G Model

NEUCHI-924; No. of Pages 11 ARTICLE IN PRESS

Neurochirurgie xxx (2017) xxx–xxx

Disponible en ligne sur

ScienceDirect
www.sciencedirect.com

Original article

Microvascular decompression for hemifacial spasm : Surgical

techniques and intraoperative monitoring
M. Sindou a,b , P. Mercier c,∗
a
University Lyon 1, domaine Rockefeller, 69000 Lyon, France
b
Groupe ELSAN, clinique Bretéché, 44000 Nantes, France
c
CS74521, department of anatomy, faculté de Santé, 28, rue Roger-Amsler, 49045 Angers cedex 1, France

a r t i c l e i n f o a b s t r a c t

Article history: Primary hemifacial spasm with few exceptions is due to the vascular compression of the facial nerve that
Received 22 January 2018 can be evidenced with high resolution MRI. Microvascular decompression is the only curative treatment
Received in revised form 22 February 2018 for this pathology. According to literature review detailed in chapter “conflicting vessels”, the compres-
Accepted 6 April 2018
sion is located at the facial Root Exit Zone (REZ) in 95% of the cases, and in 5% distally at the cisternal or
Available online xxx
the intrameatal portion of the root as the sole conflict or in addition to one at brainstem/REZ. Therefore,
exploration has to be performed on the entire root, from the ponto-medullary fissure to the internal
Keywords:
auditory meatus. Because microvascular decompression is functional surgery, the procedure should be
Hemifacial spasm
Neurovascular conflict
as harmless as possible and with a high probability of permanent efficacy. Besides facial palsy, main com-
Microvascular decompression plications are hearing loss, tinnitus and gait disturbances. Causes are cochlea/labyrinth ischemia due to
Surgical techniques manipulations of their nutrient arteries and/or stretching of the eight nerve complex. To minimize the
Neuromonitoring latter, the approach should not be with lateral-to-medial retraction of the cerebellar hemisphere, but
along an infra-floccular trajectory, from below. In fact, most of the neurovascular conflicts are situated
ventro-caudally to facial REZ at the brainstem, particularly those from a megadolicho-vertebrobasilar
artery and its posterior inferior-cerebellar branch. Also, care should be taken not to cause any injury of
the manipulated vessels or stretching of their perforators to brainstem. Heating from bipolar coagula-
tion must be avoided. The inserted material used to maintain the offending vessel(s) away must not be
neo-compressive. Intraoperative neuromonitoring is considered to be useful for achieving safe surgery at
least until the learning curve has reached an optimal level, particularly BrainstemAuditory Evoked Poten-
tials recordings. Increase in latency and/or decrease in amplitude of wave V warn excessive stretching
or damage to the cochlear nerve, and decrease in amplitude of wave I signals possible ischemia of the
cochlea. Free-running EMG of the facial muscles may warn against excessive manipulation of the facial
nerve. Recording of the lateral spread responses – which are a sign of hyperexcitabilty of the facial motor
system – may provide information on completeness of the decompression.
© 2018 Elsevier Masson SAS. All rights reserved.

The principles of surgical techniques for hemifacial spasm
(HFS) are the same as for all other cranial nerve (CN) hyperac-
tive syndromes. MicroVascular Decompression (MVD) should be
performed from the brainstem to the entry from/exit to the corre-
sponding skull base foramen. Decompression should be atraumatic
for the root and harmless for the neighboring vasculature. Main-
taining the conflicting vessel(s) away must not be neo-compressive.
Good knowledge of the anatomy of the potentially offending ves-
sels is a prerequisite for surgery of the facial nerve (FN) affected
with a neurovascular conflict (NVC) (see corresponding chapter
on surgical anatomy). Intra operative monitoring can be useful for
effectiveness and safety of facial MVD.
1. Part 1: Surgical techniques
This first part of the chapter details the techniques that we are
currently using as shown in the schematic drawings of Figs. 1–3.
Main alternatives are given in italic characters.
1.1. Installation and positioning

The patient is installed in the lateral decubitus position, head

∗ Corresponding author. slightly flexed and fixed in a three-pin headholder, the chin approx-
E-mail address: philippe.mercier@univ-angers.fr (P. Mercier). imately two fingers from the sternum, with contralateral rotation

https://doi.org/10.1016/j.neuchi.2018.04.003
0028-3770/© 2018 Elsevier Masson SAS. All rights reserved.

