-------------------------------------------------------------------------------
TITLE:   FACIAL NERVE PARALYSIS 1996
SOURCE:  Dept. of Otolaryngology, UTMB, Grand Rounds
DATE:    March 13, 1996
RESIDENT PHYSICIAN:     Kelly D. Sweeney, M.D.
FACULTY:                Jeffrey T. Vrabec, M.D.
DISCUSSANT:             Kedar K. Adour, M.D.  Sir Charles Bell Society
                         San Francisco, CA
SERIES EDITOR:          Francis B. Quinn, Jr., M.D.
-------------------------------------------------------------------------------

"This material was prepared by resident physicians in partial fulfillment 
of educational requirements established for the Postgraduate Training 
Program of the UTMB Department of Otolaryngology/Head and Neck Surgery 
and was not intended for clinical use in its present form.  It was 
prepared for the purpose of stimulating group discussion in a conference 
setting.  No warranties, either express or implied, are made with respect 
to its accuracy, completeness, or timeliness.  The material does not 
necessarily reflect the current or past opinions of members of the UTMB 
faculty and should not be used for purposes of diagnosis or treatment 
without consulting appropriate literature sources and informed professional 
opinion." 

Anatomy

Acute facial nerve paralysis is a common clinical entity with which all
practicing otolaryngologist should be familiar. In order to diagnose and
treat the many causes of facial nerve paralysis, it is important that the
clinician have a good understanding of the anatomy and function of the
facial nerve.  The facial nerve contains approximately 10,000 fibers.  Of
these, 7000 myelinated fibers innervate the muscles of facial expression,
the stapedius muscle, the postauricular muscles, the posterior belly of
the digastric muscle, and the platysma.  The remaining 3000 fibers form
the nervus intermedius (Nerve of Wrisberg) which contains sensory fibers
(taste) from the anterior 2/3 of the tongue, and parasympathetic
secretomotor fibers to the parotid, submandibular, sublingual, and
lacrimal gland. The facial nerve also contains a few general somatic
afferent fibers which join the auricular branch of the vagus to supply
sensation to the external auditory meatus, and visceral afferents which
innervate the mucous membranes of the nose, palate, and pharynx via the
greater palatine nerve.

The motor nucleus of the facial nerve lies deep within the reticular
formation of the pons where it receives input from the precentral gyrus of
the motor cortex, which innervates the ipsilateral and contralateral
forehead.  The cerebral cortical tracts also innervate the contralateral
portion of the remaining face.  This accounts for the sparing of the
forehead motion in supranuclear lesions of the facial nerve.

The parasympathetic secretory fibers of the nervous intermedius arise from
the superior salivatory nucleus.  These preganglionic fibers travel to the
submandibular ganglion via the chorda tympani nerve to innervate the
submandibular and sublingual glands, and to the sphenopalatine ganglion
via the greater superficial petrosal nerve to innervate the lacrimal,
nasal, and palatine glands.  The secretory fibers of the lesser
superficial petrosal nerve traverse the tympanic plexus, synapse in the
otic ganglion, and travel via the auriculotemporal nerve to innervate the
parotid gland.  Taste fibers from the anterior 2/3 of the tongue reach the
geniculate ganglion via the chorda tympani nerve and from there travel to
the nucleus of the tractus solitarius.

The facial nerve and nervus intermedius exit the brain stem at the
pontomedullary junction and travel laterally 12 - 14 mm with the eight
cranial nerve to enter the internal acoustic meatus.  The meatal segment
of the nerve then travels 8 - 10 mm within the internal auditory canal
(anterosuperior quadrant) to the meatal foramen where the canal narrows
from 1.2 mm in diameter to 0.68 mm in diameter (the narrowest part of the
canal).  The labyrinthine segment then runs 2 - 4 mm to the geniculate
ganglion.  Here the greater superficial petrosal nerve exits to carry
parasympathetic secretomotor fibers to the lacrimal gland.  Just distal to
this branch, the lesser superficial petrosal nerve exits to supply
parasympathetic secretomotor fibers to the parotid.  The tympanic segment
begins just distal to the geniculate ganglion where the nerve turns 40 to
80 degrees (first genu) and runs posteroinferiorly 11 mm across the
tympanic cavity to the second genu.  A branch leaves the segment near the
pyramidal eminence to supply the stapedius muscle.  The nerve then turns
about 90 degrees at the second genu inferiorly where the mastoid segment
travels for 12 - 14 mm inferiorly in the anterior mastoid to exit the
stylomastoid foramen.  The terminal branch of the nervus intermedius, the
chorda tympani, leaves the mastoid segment 5 mm proximal to the foramen
and travels lateral to the incus, medial to the malleus to exit at the
petrotympanic fissure.  The extratympanic segment is composed entirely of
motor fibers and enters the parotid gland after giving off the posterior
auricular branch and a branch to the posterior belly of the digastric
muscle.  The pes anserus forms 20 mm from the stylomastoid foramen and
further divides the nerve into the upper (temporal, and zygomatic) and
lower (buccal, mandibular, and cervical) branches.
                      

Physiology of Nerve Injury

When an axon is injured, biochemical and histological changes occur in the
cell body proximal and distal to the site of the injury.  The severity of
the changes depend upon the distance from the injury to the cell body
(proximal injuries usually more severe than distal injuries), the type of
injury (crush injuries more severe than clean transections), the age of
the patient (older individuals sustain more severe injury than younger
patients), and the nutritional and metabolic status of the patient.

Sunderland's classification of nerve injury describes five degrees of
injury.  The first degree (neuropraxia) involves a localized conduction
block in the nerve with the nerve fibers responding to electrical stimuli
proximal and distal to the lesion, but not across the injured segment.
Axonal continuity is preserved, wallerian degeneration does not occur, and
recovery is usually complete.

The second degree of nerve injury is called axonotmesis.  This refers to
disruption of the axon into proximal and distal portions with interrupted
axoplasmic flow.  Wallerian degeneration occurs within 24 hours in the
distal portion of the axon and to a slight degree in the proximal portion.
The connective tissue elements remain intact, however, and the axon may
regenerate at a rate of 1 mm/day to the original end organ with the
potential for complete recovery.

