Abstracts

Rationale:
The presence of frequent interictal spike waves in electroencephalogram (EEG) and intracranial EEG data is a hallmark of focal epilepsy. While increases in the prevalence of interictal spikes during sleep is commonly reported in scalp studies, little is known about how the frequency of interictal epileptiform discharges changes during periods of sleep vs. wakefulness intracranially.  The analysis of intracranial recordings also provides a unique opportunity to investigate differences in the frequency of spike waves over an extended period of time during periods of both sleep and wakefulness, without contamination by muscle artifacts (which may especially bias spike detection during wakefulness).
Method:
One night of sleep free of seizures was selected for 6 patients implanted with intracranial electrodes. 3 patients were implanted with depth electrodes (59, 16, 24 electrodes in cingulate and hippocampal areas) and 3 patients with a combination of grid and strip subdural electrodes (55, 94, 80 electrodes in fronto-temporal and parietal areas). After manual sleep scoring (identifying non-rapid eye movement sleep by increased delta/beta ratio and the presence of slow waves and spindles), noisy channels and epochs were rejected, and interictal spikes were manually marked for each patient. Average spike frequencies were calculated for the first and last hours of both sleep and wake. We then assessed if the frequency of interictal spikes was higher during sleep or during wake and if there were consistent changes in their frequency overnight.
Results:
When comparing the overall spike frequency during sleep versus wake, we found that 3/6 patients displayed a higher spike rate during wake (20.5 vs 9.7, 9.9 vs 3.8, 1.1 vs 0.3 spikes/min) and 3/6 a higher frequency during sleep (10.9 vs 13.8, 20.6 vs 32.3, 4.1 vs 10.8 spikes/min). Visual analysis suggested that that patients with higher frequency of spikes during sleep compared to wake displayed higher amplitude spike waves. In contrast, interictal spike frequency showed a consistent decline between the first and last hours of sleep in 5/6 patients (18.1 vs 8.3, 3.9 vs 2.0, 41.2 vs 28.6, 14.0 vs 7.3, 0.7 vs 0.4 spikes/min) – with the only patient without significant decline (5.66 vs 9.14 spikes/min) displaying smaller amplitude spike wave patterns localized to only a few of the intracranial channels. Spike frequency was also decreased during wake just after sleep compared to wake before sleep (average: 8.0 vs 4.3 spikes/min) in 4/6 patients.
Conclusion:
We observed that state-dependent changes in the frequency of interictal discharges displayed some inter-individual variability, with only half of the patients showing higher frequency of spikes during sleep compared to wake. Overnight decline in spike frequency was more consistently observed (in 5/6 subjects). Future studies will attempt to link quantitative features of time-frequency decomposition and spatial topography of interictal discharges with the observed inter-individual variability.
Funding:
:Tiny blue dot foundation, Lily fund for epilepsy research