Please cite this article in press as: Sindou M, Mercier P. Microvascular decompression for hemifacial spasm : Surgical techniques and
intraoperative monitoring. Neurochirurgie (2017), https://doi.org/10.1016/j.neuchi.2018.04.003
G Model
NEUCHI-924; No. of Pages 11 ARTICLE IN PRESS
2 M. Sindou, P. Mercier / Neurochirurgie xxx (2017) xxx–xxx
Fig. 1. Head positioning for microvascular decompression of facial nerve. Incision
(dotted curved line) encounters the occipital arterial vessels and should avoid tra-
jectory of the occipital nerve.
of 15◦ and lateral inflexion towards the opposite side, i.e. the floor.
Lateral inflexion is important so that the shoulder does not obstruct
surgeon’s vision. The shoulder is maintained pulled by a strap, but
with care to avoid brachial plexus stretching (Fig. 4). Absence of
compression on the operating table should be verified especially at
the axillary region where brachial plexus, axillary artery and venous
return are at risk. Peripheral nerves, namely radial, median, ulnar
nerves, should also be checked so they are not compressed. Risk
was considered such that it has been advocated by some authors to
monitor somato-sensory evoked potentials of both upper limbs to
warn against potential damage to the brachial plexuses and nerves
and prevent from incorrect positioning on installation [1].
Regarding patients’ position/installation many variants have been
described. As classically said, “the best one is the one that the surgeon
and the anesthesiologists are well-experienced with”. One very
different variant is the supine position with latero-basal approach, as
advocated among others by Shimizu et al. [2]. The patient is installed
in supine position, neck flexed, head rotated contralaterally in the
order of 40 to 60◦ , the vertex slightly tilted toward floor. Craniotomy
is located on the lateral and basal side of the occipital cranium.
Then a subfloccular approach enables the surgeon to reach the facial
nerve REZ and root, without the need of cerebellar retraction with
a spatula. For these authors this installation allows to escape the
“complicated” preparation for the lateral or park-bench position and
is more propitious to avoid using a retractor. However, our concern
is that the, needed, important rotation of the patient’s head entails a
significant risk of twisting the carotid and vertebral arteries and of
impairing the intracranial venous return.
1.2. Retromastoid–retrosigmoid approach
The surgical approach aims to target the facial REZ ventro-
caudally together with the ventrolateral aspect of brainstem.
This can be done through a small retromastoid bone flap or an
extreme-lateral “keyhole” craniectomy, in the order of 2 cm in
diameter, just posterior to the tip of the mastoid process so that
the facial nerve (FN) can be reached from below, passing infero-
laterally to the cerebellar hemisphere. Reaching the FN from below
along an infra-floccular route is important for two reasons. First,
NVCs are currently located ventro-caudally at the REZ. Second, a
Please cite this article in press as: Sindou M, Mercier P. Microvascular decompression for hemifacial spasm : Surgical techniques and
intraoperative monitoring. Neurochirurgie (2017), https://doi.org/10.1016/j.neuchi.2018.04.003
lateral–medial retraction of the cerebellar hemisphere would exert
stretching of the VIIIth nerve and lead to hearing loss, as demon-
strated by BAEP monitoring [3–6].
Landmarks for skin incision are traced taking into account tra-
jectories of the sensory nerves especially the great occipital nerve
and branches. After limited shaving, a linear oblique or a curved
skin incision is performed posterior to the mastoid (Fig. 2a). Then
after musculo-aponevrotic dissection with if necessary ligation and
division of the occipital artery at the bottom of the approach, the
craniotomy/craniectomy is carried out posterior to the tip of the
mastoid.
Bone opening is semilunar-shaped (2 cm × 1.5 cm) and posterior
to the sigmoid sinus, the edge of which is exposed to make the
approach as lateral as possible (Fig. 2b). However skeletonization
of the sigmoid sinus must be performed with caution because its
wall is fragile, often reduced to a thin endothelial layer. Minimally
invasive surgery could lead to serious bleeding. Its occlusion for
hemostatic reasons could be most harmful, particularly when the
lateral sinus is of the dominant type for the Superior Sagittal Sinus
(SSS) drainage [7]. While working adjacent to the sigmoid sinus,
care should be taken not to avulse the emissary vein from its sinus
insertion, which could produce significant blood loss and require
sinus repair, ideally with patching.
In a majority of cases the mastoid cells are opened; if so, they
should be tightly occluded on closure.
Then a curved dural incision is made (Fig. 2c) and the (small)
dural flap is reflected laterally based on the sigmoid sinus (Fig. 2d).
As soon as the dural flap is retracted by attaching it to the neigh-
boring periosteum, the cerebellum is exposed and protected by a
cottonoid with a thin piece of rubber interposed to avoid adhesive-
ness to the cerebellar surface (Fig. 2e).
Once the microscope is in position, the arachnoid is incised in a
punctiform manner and CSF slowly extracted to obtain progressive
cerebellar relaxation. If the intradural space remains insufficient,
even more if cerebellum herniates, adaptation of the ventilation
and pharmalogical measures are discussed with the anesthesiol-
ogist to reduce CSF and brain volumes and decrease intracranial
pressure. Generally CSF lumbar drainage is not necessary.
Depending on the surgeon’s preference the lower cerebello-
pontine angle (CPA) cistern is exposed by simply using the
micro-suction-tube as a mobile retractor or by applying a (nar-
row) self-retractor blade of the Sugita-Fukushima type, of 2 mm
in width at the tip. Then the arachnoid covering the Xth–IXth CNs,
the choroid plexus (that emerges from the lateral recess through
the foramen of Luschka) and the flocculus, is extensively opened
(Fig. 2f). Care must be taken not to tear bridging veins going from
the cerebellum surface to their dural drainage points, especially the
inferior petrosal vein. If at risk of rupturing, it is advised to coag-
ulate and divide it. Based on our experience, interruption of the
inferior petrosal vein, which most often is of a small caliber, has no
patent consequence.
Even more important, on opening arachnoid of the Cerebello-
Pontine Angle (CPA), in around half of the cases, the surgeon is
confronted with the sub-arcuate artery [8,9]. If stretching occurs
the risk is that the artery be avulsed at its dural entry point at
porus of the internal auditory canal. Provided this artery is well
differentiated from the one (or the two) labyrinthine artery(ies), it
is advisable to coagulate and divide it rather than causing its rup-
ture. In our experience there is no functional consequence by doing
it. Noteworthy, in some cases, the sub-arcuate artery has a common
trunk with the labyrinthine artery.
Because the majority of the offending vessels are located ventro-
caudally to the facial REZ basically at brainstem, trajectory should
be from below, inferior-laterally to the cerebellar hemisphere
(Fig. 5). In most cases the flocculus does not need to be significantly
mobilized and therefore dissected free. Because the flocculus is
G Model
NEUCHI-924; No. of Pages 11 ARTICLE IN PRESS
M. Sindou, P. Mercier / Neurochirurgie xxx (2017) xxx–xxx 3

Fig. 2. Various steps of the retromastoid–retrosigmoid approach for microvascular decompression of the facial nerve. Approach should primarily target the facial Root Exit
Zone (REZ), ventro-caudally from below, via an infra-floccular route; a: landmarks of the skin incision and of craniectomy/craniotomy; b: craniotomy/craniectomy should
be enlarged laterally by removing bone to the posterior margin of the sigmoid sinus; c: dural incision; d: dural flap reflected as laterally as possible; e: infra-floccular route
to reach facial REZ, after successively dissecting the lower cranial nerves; f: the choroid plexus and the flocculus.

Fig. 3. Decompression of the facial REZ on right side; a: exposure of the facial REZ compressed in this illustration by a voluminous loop of the posterior inferior-cerebellar
®
artery (PICA); b: transposition of PICA with interposition of a piece of (semi-rigid) knitted Teflon between the facial REZ (and also the VIIIth nerve) and the PICA loop.

generally strongly adhesive to the eight nerve, its excessive retrac-
tion would result in stretching the nerve and consequently provoke
vestibular and hearing disturbances. This has been demonstrated
by BAEPs recordings under the form of an increase in latency of
peaks III/V [5].
In the eventuality of an hypertrophied flocculus that would
make exposure of the facial REZ difficult, its separation from the
eight nerve can facilitate access to the REZ. When dissection of the
flocculus from the cochleo-vestibular complex is undertaken, the
surgeon must be very careful not to compromise the tiny vascula-
ture of the nerves and not to injure the nutrient vessels which arise
from the surrounding arteries or the surface of the cerebellum to
the eight complex. Also, diffusion of heat from the bipolar coagu-
lation must be absolutely avoided. If detachment of the flocculus
from the eight nerve is considered too hazardous, partial resection
of the flocculus would be preferable.
If the choroid plexus is found to be hypertrophied and
makes an obstacle, as in 5% of our cases (unpublished data
from personal operative observations), its coagulation and resec-
tion after shrinkage may help to better access the facial
REZ.
After the flocculus and choroid plexus are elevated and the pia-
arachnoid filaments of the cistern divided, the ponto-medullary
fissure is reached with the REZ of the eight complex posteriorly and
the ventro-caudal aspect of the facial REZ together with the adja-
cent brainstem anteriorly. Because the offending vessels almost
always compress the facial REZ ventro-caudally,at level of the
cerebello-medullary cistern, access should be made from below,

Please cite this article in press as: Sindou M, Mercier P. Microvascular decompression for hemifacial spasm : Surgical techniques and
intraoperative monitoring. Neurochirurgie (2017), https://doi.org/10.1016/j.neuchi.2018.04.003
G Model
NEUCHI-924; No. of Pages 11 ARTICLE IN PRESS
4 M. Sindou, P. Mercier / Neurochirurgie xxx (2017) xxx–xxx

Fig. 4. Installation in lateral decubitus position with anesthetic equipment for retrosigmoid surgery of HFS. Upper left: note the contralateral inflexion of head and neck to
the left side, and traction on the right shoulder to access facial REZ from below, so that vision of surgeon is not obstructed by the shoulder. In Lyon University installation,
left contralateral upper limb is placed under level of the operating table, as seen in the corresponding vertical photography. Lower left: note traction on the right shoulder
to avoid obstruction of surgeon’s vision. In Angers University, installation, left contralateral upper limb is placed on the operating table.

