The facial nerve is directly and indirectly involved in numerous pathological conditions affecting the temporal bone, ranging from infection to neoplasia. In each instance, a solid understanding of its complex anatomy is crucial to the physician's ability to both diagnose and treat disorders of the facial nerve.
The facial nerve (Figure 69–1) begins its development near the end of the first month of gestation, when the acousticofacial primordium, giving rise to both the facial and acoustic nerves, develops adjacent to the primordial inner ear, the otic placode. The geniculate ganglion, which arises from the second branchial arch, develops early in the second month of the gestation. Adjacent to the developing geniculate ganglion, the acousticofacial primordium differentiates into a caudal and a rostral trunk. The caudal trunk progresses into the mesenchyme of the second branchial arch, becoming the main trunk of the facial nerve. The rostral branch becomes associated with the first arch, eventually developing into the chorda tympani nerve, providing taste to the anterior two-thirds of the tongue. This development partially explains the close association of the chorda tympani with the facial nerve.
A schematic illustration demonstrating the embryology of the facial nerve. (A) The location of the primitive facial nerve in the developing embryo is shown in relation to other important nerves in the head and neck. (B) The location of the second branchial arch giving rise to the main trunk of the facial nerve is shown in relation to the other branchial arches. (C) Other derivatives of the second branchial arch are shown and help explain the complex innervation pattern of the facial nerve. (Reproduced with permission from Langman J & Sadler TW, Langman's Medical Embryology, 5th ed. Williams and Wilkins: Baltimore, MD; 1985.)
Both the geniculate ganglion and the nervus intermedius, arising from the second branchial arch, form independently of the motor division of the seventh nerve. During the sixth week of gestation, the motor division of the facial nerve establishes its position in the middle ear between the membranous labyrinth (an otic placode structure) and the developing stapes (a second arch structure). The nerve then passes into the mesenchyme of the second arch. During this time, the chorda tympani nerve becomes associated with the trigeminal nerve, which will carry the chorda tympani on its way to the tongue via the lingual nerve. The greater superficial petrosal nerve, which carries preganglionic parasympathetic fibers toward the pterygopalatine ganglion, also develops during this time period.
Most of the anatomic relationships of the facial nerve are established by the end of the second gestational month. Although the fallopian canal, the bony canal that transmits the facial nerve through the temporal bone, begins its development in the fifth gestational month, it is not complete until several years after birth. The incomplete development of this canal is though to be responsible for the natural dehiscences that may contribute to facial palsies that are associated with childhood otitis media.
During the sixth gestational week, the extratemporal portion of the facial nerve begins development. By the end of the second gestational month, all five divisions of the extratemporal nerve—the temporal, zygomatic, buccal, mandibular, and cervical branches—are present. Over the third month, the nerve becomes enveloped by the parotid gland. The facial muscles (Figure 69–2), developing independently, areÂ formed at 7–8 week gestation and must be innervated by the distal facial nerve branches or else will degenerate. By the end of the third gestational month, a majority of the facial musculature is identifiable and functional.
An adaptation of Sir Charles Bell's classic illustration of the muscles of facial expression, with the muscles labeled. (Reproduced with permission from Bell C, Essays on the Anatomy of Facial Expression, 2nd ed. Murray; 1824.)
At birth, the facial nerve is located just beneath the skin near the mastoid tip, as it emerges from the temporal bone. This nerve is thus placed at risk when a postauricular incision is made in a young child, as is often done for ear surgery. As the mastoid tip forms and elongates during childhood, the facial nerve assumes its more medial position. Individual axons of the facial nerve also undergo myelination until the age of 4Â years, an important consideration during electrical testing of the nerve during this time period.
Central Neuronal Pathways
The primary somatomotor cortex of the facial nerve is located in the precentral gyrus, corresponding to Brodmann areas 4, 6, and 8. It is from this region that the complex voluntary motor functions of the facial nerve, such as facial expression, are controlled (Figure 69–3). Neural projections from this area combine into fascicles of the corticobulbar tract during their descending course through the internal capsule. These neural projections continue through the pyramidal tracts within the basal pons. In the caudal portion of the pons, most of the facial nerve fibers cross the midbrain to reach the contralateral facial nucleus. A small number of facial nerve fibers innervate the ipsilateral facial nucleus, a majority of which are destined for the temporal branch of the nerve. This distinction becomes important when the clinician is trying to determine if a facial paralysis is due to a central or peripheral lesion: central lesions spare the forehead muscle since they receive input from both cerebral cortices, whereas peripheral lesions will involve all branches of the facial nerve.
A schematic illustration of the complete pathway of the motor division of the facial nerve. (Reproduced with permission from Wilson-Pauwels L, Akesson EJ, Stewart PA, Cranial Nerves: Anatomy and Clinical Comments. B.C. Decker, 1988.)
In addition to these voluntary neural projections to the facial nerve, there is also an extrapyramidal cortical input to the facial nucleus from the hypothalamus, the globus pallidus, and the frontal lobe, all of which control involuntary facial expression associated with emotion. Additional projections to the facial nuclei from the visual system are involved in the blink reflex. Projections from the trigeminal nerve and nuclei contribute to the corneal reflex, whereas those from the auditory nuclei help the eye close involuntarily in response to loud noises.
Facial Nucleus & Brainstem
The efferent projections from the facial motor nucleus emerge dorsomedially to form a compact bundle that loops over the caudal end of the abducens nucleus beneath the facial colliculus or internal genu (or turn). The neurons then pass between the facial nerve nucleus and the trigeminal spinal nucleus, emerging from the brainstem at the pontomedullary junction (Figure 69–4).