The third degree of nerve injury refers to endoneurotmesis.  In this type
of nerve injury, the endoneurium and axon are destroyed, but the
perineurium remains intact.  Wallerian degeneration occurs.  Axons may
regenerate, but can be blocked by scar tissue.  This will result in
partial reinnervation.  In addition, misdirection of fibers can occur with
resultant synkinesis (abnormal mass movement of muscles which do not
normally contract together) and incomplete recovery.
   
Fourth degree nerve injury is called perineurotmesis.  In this type, only
the epineurium remains intact, while the axon, endoneurium, and
perineurium are disrupted.  With this type of injury, wallerian
degeneration occurs, and there is much greater chance for aberrant
regeneration, synkinesis, and incomplete recovery.

Fifth degree injury or neurotmesis refers to complete disruption of neural
continuity.  Without careful repair, there is little to no chance of
regeneration and recovery.  In addition, axonal sprouts may escape the
confines of the nerve sheath and produce painful neuromas adjacent to the
injured nerve.  Except in cases of complete transection, nerve injury is
usually a combination of degrees of injury.


Clinical Evaluation

The first step in evaluating any patient who presents with facial nerve
paralysis involves taking a careful and thorough history.  It is important
to determine the onset of the paralysis (sudden vs delayed), the duration,
and the rate of progression.  It is especially important to determine
whether the paralysis is complete verses incomplete.  Patients should be
questioned regarding previous episodes, family history, associated
symptoms (hearing loss, otorrhea, otalgia, vertigo, headaches, blurred
vision, parasthesias), associated medical illnesses (diabetes, pregnancy,
autoimmune disorders, cancer), history of trauma (recent or remote), and
previous surgery (otologic, rhytidectomy, parotidectomy).

A complete head and neck examination must be performed, including
microscopic examination of the ears, careful palpation of the parotid
glands and neck, ophthalmologic examination (r/o papilledema),
auscultation of the neck ( r/o carotid bruits), and a thorough
neurological examination.  It is important to assess the degree of
voluntary movement present in order to document the grade of facial
paralysis as described in the House classification system:

Grade               Degree         Description 

I                   Normal         Normal facial movements. No synkinesis

II                  Slight         Mild deformity, mild synkinesis, good 
                                   forehead function, slight asymmetry.  

III                 Moderate       Obvious facial weakness, forehead motion 
                                   present, good eye closure, asymmetry, 
                                   Bell's phenomenon present. 

IV                  Moderately     Obvious weakness, increasing synkinesis, 
                     Severe        no forehead motion.

 V                  Severe         Very obvious facial paralysis, some tone 
                                   present, cannot close eye.

 VI                 Total          Complete facial paralysis, absent tone. 
                     Paralysis                                            
          
It is also important to determine if the paralysis is central versus
peripheral.  Supranuclear (central) lesions produce contralateral
voluntary lower facial paralysis.  The frontalis muscle is spared because
of the bilateral innervation as described previously.  Emotional response
(facial motion on laughing or crying) may also be preserved with central
lesions.  Presence of Bell's phenomenon (upward outward turning of the
eyeball as the patient attempts to close the eyelids) indicates a
peripheral lesion.

Any patient presenting with facial paralysis should undergo formal
audiological testing, including pure tone, air and bone conduction, speech
discrimination, reflexes, and tympanometry.  If asymmetry is found on the
audiogram, an ABR and/or MRI should be obtained.  Electronystagmography
(ENG) is usually not indicated unless vertigo or other balance disturbance
is part of the clinical picture.

Radiologic evaluation may be undertaken in patients with a history of
recurrent paralysis, associated neurological symptoms, suspected CPA
lesions, concurrent otologic findings (AOM, COM, suspected cholesteatoma),
history of trauma, gradually developing facial nerve paralysis, atypical
presentation, or if patients show no evidence of recovery after one month
from onset.  Gadolinium enhanced MRI is superior for soft tissue
evaluation and will usually reveal the inflammation and edema associated
with Bell's palsy and with Herpes Zoster oticus.  It is also considered to
be the procedure of choice to rule out a cerebellopontine angle tumor or
other brain tumors.  High- resolution computed tomography provides
excellent bony assessment and is the study of choice to rule out a
temporal bone fracture, or to evaluate the middle ear and mastoid.

 
Topognostic Testing
    
The principle behind topognostic testing is that lesions distal to the
site of a particular branch of the facial nerve will spare the function of
that branch.  Moving distally from the brainstem, these tests include: the
schirmer test for lacrimation (GSPN), the stapedial reflex test (stapedial
branch), taste testing (chorda tympani nerve), salivary flow rates and pH
(chorda tympani).

The Schirmer test evaluates the function of the greater superficial
petrosal nerve by determining the rate of lacrimation.  Filter paper is
placed in the lower conjunctival fornix bilaterally.  After 3 - 5 minutes,
the length of the strip that is moist is compared to the normal side.  A
value of 25% or less on the involved side or total lacrimation less than
25 mm is considered abnormal.  An abnormal result can indicate injury to
the GSPN or to the facial nerve proximal to the geniculate ganglion and
may predict patients at risk for exposure keratitis.

Stapedial reflex testing is routinely done during the audiological
evaluation.  This test evaluates the stapedius branch of the facial nerve
which leaves the main trunk just past the second genu in the mastoid.  Of
all the topographic tests, this one is the most objective and
reproducible.  A loud tone is presented to either the ipsilateral or
contralateral ear which should evoke a reflex movement of the stapedius
muscle.  This changes the tension on the TM (which must be intact for a
valid test) resulting in a change in the impedance of the ossicular chain.
If the tone is presented to the opposite ear (normal hearing) and the
reflex is elicited, the seventh nerve is considered to be intact up to
that point.  In the case of Bell's palsy with an intact stapedial reflex,
complete recovery can be expected to begin within six weeks.  Absence of
the reflex when either ear is stimulated with normal VIII nerve function
suggests an abnormality of the facial afferent.  In Bell's palsy, however,
absence of the stapedial reflex during the first two weeks is common and
is usually of no prognostic significance.

