s1042368008000120.pdf

Neurosurg Clin N Am 19 (2008) 289–315

Intraoperative Monitoring of Facial and Cochlear

Nerves During Acoustic Neuroma Surgery

Charles D. Yingling, PhD a , * , John N. Gardi, PhD b

a Department of Otolaryngology/Head and Neck Surgery Stanford School of Medicine and California

Neuromonitoring Services, San Francisco, CA, USA

b California Neuromonitoring Services, San Francisco, CA

COMMENTARY

irritation. While methods based on direct detection

of facial motion by attached sensors have been at-

tempted [3] , the best alternative to EMG may be

a video-based system [4] .

Another recent development is the identification

of a specific EMG response to stimulation of the

nervus intermedius [5] . This response has a charac-

teristic low amplitude, prolonged latency, and re-

stricted distribution compared to stimulation of

the facial nerve itself. If this distinction is not recog-

nized, the n. intermedius may be mistaken for the

facial nerve; since these nerves are sometimes widely

separated by the growth of the tumor this may lead

to inadvertent section of the facial nerve.

Since the NIH Consensus Statement on Acoustic

Neuroma [6] unequivocally recommended routine

intraoperative monitoring of the facial nerve, there

have been no formal clinical trials to assess the effi-

cacy of facial nerve monitoring in improving out-

come. However, numerous studies have shown

than parameters derived from responses to intra-

operative stimulation (ie, as threshold, amplitude,

pre-post surgery, or proximal/distal ratios) are

strong predictors of postoperative facial function

[7–20] . While the consistency of these reports is en-

couraging, the optimum predictive variables have

yet to be determined; this will require a larger pop-

ulation studied with a consistent set of parameters.

In a different context (middle ear and mastoid sur-

gery), Wilson and colleagues [21] demonstrated

the cost-effectiveness of facial nerve monitoring.

Finally, one of the remaining issues in facial nerve

monitoring is the necessity for surgeon-applied stim-

ulation of the nerve itself, which is often difficult in

larger tumors until substantial resection has taken

place, sometimes without knowledge of the location

of the nerve until it is too late. A method for contin-

uous assessment of facial nerve function without the

necessity for direct intracranial stimulation would

help mitigate this problem. An obvious candidate

While most of the information in this article is still

useful and relevant, there have been several develop-

ments in cranial nerve monitoring during the 16 years

since its initial publication. This brief addendum is not

meant to be a comprehensive update, but rather

a concise summary of some of the most relevant de-

velopments with references to more recent literature.

For a more comprehensive treatment of many of

these topics, see Yingling and Ashram [1] . More in-

formation specific to cochlear nerve monitoring

may be found in Martin and Stecker [2] .

In 1992, many of the systems used in intraoperative

monitoring were cobbled together from laboratory

equipment. Today, commercial systems specifically

designed for intraoperative monitoring are available

from several manufacturers, including Cadwell Labo-

ratories, Axon Systems, and Nihon-Kohden. These

systems typically include specialized low voltage

and/or current stimulators suitable for direct cranial

nerve stimulation, the ability to perform EMG and

evoked potential recordings simultaneously, and can

be configured to monitor many other types of surgery

with appropriate software protocols.

A major limitation of EMG-based methods for fa-

cial nerve monitoring has been their inability to be

used during electrocautery, which creates large elec-

trical artifacts that obliterate EMG signals at times

when cranial nerves may be at significant risk from

thermal injury if cautery is applied in the vicinity of

the nerve. Thus methods not based on electrical re-

cordings are a useful adjunct to EMG, which re-

mains the most sensitive indicator of facial nerve

This article originally appeared in Otolaryngologic

Clinics of NA, volume 25, issue 2, April 1992;

p. 413–48.

* Corresponding author. University of California,

401 Parnassus Avenue, San Francisco, CA 94143–0984.

doi:10.1016/j.nec.2008.02.011

neurosurgery.theclinics.com

290

YINGLING & GARDI

is the blink reflex, which is elicited by stimulation of

the supraorbital branch of the trigeminal nerve and,

via a polysynaptic reflex arc, elicits responses in

muscles innervated by the facial nerve. Unfortu-

nately, although the blink reflex has shown promise

as a prognostic indicator in the clinical setting [5] ,

it has proven difficult to reliably elicit under general

anesthesia [22] .

The most promising solution to this problem may

be elicitation of facial nerve motor evoked potentials

by transcranial electrical stimulation of the contra-

lateral face area of motor cortex [23] . Questions re-

main as to the specificity of this response to facial

nerve activation (for example, trigeminal nerve acti-

vation could produce similar responses, and latency-

based techniques for differentiation of V vs. VII as

described in the main article may not be applicable

to transcranial stimulation). Nevertheless, the

method of transcranial stimulation has come into

widespread use for monitoring corticospinal tracts

during spinal surgery, and its application to facial

nerve monitoring may prove to be the most signifi-

cant advance in this field since the advent of EMG

monitoring in the late 1970’s.

