3 Neurophysiological Monitoring in Skull Base Surgery
Abstract
Skull base surgery allows for the treatment of a plethora of cranial base pathologies but carries with it the risk of damage to neurovascular structures found in the area. Due to the complex nature of many skull base pathologies, it is important to properly identify and monitor neural structures intraoperatively. Various intraoperative neuromonitoring methods can be utilized during skull base surgery to not only help prevent damage to neural structures but also to increase the efficacy of the surgical intervention. In this chapter, we review the various methods, including electromyography, brainstem auditory evoked potentials, and visual evoked potentials. We discuss methods that are used to monitor the different cranial nerves and review the efficacy of neuromonitoring.
Keywords: Keywords: skull base surgery, neuromonitoring, cranial nerves, electromyography, lateral spread response, BAEP, VEP
3.1 Introduction
Intraoperative monitoring allows surgeons the possibility of decreasing morbidity and improving outcomes in neurological surgery, especially with skull base pathologies. Intraoperative neuromonitoring allows surgeons to receive continuous feedback about surrounding cranial nerves (CNs). Pathologies such as tumors located in or around the cavernous sinus or superior orbital fissure, which can place CNs III, IV, and VI at risk of injury, or hemifacial spasm (HFS) involving CN VII can all be treated surgically but require intraoperative neuromonitoring to help identify the CNs and prevent new damage from occurring. Intraoperative neuromonitoring can also aid in the identification of certain pathologies and might serve a prognostic role. Understanding which CNs are at risk for injury and how to monitor them in various skull base cases are essential. In this chapter, we review the various forms of intraoperative neuromonitoring, relevant CN anatomy, intraoperative monitoring techniques, and evidence of whether the monitoring has been found to improve surgical outcomes. The covered intraoperative neuromonitoring methods for skull base surgery include electromyography (EMG), brainstem auditory evoked potentials, and visual evoked potentials (VEPs). The monitoring covered will be in regard to common skull base extra-axial pathologies.
3.2 Electromyography (EMG)
3.2.1 Background
Electromyography works by measuring the electrical activity of muscles, which allows it to monitor the integrity of any CN with motor function including III, IV, V, VI, VII, X, XI, and XII. Free-running EMG (f-EMG), or ongoing spontaneous motor activity (SMA), measures motor unit potentials (MUPs). Conversely, triggered-EMG (t-EMG), which produces compound muscle action potentials (CMAPs), consists of electrically stimulating a CN and recording the CMAPs when its corresponding motor neurons are activated.46 Free-running EMG allows the surgical team to receive continuous real-time feedback on the functional integrity of the CNs and whether certain actions may have adverse effects on the nerves, with a high specificity for postoperative deficits.10
EMG tracings can consist of spikes and bursts, A, B, and C trains or neurotonic discharges.40,45 Spikes have a single peak with an amplitude of less than or equal to 2,000 µV and are bi- or tri-phasic potentials, while bursts are a discrete grouping of numerous superimposed spikes, or MUPs, with amplitudes equal or less than 5,000 µV.45 Trains are patterns of electrical activity consisting of a chain of MUPs that may last for seconds and can be grouped into three patterns: A, B, and C. They can be triggered by mechanical manipulation and irrigation.20 The A train pattern has a sinusoidal waveform that can last from milliseconds to seconds with amplitudes averaging 100 to 200 µV, while the B train pattern consists of individual components that can take on a regular or irregular sequence lasting minutes to hours and containing spikes and/or bursts, and the C train pattern consists of an irregular sequence with amplitudes that can range from 20 to greater than 500 µV and is usually emitted by pharyngeal muscles innervated by CN IX and CN X.45 In a study of patients undergoing surgery for cerebellopontine angle (CPA) tumors, spikes, bursts, B trains, and C trains were not found to correlate with postoperative facial nerve paresis; however, intraoperative presence of the A train pattern was associated with paresis as measured by House-Brackmann grade.45
Triggered EMGs allow surgeons to identify the position and functionality of CNs intraoperatively so that the surgeon can avoid and detect iatrogenic damage quickly.8,28 The currents used to trigger the EMGs are low, starting around 0.05 mA and rarely rising above 5 mA to maintain the specificity of the responses. Anesthesia may interfere with EMG monitoring, and it is helpful to verify whether or not there is pharmacologic muscle relaxation before monitoring with a train of four. CMAPs can be stimulated using bipolar or monopolar probes, with bipolar probes delivering a more focused delivery of the current due to the placement of both the cathode and anode on the nerve, while monopolar stimulation is less likely to have current shunting.28,63 The amplitude of CMAP responses is viewed as a reflection of the functional status of the nerve, how many motor units are activated by the trigger, and how far away they are from the electrode.
