CLINICAL STUDIES

UNUSUAL CAUSES AND PRESENTATIONS OF

HEMIFACIAL SPASM
In-Bo Han, M.D.
Department of Neurosurgery,
Pochon CHA University
College of Medicine,
Sungnam, Korea
Jong Hee Chang, M.D.
Department of Neurosurgery,
Yonsei University
College of Medicine,
Seoul, Korea
Jin Woo Chang, M.D.
Department of Neurosurgery,
Yonsei University
College of Medicine,
Seoul, Korea
Ryoong Huh, M.D.
Department of Neurosurgery,
Pochon CHA University
College of Medicine,
Sungnam, Korea
Sang Sup Chung, M.D.
Department of Neurosurgery,
Pochon CHA University
College of Medicine,
Sungnam, Korea
OBJECTIVE: To evaluate unusual possible causes and clinical presentations of hemifa-
cial spasm (HFS).
METHODS: The authors reviewed 1642 cases of HFS. Assessments were based on clin-
ical features, 3-dimensional time-of-flight magnetic resonance angiography, and sur-
gical findings. Causes other than neurovascular compression at the root exit zone of
the facial nerve were investigated and unusual clinical presentations were noted.
RESULTS: Nine (0.5%) patients had a secondary causative structural lesion, 7 patients
had a tumor, and the remaining 2 had a vascular malformation. Direct compression by
dolichoectatic vertebrobasilar artery was noted in 12 (0.7%) patients. In 7 (0.4%)
patients, only the distal portion of the facial nerve was compressed, and five (0.3%)
had only venous compression. Bilateral HFS and tic convulsif were encountered in 7
(0.4%) and 6 (0.37%) patients, respectively. Fifty-six (3.4%) patients were younger than
30 years old at the time of microvascular decompression.
CONCLUSION: HFS can result from tumor, vascular malformation, and dolichoectatic
artery. Therefore, appropriate preoperative radiological investigations are crucial to
achieve a correct diagnosis. The authors emphasize that distal compression or only
venous compression can be responsible for persistent or recurrent symptoms postop-
eratively. In cases of bilateral HFS, a definite differential diagnosis is necessary for appro-
priate therapy. MVD is recommended as the treatment of choice in patients younger
than 30 years old or patients with painful tic convulsif.
KEY WORDS: Bilateral, Dolichoectasia, Facial nerve, Hemifacial spasm, Neurovascular compression, Tic
convulsif
Neurosurgery 65:130137, 2009 DOI: 10.1227/01.NEU.0000348548.62440.42 www.neurosurgery-online.com

Reprint requests:
Ryoong Huh, M.D.,

or Sang Sup Chung, M.D.,
Department of Neurosurgery,
Pochon CHA University,
Bundang CHA Hospital,
351 Yatap-dong, Bundang-gu,
Sungnam, 463-712, Korea.
Email: haninbo@naver.com
Received, September 1, 2008.
Accepted, March 12, 2009.
Copyright 2009 by the
Congress of Neurological Surgeons
H
emifacial spasm (HFS) is characterized
by involuntary contractions of the
facial muscles and is generally caused
by vascular compression of the root exit zone
(REZ) of the ipsilateral facial nerve (1, 4).
However, a number of rare causes of HFS have
also been identified, which include tumors,
aneurysms, and arteriovenous malformations.
Furthermore, a number of large series have
reported that symptomatic HFS resulting from
mass lesions occurs infrequently in 0.4% to
2.2% of cases (1, 810, 15, 19). In addition,
direct compression by dolichoectatic verte-
brobasilar arteries (VBAs), characterized by
ABBREVIATIONS: HFS, hemifacial spasm; MVD,
microvascular decompression; REZ, root exit
zone; TN, trigeminal neuralgia; VBA, verte-
brobasilar artery
artery enlargement and tortuosity, is also a rare
cause of HFS (13). Vascular compression usu-
ally occurs at the REZ of the facial nerve by
arterial vessels. However, Ryu et al. (14)
reported that only the distal portion of the
facial nerve was compressed in 2.1% of pa-
tients, and Barker et al. (1) reported that the
facial nerve is compressed only by a vein in
2.9% of patients.
Bilateral involvement in HFS is extremely
rare and occurs in only an estimated 0.6% to
5% of patients (7, 17, 18). The coexistence of
HFS and ipsilateral trigeminal neuralgia (TN),
also known as tic convulsif, is an additional
possible clinical manifestation. Barker et al. (1)
reported that 4.1% of HFS patients had tic con-
vulsif. HFS usually occurs in middle-aged and
older patients, but HFS has been reported in
patients younger than 30 years old (3).

