J Neurol Neurosurg Psychiatry: first published as 10.1136/jnnp.40.4.342 on 1 April 1977. Downloaded from http://jnnp.bmj.com/ on February 10, 2020 by guest.

Journal ofNeurology, Neurosurgery, and Psychiatry, 1977, 40, 342-350

Delayed facial palsy after head injury

K. PUVANENDRAN, M. VITHARANA, AND P. K. WONG
From the University Department of Medicine, and the Department of Otorhinolaryngology,
Singapore General Hospital, Singapore

SUMMARY Where facial palsy follows head injury after many days, the mechanism is not clear, and
there has been no detailed study on this condition. In this prospective study, an attempt is made to
estimate this complication of head injury, and to study its pathogenesis, natural history, prognosis,
and sequelae which differ markedly from Bell's palsy. It has a much worse prognosis and so surgical
decompression should be considered early in this condition.

The facial nerve is the motor cranial nerve which is studies, for prediction of prognosis at a time when
most commonly affected in closed head injuries surgical intervention seems most advantageous.
(Turner, 1943). In facial palsy which immediately
follows a head injury, the mechanism is obvious, but Patients and methods

Protected by copyright.
it is not clear when the facial palsy follows the head
injury after many days (Potter and Braakman, 1976). During the period May 1974-April 1975, there were
Traumatic facial palsy has received much attention 6304 cases of head injury admitted to government
but few authors distinguish between immediate and hospitals in Singapore. The chief criterion for
delayed palsy. admission to hospital was the occurrence of traumatic
Turner (1944) studied a selected group of war-time amnesia or unconsciousness, indicating concussion
head injuries from a military hospital for head of the brain. Knowing that most post-traumatic
injuries, and found an incidence of 2.2 % developing facial palsies were associated with bleeding ears
a delayed facial palsy. Potter (1964), however, (Turner, 1944; Briggs and Potter, 1967), all cases of
estimated the incidence of this complication of head bleeding ear after such closed head injury were
injury as about 0.6 %. Both authors have cautioned studied from the day of head injury to the evolution
about the difficulty in detecting an immediate facial of the delayed facial palsy in some. Cases of bleeding
palsy and which may result in wrong classification as ear referred for study were those where the bleeding
delayed facial palsy. Turner's estimate of this was thought to come from the middle or inner ear.
condition would appear to be high for this reason These cases were seen initially daily, and examined
and because of the selected type of cases studied. clinically and by electrodiagnostic methods. This
We studied these cases in detail clinically and took gave us an accurate indication of the exact time of
advantage of electrodiagnostic methods (Puvanen- onset of the facial palsy, and allayed arguments as
dran et al., 1977). Grove (1939), Turner (1944), and to whether these were in fact cases of immediate
Miehlke (1973b) state that 80-90% of delayed facial palsy which was not obvious because of the
facial palsies make a spontaneous and complete obtunded state of the patient immediately after the
recovery. Our study shows results which are unlike head injury.
these findings. We have tried to explain the patho- Every case of bleeding ear with facial palsy had
genesis of this condition, and also reviewed other radiographs of the skull, including special views to
hypotheses on the causation of this palsy. show the petrous temporal bones on both sides.
There is much disagreement about the precise Where no fracture was seen tomograms were
place for surgical exploration of the injured nerve prepared.
because of lack of knowledge of its natural history Cochlear, vestibular, and middle-ear functions
and the lack of any method, including electrical were also studied in these cases in detail because of
their close anatomical relation to the 7th nerve.
Address for reprint requests: Dr K. Puvanendran, University Depart- All patients had a detailed neurological examina-
ment of Medicine (1), Singapore General Hospital, Singapore 3,
Republic of Singapore. tion with special regard to other cranial nerve palsy.
Accepted 4 November 1976 The site of the lesion of the facial nerve was
342
 J Neurol Neurosurg Psychiatry: first published as 10.1136/jnnp.40.4.342 on 1 April 1977. Downloaded from http://jnnp.bmj.com/ on February 10, 2020 by guest.
Delayedfacial palsy after head injury 343
ascertained by testing the function of its branches, Table 1 Delayedfacial palsy
such as lacrimation by the Shirmer test (Miehlke,
1973a), and taste function from the anterior two- Males Females Total
12 4 16
thirds of the homolateral side of the tongue, in Side ofPalsy Right Left (Bilateral) Total
addition to the facial nerve proper. All patients had 9 10 (3) 19
a complete or partial paralysis of muscles of Type of Palsy Clinical
17
Latent
2
Total
19
expression of the whole of one side of the face. The Clinical palsy Complete Partial Total
14 3 17
degree of paralysis was estimated visually as a Site of lesion in Geniculate Distal to Total
percentage of the normal side (Taverner, 1955). This clinical palsy ganglion -chorda
method was fairly consistent in studying the progress tympani
10 7 17
of the palsy, but we often supplemented this with
photographs which made comparison easy. By
comparison with the opposite side, the weakness was
estimated as 25o%, 50%, 75%, 1000%, using the developed after a long delay, after the pain of the
frontalis, orbicularis oculi, and orbicularis oris head injury had disappeared.
muscles, and the average of these estimations was
taken. We studied the orbicularis oculi, measuring SITE OF IMPACT
the palpebral fissure and, on recovery of power, This was judged from the history and from external
estimating how much more of the eyelash jutted out injuries. In nine cases the site of impact was around
on attempting tight closure of the eye. We defined as the ear, in two it was over the mastoid bone, and in
subclinical or latent palsy those cases with abnormal three over the occipital area. In each of these cases it
electrical tests but without clinically evident palsy. was on the side of the facial palsy which also
All cases were followed up to study the progression corresponded with the side of the bleeding ear. In

