Abstracts

Rationale: By analyzing High Frequency Oscillations in the SEEGs of epileptic patients, we have observed a continuous/semi-continuous (C/SC) ripple activity in some mesial temporal structure (MTS) channels, which we defined continuous High Frequency Activity (HFA)(Mari et al., Epilepsia 2012;53:797-806). We now extend the search for this activity evaluating its potential occurrence also in neocortical (NC) areas. We systematically reviewed and classified the background of each SEEG channel of 20 patients implanted in the MTS and NC regions, defining correlations with neurophysiological and neuro-radiological features to provide an accurate definition of the clinical significance of this particular SEEG pattern. Methods: 20 patients with intracranial electrodes placed in the MTS and NC areas were selected, enabling an extensive coverage of many different brain regions. The anatomical localization of the electrodes in each lobe was defined by post-implantation CT and post-explantation MRI. One minute of slow-wave sleep, low-pass filtered at 500-Hz and sampled at 2000 Hz, was reviewed. The background was analyzed on segments 1 second away from spikes (time resolution: 0.92 s/page; LF:80 Hz). The background was classified as C/SC, Irregular or Sporadic, based on the length of the fast oscillations and on the presence of a clear separation between the transient elements. Ripples and fast ripples were marked independently of background classification. Each channel was classified as inside the Seizure Onset Zone (SOZ) or in a visible lesion. Results: 790 bipolar channels (688 NC, 102 MTS) were analyzed. The C/SC background was found in 54 channels (22 MTS, 32 NC). Considering the anatomical distribution, the most frequent localization in the MTS was the hippocampus (χ2(2,22)= 14.5, p<0.001); in the NC, this pattern was not randomly distributed (χ2(4,32)= 60.6, p<<0.001), being the most common the occipital lobe (fig.1); this region showed 63% of all the NC C/SC background. In the occipital lobe the C/SC background was independent from the occurrence of the alpha rhythm (no correlation between the power spectrum of alpha and ripple band during eyes open/closed). The C/SC background showed low sensitivity (16%) and high specificity (96%) for the SOZ, and no correlation with lesional channels; it had the highest rates of spikes, ripples and fast ripples compared to the Irregular and Sporadic backgrounds (p<<0.01, K-W test). Conclusions: This study provides a further description of a particular type of HFA as a continuous oscillatory activity, and defines its occurrence in neocortical areas as well as in MTS. This continuous HFA, quite rare (7% of the channels), shows a clear preferential distribution for the hippocampus and the occipital lobe. The continuous HFA characterizes ‘normal' channels as well as those inside the SOZ or in a lesion, showing no correlation with behavioral changes (appearance of the alpha rhythm). These findings suggest that this pattern may be an intrinsic characteristic of specific brain regions. Supported by Robb fellowship and CIHR MOP-102710