Measurement of taste by the anterior 2/3 of the tongue can be done by
placing a small amount of salt, sugar, or lemon juice on the tongue.  Loss
of taste may indicate interruption of the ipsilateral chorda tympani
nerve.  This test is extremely subjective.  A more reliable indicator of
interruption of the chorda tympani nerve involves microscopic detection of
the absence of taste papillae on the involved side of the tongue.
Papillae generally disappear within 10 days post injury.  Examination of
the middle 1/3 of the tongue is most indicative, because the anterior 1/3
may receive bilateral input.

Salivary flow rates can also be assessed to evaluate functional integrity
of the chorda tympani nerve.  This test involves cannulation of Wharton's
ducts bilaterally with measurement of output after five minutes.  A 25%
reduction in flow of the involved side as compared to the normal side is
considered significant.  This test has largely been abandoned secondary to
technical difficulty of cannulating the ducts and patient discomfort.
Salivary ph may be examined as an indirect measure of flow. As the rate of
flow increases, the ph increases.  Therefore, a ph of less than 6.1 may
predict loss of function of the chorda tympani.

Although these tests are of historical interest, they have not been found
to be of much use clinically for determining the site of the lesion in
facial paralysis or for predicting the outcome.  Marked discrepancies are
often seen.  For example, patients may exhibit a marked decrease in
lacrimation with a normal stapedial reflex and intact taste, or they may
have absent lacrimation and an absent stapedial reflex with normal
salivation.  These discrepancies are easily explained in Bell's palsy,
where there can be multiple sites of inflammation and demyelinization from
the brainstem to the peripheral branches of the nerve.  In addition, in
temporal bone fractures, the GSPN and chorda tympani nerves are very
vulnerable to injury and may be disrupted with an intact facial nerve.
Also, with tumors, transmission of nerve impulses can occur through the
tumor mass itself until late in the disease with different areas of the
nerve being affected at different times.  These tests may be helpful,
however, for predicting the likelihood of development of exposure
keratitis.


Electrophysiologic Tests

These tests are useful for patients with complete paralysis for
determining prognosis for return of facial function and the endpoint of
degeneration by serial testing.  They are most useful when considering
decompression surgery and are of no value in patients with incomplete
paralysis.

The nerve excitability test (NET), maximal stimulation test (MST), and
electroneuronography (ENoG) are most useful in the degenerative phase.
These tests will give normal results during the first 72 hours after
injury due to the stimulating and recording electrodes both being distal
to the site of the injury.  After 3 - 4 days, the nerve degeneration
reaches the site of stimulation and useful results will be obtained.
These tests can only be used for unilateral paralysis because all three
involve comparison to the contralateral side which must be normal for
valid results.

The nerve excitability test (NET) is the most commonly used secondary to
the low cost, readily available equipment, and ease of performance.  This
test involves placement of a stimulating electrode over the stylomastoid
foramen.  The lowest current necessary to produce a twitch on the
paralyzed side of the face (threshold) is compared with the contralateral
side.  A difference of greater than 3.5 milliamps indicates a poor
prognosis for return of facial function.  The major draw back to the use
of this test is its subjectivity, with reliance entirely on a visual end
point.  In addition, since such a small amount of current is used with
this test, a few intact axons may give a visible response leading the
clinician to predict a good prognosis, when in reality most of the fibers
are degenerating.

The maximum stimulation test (MST) is a modified version of the NET.  A
maximal stimulus is used to depolarize all facial nerve branches.  The
paralyzed side is then compared to the contralateral side and the
difference is graded as equal, slightly decreased, markedly decreased, or
absent.  Testing begins on the third day post onset and is repeated
periodically until return of facial function or absent response.  An equal
or slightly decreased response on the involved side is considered
favorable for complete recovery.  An absent or markedly decreased response
denotes advanced degeneration with a poor prognosis.  The response to this
test becomes abnormal sooner than the response to the NET and is therefore
considered superior.  However, like the NET, this test is also subjective.


Electroneuronography (ENoG) is considered to be the most accurate
prognostic test because it provides an objective, qualitative measurement
of neural degeneration.  The facial nerve is stimulated with an impulse
transcutaneously at the stylomastoid foramen using bipolar electrodes.
The muscular response is then recorded using bipolar electrodes placed
near the nasolabial groove.  The peak to peak amplitude of the evoked
compound action potential is considered proportional to the number of
intact axons.  The two sides are then compared with the response on the
paralyzed side of the face expressed as a percentage of the response on
the normal side of the face.  A reduction in amplitude on the involved
side to 10% or less of the normal side indicates a poor prognosis for
spontaneous recovery.  A maximal reduction of less than 90% within 3 weeks
of onset gives an expected spontaneous rate of recovery of 80 - 100%.
Disadvantages of ENoG include discomfort, cost, and test-retest
variability which is due to positioning of the electrodes and excitation
of the muscles of mastication (V).  

Electromyography (EMG) is of limited value early in the evaluation of
facial paralysis because fibrillation potentials indicating axonal
degeneration do not appear until 10 to 14 days post onset.  However, EMG
becomes important for assessing reinnervation potential of the muscle two
weeks after onset.  By using needle electrodes placed transcutaneously
into the muscles of facial expression, muscle action potentials generated
by voluntary activity can be recorded.  Electrical silence can indicate
normal muscle in a resting state, severe muscle wasting and fibrosis or
acute facial paralysis in the early stages.  During normal voluntary
contraction organized diphasic or triphasic potentials are seen.
Fibrillation potentials indicate degeneration of the neural supply to the
muscle in question.  Polyphasic potentials indicate reinnervation.  These
are important because they usually appear 6 - 12 weeks before clinical
return of function.

   
Bell's Palsy

Bell's palsy is the most common cause of facial paralysis and accounts for
more than half of all cases.  Traditionally, this was considered to be a
diagnosis of exclusion after ruling out all other possible causes.
However, it has recently been considered a positive diagnosis if the
following are present: unilateral weakness of all facial muscles of sudden
onset, possibly associated with a viral prodrome, no evidence of central
nervous system pathology, no evidence of a CPA lesion, no history of
otologic disease.  Patients may exhibit evidence of concomitant sensory
cranial polyneuritis with otalgia or postauricular pain, dysacousis or
hyperacusis, dysgeusia, decreased tearing or epiphora, and facial
hypesthesias/dysesthesias of V or IX .  Although the exact etiology of
Bell's palsy is still unclear, most clinicians believe that herpes simplex
infection is the most likely agent.  This belief is supported by an
increased incidence of HSV antibodies in patients with Bell's palsy when
compared to age-matched controls.