Protected by copyright.
of the facial palsy, the recovery, the development of the three cases who developed bilateral palsy, the site
sequelae such as abnormal movement, the cosmetic of impact was only over one ear.
result, and the patient's satisfaction with the recovery.
SITE OF LESION
Results In 10 cases the site of the lesion was found to be at
the geniculate ganglion, and in these cases secretion
Of the 6304 cases of head injuries admitted to the of tears, which is innervated by the greater superficial
various hospitals, 39 had bleeding ears. Sixteen of petrosal nerve, was lost on the side of the lesion,
these patients developed delayed facial palsy while with loss of taste sensation in the anterior two-thirds
three were noted to have an immediate facial palsy at of that side of the tongue which is innervated by the
time of injury. Three of these 16 patients had bilateral chorda tympani nerve. In seven cases the taste
facial palsies, two of the latter each having a sensation and lacrymation were intact, and the lesion
subclinical palsy as defined above. The incidence of was considered to be distal to the chorda tympani
bleeding ears in these patients with closed head branch.
injuries was 0.6 % and the incidence of delayed
facial palsy 0.3 %. The chance of a case of bleeding ONSET
ear developing a delayed facial palsy after head The delay in onset of facial palsy after head injury
injury was 49 %. There was no case of delayed facial varied from two to 21 days (Fig. 1). In most cases
palsy who did not have a bleeding ear after head the facial palsy came on between two and seven
injury, except one patient who had bilateral facial days. In two instances the onset of delayed facial
palsy but with bleeding from the right ear only. palsy was indicated by a sudden abnormal prolonga-
There were 12 males and four females with ages tion in the terminal latency of facial muscle evoked
anging from 9 to 58 years (Table 1). Eleven patients response, previously shown to be normal in the
were involved in traffic accidents while five had serial electrodiagnostic tests. In both instances, the
fallen. In 14 cases the facial palsy was complete, in subclinical facial palsy occurred 13 to 14 days after
three it was partial, while two patients had sub- the onset of a clinically obvious facial palsy on the
clinical palsy. One of the cases of complete facial opposite side, a delay of 16 to 19 days after head
palsy was partial at first, and it took six days before injury. The cases of conduction block had a delayed
it became complete. onset of seven to nine days. In all five cases of
None of these patients complained of any pain in complete denervation, who showed the most severe
the appropriate ear apart from pain at the site of facial palsy, the delay in onset of the facial palsy
impact before the onset of facial palsy. This was was between two and six days. The cases of partial
looked for especially in cases where facial palsy denervation had a wider scatter in the day of onset
 J Neurol Neurosurg Psychiatry: first published as 10.1136/jnnp.40.4.342 on 1 April 1977. Downloaded from http://jnnp.bmj.com/ on February 10, 2020 by guest.
344 K. Puvanendran, M. Vitharana, and P. K. Wong

