Facial Nerve Repair

This article describes facial nerve repair for facial paralysis. Paralysis of the facial nerve is a cause of significant functional and aesthetic compromise.[1] Functional concerns primarily involve adequate protection of the eye, with a real risk of exposure keratitis if not properly addressed. In addition, swallowing, drooling, and speech difficulties may arise.

The degree of suffering these patients feel, however, is often far greater than these functional problems alone would produce. Patients with facial paralysis, especially younger ones, may experience tremendous psychosocial distress about their condition. Poor self-image and difficulty interacting with peers and family members can be devastating.

Repair of the facial nerve is generally a concern in cases of permanent complete facial paralysis. This may involve all or selected branches of the facial nerve.

Facial paralysis has many causes, which may be conveniently divided into the following 5 categories:

• Congenital

• Idiopathic

• Traumatic

• Neoplastic

• Inflammatory

Congenital facial paralysis, as in the well-described although poorly understood Möbius syndrome, is uncommon.

Idiopathic facial paralysis (Bell palsy) is the most common type. It is often thought to be due to virally induced inflammation of the nerve that results in functional compromise, swelling, and vascular compromise. Facial nerve repair is infrequently required, because most of these patients regain function spontaneously. When paralysis is permanent, some advocate facial nerve decompression in selected cases. This article focuses on cases requiring various nerve grafting techniques, rather than procedures of decompression.

Traumatic facial paralysis (from blunt and penetrating trauma or intraoperative iatrogenic injury) is the next most common type. The site of facial nerve injury may be intracranial, intratemporal, or external to the stylomastoid foramen. Acoustic neuroma surgery is an example of a procedure that puts the facial nerve at risk (in the cerebellopontine angle, in this case). Any injury sites may be amenable to facial nerve repair, except for sites near the nerve’s root entry zone, where the available facial nerve stump may not be long enough to allow repair.

Neoplastic causes of facial paralysis include tumors of the parotid gland, typically malignant. Facial nerve schwannomas, acoustic neuromas, and neoplasms of the brain are among the less common causes of facial paralysis.

Inflammatory and infectious causes of facial paralysis can occur. Infectious agents implicated include virally mediated conditions such as herpes zoster (eg, Ramsey Hunt syndrome), mumps, Coxsackie virus, and mononucleosis. Bacterial infections include sequelae of otitis media and Lyme disease. Inflammatory causes of facial paralysis include sarcoidosis.

The etiology of the facial paralysis determines the likelihood of spontaneous return of function, as in most cases of idiopathic facial paralysis (Bell palsy). The transected or severely damaged nerve must be repaired if satisfactory return of function is to be achieved. The problem remains a frustrating one because residual weakness and synkinesis are inevitable components of the healing process. Nonetheless, most patients benefit greatly from modern techniques of repair.[2, 3]

Today, several different procedures are available for repair of the facial nerve, including direct repair, cable nerve grafting, and nerve crossover techniques. Facial nerve decompression for cases of intact but damaged nerves and procedures of facial reanimation distinct from the repair of the facial nerve are discussed elsewhere (see Dynamic Reanimation for Facial Paralysis).[4, 5]

Active research continues to find ways to improve the results of treating facial nerve injuries. These include advances in the molecular biology of nerve regeneration and improved techniques of repair. A variety of neurotrophic and neurotropic factors have been investigated for their effects on the facial nerve. In addition, humoral factors have been identified as having a role in facial nerve regeneration.

Newer techniques proposed as possible alternatives to suture repair include laser neurorrhaphy and tissue adhesive repair. Synthetic and biologic tubules have been created to provide a path for the regenerating axons, even spanning small gaps in the nerve. Although these have not been clearly shown to be superior to standard nerve grafting techniques, they remain important areas of investigation.

Facial nerve repair is an option for cases of facial paralysis in which there is no reasonable likelihood of spontaneous return of function. Essential in determining whether repair is indicated are the cause of the paralysis and the duration of time since the injury. In this regard, some general principles will be helpful.

In general, transected nerves produce the best result when reapproximated. This produces an intact motor nerve supply from the facial motor nucleus in the pons to the muscle endpoint and is preferable whenever possible. If direct repair without tension is possible, it should be performed. Otherwise, a cable graft may be inserted to produce a tensionless coaptation of the proximal nerve stump to the distal branch or branches. Nerve crossover techniques are used when the proximal nerve stump is inadequate or inaccessible and cannot be used for grafting.