Fig. 5. Surgical approach of the facial REZ for left HFS; a: first step is exposure of lower cranial nerves, successively XIth, Xth and IXth rootlets; b: retraction of choroid plexus
(ch. pl.) andthe flocculus (Fl) to access facial REZ (not seen) compressed by PICA (VIIIth: cochleo-vestibular nerves); c: PICA loop is pushed away caudally with a microsucker,
in order to free the facial REZ (visible: *); d: recommended instruments for keyhole vascular decompression: fine retractor spatula of Sugita-Fukushima type, microdissector
and fine curved microscissors in addition to microsuckers (not photographed).

in between the IXth rootlets and the inferior aspect of the VIIIth
nerve REZ (Fig. 6).
The conflicting arteries which are the AICA, the PICA and/or
the VBA must be identified (which is often difficult) then care-
fully transposed laterally not to damage their perforating branches
to the brainstem. Since these arteries are often atherosclerotic,
their mobilization may be dangerous, especially when perforators
arise from the offending loops. (See: chapter on the conflicting
vessels).
When the conflicting vessel(s) is (are) located more distally,
MVD is more problematic. Transposition and maintaining away can
be difficult without exerting a certain degree of neo-compression
to the neighboring neural structures. The way to secure vessels in
place depends on the anatomical features of each particular neu-
rovascular conflict and is a tailored operation.
Dissection of neural structures and manipulation of vessels
require adapted microsurgical instruments as shown in Fig. 5d and
6b.
1.3. Vascular decompression
For decompression we routinely have recourse to three different
types of procedures depending on the anatomical situations that
are encountered:
• when elongated arteries have ample loops and are not arte-
®
riosclerotic, we used traction on the artery with Teflon tapes
passed around the loops (sling effect). Extremities of the tape
are then blocked by squeezing them in between the surface of
the cerebellum and the posterior aspect of the petrous pyramid.

Please cite this article in press as: Sindou M, Mercier P. Microvascular decompression for hemifacial spasm : Surgical techniques and
intraoperative monitoring. Neurochirurgie (2017), https://doi.org/10.1016/j.neuchi.2018.04.003
G Model
NEUCHI-924; No. of Pages 11 ARTICLE IN PRESS
M. Sindou, P. Mercier / Neurochirurgie xxx (2017) xxx–xxx 5

Fig. 6. Microvascular decompression on left side of a neurovascular conflict by a megadolicho-posterior-inferior-cerebellar artery (PICA). Decompression should start from
below, to dislodge PICA from the cerebello-medullary fissure and brainstem, passing in between the IXth (seen) and the Xth (not seen) rootlets (a). After detachment of the
PICA from the facial REZ, one can see a greyish zone in the REZ that corresponds to a focal demyelination due to the chronic compression by the PICA loop (b). PICA will be
® ®
maintained apart by inserting a piece of 15 × 10 mm of (semi-rigid) knitted Teflon (c); The Teflon piece lies in between anteriorly: the ventrolateral aspect of brainstem
®
ventrally and the flocculus dorsally, and the PICA posteriorly (d). Thus the Teflon piece bridges the facial REZ and is not compressive for the VIIIth nerve complex either.

® ®
Tapes are made of Teflon felt fibers teased from the original
®
pieces of Teflon (PTFE felt). Tapes are approximatively 4 cm in
length × 3 mm in width. Inconvenience is some difficulty to main-
tain the artery pulled enough away from the facial nerve so that
the vessel does not come back in the initial harmful situation;
• when loops have short perforators to brainstem, we use ball(s)
®
of Teflon – approximatively 5 to 7 mm in diameter – interposed
between the transposed vessel and the facial REZ. Inconveniences
are that balls may become flat and still transmit pulsations or, if
too big and compact, exert a neo-compression to the facial and
also the VIIIth CNs;
• our preferred procedure – that we find especially adapted for
compression of the REZ by PICA, AICA, VBA or several of those
types, especially when vessel(s) originate from a ventro-caudal
location(which corresponds to the commonest encountered
situation)– is as follows. It consists of inserting a rectangular
plaque of approximatively 15 × 4 mm in between the transposed
compressive artery(ies) on one side and on the other side the
lateral aspect of the brainstem anteriorly and the flocculus poste-
riorly (Fig. 6c,d). By doing so, the prosthesis bridges the VIIth and
VIIIth REZ without any compression of the nerves. This bridging
®
plaque can be semi-rigid if taken from knitted Teflon (Edwards
Outflow Tract Knitted PTFE) or more supple if taken from knitted
®
Dacron (Sauvage Knitted Polyester). These pieces should not be
taken from Teflon felt because of the tendency of felt when used
as plaque to aggressively adhere to the neural structures. Felt
is only used in our practice for making tapes or balls, by teas-
ing constituting fibers. Fig. 7 shows the main variety of implant
materials.
There are as many ways to maintain decompression as authors
reporting:
when using a piece of Teflon felt especially when implant is
touching the nerve [12,13]. We could observe these types of
harmful adherences not only to the facial nerve but also to the
VIIIth nerve due to repeated surgery in patients operated else-
where. Interposition technique may in addition be compressive
for the neural adjacent structures;
• the sling transposition method consists of passing ribbon-shaped
tapes around the offending vessel and to attach them to the
dural wall of the CPA. Slings may be made of autologous free
graft like fascia or of synthetic material. The advantage is that no
material touches the nerve, but drawback is that the sling may
not sufficiently pull away the arterial loop so that the artery may
return to its conflicting position. A pedicled dural flap can be
used, as successfully reported in seven cases [14]. Drawback of
this elegant technique is that it involves a certain surgical risk;
• transposition using surgical glue may be used especially when
the offending vessel is arteriosclerotic. After separation from
the facial nerve is completed, the artery is stuck to the dura of
the petrous pyramid using a biological glue. EC Transglutami-
nase biological glue (for tissue surface adhesion) has a higher
adhesive strength than the currently used fibrin biological glue.
A few cases with successful outcome have been reported to
date [15,16]. Advantage is that there is no implanted material
touching the nerve. Inconvenience is that long-term solidity of
the glue attachment cannot be warranted.
In conclusion, regarding this section there is still no consen-
sus on the best way to use materials for decompression. To our
knowledge no systematic comparative studies either experimen-
tal or clinical have been reported in the literature regarding this
subject.

• interposition of an implant is the most commonly used tech-
nique. It is called “Isolation method” [1]. The implant is placed in
between the conflicting vessel(s) and the facial nerve [10,11]. The
advantage is its simplicity. Main drawback is the development
of adhesions to the nerve that may be responsible for recurrence
1.4. Implant material for Interposition/Transposition
During the past several decades of the history of MVD-surgery,
many technical tools have been attempted to achieve the most
effective and safest decompressive effect.

Please cite this article in press as: Sindou M, Mercier P. Microvascular decompression for hemifacial spasm : Surgical techniques and
intraoperative monitoring. Neurochirurgie (2017), https://doi.org/10.1016/j.neuchi.2018.04.003
G Model
NEUCHI-924; No. of Pages 11 ARTICLE IN PRESS
6 M. Sindou, P. Mercier / Neurochirurgie xxx (2017) xxx–xxx

®
Fig. 7. Some types of material to implant: a left: Shredded PTFE Teflon felt prepared by teasing constituting fibers (upper view), to make balls or ribbons (lower view).
Simple pieces (15 × 8 mm in this illustration) are not recommended because of their propensy for (harmful) adhesiveness to neurovascular structures; b center: Edwards
®
Outflow Tract knitted PTFE Teflon is semi-rigid material. Picture shows insertion in between the PICA to maintain it away and the facial REZ as well as the VIIIth complex.
The plaque (10 × 15 mm) is placed on the ventrolateral aspect of brainstem (pons) ventrally and the flocculus dorsally, so that it has a “bridging effect” over the facial and
® ®
VIIIth cranial nerves; c right: Sauvage knitted Polyester Dacron is a more supple material than the two others.