130 | VOLUME 65 | NUMBER 1 | JULY 2009 www.neurosurgery-online.com

 UNUSUAL CAUSES AND PRESENTATIONS OF HEMIFACIAL SPASM
In the present study, we investigated unusual causes of HFS
(i.e., other than vascular compression at the REZ of the facial
nerve) and its unusual clinical manifestations. In addition, the
literature on the subject was reviewed and possible mecha-
nisms are discussed.
PATIENTS AND METHODS
Patient Population
We analyzed 1642 HFS patients who were treated at Yonsei
University, Severance Hospital (September 1978February 2003) and at
Pochon CHA University, Bundang CHA Hospital (March 2003March
2007). A total of 1658 operations were performed by the senior author
(SSC). Decisions to perform surgery were based on clinical assessments
and preoperative imaging findings.
Pre- and Intraoperative Evaluation
Preoperative 3-dimensional time-of-flight magnetic resonance
angiography (Sonata; Siemens, Erlangen, Germany) was first utilized in
1992 and preoperative 3-dimensional-constructive interference in
steady state magnetic resonance imaging in 2003. As compared with
3-dimensional time-of-flight images, more vascular contact at the root
entry zone can be detected by constructive interference in steady state
magnetic resonance imaging. Brainstem auditory evoked potentials
were monitored during operations, but facial electromyography was
not performed intraoperatively.
Clinical Assessments
Clinical results were assessed by personal interview or by telephone,
and were classified as excellent, good, fair, or poor, where excellent
meant no residual spasm following microvascular decompression
(MVD), good meant spasm relief of more than 90%, fair indicated
50% to 90% spasm relief, and poor described spasm relief of less
than 50%. Recurrence was defined as reappearance of facial spasm
after complete symptom resolution at more than 6 months after MVD.
RESULTS
Clinical Features and Surgical Outcomes
The mean patient age was 49.5 years (age range, 15 78
years). The male-to-female ratio was 1:3.4, and no significant
laterality differences were observed. The mean symptom dura-
tion was 7.0 years (range, 6 months40 years), and the mean
follow-up duration was 31.7 months (range, 6.4196.8 months).
Excellent and good results were achieved by 87.7% and
5.9% of patients, respectively, resulting in an overall success
rate of 93.7%. The rate of HFS recurrence was 0.8% (n 13).
Postoperative complications included infarction or hemorrhage
(n 18, 1.1%), permanent facial palsy (n 21, 1.3%), permanent
hearing deficit (n 36, 2.2%), wound infection (n 64, 3.9%),
lower cranial nerve palsy (n 4, 0.2%), and a cerebrospinal
fluid leak (n 21, 1.3%). There was no operative death.
Secondary HFS
In patients with neurovascular compression, the main cul-
prits were the anterior inferior cerebellar artery (42.3%), the pos-
terior inferior cerebellar artery (39.9%), and the vertebral artery
NEUROSURGERY
TABLE 1. Type of offenders
Type No. of patients (%)
Vessels 1633 (99.4 %)
Artery 1628 (99.1%)
Only vein 5 (0.3 %)
Mass 9 (0.5 %)
Epidermoid 3
Vestibular schwannoma 1
Lipoma 2
Arachnoid Cyst 1
Arterivenous malformation 1
Developmental venous anomaly 1
Total 1642 (100%)
(1.6%). Multiple offenders were found in 15.5% of cases.
However, 7 patients (0.4%) had HFS secondary to a cerebello-
pontine angle tumor (3 epidermoids, 1 vestibular schwannoma,
1 arachnoid cyst, and 2 lipomas), and 2 patients (0.1%) had HFS
resulting from vascular malformations (1 from an arteriovenous
malformation and 1 due to a developmental venous anomaly).
Types of offenders are summarized in Table 1. In 3 cases (2 epi-
dermoids and 1 arachnoid cyst), the facial nerve was com-
pressed by the tumor and vascular compression was observed
at the REZ after tumor resection. Therefore, in these cases,
Teflon felt (Bard, Tempe, AZ) was also placed to decompress the
facial nerve (Fig. 1). In the remaining 3 cases (1 epidermoid, 1
vestibular schwannoma, and 1 lipoma), surgical findings
showed an incorporated and distorted facial nerve within the
tumor without any obvious vascular compression; thus, only
tumor removal was performed (Fig. 2). Postoperatively, HFS
completely disappeared in all 6 patients. In 1 case with worsen-
ing HFS, we excised the lipoma; the subsequent HFS completely
resolved. However, the patient developed transient facial weak-
ness and a severe permanent hearing loss after resection.
Another lipoma patient was managed conservatively due to a
