Ictal central apneas in temporal lobe epilepsies

Abstract

The objective of the study was to examine the frequency and characteristics of ictal central apnea (ICA) in a selective cohort of patients with mesial or neocortical temporal lobe epilepsy (TLE) undergoing surface video-electroencephalography (EEG) and multimodal recording of cardiorespiratory parameters. We retrospectively screened 453 patients who underwent EEG in a single center including nasal airflow measurements, respiratory inductance plethysmography of thoracoabdominal excursions, peripheral capillary oxygen saturation, and electrocardiography. Patients with confirmed TLE subtype, either by magnetic resonance imaging (MRI) lesions limited to the temporal neocortex or mesial structures and concordant neurophysiologic data, or patients who underwent invasive explorations were included. Ictal central apnea frequency and characteristics were analyzed in 41 patients with 164 seizures that had multimodal respiratory monitoring. The total occurrence of ICA in all seizures in this cohort was 79.9%. No significant difference was seen between mesial and neocortical temporal lobe seizures (79.8% and 80.0%, respectively). Ictal central apnea preceded EEG onset by 13 ± 11 s in 33.3% of seizures and was the first clinical sign by 18 ± 14 s in 48.7%. Longer ICA duration trended towards a more severe degree of hypoxemia. In a selective cohort of TLE defined by MRI lesion and/or intracranial recordings, the frequency of ICA was higher than previously reported in the literature. Multimodal respiratory monitoring has localizing value and is generally well tolerated. Ictal central apnea preceded both EEG on scalp recordings as well as clinical seizure onset in a substantial number of patients. Respiratory monitoring and ICA detection is even more paramount during invasive monitoring to confirm that the recorded seizure onset is seen before the first clinical sign.