Simple opening of the arachnoid to provide more space in the
CPA cistern for the neurovascular structures, with simple separa-
tion of the conflicting vessel(s) – although the most elegant – has
been shown to be insufficient in the majority of cases.
Implants can be used either for the simple interposition between
the vessel and the nerve, or to maintain the conflicting vessel
apart (=transposition technique). Prevention of arteries to become
conflicting again is important as arteries are “shape-memory”
structures that tend to return to their previous offending position.
There has been a long list of implanted material trials:
• absorbable gelatin (Gelfoam*, Pfizer), knitted cellulose (Surgicel
*, Johnson and Johnson) and fibrin glue as well, resorbed within
only a few days;
• autologous fat or muscle, that was frequently utilized, was found
to generate fibrosis and adherences;
• ivalon sponge (Unipoint Industries, Hight Point, NC), after large
use in the US, was shown to be aggressive and neo-compressive
for the neighboring neural structures;
• in spite of the absence of experimental studies in MVD indica-
®
tion, shredded Teflon Felt (PTFE felt, C.R. Bard Co, Bard Implants
Division, Billerica, MA) has become the preferred product for
®
MVD. To be recalled TEFLON-PTFE means Polytetrafluorethy-
lene. Introduced for use in the treatment of paralyzed vocal cords,
®
TEFLON-PTFE is a biologic implant on the market for the past 40
®
years. Teflon for cardiovascular patching is the same product
®
used for MVD. According to literature, Teflon Felt appears to
be a good inert material [17]. However, a few publications have
reported inflammatory reactions or granulomas [13,18–28].
®
Teflon Felt utilized in the form of a compact piece may gener-
ate strong adhesions to the adjacent neurovascular structures due
to its filamentous nature. This has been reported in the literature
concerning repeated surgery, and is also our observation.
®
The most current way of using Teflon felt is to tease its fibers
from the felt; this can be done with a pair of forceps or by pulling
on the fibers between the gloved thumb and index fingers, which is
not easy (Fig. 7a). Then the fibers, after being moistened with saline
solution, can be shaped into different tailored forms according to
different situations: rolled into a ball or in an oval or peanut shape,
more or less compact and of various sizes. Fibers can be agglomer-
ated into ribbon-shaped bundles (of ± 2 mm in width and 2 to 4 cm
in length). These ribbon-shaped tapes are passed around the ves-
sel loop(s) to be transposed; they are used to exert a pulling effect
and to maintain the vessel away so that it has no contact with the
decompressed root.
®
In addition to the shredded Teflon felt material, that we are
preferentially using, two other products, of the knitted type, can be
utilized as plates (plaques) for interposition:
• the “Edwards Outflow Tract Knitted PTFE” when a relatively rigid
plate is needed (Fig. 7b);
• the “Sauvage knitted Polyester DACRON® ” when a more supple
sheet is preferable. (Fig. 7c).
Both products are manufactured by BARD Peripheral Vas-
cular Inc. 1625 West 3rd street, Tempe, AZ 85281, USA
(www.bardpv.com).
It is important to underline that until now there has been no real
consensus on the types of implant material to be used.
1.5. Closure
After decompression completion, it is important to check that
the transposed and manipulated vessels are not kinked or com-
pressed. If arteries manifest spasm, a few droplets of papaverine
in a saline solution, with a dilution of 10%, are topically applied.
Not too many droplets should be used as papaverine has a very
acid pH, in the order of 2.8. This might be at least temporarily toxic
for the neurovascular structures, as we observed in our first cases.
In addition to its beneficial effect on spasm, the dilatation effect
of papaverine, together with irrigation with warm saline solution,
helps to assure the quality of arteriolar hemostasis.
Jugular veins are then compressed at neck and also Valsalva
maneuvers are performed to check the venous hemostasis. At the
end, still under the microscope, CPA cistern is irrigated with warm
saline solution until one can be certain that CSF remains clear.
There are almost as many ways for closure as teams dealing with
this type of surgery! However,everyone agrees that dural closure
should be watertight. Due to some degree of dural retraction, this is
not easy when the craniectomy is small. To achieve tightness some
surgeons advise to use artificial dura together with biological glue
In our procedure we extradurally fix a piece of aponevrosis and fat
over the suture line. When mastoid cells have been largely opened,

Please cite this article in press as: Sindou M, Mercier P. Microvascular decompression for hemifacial spasm : Surgical techniques and
intraoperative monitoring. Neurochirurgie (2017), https://doi.org/10.1016/j.neuchi.2018.04.003
G Model
NEUCHI-924; No. of Pages 11 ARTICLE IN PRESS
M. Sindou, P. Mercier / Neurochirurgie xxx (2017) xxx–xxx 7

