Abstracts

Rationale: An increasing number of studies suggest that high frequency oscillations (HFO) in the ripple (80-250Hz) and fast ripple band (250-500Hz) may represent surrogate markers of epileptogenic processes. HFO are investigated using invasive EEG recordings during evaluation for epilepsy surgery. Results are used for preoperative focus localization and determination of resection volumes. However, studies investigating HFO detection during the surgical procedure, i.e. in general anaesthesia in human, are lacking. Our study aims to demonstrate existence of intraoperative epileptic ripple HFO and to evaluate the correlation to hippocampal pathologies. Methods: Electrocorticographic data (ECoG) from 15 patients with pharmacoresistant temporal lobe epilepsy were analyzed retrospectively: 9 with hippocampal sclerosis (HS) verified by histology and 1 with mesial ganglioglioma (group 1) versus 6 with no or unspecific hippocampal changes (atrophy/gliosis) (group 2). Data were recorded intraoperatively before resection using subdural electrodes: 3 temporobasal strips(4 contacts each), 1 hippocampal strip (4 contacts) and 20 electrodes over temporolateral areas. In three patients, an additional amygdala electrode (4 contacts) was used. Data was acquired with a sample rate of 1024Hz after application of an analogue highpass filter (0.08Hz) using an IT-med amplifier (Usingen, Germany). Raw data were processed according to the short time line length method described by Gardner et al. (Clin Neurophysiol. 2007; 118(5): 1134-1143), however with the baseline noise level calculated from a user specified segment containing no detectible artefacts or ripple activity. For comparison of resulting ripple rates, spikes were detected using a visually controlled template search procedure. Results: Ripple activity was observed in all patients, however rates of >0.5 ripples/second were achieved only in 13 of all 15 patients. In 14 patients, electrode with maximum ripple rate was located over resected volumes, however no association with a particularly good or bad outcome was seen. Ripples occured superimposed on and independent of spikes. Total rate of ripples and spikes were highly correlated (r=0.91, p<0.0001, pearson correlation). No difference between the groups was observed in regard to total rate of spikes (p=0.51, ranksum test). However, total ripple rate was higher in group 1: 0.24/s (0.13/s-0.40/s) vs. 0.08/s (0.01/s-0.11/s) (p=0.055, ranksum test). Furthermore, a significant difference in hippocampal ripple rates (p<0.005, ranksum test) was observed between groups: 0.03/s (0.02/s-0.17/s, 1st and 3rd quartile) vs. 0/s (0/s-0.002/s).