Another simple rule of thumb is that the sooner an injured facial nerve can be repaired, the better the long-term result. Once regarded as somewhat controversial, this rule is now believed to hold true for most cases.

In cases of trauma where the continuity of the nerve is in question, exploring within the first 3 days after injury is extremely desirable. Within this timeframe, the surgeon can use a nerve stimulator intraoperatively to identify the branches of the facial nerve. Once degeneration has occurred, stimulation of the nerve is not possible, and identifying the branches in an inflamed field can be extremely difficult.

Fibrosis of the nerve, fibrosis of the motor endplate, and atrophy of the muscle all ensue after injury. The surgeon is in a race against this inevitable process. Thus, a repair performed 18 months after the onset of paralysis is sure to have a poorer chance of success than a repair performed in the first month after injury.

Waiting may be appropriate in specific instances, depending on the health of the patient, oncologic surveillance issues, and other concerns. Nonetheless, when possible, maximizing the patients’ chance of a satisfactory outcome is desirable. In such cases, it is worthwhile to consider alternatives to facial nerve repair, such as facial reanimation procedures.

In addition, patient-specific factors influence clinical decision-making. In an elderly patient, slower nerve regeneration is expected, and the result of repair is likely to be poorer than could be achieved in a younger patient. If the patient’s life expectancy is short, one may elect to perform an adjunctive procedure that produces an immediate improvement (eg, a dynamic muscle sling) rather than a nerve repair that, even in the best of circumstances, takes some time to yield results.

Repair of the facial nerve is contraindicated when the motor endplate muscle unit is no longer functional. This occurs after long-standing paralysis in which fibrosis develops along with atrophy of the facial musculature. In such instances, reinnervation is not successful. The motor endplate muscle unit may fuse in cases of long-standing paralysis (ie, >1-1.5 years). Electrophysiologic testing can help determine whether this has occurred.

Advanced patient age is considered a relative contraindication by some. Anecdotal evidence suggests that the results of reinnervation techniques are poorer in elderly patients. The nerve regenerates more slowly, and results ultimately are not as good as those achieved with procedures performed on younger patients.

Planned radiation therapy is not a contraindication to facial nerve repair. Regeneration of nerve function has been demonstrated despite subsequent ablative doses of radiation.

Facial nerve repair may be contraindicated in other situations. These would include instances when the patient’s general health status prevents elective surgery.

The facial nerve may be divided into intracranial, intratemporal, and extratemporal components (see the image below). The intracranial portion of the facial nerve may be considered to include the supranuclear component (ie, the voluntary motor cortex, internal capsule, extrapyramidal system, midbrain, and pons) and the facial nerve nucleus and intracranial facial nerve. (See also Facial Nerve Anatomy and Brain Anatomy.)

The intratemporal portion of the facial nerve begins as the nerve enters the internal acoustic meatus and includes the well-described meatal, labyrinthine, tympanic, and vertical segments. The nerve then exits the stylomastoid foramen and soon divides at the pes anserinus. The subsequent branching to the temporal, zygomatic, buccal, marginal mandibular, and cervical branches shows some variability from person to person.

The blood supply to the facial nerve begins with the middle cerebral artery supplying the motor cortex. The facial nucleus in the pons is supplied by the anterior inferior cerebellar artery and the short and long circumferential arteries.

The facial nerve proper is then supplied by the anterior inferior cerebellar artery, the middle meningeal artery, and the stylomastoid branch of the postauricular artery. These tend to overlap; however, the region just proximal to the geniculate ganglion is thought to be somewhat susceptible to vascular compromise secondary to the poorer redundancy present there compared with other areas.

Some discussion of the microanatomy is warranted. Approximately 7000 neuron cell bodies make up the facial nerve, each of which innervates approximately 25 muscle fibers. The axons are surrounded by myelin, produced by the Schwann cells surrounding the axons. The nerve sheath is composed of the following 3 membranes:

• Epineurium – This is the outer covering, composed of loose areolar tissue, which separates the fascicles and holds them together

• Perineurium – This is the next inner layer, consisting of cells that are metabolically active and function as a diffusion barrier; the perineurium provides considerable strength to the nerve sheath

• Endoneurium – This membrane, the innermost layer, surrounds each of the individual nerve fibers

The spatial orientation of the facial nerve has been debated. In the cortex and brainstem, the nerve is spatially oriented. If it continued in an organized spatial orientation through its extra-axial course, this would have implications for the surgeon’s repair technique and ability to help prevent synkinesis; it would also allow some identification of the injured area of the nerve on the basis of clinical findings. Although such clinical observations have been made, the evidence overall suggests that the spatial orientation is not present in the extra-axial facial nerve.[6]

After significant injury, the facial nerve undergoes degeneration of the distal segment, as described by Waller. Sunderland classified such injuries into 5 types or degrees. These 5 types describe the pathophysiologic events associated with each of the disorders that may affect the nerve.

