Key-words:
Hearing loss - microvascular decompression - retromastoid
Introduction
Hemifacial spasm is a clinical syndrome manifesting as irregular, involuntary, and
recurring unilateral facial muscles contractions.[[1]] Medical treatment and botulinum toxin are less invasive alternates; however, microvascular
decompression (MVD) had been proven as a more effective and long-lasting option.[[2]],[[3]] The complications of MVD procedure for hemifacial spasm include cerebrospinal fluid
(CSF) leak, facial palsy, hearing loss, lower cranial nerve (LCN) palsy, cerebellar
or brainstem infarcts, and hematomas.[[4]],[[5]] Hearing loss is one of the most dreadful complications following MVD. Sensorineural
hearing loss (SNHL) can be caused due to stretching of CN VIII, whereas cerebellar
retraction, direct manipulation of cranial nerve (CN) 8, injury or vasospasm of the
labyrinthine artery and AICA, and acoustic trauma due to drill noise[[6]] may also be included in causative factors. Fluid entering the mastoid air cells
may cause middle ear effusion leading to conductive hearing loss.[[7]]
Hearing loss may have a very significant impact on an individual's social and professional
life. The aim of our study was to evaluate the incidence and severity of hearing loss
following MVD in hemifacial spasm. We also tried to analyze the factors which might
be responsible for this complication, so that it can be further minimized in future.
The role of intraoperative adjuncts such as neurophysiological monitoring (brainstem
auditory evoked responses [BAER]), dual-image video angiography (DIVA), and endoscope
was evaluated in terms of outcome.
Methods
An retrospective analysis of records of all patients who underwent MVD for trigeminal
neuralgia at Bantane Hospital, Fujita Health University from January 1, 2014, to December
31, 2018, was conducted. This included patients with at least 3 months of follow-up.
Preoperative evaluation included magnetic resonance imaging (MRI) brain spoiled gradient
recalled echo (SPGR) sequence and computed tomography (CT) angiography. MRI and CT
angiography images were fused to clearly demonstrate the neurovascular conflict [[Figure 1]]. Pure-tone audiometry (PTA) was not performed routinely in all patients.
Figure 1: Fusion image (magnetic resonance imaging brain SPGR sequence and computed tomography
angiography) demonstrating anterior Inferior cerebellar artery (AICA) loop compressing
right side VII cranial nerve root entry zone
Surgical technique
All patients were operated in a lateral decubitus position. A small retromastoid craniotomy
was made. Dura opened in the curvilinear fashion based on sigmoid sinus and laterally
tented with tack-up sutures. Cisterns were fenestrated to release CSF and relax the
cerebellum. Sharp arachnoid dissection was performed around CNs except CN 8. The endoscope
was used at this time for identification and better anatomical orientation of offending
vessel loop. Machida right angle endoscope was used, which is easy to introduce deep
through narrow working corridor. Teflon pledget was placed between the vessel loop
and VII nerve. Transposition of the vessel loop was the preferred method whenever
feasible. The primary dural closure was performed in all cases [[Figure 2]].
Figure 2: Intraoperative images. (a) Craniotomy and dural opening. (b) Lax cerebellum after
draining cerebrospinal fluid from cerebellopontine cistern. (c) AICA in close proximity
to VII-VIII nerve complex. (d) Endoscopic view showing AICA loop compressing VII nerve
root entry zone. (e) AICA transpositioned away from VII to VIII nerve complex and
fixed to petrous dura with gel foam soaked in glue. (f) Teflon pledget placed between
AICA and VII nerve root entry zone
Intraoperative brainstem auditory evoked potential
Brainstem auditory evoked potential (BAEP) monitoring was performed using Nihon Kohden's
system. The auditory stimulus was delivered at a frequency and intensity 11.1 Hz and
105 db, respectively, through earphones. Baseline recordings were done. An increase
in wave V latency by >1.0 ms and fall in amplitude by more than 50% were considered
significant.