In 1995, Sugita et al were successful in producing an acute and transient
facial paralysis in mice by inoculating herpes simplex virus into their
auricles (104) or tongues (30).  Facial paralysis developed in the mice
between six and nine days after inoculation, lasted for three to seven
days, and then resolved spontaneously.  Histopathological studies of the
facial nerve and nuclei from these mice revealed severe nerve swelling,
vacuolar degeneration, and infiltration of inflammatory cells.  HSV
antigens were detected in the facial nerve, geniculate ganglion, and the
facial nerve nucleus. They concluded that HSV could produce an acute and
transient facial paralysis through a natural infectious route from the
auricle or tongue to the geniculate ganglion.

Murakami et al (1996) also investigated the role of herpes simplex virus
in the pathogenesis of facial paralysis in mice by inoculating mouse
auricles with HSV.  On the third day following inoculation, HSV DNA was
noted in the ipsilateral facial nerve.  On the tenth day, HSV DNA was
noted in both facial nerves and brain stem in the mice with facial
paralysis, but absent in these tissues in the mice without facial
paralysis.  Between days 4 and 20, the neutralization antibody titer was
elevated in all of the mice.  In addition, facial paralysis developed only
on the inoculated side.  They concluded that HSV infection in the facial
nerve and brain stem must be a prerequisite for the development of facial
paralysis and suggested that an immunologic reaction after a viral
infection plays a role in the pathogenesis.

The incidence of Bell's palsy is estimated to be 20 to 30 per 100,000, but
appears to increase with age.  There is an equal male to female ration and
a 3.3 times greater incidence in pregnant females.  The left and right
sides of the face are equally involved, and less than 1% of cases are
bilateral.  The recurrence rate is about 10% and can be ipsilateral or
bilateral.  Patients with diabetes have 4 - 5 times more risk of
developing the disease.  A family history is positive in about 10% of
patients with Bell's palsy.

The most likely site of lesion in Bell's palsy is the meatal foramen
(junction of the internal auditory canal portion of the nerve and the
labyrinthine segment of the nerve), which is considered to be the
narrowest portion of the fallopian canal.  MRI with gadolinium will
usually show enhancement of the labyrinthine portion of the nerve.  As the
edema within the nerve increases, axonal flow and circulation are
inhibited resulting in varying degrees of nerve injury (first, second, and
third degree).  Patients who are most severely affected develop a high
level of third degree injury which can result in the loss of endoneural
tubules and misdirected axonal regeneration. Histological studies from
patients with Bell's palsy who died of nonrelated causes reveal diffuse
demyelination of the facial nerve with lymphocytic infiltrates.

The prognosis for Bell's palsy is generally good with 85 to 90% of
patients recovering completely within one month.  The remaining 15%
progress to complete degeneration and will not usually show signs of
recovery for three to six months.  The longer the time needed for
recovery, the greater the probability of sequelae.  The single most
important prognostic factor is the degree of paralysis.  Patients with
incomplete paralysis will recover with no sequelae 95% of the time.

The treatment of Bell's palsy is variable, ranging from observation to
surgical decompression.  Regardless of treatment given, all patients must
be counselled regarding proper eye care to prevent exposure keratitis.
Patients should use natural tears liberally during the day and should
place lacrilube ointment in the eye at night.  Taping of the eye lids
during sleep may be helpful as well as the use of a moisture chamber.
Patients should avoid fans and dust, and should consider wearing eye
protection when outside in the wind.

Oral prednisone in a divided dosage of 1 mg/kg/day may be helpful in
preventing or lessening degeneration, decreasing synkinesis, and relieving
pain, and may result in earlier recovery.  Patients should be reevaluated
within five days after starting steroids.  If some function is present
(paresis), taper the steroids over the next five days.  If no improvement
is noted, the full dose should be given for an additional ten days, then
tapered over five days.  Oral acyclovir may help improve recovery in
Bell's palsy.  The usual dosage is 500 mg po four times a day for ten
days.  For patients in whom steroids or acyclovir is contraindicated,
observation and eye care may be all that is possible.

Surgical decompression for Bell's palsy is somewhat controversial.  Most
surgeons agree, however, that in patients progressing to total paralysis
within two weeks, with an ENoG demonstrating 90% or greater degeneration,
decompression of the facial nerve may prevent further degeneration and may
improve outcome.  The rationale behind surgical decompression is based on
the assumption that the site of maximal facial nerve injury in Bell's
palsy is within the meatal foramen.  With increasing edema and decreasing
axoplasmic flow and microcirculation a pathological compression injury of
the nerve occurs at this point of maximal constriction.  This can range
from first degree to third degree.  Removal of the compression, if
performed before irreversible injury to the endoneural tubules occurs (two
weeks), will allow for axonal regeneration to occur.  This is usually
accomplished via a middle fossa approach.  Surgical decompression should
not be done in an only hearing ear.

In a retrospective study, Fisch (1981) compared fourteen patients with
>90% degeneration within 1 to 14 days after the onset of facial paralysis
who underwent decompression using the middle fossa approach to thirteen
similar patients who refused surgical decompression.  A subtle but
statistically significant improvement in long-term facial recovery was
noted in the operative group as compared to the patients who refused
surgery.  Fisch concluded that in order to obtain a satisfactory return of
facial function in all cases of Bell's palsy, surgical decompression of
the facial nerve should be performed within 24 hours when results of ENoG
indicate >90% degeneration has occurred.  

Trauma

Trauma is the second most common cause of facial nerve paralysis.
Longitudinal fractures of the temporal bone make up 80% of all temporal
bone fractures.  In this type, the fracture line extends along the
longitudinal axis of the temporal bone resulting in an external auditory
canal laceration, TM perforation, and possible ossicular disruption or
hemotympanum.  Facial nerve injury occurs in 10 to 20% of these fractures
with the injury most common in the perigeniculate region.  Transverse
fractures are much less common (20% of all temporal bone fractures).
These fractures extend along the transverse axis of the temporal bone
across the vestibule, resulting in sensorineural hearing loss, possible
loss of vestibular function, and a 50% chance of facial nerve injury which
is also usually in the perigeniculate region.