3.2.2 Cranial Nerves III, IV, and VI
The CNs involved in extraocular movements are at risk of injury during skull base surgeries involving the parasellar, cavernous sinus, superior orbital fissure, or petroclival regions. These CNs include the oculomotor nerve (CN III) which innervates the levator palpebrae superioris; the inferior rectus, inferior oblique, medial rectus, and superior rectus muscles; the trochlear nerve (CN IV) which innervates the superior oblique muscle; and the abducens nerve (CN VI) which innervates the lateral rectus muscle.28 Damage to these nerves during surgery can lead to ptosis and/or diplopia.47,48 In order to prevent iatrogenic damage, it is helpful to monitor them intraoperatively. Studies have found that f-EMG intraoperative monitoring allows surgeons to properly locate the anatomical position of these extraocular CNs.14,47 One study evaluating the role of CMAP amplitudes of triggered EMG for CN III and CN VI found no association in predicting postoperative paresis; however, the sample size was small. More studies are needed to evaluate the usefulness of triggered EMG in predicting postoperative functional outcomes of extraocular CNs.28
The oculomotor nucleus is found in the rostral midbrain and courses in the interpeduncular fossa exiting between the posterior cerebral artery (PCA) and superior cerebellar artery (SCA) to the roof of the cavernous sinus. It ultimately travels through the superior orbital fissure and enters the orbit just past the annulus of Zinn after which it splits into superior and inferior branches.17 To monitor CN III, two to three pairs of needle electrodes are placed subdermally, through the eyelid, targeting the superior, medial, and inferior rectus muscles separately.28 The globe is covered with lubricant and a corneal eye shield. During placement of electrodes, the globe is depressed with the contralateral hand, while the other hand places the electrodes through the eyelid into the orbit, aiming away from the globe itself. The electrode pair targeting the superior rectus must not be too close to the superior oblique or alternatively can be entirely forgone to avoid cross contamination in monitoring of electrical activity between the superior rectus (CN III) and superior oblique (CN IV) muscles. CN IV, which also originates in the caudal midbrain, decussates and exits the brainstem posteriorly and passes through the ambient and crural cisterns, and enters the free edge of the tentorium just before it enters the posterior wall of the cavernous sinus. The nerve then runs along the lateral wall of the cavernous sinus initially under CN III, then crossing superiorly and medially in the superior orbital fissure to reach the superior oblique muscle.17 To monitor CN IV, electrodes can be placed into the superior oblique muscle in the superomedial portion of the orbit. CN VI originates at the caudal pons and exits at the pontomedullary fissure, and, just as with the other extraocular CNs, it passes through the superior orbital fissure into the orbit.17 For CN VI, needle electrodes can be bent 90 degrees and inserted into the lateral rectus muscle along the lateral orbital rim.28 Most surgeons place electrodes free-hand, taking care to avoid the globe,28 while some advocate for image-guided placement.3 All electrodes should be secured to the patient’s skin, usually with tape, to minimize movement of the electrodes throughout the case (Fig. 3.1).
Fig. 3.1 Extraocular nerve electromyography (EMG). Representative image of placement of stimulation electrodes to monitor cranial nerves (CNs) III, IV, and VI. For CN III, electrodes target the superior, medial, and inferior rectus muscles, separately. For CN IV, electrodes (yellow/black) target the superior oblique muscle in the superomedial portion of the orbit. For CN VI, the electrodes target the lateral rectus muscle along the lateral orbital rim.