LA DISTRIBUTIONOF CASES
w 5- OF COMPLETE DENERVATIDN

4
Ui.
0
w

zEm 321

O
2 4 7 1o 13 16 19 21
DAYS AFTER HEAD INJURY

Fig. 1 Delayedfacial palsy-day of onset after head

injury.

of the palsy, between five and 21 days. In one patient

with obvious facial weakness of both sides (Fig.
2a-c), the second facial weakness appeared nine days
after the first and was less severe, the first palsy

Protected by copyright.
showing complete denervation.
Eight of the 16 patients with delayed facial palsy
showed a definite fracture of the petrous bone on the
side of the facial palsy (Table 2). Seven cases had a
longitudinal fracture while only one was transverse.
In two cases with longitudinal fracture the tomogram
showed a fracture of the external auditory canal on
the side of the facial palsy. In the other eight patients
no fracture could be demonstrated despite special
petrous radiographs and tomography. This finding
contrasted with the three immediate facial palsies
where each patient showed a definite petrous bone
fracture, two being transverse fractures and one
longitudinal.
The 20 cases of ear bleeding without a facial palsy
did not show any obvious fracture of the petrous
bone. No tomograms were done on these cases.
OUTCOME OF THE FACIAL PARALYSIS
This could be divided into three patterns depending
on the degree of improvement and on the electrical
reactions (Table 3): (1) conduction block, (2) com-
plete denervation, and (3) partial denervation. In
conduction block, recovery of facial movements was
complete and there were no sequelae. In denervation,
the facial movements did not recover completely
and these patients were left with sequelae. Electrical
reactions used to divide these patients into the three
groups are described by Puvanendran et al. (1977).
Conduction block occurred in five examples of
facial palsy, of which there were three clinical palsies
and two subclinical. The other 14 facial palsies
developed denervation, six complete and eight partial.
Of the three clinical palsies with conduction block,
two had only a partial palsy and the other a complete
 J Neurol Neurosurg Psychiatry: first published as 10.1136/jnnp.40.4.342 on 1 April 1977. Downloaded from http://jnnp.bmj.com/ on February 10, 2020 by guest.
Delayedfacial palsy after head injury 345
Table 3 Outcome of delayedfacialparalysis-conduction
block 17.6 %, denervation 82.4%
Conduction Complete Partial
block denervation denervation Total
5 6 8 19
Clinical palsy 3 6 8 17
Subclinical palsy 2 0 0 2
Partial palsy 2 0 1 3
Complete palsy 1 6 7 14
Site
Distal to chorda
tympani 3 1 3 7
Geniculate ganglion 0 5 5 10

paralysis, the site of the lesion in all three being

distal to the chorda tympani branch. All six facial
palsies which developed complete denervation had
a complete paralysis clinically, the site of the lesion
being at the geniculate ganglion in all except one
where the lesion was distal to the chorda tympani
branch. In five cases of partial denervation the site
of the lesion was at the geniculate ganglion, and in
the other three cases it was distal to the chorda

