Keywords
facial nerve - local anesthesia - otologic surgical procedures
Introduction
Surgical procedures in modern medicine have progressively evolved toward the use of
minimally-invasive and more conservative techniques, both as a demand of the patients
and as a possibility brought by state-of-the-art technologies and avant-garde studies.
These contemporary approaches would be considered effective if they incurred in reduced
stress response to trauma and lower intraoperative risks without compromising the
surgical outcomes. Within this context, the advent of surgical techniques with smaller
skin incisions, video-assisted procedures and, more recently, robotic surgery has
been observed. Developments in the anesthetic practice would also raise the prospect
of performing surgical procedures with decreased risks and morbidity along with improved
comfort for the patients. Hence, integrating safe and high-quality surgical care with
low-risk anesthesia in a clinically-applicable fashion should continuously be the
ultimate goal of surgical innovations.
In otolaryngology, and more specifically in otology, one of the most feared complications
is facial nerve injury, which is associated with various grades of muscle weakness
and subsequent impairment in facial expression. This chain of events might lead to
significant esthetic and functional consequences. The facial nerve runs a complex
route within the temporal bone in such a way that otologic procedures are often performed
near the nerve canal, which is known as the Fallopian canal, or use it as a reliable
landmark.
The concerns regarding facial nerve injury have led to the emergence and wide use
of intraoperative neuromonitoring systems, which have been thoroughly studied and
validated for surgeries under general anesthesia.[1]
[2] Cochlear implantations, for example, are among the surgeries during which injuries
to the seventh cranial nerve are most likely and feared. When performing posterior
tympanotomy, the surgical step in which the facial recess is opened and widened to
enable the insertion of the electrode array, the nerve lies just underneath the drilling
site. Thus, cochlear implant surgeries are regarded as an intervention in which facial
nerve monitoring is classically advocated. However, other surgeries such as those
for treating chronic otitis also pose a risk to this nerve, as the pathological tissue
might surround the Fallopian canal and sometimes the nerve sheath itself.
Local anesthesia with sedation has been successfully employed for an increasingly
number of procedures across otolaryngology. Although this anesthetic management was
initially established for nasal surgeries, it proved effective in such otologic procedures
as tympanoplasty, tympanomastoidectomy, stapedectomy, and cochlear implantation.[3]
[4] Nonetheless, the principles, techniques and parameters of use of facial nerve monitoring
under local anesthesia with conscious sedation have not yet been studied, determined
or reported in the medical literature.
The present study has the goal of reporting the preliminary outcomes of facial nerve
intraoperative monitoring (FNIM) during otologic procedures under sedation and local
anesthesia to suggest methods and techniques for its safe and effective use.
Methods
Herein we will describe and discuss the perioperative electrophysiological data obtained
from five otologic procedures in four adult patients. These procedures were performed
by the same surgeon, with the patients under local anesthesia and sedation. The preoperative
imaging assessment included a computed tomography scan for each case and the images
did not show any temporal bone malformation in the patients included. Moreover, upon
clinical examination, they did not present any limitation in facial muscle tone or
expression.
The extensive preoperative anesthetic consultation consisted of evaluation of the
patients' clinical condition, investigation of possible coexisting diseases, and a
thorough explanation of the anesthetic procedures required intraoperatively.
No premedication was administered. The intraoperative basic monitoring included electrocardiogram
and pulse oximetry. Fentanyl 1 ucg/kg, meperidine 0.5 mg/kg, midazolam 5 mg, and clonidine
2 ucg/kg upon induction of the anesthesia were administered to both patients. An oxygen
flow of 3 L/min was delivered by nasal cannula. Local infiltration anesthesia was
then performed by the surgeon, using a solution of 2% lidocaine with 1:50.000 epinephrine.