3.2.3 Cranial Nerve V
The trigeminal nerve (CN V) has three branches: ophthalmic (V1), maxillary (V2), and mandibular (V3), and, of the three branches, only V3 has motor components.61 The motor root of CN V is in the rostral pons, and courses through the prepontine cistern just superomedial and parallel to the sensory portion of CN V, which run together into Meckel’s cave, at which point the motor root separates with the sensory portion V3 to become the mandibular nerve.5,7,23,31 The masseter, temporalis, medial, and lateral pterygoid, the mylohyoid, and the anterior belly of the digastric muscle, tensor tympani, and tensor veli palatini are all innervated by the motor component of CN V.5,31 For intraoperative monitoring of CN V, the masseter and/or temporalis muscles are typically monitored.
3.2.4 Cranial Nerve VII
In skull base pathologies involving the prepontine or CPA regions, monitoring of the facial nerve (CN VII) can be helpful.34 The facial nerve originates on the ventral side of the brainstem at the pontomedullary junction just inferomedially to the flocculus and runs through the internal auditory canal (IAC) and temporal bone, and then exits through the stylomastoid foramen. Once extratemporal, the facial nerve splits into an upper trunk and lower trunk. The upper trunk further divides into the frontal branch (innervation for the occipitofrontalis, orbicularis oculi, the corrugator supercilii, and the anterior and the superior auricular muscles), the zygomatic branch (innervation for the orbital, infraorbital, and zygomatic muscles), and the buccal branch (innervation for buccinator, levator labii, anguli oris, and orbicularis oris).9,15,26 The lower trunk further divides into the marginal mandibular branch (innervation for the depressor labii inferioris and depressor anguli oris) and the cervical branch (innervation for the platysma).2,26 It is important to understand the location and innervation patterns of the branches of the facial nerve for intraoperative monitoring.
Neuromonitoring can have a unique benefit in microvascular decompression (MVD) for HFS. HFS is thought to be due to hyperexcitability of CN VII and related to vascular compression of the nerve.12 MVD is an accepted surgical treatment for HFS. Intraoperative neuromonitoring of the facial nerve allows for observation of the lateral spread response (LSR). The LSR occurs when one branch of CN VII is stimulated and activates the muscles associated with another branch of the facial nerve, which is characterized as an abnormal motor response.33,36 Most commonly, one stimulation electrode is placed near the zygoma to stimulate the zygomatic branch and another is placed near the angle of the mandible to stimulate the mandibular branch. LSR occurs when a response is detected, via bipolar subdermal needle electrodes, in mentalis or orbicularis oris when the zygomatic branch is stimulated, or when a response is detected in frontalis or orbicularis oculi when the mandibular branch is stimulated (Fig. 3.2).24,49,57
It has been shown that the LSR can disappear intraoperatively when the dura mater is opened, when the arachnoid is opened, or when the compressing blood vessel is moved off the nerve, with the LSR returning when the vessel is moved back onto the facial nerve (Fig. 3.3).12,33 Due to these observations, LSR can be used as a surrogate for decompression of the facial nerve, and intraoperative monitoring can help predict outcomes after MVD for patients with HFS, particularly given that HFS may take several weeks to resolve after surgical decompression. It is unknown whether resolution of LSR intraoperatively correlates with better long-term resolution of symptoms. Some studies have found no difference in spasm relief at follow-up between those with or without residual LSR after decompression intraoperatively,57,58 while others found a higher symptom relief success rate in patients whose LSR disappeared with the vascular decompression.13,50
Fig. 3.2 Facial nerve electromyography (EMG) for lateral spread response (LSR). Representative image of placement of stimulation electrodes at the zygoma (red/white) and near the angle of the mandible (red/black) with subdermal electrodes measuring the response from the frontalis, orbicularis oculi, orbicularis oris, and the mentalis.