For complete facial nerve paralysis with clinical evidence of a temporal
bone fracture, obtain a high resolution CT scan/temporal bone protocol.
If an obvious fracture is present, surgical exploration of the facial
nerve should be undertaken as soon as possible via either a
transmastoid/translabyrinthine (+ SNHL) or transmastoid/middle fossa
approach (- SNHL). During exploration the nerve must be fully exposed in
order to identify all injured segments, and remove any compression from
fracture fragments.  The nerve sheath should be incised and any hematomas
within the sheath must be carefully evacuated.  If complete transection of
the nerve is found during exploration, a direct end-to-end anastomosis
should be performed if possible. It is important to handle all neural
tissue as atraumatically as possible, using microvascular instruments and
techniques.  The nerve endings should be prepared by sharply cutting at a
ninety degree angle and reapproximating under no tension with two to three
9.0 nylon sutures through the epineurium.  When a direct end-to-end
anastomosis creates tension, or when segments of the nerve are missing or
severely damaged, interpositional grafts from the greater auricular,
medial antebrachial cutaneous, or sural nerve should be used.  

For incomplete facial nerve paralysis or for delayed onset paralysis
associated with a temporal bone fracture, facial nerve testing should be
obtained on day 4 after onset.  If advanced degeneration has occurred, the
nerve should be surgically explored and decompressed. 

Gun shot wounds of the temporal bone cause facial paralysis in over 50% of
cases.  The nerve may be transected, or may be secondarily injured by the
kinetic injury from the bullet or from bony fragmentation of the temporal
bone.  The most common sites of injury are the tympanic and mastoid
segments of the nerve and the stylomastoid foramen area.  For a complete
paralysis after a gun shot wound, surgical exploration and repair should
be undertaken as soon as the patient is medically stable.  Generally,
outcome of facial function is much worse with gun shot wounds to the
temporal bone than with temporal bone fractures.

The facial nerve can also be injured during middle ear and mastoid
surgery.  If iatrogenic transection occurs during surgery, the nerve must
be repaired.  If paralysis occurs postoperatively and the surgeon is
confident that the nerve was intact at the conclusion of the case, a high
resolution CT scan should be obtained and facial nerve testing should be
done on POD #4 - 6.  If advanced degeneration is evident, surgical
exploration and decompression should be done.  If there is any question as
to the integrity of the nerve, exploration should be done as soon as
possible by a surgeon familiar with the intratemporal anatomy of the
facial nerve and reconstructive techniques.

Penetrating facial injuries with immediate facial nerve paralysis should
be explored and repaired as soon as possible while electrical stimulation
can still be used to help locate the distal branches.  When the injury is
medial to the lateral canthus of the eye, aggressive exploration is not
usually mandatory because the nerve endings are small and a rich
anastomotic network exists in this area.


Herpes Zoster Oticus

Herpes zoster oticus (Ramsay-Hunt Syndrome) is the third most common cause
of facial nerve paralysis.  This is a manifestation of a dormant varicella
zoster virus reactivating in extramedullary cranial nerve ganglia during a
period of decreased cell mediated immunity. The prodromal symptoms are
very similar to those seen in Bell's palsy, but are usually more severe.
Symptoms include severe otalgia, facial paralysis, facial numbness, and a
vesicular eruption on the concha, external auditory canal, and palate.
Patients may also have varying degrees of SNHL, vestibular symptoms, and
associated cranial nerve symptoms.  Laboratory evaluation will often show
the rise and fall of antibody titers specific for the virus.  The palsy is
usually more severe and the prognosis much poorer compared to Bell's
palsy.  Only about 50% of these patients regain normal facial function as
opposed to 90% with Bell's palsy.  The degeneration occurs more slowly,
usually over three weeks instead of two.  Treatment includes steroids,
valacyclovir 1 gm po tid, and proper eye care.  Surgical decompression has
not been proven beneficial but may be considered for ENoG showing greater
than 90% degeneration.


Otitis Media

In patients with evidence of acute otitis media, dehiscences in the
fallopian canal may serve as portals for direct bacterial invasion and
inflammation along the nerve.  Facial paralysis may begin within a few
days of onset of an acute otitis media and is usually incomplete.
Treatment includes a wide myringotomy, drainage, and culture with
antibiotic coverage for gram positive cocci and H. flu.  The facial palsy
associated with acute otitis media generally resolves with aggressive
management of the infection.  However, if a total paralysis is present,
serial ENoG should be obtained.  If axonal degeneration reaches > 90%,
surgical exploration and decompression should be performed.

Patients with chronic otitis media may also develop facial paralysis which
is usually secondary to cholesteatoma or from inflammation/osteitis
compressing the facial nerve.  In these cases a high resolution CT should
be obtained, and surgery should be performed as soon as possible
(tympanomastoidectomy, facial nerve exploration and decompression). 


Tumors

About 5% of cases of facial nerve paralysis are caused by tumors.  Factors
which should increase the clinicians suspicion of a possible tumor
include: a slowly developing paresis over more than three weeks, facial
twitching, additional cranial nerve deficits, recurrent ipsilateral
involvement, associated adenopathy, or a palpable neck or parotid mass.
In these cases, MRI and CT should be obtained.  The most common benign
tumor causing facial nerve paralysis is a facial nerve schwanomma.  The
most common malignant tumors causing facial paralysis are mucoepidermoid
carcinoma and adenoid cystic carcinoma of the parotid gland.  The
management of facial nerve paralysis caused by tumors depends upon the
lesions location, size, and malignant potential and may include
transposition of the nerve, division and reanastomosis, interposition
grafting, and cranial nerve crossover.  After nerve grafts, the best
possible result that can be expected is a House grade III paresis.