the piece has to be large; we prefer to harvest fascia-lata and fat at

the thigh rather than in the retromastoid region that would entail
the risk of local tissue devascularization and also of injury of the
occipital nerve branches.
If opened, mastoid cells should be tightly sealed. Most teams use
wax. We prefer to occlude the cells by inserting or apposing pieces
of fat. We have observed that impacting wax, with or even with-
out putting back the bone powder, was not infrequently followed
by local inflammatory reactions and granulomas, and secondarily
repetitive acute otitis.
To treat the bone defect the majority of surgeons routinely put
back the bone flap or the bone powder or perform a cranioplasty, on
the basis that this “dramatically decreases the incidence of postop-
erative CSF leaks and also headaches” [1,29,30]. Until recently our
policy was to put back the bone powder inside the key-hole only
when the mastoid cells were estimated not opened. At present, ten-
dency is to put back the bone powder even when cells were opened,
provided cells are considered satisfactorily occluded.
After tight closure of craniotomy/craniectomy, wound is washed
with Sodium hypochlorite. Then the musculo-aponevrotic, sub-
cutaneous and cutaneous layers are successively closed by Fig. 8. Manipulations of VIIIth nerve and/or nutrient vessels are manoeuvers at risk
interrupted sutures. Finally, a protective bandaging is applied on during MVD of cranial nerves in the cerebello-pontine angle. Figure shows sudden
decrease in amplitude of wave I (arrow), with subsequent disappearance of wave III
the operated region. A certain degree of compression is wise to pre-
and V. Such a BAEP event likely resulted from (severe) cochlear ischemia or (major)
vent “dead-spaces” and therefore minimize the risk of CSF leak. In indirect injury (rupturing?) of the cochlear nerve at the level of its junction with
our practice a soft cervical collar to exert nuchal and retromastoid cochlea at fundus of internal auditory canal. Patient remained with deafness.
compression is offered to the patient for the following ten days.
2. Part 2: intraoperative monitoring
1.6. Retromastoid–retrosigmoid approach variant
Intraoperative monitoring is an important part of this chap-
An alternative to keyhole craniectomy is to perform a bone flap
ter: Good knowledge of microsurgical anatomy, advanced imaging
of larger size: 4 × 4 cm, posterior to the tip of the mastoid process.
and skills in microsurgical/endoscopic techniques have made
This latter way is the preferred approach by one of the authors
MVD-surgery an effective method for treating cranial nerve (CN)
(PM). Reasons are both considerable experience of this type of
neurovascular compression (NVC) syndromes in the CPA region.
approach for CPA tumor surgery and the following surgical argu-
Information from intraoperative neuromonitoring (ION) has been
ment. Because a craniotomy with bone flap is bigger than a key-hole
recognized to help preventive functional damage to those nerves in
opening, the exposed dura is larger and dural closure easier. If
anesthetized patients. This is particularly true when dealing with
properly moistened during surgery, dural retraction is reduced and
decompressive surgery of the facial nerve (FN), known to have a
tightness of closure is easier, by simple suturing.
close relationship with the cochleo-vestibular neurovascular com-
The bone flap is turned posteriorly to the tip of the mastoid
plex. Also it has been shown that the effect of MVD on the hemifacial
and extended laterally and inferiorly by drilling off the bone to the
spasm (HFS) can be tested intraoperatively by EMG-recordings of
sigmoid sinus, which allows to reach the facial nerve from below
the abnormal responses, especially the so-called Lateral Spread
passing inferior-laterally to the cerebellar hemisphere. Then the
Responses (LSR) which are indicative of the degree of hyperex-
dura is obliquely opened from the inferior-medial corner to the
citability of the altered facial motor system.
superior-lateral corner in the form of a lateral flap which is divided
in two and sutured to the neighboring periosteum. The medial
flap protects the cerebellum and limits the possibility of cerebellar 2.1. Brainstem Auditory Evoked Potentials (BAEPs)
retraction.
Then the dura is sewed all around the bone flap opening. If Hearing loss (Fig. 8) and auditory-vestibular disturbances,
opened, mastoid cells are tightly sealed. We do not use wax since among which tinnitus, gait imbalance and vertigo dominate, are
an histological study on the occasion of mandible surgery showed the most frequently encountered CN complications in CPA surgery.
that wax produced local chronic inflammation (personal communi- The cochlear-vestibular neurovascular structures may be dam-
cation from maxillo-facial surgeons). We prefer to occlude the cells aged through various mechanisms:
by inserting bone powder covered by pieces of fat and biological
glue. • Stretching of the VIIIth nerve due to excessive retraction of the
To treat the bone defect, we routinely put back the bone flap flocculus;
and maintain it by sutures. Bone powder is used to fill the defects • Manipulation of the labyrinthine artery or the anterior inferior-
produced by the trepan opening. Finally the musculo-aponevrotic cerebellar artery, its parent artery, that can generate vasospastic
and subcutaneous layers are successively closed with interrupted reactions;
sutures, and the skin with intradermic resorbable thread. A pro- • Direct trauma by an instrument or heating due to nearby coagu-
tective bandaging is applied on the operated region with a certain lation;
degree of compression during 3 to 4 days to minimize the risk of • And at end of surgery neo-compression of the VIIIth CN by an
CSF leak. interposed material inserted between the offending vessel and
In conclusion, meticulous technical precision is required to the VIIth–VIIIth complex [31].
achieve effective and safe surgery at same time. Complications
should be of minimal probability in this delicate functional neu- In current BAEP monitoring the waves most classically studied
rosurgical field. are the following:

Please cite this article in press as: Sindou M, Mercier P. Microvascular decompression for hemifacial spasm : Surgical techniques and
intraoperative monitoring. Neurochirurgie (2017), https://doi.org/10.1016/j.neuchi.2018.04.003
G Model
NEUCHI-924; No. of Pages 11 ARTICLE IN PRESS
8 M. Sindou, P. Mercier / Neurochirurgie xxx (2017) xxx–xxx
Fig. 9. BAEP recording during MVD for HFS. Note the clear-cut increase in latency
of wave V soon after application of the retractor (arrow) and also its decrease in
amplitude although more difficult to evidence from baseline. This pattern is very
representative of an excess of stretching of the VIIIth nerve in its cisternal portion.
Note the slow tendency to normalization after retraction off.
• Wave I, the source of which is the cochlea-distal part of the
cochlear nerve junction and,
• Waves III and V, generally linked, which are generated in brain-
stem at the superior olivary nucleus and the inferior colliculus,
respectively. In our practice we only considered wave V for sim-
plicity purposes.
The warning-signals for hearing function at risk should be given
by the neurophysiologist – while watching the neurosurgeon on
video – on real time. However this warning from neurophysiolo-
gist should not be manifested too abruptly as not to disrupt the
surgeon. Although there are no official basic criteria as regards any
recommendations, and remaining controversies, we adopted the
following process.
• Wave V
Studies of correlations between outcome on hearing and intra-
operative modifications of BAEPs have shown that auditory
function is at risk when amplitude of wave V decreases in the
order of 50% [32] or latency increases in the order of 1 msec [5].
Warning-signals should be given either when amplitude progres-
sively decreases to 30%–50% or when latency evolves from + 0.5
to + 1 msec., in the absence of anesthetic causes for these mod-
ifications. Some authors consider that a decrease in amplitude
is a more reliable indicator of hearing at risk [33,34], while
other authors suggest that it is an increase in latency [5,6,35,36].
Whatever the case, because of a likely excess of stretching, the
surgeon should stop all manoeuvers, release retraction, irrigate
with warm saline solution and wait for normalization of BAEPs
before resuming surgery (Fig. 9). Also neo-compression of the
cochlear nerve by the interposed material can be the cause of a
delay in wave V; if so, material should be repositioned.
• Wave I
Even a slight decrease in the amplitude of wave I, if persistent,
should be considered a warning-signal of cochlear ischemia. This
may happen on manipulating the anterior inferior-cerebellar
artery and/or the labyrinthine artery, its branch to the cochlea. In
this eventuality arteries must be irrigated with warm saline solu-
tion and if a vasospasm is observed some droplets of papaverine
applied on the vasospastic segments. Generally the spasm
Please cite this article in press as: Sindou M, Mercier P. Microvascular decompression for hemifacial spasm : Surgical techniques and
intraoperative monitoring. Neurochirurgie (2017), https://doi.org/10.1016/j.neuchi.2018.04.003
disappears within a few minutes and the amplitude of wave I
returns to its normal baseline value soon afterwards [5] (Fig. 10).
Wave I observation may also be useful to check integrity of
the vestibular CN. An increase in amplitude could likely corre-
spond to impairment in vestibular conduction. We observed such
an increase when an interposed prosthesis exerted a compressive
effect on that nerve [5] (Fig. 11).
* Practice of MVD either for trigeminal neuralgia or hemifa-
cial spasm under BAEPs monitoring has taught us that the main
harmful manoeuver was lateral-to-medial retraction of the cere-
bellum, with as its consequence stretching of the VIIIth nerve [37].
We also learned that this could be circumvented by approaching
the Vth nerve for trigeminal neuralgia through an infratentorial-
supracerebellar trajectory, and for the VIIth–VIIIth complex, as well
as for the IX–Xth nerves in cases with vago-glossopharyngeal neu-
ralgias, via an infra-floccular approach. Since those principles were
followed, hearing and vestibular complications were low, so that
we considered that BAEPs monitoring was not anymore mandatory
except in some rare clinical conditions, such as in patients with pre-
operative hearing deficit or in risky anatomical situations that may
be predicted on pre-operative imaging explorations.
2.2. EMG-recordings
For EMG-recordings, as well as for BEAPs monitoring, anesthe-
siologists should be aware of the surgical team’s purposes in order
to adapt the anesthetic protocol accordingly. For EMG-recordings
myo-relaxants must be avoided or only those short-lasting can be
used.
• free-running – EMG-recording (fr-EMG) consists of recording
spontaneous muscular activity in the main facial muscles, namely
the orbicularis oculi, orbicularis oris and mentalis. Addition of a
loudspeaker is most useful to alert the surgeon when provoking
certain abrupt manoeuvers. This is especially true when firing
tends to last after cessation of manipulations.
• Triggered-EMG-recording, also with audio-system, consists of
direct stimulation of the FN (with pulse duration of 0.1–0.2 msec,
frequency of 3 ± 1 Hz and intensity ranging from 200 micro-A to
1–2 milli-A) and of recording elicited Compound Muscular Action
Potentials (CMAPs).
This technique is not very useful in MVD-surgery as opposed
to surgery for tumors in which identification of the FN can be
difficult. However the method can be used to verify the integrity
of the nerve after “difficult” decompression or to ascertain that an
eventual interposed prosthesis is not excessively compressive.
• The Lateral Spread Response (LSR) phenomenon corresponds to
the fact that the stimulation of one branch of the FN produces
EMG-responses not only in the muscles innervated by the stim-
ulated branch but also in the muscles innervated by the other
branches.
This phenomenon – considered typical of primary HFS [38] is
thought to be related to ephaptic transmission at the lesion site,
alone [39] or in combination with motor nucleus induced hyper-
activity [40]. LSR has been particularly studied intraoperatively
by Moller and Jannetta [41].
In Moller’s technique the stimulation electrodes are placed
at the temporal branch and the mandibular branch of the FN.
Recording electrodes are inserted subcutaneously into the men-
talis muscle and the orbicularis oculi muscle. Recommended
parameters of stimulation are: pulse duration of 0.1–0.2 msec,
frequency of 3 ± 1 Hz, intensity of 1–2 mA to up to 5–15 mA.
Abnormal responses are recorded in the mentalis muscle when
the temporal branch is stimulated and in the orbicularis oculi
muscle when the mandibular branch is stimulated (Fig. 12).
G Model
NEUCHI-924; No. of Pages 11 ARTICLE IN PRESS
M. Sindou, P. Mercier / Neurochirurgie xxx (2017) xxx–xxx 9