Protected by copyright.
tympani branch of the facial nerve. Only one case of
the eight with partial denervation electrically showed
a partial facial palsy clinically. The seven others
showed a complete facial palsy on clinical testing.
Thus conduction block occurred in 17.6 % of patients
and denervation in 82.4 % of patients.
CLINICAL PROGRESS
(c) In cases with a conduction block clinical recovery of
facial weakness started by about the fifth day (Fig. 3),
Fig. 2 Case of bilateralfacialpalsy after head injury. and was complete by the 36th day (mean). In the
(a) Right sidedfacial weakness came on two days after two examples of subclinical palsy, the terminal
head injury. Left side showed normal power. (b) Left
sided weakness came on nine days after right. (c) By the motor latencies of facial muscles returned to normal
eighth week the left face recovered almost complete by the 22nd day and the 56th day after onset
power leaving a complete right sidedpalsy with the mouth respectively.
pulled to the left side. The mean time for the onset of recovery in partial
denervation (assessed electrically) was 13 days and
complete recovery occurred by the 95th day.

Table 2 Delayedfacial palsy-petrous fracture- 100141

bleeding ear - 100%
'
751,. .x
5 Days
Npa imber of
pat,tients 0
0
50'1.
x-'
'!~~~~~ ~,
Fracture only 0 25%.
Bleeding ear only 20 < I
Fracture and bleeding ear 8 I I90 Days
1 30 t 40
Paralysis without bleeding or fracture
Patients with delayed facial palsy 16
DAYS 1i t
PARTIAL
20
COMPLETE
50 60
Patients with ear bleeds 39 DENE RVAT ION DENERVATION
Description of petrous fracture CONDUCTION
BLOCK
longitudinal 7
transverse I
Fig. 3 Clinicalprogress in delayedfacialpalsy.
 J Neurol Neurosurg Psychiatry: first published as 10.1136/jnnp.40.4.342 on 1 April 1977. Downloaded from http://jnnp.bmj.com/ on February 10, 2020 by guest.
346 K. Puvanendran, M. Vitharana, and P. K. Wong
Where denervation was assessed as complete, movement of one part of the face on voluntary
clinical recovery began on the 35th day (mean). By movement of another part. A common example is
the 90th day (mean) only a 25 % recovery was seen. dimpling of the chin or retraction of the angle of the
None of these cases have so far shown a complete mouth on the side of the palsy when closing the eyes
recovery of the facial weakness. Cases 2 and 3 have
not shown more than a 75% recovery by 212 days
and 248 days respectively. Case 1, seen on the 270th
day after injury, has not shown any improvement
beyond 25 %.
It was easy to detect initial improvement because
the first facial muscle to show improvement was the
orbicularis oculi muscle, and this is easily estimated
by measuring the palpebral gap. It was very constant
for recovery to start in that muscle in all cases of
complete denervation. It is a long time before the
orbicularis oris and frontalis muscles start to show
improvement. In most cases with partial denervation
recovery starts in the orbicularis oculi but quickly
follows in the other facial muscles. In most cases of
partial denervation, orbicularis oris was the last
muscle to recover whereas after complete denervation
the frontalis was the last to recover.
Even though all cases of partial denervation