First-degree injury is referred to as neurapraxia, in which a physiologic block is produced by increased intraneural pressure (such as may be produced by external compression). The covering layers of the nerve (ie, endoneurium, perineurium, and epineurium) are not disrupted, and the nerve is capable of stimulation. Full return of function without synkinesis is observed.

Second-degree injury involves a similar mechanism, but the compression is unrelieved and results in degeneration of the nerve axons. This injury is termed axonotmesis, and again, excellent return is expected, though recovery may take several months. Because nerve stimulation is compromised, distinguishing axonotmesis is difficult.

Third- through fifth-degree injuries involve loss of endoneurial, perineurial, and epineurial tubes, respectively. Fourth- and fifth-degree injuries imply partial or complete transection of the nerve. Regeneration is incomplete, and synkinesis is inevitable. Repair of the facial nerve is generally performed in cases of complete paralysis.

After facial nerve repair, residual weakness and synkinesis are guaranteed. A House-Brackman grade III is the best possible result. Certain factors, including older age at time of repair, long grafts, and extended delay between time of injury and repair appear to limit the functional outcome of repair.

The vast majority of patients experience improved symmetry and tone after primary nerve repair or cable nerve grafting. More than 90% have better tone and symmetry after hypoglossal facial crossover. Training is required to help these patients learn to produce a smile by stimulating the hypoglossal nerve.

The history reveals the cause of the facial paralysis and is extremely important for treatment planning. The signs and symptoms of facial paralysis are obvious. The House-Brackman grading scale for facial paralysis is used to provide an objective description of the degree of paresis or paralysis. Patients demonstrate a lack of tone in addition to no movement on the affected side.

Of particular importance is the inability to protect the eye adequately. This should be assessed by careful inspection for signs of exposure keratitis, ability to close the eye, and the presence or absence of a Bell phenomenon. Questioning of partners and family members can reveal the adequacy of eye closure during sleep. Measures to protect the eye (eg, lubricants, artificial tears, or eye taping at night) must be instituted as appropriate.

In selected cases in which either the location or site of injury is unknown, both computed tomography (CT) and magnetic resonance imaging (MRI) may be helpful. High-resolution CT of the temporal bone provides the best imaging of the bony confines of the facial nerve and may reveal the pathologic site at any point along its course. MRI is superior for delineating the details of the soft tissues, including the nerve itself. Accordingly, it is the study of choice for diagnosing lesions such as acoustic neuromas and facial nerve schwannomas.

Several electrodiagnostic tests are used to evaluate the status of the facial nerve. These may be helpful in instances when the continuity of the nerve is unknown (eg, after penetrating trauma in which the function of the nerve immediately after injury is not documented).

More often, nerve testing is helpful in cases where transection is not a concern but the viability of the nerve is questionable. For example, testing would be helpful for patients with complete paralysis after acoustic neuroma surgery in which the nerve was maintained and patients with idiopathic facial nerve paralysis. Of course, these tests are useful for cases of complete facial nerve paralysis.

The tests available for evaluating the status of facial nerve function include nerve excitability testing (NET), maximal stimulation testing (MST), electromyography (EMG), and electroneurography (ENoG). Of these, the 2 that are currently considered most helpful are ENoG and EMG.

ENoG is an objective quantitative measurement of nerve function. It measures the summation potential of the response on the normal side of the face upon stimulation and compares the amplitude of the response to that on the paralyzed side. The measurement is thought to correspond to the number of remaining functional nerve fibers and thus to be predictive of the likelihood of spontaneous recovery. For example, in Bell palsy, patients who demonstrate greater than 90% degeneration within 14 days of onset have a decreased chance of recovery.

It should be kept in mind that ENoG (like MST and NET) is not useful in the first 3 days after injury, because Wallerian degeneration has not yet occurred to a significant enough extent. Tests performed during this timeframe overestimate the level of function.