This infiltration was performed at the postauricular area, near the postauricular
sulcus, and at the four quadrants of the external auditory canal, as traditionally
described for otologic procedures. A total volume of 6–8 mL of anesthetic solution
was injected for each patient. During surgery, opioids would be re-administered if
any signs of pain or discomfort were perceived. Other drugs routinely administered
were: ondansetron 4 mg, metoclopramide 10 mg, cefazolin 1 g, dexamethasone 1 mg/kg,
dypirone (metamizole) 1 g, and ketorolac 30 mg. Reversal of morphine anesthesia, if
needed, could be achieved by administering naloxone 0.2 mg.
Facial nerve monitoring was performed using the NIM-Neuro 3.0 (Medtronic Xomed Inc.,
Jacksonville, FL, US) nerve integrity monitor by a specialized electrophysiologist.
Previously to the local infiltration, but after sedating the patient, intramuscular-type
electrodes were inserted to capture the facial muscle responses. Paired electrodes
were inserted at least into the frontalis, orbicularis oris and mentalis muscles,
along with a ground electrode. In some cases, an additional pair was inserted into
the orbicularis oris muscle. During surgery, the electrophysiologist was responsible
for evaluating the electrophysiological parameters obtained, that is, the electric
signal recorded and originated from the muscle membrane potential.
The monitor's sensitivity was set at 100 mV. This means that any source of irritation,
regardless of its origin, that led to stimuli of the facial nerve higher than 100 mV
would result in typical oscillations around the baseline, with the concomitant emission
of a warning noise. If necessary, a specific type of probe could also be used to identify
the nerve within the surgical field. For this purpose, the tip of the probe is placed
on any structure, and then discharges an electric current whose impact on the baseline
register helps to define if the targeted structure corresponds to the nerve itself.
Moreover, an analysis of the repercussion of the changes in the amperage of the emitted
current on the latency and amplitude of the neural response might be helpful in diagnosing
potential lesions suspected perioperatively.
Results
The five procedures (four patients in total) herein described were uneventful, with
satisfactory surgical and anesthetic outcomes. The summary of the data obtained are
shown in [Table 1].
Table 1
Summary of the information regarding the five procedures with intraoperative facial
nerve electrophysiological records described
|
Patient
|
Age (years)
|
Gender
|
Procedure
|
Otologic background
|
Electrophysiological findings
|
|
Case 1
|
51
|
Female
|
Endoscopic cholesteatoma removal and tympanoplasty
|
Chronic otitis
|
Progressively decreasing low-amplitude baselines, with irregular initial moderate
oscillations in the mentalis electrode compatible with voluntary muscle activity.
Major alterations during surgical manipulation and probe-induced nerve stimulation.
|
|
Case 2
|
40
|
Male
|
Left cochlear implantation
|
Temporal bone fracture
|
Progressively decreasing low-amplitude baselines. Mild alterations close to the skin
incision. No oscillations detected during other specific surgical steps.
|
|
Case 3
|
40
|
Male
|
Right cochlear implantation
|
Temporal bone fracture
|
Low-amplitude and more stable baselines since the beginning of the surgery. The use
of the stimulation probe led to a short-latency and high-amplitude neuromuscular response.
|
|
Case 4
|
68
|
Male
|
Right cochlear implantation
|
Idiopathic sensorineural hearing loss
|
Progressively decreasing low-amplitude baselines. Mild alterations close to the posterior
tympanotomy, but without triggering the alarm.
|
|
Case 5
|
34
|
Male
|
Left cochlear implantation
|
Advanced otosclerosis
|
Progressively decreasing low-amplitude baselines, with mild non-significant oscillations
during the use of the drill. Major alterations during voluntary manipulation of an
area of facial nerve dehiscence and the use of the stimulation probe.
|
The first patient, NSCBR, a 51-year-old female with clinical history of controlled
hypertension (American Society of Anesthesiologists [ASA] grade II), was submitted
to the endoscopic management of a limited attic cholesteatoma with concomitant tympanoplasty.