Postoperatively, the outcome was assessed using the scoring system given by Kondo
et al. Hearing was assessed subjectively; however, PTA was performed in patients who
complained hearing loss.
Statistical analysis
Quantitative data were described in mean, and qualitative data were described in proportions.
The factors responsible for good outcomes were analyzed as well as those for complications.
Results
Patient demographics
A total number of patients included in the study was 30 (n = 30). The mean age was
60.7 years (range = 36–83 years, standard deviation = 15.5). There were 11 (36.7%)
male and 19 female patients (63.3%). All the patients presented with hemifacial spasm
[[Table 1]].
Table 1: Patient demographics and operative findings
Operative findings
The most commonly encountered offending vessel was AICA (40%), followed by PICA (33.3%).
In seven cases (23.3%), two loops were found compressing the nerve. The vein was not
found culprit in any case. The late phase of fusion images can identify venous compression.
Fusion imaging was able to correctly localize the culprit vessel in all cases except
in the two patients which did not showed any benefit in spasm, probably due to thick
arachnoidal bands present at the root entry zone (REZ). Out of total 37 vessel loops,
interposition and transposition were done in 16 (43.2%) and 21 (56.7%) cases, respectively.
The endoscope and DIVA were used in all cases.
Outcomes
Freedom from spasm
The mean follow-up period was 12.9 months. Freedom from spasm (E0 and E1) was achieved
in 27 patients (90%) at the time of discharge. Moderate spasm (E2) persisted in one
patient (3.3%) which was controlled with medicines. There was no relief from spasm
in two (6.7%) patients, probably due to the thick arachnoidal bands present at REZ
in them.
Complications
The recurrence was noted in 3 patients (10%) at 6 months follow-up. In the recurrence
group, two of three patients were managed successfully with botulinum injection. Facial
palsy was noted in one patient (3.3%).
Hearing loss
Postoperative hearing loss was found in one female patient, which was confirmed by
audiometry. The offending vessel, in this case, was both AICA and PICA loop which
were transpositioned during surgery. This patient was completely spasm free postoperatively.
Postoperative audiogram showed left side severe SNHL [[Figure 3]].
Figure 3: Pure-tone audiometry showing left side severe sensorineural hearing loss
The retractor was not applied during the surgery. During arachnoid dissection, there
was an increase in wave V latency by 1 ms and drop in amplitude by more than 50%,
after which dissection was stopped till wave V latency and amplitude came back to
normal [[Figure 4]]. At the end of the procedure, V-wave latency and amplitude were the same as baseline.
Figure 4: Intraoperative brainstem auditory evoked potential showing drop in amplitude and
increase in latency of wave V
Discussion
Hemifacial spasm is a movement disorder, characterized by progressive, involuntary,
and irregular contractions of facial muscles.[[1]],[[8]] It typically begins with the periocular muscles involvement and gradually involves
other ipsilateral facial muscles.[[9]] In severe cases, the platysma and stapedius can also be involved. Differential
diagnoses include blepharospasm, oromandibular dystonia, facial tic, and focal seizures.[[1]] Disease prevalence reported in the literature is 11 cases/100,000 population, with
a female-to-male ratio of 2:1.[[10]]
The most common cause of HFS is compression of VII CN by vessel loop at its REZ.[[11]] REZ is the transition between central and peripheral myelination.[[12]] Other less common causes of compression include cerebellopontine angle tumors,
arachnoid cysts, and thick arachnoid bands.[[1]],[[13]] Rarely multiple sclerosis and brainstem infact can also cause HFS.
Several hypotheses had been proposed to explain the pathogenesis of HFS.
Peripheral hypothesis advocates ectopic and ephaptic transmission in VII CN, whereas
the central hypothesis assumes hyperexcitability of the facial nerve nucleus.[[12]]
The treatment options include medical treatment, botulinum injection, and MVD. Medical
treatment is effective only for mild form of the disease.[[1]],[[3]] The drugs commonly used include carbamazepine, oxcarbazepine, baclofen, and gabapentin.