perceived high surgical risk (Fig. 3). In a case with a cerebellar
arteriovenous malformation, an engorged draining vein com-
pressed the superior aspect of the REZ of the facial nerve (Fig.
4). This patient was treated effectively by gamma knife surgery.
One patient (patient 25) with a developmental venous anomaly
presented with a 5-year history of left trigeminal neuralgia and
an 18-month history of left HFS. In this case, the HFS was
caused by a draining vein which compressed the REZ of the
facial nerve, and the TN probably resulted from the involve-
ment of the spinal trigeminal nucleus of the trigeminal nerve by
the developmental venous anomaly (Fig. 5). This patient was
treated conservatively due to the high risk of surgery.
Vertebrobasilar Dolichoectasia
Direct compression by dolichoectatic VBA was noted in 12
patients (0.7%) who underwent MVD. The mean patient age
VOLUME 65 | NUMBER 1 | JULY 2009 | 131
HAN ET AL.
A B
FIGURE 1. Preoperative magnetic C
resonance imaging (MRI) series in
a 75-year-old woman with a 5-year
history of right hemifacial spasm.
A, axial T2-weighted MRI scan
showing a hyperintense lesion in
the right cerebellopontine angle cis-
tern. B, axial 3-dimensional time-
of-flight (3D-TOF) magnetic reso-
nance angiogram (MRA) showing
the artery (arrow) compressing the
root exit zone. C, diffusion weighted
image revealing the hyperintensity
of the lesion (arrow), consistent
with an epidermoid tumor.
A B
FIGURE 2. Preoperative MRI series in a 34-year-old woman with a 2-year
history of left hemifacial spasm. A, T1-weighted scan showing a
hypointense lesion in the left cerebellopontine angle cistern. B, T2-
weighted scan revealing a hyperintense lesion, consistent with an epider-
moid tumor.
was 58.9 years (age range, 4271 years) with a male-to-female
ratio of 7:5. Ten patients had symptoms on the left side, and the
mean symptom duration was 7.0 years (range, 122 years).
Dolichoectatic VBA proved difficult to mobilize and a Teflon
ball (Bard, Tempe, AZ) was introduced step-by-step from the
proximal part of the dolichoectatic VBA to separate the facial
nerve from the offender. After performing this maneuver, all
the patients achieved satisfactory symptom relief. However,
the postoperative complication rate among these 12 patients
was greater than the overall complication rate. Transient facial
palsy and transient hearing impairment were noted in 25% and
132 | VOLUME 65 | NUMBER 1 | JULY 2009
A B
C D
FIGURE 3. Axial MRI series in a 37-year-old woman with a 2-year his-
tory of left hemifacial spasm. T2-weighted (A), T1-weighted (B), and
fluid-attenuated inversion recovery (C) images demonstrating a hyperin-
tense lesion. D, the lesion appears hypointense on fat suppression
sequences, consistent with a lipoma.
26%, respectively, and 8.3% had a permanent hearing deficit. Of
these 12 patients, one (patient 17) had bilateral HFS and 2
patients (patients 20 and 24) had tic convulsif. The patient with
bilateral symptoms underwent 2-stage MVD and achieved
complete recovery, and the 2 patients with tic convulsif
achieved satisfactory results after MVD.
Only Venous Compression of the Facial Nerve
Venous compression alone was identified in 5 patients (0.3%).
Table 2 shows the clinical features of these patients. Two patients
(patients 1 and 4) exhibited a prominent large vein crossing over
the REZ of the facial nerve. Teflon felt was placed to separate this
from the facial nerve and to elevate coagulated venous branches
off the facial nerve following coagulation. The remaining 2
patients (patients 2 and 3) had a smaller vein, which had coagu-
lated and divided, compressing the facial nerve REZ. As
described in our previous report (11), 1 patient (Patient 5) had no
improvement after the first MVD at another hospital and under-
went a second operation. During the second MVD procedure, a
large vein compressing the facial nerve in the distal region only
was located under Teflon felt. The patient achieved complete
relief from HFS after the second MVD procedure.
Distal Compression of the Facial Nerve
The authors treated 7 patients (0.4%) who exhibited com-
pression only at the distal portion of the facial nerve. Table 3
www.neurosurgery-online.com
 UNUSUAL CAUSES AND PRESENTATIONS OF HEMIFACIAL SPASM
A B
FIGURE 4. 3D-TOF MRA in a 40-year-old man with a 5-year history of
left hemifacial spasm, revealing a nidus (arrow) in the left cerebellum (A)
and dilated vein (arrow) compressing superior aspect of root exit zone of