A possible dehiscence of the facial nerve in the tympanic segment was suspected in
the preoperative assessment and confirmed intraoperatively. [Fig. 1] summarizes the intraoperative record of the facial neuromuscular electric activity.
As the procedure began, with the patient still under mild sedation, low-amplitude
(20–8 + 0 µV) baseline values were detected. However, the amplitudes were higher than
those habitually observed during facial nerve monitoring upon general anesthesia.
The mentalis baseline value showed an irregular pattern, with much greater amplitude,
varying from 200 to 300 µV ([Fig. 1a]). This finding was compatible with the presence of correspondent muscle activity
and noisy breathing. To avoid repetitive false alarms, the data provided by this electrode
were excluded from analysis in the beginning of the procedure.
Fig. 1 Electrical records obtained for facial nerve monitoring during different surgical
steps of an endoscopic approach to a limited attic cholesteatoma. (A) Beginning of the surgery, with the patient under mild sedation; (B) patient under deeper anesthetic effect; (C) use of the drill in the surgical site; (D) debulking of the cholesteatoma over an area of facial nerve dehiscence; (E) direct stimulation of the facial nerve with the testing probe.
As the surgery progressed and the patient achieved a stronger anesthesia effect, there
was a perceivable reduction in the overall mean amplitude for all electrodes (5–60
µV for the 3 upper electrodes, 60–90 µV for the mentalis electrode) ([Fig. 1b]). The use of the drill in the attic/epitympanum led to a mild increase in the mean
amplitudes, although it remained inferior to 100 µV ([Fig. 1c]). Nonetheless, manipulation of the facial nerve sheath in the tympanic segment while
debulking squamous debris generated great oscillations on the values for the frontalis
and orbicularis oculi electrodes – the latter reaching 143 µV and determining the
alarm activation and concomitant sound emission ([Fig. 1d]). The use of the testing probe in the facial nerve pathway resulted in high-amplitude
(> 700 µV) and short-latency (4 ms) oscillations ([Fig. 1e]).
The second patient (second and third procedures) was a 40-year-old male, JAP, who
underwent a simultaneous bilateral cochlear implantation (Nucleus, Cochlear Ltd.,
Sydney, Australia) to address a bilateral anacusis secondary to head trauma with temporal
bone fractures. He had no preoperative comorbidity (ASA I), and, despite the complex
bilateral fractures, he showed no paralyzed facial muscles at any moment before the
procedure. The surgery began with the left side. A continuous low-amplitude baseline
electrical value was initially recorded. As the procedure progressed and the anesthesia
intensified, the mean baseline amplitude descended. Isolated major transitory oscillations
were observed close to the incision time, with the patient still mildly sedated, which
were concomitant and compatible with voluntary movements. No specific alterations
could be detected during the detachment of the soft tissues, the bone drilling, and
the posterior tympanotomy. As the patient had already been properly sedated, in the
beginning of the procedure on the right side, no oscillations in the recordings could
be noticed. The baseline amplitude for this procedure remained comparatively lower
and more stable. Similarly, there were no significant oscillations during specific
surgical steps. Upon the use of the stimulation probe in a presumed area of the Fallopian
canal on the right side, a high-amplitude, short-latency oscillation was observed
for all of the electrodes inserted.
The fourth patient, AZ, was a 68-year-old adult male submitted to a cochlear implant
surgery (Nucleus). He had been suffering from profound bilateral progressive hearing
loss of unknown cause. He had only mild drug-controlled systemic hypertension (ASA
II). In this case, stable, low-amplitude values were recorded throughout the whole
procedure. During the manipulation of the nerve behind the posterior tympanotomy,
an isolated wave of moderate amplitude arose in record of the orbicularis oris muscle.
This timely oscillation enabled a controlled and safe drilling of the facial recess.
