Abstracts

Rationale: Interictal high frequency oscillations (HFOs) are one of the possible biomarkers of epileptogenicity. The occurrence rate of HFOs depends on the state of vigilance level. Previous studies have reported that the rate of HFOs is the highest in non-REM sleep, but is low in wakefulness and in the REM sleep. Propofol is a short acting intravenous agent used for induction and maintenance of general anesthesia, and has two opposite or reciprocal, anticonvulsive vs. proconvulsive, properties (D San-juan et al. 2010 Clin Neurophysiology). Some previous reports presented that occurrence rate of HFOs increased after propofol cessation probably reflecting its anticonvulsive action. However, it has not been elucidated the influence of natural sleep and general anesthesia on HFOs in the same patients. In this study, by comparing the rate of HFOs of the seizure onset zone (SOZ) between natural sleep and intraoperative general anesthesia in the same patients, we aimed to clarify the adequate vigilance state of patient for epileptic focus examination. Methods: We recruited 4 patients (2 males and 2 females, mean age of 26.8 years, ranged from 16 to 39) with medically intractable partial epilepsy who underwent chronic subdural electrode implantation. One patient suffered from mesial temporal lobe epilepsy and the three from neocortical epilepsy. We recorded electrocorticograms (ECoGs) for wakefulness (W), light sleep (L), slow-wave sleep (S), REM sleep (R), and under propofol anesthesia (Pro) with 1-2 kHz of sampling rate. The staging of natural sleep was done based on the standard Rechtschaffen and Kales criteria. During the recording of ECoGs of Pro, we kept its blood concentration between 2.0 to 3.0μg/ml by Target Control Infusion (TCI) function. We excluded the periods with clear artifacts from the subsequent analysis. By using modified automatic HFO detection algorithm (von Ellenrieder et al. 2012 Clinical Neurophysiology) to five-minute ECoGs of each state, we calculated the rate (events/min) of ripples (80-250 Hz), and fast ripples (>250Hz) in 1-2 pairs of bipolar electrodes on the seizure onset zone. In the algorithm, we applied two finite impulse response filters, and extracted HFOs presenting no less than 4 oscillatory activities in comparison with background activities. Results: In 4 natural states (W, L, S, and R), the ripple rate (events/min) were 10.0±4.1, 12.5±3.5, 23.3±9.3, and 3.7±0.8, and the fast ripple rates were 3.1±1.7, 4.5±2.0, 11.5±5.9, 1.1±0.2 respectively (mean±S.E.). They were highest in S and lowest in R. In Pro, ripple rate was 26.1±11.9 and fast ripple rate was 8.0±5.4, showing the similar behavior to those in S. Conclusions: We investigated the influence of natural sleep and general anesthesia on HFOs in 4 patients. The occurrence rate of HFOs on the SOZ in Pro was as high as that in S in the same patients. Funding: ‘Non-linear Neuro-oscillology: Towards Integrative Understanding of Human Nature’ (KAKENHI 15H05874) to A.I. and KAKENHI 15K10361 to T.K.