When the identity of a cranial nerve is unclear to a surgeon, we are frequently called upon to identify nerves through the use of intraoperative neurophysiology. If the nerve in question is a motor cranial nerve, or mixed sensory-motor cranial nerve, then we can identify it with the use of electrical nerve stimulation and concurrent recording from muscles innervated by that nerve. We call this test stimulus-triggered electromyography (T-EMG) and the responses are called compound muscle action potential (CMAPs).
- Stimulus Parameters
- Recording Parameters
- Recording Locations
- Important Notes
- Differentiating Oculomotor, Trochlear and Abducens Nerves
- Differentiating Trigeminal and Facial Nerves
- Differentiating the Lower Cranial Nerves (IX-XII)
- Bonus Round
Identification and differentiation of motor cranial nerves can be rather straight-forward, or it can just as well be complex and challenging. Intimate knowledge of cranial nerve anatomy, including anatomical variations, is essential to making the correct decision. I also believe that direct visualization of the operative field (integrating microscope feed onto your computer screen) and close/clear communication with the surgeon are extremely important.
Here I will review the motor and mixed sensorimotor cranial nerves and the techniques that you can use to help identify and differentiate them in surgery. Keep in mind that not all surgical procedures can be reviewed and variations in anatomy, physiology, conduction velocity, latency, amplitude and required stimulation parameters will inevitably exist.
It is strongly recommended that you do not perform electrical stimulation of the nervous system unless properly trained and credentialed. These techniques should only be performed under the supervision of a neurophysiologist with a professional board certification (e.g., D.ABNM, or a neurologist fellowship-trained in IONM).
Back to the Top
Stimulus parameters will depend on the type of probe you are using (i.e., monopolar v. bipolar) and the type of tissue you are stimulating (e.g., bone, tumor, muscle or direct nerve). Importantly, stimulation parameters will also depend on the goal of stimulation. The following parameters assume direct stimulation of cranial nerves for the purpose of differentiating one nerve from another:
- Intensity: ≤ 5.0 mA. Start at 0.0 mA and increase intensity by increments of ~0.1 mA until you record CMAPs from the appropriate muscles at the expected latency.
- Pulse: cathodal, square wave.
- Pulse Width: commonly ≤ 200 μsec, but I recommend ≤ 100 μsec.
- Charge: 1.0 μC/cm^2/ph (assuming 5 mA and 200 μsec).
- Frequency: 2-5 Hz.
Note: there are a number of different methods for evaluating the function of a nerve in the context of making a prognosis. These methods differ from the above and are beyond the scope of this particular post.
- High Pass Filter: 1-10 Hz, I use 10 Hz.
- Low Pass Filter 1500-5000 Hz, I use 3000 Hz.
- Time Base: 20-50 msec (2-5 msec/division), I use 2-3 msec/div.
In the table below, I have listed some common muscle recording sites for monitoring and mapping cranial nerves and their branches during surgery. The list of muscles is not all-inclusive, and some muscles may not make sense for all procedures and/or patients. This table is provided only as a guide.
Many people make the mistake of marking the latency of a CMAP at it’s onset. You can get into some trouble doing this, particularly when differentiating the lower cranial nerves. Make sure you mark the latency of a CMAP at the peak of it’s first deflection, just like other evoked potentials. When you measure the amplitude, you measure peak-to-peak.
The reliability of T-EMG is largely dependent upon the integrity of the neural elements being tested. Thus, a patient presenting with pre-operative neurological deficits confounds the monitoring. Often, a nerve that is compromised will have an altered threshold, decreased amplitude, increased latency, poor morphology, or no response at all (Singh et al., 2016).
In order to optimize interpretation of T-EMG, the patient must be sufficiently free of pharmacological blockade of the neuromuscular junction. Absence of neuromuscular blockade, or sufficient clearance/reversal for reliable monitoring, should be documented by the neurophysiologist using “train-of-four” (TOF) monitoring, which records muscle twitches in response to stimulation of a peripheral nerve. For cranial nerve monitoring, I recommend a TOF ratio (T4:T1) ≥ 0.70. Keep in mind that partial blockade affects different muscle groups to varying degrees (Gavrancic et al., 2014; Saitoh et al., 1996), and can be especially variable in a patient with pre-existing neurological dysfunction (Sloan, 2013; Sloan & Heyer, 2002).
Back to the Top
Differentiating Oculomotor (III), Trochlear (IV) and Abducens (VI) Nerves:
It is commonly necessary to differentiate CNs III, IV and VI in the middle fossa and cavernous sinus. Recording from extraocular muscles requires knowledge of ocular anatomy and associated vasculature. Careful electrode placement will ensure the integrity of the sclera. Several recording methods have been reviewed by López (2011). In a recently-published paper, my colleagues and I described our methodology for placing extra-ocular recording electrodes (Singh et al., 2016).