The fifth procedure was a unilateral cochlear implant surgery (Nucleus). The patient,
AP, a 34-year-old male, developed profound bilateral hearing loss due to advanced
otosclerosis. He had no other comorbidities (ASA I). A total of four pairs of electrodes
were placed in the same way as described for the first patient. [Fig. 2] summarizes the electrical values during the main surgical events. Initial low-voltage
waveforms were obtained for all electrodes, although they were slightly superior for
the orbicularis oculi muscle ([Fig. 2a]). During skin incision, as the patient was under deeper sedation, the baseline amplitude
had noticeably diminished ([Fig. 2b]). A mild non-significant increase could be observed during the use of the drill
([Fig. 2c]). The use of the stimulation probe in the facial nerve pathway provoked an immediate
neuromuscular response with great amplitude and short latency ([Fig. 2d]) – similar to that obtained during the voluntary manipulation of a dehiscent segment
of the facial nerve with a Rosen elevator ([Fig. 2e]). However, the amplitudes recorded during the mechanical manipulation were distinctly
lower than those induced by the electrical stimulus.
Fig. 2 Electrical records obtained for facial nerve monitoring during different surgical
steps of a cochlear implantation. (A) Beginning of the surgery, with the patient under mild anesthesia; (B) patient under deeper anesthesia; (C) use of the drill when performing the posterior tympanotomy; (D) direct stimulation of the facial nerve with the probe; (E) voluntary facial nerve manipulation over an area of dehiscence.
Discussion
Intraoperative neurophysiological monitoring or intraoperative neuromonitoring (IOM)
is a well-established tool used during procedures involving the possibility of iatrogenic
nerve injuries. The first records of the practical use of this technique date back
to 1930, during epilepsy surgeries.[5] Nevertheless, the use of IOM only became popular and widespread in the early 1980s.[6] Since then, major improvements have been achieved due to the scientific and technological
developments in the field of electrophysiology. This surgical tool is nowadays broadly
used preventively to lower iatrogenic risks, as it enables the identification of important
neural structures in the operative field.[7] The neural response may be assessed intraoperatively by different methods: motor-evoked
potentials (MEPs), somatossensory-evoked potentials (SSEPs), electroencephalography
(EEG), electromyography (EMG), brainstem auditory-evoked potentials (BAEPs) and visual-evoked
potentials (VEPs).
Among these surgical tools, EMG is the most frequently used during otologic procedures,
and it consists of recording skeletal muscle activity by measuring and tracking changes
in the muscle cell membrane potential. It might be obtained with either intramuscular
or surface electrodes. Within this context, the muscular fibers innervated by a single
motor axon would form a motor unity,[8] whose electrical activity would collectively determine the motor unity action potentials.[9] The motor-evoked responses obtained in the EMG, therefore, mirror the anatomical
and physiological properties of the muscle under evaluation. Thus, it might be employed
to investigate abnormalities in the skeletal system and determine signs of muscular
dysfunction, evaluate central or peripheral neuropathies, estimate the prognosis and
severity of neuromuscular disorders, and monitor the efficacy of different therapeutic
approaches.[10]
Previous studies have established the accuracy of intraoperative monitoring in facial
nerve identification,[11] as well as the cost-effectiveness when used for ear and temporal bone surgeries.[12] The indications for facial nerve monitoring have become increasingly diverse, and
include parotidectomies, revision tympanoplasties and/or tympanomastoidectomies, cochlear
implantations, surgical treatment of middle or posterior fossa tumors, and any other
otologic procedure that may pose a risk to the VII cranial nerve.[1]
[13] Besides, a study by Hu et al[14] suggests a contemporary trend toward the increased use of nerve monitoring by younger
surgeons. The researchers also found a growing inclination among otologic surgeons
towards routine intraoperative monitoring when treating chronic ear diseases.[15] The main drawback in electromyography-based facial nerve monitoring is the inability
to discriminate the proximity of the nerve within the submillimeter range. Although
this does not significantly impact the regular otologic procedures performed nowadays,
future techniques such as robotic cochlear implantation might need improved automated
systems. Previous studies have already shown that an optimized neuromonitoring protocol
using an integrating bipolar and monopolar stimulating probe could be safe during
the robotic drilling in animal models and for the clinical use.[16]