Fig. 10. BAEP recording during MVD for HFS; a: retractor on; b: retractor off. Note the transient decrease in amplitude of wave I on trace 1 (arrow) likely to correspond to
a vasospasm of the labyrinthine artery, as illustrated in the operative upper view (1). Also note the progressive reappearance of normal amplitude of wave I since trace 2
(arrow) soon after spasm disappeared after irrigation with papaverine in warm saline solution as seen in the lower operative view (2).

Fig. 11. BAEP recording during MVD for HFS. Note the increase in amplitude of
wave I (arrow on punctuated trace compared to initial trace). Interpretation was
that the vestibular nerve was compressed by the material implanted in between the
conflicting arteries (PICA and VBA) and the VII–VIII nerve complex.
In 1987, Moller and Jannetta advocated routine intraoperative
EMG-recording of LSR to ensure that adequate decompression is
achieved [41]. They concluded that spasms are likely to persist
if LSR is still observed at the end of the procedure. In contrast,
the absence of LSR indicates that spasms are likely to disappear.
If a majority of authors favor LSR monitoring [42–48], other
authors including our team do not find it very reliable [49–53].
As a matter of fact LSRs may disappear due to manipulations of
the FN. However LSRs may persist at the end of surgery in spite
of an apparently valuable decompression whilst patients have a
(more or less delayed) total relief of their spasm [49]. It has been
stressed on the particular usefulness of LSR-recordings in the
cases in whom failure necessitate to repeat surgery [54,55].
• Free-running EMG under saline irrigation can be used to evaluate
excitability of the FN [56]. Background is that activity gradually
increases, in amplitude and frequency, 20 to 30 sec following
saline irrigation and gradually decreases to baseline level 1
to 2 min after completion. To be mentioned EMG-responses
are much larger with saline than with lactic solution or the
CSF, especially in the mentalis muscle. Fr-EMG can be used to
compare pre- and post- MVD activity. In the Fukuda et al. study,
patients who did not exhibit decrease of fr-EMG activity ratios in
response to saline irrigation tended to have persistent abnormal
Lateral Spread Responses, despite successful completion of
MVD-surgery. This test is rarely employed in current practice.
• The ZLR-response
The ZL EMG response to stimulation was first studied by Zheng
et al. [57] and proposed by Zhong and Li as a complement of
the LSR-EMG monitoring to determine the real culprit vessel(s)
when several offending vessels are suspected [58].
The technique consists of placing a (concentric) stimulating
electrode on the offending vessel and to stimulate with the same
parameters as the ones used for the LSR-technique. Decompres-
sion is considered effective when EMG activities disappear. In our
experience this method is similar to an indirect triggered-EMG
response. Recently the combination of LSR and ZLR recordings
has been advocated especially in cases with multiple offending
vessels [59].
2.3. Other methods
Several other methods that belong to the general ION arma-
mentarium [60] have been probed in MVD-surgery, among them:
recordings of the facial F-waves, the blink-reflex, the facial motor
evoked potentials, especially for verifying FN integrity. Interesting
for research, these methods did not become of use for HFS surgery.
Somato-Sensory Evoked Potentials (SSEPs) monitoring of the
upper limbs has been proposed for checking safety of patients’
installation [61]. Heavy to handle, SSEPs recordings did not gain
wide application for MVD-surgery; however it can be useful during