We use a commercially-available subdermal needle electrode which is pre-bent to approximately 90°. With the eye in the closed position, the neurophysiologist inserts the needle into the extraocular muscle of interest and carefully advances the electrode along the bony ridge of the orbit until it is fully inserted. A small piece of TransporeTM tape is used to secure each electrode as the others are placed. One electrode is placed for each of the extraocular muscle monitored, and each recording is referenced to a needle inserted into the frontalis muscle of the forehead. TegadermTM film is applied to protect the eyes during surgery.
Once the electrodes are in place, differentiating the nerves is rather straight-forward. For the most part, the CMAP latencies following stimulation of each nerve will be identical. So, you will be relying solely on the muscle from which the CMAP is recorded. Thus, when a CMAP is recorded from only the inferior rectus, the nerve in question in the oculomotor nerve (CN III). When a CMAP is recorded from only the superior oblique, the nerve in question in the trochlear nerve (CN IV). When a CMAP is recorded from only the lateral rectus, the nerve in question in the oculomotor nerve (CN VI).
What about the situation in which you stimulate a single nerve, and get CMAPs from multiple extra-ocular muscles, such as the inferior rectus and the superior oblique? Assuming you’ve ruled out technical factors, like current spread and current shunting, the most likely interpretation is that you are stimulating the oculomotor nerve (CN III). This is because your superior oblique electrode, which you are using to monitor CN IV, is so close to the superior rectus muscle (innervated by CN III) that you are likely to record CMAPs from this muscle, which will obviously coincide with CMAPs recorded from the inferior rectus muscle.
Back to the Top
Differentiating Trigeminal (V) and Facial (VII) Nerves:
It is typically only necessary to differential the trigeminal and facial nerves when working at the brainstem, because they have very different anatomical courses. The trigeminal nerve goes anterior through Mekel’s Cave and the mandibular division (V3) exits the skull through the foramen ovale. The facial nerve goes lateral with the vestibulocochlear nerve (VIII), enters the internal auditory meatus, winds through the middle/inner ear and exits the skull through the stylomastoid foramen.
Following electrical stimulation of cranial nerves V or VII, CMAPs are usually recorded simultaneously from muscles innervated by both nerves. Because all of these muscles are in close proximity in the face, the CMAP from one muscle will be recorded by electrodes placed in nearby muscles. This is called volume conduction (some people call it cross-talk). So, you can’t always use the location of the CMAP to identify the nerve.
Because responses come from both muscle groups, he best way to differentiate CNs V and VII is through the latency of the CMAP. When working at the brainstem, stimulation of the trigeminal nerve (CN V) will normally produce CMAPs with a latency of 4-5 msec, while stimulation of the facial nerve (CN VII) will normally produce CMAPs with a latency of 6-7 msec.
Here are a couple of examples from a retrosigmoid craniectomy for resection of acoustic schwannoma.
In addition to the short, ~5 msec CMAP latency, notice that the amplitude of the response recorded from the masseter muscle is much larger as compared to the rest of the muscles, all of which are innervated by CN VII.
In addition to the long, ~7 msec CMAP latency, notice that the amplitude of the responses recorded from the facial muscles are much larger as compared to the masseter muscle, which is innervated by CN V.
Keep in mind that the latency of a CMAP is a function of the nerve’s 1) health, 2) conduction velocity, and 3) length. In some situations, it may be possible to record CMAPs of similar latency/amplitude, particularly if in the context of per-operative motor deficits secondary to chronic nerve compression/injury. So, you may not always be able to rely on CMAP latency and amplitude.
If you want to make your life a lot easier, I have a tip for you: Rather than use subdermal needle electrode for recording from the mimetic (facial) and mastication (chewing) muscles, use intramuscular needle electrodes! This will purify your recordings and virtually eliminate volume conduction.
Here is an example of CN VII stimulation during a translabyrinthine craniectomy for resection of acoustic schwannoma.
It has become my preference to use intramuscular needle electrodes for any surgery in which I need to differentiate between the trigeminal and facial nerves. Give it a shot and let me know what you think. You can get these electrodes here.
Back to the Top
Differentiating the Lower Cranial Nerves (IX, X, XI and XII):
The locations where you are most likely to have to differentiate the lower cranial nerves are at the brainstem and in the jugular foramen.