facial nerve (B). The patients hemifacial spasm improved significantly
after gamma knife radiosurgery.
A B
FIGURE 5. Two-dimensional fast C 21, and 22), multiple vessels were responsible for tic convulsif.
spin echo image (A) and T1-
weighted image after the adminis-
tration of gadolinium (B) showing
dilated draining vein (arrow) com-
pressing the facial nerve (arrow-
head). C, T1-weighted image with
contrast demonstrating an involve-
ment of the spinal trigeminal
nucleus (arrow) of the trigeminal
nerve by the developmental venous
anomaly.
shows the clinical features of these patients. All patients expe-
rienced typical HFS, and no vascular compression was found at
the REZ in these 7 patients. All but 2 patients (patients 5 and 6)
had an excellent result after MVD. In 1 patient (patient 5) with
persistent spasm after the first MVD procedure, the facial nerve
was found to be distally compressed by a vein. This patient
achieved complete recovery after repeat MVD.
Bilateral HFS
The prevalence of bilateral HFS in our series was 0.4% (7
patients), and its main clinical features are summarized in Table
4. All patients had unilateral HFS followed by bilateral and
asymmetric facial contractions. The mean age of these 7
patients was 50.3 years (age range, 3569 years), and the mean
NEUROSURGERY
symptom duration was 8.6 years (range, 120 years). HFS
started on the left side in 5 of the 7, and the mean latency to
contralateral facial nerve involvement was 6.1 year (range,
0.510 years). The sides affected initially usually remained
more severely involved. The upper face was the site of onset in
all patients. All but 1 of the 7 patients underwent unilateral
MVD for the initial more severe spasm, because contralateral
HFS was mild and satisfactory results were achieved post uni-
lateral MVD. However, 1 patient (patient 17) with dolicho-
ectatic VBA compressing the facial nerve had sufficiently severe
contralateral symptoms to require bilateral MVD (Fig. 6). This
patient developed left eyelid twitching followed by lower facial
muscles. Operative findings on the side of initial onset showed
REZ compression of the facial nerve by dolichoectatic VBA and
posterior inferior cerebellar artery. Approximately 6 years after
HFS abolition by initial MVD, the patient developed twitching
of right facial muscles. A second operation revealed neurovas-
cular compression at the right REZ by the anterior inferior cere-
bellar artery (Fig. 6). This patient was successfully treated by 2-
stage bilateral MVD.
Tic Convulsif (Concurrent HFS and Ipsilateral TN)
Six patients (0.4%) were found to have tic convulsif. Table 5
shows the clinical features of these 6 patients. Two patients
(patients 20 and 24) noted facial muscle contraction and ipsilat-
eral severe TN. These patients were successfully treated by
simultaneous decompression of Cranial Nerves V and VII. One
patient (patient 24) had a single dolichoectatic VBA compressing
both the trigeminal and facial nerve. In 3 patients (patients 20,
Because complete remission from TN occurred after botulinum
toxin injections, 2 patients (patients 21 and 22) underwent MVD
for HFS alone. Another patient (patient 23) exhibited no visible
vessels compressing the REZ of the trigeminal nerve and MVD
was performed for only HFS. Patient 25 had a developmental
venous anomaly, namely, an engorged draining vein compress-
ing the REZ of the facial nerve and the spinal trigeminal nucleus
of the trigeminal nerve (Fig. 5). This patient was managed con-
servatively due to the high risk of surgery.
HFS in Youth
Fifty-six (3.4%) patients were younger than the age of 30
years at the time of MVD. Thirty-three (58.9%) were women,
and 31 (55.4%) had right side symptoms. The mean symptom
duration was 5.7 0.5 years. The patients age at HFS onset
ranged from 11 to 27 years (mean, 22.6 0.7 years) and the age
at time of MVD ranged from 21 to 30 years (mean, 26.3 0.3
years). All patients had definite offending vessels as follows:
anterior inferior cerebellar artery in 48.2% (27 of 56), posterior
inferior cerebellar artery in 42.9% (24 of 56), or multiple vessels
in 8.9% (5 of 56). No anatomic variations were found of vessels
at the REZ. The overall success rate (excellent or good) was
92.9% (n 52) and 1 (1.8%) recurred. No significant difference
was found between young patients and older patients in terms
of symptom duration and surgical outcome.
VOLUME 65 | NUMBER 1 | JULY 2009 | 133
HAN ET AL.

