BELL’S PALSY

Bell (1774-1842)¹ first described a patient with a facial paralysis in 1818; subsequently, all patients with facial palsy of unknown etiology have come to bear his name. The etiology of this “idiopathic” disorder has become much clearer in recent years. Although first proposed in 1972 by McCormick,² herpes simplex virus (HSV) has been identified as the disease vector only recently,^3–5 and an animal model has been designed.⁶ Murakami and associates⁵ identified HSV type 1 (HSV-1) DNA fragments in perineural fluid in 11 of 14 patients undergoing facial nerve decompression. In this study, no control subjects had HSV-1 DNA in perineural fluid. Using polymerase chain reaction to analyze the saliva of patients with Bell’s palsy, Furuta and colleagues³ identified HSV-1 DNA in 50% of patients, which was significantly more often than controls, a finding confirmed by other groups.⁷ Polymerase chain reaction has also been used to isolate HSV-1 genomic DNA from the geniculate ganglion of a temporal bone in a patient dying during the acute phase of Bell’s palsy.⁴

Sugita and coworkers⁶ proposed an animal model of Bell’s palsy. Six days after inoculation of HSV-1 into either the auricle or the tongue of mice, a temporary ipsilateral facial paralysis was identified that recovered spontaneously within 3 to 7 days. Histopathologically, neural edema, vacuolar degeneration, and inflammatory cell infiltrate with associated demyelination or axonal degeneration were shown in the affected facial nerve and nucleus.⁸ HSV-1 antigens were identified within the facial nerve, geniculate ganglion, and facial nucleus 6 to 20 days after inoculation.⁶ Similar pathologic findings have been shown in rabbits after HSV-1 inoculation, but without the associated facial paralysis.⁹

The histopathologic changes seen in autopsy specimens of patients who died with acute idiopathic facial paralysis have provided some insight into the underlying cellular mechanisms in Bell’s palsy. Reddy and coworkers¹⁰ found degeneration of the myelin sheath and axons, perivascular inflammation, and a phagocytic cell infiltrate in 10% to 30% of nerve fibers in a patient 17 days after the onset of an acute idiopathic facial paralysis. Fowler¹¹ found diffuse vascular engorgement throughout the intratemporal facial nerve and evidence of acute hemorrhage within the intracanalicular, labyrinthine, and geniculate portions of the facial nerve in a patient who died shortly after the onset of Bell’s palsy. In examining an autopsy specimen of a patient who died 13 days after the onset of Bell’s palsy, McKeever and colleagues¹² found a diffuse lymphocytic infiltration of the intratemporal facial nerve with ongoing myelin phagocytosis. They later re-examined the same specimen and noted the most pronounced lymphocytic infiltration of the nerve at the labyrinthine segment, findings that these authors believed were most consistent with an ongoing compression-type injury to the facial nerve.

Multiple reports have found that there seems to be evidence of constriction of the nerve at the meatal foramen in cases of Bell’s palsy. In specimens examined during the acute phase of Bell’s palsy, lymphocytic infiltration, perineural edema, and myelin degeneration have been noted.^13,14 In a nerve specimen examined 1 year after the diagnosis of Bell’s Palsy lymphocytic infiltration, perineural edema, and fibrotic changes were noted.¹⁵ Intraoperative biopsy specimens of the greater superficial petrosal nerve from patients undergoing nerve decompression procedures for Bell’s palsy have shown axonal demyelination and degeneration with lymphocytic infiltration.¹⁶ Together, the histopathologic findings in Bell’s palsy show demyelination and axonal loss with lymphocytic or phagocytic infiltration, findings that seem to be more pronounced at the labyrinthine segment of the facial nerve.

In considering this evidence, the pathogenesis of Bell’s palsy becomes more apparent: a virally induced, inflammatory response that produces edema within the nerve. Fisch and Felix¹⁶ first proposed that the facial nerve was entrapped at the meatal foramen as a result of neural edema. Intraoperative conduction studies have shown an electrophysiologic blockage at this site.¹⁷ The constriction imposed produces a conduction block at first; however, with prolonged or increased constriction, ischemia results. Subsequently, wallerian degeneration occurs, producing axonotmesis or neurotmesis or both. A spectrum of injury within the nerve from neurapraxia to neurotmesis may occur in Bell’s palsy.^11,18 The proportion of each of these determines the amount of facial function that returns when the acute phase of the disease subsides.

