A Novel Method for Capturing the Visual Evoked Potential During Spine Surgery Under Total Intravenous Anesthesia

Spine surgery is associated with a variety of post-operative visual complications, ranging from mild, transient deficits in acuity to full visual loss. Known risk factors for visual loss include anemia, hypotension, and long surgical times in the prone position under general anesthesia. Despite recognition and control of risk factors, the prevalence of visual loss after prone spine surgery is estimated at approximately 0.1-1% . The pathophysiology of visual loss most often reflects posterior ischemic optic neuropathy, but anterior ischemic optic neuropathy and posterior reversible encephalopathy syndrome have also been described.

These reports suggest that the entire pathway from retina to visual cortex is vulnerable to injuries causing post-operative visual deficits. The potential for severe injury during spine surgery mandates improved detection of evolving optic tract injury during prone spine surgery.

Theoretically, the flash visual evoked potential (VEP) represents a useful intraoperative tool to monitor global visual system function. However, its application to spine surgery has been limited due to challenges in acquiring and interpreting signals intraoperatively. Several devices for intraoperative VEP-recording have been described . However, most of these have been trialed with patients in the supine or beach chair position, undergoing neurosurgical procedures. Goggle or eyepatch-designed devices have the potential to move during surgery, changing the relationship between the goggle and globe. The goggle itself may potentially compress the globe, producing inconsistent illumination to the retina or causing mechanical injury. These are particular liabilities during changes in position from supine to prone, as is required for many spine surgeries. Additionally, VEPs are extremely variable and exquisitely sensitive to anesthetics- particularly inhaled agents. Consistent with this, intravenous (rather than inhaled) agents have been associated with successful intraoperative VEP monitoring patients, particularly in patients without baseline visual impairment. Nonetheless, there are multiple reports in the literature showing poor correlation/prediction of postoperative visual status by intraoperative VEP measurements. A novel technology for capturing VEPs has recently been developed and may address several of the liabilities described above. The device is constructed from soft foam padding in a "ski mask" design, which we hypothesize may add a measure of protection to the eyes compared to traditional VEP goggles or patches. This design may also provide stability during changes in patient positioning from supine to prone. The front of the mask consists of an opaque circuit board with six bright light emitting diodes positioned in front of each eye, allowing illumination of the retina through closed, protected eyes. Opaque foam separates right and left sides of the mask, allowing each eye to be stimulated individually. To date, the mask has not been tested for sensitivity to elicit and record VEPs in patients under anesthesia. In this pilot study, we tested the feasibility of using the mask during prone spine surgery. Specifically, we asked if VEPs can be detected throughout spine surgeries of varying complexity and duration, whether the signals change with changes in patient positioning and anesthetic conditions, and whether the mask was associated with any adverse effects. Here we show that intraoperative VEPs can be recorded throughout spine surgery using this device. We also describe effects of several anesthetic agents on VEP stability, and ask if there is any qualitative relationship between changes in VEPs and SSEPs recorded simultaneously.