TABLE 2. Summary of five cases of hemifacial spasm resulting from only venous compressiona
Symptom Side of
Patient no. Age (y)/sex Treatment Result
duration (y) symptom
1 38/F 8 Left MVD Excellent
2 61/F 7 Right Coagulation Poor
3 63/F 10 Right Coagulation Excellent
4 35/F 7 Right MVD Excellent
5 49/F 5 Right First operation: MVD of artery at REZ No relief
Second operation: 5 years later Excellent
No NVC at REZ
Vein

a
MVD, microvasular decompression; REZ, root entry zone; NVC, neurovascular compression.

TABLE 3. Summary of seven cases of hemifacial spasm resulting from compression of the distal portion of the facial nervea
Age Symptom Surgical
Patient no. Side Offenders
(y)/sex duration (y) outcome
6 45/F 3 Left No NVC at REZ Poor
MVD of PICA at distal portion of the nerve
7 60/M 8 Left No NVC at REZ Good
MVD of AICA at midportion of the nerve
8 42/F 6 Left No NVC at REZ Excellent
MVD of PICA at distal portion of the nerve
9 46/M 2 Left No NVC at REZ Good
MVD of AICA at distal portion of the nerve
5 49/F 5 Right First operation: MVD of artery at REZ No relief
Second operation: 5 years later Excellent
No NVC at REZ
Vein at distal portion of the nerve
11 62/F 1 Left No NVC at REZ Good
MVD of AICA at distal portion of the nerve
12 56/M 5 Right No NVC at REZ Excellent
MVD of PICA at distal portion of the nerve

a
NVC, neurovascular compression; REZ, root entry zone; MVD, microvasular decompression; PICA, posterior inferior cerebellar artery; AICA, anterior inferior cerebellar artery.

DISCUSSION
HFS is characterized by a painless, involuntary twitching of
one side of the face. Vascular compression at the REZ of the
facial nerve has been suggested to be responsible for HFS (1, 4),
although other less common etiologies and presentations have
been described (2, 6, 810, 12, 15, 16, 19).
Secondary HFS
HFS may result from direct neural compression by a second-
ary mass lesion causing neural dysfunction or irritation of the
facial nerve pathway. Coexistent vascular compression of the
facial nerve is a possible cause in secondary HFS. Therefore,
MVD of the facial nerve should be performed after tumor exci-
sion if a vessel loop is noted in close proximity to the nerve and
is compressing its REZ (6). In cases of cerebellopontine angle
lipoma, which is a slow-growing tumor with infiltrative
growth, conservative treatment is usually recommended due to
the high risk of surgery. However, in cases with intractable
clinical symptoms, partial resection is usually advocated to
minimize postoperative deficits (2). In our series, 1 lipoma
patient with worsening HFS achieved complete relief from
spasm after surgical resection, but at the expense of transient
facial weakness and permanent severe hearing loss.
Vertebrobasilar Dolichoectasia
HFS resulting from dolichoectatic VBA has been reported to
be treatable by MVD. However, manipulation of ectactic arter-

134 | VOLUME 65 | NUMBER 1 | JULY 2009 www.neurosurgery-online.com

 UNUSUAL CAUSES AND PRESENTATIONS OF HEMIFACIAL SPASM

TABLE 4. Clinical features of patients with bilateral hemifacial spasma

Age at onset Symptom Side Symptom Side of
Patient no. Latency (y) Severity Imaging
(y)/sex duration (y) onset type MVD
13 39/F 7 L Typical 6 LR Left PICA compression L
Right AICA contact
14 60/F 20 L Typical 10 LR Left AICA compression L
Right AICA contact
15 35/F 3 R Typical 7 RL Right PICA compression R
Left: normal
16 54/F 6 R Typical 7 RL Right AICA compression R
Left AICA contact
17 51/F 1 L Typical 4 LR Dolichoectatic VBA L, R
18 44/F 8 L Typical 8 LR Left AICA compression L
Right: normal
19 69/F 15 L Typical 0.5 LR Left AICA compression L
Right PICA contact

a
MVD, microvasular decompression; L, left; R, right; PICA, posterior inferior cerebellar artery; AICA, anterior inferior cerebellar artery; VBA, vertebrobasilar artery.

(4, 13). In our series, the postoperative complication rate was

A B greater than the overall complication rate, although satisfactory
symptom relief was achieved. To reduce complications and
decompress the REZ successfully, we recommend a stepwise
decompression method, that is, incremental decompression
from the proximal part of the dolichoectatic VBA using a Teflon
ball. Furthermore, an effort should be made to leave enough
working space to decompress offenders related to dolichoecta-
sia and to reduce vascular manipulation to a minimum.