Protected by copyright.
ultimalely achieve normal voluntary power of the
facial muscles, the spontaneous blink on the side of
the palsy remains sluggish for a long time, and may
even be permanently slow.
SEQUELAE
The sequelae of the delayed facial palsy noted were:
(1) incomplete return of voluntary movement of
facial muscles, (2) asymmetry of the face on repose,
and contracture formation; sequelae of abnormal
reinnervation following denervation such as (3)
associated movements, (4) spontaneous movements,
(5) 'crocodile tears'.
All facial nerves showing only a conduction block
in our patients recovered completely without any of
these sequelae. Even though the six cases of partial
denervation recovered full voluntary movement,
they were not without the sequelae of abnormal
reinnervation. Sequelae in a case of complete
denervation (case 1) are shown in Fig. 4, a, b and c.
Contracture of the affected facial muscle often
shows as a deepening of the nasolabial fold of that
side on repose, and this often improved the facial
symmetry on smiling, to the patient's satisfaction.
At first glance, the side of the contracture gives the
impression of being the normal side and the other as
the weak side. Contracture of the orbicularis oculi
gives the patient a narrow palpebral fissure which is
brought out well on getting the patient to look up.
The eyebrow on the side of the palsy may appear to
be elevated due to contracture in the ipselateral
frontalis muscle.
Due to perversion of reinnervation there may be
 J Neurol Neurosurg Psychiatry: first published as 10.1136/jnnp.40.4.342 on 1 April 1977. Downloaded from http://jnnp.bmj.com/ on February 10, 2020 by guest.
Delayedfacial palsy after head injury 347
expldins why these patients were spared from
contractures. Only six of the 14 cases of denervation
had obvious contracture. The case with a bilateral
facial paralysis maintained a symmetrical mouth
until about the eighth week when the left face
recovered almost complete power, the mouth
became pulled to the left side, and the right face was
still completely paralysed (Fig. 2).
While no case with a conduction block developed
abnormal movement, all cases with denervation
except one developed some form of abnormal
movement. One case developed crocodile tears after
the 30th week, and it is interesting to note that in
this patient the lesion in the facial nerve was at the
level of the geniculate ganglion as indicated by loss
of tear secretion and taste sensation in the anterior
two-thirds of the tongue on that side at the onset.
This patient, who used to have tears pouring down
her cheek, showed some improvement possibly with
returning tone in the orbicularis oculi which increases
the efficiency of the lacrimal duct in draining all that
outpouring of tears.
All cases, except the five palsies with conduction