The anatomy of the facial nerve (CN VII), cochlear nerve, and vestibular nerve (CN VII), as they exit the brainstem at the level of the pontomedullary junction. (Reproduced, with permission from Wilson-Pauwels L, Akesson EJ, Stewart PA, Cranial Nerves: Anatomy and Clinical Comments. B.C. Decker Inc.: Toronto; 1988.)
In addition to supplying motor innervation to the muscles of facial expression, other neuronal projections found in association with the facial nerve are partially responsible for taste, cutaneous sensation of the external ear, proprioception, lacrimation, and salivation (Table 69–1; Figure 69–5). This bundle of these nerves, termed the nervus intermedius, or Neve of Wrisberg, exits the brainstem adjacent to the motor branch of the facial nerve. The general visceral efferent fibers of the nervus intermedius are preganglionic parasympathetic neurons that innervate the lacrimal, submandibular, sublingual, and minor salivary glands. The cell bodies of these nerves arise in the superior salivatory nucleus and join the facial nerve after it has passed the abducens nucleus. They travel together until reaching the geniculate ganglion in the temporal bone. At this point, the greater superficial petrosal nerve branches offs, containing the neurons destined for the pterygopalatine ganglion. The greater superficial petrosal nerve ultimately innervates the lacrimal, minor salivary, and mucosal glands of the palate and nose. The remaining fibers form part of the chorda tympani nerve, proceed to the submandibular ganglion, and eventually proceed to the submandibular and sublingual salivary glands.
Table 69–1. Subdivisions and Functions of the Facial Nerve. ||Download (.pdf)
Table 69–1. Subdivisions and Functions of the Facial Nerve.
|Facial Nerve Subdivision||Function|
- Muscles of facial expression
- Posterior belly of digastric muscle
- Stylohyoid muscle
- Stapedius muscle
- Salivation—lacrimal, sub-mandibular, and sublingual
- Nasal mucosa or mucous membrane
- Sensory to auricular concha
- External auditory canal
- Tympanic membrane
|Special sensory||Chorda tympani nerve—taste to anterior two-thirds of the tongue|
The anatomy of the visceral motor portion of the facial nerve, making up the nervus intermedius, or Wrisberg nerve. The preganglionic, parasympathetic portions of this nerve have cell bodies located in the abducens nucleus. From there they travel toward the geniculate ganglion in the temporal bone, located at the first genu of the facial nerve on the floor of the middle cranial fossa. Fibers from this nerve are destined to innervate the lacrimal gland, the minor salivary glands, and the mucosal glands of the palate and nose. (Reproduced with permission from Wilson-Pauwels L, Akesson EJ, Stewart PA, Cranial Nerves: Anatomy and Clinical Comments. B.C. Decker Inc.: Toronto; 1988.)
The special visceral afferent fibers, which also form a portion of the chorda tympani nerve, receive input from the taste buds of the anterior two-thirds of the tongue, as well as the hard and soft palates (Figure 69–6). These sensory afferents for taste have their cell bodies in the geniculate ganglion and will eventually synapse in the medulla, in the nucleus solitarius.
The anatomy of the special sensory component of the facial nerve, comprising the chorda tympani nerve. (Reproduced with permission from Wilson-Pauwels L, Akesson EJ, Stewart PA: Cranial Nerves: Anatomy and Clinical Comments. B.C. Decker Inc.: Toronto; 1988.)
The general sensory afferent neurons of the nervus intermedius are responsible for cutaneous sensory information from the external ear canal and postauricular region. These cutaneous sensory fibers enter the spinal trigeminal tracts without synapsing in the geniculate ganglion.
The facial nerve leaves the brainstem at the pontomedullary junction (see Figure 69–4). At this location, it lies in close approximation to the eighth cranial nerve (the vestibulocochlear nerve). This intimate relationship takes on critical importance when disease, most commonly a vestibular schwannoma, arises in the region of the cerebellopontine angle. In this location, the facial nerve is placed in jeopardy both during the growth of the tumor and during attempted surgical resection in this area.
During its lateral course through the cerebellopontine angle and internal auditory canal (IAC), the relative positions of the facial and cochleovestibular nerves change by rotating 90Â°. In the cerebellopontine angle, the facial nerve is covered with pia, bathed in cerebrospinal fluid, and is devoid of epineurium. As a result, the nerve is very susceptible to trauma or manipulation in this region, such as during intracranial surgery.
Intratemporal Nerve Pathways
After traversing the cerebellopontine angle, the facial nerve enters the temporal bone along the posterior face of the petrous bone. Within the temporal bone, the facial nerve successively passes through 4 regions prior to its exit out of the stylomastoid foramen: (1) the IAC, (2) the labyrinthine segment, (3) the intratympanic segment, and (4) the descending segment (Figures 69–7, 69-8, and Figure 69–9). From the lateral end of the IAC to its exit out the stylomastoid foramen, the nerve travels approximately 3 cm within the facial canal, also known as the fallopian canal.
The intratemporal divisions of the facial nerve. After passing through the internal auditory meatus on the posterior face of the petrous temporal bone, the nerve enters its canalicular segment, within the IAC. It then becomes the labyrinthine segment, as it traverses between the cochlea and the vestibular labyrinth. After making its first genu (bend) at the geniculate ganglion, it becomes the tympanic segment, coursing through the middle ear space, just superior to the oval window. It then makes its second major genu at the level of the horizontal semicircular canal and becomes the vertical or descending segment. After passing through the stylomastoid foramen, it becomes extracranial.
A stylized representation of the anatomy of the lateral aspect of the internal auditory canal. The facial nerve lies at the most anterior and superior location at this level.