Bell’s palsy accounts for nearly three quarters of all acute facial palsies.¹⁹ The incidence of Bell’s palsy is 20 to 30 cases per 100,000 per year.²⁰ The median age is 40 years, but it can occur at any age.²¹ The incidence is highest in patients older than 70 years and lowest in children younger than 10 years. The left and right sides are affected equally. Men and women are equally affected, but the incidence of Bell’s palsy is higher in pregnant women (45 cases per 100,000).

The clinical presentation of Bell’s palsy is well known; however, the clinician must exclude other causes of facial paralysis based on history and physical examination findings. Patients describe an abrupt onset of unilateral paresis that occurs over 24 to 48 hours. The paresis can progress over 1 to 7 days to complete paralysis. Bilateral involvement, either simultaneously or consecutively, has been described.²³ A history of progression of weakness over weeks to months, repeated episodes of paralysis, and twitching of the facial muscles should not be considered symptoms of Bell’s palsy. Other associated symptoms of hearing loss, vestibular symptoms, or other cranial nerve neuropathies also rule out the diagnosis of Bell’s palsy.

On physical examination, the patient displays unilateral weakness or flaccid paralysis of all branches of the facial nerve. If forehead movement is normal, and there is strong eye closure and symmetric blinking, a central origin of paralysis should be suspected. The tympanic membrane should be of normal color and mobility. Careful bimanual palpation of the parotid gland may reveal a deep lobe parotid neoplasm. Oral cavity examination may show loss of papillae on the ipsilateral tongue. Cranial nerve testing is normal with the exception of the involved CN VII. Serial examinations are essential; if some evidence of recovery is not noted within 3 to 6 months, an aggressive search for an underlying neoplasm should be undertaken.

Audiometric evaluation should reveal symmetric function except for an absent ipsilateral acoustic reflex. If unilateral hearing loss or acoustic reflex decay is present, further evaluation for retrocochlear pathology is necessary. If vestibular complaints are present, an electronystagmogram is performed, and diagnosis of Bell’s palsy should be questioned. If the history and clinical presentation are highly suggestive of Bell’s palsy, magnetic resonance imaging (MRI) and computed tomography (CT) are not performed. High-resolution CT scans and MRI are obtained if patients have associated symptoms of otorrhea, vestibular complaints, and hearing loss. Planned surgical decompression and persistent dense paralysis after 6 months are also indications for imaging. Serologic tests for Lyme disease (IgG, IgM) are an important part of the work-up for unexplained facial paralysis in endemic areas.²⁴ Serologic testing for antibodies to HSV or varicella zoster have provided some correlative evidence for a viral etiology in Bell’s palsy, although laboratory investigations in general have not been found to provide clinically relevant data.

Electrodiagnostic testing is an important element of the diagnostic evaluation of facial paralysis. Testing can determine the extent of facial nerve injury and provide useful prognostic information for the development of management strategies. The technique of electroneuronography (ENoG), developed by Esslen,²⁵ can distinguish the nerve fibers that have undergone wallerian degeneration from fibers that are temporarily blocked (neurapraxia). ENoG testing is not performed until 3 to 4 days after the development of complete unilateral paralysis because wallerian degeneration does not become apparent until 48 to 72 hours after an acute injury to the nerve. Electrodiagnostic testing is not performed when the patient exhibits paresis only. If the paresis progresses to total paralysis, electrodiagnostic testing is performed 3 days after the onset of total paralysis. Presence of voluntary movement 4 to 5 days after the onset of paresis indicates only minor injury, and complete recovery should be anticipated.

ENoG is most accurate when it is performed within 3 weeks of the acute injury. The test is performed using standard electromyography (EMG) equipment, but requires the use of special surface stimulating and recording electrodes.²⁶ The recording electrodes are in a hand-held carrier and are manipulated in the nasolabial fold with the Esslen technique. The recording electrodes are not taped to the skin, as has been described by others.²⁷

An evoked electric stimulus generates synchronous facial muscle movement that can be recorded from the skin surface (termed the compound muscle action potential [CMAP]). The amplitude of the biphasic CMAP has been found to correlate with the number of blocked or neurapraxic nerve fibers. As the percentage of degenerated fibers within the nerve increases, the amplitude of the CMAP decreases compared with the normal side of the face. Fisch and Esslen²⁸ determined that if 90% or more of the fibers within the facial nerve degenerate within the first 14 days of an acute paralysis, a severe injury has occurred, and the chances of complete recovery are less than 50%. Patients who do not reach the 90% degeneration level by 3 weeks have a very good prognosis and are likely to regain normal facial motion without synkinesis. The time course of degeneration is also important; the more rapid the degeneration, the more severe the injury.²⁹ A patient showing greater than 90% degeneration at 5 days would have a worse prognosis than a patient with 90% degeneration at 14 days.

