Abstracts

Rationale: Insular cortex epilepsy is the great mimicker due to seizures presenting as frontal or temporal. However, there are some specific signs suggestive of insular seizure onset such as gustatory sensation or pain.

Methods: We retrospectively included patients with drug resistant epilepsy that underwent invasive Stereo-electroencephalographic (SEEG) presurgical work-up at the University Emergency Hospital Bucharest. Selecting criterion was based on an SEEG exploration comprising at least one pair of contacts sampling the insular cortex. As part of the presurgical evaluation patients underwent high frequency direct electrical stimulation (43 Hz, 1 ms biphasic pulse duration, alternating polarity, 5 seconds train duration, amplitude varied based on clinical effects from 0.25 up to 3 mA) to map functional cortex, trigger ictal symptoms and compute effective connectivity. Clinical responses elicited by electrical stimulations were classified intro five physiologic categories (cognitive, emotion, autonomic, motor, sensory) with second and third order subcategories (e.g., sensory, auditory, hallucination, elementary). The effective connectivity metric was chosen as the RMS of the signal filtered using a bandpass comb filter at the stimulation frequency and up to 5th order harmonics, frequencies that for alternating polarity pulses are disjunct from the artefactual components (Barborica et al. Hum Brain Mapp. 2022;43:1657). A threshold of Z=3 for the Z-test statistic comparing the set of RMS values for the responses over a set of 0.2 s subintervals (50% overlap) during the response versus baseline intervals (2 s each) was used to select most robust connections between stimulated site and the rest of the sampled cerebral structures.

Results: We included 35 patients in whom 461 contacts sampled the insular cortex bilaterally. The physiologic clinical effects elicited by stimulations were somatosensory, autonomic (nausea/vasomotor/hypoventilation), cognitive (language), sensory (olfactory, gustatory, auditory) and vestibular. Effective connectivity analysis highlights recruitment of extra-insular region such as opercular, parietal, premotor and motor cortex, mesial temporal lobe, primary somatosensory area. In seven patients, the insular cortex was part of the epileptogenic zone and in five patients the ictal symptoms elicited by stimulation were gustatory, fear/vertigo, gustatory/happiness/heat, and somatosensory. When comparing the networks of similar clinical effects (e.g., gustatory sensation) elicited as a physiologic or epileptogenic sign we identified that there is a partial overlap occurring over the insular cortex, rolandic and frontal operculum (response 5-12 uV) in the example of gustatory hallucinations.

Conclusions: Generation of signs and symptoms during high-frequency stimulation of the insular cortex engages networks involving extra-insular regions. Our analysis highlights that the same network is recruited by the high frequency stimulation when generating a clinical effect in both a physiologic and pathologic setting.  
Funding: PN-III-P4-ID-PCE-2020-0935