EMG is often complementary to ENoG. It can be used to determine if the nerve in question is in fact in continuity (as evidenced by recorded volitional activity), shows evidence of Wallerian degeneration (as reflected by fibrillation potentials), or exhibits signs of reinnervation (as reflected by polyphasic innervation potentials). Fibrillation potentials typically arise 2-3 weeks after injury, and polyphasic reinnervation potentials may precede clinical signs of recovery by 6-12 weeks.

Currently, the following 3 options are available for surgical repair of the facial nerve:

• Direct nerve repair

• Cable nerve grafting

• Nerve substitution or techniques

Primary facial nerve repair

Direct repair of the facial nerve is the best method for rehabilitating the paralyzed face. It provides a chance of restoring spontaneous emotional expression to the face. Restoring continuity of the nerve through coaptation of the 2 nerve ends is indicated whenever the available nerve length is sufficient to allow this to be accomplished without undue tension.

Primary nerve repair may be performed virtually anywhere along the course of the nerve. An exception is the most proximal part of the nerve in the cerebellopontine angle if the length is inadequate for grafting. In addition, if the nerve has been transected distal to the lateral canthus, repair is usually unnecessary, because the nerve typically regenerates spontaneously in the more medial areas of the face.

If the nerve repair is to be successful, functional motor units must be available to receive reinnervation. That is, the facial musculature must not have atrophied excessively, and the motor endplates must be functional without significant fibrosis that would prevent reinnervation.

There remains some controversy as to how long after injury these conditions may still be met. Most authors consider reinnervation by direct repair to be possible up to 1 year after the injury. Some, however, argue that less than or more than a year is reasonable. In selected cases, EMG may be helpful to determine whether the nerve and muscle can be stimulated distal to the site of injury. If stimulation is possible, a functioning motor endplate muscle unit is likely.

The optimal timing of repair has been discussed extensively in the literature. The commonly accepted belief today is that an injured nerve should be repaired as soon as possible, not in a delayed fashion. This belief was initially promoted in the 1970s on the basis of experimental data suggesting that repair ideally should be performed 3 weeks after injury. Subsequent data, however, reinforced the idea that the best long-term results are achieved when the repair is performed as soon as possible.

Cable nerve grafting

When the length of facial nerve available is insufficient to permit primary repair, the best option is cable grafting. The same general principles of microsurgery that apply to direct repair also apply to cable grafting. Preoperative considerations must include a discussion with the patient regarding the options for donor nerves for repair. Commonly used nerves include the great auricular nerve and the sural nerve.

The great auricular nerve has the advantages of proximity to the operative field and ease of harvest. The diameter is usually an appropriately sized match for the facial nerve. Because up to 10 cm may be harvested, this nerve is a good choice for most cable nerve grafts. The main disadvantage is the numbness of the ear that inevitably results.

In addition, use of the great auricular nerve has been discouraged in cases of malignant disease because of the possibility of microscopic involvement of the nerve. Those who are concerned about this possibility suggest using an alternative donor out of the field of involvement in cases resulting from facial nerve sacrifice secondary to malignancy, particularly when the tail of the parotid is involved or metastases to the neck are present.

The sural nerve has some advantages over the great auricular nerve, in that it is longer (up to 40 cm) and has a greater number of neural fascicles. Use of this nerve results in lateral foot numbness.

Another nerve that may be used for cable nerve grafting is the medial antebrachial cutaneous nerve of the upper arm. This can provide 15 cm of length and has a branching pattern that can be used to graft to multiple facial nerve branches.

Nerve substitution techniques

Cranial nerve substitution techniques include hypoglossal-facial (XII-VII) anastomosis and cross-face grafting. The spinal accessory–facial anastomosis is another well-described procedure, but it is largely relegated to historical status or situations in which other options are not available.

Hypoglossal-facial anastomosis

The hypoglossal-facial (XII-VII) anastomosis is the preferred approach to reanimating the face when the proximal end of the facial nerve is not available and the peripheral system is still viable. This situation may arise after surgery of the skull base, as in resection of an acoustic neuroma.

If sacrifice of the facial nerve is known at the time of tumor removal, hypoglossal-facial anastomosis is best performed at that time. In cases in which paralysis becomes evident postoperatively, the same principles of timing apply. If no return of function becomes apparent, performing the cranial nerve substitution procedure is appropriate. A successful result is possible many months later, but after 12-18 months, the likelihood of such an outcome is small.

