Neuromonitoring in Endoscopic Skull Base Surgery

Electromyography (EMG)

EMG was first used intraoperatively in the 1960s for the monitoring of facial nerve function during exploratory parotid surgery. During endoscopic skull base surgery, EMG can be used for monitoring of any cranial nerve with motor function including cranial nerves III-VII and X-XII. The pathologies involving the cavernous sinus and/or superior orbital fissure often threaten cranial nerves III, IV, & VI. They are monitored by performing an EMG of the extraocular muscles. Because of their relative frequency of use in endoscopic skull base surgery, EMG monitoring of the extraocular muscles will be a particular focus of this chapter. For transclival approaches to prepontine or cerebellopontine angle pathologies, the facial nerve, vagal nerve, accessory nerve, and hypoglossal nerve may additionally be monitored. The functional status of the facial nerve is monitored by recording EMG of the orbicularis oris and orbicularis oculi muscles. Similarly for the monitoring of glossopharyngeal, vagus, accessory, and hypoglossal nerves, EMGs of stylopharyngeus, laryngeal muscles, trapezius, and tongue are recorded, respectively.

Two types of EMG activity are recorded: free running and triggered EMG. Free running EMG continuously records the motor unit potentials (MUP) of the muscle fibers. It has high specificity and negative predictive value regarding postoperative cranial nerve deficits. This provides some degree of confidence to the surgeon that these cranial nerves are not being disrupted during tumor exposure and removal. Based on the amplitude and frequency of discharges, the free running EMG signals can be classified into spikes, bursts, trains and neurotonic discharges. A single MUP wave is called a “spike.” A short chain of MUPs firing at 30–100 Hz and less than 200 ms in duration is called a “burst.” When a persistent chain of MUPs is recorded, it is referred to as a “train.” Bursts and spikes are typically triggered by touching, rubbing, or other mechanical manipulations of the nerve with no correlation to nerve injury. Trains are elicited by mechanical stimuli, saline irrigation, and possibly nerve ischemia

The neurotonic discharges are of primary interest to the neuromonitoring technician. They were first described in the 1980s and are defined as a train of MUPs at high frequency (> 30 Hz) recorded from a muscle in response to mechanical or metabolic stimulation. Since neurotonic discharges are triggered by mechanical stimulation of motor axons, they act as sensitive indicators of nerve injury. But absence of neurotonic discharges doesn’t necessarily exclude nerve injury and presence of neurotonic discharge doesn’t always signify nerve injury. Sharp transection of a nerve elicits negligible neurotonic discharges as compared to mechanical irritation or manipulation. The signal voltage is set between 50 and 200 μV, the frequency filter between 30 Hz to 20 kHz, and the sweep speed is at 100 ms per division for recording the responses.

Triggered EMG activity is seen when the cranial nerve is electrically stimulated. This leads to recording of compound muscle action potentials (CMAPs) from the muscle fibers. Triggered EMGs are needed to check the integrity of peripheral motor axons. CMAPs can be produced by either bipolar or monopolar stimulation. In bipolar stimulation both the cathode and anode are directly on the nerve, which reduces current spread to adjacent nerves leading to localized flow of current. But the localized flow of current may lead to submaximal stimulation if fluid causes current shunting. In monopolar stimulation the cathode is directly on the nerve and the anode is kept away from the nerve by at least several centimeters. This lowers the chances of current shunting but increases the probability of activating nearby neural structures by current spread. Nevertheless, monopolar stimulation is mostly preferred as it is easier to use in confined spaces of the brain. A current of very low intensity (0–2 mA) and duration (0.05– 0.1 ms) is typically used for cranial nerve stimulation during surgery. Higher intensities may be needed if the nerve is less responsive due to damage, insulated by tissue or fluid, or at a distance from the stimulating electrodes. Intensities stronger than 5 mA can spread and lead to unintended activation of nerves. The anesthetic regimen has to be optimized before recording intraoperative EMG. After the induction of anesthesia, muscle relaxants (e.g., vecuronium or pancuronium) are ceased once intubation has been performed, and a train of four should be performed to confirm absence of physiologic muscle relaxant.

Somatosensory Evoked Potentials (SSEP)

Intraoperative monitoring of somatosensory evoked potentials (SSEPs) is one of the most commonly used modalities for predicting and preventing postoperative neurological deficits. SSEPs have been reported to detect the presence of cortical ischemia during cerebrovascular procedures, and their utility during skull base procedures is well recognized. In endoscopic skull base surgery, SSEPs are most commonly utilized when working very closely on the carotid artery. In the event of a carotid artery injury, SSEPs can notify the surgeon if cerebral ischemia is occurring during use of temporary clipping or if too much packing or compression has been performed. SSEPs monitor the integrity of the spinal cord dorsal columns, medial lemniscus pathways to the thalamus, and its connections to the primary sensory cortex by detecting a stimulus—administered to a peripheral nerve—at the somatosensory cortex.

After the induction of anesthesia, baseline SSEPs are recorded. It can be recorded prior to patient positioning or after positioning when lateral, three-quarter, or prone positioning is used. Recording baseline before positioning is preferable, as pressure on the brachial plexus or peripheral nervous system can be detected and corrected. For upper extremity SSEP recording, bilateral stimulation of the median or ulnar nerve is performed in an alternate fashion at the wrist with a pair of subdermal needle electrodes. For the lower extremities, bilateral alternate stimulation of the tibial nerve is performed. In case one cannot elicit a reliable tibial nerve response, the peroneal nerve can be stimulated. The stimulation of the tibial nerve is performed by a pair of subdermal needle electrodes placed at the medial malleolus of the ankle with a proximal cathode and distal anode separated by a gap of 1 cm. The stimulation of the peroneal nerve is carried out by a pair of subdermal needle electrodes placed at the head of the fibula and medially in the popliteal fossa.

The SSEPs resulting from the stimulation of ulnar or median nerves are recorded by P4/Fz and P3/Fz scalp electrodes (cortical) and a cervical electrode localized at the C7 spinous process (subcortical) and referenced to Fz. The SSEPs resulting from the stimulation of peroneal or tibial nerve are recorded by Pz/Fz and P4/P3 scalp electrodes, and a cervical electrode is localized at the C7 spinous process and referenced to Fz. Band-pass filters set at 30 to 300 Hz are used for cortical recordings, and band-pass filters set at 30 to 1000 Hz are used for subcortical (cervical) recordings. The alarm threshold is a sustained 50% decrease in primary somatosensory cortical amplitude or an increase in response latency by > 10% from baseline. Changes in amplitude or latency of SSEPs in > 2 averaged trials qualify as sustained changes. SSEPs have certain limitations including their inability to detect subcortical ischemia and lack of information about the integrity of motor pathways.