Fig. 3.3 Facial nerve electromyography (EMG) demonstrating the lateral spread response (LSR): (a) When the facial nerve is compressed, the LSR is noted. The left panel EMG tracing demonstrates a response in the mentalis and orbicularis oris when the zygomatic branch is stimulated, while the right panel demonstrates a response in the frontalis and orbicularis oculi when the mandibular branch is stimulated. (b) When the facial nerve is decompressed, the LSR disappears. The red tracing is the baseline signal obtained at the beginning of the case. The green tracings represent recently acquired averages, and the black tracing is the currently acquired average.
Tumors of the CPA including meningioma, schwannoma, and epidermoid cyst resections also benefit from intraoperative neuromonitoring. With vestibular schwannoma resections, facial nerve preservation rates in large tumor resections were found to be greater in cases with facial nerve monitoring, with one older study reporting an increase in anatomic facial nerve preservation from 41 to 71% when intraoperative monitoring was used.1,18,19 Studies have found that, intraoperatively, the A train pattern in EMG monitoring correlates with damage to the facial nerve, and increased A train activity corresponds to a higher risk of postoperative facial nerve palsy.41,42,43,44,45 The House-Brackmann grade can also be predicted by the minimum current intensity required to stimulate the facial nerve at the brainstem after tumor resection. Some studies have found that low stimulus thresholds along with high response amplitudes are predictive of a low grade House-Brackmann facial nerve palsy, with one study reporting that a stimulus threshold of 0.05 mA or less along with a response amplitude of at least 240 µV were predictors of facial nerve palsy consistent with a House-Brackmann grade I or II, 1 year postoperatively.29,35 Alternatively, other studies have found a correlation between higher stimulus thresholds and worse immediate facial nerve palsy.62
3.2.5 Cranial Nerves IX, X, and XII
The anatomical and functional integrity of the lower CNs (IX–XII) is often endangered when solid tumors, like meningiomas, located in the petroclival region, CPA, brainstem, or in the proximity of foramen magnum are surgically treated.46 Depending on the nerve damaged, the patient can suffer from postoperative neurological deficits like dysarthria, dysgeusia, dysphagia, and dysphonia. Intraoperative neuromonitoring of the lower CNs helps prevent such adverse outcomes. This section shall focus on the anatomy and neuromonitoring technique of three CNs: IX, X, and XII. The spinal accessory (CN IX) is discussed separately.
Both the glossopharyngeal and vagus nerves originate from the medulla oblongata, carry sensory, parasympathetic, and motor fibers, and exit the cranium via the jugular foramen.21 The glossopharyngeal nerve (CN IX) originates from the upper part of the preolivary sulcus of the medulla oblongata as multiple rootlets that merge relatively quickly to form a single nerve.39 The glossopharyngeal nerve provides sensation to the posterior third of the tongue and the entire palate.
The vagus nerve (CN X) has a long and spread out course that is reflected in its name (Latin vagary: to wander). It is the longest CN. Like CN IX, it originates from the medulla oblongata as multiple rootlets but becomes a single nerve more distal in the jugular foramen. The motor fibers supply all of the striated muscles of the pharynx, larynx, and the soft palate except the tensor veli palatini and stylopharyngeus muscles.52 The hypoglossal nerve (CN XII) originates from the preolivary sulcus of the medulla oblongata as multiple rootlets.4,22 These rootlets merge into two to four trunks that course forward as the hypoglossal nerve and then exit the cranium via the hypoglossal canal.22 The hypoglossal nerve is a motor nerve that supplies all of the intrinsic and extrinsic muscles of the tongue, except for the palatoglossus muscle, which is supplied by the vagus nerve.
The CN IX is generally not monitored as it is frequently not clinically relevant to separately monitor CN IX. Damage to CN IX is not debilitating as demonstrated by historical cases in which the nerve has been sectioned rather than decompressed in glossopharyngeal neuralgia without adverse neurological consequences.
Intraoperative neuromonitoring of the vagus nerve is performed by recording EMG from the true vocal cords.60 An endotracheal tube that has built-in steel wire electrodes is placed under direct vision by the anesthesiologist so that the electrodes are in close contact with the true vocal cords (Fig. 3.4 and Fig. 3.5).30