Only Venous Compression of the Facial Nerve

C D
FIGURE 6. Preoperative 3D-TOF MRA showing left vertebral artery
compressing the facial nerve at the root exit zone (A) and tortuosity of the
vertebral artery (B). C and D, MRA source images after first operation
revealing loop of right anterior inferior cerebellar artery (arrow) causing
neurovascular compression.
ies and traction of small perforators during MVD have been
reported to present serious risks, which included permanent
facial weakness, deafness, stroke, or death. Furthermore, fre-
quent recurrence may occur due to poor mobility of the artery,
which makes it difficult to achieve sufficient decompression
Venous compression alone can cause persistent symptoms or
recurrence after the first MVD procedure (11, 14). In our series,
1 patient (patient 5) with persistent spasm had a large vein in
the distal portion that was overlooked during the first MVD
procedure. This serves as a reminder that every effort should be
made to identify veins that compress the facial nerve during
operation. Furthermore, recurrence of HFS attributable to veins
can be caused by venous recollateralization. Therefore, Teflon
felt should be placed to elevate coagulated venous branches,
especially if a large vein compressing the facial nerve is coag-
ulated and cut.
Distal Compression of the Facial Nerve
Neurosurgeons can easily overlook vascular compression of
the distal portion of the facial nerve, and as a result, spasms
may persist after MVD. Although Schwann cells of peripheral
axons are more resistant to demyelination, distal compression
of the facial nerve may play an important role, especially in
cases with no offender at the REZ (14). Therefore, it should be
borne in mind that compression of the distal portion of the
nerve may be responsible for HFS in cases with no vascular
compression at the REZ during operation, and in patients with
a persistent spasm after first MVD at the REZ.

NEUROSURGERY VOLUME 65 | NUMBER 1 | JULY 2009 | 135

HAN ET AL.

TABLE 5. Summary of five cases with tic convulsifa

Offender
Age Trigeminal Facial Follow-up Surgical
Patient no. Side Surgery
(y)/sex nerve nerve duration (y) outcome
20 59/M Vein, SCA Dolichoectactic L 5 MVD for TN, No pain,
VA, AICA HFS HFS: good
21 40/F SCA AICA R 2 MVD for No pain, HFS:
HFS excellent
22 44/F SCA PICA R 2 MVD for No pain, HFS:
HFS excellent
23 52/F No AICA R 1 MVD for Mild pain, HFS:
offender HFS excellent
24 75/F Dolichoec- Dolichoec- L 2 MVD for TN, Mild pain,
tactic VA, SCA tactic VA, PICA HFS HFS: good
25 41/F Developmental Draining vein L 3 Conservative
venous anomaly of developmental treatment
venous anomaly

a
SCA, superior cerebellar artery; VA, vertebral artery; AICA, anterior inferior cerebellar artery; L, left; MVD, microvasular decompression; TN, trigeminal neuralgia; HFS, hemifacial
spasm; R, right; PICA, posterior inferior cerebellar artery.

Bilateral HFS
In our series, no significant clinical difference was found
between unilateral and bilateral HFS. All bilateral patients had
typical HFS, and the side affected initially usually remained the
more severely involved. Bilateral HFS may sometimes be difficult
to distinguish from other facial dyskinesias, such as ble-
pharospasm, facial tics, and facial myokymia, but a misdiagnosis
can lead to inadequate therapy a progressive symptom worsen-
ing. Thus, differential diagnosis should be undertaken promptly
to ensure appropriate treatment. The protracted latency of con-
tralateral involvement and asymmetrical contractions may help
to differentiate bilateral HFS from other facial dyskinesias (7).
Tic Convulsif
Previous reports have described various causes of tic convul-
sif. These include dolichoectatic vertebral artery, vascular mal-
formations, and tumors (5, 6). In our series, multiple vessels
compressed the REZ of the trigeminal and facial nerve in most
cases, but only 1 patient had a single dolichoectatic VBA com-
pressing both facial and trigeminal nerves. When neurovascu-
lar compressions are found at the REZs of trigeminal and facial
nerves, MVD of the 2 cranial nerves is recommended.
HFS in Youth
HFS is generally a disease of the elderly, and vascular com-
pression is believed to result from atherosclerotic vessel
changes. As described in our previous report (3), in young
patients (30 years) undergoing MVD, we rarely observed a
tortuous vertebral artery compressing the facial nerve and
found no anatomic variations of vessels or any arachnoid thick-
ening. The findings of the present study show that satisfactory
surgical results are achieved in both younger and older patients.
Furthermore, they also suggest that HFS may be caused by eti-
ologies other than congenital anomalies of offenders or the ves-
sel aging process. Moreover, although the pathogenesis of early
onset HFS remains unclear, the causes and progress of HFS
appear to be identical in young and elderly HFS patients.
CONCLUSIONS
HFS can result from tumor, vascular malformation, or
dolichoectatic artery. Therefore, appropriate preoperative radi-
ological investigations are mandatory to reach a correct diagno-
sis. Distal compression can be responsible for HFS, especially if
vascular compression at the REZ is not identified during oper-
ation and if MVD at the REZ does not result in cure. Further-
more, it should be borne in mind that only venous compression
can be responsible for persistent symptoms or recurrence after
first MVD. In cases of bilateral HFS, a definite differential diag-
nosis is necessary to facilitate appropriate therapy. Moreover, in
young patients under 30 and painful tic convulsif patients,
MVD is recommended as the treatment of choice.
Disclosure
The authors have no personal financial or institutional interest in any of the
drugs, materials, or devices described in this article.
REFERENCES
1. Barker FG 2nd, Jannetta PJ, Bissonette DJ, Shields PT, Larkins MV, Jho HD:
Microvascular decompression for hemifacial spasm. J Neurosurg 82:201210,
1995.
2. Brodsky JR, Smith TW, Litofsky S, Lee DJ: Lipoma of the cerebellopontine
angle. Am J Otolaryngol 27:271274, 2006.
3. Chang JW, Chang JH, Park YG, Chung SS: Microvascular decompression of the
facial nerve for hemifacial spasm in youth. Childs Nerve Syst 17:309312, 2001.