Essentially, 2 options exist for this procedure. Either the hypoglossal nerve is completely transected and connected to the facial nerve, or a partial transection of the hypoglossal nerve is attached to the facial nerve. The latter may be accomplished by means of an interposition cable graft. An alternative is to mobilize the facial nerve via a mastoidectomy and nerve dissection, allowing the facial nerve to be transected more proximally. The facial nerve can then be brought to the partially transected hypoglossal nerve for repair.

The first option, a classic end-to-end anastomosis, involves sectioning the hypoglossal nerve distal to the takeoff of the descendens hypoglossi. This produces reliably improved facial tone and symmetry in more than 90% of patients. The improvement occurs over a period of 4-6 months; it is largely observed in the midface, is less noticeable in the lower face, and is even less significant in the upper face.

On the other hand, hemitongue mobility is impaired. This loss of function must be weighed against the expected gain in facial tone, symmetry, and movement. Spontaneous mimetic function is not expected to return, and some training is necessary to teach the patient to smile by stimulating the hypoglossal nerve. Nevertheless, this procedure is generally well tolerated and fairly successful.

The second option involves the use of jump grafts, in which the hypoglossal nerve is only partially transected and a graft (eg, the great auricular nerve) is used to connect the transected portion of the hypoglossal nerve to the end of the facial nerve. The advantages of this technique include minimization of tongue weakness and a purported decrease in synkinesis. The disadvantages include the need for 2 anastomoses and the availability of fewer nerve cells to “drive” the face.[7]

Cross-face grafting

A facial nerve cross-face anastomosis attempts to connect branches of the facial nerve of the normal side with corresponding branches of the paralyzed side. This procedure may be chosen in cases where the proximal facial nerve of the involved side is unavailable for repair. It is made possible by the redundancy of facial nerve innervation that is often present in many areas of the face, particularly the midface.

Alternatively, a donor nerve with less associated morbidity (eg, the marginal mandibular nerve) may be sacrificed for the sake of the other side. Some iatrogenic weakness of the donor side is expected, and the patient must be well informed of this likely result before consenting to undergo the procedure.

Some facial nerve surgeons believe that donor side weakness can contribute to success, noting that many patients are happier with a more symmetric-appearing face, even at the cost of some function. This is a matter of debate. As in all reinnervation procedures, success depends on the presence of functioning motor end plates in the paralyzed side. Thus, when paralysis has lasted longer than 1 year, the likelihood of success may be compromised. The cross-face technique has had variable success in different reports and remains controversial.

Early follow-up relates to typical wound care. Long-term follow-up is very important. With most procedures, the results are delayed for several months to a year. In some cases, results are delayed for as long as 2 years, particularly with nerve transposition repairs.[8] For this reason, adjunctive procedures (eg, gold weight eyelid implant and brow lift) may be considered early.

If no improvement is observed 1 year after the repair, reexploration of the site of anastomosis is acceptable, and revision is performed if the repair is found to be inadequate. Another option is to consider adjunctive functional and cosmetic procedures or even alternative measures such as dynamic or static slings and free tissue transfer techniques.

Currently, no medical treatment exists for facial nerve repair. Systemic corticosteroids and/or anti-virals are advocated by some to minimize swelling of the nerve in certain cases. In animal models, the use of electrical stimulation therapy appears to be beneficial for initiating and accelerating facial nerve recovery.[9] A number of metabolic factors (eg, neurotrophic factors, growth factors, and stem cells) have shown some promise for facilitating nerve repair in the laboratory, although these remain in the investigational phase and results to date have been mixed.

In animal experiments, glial cell line–derived neurotrophic factors promoted facial nerve regeneration in delayed grafting but inhibited immediate nerve grafting.[10] In similar experiments, stem cell therapy (eg, bone marrow–derived mesenchymal stem cells in collagen) promoted excessive growth support for axon regeneration and excessive collateral nerve branching of facial motor endplates (which was not improved by manual stimulation). Current stem cell therapy requires additional study before it can be clinically useful for facial nerve repair.[11]

Accordingly, surgical repair is the mainstay of treatment. At present, there are 3 surgical options for repair of the facial nerve: direct repair, cable nerve grafting, and nerve substitution techniques (see below). Primary end-to-end nerve anastomosis and cable graft interposition have shown to produce better functional outcomes than nerve substitution techniques.[8]

Additional techniques that have been proposed as possible alternatives to suture repair include laser neurorrhaphy and tissue adhesive repair. Synthetic and biologic tubules have been created to provide a path for the regenerating axons, even spanning small gaps in the nerve. These have not yet shown any clear superiority to standard nerve grafting techniques, but they remain under investigation.