[17] Another preliminary study on this matter has reported that the accuracy of intraoperative
tissue discrimination would also increase if impedance spectroscopy was used during
robotic mastoid drilling. Therefore, in the future, these techniques may be combined
to improve nerve distance assessment in this particular scenario.[18]
The main goals of FNIM are to track the nerve beforehand, to help identify bony canal
dehiscence that might increase the surgical risks, and to assess nerve function after
the procedure.[1]
[11]
[19] Many factors support FNIM in otology. First, the tridimensional anatomy of the facial
nerve within the temporal bone is complex, and the complete delineation of this structure
is often hindered by coexisting diseased tissue. Moreover, dehiscent segments are
not uncommon, and might not be easily told apart from the hypertrophic mucosal lining
or granulation tissue.[20] It should be highlighted that facial paralysis is utterly stigmatizing, and might
lead to great functional and esthetic handicaps. Lastly, upon an injury, nerve reconstruction
using the techniques currently described does not enable the full restoration of facial
mimicry (grades I and II of the House-Brackmann scale, which ranges from I to V).[2]
[21] This means that even after a successful reconstruction, the best result one could
achieve (grade III) would still be consistent with critical remaining sequelae.[22]
A complete FNIM is performed using paired electrodes corresponding to the following
muscles: frontalis, orbicularis oculi, orbicularis oris and mentalis. At least three
of them are usually used. A ground electrode is also needed, and often placed in the
sternum. In the procedures herein described, intramuscular electrodes were preferred.
The surgeon also has at his/her disposal a stimulation probe that may be used to intentionally
provoke an electrical stimulus on a certain area. With this maneuver, it is possible
to investigate the presence of the nerve in the region that is touched. This strategy
may work as a navigation system. Whenever an electric current is discharged onto the
nerve, an action potential is triggered, leading to facial spasms and specific, short-latency
fluctuations in the baseline. Due to this chain of events, the monitor sounds an alarm
that alerts the surgeon to facial nerve proximity.
Sustained motor unity potentials with a frequency > 30 Hz are called “neurotonic.”
This type of activity results from intense stimuli, irritation or injury. It appears
during FNIM as an unexpected oscillation, whose interpretation depends on the surgical
step in which it occurs and on the type of anesthesia to which the patient is submitted.
In case of stable general anesthesia, no spontaneous muscular contraction is expected.
Thus, in this scenario, any major fluctuation in the baseline records should be thoroughly
checked. The stimulation probe may be used to confirm the proximity of the facial
nerve to the area that was being manipulated when the abnormal event occurred.
Hence, the characteristics of the electromyographic recordings obtained during FNIM
might be significantly affected not only by individual factors, such as obesity and
previously existing neuromuscular condition, but also by the anesthetic depth and
type.[7] The most used type of anesthesia when establishing the FNIM parameters is total
intravenous anesthesia (TIVA) by combining propofol with opioids.[23] Although propofol exerts its effects primarily by modulating gamma-aminobutyric
acid (GABA) receptors, it is quickly metabolized so that its blood concentration is
swiftly lowered to levels in which the neuroelectrophysiological recording is not
usually impaired. Greater doses of propofol, however, lead to a significant negative
effect in the FNIM.[24]
[25] Classically, the administration of neuromuscular blocking agents is thought to be
detrimental to intraoperative nerve monitoring, as they act directly in the neuromuscular
junction. Choe et al[25] have showed, nonetheless, that during TIVA by combing propofol and remifentanil,
the controlled use of nondepolarizing neuromuscular blocking agents would not hinder
facial nerve monitoring.[24] Halogenated volatile compounds might completely abolish neuroelectrophysiological
responses, and are, therefore, not routinely used when intraoperative nerve monitoring
is considered during the surgery.[26] Many other substances might have a negative impact on nerve monitoring, such as
ketamine and muscle relaxants.[26] None of the drugs used during the procedures described in the present paper (fentanyl,
meperidine, midazolam, clonidine and lidocaine, for example) seemed to impair neuromonitoring,
as it could be performed without restriction for all of the patients. Therefore, it
appears that opioids and benzodiazepines are also compatible with proper intraoperative
nerve monitoring.