Protected by copyright.
block and one patient with a partial denervation, had
(C) some sequelae which disfigured the face. Most of
Fig. 4 Case ofleft facial palsy after head injury
these, especially the abnormal movement, could be
showing the sequelae ofcomplete denervation. permanent.
(a) Contracture of left side offace showing as deepening
ofnasolabialfold on that side on repose. (b) Contracture Hearing loss Four patients were left with normal
of orbicularis oculi muscle giving a narrow palpebral audiometry while 10 had a perceptive or mixed
fissure on left side. Eyebrow on this side is elevated due to deafness (Tables 4, 5). None had a pure conduction
contracture of thefrontalis muscle. (c) Dimple on left side deafness.
of chin and retraction of the angle of the mouth of that
side on closing the eye voluntarily or on blinking. Vestibular function Seven patients had abnormal
vestibular function on the side of the facial palsy.
voluntarily. What appears to be a spontaneous The only other cranial nerve palsy observed in
twitch of the mouth on the side of the paralysis is in these 16 cases was a 6th nerve palsy of the opposite
reality associated with involuntary blinking on the side in one patient. There was no instance of olfactory
side of the palsy, the eye blinking rather sluggishly.
This is easily detected on the EMG as bursts of blink Table 4 Audiometry
potentials when the orbicularis oris muscle is studied.
The effort involved in the spontaneous blink is Side offacial
probably channelled between this movement and the palsy only Bilateral
mouth, resulting in a drag on that eyelid. Mixed deafness 1 3
The crocodile tear phenomenon is seen as Perceptive deafness 4 2
Pure conduction deafness 0 0
unilateral lacrimation on eating. In one of our patients Normal 0 4
unilateral lacrimation was also observed on smiling,
on reading, and during facial nerve stimulation.
None of the patients with a conduction block of Table 5 Hearing loss
the facial nerve developed a contracture. The longest 30-40db 40-60db > 60db
time that any of these cases took to recover com-
pletely was 40 days. The earliest day any of the cases Side of facial injury
studied was noted to have a contracture was 58 days. Mixed deafness 1 3 0
Perceptive deafness 0 2 3
The cases with a conduction block improved fully, Opposite side
well before the minimum time required for develop- Mixed deafness 1 1 0
Perceptive deafness 0 1 0
ment of contracture in this series, and this possibly
 J Neurol Neurosurg Psychiatry: first published as 10.1136/jnnp.40.4.342 on 1 April 1977. Downloaded from http://jnnp.bmj.com/ on February 10, 2020 by guest.
348 K. Puvanendran, M. Vitharana, and P. K. Wong
nerve involvement. One patient had an associated nerve, and ultimately cut off its blood supply,
subdural haematoma which was evacuated. None of causing ischaemic damage to the nerve. If the
these patients had evidence of diabetes mellitus or pressure were mild, there would be only a neuro-
hypertension. praxia or conduction block due to segmental
demyelination. If the damage were more severe, there
Discussion could be axonal damage with denervation.
As half of our patients with delayed facial palsy
In facial palsy after closed head injury, a fracture of did not show a petrous fracture on special radiological
the base of the skull is usually assumed to be present. views and tomograms, it may be possible to have
In our cases, all those who had a fracture of the bleeding into the facial canal from the blood vessels
petrous bone and bleeding ear developed a facial found loosely supported around it, somewhat akin
palsy. This finding contrasts with that of Potter (1964) to a subdural haemorrhage due to a sudden shearing
who says that there is a 50 % chance of having some force (Kettel, 1958). The bleeding from the ear comes
facial weakness, and a 38.5 % chance of a delayed from the middle or internal ear, and it is coincidental
facial weakness if there is both a fracture and a in this injury.
bleeding ear (Briggs and Potter, 1967). Briggs and Potter (1967, 1971) postulate a similar
Petrous bone fractures that cause facial paralysis mechanism for this facial weakness as in Bell's palsy
are either transverse or longitudinal. Transverse with a possible inflammatory reaction in and around
fractures are uncommon and account for only 10% the nerve, or a swelling of the nerve in the canal
of temporal bone fracture (Tos, 1971). Transverse which could lead to ischaemia. On this postulate,
fracture rarely causes bleeding from the ear or they claim that prophylatic injection of ACTH after
fractures of external auditory canal. Facial nerve head injury can reduce the incidence of delayed
palsy occurs in 30-50% of transverse fractures, and palsy. Jongkees (1972) thinks that the nerve could

