minute), 3 without significant arrhythmia. 6/9 seizures with bradyarrythmia were complex partial, 4/9 started in right hemisphere, and seizure onset area was always hemispheric or parasaggital. Seizure duration at onset of asystole ranged from 34-206 sec. Mean duration of asystole was 24 seconds (range 4-70 sec). One tonic-clonic seizure was followed by severe bilateral EEG suppression, a 70 sec central apnea, and severe oxygen desaturation; in this seizure, a 43 sec asystole started 65 sec after seizure termination. This 'near SUDEP' was the only case of postictal asystole. Cardiac pacemakers were implanted in all cases. Seizures persisted but collapses decreased in all.Conclusions: Gross cardiac pathology does not appear to play an important role in seizure related asystole. Our observations suggest two different mechanisms for this condition. In one group more than 30 seconds of medial hemispheric seizure activity induces ictal asystole and secondary generalization is not necessary. This is consistent with a report of asystole induced by stimulation of the cingulate gyrus (Epileptic Disord 2007;9:77-81). Less frequently, severe cerebral suppression following a tonic-clonic seizure results in prolonged postictal central apnea followed by asystole. This is consistent the animal model of SUDEP and a previously described 'near SUDEP' case (Epilepsia 2000;41:1494-7).