Melkerson-Rosenthal Syndrome

Melkerson-Rosenthal syndrome is a rare disease that consists of a triad of
symptoms: recurrent orofacial edema, recurrent facial palsy, and lingua
plicata (fissured tongue).  The defining feature of this disease is the
persistent or recurrent nonpitting facial edema that cannot be explained
by infection, malignancy, or connective tissue disorder.  It usually
involves the lips and buccal area and may involve the gingiva, palate, and
tongue.  The lips become chapped, fissured, and red-brown in appearance
and may develop permanent deformity after numerous recurrences.  Facial
paralysis and lingua plicata occur in half of the patients.  The complete
triad is only present in 25% of these patients.  This condition usually
starts in the second decade of life, and the manifestations usually occur
sequentially (rarely simultaneously).  The etiology of this syndrome is
unknown.  Some consider it a variant of sarcoidosis secondary to biopsies
of the lips which may reveal noncaseating granulomas.  Facial paralysis
occurs in 50 to 90% of these patients and tends to be abrupt.  A history
of bilateral sequential paralysis and relapse after initial recovery is
common.  The site of the paralysis usually corresponds to the facial
swelling.  Facial nerve decompression may be indicated if episodes of
facial paralysis are frequent and progressive.


Congenital Facial Paralysis

Newborn facial paralysis is estimated to have an incidence of about 0.23%
of live births.  The most common cause is birth trauma (80%), which is
usually evident by other signs of injury.  These include a history of a
difficult delivery with or without forceps, facial swelling, bruising over
the mastoid or extratemporal course of the nerve, and hemotympanum.  The
mechanism of injury is thought to be from direct pressure during forceps
use or from pressure on the infants face by the sacral prominence during
delivery.

Congenital causes of newborn facial paralysis are much less common.
Mobius' syndrome consists of a broad spectrum of clinical findings which
can range from an isolated unilateral facial paralysis to bilateral
absence of facial and abducens nerve function.  In this syndrome, the
facial nerve forms but consists of only a fibrotic tract.  The muscles of
facial expression may form in some cases, but degeneration to fibrosis
generally occurs rapidly.  These children will have no response on EMG at
birth and have no chance of spontaneous recovery of facial function.  Many
other cranial nerves may be involved (III, IV, IX, X, XII) and skeletal
abnormalities may be present.  Treatment for these causes of newborn
facial paralysis is generally delayed until late childhood and usually
requires static slings and muscle transfers.

A rare cause of isolated newborn facial paralysis is dysgenesis of the
intratemporal facial nerve.  CT and surgical findings show that the most
common site of the lesion is the distal part of the mastoid segment in the
fallopian canal.  The nerve is usually very thin and fibrotic at this
area.  EMG findings in these cases usually show a few functional motor
units but useful facial function is not usually present.

Newborns who present with a complete facial nerve paralysis should undergo
electrical testing within the first three days of life to differentiate
between congenital and traumatic causes.  After birth trauma, the nerve
can be stimulated for up to five days post injury and fibrillation
potentials will be seen on EMG at ten to fourteen days.  In congenital
cases, the nerve will usually not stimulate and no fibrillation potentials
will be seen on EMG. The prognosis for trauma related facial nerve
paralysis at birth is usually excellent.  Surgical decompression should
not be considered until the nerve has had a chance to recover or until
>90% degeneration has occurred.


Other Causes

Sarcoidosis consists of a chronic noncaseating granulomatous disease
usually characterized by hilar or peripheral adenopathy, polyarthralgias,
anergy, elevated serum calcium, and hepatic dysfunction.  One variant of
sarcoidosis, Heerfordt's disease (a.k.a. uveoparotid fever) consists of
uveitis, mild fever, nonsuppurative parotitis, and cranial nerve
paralysis.  The facial nerve is the most commonly involved cranial nerve
and the paralysis usually starts abruptly days to months after the
parotitis.  In these cases, the paralysis is considered to be from direct
invasion of the nerve by the granulomatous process.  An elevated serum
angiotensin-converting enzyme (ACE) generally confirms the diagnosis, and
treatment consists of steroids.

Lyme disease is an infection caused by the tick-borne spirochete Borrelia
burgdorferi.  Several species of Ixodes ticks carry the spirochete, and
the primary reservoir is the white-tailed deer and white-footed mouse.
The disease generally occurs in several stages.  The first stage consists
of a flu-like illness with regional lymphadenopathy, general malaise, and
erythema migrans (erythematous enlarging annular skin lesions found
anywhere on the body).  The second stage usually starts several weeks to
months later with neurologic abnormalities.  These include meningitis and
cranial and peripheral nerve neuropathies.  The final stage occurs months
to years later in the form of recurrent meningitis, subtle mental
disorders or neurologic deficits, and chronic arthritis.  Ten percent of
these patients develop facial paralysis (unilateral or bilateral) and
hearing loss.  The facial paralysis typically resolves completely, but may
rarely result in mild permanent facial weakness.  Treatment for Lyme
disease consists of IV ceftriaxone (2 g/day) for fourteen days.  


Conclusion

Although Bell's palsy remains the most common cause of acute facial nerve
paralysis, it is important for clinicians to consider all causes to avoid
overlooking potentially life-threatening diseases.  A good history and
physical is mandatory in the work-up of these patients and will usually
help to significantly narrow the possible causes.  Although topognostic
testing is of historical interest, it has not proven to be of much use
clinically.  ENoG is the most useful electrophysiologic test and should be
performed within 4 - 6 days post-onset in patients with a complete
paralysis with surgical decompression considered for > 90% degeneration.
Treatment plans should be individualized in each patient, but must include
education on eye protection.
--------------------------------------------------------------------------

Editor's Comment:

With one exception, the article by Murakami stands as the most recent
and perhaps most significant contribution to the understanding of
"idiopathic" facial nerve paralysis.  The exception is the section in
Conn's Current Therapy - 1996 on Acute Facial Paralysis, by Kedar Adour.
Dr. Adour discusses the etiology and natural history of the disorder
with a clarity rarely matched in our literature.  He proposes a diagnostic
protocol which discusses manifestations which do NOT permit the diagnosis
of Bell's palsy and Herpes zoster paralysis ("exclusion criteria.")
He proposes a compellingly rational plan of treatment, in which he suggests
that "electrotherpy is of no benefit and may be harmful."