Protected by copyright.
the paralysis is likely to be immediate in onset and be compressed by its swelling within its fibrous
complete because of nerve disruption (McHugh, sheath or epineurium, and this swelling may be a
1959; Harker and McCabe, 1974; Miehlke, 1973b). delayed response to trauma to the nerve itself, or
Up to 90 % of all petrous fractures are longitudinal. secondary to damage to its surrounding vasculature
Unlike the former type, longitudinal fracture very resulting in delayed arterial spasm or arterial or
often passes through the external auditory canal, and venous thrombosis. Turner (1944) says that external
usually tears the tympanic membrane producing pressure on the nerve by blood in the Fallopian
bleeding from the external auditory canal. Facial aqueduct is the likely cause of delayed facial palsy.
paralysis occurs in 10-25 % of these cases (McHugh, Robson and Dawes (1960) attribute this palsy to
1959; Kettel, 1958; Harker and McCabe, 1974), and pressure on the facial nerve from oedema of the
the onset of paralysis is usually delayed (Ulrich, mucosa of the facial canal which was non-bacterial
1926). and was a reaction to the presence of blood. However,
In our cases, 50 % of delayed facial palsies had a this would not explain an onset of facial palsy as
fracture of the petrous bone, 12 % transverse and early as one hour after head injury in one of his cases.
88% longitudinal. This contrasted with the three Such rapid onset could only be explained by rapid
cases of immediate facial palsy where all had bleeding into the facial canal.
demonstrable petrous fracture, transverse in two and We note many dissimilarities between delayed
longitudinal in one. This leads us to doubt the facial palsy and Bell's palsy, and hence assume a
assumption of earlier authors that in facial palsy difference in their mechanisms. Unlike Bell's palsy,
after closed head injury, a fracture of base of skull the upper limit of the lesion extended very often to
may be assumed. It leads also to the question as to the geniculate ganglion in 59 % of cases, and ear pain
what causes the delayed facial palsy in these cases. at onset was not seen. The outcome of the facial
In the facial canal, the area occupied by the facial palsy was much worse than in Bell's palsy. Only
nerve is only 30-50% (at its greatest width) of the 17.6% of our patients had a conduction block and
cross-sectional area of the canal. The remainder of 82.4% had denervation, whereas more than 60% of
the facial canal is occupied by blood vessels with cases of Bell's palsy show a conduction block
connective tissue loosely arranged around it (Sunder- (Taverner, 1959; Groves, 1973; Kristensen, 1968),
land and Cosser, 1953). In immediate facial palsy and thus a better chance for complete recovery
it is easily assumed that the fracture damages or without abnormal reinnervation. Miehlke (1973a, b)
causes severance of the facial nerve. Delayed facial gives his impression that 90% of cases of late
palsy is possibly the result of bleeding into the facial traumatic facial paralysis resolve, regaining satis-
canal. Increasing size of a haematoma in the limited factory nerve function without surgical intervention.
non-expanding bony tube could press on the facial This figure, however, is a clinical impression not
 J Neurol Neurosurg Psychiatry: first published as 10.1136/jnnp.40.4.342 on 1 April 1977. Downloaded from http://jnnp.bmj.com/ on February 10, 2020 by guest.
Delayedfacial palsy after head injury 349
supported by electrical studies. He does not mention Briggs, M. and Potter, J. M. (1971). Prevention of delayed
how much of nerve function recovery and sequelae traumatic facial palsy. British Medical Journal, 3,
is acceptable. The duration of weakness was more 458-459.
protracted with delayed facial palsy showing either Grove, W. E. (1939). Skull fracture involving the ear: a
denervation or conduction block when compared to clinical study of 211 cases. Laryngoscope (St. Louis), 49,
Bell's palsy (Taverner, 1959). 678-833.
The electrical reaction after complete traumatic Groves, J. (1973). Facial palsies: Selection of cases for
denervation is different from that after Bell's palsy treatment. Proceedings of the Royal Society of Medicine,
66, 545-549.
with complete denervation: the conduction time Harker, L. A. and McCabe, B. F. (1974). Temporal bone
increased before excitability finally disappeared. fractures and facial nerve injury. Otolaryngologic
There were also differences in electrical reaction Clinics of North America, 7, 425-431.
between these disorders in partial denervation and Howe, H. A., Tower, S. S., and Duel, A. B. (1937). Facial
conduction block. tic in relation to injury of the facial nerve. Archives of
We have discussed in detail the natural history, Neurology and Psychiatry (Chicago), 38, 1190-1198.
including the rates of clinical recovery, and the Jongkees, L. B. W. (1965). Facial paralysis complicating
sequelae including those from abnormal reinnerva- skull trauma. Archives of Otolaryngology, 81, 518-522.
tion. Lipschitz (1906) explained these phenomena as Jongkees, L. B. W. (1972). On peripheral facial paralysis.
a misdirection of a branch of the regenerating axon Archives of Otolaryngology, 95, 317-323.
into an axon sheath leading to a different part of the Kettel, K. (1958). Peripheral Facial Palsy. Pathology and
face. This theory is supported by experimental Surgery, p. 246. Blackwell Scientific Publications:
Oxford.
studies by Howe et al. (1937) in the monkey. Kristensen, H. F. (1968). Discussion on facial paralysis.
A possibility of complete denervation arises if the Journal of Laryngology and Otology, 82, 665-666.
facial palsy comes on two to six days after head Lipschitz, R. (1906). Beitrage zur lehre von facialislah-