Patients exhibiting 90% to 100% degeneration or no response to electric stimulation, in addition to ENoG, must undergo EMG testing. An EMG needle is placed in the orbicularis oris and oculi muscles, and the patient is asked to make a forceful contraction. Any voluntary motor unit activity indicates deblocking of the conduction block and a favorable prognosis. When deblocking occurs, fibers may not discharge at the same rate because of the previous injury, and may fail to generate a surface CMAP, resulting in a false-positive test result on ENoG testing. As movement returns to the face, the surface CMAP may also be absent for the same reason. Voluntary evoked EMG testing is mandatory if a surgical decompression is planned. If a facial paralysis has been slowly progressive over weeks to months, degeneration and regeneration of nerve fibers occur within the nerve, resulting in similar dyssynchronous discharge of evoked impulses and an inaccurate CMAP recording.

Topognostic testing is widely reported to be useful in determining the site of injury in acute facial paralysis; however, intraoperative studies have shown that the Schirmer test is not accurate in diagnosing Bell’s palsy.²⁶ The Schirmer test may be used to determine the extent of lacrimation and the need for eye protection.

A review of the natural history of Bell’s palsy shows that approximately 85% of patients begin to display some return of facial movement within 3 weeks of the onset of paresis.³⁰ The remaining 15% begin to improve 3 to 6 months after the onset of the disease. Most patients show a complete return of facial function, but 10% to 15% have residual unilateral weakness and develop secondary deformities of synkinesis, tearing, or contracture. Some motion returns in almost all individuals with Bell’s palsy by 6 months. If no movement returns, a vigorous search for another etiology should begin anew.

The management of patients with Bell’s palsy varies, depending on the type of specialist initially seeing the patient and on the training of the individual specialist. Figure 28-1 presents an overview of our management strategy. Patients presenting within the first week of facial weakness with paresis are placed on steroid therapy (prednisone, 60 to 80 mg/day for 7 days) and an antiviral agent (valacyclovir, 500 mg three times per day for 10 days). If patients are seen 7 to 10 days after onset and motor function is stable or improving, medical treatment is unnecessary. Patients are instructed to return in 1 week for re-evaluation to determine if neural degeneration has occurred. Electrodiagnostic testing is unnecessary as long as voluntary facial movement is present. If total paralysis occurs in the interim, the total paralysis protocol of Figure 28-1 is followed. Stable or improving patients are seen in 1 month.

FIGURE 28-1 Algorithm for management of Bell’s palsy. ENoG, electroneuronography; MCF, middle cranial fossa.

The use of steroids in Bell’s palsy has been the subject of much debate. Numerous studies of varying designs in adults have shown better outcomes in patients treated with steroids,^21,31,32 especially when initiated early in the course of the disease.^33–35 Other randomized studies³⁶ and meta-analyses,³⁷ including many studies in children,^38–40 have concluded that steroids did not affect the ultimate facial function outcomes in Bell’s palsy. Grogan and Gronseth,⁴¹ in a comprehensive, evidence-based review, concluded that there seemed to be a beneficial effect in the use of steroids in Bell’s palsy. Ramsey and coworkers⁴² performed a meta-analysis of 47 trials of steroid therapy for Bell’s palsy, and concluded that there seemed to be an improved odds of recovery in patients treated with steroids (49% to 97%) compared with untreated controls (23% to 64%). We use prednisone in a dose of 1 mg/kg daily for 7 days in all cases in which it is not medically contraindicated, in anticipation of speeding recovery, reducing the number of degenerating axons, and reducing the number of patients needing decompression.