Histological cross sections of the facial nerve at 3 points along its course within the temporal bone. Proximally within the temporal bone at the level of the internal auditory canal, (A) the individual nerve fascicles are not defined and nerve elements appear homogenous. (B) As the nerve proceeds through the tympanic segment and (C) at the level of the stylomastoid foramen, individual nerve fascicles becomes increasingly defined.
After traversing the cerebellopontine angle, the facial nerve enters the temporal bone along the posterior face of the petrous bone, piercing the internal auditory meatus. At the lateral end of the IAC, a ridge of bone, the traverse crest, divides the IAC into superior and inferior portions. It is at this lateral portion of the IAC that the anatomy is most consistent: the superior portion is occupied by the facial nerve anteriorly and the superior vestibular nerve posteriorly (see Figure 69–8). These two nerves are additionally divided by a bony ridge, the vertical crest or “Bill's bar.” The inferior portion of the IAC, below the transverse crest, contains the cochlear nerve (anterior) and the inferior vestibular nerve (posterior). Within the IAC, the dural covering of the facial nerve is transformed to epineurium.
At the lateral portion of the IAC, the facial nerve pierces the meatal foramen to enter the labyrinthine segment. The labyrinthine segment is notable in that it is the narrowest portion of the fallopian canal, where it averages <0.7 mm in diameter and occupies the canal to the greatest proportional extent. As a result, it is believed that infections or inflammations of the facial nerve within this region can lead to temporary or permanent paralysis of the nerve, such as in Bell's palsy.
The distal end of the geniculate ganglion is considered the end of the labyrinthine segment of the nerve and lies just superior to the nerve. While it is generally bone-covered just below the floor of the middle cranial fossa, the geniculate ganglion is dehiscent into the middle fossa up to 15% of the time. Arising from the geniculate ganglion is the greater superficial petrosal nerve.
At the geniculate ganglion, the facial nerve makes its first genu and becomes the tympanic segment of the facial nerve, so called because it travels within the middle ear space. This portion of the nerve is approximately 10 mm long. As the nerve enters the tympanic space, it is positioned just superiorly, medially, and anteriorly to the cochleariform process, which serves as an excellent anatomical landmark during surgical identification of the nerve. An additional useful landmark, the “cog,” which is a small bony prominence projecting from the roof of the epitympanum, lies just superior to the facial nerve as it enters the tympanic cavity. The facial nerve then travels posteriorly along the medial portion of the epitympanum, passing superior to the oval window and stapes. The nerve then curves inferiorly at its second genu, just posterior to the oval window, pyramidal process, and stapedial tendon, and anterior to the horizontal semicircular canal. It is this portion of the nerve that is susceptible to injury during surgery since processes such as cholesteatoma will frequently erode the bone covering the facial nerve in this region, leaving it precariously exposed.
In addition to bony dehiscence from pathology, natural fallopian canal dehiscences have also been described in cadaver specimens, a majority of which occurred in the tympanic segment. In more than 80% of cases, the dehiscences involved the portions of the canal adjacent to the oval window.
Vertical, Descending, or Mastoid Segment
After the second genu, the nerve traverses the synonymously named vertical, descending, or mastoid segment en route to the stylomastoid foramen. As the facial nerve descends inferiorly in this portion, it gradually assumes a more lateral position. The facial nerve branch of the stapedius muscle arises in this segment, traveling a short distance to the stapedius muscle. More inferiorly, the chorda tympani nerve, which carries preganglionic parasympathetic fibers to the submaxillary and sublingual glands and taste fibers to the ipsilateral anterior two-thirds of the tongue, arises from the facial nerve and travels superiorly, laterally, and anteriorly back toward the middle ear space. The angle between the chorda tympani nerve and the descending portion of the facial nerve is approximately 30Â° and delineates a triangular space known as the facial recess. The facial recess is an important surgical route of entry into the middle ear space, enabling access of the stapes superstructure, the promontory, and the round window niche.
In its most inferior portion, the facial nerve takes on a close proximity to the digastric ridge and muscle, where the nerve is consistently medial and anterior to these structures. An additional close anatomic landmark in this region includes the sigmoid sinus, where it passes deep to the facial nerve in this region. Upon exiting the stylomastoid foramen, the nerve becomes encased in the thick fibrous tissue of the cranial base periosteum and digastric muscle.
Although the facial nerve most commonly descends in its vertical segment as a single nerve, bifurcations, trifurcations, and hypoplasia of the facial nerve have been found within the mastoid segment. Additionally, the chorda tympani nerve has been noted to arise from the facial nerve anywhere from the stylomastoid foramen to the geniculate ganglion.
Peripheral Facial Nerve Anatomy
The facial nerve exits the skull base through the stylomastoid foramen, between the mastoid tip laterally and the styloid process medially (Figure 69–10). At the stylomastoid foramen, the facial nerve passes into the parotid gland, typically as a single large trunk. The nerve then divides within the parotid gland into its temporofacial and cervicofacial branches. Rarely, this division can occur within the temporal bone and exit the stylomastoid foramen as separate branches. One branch is the posterior auricular nerve that courses lateral to the mastoid and is joined by a filament of the auricular branch of the vagus nerve.
A portion of an illustration from Sir Charles Bell, demonstrating the exit of the facial nerve from the stylomastoid foramen. (Reproduced with permission from Bell C, The Nervous System of the Human Body. Longman; 1830. Plate VII.)
