Abstracts

Rationale: Because of the interaction between sleep and epilepsy, it is important for both clinical and scientific purposes to understand sleep structure and any changes to it over the course of intracranial EEG (icEEG) monitoring. We evaluated sleep structure over multiple nights during icEEG monitoring. Methods: We studied a total of 28 epilepsy patients who underwent continuous icEEG monitoring for presurgical evaluation, and were monitored for a median of 12 nights (range 6-27). For each patient, we selected a single electrode contact for analysis. We then calculated relative delta power (RDP) at a one-second resolution and used a metric calculated from RDP to evaluate sleep structure during continuous icEEG monitoring. We used multiple methods to verify that RDP and our sleep metric accurately measure sleep structure and not unrelated low-frequency activity in the icEEG. We visually compared RDP to EMG and EOG, related RDP to approximate entropy (ApEn) and the ICU depth of sleep (IDOS) index, and examined RDP and metric values during daytime recordings to confirm that these oscillations in RDP occur exclusively during nights. As a number of factors can influence sleep structure, including seizures and AEDs, we also analyzed the effect of these. We analyzed subsets of the 28 patients to determine the effect of daytime seizures, nighttime seizures, and seizures anytime on sleep structure. In order to assess the effect of AEDs, we grouped relevant AEDs into those that disrupt sleep, improve sleep, and have no significant effect on sleep. We then grouped patients into those that received AED combinations that disrupt sleep (n = 19) and those that received combinations that improve or have no significant effect on sleep (total n = 9: n = 6 for improve, n = 3 for no significant effect). Results: There was a significant improvement in sleep quality as measured by our metric over the term of icEEG monitoring. Seizures and AEDs did not have a significant effect on this robust improving trend. Conclusions: Sleep quality improves during icEEG monitoring, regardless of seizures and AEDs. It is possible that this improvement is due to the patient recovering from surgical implantation of intracranial electrodes, and that the effect of this considerable adjustment outweighs effects due to seizures and AEDs. Our results indicate that sleep is not uniform for every night of icEEG monitoring, and that there is a considerable improving, multi-night trend in sleep structure.