The combination of steroids and antivirals may be superior to either one alone. A double-blind, randomized, controlled trial of acyclovir and prednisone versus prednisone alone in the treatment of Bell’s palsy showed better results with the combination therapy.⁴³ This study documented poor facial function recovery in 23% of the prednisone-only group compared with 7% in the acyclovir-plus-prednisone group. Interpreting the results of this study, Grogan and Gronseth⁴¹ reported that patients who receive antiviral therapy in addition to steroids were 1.22 times as likely to have a good facial nerve outcome. A multicenter, randomized, placebo-controlled study comparing treatment of patients with Bell’s palsy with steroids and antivirals with treatment with steroids alone concluded that the addition of valacyclovir improved the recovery rate from 75% to 90% in cases of complete palsy, and from 89% to 96% in all cases of facial palsy.⁴⁴ Many other reviews of the topic have made similar conclusions.^45–48

Other studies have found no significant difference between this combination of drugs and the natural history of the disease.^24,49 One study found a negative impact of treatment of patients with Bell’s palsy with antivirals alone versus steroids alone.⁵⁰ Because of the paucity of potential side effects and good patient tolerance of antiviral medications, the addition of antiviral medications to steroids in the treatment of patients with Bell’s palsy seems prudent.

Patients presenting within 1 week of the onset of total unilateral paralysis undergo electrodiagnostic testing (if at least 3 days have passed since the onset of paralysis) and are started on medical therapy. If the patient is seen in the first 3 days after the onset of paralysis, steroid and antiviral therapy are initiated, and follow-up electrodiagnostic studies are arranged. The frequency of follow-up electrodiagnostic examinations is determined by the result of testing and the time interval after paralysis, as suggested by Fisch⁵¹ (Fig. 28-2). Patients exhibiting nearly 90% neural degeneration on ENoG examination, or who are degenerating quickly, undergo frequent electrodiagnostic testing (every 1 or 2 days). If greater than 90% degeneration is reached, and there are no motor unit potentials on voluntary EMG testing, the patient is considered a candidate for middle cranial fossa decompression. When 90% degeneration is not reached within 2 weeks (14 days) after the onset of paralysis, no further electrodiagnostic studies need to be performed.

FIGURE 28-2 Recommended electroneuronography (ENoG) testing schedule in acute facial paralysis.

Patients seen for the first time more than 2 weeks after the onset of paralysis undergo EMG evaluation to determine if regeneration has begun. They are scheduled for a 6 month follow-up to ensure that some motor function has returned. If no movement is evident at 6 months, it must be assumed that Bell’s palsy was an incorrect diagnosis, and a search for another disease process is begun.

RAMSAY HUNT SYNDROME

Herpes zoster oticus refers to a syndrome of acute otalgia accompanied by a herpetic, vesicular rash. When accompanied by facial paralysis, the syndrome is known as Ramsay Hunt syndrome.⁵² Ramsay Hunt syndrome is the second most common cause of facial paralysis (after Bell’s palsy), and is induced by the reactivation of the varicella-zoster virus that remains latent in the geniculate ganglion after primary infection with chickenpox.⁵³ Classically, patients present with severe otalgia and unilateral facial paralysis. Vesicular eruptions may or may not be present initially, but usually appear within 3 to 5 days of the paralysis. The vesicular lesions can appear on the concha, ear canal, postauricular skin, and tympanic membrane. Occasionally, the oral cavity, neck, and shoulder are also involved. The disease can affect other cranial nerves, including auditory, vestibular, trigeminal, glossopharyngeal, and vagus, prompting the name herpes zoster cephalicus.⁵⁴

Also in contrast to Bell’s palsy, the symptoms are more severe, and the prognosis is worse in Ramsay Hunt syndrome. The frequency of complete neural degeneration of the facial nerve is substantially higher than with Bell’s palsy, and complete recovery of facial motor function has ranged from 10% to 31% in several studies.^55–57 Patients with auditory and vestibular dysfunction in addition to facial paralysis generally have a worse prognosis. In addition, patients with diabetes, hypertension, and advanced age have been reported to have a poorer prognosis in Ramsay Hunt syndrome.⁵⁸

The diagnosis of Ramsay Hunt syndrome is based on the history of otalgia, vesicular lesions or eschars, and facial paralysis. MRI shows enhancement of a large portion of the facial nerve, often the vestibular and cochlear nerves, the labyrinth, and the dura lining the internal auditory canal.⁵⁹ Imaging as part of routine evaluation is unnecessary. Electrodiagnostic studies have been unreliable in herpes zoster oticus.

