Abstracts

Rationale: In studies of cortical respiratory control, direct electrical stimulation (DES) of amygdala and hippocampus reliably induces central apnea. Although extra-temporal limbic structures have been associated with breathing control, whether DES of these regions can elicit transient hypoxemia remains unclear. Methods: Evolution of pulse oximetry (SpO2) after 1352 DES was analyzed in 75 patients with refractory focal epilepsy who underwent stereo-electroencephalography recordings. For each DES, we assessed the change in SpO2 from 30 seconds prior to DES onset to 120 seconds following the end of the DES. The primary outcome was occurrence of stimulation induced transient hypoxemia (SITH) as defined by decrease of SpO2  5% within 60 seconds after stimulation onset as compared to pre-DES SpO2 and/or SpO2 nadir <90% during at least 5 seconds. Localization of the stimulated contacts was defined according to MarsAtlas brain parcellation and Freesurfer segmentation. For each spontaneous seizure which occurred during the SEEG monitoring with informative ictal and post-ictal SpO2 data, we analyzed the evolution of the SpO2 during the 3 minutes preceding the seizure onset, during the seizure and up to 10 minutes after the end of the seizure. Post-ictal hypoxemia was defined as SpO2 <90% during at least 5 seconds Results: A stimulation induced transient hypoxemia was observed after 16 DES (1.2%) in ten patients (13%), including six where SpO2 nadir was <90%. Among the 16 DES associated with SITH, seven (44%) were localized within the perisylvian cortex, including two within the circular sulcus of the insula, two within the subcentral gyrus, two within the temporal operculum and one within the frontal operculum. After correction for individual effects and the varying number of DES contributed by each person, significant decrease of SpO2 was significantly associated with the localization of DES (p=0.019). Specifically, SITH was significantly more frequent for DES performed in the perisylvian cortex than in the frontal lobe (Odds ratio (95% CI) 0.10 (0.16-0.67), p=0.017). In comparison with perisylvian cortex, the rate of significant decrease of SpO2 after DES in the temporal neocortex was lower, but this difference did not reach significant value (OR [95% CI] 0.42 [0.16-1.14], p=0.088). There was no difference between perisylvian cortex and orbito-frontal cortex (OR [95% CI] 0.56 [0.08-3.91], p=0.56) or occipital cortex (OR [95% CI] 0.25 [0.03-1.96], p=0.187). There was no relation between evolution of SpO2 following DES and occurrence of hypoxemia during seizure Conclusions: Though rare, a significant decrease of SpO2 could be elicited by cortical direct electrical stimulation outside the temporo-limbic structures, most commonly after stimulation of the perisylvian cortex. Funding: The REPO2MSE study, the SAVE study and the ENALEPSY study have been funded by the French Ministry of Health (PHRC National 2009, 2013 and 2013).B. Chatard has received funding from the “Programme Avenir Lyon Saint-Etienne” of Université de Lyon, within the program “Investissements d’Avenir”, ANR convention n◦ ANR-11- IDEX-0007; the LabEx CORTEX, ANR convention n◦ ANR-11-LABX- 0042; and European Union Seventh Framework Programme (FP7/ 2007-2013) under Grant Agreement No. 604102 (Human Brain Project)