Temporal bone surgeries in which FNIM is indicated are performed under general anesthesia
in most services throughout the world. Some patients, however, might benefit from
having their procedures under local anesthesia and sedation. Elderly individuals with
preexisting multiple comorbidities, patients with predictable difficult airway management,
and individuals with a risk of having arrhythmic events are a few illustrative examples.
Besides, general anesthesia incurs in increased hospital stay and costs, slower postoperative
recovery and higher anesthetic risks. Among young adults, the major advantages of
local anesthesia are the delayed and reduced need for symptomatic therapy and earlier
hospital discharge.[3]
[4]
When the patients are under sedation, some level of consciousness and spontaneous
muscular activity might be present. This might explain isolated oscillations in the
intraoperative facial nerve recordings during the surgery. Whenever an electrical
event occurs, the surgical team must evaluate the possibility of potential harm to
the nerve. This evaluation should include the surgical step in which the event occurred,
the corresponding anesthetic depth, the characteristics of the oscillation, and the
concomitance of voluntary facial movements. In the beginning of the procedure, before
any drilling, for example, the facial nerve is not under major risk, and the patient
is not under deep sedation yet. Therefore, minor and temporary oscillations in the
electric recording are more frequently observed due to voluntary activity. On the
other hand, any alteration appearing during manipulation near the facial nerve canal
must serve as an alarm of potential injury, unless there is compelling evidence to
the contrary. Tailoring the degree of anesthesia when facing situations of uncertainty
might help the team determine the actual neural risk associated with the event. The
present study did not aim to compare the objective latency and amplitude levels of
facial nerve stimulation under sedation with those under general anesthesia. Nonetheless,
within the surgery under sedation and local anesthesia, we could observe that minor
interferences and oscillations were not uncommon, although they exhibited different
characteristics from those seen during risky nerve manipulation. The latter seems
to occur more persistently as the surgical step proceeds and has apparently shorter
latencies and greater amplitudes. However, if any uncertainty persists, it should
be underlined that the surgeon could also use the stimulation probe to verify the
presence of the nerve on the site under manipulation. Hence, in opposition to otologic
surgeries under general anesthesia, when the patient is under local anesthesia and
sedation, minor oscillations in the electrical recording might ensue without any pathological
significance. This particularity demands close attention on the part of the surgical
team. The present case series preliminarily suggests that FNIM during otologic procedures
is also feasible when they are performed under local anesthesia and sedation. It should
be emphasized that, in this scenario, any changes in the basal pattern of the EMG
activity must be taken into consideration, and that the surgeon's experience and anatomical
knowledge are paramount.
There is no other available data in the literature with regard to FNIM during surgeries
under local anesthesia and sedation. Nevertheless, the validation of nerve monitoring
in this setting is utterly important, as it would increase the safety of the procedures
performed under anesthetic techniques with lower morbidity, and reduce the overall
costs. Standard facial nerve monitoring systems are suitable even for procedures usually
performed under general anesthesia, but that have been already validated under less
aggressive anesthetic management, such as cochlear implantations in adults. Thus,
the need for intraoperative facial nerve surveillance should not be the factor hindering
the choice of type of anesthesia for otologic procedures.
Conclusion
Intraoperative facial nerve monitoring during otologic procedures under local anesthesia
with sedation is feasible and reliable, although it requires a surgical team with
proper procedural knowledge and up-to-date technical skills. The interpretation of
abnormalities within the recordings in this setting must be timely, considering the
surgical anatomy and steps, and the depth of the anesthesia.
