Abstracts

Rationale: Epileptic spasm (ES) has slow waves (SWs) as characteristic components on both scalp and intracranial EEG. Interictal SWs are well known to follow the spike in the interictal epileptiform discharge (IIED). Kobayashi et al. reported SWs during ES with negativity over the posterior head regions or abnormal brain lesions on scalp EEG. There was no report analyzing SWs during ES on intracranial EEG. We analyzed spatial distributions of SWs and high-frequency oscillations (HFOs) on intracranial EEG during both ES and IIED. We hypothesize the SWs during ES differ from those during IIED.Methods: We studied 12 patients with intractable epilepsy with ES, who underwent intracranial EEG prior to epilepsy surgery. EEG was recorded at 1 kHz sampling rate. The number of EEG electrodes ranged from 101 to 125 (mean; 114). We selected ESs with definite deflections of deltoid EMG for time lock of SWs and HFOs. We selected IIEDs with similar locations to ESs. To analyze SWs, we applied 5Hz low-pass filter to the EEG epochs (2 seconds associated with deltoid EMG discharges for ES, and 500ms during SW of spike-wave complex for IIED). To analyze HFOs, we applied 80-200Hz band-pass filter to the EEG epochs (500ms associated with deltoid EMG discharges for ES, and 150ms during spike of spike-wave complex for IIED). To obtain power topographic maps, we overlaid the root-mean-square (RMS) amplitude of SWs and HFOs during the EEG epochs for each electrode onto the picture of brain surface. We applied Kittler s threshold method to detect electrodes with prominent RMS amplitude of SWs and HFOs. We calculated the presence rate of HFOs and SWs with prominent RMS amplitude for each electrode. We evaluated spatial overlap between HFOs and SWs during both ESs and IIEDs using p-value (<0.05) of the correlation analysis. We obtained approval from The Research Ethics Board at The Hospital for Sick Children.Results: We analyzed 7-30 ESs (mean; 18) and 20-39 IIEDs (mean; 29) for each patient. The average number of electrodes with prominent HFOs during ESs ranged 6-38 (mean; 22), and those during IIEDs ranged 4-26 (mean; 13).The average number of electrodes with prominent SWs during ESs ranged 3-42 (mean; 18), and those during IIEDs ranged 5-36 (mean; 14). In ESs, there was no overlap between SWs and HFOs distributions in seven (58 %) of 12 patients, while the other five patients showed the overlap (r=0.22 to 0.60, mean; 0.334, p<0.0001 to p= 0.031). In IIEDs, 11 (92%) of 12 patients showed the overlap between SW and HFO distributions with high correlation coefficient (r=0.35 to 0.88, mean; 0.53, p <0.0001).Conclusions: The spatial distributions of SWs and HFOs less overlapped during ESs than those during IIEDs. While SWs in IIEDs correspond to local inhibition in the cortex, ESs may have unique mechanisms to project characteristic SWs.