The management of Ramsay Hunt syndrome has changed with the development of antiviral agents. The natural history of the disease was evaluated by Devriese and Moesker,⁵⁶ who identified only a 10% rate of complete facial nerve recovery. Many studies have identified a superior recovery rate of 75% using a combination of steroids and antiviral agents.^45,60–62 The benefit of early initiation of steroid and antiviral treatment was clearly shown in a study by Murakami and colleagues,⁶⁰ in which 75% of patients treated within 3 days of the onset had complete recovery, whereas only 30% of patients treated after 7 days experienced a complete recovery. We have experienced similar successful results using intravenous acyclovir (10 mg/kg three times daily), oral acyclovir (800 mg five times daily), or oral valacyclovir (500 mg three times daily) for 10 days, in combination with a 3 week tapering course of prednisone (60 to 80 mg/kg daily). Patients report rapid reduction in pain, and occasionally experience return of facial movement during the medical therapy. Because of the presence of “skip” regions and diffuse neuritis of the facial nerve,^63,64 surgical decompression of RHS is not recommended.

FACIAL NERVE DECOMPRESSION FOR BELL’S PALSY

Decompression of the facial nerve for Bell’s palsy has been reported since the 1930s,⁶⁵ and is recommended in patients who exhibit greater than 90% degeneration on ENoG and no voluntary EMG activity within 14 days of the paralysis. Preliminary results suggested that early decompression of the labyrinthine, geniculate, and tympanic segments of the facial nerve through a middle cranial fossa approach improves recovery of severely degenerated cases,²⁵ whereas decompression of the mastoid segment alone did not alter the natural history of the disease.⁶⁶ Gantz and associates⁶⁷ showed a statistically significant difference in facial nerve outcome when decompression is performed within 2 weeks of onset of the paralysis. The patients undergoing decompression exhibited House-Brackmann grade I (n = 14) or II (n = 17) in 91% of cases; two patients had a grade III, and one patient had a grade IV outcome. There were no grade V or VI results in the surgical group. Patients who met surgical criteria but elected not to undergo surgery had a 58% chance of a poor outcome. Within the nonsurgical group, 19 patients had a House-Brackmann grade of either III or IV at 7 month follow-up.

Intraoperative evoked EMG is used to localize the region of the conduction block.¹⁷ Intraoperative direct nerve stimulation is useful when the nerve is not completely degenerated, which includes most instances when preoperative ENoG reveals 100% degeneration. This finding suggests that a few nerve fibers remain capable of stimulation even when ENoG shows total nerve degeneration. In most cases, the nerve conduction block is localized between the geniculate ganglion and the internal auditory canal segments of the facial nerve. Failure to generate a motor unit potential when the tympanic segment of the nerve is stimulated indicates that the nerve is 100% degenerated, or that the conduction block is more distal. If the nerve appears normal in the tympanic portion, while there is edema and erythema of the internal auditory canal segment, total degeneration should be suspected. If erythema and edema are apparent in the tympanic segment, a mastoid decompression is added.

Preoperative Preparation

The risks of middle fossa decompression of the facial nerve are discussed with the patient, and include cerebrospinal fluid leak (4% to 6%), infection (1%), hearing loss (1%), dizziness (1%), intracranial hemorrhage (<1%), and aphasia (<1%). When the decision is made to proceed with surgical decompression, the procedure should be performed as soon as possible. Appropriate preoperative laboratory and imaging studies are obtained, along with a Stenvers projection plain radiograph of the temporal bone to identify the floor of the middle cranial fossa and superior semicircular canal. Cefazolin and dexamethasone are given before the skin incision and are continued for six doses.

General anesthesia and transoral endotracheal intubation are accomplished; thereafter, the bed is rotated 180 degrees for access (Fig. 28-3). Paralytic agents must be reversed before the skin incision. A urinary catheter is placed to monitor fluids and diuresis. Hair is shaved approximately 10 cm above and 5 cm behind the ear. The entire side of the head and face are prepared, and EMG needles are placed in the orbicularis oculi and oris muscles (Fig. 28-4A) for facial nerve monitoring. A clear drape is placed over the prepared area to allow visualization of the entire side of the face in the case of monitoring equipment failure (Fig. 28-4B). Standard auditory brainstem–evoked recording electrodes are placed in the right and left mastoid tips and the vertex and forehead (ground), and insert headphones are placed in both external auditory canals. The auditory brainstem response is monitored throughout the procedure.