Controversies persist regarding the type (epineural versus perineural) and the timing of repair. The cross-facial nerve graft has very good results in certain reports but has not been as effective in other hands.

The facial nerve is exposed to a length sufficient to permit the exposure and mobility required for performance of the grafting procedure. A parotidectomy incision is made, and the facial nerve is identified as it exits the stylomastoid foramen by using the traditional landmarks (ie, the tragal pointer, the sternocleidomastoid, the posterior belly of the digastric, and the stylomastoid suture).

Once the nerve is identified, it is followed distally as necessary. Meticulous, gentle technique is employed, as in a parotidectomy. If the intratemporal portion of the nerve is involved or requires exposure, it is exposed via a mastoidectomy.

The key to successful nerve grafting is careful coaptation of the nerve ends without tension. Several principles apply, regardless of which specific type of procedure is chosen. The nerve tissue must be handled atraumatically with microinstruments. Magnification with an operating microscope is essential to allow meticulously precise alignment of the nerve ends. A colored background is helpful for facilitating clear visualization of the anatomy.

Typically, 2-3 sutures are carefully placed through the epineurium. Fine (8-0 to 10-0) monofilament permanent suture material is used; this minimizes tissue reactivity. Although epineurial suturing is generally considered the standard, there is some controversy on this point. Perineural and endoneurial or intrafascicular repairs have been advocated, but the data are unclear as to whether these techniques actually offer significant advantages.

The important relation to keep in mind is the size match between the endoneurial surfaces. This must be inspected with magnification; if a mismatch is evident, then one end may be trimmed in a beveled fashion to yield a better match between the surface areas of the ends to be approximated.

In some cases, a nerve segment may be unavailable for reapproximation, either because of resection of a malignancy or because of destruction by trauma. This prevents primary grafting without tension. The problem may be addressed in a number of ways. For example, if the injury site is intracranial or intratemporal, the facial nerve must be rerouted, repaired with a cable graft, or both. In this instance, one may mobilize the facial nerve from the fallopian canal by decorticating it via a transmastoid approach.

If hearing preservation is not a concern, a translabyrinthine approach allows exposure of the entire length of the nerve up to the internal auditory canal. The mastoid tip may be removed to afford the nerve greater mobility. Furthermore, the entire nerve can be rerouted by combining the mastoidectomy with a middle fossa approach, which allows grafting of the proximal portion of the nerve.

Peripheral injuries likewise may necessitate mobilization of the nerve through one of the abovementioned techniques in order to allow primary repair. Repairing the nerve with primary grafting is always preferable when feasible. If it is not feasible, then a cable graft may be used.

The great auricular nerve is found by drawing a line between the angle of the jaw and the mastoid tip. This line is bisected at a right angle by the great auricular nerve as it passes around the posterior border of the sternocleidomastoid muscle just behind the external jugular vein. The nerve is the largest of the ascending branches of the superficial cervical plexus and arises from C2-3. Extra branches can be found by following the nerve toward its origin behind the sternocleidomastoid muscle.

The sural nerve can be located between the lateral malleolus and Achilles tendon. It lies just deep or posterior to the saphenous vein. It then runs superiorly up the back of the lower leg in a subcutaneous plane until it descends between the 2 heads of the gastrocnemius toward the popliteal fossa and its origin off the tibial nerve.

The sural nerve may be harvested either by making a single long incision from the ankle to the popliteal fossa (depending on the length of nerve required) or a series of shorter transverse incisions. The nerve may be dissected under direct vision with the single incision or by using a fascia stripper and making the stepwise incisions.

The technique of nerve grafting is the same as for primary repair. In the case of cable grafting, obtaining enough nerve graft length to allow the graft to have some redundancy between the ends of facial nerve may be helpful. This would create a C or S shape and ensure tension-free coaptation.

Hypoglossal-facial anastomosis

A parotidectomy-type incision is made. The facial nerve is identified as it exits the stylomastoid foramen and is followed to just beyond the pes anserinus. It is sharply transected where it exits the stylomastoid foramen.

