Abstracts

Rationale: Both spikes and high-frequency oscillations (HFOs) can serve as electrocorticographic markers of the epileptogenicity. Both appear less frequently in rapid eye movement (REM) than in non-REM sleep. Interestingly, spikes were more lateralized to the epileptogenic hemisphere during REM sleep in a recent study of children with tuberous sclerosis complex (Ochi A, et al. Epilepsia 2011;52:1992). However, another recent study demonstrated no such difference in the HFO distribution between REM and non-REM sleep (Bagshaw AP, et al. Epilepsia 2009;50:627), possibly because of the small number of subjects and/or small cortical sampling with depth electrodes. We tested whether REM sleep not only decreases the occurrence of HFO but also concentrates the HFO distribution toward the epileptogenic zone. Methods: The subjects comprised 13 patients (average age, 25.4 years; range, 12-41 years; 5 males) with drug-resistant epilepsy who underwent extraoperative intracranial electroencephalography (EEG) monitoring by a combination of depth electrodes (median, 18 contacts per patient; range, 0-30 contacts) and subdural electrodes (median, 80 contacts per patient; range, 8-130 contacts). Twelve patients underwent surgical resection and seven achieved freedom from seizures postoperatively. Intracranial EEG signals were sampled and recorded at 1000 Hz simultaneously with scalp EEG and electromyography for sleep staging. The recorded signals were filtered between 80 and 200 Hz, and interictal HFOs were automatically detected on 5- to 15-min EEG samples derived from different sleep stages. HFOs were defined by events above five times the standard deviation of baseline activities and containing at least six consecutive oscillations. The occurrence rate of HFOs was compared between REM sleep and stage N3 sleep (NREM3). The correlation between the areas of surgical resection and electrodes presenting HFOs at a higher occurrence rate during REM than NREM3 sleep was associated with postoperative freedom from seizures. Results: In total, 22,158 and 5,396 HFOs were identified during NREM3 and REM sleep, respectively, from 730 and 437 electrodes of a total of 1,071 intracranial electrodes across all patients. The occurrence rate of HFOs was significantly lower during REM sleep (mean, 0.3/min; range, 0.0-11.7/min) than during NREM3 sleep (mean, 1.4/min; range, 0.0-47.4/min) (p < 0.0001, paired t-test). The occurrence rate of HFOs was higher during REM sleep than during NREM3 sleep in a total of 21 electrodes among 6 patients. Inclusion of these electrodes in surgical resection was significantly associated with postoperative freedom from seizures (p < 0.01, chi-square test). Conclusions: Sleep stages influence the occurrence of HFOs. Generally, the occurrence rate of HFOs was lower during REM sleep than during NREM3 sleep. However, a minority of electrodes exhibiting a less suppressive effect of REM sleep on the occurrence of HFOs can serve as a specific marker of the epileptogenic zone.