The faculty have invited Dr. Adour to present a Discussion of this Grand 
Rounds topic, and we are honored that he has agreed to do so.  
-----------------------------------------------------------------------------

Discussion by Kedar K. Adour, M.D. President, Sir Charles Bell Society:

Recent published articles have changed the thinking regarding diagnosis and
treatment of acute facial paralysis:

Bell Palsy and Herpes Simplex Virus: Identification of Viral DNA in
Endoneurial Fluid and Muscle

Shingo Murakami, MD; Mutsuhiko Mizobucbi, MD; Yuki Nakashiro, MD; Takashi
Doi, MD; Naohito Hato, MD; and Naoaki Yanagihara, MD 

Objective:
To determine whether herpes simplex virus type 1 (HSV-1) causes Bell palsy. 

Design: Prospective study.
Setting: University inpatient service.

Patients: 14 patients with Bell palsy, 9 patients with the Ramsay-Hunt
syndrome, and 12 other controls. 

Measurements:
Viral genomes of HSV-1, varicella-zoster virus, and Epstein-Barr virus were
analyzed in clinical samples of facial nerve endoneurial fluid and posterior
auricular muscle using polymerase chain reaction (PCR) followed by
hybridization with Southern blot analysis. 

Results:
Herpes simplex virus type 1 genomes were detected in 11 of 14 patients (79%)
with Bell palsy but not in patients with the Ramsay-Hunt syndrome or in other
controls. The nucleotide sequences of the PCR fragments were identical to
those of the HSV-1 genome. 

Conclusions:
Herpes simplex virus type 1 is the major etiologic agent in Bell palsy. 

                 ------------------------------
The editorial comment: "One might now question whether we should continue
using the term "Bell's palsy" to mean "idiopathic facial paralysis" or
whether we should now recognize Bell's palsy as "herpetic facial paralysis,"
confirming the hypothesis advanced initially in 1972 by McCormick. 
It is also documented that Adour in 1970 suggested to the American Academy
of Otolaryngology that herpes simplex was the cause of Bell's palsy." 
                 -------------------------------

Just as the pyramids of Egypt have guarded the riddle of the sphinx, the
petrous pyramid of the temporal bone has guarded the secrets of the facial
nerve. Even today there is question about the anatomic compartments of the
facial nerve. The authors of the above text suggest that the facial nerve
carries some parasympathetic secretomotor fibers to the parotid gland and a
few general somatic afferent fibers to the external ear. These two
suggestions are questionable and are omitted in many neuro-anatomical
textbooks. Cutting the facial nerve at the brainstem does not lead to any
deficit of parotid secretion nor somatosensory loss in the ear canal. 
Sunderland's classification of nerve injury is not applicable to herpetic
facial paralysis (herpes simplex or herpes zoster). Further, Sir Sydney
Sunderland has not written a single word about the facial nerve in his
classic textbook. The five degrees of injury are of increasing severity
according to the effect of the injury not the cause. The statement "The first
degree (neuropraxia) involves a localized conduction block in the nerve with
the nerve fibers responding to electrical stimuli proximal and distal to the
lesion, but not across the injured segment." is incorrect. Segmental
demyelination also causes a neuropraxic state. The terms neuropraxia,
axonotmesis, and neurotmesis should not be equated with the 5 degrees of
nerve injury. 

Under clinical evaluation: The House Grading System cannot be used to
document the grade of facial paralysis at the onset of a paralysis. The House
system is an "end-stage" interpretation. Grade II includes "mild synkinesis",
Grade III and IV "increasing synkinesis". Synkinesis occurs when the nerve
has been damaged and then abnormally regenerated. It is not present in the
acute stages. The Brackmann modification somewhat addresses this problem and
is a modification of the original Adour-Swanson Recovery Profile. 
"Bell's phenomenon (upward outward tuning of the eyeball as the patient
attempts to close the eyelids) indicates a peripheral lesion." is not
correct. The eyeball motion described is a normal function of eye muscle
activity which only became  apparent to Sir Charles Bell because the eyelids
did not close. A central lesion with incomplete closure of the eyelids will
also display a "Bell's phenomenon." 

We agree that topognostic testing is of historical interest only and is
invalid in herpetic facial paralysis, or in temporal bone fractures. 
The presenting symptom of dysgeusia is diagnostic for herpetic facial
paralysis. No further diagnostic tests need be done. One hundred per cent of
patients with herpetic facial paralysis demonstrate papillitis of the
fungiform tongue papillae and has led to the truism that " Bell's palsy is a
tongue blade diagnosis." Taste tests using electrogustometry are research
tools only. With regards to the stapedial reflex, it should be considered as
the "otologists' electromyography". (See below) 
When should you question the diagnosis of herpetic facial paralysis formerly
called Bell's palsy and Ramsay Hunt syndrome?? 

1. Partial facial paralysis that does not resolve in 3 to 6 WEEKS. 

2. Partial facial paralysis with electrical evidence of nerve degeneration.
     a.  MST in one or more branches with moderate (score of 2) to severe 
         (scoreof 1) decrease in muscle motion.
     b.  Global MST less than 3.
     c.  Electromyography with fibrillation (denervation) potentials.  

3.  Partial facial paralysis with ipsilateral hearing loss.  

4.  Any facial paralysis with evidence of chronic otitis media, or history of
    previous ear surgery.  

5.  Progressive facial paralysis over period of weeks or months is not Bell's
    palsy.  Note: Damage to the nerve is complete by day 14 in Bell's
    palsy but not until day 21 in Ramsay Hunt syndrome.  

6.  Total facial paralysis with no regeneration in 4 months.  Note:
    Even with total loss of nerve excitability by ENOG or Global MST 99.9% of
    all Bell's palsy or Ramsay Hunt Syndrome will regenerate with midface 
    contracture with synkinesis. The earliest sign of regeneration is the 
    return of the stapedial reflex. The next earliest sign is evidence of
    mouth motion with voluntary forced closure of the eyes (synkinesis).  