The hypoglossal nerve is then isolated in the neck. It may be identified by following the posterior belly of the digastric toward the hyoid bone. The hypoglossal nerve passes lateral to the carotid artery and medial to the internal jugular vein. The nerve may then be followed distally to gain the maximum length for anastomosis. The descendens hypoglossi is usually transected to aid in mobilization and length.

In the case of an end-to-end anastomosis, the hypoglossal nerve is transected distally and brought to meet the facial nerve by passing it either medial or lateral to the digastric muscle. The grafting of the nerve ends proceeds as previously discussed.

If a jump graft is performed, identification and transection of the facial nerve proceed in the same manner. In this case, an appropriate length of great auricular nerve is then harvested. The hypoglossal nerve is then identified, including the descendens branch. The nerve is partially incised in a beveled fashion (most authors recommending cutting one third to one half the way across). The great auricular nerve graft is anastomosed to the proximal portion of the beveled cut. The other end is grafted to the distal end of the facial nerve in a typical fashion.

Another option for preserving tongue function is the split hypoglossal nerve graft. In this technique, the hypoglossal nerve is split and dissected back to obtain a length that can be anastomosed to the facial nerve, which means that only 1 anastomosis would be required. However, given that the hypoglossal nerve also exhibits a lack of spatial orientation, with the nerve fibers interwoven randomly, some consider this approach flawed concept because splitting of the nerve any significant distance would cause significant denervation of the tongue.

Cross-facial nerve graft

Several variations of the cross-facial anastomosis have been described. These differ in terms of which donor facial nerve branches are used, how many cross-facial grafts are performed, and the path chosen to pass the graft across the face to the other side.

In each variation of the procedure, a suitable nerve to graft must be chosen. The sural nerve is most commonly used because the length of nerve required to thread across the entire face, particularly if several branches are to be anastomosed, is quite significant. Alternatively, one may perform a single graft, in which case the great auricular nerve may be sufficient.

Dissection is performed first on the paralyzed side to ensure that the distal nerve can be identified. If several branches are to be grafted, the branches may be identified by following the nerve beyond the pes anserinus and identifying the branches within or just beyond the parotid gland. The contralateral normal facial nerve is identified in the standard fashion.

The branches are individually identified. The donors are taken at the distal border of the parotid gland. A nerve stimulator is used to determine the areas innervated by the individual branches, and for each facial region, an attempt is made to find multiple nerve branches supplying a given area. This allows for the sacrifice of 1 branch with preservation of an adequate degree of function for a given anatomic region.

The sural nerve is then harvested and the branches tunneled across the face. This may be performed above and below the lip and through the neck. Passing the nerve across the forehead may have a poorer result because of the lesser blood supply to the graft lying within a relative lack of soft tissue. The nerve graft is sutured to the normal side. The graft may then be sutured to the chosen branch or branches on the paralyzed side.

Another option is to forego the anastomosis on the paralyzed side. A “babysitter” hypoglossal nerve anastomosis may be performed instead. This provides innervation to the paralyzed facial musculature and thereby prevents further degradation of the motor endplates and muscle atrophy as regeneration occurs across the considerable distance necessary for the cross-face graft.

In this situation, the Tinel sign (ie, paresthesias associated with areas of nerve regeneration) is followed across the face for several months, and the patient is returned to the operating room when this sign indicates that regeneration has completely spanned the gap. The grafts are then anastomosed to the paralyzed branches.

A simpler technique uses only the marginal mandibular branch, which is sacrificed and anastomosed to the main branch of the paralyzed nerve in a single setting. Function for each of these cases takes at least 6 months to begin to return. As with the other procedures, synkinesis is expected.

Postoperative care is similar for each of the procedures used for facial nerve repair. Meticulous hemostasis is important, with drains placed to prevent hematoma formation. Routine attention to details such as eye protection remains of paramount importance. Return of function takes months to occur.

Synkinesis is expected for all cases of facial nerve transection, regardless of the mode of repair chosen, and the best result one can hope for is a House-Brackman grade III. These consequences therefore should be viewed not as complications but as expected sequelae.

Donor-site morbidity is also expected for the hypoglossal crossover technique (tongue weakness), great auricular harvesting (ear numbness), and sural nerve harvesting (lateral leg numbness). Complications for each of the procedures include hematoma and infection.