Hybrid Fluoroscopic and Neurophysiological Targeting of Responsive Neurostimulation of the Rolandic Cortex

Abstract

Background: Precise targeting of cortical surface electrodes to epileptogenic regions defined by anatomic and electrophysiological guideposts remains a surgical challenge during implantation of responsive neurostimulation (RNS) devices.

Objective: To describe a hybrid fluoroscopic and neurophysiological technique for targeting of subdural cortical surface electrodes to anatomic regions with limited direct visualization, such as the interhemispheric fissure.

Methods: Intraoperative two-dimensional (2D) fluoroscopy was used to colocalize and align an electrode for permanent device implantation with a temporary in Situ electrode placed for extraoperative seizure mapping. Intraoperative phase reversal mapping technique was performed to distinguish primary somatosensory and motor cortex.

Results: We applied these techniques to optimize placement of an interhemispheric strip electrode connected to a responsive neurostimulator system for detection and treatment of seizures arising from a large perirolandic cortical malformation. Intraoperative neuromonitoring (IONM) phase reversal technique facilitated neuroanatomic mapping and electrode placement.

Conclusion: In challenging-to-access anatomic regions, fluoroscopy and intraoperative neurophysiology can be employed to augment targeting of neuromodulation electrodes to the site of seizure onset zone or specific neurophysiological biomarkers of clinical interest while minimizing brain retraction.

Keywords: Cortical electrode; Epilepsy; Fluoroscopy; Physiological mapping; RNS.

FIGURE 1.

Phase II intracranial implantation of grid and strip electrodes within a left fronto-parietal cortical dysplasia. During phase I evaluation, an epilepsy protocol MRI demonstrated a large left frontal-parietal cortical dysplasia. A, Axial (left) and sagittal (right) T1-weighted magnetic resonance (MR) demonstrated left parietal pachygyria and loss of grey-white differentiation (arrows). Axial T2-FLAIR MR image (middle) demonstrating white matter hyperintensity in the region of the parietal dysplasia (arrows). In addition, interictal positron emission tomography imaging (PET) demonstrated left posterior frontal and parietal hypometabolism overlapping the anatomic region of cortical dysplasia. Subtracted ictal and interictal single-photo emission computed tomography (SPECT) using injection of technetium-99m bicisate (Neurolite®, Billerica, Massachusetts) demonstrated increased cerebral blood flow colocalizing within the dysplasia. B, Intraoperative view of the cortical surface overlaying the region of dysplasia (asterisk) before (left) and after (right) phase II grid and depth electrode placement. The LIHS label indicates the introduction point of the interhemispheric lead. The silk tie (left, white arrow) represents the location of the central sulcus as demarcated by the intraoperative phase reversal. To define the depth of epileptiform activity, we used intraoperative frameless stereotaxy (StealthStation S8, Medtronic, Dublin, Ireland) to place 4 depth electrodes (Ad-Tech Medical, Oak Creek, Wisconsin) aimed perpendicular to the gyral surface and targeted to a depth of 5 cm (depth of dysplasia), at the anatomic boundaries of dysplasia (numerals on brain). C, ECoG recorded from the interhemispheric strip demonstrating maximal spiking amplitude across contacts 4 to 6. D, Postoperative 3D model of interhemispheric lead (CT imaging) fused with a patient-derived brain surface model.

FIGURE 2.

Fluoroscopic and neurophysiological guidance of interhemispheric strip electrode for RNS. Baseline fluoroscopic imaging was obtained as a reference point for placement of the RNS lead A. The lead was advanced into desired alignment B under intermittent fluoroscopic guidance. The in situ phase II interhemispheric lead was withdrawn under intermittent fluoroscopic imaging to confirm no movement of implanted RNS lead C. D, Intraoperative ECoG from the interhemispheric strip (LIHS) at the time of RNS placement demonstrated maximal amplitude epileptic spiking with phase reversal between contacts 4 and 6. Recording parameters including a high-pass (HP) filter of 1.6 Hz and a low-pass (LP) of 70 Hz. E, A total of 2 representative 3-s epochs of intraoperative ECoG tracings from the RNS lead demonstrated highest amplitude epileptic spiking from contacts 3 to 4, correlating with the location of contacts 4 to 5 on the phase II electrode (LIHS). RNS channels 1 to 2 (top) and 3 to 4 (bottom) represent bipolar montages of the respective channels. Black bars represent 1 s. F, The patient's T1 precontrast MR was utilized to generate a 3D volumetric cortical surface rendering. The implanted RNS interhemispheric lead (white) final position was determined using postoperative CT imaging and MRI colocalization (top). The phase II interhemispheric lead (LIHS) position from postimplant CT imaging was superimposed on a patient-specific 3D cortical surface rendering (bottom). Discussion of infection risks of implantation of permanent RNS electrodes concurrent with Phase II explant is provided in the Supplemental Discussion.

FIGURE 3.

RNS interhemispheric lead recording SSEPs. A, Schematic detailing placement of a 4-contact interhemispheric RNS lead, which was then connected via Oscor cables to the IONM platform for recording of SSEPs. Number at connected end of electrode corresponds to strip electrode contact number (NeuroPace Cortical Strip Lead: CL-325-10-K). B, N20 SSEPs were elicited with a mean latency of 21.6 ms after median nerve stimulation on channels 2 to 4, indicating contact 1 was nearest to primary motor cortex.