136 | VOLUME 65 | NUMBER 1 | JULY 2009 www.neurosurgery-online.com

 UNUSUAL CAUSES AND PRESENTATIONS OF HEMIFACIAL SPASM
4. Chung SS, Chang JH, Choi JY, Chang JW, Park YG: Microvascular decompres-
sion for hemifacial spasm: A long-term follow-up of 1,169 consecutive cases.
Stereotact Funct Neurosurg 77:190193, 2001.
5. Cook BR, Jannetta PJ: Tic convulsif: Results in 11 cases treated with microvas-
cular decompression of the fifth and seventh cranial nerves. J Neurosurg
61:949951, 1984.
6. Desai K, Nadkarni T, Bhayani R, Goel A: Cerebellopontine angle epidermoid
tumor presenting with tic convulsif and tinnitusCase report. Neurol Med
Chir (Tokyo) 42:162165, 2002.
7. Felcio AC, Godeiro-Junior Cde O, Borges V, Silva SM, Ferraz HB: Bilateral
hemifacial spasm: A series of 10 patients with literature review. Parkinsonism
Relat Disord 14:154156, 2008.
8. Kim Y, Tanaka A, Kimura M, Yoshinaga S, Tomonaga M: Arteriovenous mal-
formation in the cerebellopontine angle presenting as hemifacial spasm
Case report. Neurol Med Chir (Tokyo) 31:109112, 1991.
9. Kobata H, Kondo A, Iwasaki K: Cerebellopontine angle epidermoids pre-
senting with cranial nerve hyperactive dysfunction: Pathogenesis and long-
term surgical results in 30 patients. Neurosurgery 50:276286, 2002.
10. Nagata S, Matsushima T, Fujii K, Fukui M, Kuromatsu C: Hemifacial spasm
due to tumor, aneurysm, or arteriovenous malformation. Surg Neurol
38(3):204209, 1992.
11. Park YS, Chang JH, Cho J, Park YG, Chung SS, Chang JW: Reoperation for
persistent or recurrent hemifacial spasm after microvascular decompression.
Neurosurgery 58:11621167, 2006.
12. Peker S, Ozduman K, Kili T, Pamir MN: Relief of hemifacial spasm after
radiosurgery for intracanalicular vestibular schwannoma. Minim Invasive
Neurosurg 47:235237, 2004.
13. Rahman EA, Trobe JD, Gebarski SS: Hemifacial spasm caused by vertebral
artery dolichoectasia. Am J Ophthalmol 133:854856, 2002.
14. Ryu H, Yamamoto S, Sugiyama K, Uemura K, Miyamoto T: Hemifacial spasm
caused by vascular compression of the distal portion of the facial nerve.
Report of seven cases. J Neurosurg 88:605609, 1998.
15. Sato K, Ezura M, Takahashi A, Yoshimoto T: Fusiform aneurysm of the ver-
tebral artery presenting hemifacial spasm treated by intravascular emboliza-
tion: Case report. Surg Neurol 56:5255, 2001.
16. Takano S, Maruno T, Shirai S, Nose T: Facial spasm and paroxysmal tinnitus
associated with an arachnoid cyst of the cerebellopontine angleCase report.
Neurol Med Chir (Tokyo) 38:100103, 1998.
17. Tan EK, Chan LL: Clinico-radiologic correlation in unilateral and bilateral
hemifacial spasm. J Neurol Sci 222:5964, 2004.
18. Tan EK, Jankovic J: Bilateral hemifacial spasm: A report of five cases and a lit-
erature review. Mov Disord 14:345349, 1999.
19. Uchino M, Nomoto J, Ohtsuka T, Kuramitsu T: Fusiform aneurysm of the ver-
tebral artery presenting with hemifacial spasm treated by microvascular
decompression. Acta Neurochir (Wien) 147:901903, 2005.
Acknowledgment
Ryoong Huh, M.D., and Sang Sup Chung, M.D., contributed equally to the
preparation of this manuscript.
COMMENTS
H emifacial spasm (HFS) is a very specific clinical entity that is asso-
ciated with vascular compression at the root exit zone of the facial
nerve, with very rare exceptions. This article elegantly outlines those
rare exceptions and provides some insight into the incidence and clini-
cal management of other structural anomalies that may account for HFS.
Han et al. have reviewed the extensive experience of the senior author
(SSC) in managing 1642 patients with HFS who were evaluated by clin-
ical examination, 3-dimensional magnetic resonance angiography, and
intraoperative findings in the majority of patients. Nine patients har-
bored structural abnormalities, including 7 neoplasms and 2 vascular
malformations causing the HFS. Direct compression by dolichoectatic
vertebrobasilar arteries was identified in 12 patients, for an incidence of