FIGURE 28-3 Operating room setup for middle cranial fossa approach.

FIGURE 28-4 A, Electrode placement for intraoperative electromyography. B, Draping with exposure of half the face for visual monitoring of facial movement.

FIGURE 28-5. Skin incision for middle cranial fossa approach. Mastoid exposure can be obtained by extending incision postauricularly. Incision of anteriorly based temporalis muscle flap is offset to keep suture lines from being directly in line.

Surgical Technique

The skin incision is marked as shown in Figure 28-5 and carried down to the level of the temporalis fascia. Meanwhile, mannitol (0.5 g/kg body weight) and hyperventilation (partial pressure of carbon dioxide of 25 mm Hg) are initiated by the anesthesiologist to relax the brain. After the posteriorly based skin flap is elevated, a 4 × 6 cm piece of temporalis fascia is harvested and set aside in a moist gauze for use at the time of closure. An anteriorly based, inferiorly staggered muscle flap is elevated down to the level of the linea temporalis and reflected forward with an Adson cerebellar retractor. Staggering the incisions prevents dural exposure if wound dehiscence occurs. The zygomatic root identifies the floor of the middle cranial fossa and is the central landmark of the craniotomy. The skin and muscle flaps should be wrapped with moist sponges and secured with temporary retraction stitches.

The craniotomy should be approximately 4 cm in anteroposterior dimension and 5 cm cephalocaudal (Fig. 28-6A). The anteroposterior margins must be kept parallel for stability of the middle cranial fossa retractor. The craniotomy can be created with either cutting burrs or a craniotomy saw. The bone flap should be elevated with care by use of a blunt dural elevator. Occasionally, the middle meningeal artery is embedded within the bone, requiring bipolar coagulation to free it. The bone flap is wrapped in a moist gauze and set aside for use at closure.

FIGURE 28-6 A, Craniotomy for middle cranial fossa exposure. Inferior expansion of craniotomy allows more exposure during dural elevation. Vertical margins must be parallel for stability of retractor. Craniotomy should be centered on the temporal root of the zygoma (dashed line). B, Placement of House-Urban middle cranial fossa retractor.

FIGURE 28-7. Exposure of superior semicircular canal blue line and position of internal auditory canal (IAC) 60 degrees anterior to a line through the blue line. Superior canal blue line almost always is perpendicular to petrous ridge.

Dural elevation from the floor of the middle cranial fossa is accomplished with a Freer or joker elevator, always in a posterior-to-anterior direction, which prevents injury to the greater superficial petrosal nerve and geniculate ganglion. The petrous ridge is identified at the posterior margin of the craniotomy, and the dura is slowly elevated over the arcuate eminence and meatal plane. Dural reflections are cauterized and sharply transected to allow elevation to the anterior petrous ridge. The hiatus of the facial canal with the greater superficial petrosal nerve and artery is the anterior margin of elevation. Further anterior elevation exposes the foramen spinosum and the pterygoid plexus of veins, which can cause troublesome oozing throughout the procedure. After dural elevation, cottonoid sponges can be placed at the anterior and posterior margins of the elevation to help retract the dura during placement of the self-retaining middle cranial fossa retractor (Fig. 28-6B).

Before the bony exposure of the facial nerve, the Stenvers projection radiograph is re-examined to determine the depth of the superior semicircular canal in the temporal bone. The superior semicircular canal is the first structure to be located. When its blue line is identified, the remaining intratemporal structures have consistent anatomic locations. Landmarks of the middle cranial fossa floor can be quite subtle. The arcuate eminence may not be apparent and, in many instances, is not parallel with the superior semicircular canal. One consistent anatomic feature to remember is that the plane of the superior semicircular canal is almost always perpendicular to the petrous ridge (Fig. 28-7). If the arcuate eminence is not initially apparent, drilling is begun posterior to the semicircular canal, slowly removing the tegmen mastoideum with a moderate-sized diamond burr. The whitish color of the membranous temporal bone can be distinguished from the yellow, dense otic capsule bone of the superior canal. When the superior canal is identified, drilling in a parallel direction with the canal gradually exposes the blue line.