Within the parotid gland, the nerve can assume numerous configurations, with frequent anastomoses between branches. However, generally five main branches of the nerve can be identified: (1) the temporal, (2) the zygomatic, (3) the buccal, (4) the mandibular, and (5) the cervical. The temporal branch innervates the frontalis muscle, which allows for the voluntary raising of eyebrows. The zygomatic branch innervates the orbicularis oculi muscle and is critical for proper eye closure. The buccal nerve innervates the buccinator and orbicularis oris, allowing for proper mouth closure and cheek muscle activity. The mandibular branch innerves the platysma. The posterior auricular nerve, arising just after the exit of the facial nerve from the stylomastoid foramen, sends branches to the occipitalis muscle posteriorly on the skull.
Courbille J. The nucleus of the facial nerve: the relation between cellular groups and peripheral branches of the nerve. Brain
(A classic study of the anatomy of the facial nerve nucleus.)
Fisch U, Esslen E. Total intratemporal exposure of the facial nerve: pathologic findings in Bell's palsy. Arch Otolaryngol
(The study shows edema of the facial nerve in Bell's palsy during intratemporal exposure, with the most severe constriction in the region of the geniculate ganglion.)
Gasser RF. The development of the facial nerve in man. Ann Otol Rhinol Laryngol
(A classic manuscript on the embryologic development of the facial nerve.)
Gasser RF. The early development of the parotid gland around the facial nerve and its branches in man. Anat Rec
(A classic manuscript on the development of the parotid gland in relation to the facial nerve.)
Gasser RF, May M. Embryonic development of the facial nerve. In: May M, ed. The Facial Nerve. 1st ed. Thieme, 1987:3.
Hall GM, Pulec JL, Rhoton AL Jr. Geniculate ganglion anatomy for the otologist. Arch Otolaryngol
(A classic study of the anatomy of the geniculate ganglion.)
Nager GT, Proctor B. Anatomic variations and anomalies involving the facial canal. Otolaryngol Clin North Am
(A classic, comprehensive study of the anatomy of the facial canal.)
Ozbek C, Tuna E, Ciftci O et al. Incidence of fallopian canal dehiscence at surgery for chronic otitis media. Eur Arch Otorhinolaryngol
. 2009;266(3):357. [Epub 2008 Jun 20.
] (An anatomic study describing naturally occurring dehiscences of the facial nerve.)
Proctor B, Nager GT The facial canal: normal anatomy, variations and anomalies. II. Anatomical variations and anomalies involving the facial canal. Ann Otol Rhinol Laryngol Suppl
(A classic study of the anatomy of the facial canal.)
Rhoton AL Jr, Kobayashi S, Hollinshead WH. Nervus intermedius. J Neurosurg
(A classic study of the anatomy of the nervus intermedius.)
Vidic B. The anatomy and development of the facial nerve. Ear Nose Throat J
(A review of facial nerve anatomy and embryology.)
Vidic B, Wozniak W. The communicating branch of the facial nerve to the lesser petrosal nerve in human fetuses and newborns. Arch Anat Histol Embryol
(A classic anatomic study of the communicating branch of the facial nerve.)
The facial nerve trunk consists of approximately 10,000 nerve fibers, approximately 7000 of which are myelinated motor fibers. The facial nerve sheath consists of several layers. The endoneurium, closely adherent to the layer of Schwann cells of the axons, surrounds each nerve fiber. The perineurium, which is the intermediate layer surrounding groups of fascicles, provides tensile strength to the nerve and is believed to represent the primary barrier to the spread of infection. The outermost layer of the nerve is the epineurium. This outer layer contains the vasa nervorum, which provides the blood supply to the nerve.
Classification of Facial Nerve Degeneration
If the facial nerve is injured, various degrees of injury may result. Several models allow for a clinical determination of the degree of nerve fiber injury that produces an irreversible conduction block (ie, fiber degeneration). It was originally proposed that peripheral nerve injury involves varying degrees of neuropraxia (blockade), axonotmesis (division of individual fibers), and neurotmesis (division of fascicles and epineurium). A clinical–pathological classification of nerve injury, the Sunderland Classification scheme (Figure 69–11), is widely accepted and grades the extent of injury as follows:
First-degree injuries are characterized by the blockage of axoplasm flow within the axon. There is sufficient pressure to restrict its replenishment when metabolic needs dictate. This blockade is sometimes referred to as neuropraxia. Although an action potential cannot be propagated across the lesion site, a stimulus applied distal to the lesion will conduct normally to produce an evoked response.
Second-degree injuries entail axonal and myelin disruption distal to the injury site as a result of the progression of a first-degree injury. Such injuries eliminate the propagation of an externally applied stimulus as wallerian degeneration of the axon ensues.
Third-degree injuries involve complete disruption of the axon including its surrounding myelin and endoneurium.
Fourth-degree injuries entail the complete disruption of the perineurium.
Fifth-degree injuries entail the disruption of the epineurium.
Sixth-degree injuries, a proposed addition to the Sunderland classification by later authors, take into account the observed patterns of blunt and penetrating injuries of the nerve. These injuries are characterized by normal function through some fascicles and varying degrees of injury (first-degree through fifth-degree injuries), differentially involving fascicles across the nerve trunk.
A model of graded neural injury that details clinical–pathological classifications. Microanatomic changes in cranial nerve injury are demonstrated in cross section. The potential for appropriate axonal regeneration across the site of injury is dictated principally by the status of connective tissue elements.
Central to Sunderland's classification is the notion that axonal recovery depends on the integrity of the connective tissue elements of the nerve trunk. This model predicts a high likelihood for the complete recovery of peripheral innervation when endoneurial tubules remain intact to support reinnervation, as is the case with first- and second-degree injuries. In contrast, disruption of the endoneurium—a third-degree injury or worse in this model—increases the likelihood of irreversible axonal injury and aberrant patterns of regeneration.