Protected by copyright.
injury. Complete facial palsy often ends up with mung nebst Bemerkungen zur trage der nervenregenera-
denervation. In our cases, a lesion that extended up tion. Monatsschrift fur Psychiatrie und Neurologie
to the geniculate ganglion was always associated (Basel), 20, 84.
with denervation. In bilateral palsy, the second one McHugh, H. E. (1959). Surgical treatment of facial
was milder if its onset was more delayed, and with paralysis and traumatic conductive deafness in fractures
longer delay it could manifest as a latent palsy. of temporal bone. Annals of Otology, Rhinology and
Electrodiagnostic methods were most useful in Laryngology, 68, 855-889.
indicating prognosis very early in the disease. An Miehlke, edition.
A. (1973a). Surgery of the Facial Nerve, 2nd
Urban and Schwarzenberg: Munich.
interesting phenomenon noted was a neuropraxic Miehlke, A. (1973b). Recognition and management of
involvement of the 'uninvolved' side of face in two facial nerve palsies of operative and traumatic origin.
patients. Almost all cases with denervation have Proceedings of the Royal Society of Medicine, 66,
shown abnormal facial movement. 549-554.
In view of the poor outcome in our series when Potter, J. M. (1964). Facial palsy following head injury.
left to its natural course, we believe that decom- Journal ofLaryngology, 78, 645-657.
pression of the facial nerve in the facial canal should Potter, J. M. and Braakman, R. (1976). In Handbook of
be considered at the earliest indication of denervation Clinical Neurology. Edited by P. J. Vinken and G. W.
(Jongkees, 1965). As the facial nerve lesion often Bruyn, Vol. 24, 105-117. North-Holland Publishing
extends up to the geniculate ganglion, the entire Company: Amsterdam, Oxford.
facial nerve in the temporal bone should be decom- Puvanendran, K., Vitharana, M., and Wong, P. K. (1977).
Electrodiagnostic study in delayed facial palsy after
pressed, and this would involve a combined closed head injury. Journal of Neurology, Neurosur-
translabyrinthine and middle cranial fossa approach gery, and Psychiatry, 40, 351-357.
(Miehkle, 1973b). If the geniculate ganglion is intact, Robson, F. C. and Dawes, J. D. K. (1960). Delayed facial
a mastoid approach would be adequate. paralysis of lower motor neurone type following head
injury. Journal ofLaryngology and Otology, 74, 275-289.
We wish to thank Mr E. H. Goh, Head of the ENT Sunderland, the facial
S. and Cosser, D. F. (1953). The structure of
nerve. Anatomical Record, 116, 157-162.
Department, Singapore General Hospital for allow- Taverner, D. (1955). Bell's palsy, a clinical and electro-
ing us to study patients under his care. myographic study. Brain, 78, 209-228.
Taverner, D. (1959). The prognosis and treatment of
References spontaneous facial palsy. Proceedings of the Royal
Society ofMedicine, 52, 33-39.
Briggs, M. and Potter, J. M. (1967). Prevention of delayed Tos, M. (1971). Fractura ossis temporalis. Forl0bet og
traumatic facial palsy. British Medical Journal, 4, f0lger of 248 petrosafrakturer. Ugesker Laeger, 13,
464-465. 1149-1156.
 J Neurol Neurosurg Psychiatry: first published as 10.1136/jnnp.40.4.342 on 1 April 1977. Downloaded from http://jnnp.bmj.com/ on February 10, 2020 by guest.
350 K. Puvanendran, M. Vitharana, and P. K. Wong
Turner, J. W. A. (1943). Indirect injuries of the optic Ulrich, R. (1926). Verletzung des Gehororgans bei
nerve. Brain, 66, 140-151. Schadelbasisfrakturen. Acta Oto-Laryngologica (Stock-
Turner, J. W. A. (1944). Facial palsy in closed head injuries. holm), Suppl. 6, 1.
Lancet, 246, 756-757.

Protected by copyright.