The facial nerve passes through the internal auditory meatus of the temporal bone and traverses the mastoid through the fallopian canal, passing near several surgical landmarks in the middle ear, including the facial ridge, hypotympanum, and oval window. Although normally protected by bone, the nerve can be exposed due to atypical anatomy, erosion of the bone by disease, or removal of bone from prior surgery. Facial nerve irritation is detected by increased electrical activity of muscles innervated by the facial nerve (passive monitoring), and the facial nerve is located by detecting facial muscle contractions from direct electrical stimulation of the nerve from various points in the operative field (active monitoring). This electrical stimulation uses 4-per-second, 0.2-msec duration pulses, whose amplitude usually ranges between 0.1 to 0.6 mA. As with any nerve conduction study, a compound muscle action potential (CMAP) is produced (with about a 3-msec latency), and it can be readily detected with an oscilloscope triggered by the stimulus pulse (Fig. 7.1).

Aside from the stapedial nerve, the motor portion of the facial nerve has no other branches before it emerges at the stylomastoid foramen. Consequently, stimulation of the nerve trunk within the mastoid or middle ear may excite any or all of the muscles of the head and neck that are innervated by the facial nerve. Stimulation during surgery, intentional or inadvertent, can be expected to activate muscles over a large portion of the face. Differential recording electrodes placed at the outer canthus of the eye and near the nares and upper lip are sensitive to myogenic activity from all muscle groups in the central portion of the face. The myoelectric signals are relatively strong in this region, providing good sensitivity and signal-to-noise ratio of the recordings. The differential ground is typically placed on the cheek. Surface and needle electrodes perform equally well for this type of broadly spaced, nonspecific array. The monitoring device presently being used by us for mastoid and middle ear surgery, the Xomed Neural Integrity Monitor (NIM), can record two channels simultaneously, typically with one pair of electrodes near the eye and the other near the upper lip. However, this somewhat increased specificity for site of activity provides no greater sensitivity for active or passive
monitoring; furthermore, two channels have the disadvantage of dividing the monitor’s attention and thereby increasing the likelihood of missing a fleeting event. For these reasons, we routinely monitor with a single channel.

Surface electrodes are easily applied with the patient just prior to anesthesia. The recording instrument can be then be used to evaluate the amplitude of the EMG (electromyography), which is particularly helpful for patients with a preexisting facial paresis. If necessary, electrodes can be placed satisfactorily even after the start of surgery. Instruments such as the NIM display the ongoing EMG on a small screen and can provide alarms when the waveform exceeds specified criteria. However, electrical artifacts generated by instruments, irrigation, drilling, and manipulation of the patient are mixed with the EMG and are often large enough to set off the alarms, causing unnecessary confusion and irritation. Consequently, the alarms are generally silenced when (i) there is no risk to the nerve or (ii) a monitor operator is present who can observe the waveform to discriminate between artifacts and EMG, so that the surgeon is alerted only when there is evidence of facial nerve stimulation.

Unfortunately, the nerve cannot be monitored reliably under all conditions. Poorly grounded equipment such as electric drills, electrocautery, and anesthesia instruments may increase background electrical noise to such an extent that the facial EMG is obscured, particularly if the EMG is abnormally small. Electrocautery generates signals so large that electrical monitoring is not possible during cautery, even though cautery can damage the nerve. Under these conditions, partial or complete loss of facial function can occur without being detected. At these times, facial muscle contractions are detected by the anesthesiologist placing a hand on the patient’s face or by the use of mechanical tranducers to detect facial movement. It is important to note that persistent heating of the nerve by drilling without irrigation and rest will make the nerve less likely to respond if damaged.

The condition of the patient also affects the quality of the monitoring. Preexisting nerve dysfunction may not be readily discerned during a clinical evaluation, but it can reduce EMG amplitudes. Although it may be possible to monitor patients with paresis as low as House grade IV, the procedure will be much more susceptible to artifacts; consequently, reliability will be greatly diminished. These conditions can be anticipated by preoperative electroneurography. Care must be exercised in the use of muscle relaxants, because their very action is to reduce the responsiveness of the muscles being monitored. When excessive, they may prevent monitoring altogether.

During passive monitoring, the role of the operator is to detect any signs of facial nerve irritation. When the nerve is encountered, a sudden burst response, bipolar in morphology, is seen. The underlying timing of nerve firing is slow enough so that when amplified and presented through a loudspeaker a characteristic low-frequency sound can be heard. Such sounds can occur from touching or transecting the facial nerve. The differences in response between a brief touch and complete division of the facial nerve are slight. A continuous “growling” sound has been associated with irritation of the nerve (e.g., stretching of the nerve or sudden change in temperature).

Active monitoring with a stimulating probe may be used by the surgeon whenever there is a need to locate the nerve and differentiate it from surrounding tissue. The monitor assists by adjusting current levels of the stimulator, discriminating CMAPs from artifacts, and reporting on the relative sizes of CMAPs. By varying the probe location and the current, the amplitude of the CMAP can often reveal the precise location of the facial nerve. In the case of a nerve enveloped in cholesteatoma or other soft tissue, for example, the monitor operator will reduce the stimulus current to a very low level (0.1 to 0.4 mA) so that the probe will only stimulate if it is directly on the nerve. In this way the surgeon can systematically probe each piece of tissue before excising it. Because bone is a poor electrical conductor, the nerve within the intact bony canal cannot be reliably stimulated or localized with this method. Very high current levels are nonlocalizing because they can produce a maximum CMAP over a wide area. However, as the bone overlying the nerve is thinned, the stimulator becomes more effective, so that current level can sometimes be used to assess the relative thickness of the remaining bone.

Cerebellopontine angle tumor resection is done by our surgeons, employing either the posterior or middle fossa approaches. In the past the posterior fossa approach had been used almost exclusively (11); more recently the middle fossa operation is being used for small intracanalicular tumors when hearing preservation is paramount. However, there may be a small increased risk of facial nerve injury with the middle fossa approach (12).