Botulinum injection provides symptomatic relief in 85%–90% of patients; however, its
major disadvantage is needed to repeat injection at 3–4 months' interval. Patients
with high anesthetic risk and not willing for surgery are suitable candidates for
botulinum injection.[[1]],[[2]],[[14]] MVD is the most effective option as it deals directly with the etiology. Postoperative
complications include CSF leak, wound infection, stroke and hematomas, facial palsy,
hearing loss, and LCN palsy.[[4]],[[5]],[[15]]
As HFS is a benign disease, patients usually do not accept undue postoperative complications.
Hence, all the measures should be applied during surgery, which helps in minimizing
the complications. As per the literature, the success rate of MVD for HFS is 80%–88%
in the first postoperative year, which is consistent with our results.[[3]],[[4]],[[5]] As per the meta-analysis done by Bartindale et al., which included 11,140 patients
operated for TN, the overall prevalence of hearing loss was 8.25%.[[16]] Miller et al. reported a 2.3% incidence of permanent hearing loss after MVD.[[17]] As per Youn et al., the incidence of hearing loss was higher in elderly patients
(3.9% vs. 1.9%, P = 0.042).[[18]] In our study, the mean age of patients was 60.7 years and hearing loss incidence
was 3.3%. There is lots of variation in hearing outcomes after MVD in literature.
The use of routine perioperative audiogram and intraoperative BAEP had a significant
impact on hearing loss detection. HFS studies in which routine perioperative audiogram
was done reported 2.27% higher incidence of hearing loss as compared to studies not
using perioperative audiogram routinely.[[16]]
Stretching of root due to excessive cerebellar retraction is the main cause of hearing
loss in MVD. Lee et al. demonstrated a change in BAEP during cerebellar retraction
in all 12 patients who developed permanent hearing loss.[[6]] Intraoperative use of rigid retractors, especially for longer duration, should
be avoided.[[19]],[[20]] Suction cannula can be used for gentle retraction of the cerebellum. Early CSF
drainage also helps in relaxing the cerebellum and minimizing the need for retraction.
Even in the cases where retractors are required, it should not be applied along the
VIII nerve as it may produce tension on the REZ of nerve, leading to hearing disturbances.
Lee et al. studied the correlation between cerebellar retraction time and intraoperative
BAER changes and emphasized the greater distance between the cerebellar surface of
petrous temporal bone and point of neurovascular conflict as an important risk factor
for SNHL.[[21]]
Spasm of labyrinthine artery is another major cause of hearing loss after MVD. Morawski
et al. demonstrated the use of topical papaverine in the reversal of internal auditory
artery vasospasm and its impact on cochlear nerve functions in animals.[[22]] Scavo et al. also reported a prophylactic effect of diluted papaverine in preventing
hearing loss during MVD for trigeminal neuralgia.[[23]]
Intraoperative BAEP is a useful monitoring adjunct in posterior fossa MVD surgery.[[16]],[[24]] As per the American Society of Neurophysiology monitoring recommendations, surgeon
should be alerted when the wave V amplitude decreases by >50% or latency increases
by >1.0 ms.[[24]] However, Dannenbaum et al. reported 114 cases of MVD without using BAEP monitoring,
in which they found comparable results with those studies using monitoring.[[4]]
The intraoperative use of endoscope is quite helpful in the visualization of vessel
loop near REZ. It also helps in minimizing the cerebellar retraction and related complications.[[25]] Indocyanine green (ICG) angiography can be helpful intraoperatively for delineating
neurovascular conflict in the cerebellopontine angle cistern. We routinely use DIVA
during MVD.[[26]]
Limitations
We acknowledge certain limitations of our retrospective study. The most important
is the small number of cases as compared to the other studies. Perioperative audiogram
had not been performed routinely, which might have changed our results.
Conclusion
MVD is a safe and effective surgical technique for hemifacial spasm. Hearing loss
is one of the most disastrous complications having a significant impact on patients'
life. The incidence can be minimized using proper surgical techniques and various
intraoperative adjuncts such as BAER monitoring, use of endoscope, and ICG or DIVA.