When the glossopharyngeal nerve is monitored, pre-bent subdermal needle electrodes are placed in the soft palate with the aid of a curved hemostat. When the vagus nerve is monitored, a commercially-available endotracheal tube (ETT) with surface-mounted electrode contact can be used. Alternatively, some people place subdermal needle electrodes in the cricothyroid muscle. When the accessory nerve is monitored, subdermal needle electrodes are placed in the upper trapezius muscle. When the hypoglossal nerve is monitored, pre-bent subdermal needle electrodes are placed in the lateral tongue with the aid of a curved hemostat.
Differentiation of the lower cranial nerves can pose some pretty significant challenges:
The first major challenge is that the lower cranial nerves all have natural anatomical variations in the form of inter-connections (anastamoses) that makes these nerves more of a plexus (Shoja et al., 2014a,b). For this reason, even low-level stimulation of one nerve can easily coactivate a different nerve and this can result in recording CMAPs from muscles that represent seeming disparate cranial nerves.
Here’s an example in which Vagus Nerve (X) stimulation also evoked CMAPs from the tongue:
The second major challenge has to do with recording locations. The only way to monitor the glossopharyngeal nerve (CN IX) is to record from the stylopharyngeus muscle, and the only way to record from this muscle is to place electrodes in the lateral soft palate. The problem is that the electrodes aren’t actually in the stylopharyngeus muscle; rather, they are close to the muscle. The electrodes are technically in the soft palate and these muscles are innervated by the vagus nerve (CN X). So, differentiating between IX and X can be challenging.
In some cases, differentiating the lower cranial nerves is straight-forward – you stimulate a nerve, and you get a response from one muscle. In my experience, this is usually true for the accessory nerve and the hypoglossal nerve. Likewise, one can stimulate CN IX and get a CMAP only from the soft palate. Here’s an example:
It’s not always that easy, though. Sometimes you stimulate CN IX and get responses from the soft palate and the vocal folds. The best physiological explanation for this is anastamoses between the nerves. It may be best to stimulate multiple nerves and see if you can find a clear differentiation. You and the surgeon may have to put your heads together to determine what makes the most sense given the location of the surgery and the patient’s known individual anatomy.
One thing that you have to keep in mind when differentiating IX and X is the anatomy of the vagus nerve. If you are recording from the vocal folds with an endotrachial tube, then you are essentially recording from the recurrent laryngeal nerve (RLN), which has a long, winding pathway to the vocal muscles. On the right side, the RLN branches off of the vagus nerve, loops under the subclavian artery and ascends back toward the vocal cords. On the left side, the RLN has a longer course after branching from the vagus, looping under the aortic arch before ascending toward the vocal cords.
This is important because the excessive length of the vagus nerve translates to a much longer latency, at least as compared to the glossopharyngeal nerve. If you stimulate CN IX at the brainstem, the CMAP should have a latency of approximately 7 msec. If you stimulate CN X at the brainstem, the CMAP should have a latency of approximately 9-11 msec on the right side, and 11-13 msec on the left side. This is where it’s absolutely critical to measure the response latency by the peak of the CMAP, and not the onset.
I think that the difficulties associated with differentiating CNs IX and X and under-reported in the literature, and this particular topic deserves more attention. Recently, the utility of glossopharyngeal nerve monitoring has been challenged by Kartush (2015). I highly recommend reading his commentary if you can.
The challenges associated with differentiating the lower cranial nerves with the use of T-EMG are not insignificant. As with the other cranial nerves, a keen understanding of the anatomy, physiology, pathology and surgical approach, as well as having a view of the surgical field, will help you to make the best decisions. When in doubt, be sure to communicate your uncertainty. Communication is critical to identification and differentiation of cranial nerve. In some cases, you and the surgeon may find yourselves making a “best guess”, and that’s ok as long as everyone is on the same page.
Back to the Top
In the image below, CMAPs are recorded from different muscles following electrical stimulation of three different cranial nerves. Within the figure, I have labeled these CMAPs by grouping (Group 1, Group 2, Group 3). As you read below, refer to these Groups within the figure.
These recordings are made following electrical stimulation of one of the right side lower cranial nerve near the brainstem. As you can see, CMAPs are recorded from the soft palate, vocalis and cricothyroid muscles. The cricothyroid response appears small in amplitude, but the screen sensitivity is set to 2000 μV/div. From the pattern of muscles activated, we can deduce that the nerve in question is the vagus nerve.
It is also important to consider the CMAP latency and what that means in the context of the vagus nerve anatomy. As the right vagus nerve descends in the neck (inside the carotid sheath) it first gives off the external branch of the superior laryngeal nerve, and this innervates the cricothyroid muscle. The latency of this response is ~7.5 msec. The vagus nerve then gives off the recurrent laryngeal nerve, which passes under the subclavian artery and then ascends in the tracheoesophageal groove before entering the larynx and innervating the vocal cords (among many other muscles). The latency of this response is ~10.0 msec. So, the latencies of these CMAPs match the anatomical course of the nerve.