An example of abnormal neural regrowth is “crocodile tears,” or increased lacrimation associated with eating. It occurs when efferent fibers normally targeted to travel with the chorda tympani nerve to the submandibular and sublingual glands are misdirected through the greater superficial petrosal nerve to the lacrimal gland. This results in parasympathetic innervation of the lacrimal gland as well as the normal target, the salivary glands. As a result, when eating, instead of getting the normal salivary response to increase the salivation, the neuronal signal causes tearing of the lacrimal gland.
Kim J MD PhD, Moon IS MD, Shim DB MD, Lee WS MD PhD. The effect of surgical timing on functional outcomes of traumatic facial nerve paralysis. J Trauma
Sunderland S. Nerve and Nerve Injuries. 2nd ed. E & S Livingstone Ltd; 1968.
The impaired transmission of neural impulses can result from physiological blockage (in the absence of nerve fiber degeneration) and axonal discontinuity with wallerian degeneration. Because the clinical presentation of a facial paralysis does not distinguish between simple conduction block and axonal disruption, investigators have explored an array of testing procedures designed to define the extent of nerve injury (Table 69–2).
Table 69–2. Tests of Facial Nerve Function. ||Download (.pdf)
Table 69–2. Tests of Facial Nerve Function.
|Minimal excitability test||The lowest stimulus intensity that consistently excites all branches on the uninvolved side|
- Patient comfort
- Easy to perform
- Relies on visual detection
|Maximal excitability test||Compares response on involved vs. uninvolved side of face|
- Patient comfort
- Easy to perform
- Relies on visual detection
|Electroneuronography and evoked electromyography (EMG)|
- Assesses the facial motor response to a supramaximal stimulus
- Records the compound muscle action potential
- Reflects % motor fibers of the facial nerve that have under-gone degeneration
- <90% denervation prognosticates excellent recovery
- Repeated every other day to detect ongoing degeneration beyond the 90% critical level
- Useful early in the course of facial paralysis
- Some measures useful in predicting the ultimate level of spontaneous recovery
- Measures postsynaptic membrane potentials
- Motor unit potentials in five muscle groups in the first 3 days after onset of palsy associated with good outcome in >90% of patients
|Precision characterization of motor units|
- Possible pitfalls with early testing
- Sparse residual motor units that suggest a favorable outcome may be evident despite severe injury to large portions of fibers that are at risk for degeneration
- F wave represents activity in facial muscles generated by antidromically activated motor neurons
- In Bell palsy, F wave seen only after recovery begun
|Can provide direct and immediate assessment of facial nerve function|
- Limited dynamic range and prognostic value
- Primarily animal testing, research
|Magnetic simulation||Electromagnetic coil to produce neural activation||Intensity of the stimulus is minimally attenuated by intervening tissue|
- Limited clinical utility
- Difficult to interpret results
- EMG recording of the blink reflex
- Compares responses between the affected and normal sides
- Abolished R1 reflex associated with little chance of recovery in the first 2 months after onset of paralysis
|Easy to perform||May be limited by small response amplitude|
In an initial evaluation of patients with acute facial paralysis, the clinician should aim to determine the prognosis for recovery as well as the cause of the paralysis. Early determination of the prognosis for recovery may permit intervention both to minimize nerve injury and to optimize regeneration.
Topognostic test batteries are intended to determine the level of facial nerve injury by inference from which branch(es) are functional. If tearing is diminished, the lesion is assumed to be proximal to the point at which the greater superficial petrosal nerve branches from the geniculate ganglion. Abnormal stapedial muscle function, as revealed by immittance testing, presumably reflects nerve impairment above the stapedial motor branch from the facial nerve trunk distal to the posterior genu. The functioning of the chorda tympani nerve can be determined by submandibular gland secretion and taste testing. Dysgeusia and diminished salivary gland flow presumably reflect nerve impairment above the branch point of the chorda tympani nerve from the vertical segment of the facial nerve in the mastoid.
Early observations suggested that more proximal levels of dysfunction correlated with a higher risk of degeneration and incomplete recovery. However, topognostic modalities have often provided inconsistent information on the level of neural injury, and are subject to the vagaries produced by “skip” lesions of the nerve that affect the motor, sensory, and autonomic portions of the nerve differently For example, the Schirmer tear test, an apparent index of proximal nerve function establishing a lesion at or above the level of the geniculate ganglion, has been subsequently shown to have an accuracy rate of only 60% using intraoperative electrical stimulation to specify the site of nerve conduction block in Bell's palsy. However, the Schirmer test has great practical value in assessing tear production and the need for adjunctive measures for eye care.
Because topognostic testing carries inherent vagaries, electrophysiological testing (below) has emerged as the diagnostic approach of choice in assessing nerve conductivity and the risk of irreversible degeneration of nerve fibers.
The interpretation and validity of electrophysiological testing of an acute facial palsy rests on two constructs with regard to nerve fiber function:
Segmentally demyelinated fibers maintain the capacity to propagate a stimulus, albeit at a higher threshold, than that of normal fibers. Anatomically intact fibers will therefore continue to propagate an applied stimulus, whereas those that have become disrupted and subsequently degenerated will not.
By estimating the proportion of degenerated motor fibers, a clinician may distinguish palsies that will fail to recover spontaneously and will produce long-term sequelae.