When the location of the superior canal is confirmed, an anterior line 60 degrees to the blue line locates the position of the internal auditory canal. The depth of the internal auditory canal varies, but drilling medially near the petrous ridge provides a safe route to the canal. Drilling laterally near the anterior ampulla of the superior semicircular canal places the geniculate ganglion, labyrinthine segment of the facial nerve, and cochlea at risk. When the blue line of the internal auditory canal is exposed, bone is removed in a lateral direction until the vertical crest is found (Fig. 28-8). Bone can now be removed over the geniculate ganglion and tegmen tympani.

FIGURE 28-8 A-D, Middle cranial fossa exposure of CN VII and surrounding anatomy. Bone removed from tegmen tympani to expose tympanic segment of CN VII. Note edema of internal auditory canal segment of CN VII, frequently found in Bell’s palsy. SSC, superior semicircular canal.

The labyrinthine segment of the facial nerve is the narrowest portion of the fallopian canal and lies in an anterosuperior plane from the vertical crest to the geniculate ganglion. A 1 mm diamond burr is used to remove bone over the labyrinthine segment, while the area directly anterior to the segment is closely observed for the blue line of the basal turn of the cochlea. The final layer of bone is removed from the labyrinthine, geniculate, and proximal tympanic segments of the facial nerve with thin, angled hooks and blunt microelevators. At this point, marked swelling of the nerve in the internal auditory canal, labyrinthine segment, geniculate ganglion, and proximal tympanic segments is usually observed. Further swelling is apparent after neurolysis. A disposable microscalpel (Beaver No. 59-10) is used to slit the periosteum and epineural sheath.

Intraoperative EMG is used to localize the region of the nerve conduction block. Direct stimulation of the most distal exposed tympanic segment of the nerve is performed with monopolar or bipolar microforceps (Fig. 28-9). If the conduction block is medial to the point of stimulation, and the nerve is not completely degenerated, a motor unit potential is observed. Stimulating more medially toward the internal auditory canal fails to elicit a motor unit potential if the conduction block is at the meatal foramen.

FIGURE 28-9 Intraoperative evoked electromyography to identify site of nerve conduction block. If conduction block is medial to geniculate ganglions, stimulation of tympanic segment results in motor unit potential (1); stimulation of internal canal segment (3) results in no response. Conduction block usually is at meatal foramen (MF). LSC, lateral semicircular canal; SSC, superior semicircular canal; SVN, superior vestibular nerve.

On completion of the decompression and confirmation of the site of the conduction block, the opened internal auditory canal is covered with a piece of temporalis muscle. The previously harvested temporalis fascia is placed over the opened floor of the middle fossa after bone waxing the opened mastoid air cells. The dural retractor is removed, and the anesthesiologist is asked to re-establish a normal partial pressure of carbon dioxide.

A corner of the craniotomy bone flap is harvested and placed superior to the fascia to prevent herniation of the temporal lobe dura into the attic and internal auditory canal. The temporal lobe is allowed to re-expand into the middle cranial fossa floor. The remainder of the craniotomy flap is placed on the dura, and the temporalis is closed, creating a watertight seal with interrupted absorbable sutures. Skin is closed with a deep layer of interrupted absorbable sutures and an outer layer of nylon absorbable sutures. A mastoid-type dressing is applied. No drains are used.

SUMMARY

Bell’s palsy is a clinical entity encountered with relative frequency by practitioners of many specialties. Before making the diagnosis of Bell’s palsy, other causes of facial paralysis must be excluded through thorough history taking and physical examination. Bell’s palsy and Ramsay Hunt syndrome seem to be due to viral etiologies, and seem to benefit from medical therapy with steroids and antiherpetic medications. Electrodiagnostic testing in cases of Bell’s palsy with complete paralysis is indicated and provides crucial information to guide surgical decision making and patient counseling. In cases of Bell’s palsy with greater than 90% degeneration on ENoG and absence of CMAP on EMG, middle fossa craniotomy with decompression of the labyrinthine, geniculate, and proximal tympanic segments of the nerve should be offered to patients without medical contraindications to an intracranial procedure. Familiarity with the anatomy of the middle cranial fossa is essential for effective decompression of the facial nerve and avoidance of complications.