Group 1 recordings are clearly made in response to electrical stimulation of the vagus nerve, and this goes to show that you can use cricothyroid recordings to monitor CN X just as well as an endotrachial tube with surface electrodes. In this case, cricothyroid recordings are made using subdermal needle electrodes placed in the neck.
Similar to Group 1, these recordings in Group 2 are also made following electrical stimulation of one of the right side lower cranial nerve near the brainstem. CMAPs are recorded most prominently from the vocalis and trapezius muscles. The latency of the CMAP recorded from the trapezius muscle and the prominence of the response is clearly consistent with stimulation of the accessory nerve (CN XI).
What’s interesting about this grouping is that the latency of the vocalis CMAP is actually consistent with activation of vagus nerve; however, there’s no CMAP recorded from the cricothyroid muscle. Also, notice that the vocalis CMAP is inverted, small in amplitude and more protracted (wider base) as compared to the recordings from Group 1. The reversal in phase leads me to believe that the generator site has changed relative to the recording location. Also, the smaller amplitude and protracted base both lead me to believe that the CMAP is far-field. I don’t know the generator of this CMAP, but I suspect that perhaps this recording is a far-field response secondary to activation of the sternocleomastoid muscle, which is innervated by CN XI.
These recordings are made following electrical stimulation of one of the right side cranial nerve near the brainstem. CMAPs are recorded most prominently from the masseter, orbicularis oculi, orbicularis oris, mentalis and cricothyroid muscles – all using subdermal needle electrodes.
The CMAP latency is too long to be activation of CN V, so we are obviously stimulating CN VII. What about the cricothyroid response? The latency of this response is much longer than the one recorded in Group 1 when we were stimulating the vagus nerve. That’s because the generator for this particular CMAP is from a CN VII-innervated muscle – the platysma, which is innervated by the cervical branch of the facial nerve. The latency is accounted for by the long, winding route of the facial nerve.
Back to the Top
Ok, that’s all, folks! I hope you enjoyed this post. Feel free to leave your comments below.
- Gavrancic B, Lolis A, Beric A. Train-of-four test in intraoperative neurophysiologic monitoring: differences between hand and foot train-of-four. J. Clin. Neurophysiol. 2014;31(6):575–9.
- Kartush JM. Commentary: Endotracheal Tube Electrodes to Assess Vocal Cord Motor Function During Surgery in the Cerebellopontine Angle. Neurosurgery. 2015;77(5):E838–41.
- López JR. Neurophysiologic Intraoperative Monitoring of the Oculomotor, Trochlear, and Abducens Nerves. J. Clin. Neurophysiol. 2011;28(6):543–550.
- Saitoh Y, Tanaka H, Toyooka H, Amaha K. Recovery of post-tetanic and train-of-four responses at the first dorsal interosseous and adductor pollicis muscles in patients receiving vecuronium. Can. J. Anaesth. = J. Can. d’anesthésie. 1996;43(4):362–7.
- Shoja MM, Oyesiku NM, Griessenauer CJ, Radcliff V, Loukas M, Chern JJ, Benninger B, Rozzelle CJ, Shokouhi G, Tubbs RS. Anastomoses between lower cranial and upper cervical nerves: a comprehensive review with potential significance during skull base and neck operations, part I: trigeminal, facial, and vestibulocochlear nerves. Clin Anat. 2014a Jan;27(1):118-30.
- Shoja MM, Oyesiku NM, Shokouhi G, Griessenauer CJ, Chern JJ, Rizk EB, Loukas M, Miller JH, Tubbs RS. A comprehensive review with potential significance during skull base and neck operations, Part II: glossopharyngeal, vagus, accessory, and hypoglossal nerves and cervical spinal nerves 1-4. Clin Anat. 2014b Jan;27(1):131-44.
- Singh H, Vogel RW, Lober RM, Doan AT, Matsumoto CI, Kenning TJ, & Evans JJ. (2016). Intraoperative neurophysiological monitoring for endoscopic endonasal approaches to the skull base: A technical guide. Scientifica Neuroscience. 2016;2016:1-20.
- Sloan TB. Muscle relaxant use during intraoperative neurophysiologic monitoring. J. Clin. Monit. Comput. 2013;27(1):35–46.
- Sloan TB, Heyer EJ. Anesthesia for intraoperative neurophysiologic monitoring of the spinal cord. J. Clin. Neurophysiol. 2002;19(5):430–43.