Electrophysiological testing ideally provides an index of the severity of injury to the total nerve trunk by reflecting the proportion of motor fibers that have progressed beyond a first-degree injury. Correlation of the ultimate level of recovery with early electrophysiological findings determines the prognostic value of the test in identifying the subset of facial palsy patients who will not obtain satisfactory, spontaneous recovery.
Clinically available electrophysiological tests indirectly assess the severity of injury to the intratemporal facial nerve. Given its course within the Fallopian canal, electrical stimulation proximal to the site of conduction blockade is possible only when the nerve is activated intracranially. For this reason, the ability of a nerve to propagate an impulse is assessed distal to the stylomastoid foramen. Even in the presence of severe neural injury, conduction distal to a lesion will continue until its axoplasm is consumed and wallerian degeneration ensues. This process requires 48–72 hours to progress from intratemporal to extratemporal segments, thereby rendering electrical stimulation tests falsely normal during this period. Routine electrophysiological tests therefore fail to detect nerve conduction as it occurs, thereby delaying the differentiation of neuropraxia from degeneration.
Nerve Excitability Testing
Minimal excitability testing with the Hilger nerve stimulator has provided a readily accessible method of facial nerve assessment. The test is indexed according to the thresholds for visually detectable activity generated by surface stimulation of a facial nerve branch. The test reflects elevated thresholds for neuromuscular stimulation produced by axonal disruption and degeneration. The lowest stimulus intensity that consistently excites all branches on the uninvolved side establishes the normal threshold. A 2.0–3.5 mA difference between the uninvolved and involved sides is reported to suggest impending denervation.
This test offers technical advantages in the portability of the necessary equipment and the use of minimal stimulation, which is more comfortable for the patient than maximal stimulation tests. The test, however, introduces subjectivity in that it relies on the visual detection of a response of a limited number of facial muscles. In addition, current threshold levels for peripheral branches are likely to selectively activate large nerve fibers with lower thresholds and those fibers closer to the stimulating electrode, thereby excluding an unknown proportion of motor fibers from the assessment.
Fisch U. Prognostic value of electrical tests in acute facial paralysis. Am J Otol
Gantz BJ, Gmur A, Fisch U. Intraoperative evoked electromyography in Bell's palsy. Am J Otolaryngol
(The technique of intraoperative evoked electromyography is described in detail. The limited extent of the blocked motor fibers suggests that segmental, rather than total, intratemporal decompression is needed in Bell's palsy.)
Lewis BI, Adour KK, Kahn JM et al. Hilger facial nerve stimulator: A 25-year update. Laryngoscope
. 1991;101(1 Pt 1):71–74.
A test of maximal electrical stimulation can be used to determine whether nerve degeneration has developed in the course of an acute facial paralysis. It involves a transcutaneous electrical impulse designed to saturate the nerve with current, activating all functioning fibers. The response on the involved side is characterized as being (1) equal to the contralateral side, (2) minimally diminished (50% of normal), (3) markedly diminished (<25% of normal), or (4) absent.
When the response is markedly diminished or absent within the first 2 week of the clinical paralysis, it has been found that there is a 75% chance of incomplete facial nerve recovery. When the response completely disappeared within the first 10 days, recovery was typically incomplete and significant sequelae ensued. Conversely, if responses were symmetric during the first 10 days of a clinical paralysis, complete return was found in more than 90% of patients tested. The use of supramaximal stimulation provides sensitivity and consistency in testing when used early in the course of an acute facial paralysis. However, the interpretation of the maximal stimulation test relies on a subjective evaluation of the visually graded evoked response.
May M. Nerve excitability test in facial palsy: limitations in its use based on a study of 130 case. Laryngoscope
(This study describes the utility and drawbacks of the facial nerve excitability test based upon the author's personal experience.)
May M, Blumenthal F, Klein S. Acute Bell's palsy: Prognostic value of evoked electromyography, max stimulation and other electrical tests. Am J Otol
(Evoked electromyography and maximal stimulation tests were the most accurate electrical tests for predicting the course of acute facial paralysis when they were performed serially within the first 10 d after onset.)
Ushio M, Kondo K, Takeuchi N et al. Prediction of the prognosis of Bell's palsy using multivariate analyses. Otol Neurotol
Evoked Electromyography & Electroneuronography
Similar to the maximal stimulation test, evoked electromyography (EEMG) or electroneuronography (ENoG) assesses the facial motor response to a supramaximal stimulus. In contrast to maximal stimulation testing, the EEMG technique records the compound muscle action potential (CMAP) with surface electrodes placed in the nasolabial fold. The CMAP can be graphically displayed for quantitative analysis and printed for the medical record (Figure 69–12). Waveform responses are analyzed to compare peak-to-peak amplitudes between normal and involved sides.
Placement of recording and stimulating electrodes for facial EEMG recordings. The compound muscle action potential is reflected in the biphasic electromyographic response.
Patients with incomplete paralyses due to Bell's palsy invariably recover function to normal or near-normal levels and do not require EEMG evaluation. The reappearance of facial movement within 3–4 week after onset also predicts an excellent prognosis for functional recovery. EMG sampling of motor activity to detect visually imperceptible facial function is advised.
When assessed within a critical time window, reductions in the amplitude of the EEMG response of the affected side are considered to reflect the percentage of motor fibers of the facial nerve that have undergone degeneration. Facial EEMG is most reliable during the initial phase of accelerated denervation when reliable results can be obtained (ie, in the first 2–3 week following the onset of a paralysis due to Bell's palsy or herpes zoster oticus). When neuropraxic fibers become “deblocked” either in the recovery phase or later on, as axons regenerate peripherally, stimulated nerve fibers discharge asynchronously. Because regenerated fibers do not discharge in synchrony, the response is disorganized and consequently diminished. This phenomenon imposes a time constraint on the reliability of EEMG testing that must be considered in interpreting the test results.