1.2% of patients having structural compression of the facial nerve attrib-
utable to an anomaly other than microvascular compression.
NEUROSURGERY
Although rare, the presence of other structural abnormalities causing
HFS underscores the importance of preoperative high-resolution imag-
ing. This is important not only for planning surgical intervention but
also for avoiding surgical intervention in unusual circumstances in
which the pathology may not be readily amenable to surgical therapy.
This is the situation for patients with lipomas and developmental
venous anomalies.
The authors also reviewed patients with unusual manifestations of
HFS, including 7 patients with a compression of the distal portion of the
facial nerve, 5 with venous compression alone, 7 with bilateral HFS, and
6 with tic convulsif. Three of the patients underwent microvascular
decompression when they were under 20 years of age. This large experi-
ence has been analyzed in detail to provide further insight into the
unusual manifestations and unusual etiologies for HFS. These anomalies
and variants should be kept in mind when evaluating patients with HFS.
Daniel L. Barrow
Atlanta, Georgia
I n this report, Han et al. describe their collective experience with a very
large series of patients operated on for HFS during a 30-year period. A
total of 1642 patients over this time period were analyzed, and the
authors culled out those patients who had unusual clinical or surgical
findings to obtain a guess on the incidence of these findings as well as
the results of surgery in these unusual circumstances. Specifically, the
authors examined patients with secondary HFS, bilateral HFS, HFS
attributable to venous compression, HFS attributable to distal nerve com-
pression, and HFS in the pediatric population. These scenarios are admit-
tedly uncommon. In fact, the small numbers of patients in each of these
categories in this large series underscores just how uncommon they are.
The first message, therefore, is that whenever a practitioner encounters
a patient with, say, bilateral HFS, another cause for these symptoms
should be thoroughly sought. The same can be said of patients thought
to have venous compression as a cause of HFS; these patients are far
more likely to have an occult arterial compression.
The overall results reported in this large series largely reflect those
of other previously published studies. The complications outlined by
the authors also reflect the usual array of complications seen elsewhere.
Therefore, with the usual caveats of a historical, uncontrolled series,
this series confirms earlier reports. There is not enough information in
this type of report on the results of surgery in the setting of any of the
unusual causes the authors examined; there are simply too few patients
in each of these groups. Nevertheless, specific attention to the unusual
causes of HFS in such a large series allows the reader to glean insight
into their true incidence.
Oren Sagher
Ann Arbor, Michigan
T he surgical treatment of choice for HFS is microvascular decompres-
sion. However, the diverse anatomic variants and causes of compres-
sion of the facial nerve do not always make it easy. The authors present a
series of 1642 patients with HFS treated during a period of more than 29
years. Through their distinct experiences, they describe the unusual cases,
including HFS attributable to venous compression, compression of the
distal portion of the facial nerve, bilateral HFS, and tic convulsif.
Although this article is rather descriptive for the paucity of unusual
causes, it may give the reader information on the incidence and operative
results of these entities. I wish that the overall results, including compli-
cations, of their vast experience were presented so that one could compare
and understand the surgical outcomes of these unusual cases.
Dong Gyu Kim
Seoul, Korea
VOLUME 65 | NUMBER 1 | JULY 2009 | 137