Abstracts

Rationale: Epileptic seizures produce clinical signs by interacting with the physiologic resting state networks (RSN) of the brain. This study aims to quantify this phenomenon by analyzing the SPES induced gamma deactivation of these networks while stimulating in various lobar epileptic structures. Methods: We analyzed 20 patients with focal epilepsy, implanted with depth electrodes. SPES (20 square biphasic pulses having variable amplitude, 3 ms pulse duration, stimulation currents varied in the 0.25–5 mA range in steps of 0.25 mA) was applied to each pair of adjacent contacts, and responses were recorded from all other contacts. To make sure we included only stimulation induced effects we included only the stimulation-response pairs significantly different from the baseline activity (Mann–Whitney U test of the differences between baseline and poststimulation periods for the 20 response values with a cutoff of p < 0.05). The mean response amplitude value was quantified in two time-periods after stimulation (60-255, 255-500 ms) for the Gamma frequency-range (30-110 Hz), and compared to baseline. Overall we had 32125 stimulation-recorded pairs: 8 % in the limbic network, 9 % in the reward RSN, 0.3% in the subcortical RSN, 6 % in the cingulo-opercular RSN, 9% in the premotor RSN, 6 % in the motor RSN, 9 % in the top-down attention network, 13% in the ventral stream RSN, 5 %  in the visual RSN, 7% in the auditory RSN,  15 % in the default mode network, 6 % in the task positive network, 5% in the verbal network. Effects obtained while stimulating the seizure onset zone (SOZ) were pooled together – 5588 stimulation-response pairs. They were divided by origin and compared with the ones produced by a non-specific stimulation on non-SOZ contacts with two tailed Mann Whitney U test and z-scoring. Results: Overall, we observed a significant global a delayed inhibition (60-500 ms) to 19% (SD 11 %) of baseline gamma power for period one and 17% (SD 10%) for period two.  Concerning the second period, frontal and temporal SOZ contacts deactivated preferentially the ventral stream network (z =  6 and  3). Insular SOZ contacts had a more significant effect on the reward network (z = 1.9) while the occipital on the task positive network and ventral stream (z = 3.3) and the visual RSN (z = 6.2). The parietal SOZ contacts produced a significant lesser deactivation than the average on all RSN, maximally on the verbal and motor RSN (z = - 4). Conclusions: Depending on their anatomic location, epileptic contacts produce a characteristic deactivation of the RSN, possibly explaining both seizures’ semiology and interictal neurocognitive impairment in patients Funding: SEEG procedures, data collection and analysis supported by Romanian UEFISCDI research grants PN-III-P4-ID-PCE-2016-0588, PN-III-P1-1.1-TE-2016-0706 and COFUND-FLAGERA II-SCALES and COFUND-FLAGERA II-CAUSALTOMICS.