ENoG is most useful early in the course of facial paralysis. More than 50% of patients with complete paralysis who exhibit a ≥90% reduction in CMAP amplitude have less than a satisfactory, spontaneous return of facial function. When results demonstrate <90% denervation (>10% in CMAP amplitude relative to the normal side), excellent recovery has been uniformly observed.
It is recommend that EEMG testing should be repeated on an every-other-day basis to detect ongoing degeneration beyond the 90% critical level. The time span of reduced electrical excitability (ie, the velocity of denervation as demonstrated by repeated testing) and the degree of degradation of the CMAP response (ie, the nadir of the response) are most useful in predicting the ultimate level of spontaneous recovery. The earlier the EEMG response drops to ⩽10% of normal, the worse the prognosis is.
Chung WH, Lee JC, Cho DY et al. Waveform reliability with different recording electrode placement in facial electroneuronography. J Laryngol Otol
Coker NJ. Facial electroneuronography: analysis of techniques and correlation with degenerating motoneurons. Laryngoscope
(A comprehensive review of electroneuronography.)
Linder TE, Abdelkafy W, Cavero-Vanek S. The management of peripheral facial nerve palsy: “paresis” versus “paralysis” and sources of ambiguity in study designs. Otol Neurotol
Ushio M, Kondo K, Takeuchi N et al. Prediction of the prognosis of Bell's palsy using multivariate analyses. Otol Neurotol
The electromyographic (EMG) response reflects postsynaptic membrane potentials that may be either initiated at the neuromuscular junction with voluntary activation or generated spontaneously across the muscle membrane.
Voluntarily and spontaneously generated facial motor responses can help to characterize the condition of motor units with precision. However, the results obtained with testing in any single field should be buttressed with testing in adjacent fields. Motor unit potentials in four of five muscle groups in the first 3 days after the onset of an acute facial paralysis is associated with a satisfactory outcome in more than 90% of patients. Motor units in two of three muscle groups predicted a satisfactory outcome in 87% of patients. When motor units were either limited to one muscle group or abolished, satisfactory recovery was found in only 11% of cases.
Although these findings suggest a role for early EMG testing in prognosticating functional recovery, others have noted potential pitfalls of early EMG testing that may mislead the examiner. Sparse residual motor units that suggest a favorable outcome may be evident despite severe injury to large portions of fibers that are at risk for degeneration. The clinical evidence of this was noted as an unsatisfactory recovery despite voluntary motor potentials in 38% of Bell's palsy patients. These observations suggest that EMG assessment should be performed within at least two muscle groups to more accurately assess the degree of denervation.
Early in the course of an acute facial paralysis, preserved facial motor activity may escape clinical inspection and yet provide prognostic information when combined with other testing modalities. For example, subclinical motor activity that is still detectable by the EMG may complement the use of EEMG in the early phase of a clinical paralysis. EMG monitoring is of limited use in detecting early degeneration since electrical evidence of nerve degeneration is absent in the first 10 days of the paralysis. Ten to 14 days following the onset of a clinical paralysis, EMG recordings reflect the dynamic resting membrane potentials of postsynaptic elements. In thisÂ phase, muscle membrane, deprived of “trophic” substances that are normally transported through the axon, undergoes changes that destabilize the resting potential. These changes produce spontaneous depolarizations reflected in the EMG as fibrillation potentials. Such changes are interpreted as indicative of persistent denervation.
Substantial axonal loss and impaired reinnervation yield fibrillation potentials as long as postsynaptic membranes remain electrically active. With persistent denervation, EMG recordings are silent and the short burst of discharges normally found on needle insertion is absent. Conversely, successful reinnervation generates high-frequency polyphasic potentials that increase in amplitude and duration and replace fibrillation potentials. In rare cases of protracted paralysis due to Bell's palsy, longitudinal EMG evaluations detect persistent nerve degeneration or reinnervation.
Granger C. Prognosis in Bell's palsy. Arch Phys Med Rehab
(Using clinical and electromyographic methods, it should be possible to forecast recovery within 3 d after onset in order to preselect patients in need of any proposed curative treatment program designed to salvage the facial nerve.)
Grosheva M, Guntinas-Lichius O. Significance of electromyography to predict and evaluate facial function outcome after acute peripheral facial palsy. Eur Arch Otorhinolaryngol
. 2007;264(12):1491. [Epub 2007 Jul 5.
May M, Blumenthal F, Klein S. Acute Bell's palsy: prognostic value of evoked electromyography, max stimulation and other electrical tests. Am J Otol
(Evoked electromyography and maximal stimulation tests were the most accurate electrical tests for predicting the course of acute facial paralysis when they were performed serially within the first 10 days after onset.)
Sillman JS, Niparko JK, Lee SS et al. Prognostic value of evoked and standard electromyography in acute facial paralysis. Otolaryngol Head Neck Surg
(The findings from this study support previous reports of the prognostic value of EEMG in idiopathic facial paralysis, but suggest that this test may have less predictive value in the evaluation of facial paralysis as a result of trauma.)
Facial Nerve Assessment with Central Activation
The previously described electrodiagnostic tests indirectly assess the severity of injury to the intratemporal segment of the facial nerve. Investigators have explored alternate testing procedures in which the facial nerve is activated central to the presumed site of involvement within the temporal bone.