7.  Total facial paralysis with electrical evidence of nerve degeneration
    that resolves WITHOUT midface contracture with synkinesis.  Note:
    Facial nerve degeneration with regeneration in Bell's Palsy or Ramsay 
    Hunt syndrome is followed by the sequelae of midface contracture with
    synkinesis.  

8.  Recurrent facial paralysis on the same side with electrical evidence of
    nerve degeneration and recovery WITHOUT contracture with synkinesis in 4
    months.  

9.  Stapedial Reflex: The Otologists' Electromyography. Total facial
    paralysis with intact stapedial reflex is a Partial paralysis! An absent
    stapedial reflex one week post onset with return of the stapedial reflex 
    by 3 weeks post onset indicates a good prognosis. The Global MST score 
    will allow you to predict rate and degree of recovery.  

10. When reviewing Bell's palsy treatment results be suspicious of any
    article when the number of patients showing electrical evidence of nerve
    degeneration does not equal the number of patients with sequelae of
    contracture with synkinesis.  

Electrical stimulation continues to be widely used in the treatment of
facial paralysis (20, 23) although there is mounting evidence that it may be
contraindicated. It has been suggested that electrical stimulation may
interfere with neural regeneration post peripheral nerve injury (24,25) and
studies proving its efficacy with facial muscles are lacking in the
literature. A 1984 report by the National Center for Health Services Research
concluded that "Electrotherapy treatment for Bell's palsy. . . has no
demonstrable beneficial effect in enhancing the functional or cosmetic
outcomes in patients with Bell's palsy." (26) 

Additionally, patients who undergo electrical stimulation acutely may
demonstrate more synkinesis and mass action than those who do not (27). It is
difficult to produce an isolated contraction of the facial muscles using
electrical stimulation due to their small size and close proximity to each
other. The contraction produced causes mass action which reinforces abnormal
motor patterns and can be painful. (20) 

20. Cole J, Zimmerman S, Spector G: Nonsurgical neuromuscular rehabilitation
of facial muscle paresis, in Rubin LR (ed): The Paralyzed Face. St Louis,
Mosby-Year Book Inc., 1991, pp 107-112. 

23. Farragher DJ: Electrical stimulation: A method of treatment for facial
paralysis, in Rose FC, Jones R, Vrbova G (eds): Neuromuscular Stimulation:
Basic Concepts and Clinical Implication. New York, Demos, 1989 vol 3, pp
303-306.

24. Cohan CS, Kater SB: Suppression of neurite elongation and growth cone
motility by electrical activity. Science 1986; 22:1638-1640. 25. Brown MC,
Holland RL: A central role for denervated tissues in causing nerve
sprouting. Nature 1979; 282:724.

26. Waxman B: Electrotherapy for treatment of Facial Nerve Paralysis (Bell's
Palsy). Health Technology Assessment Reports, National Center for Health
Services Research 1984; 3:27. 

Added Reference: 

Jansen JKS, Lomo T, Nicolaysen K, et al: Hyperinnervation of skeletal muscle
fibers: Dependence on muscle activity. Science 181:559-561, 1973. 

Lack of muscle spindles....

Brodal A: Neurological Anatomy: In Relation to Clinical Medicine, ed 3. New
York, Oxford University Press, 1981, pp 495-508. 

Dubner R, Seddie BJ, Storey AT: The Neural Basis of Oral and Facial Function.
New York, Plenum Press, 1978, pp 222-229. 

March 27, 1996
Kedar K. Adour, MD
President, Sir Charles Bell Society
1000 Green Street #1203
San Francisco, CA 94133
(415) 474-4046   Fax: (415) 474-3541 E-mail: kadour@aol.com 

-----------------------------------------------------------------------------
                          BIBLIOGRAPHY 

Bailey, Byron J., ed.  Head and Neck Surgery - Otolaryngology.
Philadelphia, PA.: J.B. Lippincott Co., 1993.

Ballenger, John J., ed.  Diseases of the Nose, Throat, Ear, Head, and
Neck.  Philadelphia, PA: Lea & Febiger, 1991.

Boies, Lawrence R., ed.  Fundamentals of Otolaryngology.  Philadelphia,
PA: J.B.  Lippincott Co., 1989.

Cummings, Charles, ed.  Otolaryngology - Head and Neck Surgery.  St.
Louis, MO.: Mosby - Year Book, Inc., 1993.

Fisch U.  Maximal Nerve Excitability Testing vs Electroneuronography.
Arch Otolaryngol 106: 352 - 357, 1980.

Fisch U.  Surgery for Bell's Palsy.  Arch Otolaryngol 107: 1 - 11, 1981.

Gates, George A.  Current Therapy in Otolaryngology - Head and Neck
Surgery.  St.  Louis, MO.: Mosby - Year Book, Inc., 1994.

Green JD, et al.  Surgical Management of Iatrogenic Facial Nerve Injuries.
Otolaryngol Head Neck Surg 111(5): 606 - 610, 1994.

Mattox, Douglas E., ed.  The Otolaryngology Clinics of North America:
Management of Facial Nerve Disorders.  Philadelphia, PA: W. B. Saunders,
Co., 1991.

May, Mark.  The Facial Nerve.  New York: Thieme - Stratton, 1986.

May M., et al.  Natural History of Bell's Palsy: The Salivary Flow Test
and Other Prognostic Indicators.  Laryngoscope 86: 704 - 712, 1976.

Murakami S, et al.  Role Of Herpes Simplex Virus Infection in the
Pathogenesis of Facial Paralysis in Mice.  Ann Otol Rhinol Laryngol 105:
49 - 53, 1996.

Peitersen E.  The Natural History of Bell's Palsy.  Am J. Otol 4: 107,
1982.

Rubin, Leonard R., The Paralyzed Face.  St. Louis, MO.: Mosby - Year Book,
Inc., 1991.  

Selesnick SH, Patwardhan A.  Acute Facial Paralysis: Evaluation and Early
Management.  Am J Otolaryngol 15 (6): 387 - 408, 1994.
  
Sugita T, et al.  Facial Nerve Paralysis Induced by Herpes Simplex Virus
in Mice: An Animal Model of Acute and Transient Facial Paralysis.  Ann
Otol Rhinol Laryngol 104:574 - 581, 1995.
--------------------------------END----------------------------------------