Please cite this article in press as: Sindou M, Mercier P. Microvascular decompression for hemifacial spasm : Surgical techniques and
intraoperative monitoring. Neurochirurgie (2017), https://doi.org/10.1016/j.neuchi.2018.04.003
G Model
NEUCHI-924; No. of Pages 11 ARTICLE IN PRESS
10 M. Sindou, P. Mercier / Neurochirurgie xxx (2017) xxx–xxx
Fig. 12. Intraoperative Lateral Spread Response (LSR) recording. In this illustrative
case stimulation was applied to the zygomatic branch (i.e., upper branch) of the facial
nerve, with EMG-recording in the orbicularis oculi and the orbicularis oris muscles.
A (normal) direct response is seen in the orbicularis oculi (left column). A (abnor-
mal) LSR is observed in the orbicularis oris (right column). Note the progressive
disappearance of LSRs (from top to bottom), as soon as the arachnoid opening.
the learning period of installation for teams starting to develop
this surgery.
2.4. Conclusions
* MVD-surgery being functional surgery for non-vital diseases
should be as inoffensive as possible. No doubt that BAEP mon-
itoring has contributed to decrease hearing complications [62].
To be recalled, hearing deficit after MVD without monitoring –
as assessed by pure tone audiometry – ranged from 7.7% to 20%
according the articles reviewed by Wilkins [63]. Thanks to the use
of BAEP monitoring, lessons learnt helped to decrease the rate
below the level of 2.3%, as in our series [31]. The dilemma to decide
whether BAEP monitoring is obligatory or may be withdrawn as
soon as the learning curve has reached a satisfactory (i.e., safe) level
remains controversial.
* Usefulness of routine use of intraoperative EMG-recordings
remains also debated. For a majority of authors LSR-recordings
are useful, for others including us reliability for ascertaining the
effectiveness of the MVD procedure is questionable.
* Whatever, intraoperative neuromonitoring requires a close
cooperation between the neurophysiologist and the surgical as well
the anesthesiologic teams.
Disclosure of interest
The authors declare that they have no competing interest.
Please cite this article in press as: Sindou M, Mercier P. Microvascular decompression for hemifacial spasm : Surgical techniques and
intraoperative monitoring. Neurochirurgie (2017), https://doi.org/10.1016/j.neuchi.2018.04.003
References
[1] Li ST, Sun H. Surgical techniques of microvascular decompression for hemifa-
cial spasm. In: LI, Zhong, Sekula, editors. Microvascular decompression surgery
Heidelberg: Springer; 2016. p. 79–93.
[2] Shimizu K, Matsumoto M, Wada A. Lateral basal approach with a supine, no-
retractor method for microvascular decompression for hemifacial spasm. Acta
Neurochir (Wien) 2015;157:803–7.
[3] Moller AR. Vascular compression of cranial nerves: II pathophysiology. Neurol
Res 1999;21:439–43.
[4] Moller RA. Pathogenesis and treatment of hemifacial spasm. In: LI, Zhong,
Sekula, editors. Microvascular decompression surgery Heidelberg: Springer;
2016. p. 35–49.
[5] Polo G, Fischer C, Sindou MP, Marneffe V. Brain auditory evoked potential
monitoring during microvascular decompression for hemifacial spasm: intra-
operative brainstem auditory evoked potential changes and warning values to
prevent hearing loss. Prospective study in a consecutive series of 84 patients.
Neurosurgery 2004;54:97–106.
[6] Polo G, Fischer C. Intraoperative monitoring of brainstem evoked potentials
during microvascular decompression of cranial nerves in cerebello-pontine
angle. Neurochirurgie 2009;55:152–7.
[7] Sindou M, Auque J. The intracranial venous system as a neurosurgeon’s per-
spective. Adv Tech Stand Neurosurg 2000;26:131–216.
[8] Mazzoni A, Hansen CC. Surgical anatomy of the arteries of the internal auditory
canal. Arch Otolaryngol 1970;91:128–35.
[9] Rhoton AL. The cerebellar arteries. Neurosurgery 2000;47(3):S29–68.
[10] Zhong J, Li ST, Zhu J, Guan HX, Zhou QM, Jiao W, et al. A clinical analysis on
microvascular decompression surgery in a series of 3000 cases. Clin Neurol
Neurosurg 2012;114(7):846–51.
[11] Zhong J, Zhu J, Sun H, Dou NN, Wang YN, Ying TT, et al. Microvascular decom-
pression surgery: surgical principles and technical nuances based on 4000
cases. Neurol Res 2014;36(10):882–93.
[12] Payner TD, Tew JM. Recurrence of hemifacial spasm after microvascular decom-
pression. Neurosurgery 1996;38:686–91.
[13] Kureshi SA, Wilkins RH. Posterior fossa reexploration for persistent or recurrent
trigeminal neuralgia or hemifacial spasm: surgical findings and therapeutic
implications. Neurosurgery 1998;43(5):1111–7.
[14] Khoo HM, Yoshimine T, Taki T. A “sling swing transposition” technique with
pedicled dural flap for microvascular decompression in hemifacial spasm. Neu-
rosurgery 2012;71:25–30 [1 Suppl operative, discussion 30–1].
[15] Kurokawa Y, Maeda Y, Toyooka T, Inaba K. Microvascular decompression for
hemifacial spasm caused by the vertebral artery: a simple and effective trans-
position method using surgical glue. Surg Neurol 2004;61(4):398–403.
[16] Ichikawa T, Agari T, Kurozumiu K, Maruo T, Satoh T, Date I. Double-stick tape”
technique for transposition of an offending vessel in microvascular decompres-
sion: technical case report. Neurosurgery 2011;68:377–82 [2 Suppl operative,
discussion 382].
[17] Ammar A, lagenaur C, Jannetta P. Neural tissue compatibility of Teflon
as an implant material for microvascular decompression. Neurosurg Rev
1990;13:299–303.
[18] Cho DY, Chang CG, Wang YC, Wang FH, Shen CC, Yang DY. Repeat operations
in failed microvascular decompression for trigeminal neuralgia. Neurosurgery
1994;35:665–9.
[19] Megerian CA, Busaka NY, McKenna MJ, Ojemann RG. Teflon granuloma pre-
senting as an enlarging, gadolinium enhancing, posterior fossa mass with
progressive hearing loss following microvascular decompression. Am J Otol
1995;16:783–6.
[20] Fujimaki T, Hoya K, Sasaki T, Kirino T. Recurrent trigeminal neuralgia caused
by an inserted prosthesis: report of two cases. Acta Neurochir (Wien)
1996;138:1307–9.
[21] Rath SA, Klein HJ, Richter HP. Findings and long-term results of subsequent
operations after failed microvascular decompression for trigeminal neuralgia.
Neurosurgery 1996;39:938–40.
[22] Liao JJ, Cheng WC, Chang CN, Yang JT, Wei KC, Hsu YH, et al. Reoperation for
recurrent trigeminal neuralgia after microvascular decompression. Surg Neurol
1997;47:568–70 [562–8 discussion].
[23] Premsaggar IC, Moss T, Coakham HB. Teflon-induced granuloma following
treatment of trigeminal neuralgia by microvascular decompression. Report of
two cases. J Neurosurg 1997;87:454–7.
[24] Bobeck MP, Sagher O. Aseptic meningitis caused by Teflon implantation for
microvascular decompression. Case report. J Neurosurg 1999;90:145–7.
[25] Chen J, Lee S, Lui T, Yeh Y, Chen T, Tzaan W. Teflon granuloma after microvas-
cular decompression for trigeminal neuralgia. Surg Neurol 2000;53:281–7.
[26] Matsushima T, Yamaguchi T, Inoue TK, Matsukado K, Fukui M. Recurrent
trigeminal neuralgia after microvascular decompression using an interpos-
ing technique. Teflon felt adhesion and the sling retraction technique. Acta
Neurochir (Wien) 2000;142:557–61.