Testing via antidromic (retrograde) conduction provides an alternative to electrodiagnostic testing of peripheral fibers that, at least theoretically, can provide a direct and immediate assessment of facial nerve function. Antidromic conduction of electrical activity in the facial nerve can be measured with near- and far-field techniques in animals (Figure 69–13), and clinically with middle ear recording electrodes. It has been demonstrated that the far-field response to antidromic stimulation represented composite activity along the facial pathway and did not appear to reflect stimulation of the facial nerve at a specific site along the intracranial segment.
Topographical representation of mean amplitude and latency of evoked neural potentials from (A) the facial genu, (B) the dorsal region to the facial nucleus, and (C) the nucleus in experimental rodent preparation. Stimulus intensity = 0.4 mA; duration = 100 Î¼s; N = 100 stimuli for each trial. (Reproduced with permission from Niparko JK, Kartush JM, Bledsoe SC, Graham MD, Antidromically evoked facial nerve response. Am J Otolaryngol. 1985;6:353.)
The F-wave represents activity in facial muscles generated by antidromically activated motor neurons and contains no reflex components. For electrodiagnostic purposes, F-waves evoked by electrical stimulation may be recorded with intramuscular needle electrodes. This response has a long latency and is normally small in amplitude, thereby limiting its dynamic range and prognostic value. In patients with Bell's palsy, electrical stimulation of the nerve reliably produces F-wave responses only after the recovery has begun.
Transcranial magnetic stimulation employs an electromagnetic coil to produce neural activation. This method of neural activation is unique in that the intensity of the stimulus is minimally attenuated by intervening tissue. This feature enables central activation via a transcranial application of induced current. Animal studies have demonstrated that transcranial magnetic stimulation can be used to activate the facial nerve centrally, although the precise site of stimulation is difficult to determine. Observations suggest that the evoked response is likely due to the excitation of the facial nerve intratemporally or intracranially rather than via cortical or brainstem excitation.
Clinical experience with electromagnetic stimulation in pathological states, including Bell's palsy, is in keeping with observations that localize the lesion intratemporally. In 11Â patients with a recent onset of Bell's palsy, none demonstrated evoked CMAPs with magnetic stimulation. The lack of response is attributed to the elevation in threshold associated with segmental demyelination and the inability of the current generated by the electromagnetic field to reach the threshold.
Further refinement in the application and interpretation of transcranial magnetic stimulation for prognosticating facial nerve lesions awaits further understanding of the actual site of activation. The development of coils that will facilitate a more focused current offers the possibility of site-specific stimulation of the central facial motor tract and intracranial segment of the facial nerve—sites proximal to the typical sites of nerve injury for most acute facial paralyses.
The blink reflex can be tested clinically to assess the efferent arc contributed by cranial nerve VII. EMG recording of the trigeminofacial reflex provides a quantitative assessment of facial nerve conduction via activation of the facial nucleus centrally. This technique records action potentials reflexively generated in the orbicularis oculi muscle in response to an electrical stimulus applied to the supraorbital area (V1 branch). Responses between the affected and normal sides are compared to provide quantitative assessment of the reflex, thereby providing a measure of the functional integrity of the facial nerve. Trigeminofacial reflex testing of acute facial paralysis may be limited by small response amplitudes.
An abolished R1 trigeminofacial reflex response is associated with little chance of recovery in the first 2 months following the onset of paralysis. Preserved, early R1 responses predicted return to normal facial nerve function within the first month. The performance of this test in selecting those patients with an absent R1 response who have a poor long-term prognosis is yet to be evaluated.
Cocito D, Isoardo G, Migliaretti G et al. Intracranial stimulation of the facial nerve: normative values with magnetic coil in 240 nerves. Neurol Sci
Nakatani H, Iwai M, Takeda T et al. Waveform changes in antidromic facial nerve responses in patients with Bell's palsy. Ann Otol Rhinol Laryngol
Kartush JM, Bouchard KB, Graham MD et al. Magnetic stimulation of the facial nerve. Am J Otol
(Normal volunteers and one patient with acute facial paralysis were studied with both magnetic and electric stimulation of the facial nerve.)
Kartush JM, Garcia P, Telian SA. The source of far-field antidromic facial nerve potentials. Am J Otolaryngol
(This study attempts to identify the generator sites of the far-field antidromic facial nerve response in dogs and suggests the following for the montage and stimulus protocol employed: (1) the far-field antidromic response is a volume-conducted nerve action potential generated primarily from the mastoid segment, (2) complete transection of the facial nerve at the CPA has little effect on the responses, and (3) there appears to be no significant supranuclear generator site.)
Niparko JK, Kartush JM, Bledsoe SC et al. Antidromically evoked facial nerve response. Am J Otolaryngol
(Antidromic conduction testing was tested. Results suggest that the recorded potentials measured represent antidromic activation of the facial nerve, further suggesting that antidromic testing may provide a useful means of assessing proximal facial nerve function in pathological states.)
Sawney BB, Kayan A. A study of the F wave from the facial muscles. Electromyog
Schriefer TN, Mills KR, Murray NMF. Evaluation of proximal facial nerve conduction by transcranial magnetic stimulation. J Neurol Neurosurg Psych
(A magnetic stimulator was used for direct transcutaneous stimulation of the intracranial portion of the facial nerve in patients with a variety of facial nerve pathologies.)
von Dincklage F, Koppe P, Kotsch J et al. Investigation of threshold and magnitude criteria of the nociceptive blink reflex. Clin Neurophysiol
. 2010. [Epub ahead of print.
Zealar D, Kurago Z. Facial nerve recording from the eardrum. Otolaryngol Head Neck Surg
(A noninvasive technique is described for recording from the facial nerve within the fallopian canal using electrodes placed on the eardrum.)