[27] Vitali AM, Sayer FT, Honey CR. Recurrent trigeminal neuralgia secondary to
Teflon felt. Acta Neurochir (Wien) 2007;149:719–22.
[28] Capell HH, Brandis A, Tschan CA, Krauss JK. Treatment of recurrent trigeminal
neuralgia due to Teflon granuloma. J Headache Pain 2010;11:339–44.
[29] Rosseau G, Jannetta PJ. Surgical treatment of hemifacial spasm by microvascular
decompression. In: Sindou M, Keravel Y, Moller A, editors. Hemifacial spasm:
a multidisciplinary approach. New-York: Springer, Wien; 1997. p. 81–7.
[30] Ling PY, Mendelson ZS, Reddy RK, Jyung RW, Lin JK. Reconstruction after
retro-sigmoid approaches using autologous fat graft-assisted Medpor Titan
G Model
NEUCHI-924; No. of Pages 11 ARTICLE IN PRESS
M. Sindou, P. Mercier / Neurochirurgie xxx (2017) xxx–xxx
cranioplasty: assessment of postoperative CSF leak and headaches in 60 cases.
Acta Neurochir (Wien) 2014;156:1879–88.
[31] Sindou M. Microvascular decompression for primary hemifacial spasm. Impor-
tance of intraoperative neurophysiological monitoring. Acta Neurochir Wien
2005;147:1019–26.
[32] Legatt AD. Mechanisms of intra operative brainstem auditory evoked potential
changes. J Clin Neurophysiol 2002;19:396–408.
[33] Hatayama T, Moller AR. Correlation between latency and amplitude of peak
V in th BAEPs: intraoperatve recordings in MVD operations. Acta Neurochir
1998;140:681–7.
[34] Jo KW, Kim JW, Kong DS, Hong SH, Park K. The patterns risk factors of hearing
loss following MVD for HFS. Acta Neurochir 2011;153:1023–30.
[35] Raudzens PA, Shetter AG. Intraopertive monitoring of BAEPs. J Neurosurg
1982;57:341–8.
[36] Watanabe E, Schramm J, Strauss C, Fahlbusch R. Neurophysiologic monitor-
ing in posterior fossa surgery: part II-BAEP-waves I and V and preservation of
hearing. Acta Neurochir 1989;98:118–28.
[37] Sindou M, Fobe JL, Ciriano D, Fischer C. Hearing prognosis and intraoperative
guidance of BAEPs in MVD. Laryngoscope 1992;102:678–82.
[38] Moller AR, Jannetta PJ. On the origin of synkinesis in hemifacial spasm: results
of intracranial recordings. J Neurosurg 1984;61:569–76.
[39] Gardner WJ. Crosstalk: the paradoxical transmission of a nerve impulse. Arch
Neurol 1966;14:149–56.
[40] Moller AR. Hemifacial spasm: ephaptic transmission or hyperexcitability of the
facial motor nucleus? Exp Neurol 1987;98:110–9.
[41] Moller AR, Jannetta PJ. Monitoring facial EMG during microvascular decom-
pression operations for hemifacial spasm. J Neurosurg 1987;66:681–5.
[42] Mooij JJ, Mustapha MK, Van Weerden TW. Hemifacial spasm: intraop-
erative electromyographic monitoring as a guide for MVD. Neurosurgery
2001;49:1365–71.
[43] Kong DS, Park K, Shin BG, Lee JA, Eum DO. Prognostic value of the lateral spread
response for intraoperative electromyography monitoring of the facial muscu-
lature during microvascular decompression for hemifacial spasm. J Neurosurg
2007;106:384–7.
[44] Sekula Jr RF, Bhatia S, Frederickson AM, Jannetta PJ, Quigley MR, Small GA, et al.
Utility of intraoperative electromyography in microvascular decompression for
hemifacial spasm: a meta-analysis. Neurosurg Focus 2009;27(4):E10.
[45] Kim CH, Kong DS, Lee JA, Park K. The potential value of the disappearance of
the lateral spread response during microvascular decompression of predict-
ing the clinical outcome of hemifacial spasms: a prospective study. Neurosurg
2010;67:1581–8.
[46] Thirumula PD, Shah AC, Nikonow TN, Habeych ME, Balzer JR, Crammond DJ,
et al. Microvascular decompression for hemifacial spasm: evaluating outcome
prognosticators including the value of intraoperative lateral spread response
monitoring and clinical characteristics in 293 patients. J Clin Neurophysiolog
2011;28:56–66.
[47] Ying TT, Li ST, Zhong J, Li XY, Wang XY, Zhu J. The value of abnormal mus-
cle response monitoring during microvascular decompression for hemifacial
spasm. Int J Surg 2011;9:347–51.
Please cite this article in press as: Sindou M, Mercier P. Microvascular decompression for hemifacial spasm : Surgical techniques and
intraoperative monitoring. Neurochirurgie (2017), https://doi.org/10.1016/j.neuchi.2018.04.003
11
[48] Sun H, Li ST, Zhong J, Zhang WC, Hua XM, Wan L, et al. The strategy of microvas-
cular decompression for hemifacial spasm: how to decide the endpoint of an
MVD surgery? Acta Neurochir (Wien) 2014;156(6):1155–9.
[49] Hatem J, Sindou M, Vial C. Intraoperative monitoring of facial EMG responses
during microvascular decompression for hemifacial spasm. Prognostic value
for long-terrm outcome: a study in a 33-patients series. Br J Neurosurg
2001;15(6):496–9.
[50] Kiya N, Bannur T, Yamagushi A, Yoshida K, Kato Y, Kanno T. Monitoring of facial
evoked EMG for hemifacial spasm: critical analysis of its pronostic value. Acta
Neurochir 2001;143:365–8.
[51] Joo WI, Lee KJ, Park HK, Chough cK, Rha HK. Prognostic value of intraopera-
tive lateral spread response monitoring during microvascular decompression
in patients with hemifacial spasm. J Clin Neurosci 2008;15:1335–9.
[52] Von Eckardstein K, Harper C, Castner M, Link M. The significance of intraoper-
ative electromyographic lateral spread in predicting outcome of MVD for HPS.
J Neurol Surg Skull Base 2014;75:198–203.
[53] Wei Y, Yang W, Zhao W, Pu C, Li N, Cai Y, et al. MVD for HFS: can intraopera-
tive lateral spread response monitoring improve surgical efficacy? J Neurosurg
2017;12:1–6 [103171/2016;11.JNS162148].
[54] Li S, Hong W, Tang Y, Ying T, Zhang W, Li X, et al. Re-operation for per-
sistent hemifacial spasm after microvascular decompression with the aid
of intraoperative monitoring of abnormal muscle response. Acta Neurochir
2010;152:2113–8.
[55] Zhong J, Xia L, Dou NN, Ying TT, Zhu J, Liu MX, et al. Delayed relief of hemifacial
spasm after vascular decompression: can it be avoided? Acta Neurochir (Wien)
2015;157:93–8 [Discussion 98–9].
[56] Fukuda M, Takao T, Hiraishi T, Fujii Y. Free running EMG monitoring
during microvascular decompression for hemifacial spasm. Acta Neurochir
2015;157:1505–12.
[57] Zheng X, Hong W, Tang Y, Ying T, Wu Z, Shang M, et al. Discovery of a new
waveform for intraoperative monitoring of hemifacial spasms. Acta Neurochir
(Wien) 2012;154:799–805.
[58] Yang M, Zheng X, Ying T, Zhang W, Yang X, Li ST. Combined intraoperative moni-
toring of abnormal muscle response and Z-L response for hemifacial spasm with
tandem compression type. Acta Neurochir (Wien) 2014;156:1161–6 [Discus-
sion 1166].
[59] Zhang X, Zhao h, Tang YD, Zhu J, Zhon p, Yuan Y, et al. The effects of combined
intra operative monitoring of abnormal muscle response and Z-L response for
HFS. World Neurosurg 2017;108:367–73.
[60] Moller AR. Intraoperative neurophysiological monitoring. Third edition New-
york: Springer; 2011.
[61] Li ST, Ying TT. Intraoperative monitoring. In: Li ST, Zhong J, Sekula RF, editors.
Microvascular decompression surgery. Springer, Dordrecht Heidelberg; 2016.
p. 151–70 [Chap. 12].
[62] Moller AR, Moller M. Does intraoperative monitoring od auditory evoked
potentials reduce incidence of hearing loss as a complication of MVD of cranial
nerves. Neurosurgery 1989;24:257–63.
[63] Wilkins RH. Hemifacial spasm: a review. Surg Neurol 1991;36:251–77.