This early technique, in which the face was ob-

served for visible contractions after electrical stim-

ulation, remained the state of the art for facial

nerve identiﬁcation until the late 1970s, when

the use of facial electromyography (EMG) was in-

troduced by Delgado et al in 1979 [32] .

There are, to our knowledge, no reports of

VIIIth cranial nerve monitoring until the late 20th

century, undoubtedly because the development of

techniques for signal averaging and the discovery

of the human auditory brain stem response (ABR)

by Jewett and Williston in 1971 [33] were neces-

sary preconditions for attempting to monitor co-

chlear nerve function. Also, during the early

days of acoustic neuroma surgery, the generally

large size of the tumors when diagnosed and the

relatively crude state of surgical techniques made

mortality the main issue, rather than cranial nerve

preservation. As advances in diagnosis and micro-

surgical techniques have made such surgery safer,

increasing emphasis has been placed on preserva-

tion of cranial nerve function, with a resultant

growth in development of techniques for monitor-

ing these nerves during surgery.

Now VIIth cranial nerve monitoring during

acoustic neuroma surgery has become routine,

and anatomic preservation of the facial nerve is

regularly achieved in all but a few of the largest

tumors. Facial motility is still often compromised

in the immediate postoperative period, but the

prognosis for eventual recovery of function is

good if the nerve is intact and can be electrically

stimulated after tumor removal. Preservation of

hearing has been more dicult to achieve, owing

to the more intimate relationship of the tumors

with the cochleovestibular nerve, but can now

often be achieved in smaller tumors with the aid of

VIIIth cranial nerve monitoring techniques. This

article, based on our experience in over 500

posterior fossa procedures as well as a review of

the literature, describes the methods currently

available for cranial nerve monitoring, emphasiz-

ing facial and cochlear nerve monitoring during

acoustic neuroma surgery.

The ﬁrst use of cranial nerve monitoring

during posterior fossa surgery was almost a cen-

tury ago. On July 14, 1898, Dr. Fedor Krause

performed a cochlear nerve section for tinnitus

and noted that ‘‘. unipolar faradic irritation of

the (facial) nerve-trunk with the weakest possible

current of the induction apparatus resulted in

contractions of the right facial region, especially

of the orbicularis oculi, as well as of the branches

supplying the nose and mouth.’’ [24] . The pa-

tient awoke with a slight facial paresis, which

mostly resolved within a day. Krause also noted

contractions of the shoulder, which he attributed

to stimulation of the spinal accessory nerve, which

‘‘. had undoubtedly been reached by the current,

because it was, together with the acusticus, bathed

in liquor that had trickled down. .’’ He thus an-

ticipated not only the use of electrical stimulation

to locate cranial nerves but also the enduring prob-

lem of artifactual responses from current spread.

Frazier [25] described a similar technique used

in 1912 during an operation for relief of vertigo,

pointing out the importance of facial nerve preser-

vation and the fact that it could be identiﬁed by

‘‘galvanic current.’’ Similar methods were later

described by Olivecrona [26,27] , Hullay and To-

mits [28] , Rand and Kurze [29] , Pool [30] , and Al-

bin and colleagues [31] . Givre and Olivecrona [26]

and Hullay and Tomits [28] even recommended

removal of acoustic neuromas under local anes-

thesia to facilitate assessment of facial function.

Technical issues

Personnel

Successful performance of intraoperative mon-

itoring is not simply a matter of bringing another

piece of equipment into the operating room.

Applying neurophysiologic techniques in the time-

pressured and electrically hostile environment of

CRANIAL NERVE MONITORING DURING ACOUSTIC NEUROMA SURGERY

291

the operating room requires specialized skills

that may make the diﬀerence between successful

monitoring and no monitoring or, even worse,

inadequate monitoring that provides inaccurate

feedback to the surgeon. As a result, a new

specialty ﬁeld of intraoperative neurophysiologic

monitoring is evolving, and a professional

organization, the American Society of Neuro-

physiological Monitoring (ASNM), has been

founded. Specialists in intraoperative monitoring

have come from diverse backgrounds, including

neurology, neurophysiology, audiology, and an-

esthesiology; regardless of background or pro-

fessional degree, however, such personnel share

a common fund of knowledge including the

relevant neuroanatomy and neurophysiology,

principles of biomedical instrumentation, knowl-

edge of the variety of intraoperative monitoring

techniques and their uses and limitations, and

practical experience in performing these tech-

niques and interpreting their results. Given the

potentially catastrophic consequences of inappro-

priate application of monitoring techniques, we

believe that the participation of professional

monitoring personnel is highly desirable, despite

the additional costs incurred. Third-party reim-

bursement should be facilitated by the recent

addition of a CPT code (95920) speciﬁc to intra-

operative neurophysiologic monitoring.

anesthetic management because patient movement

could have disastrous consequences and must be

prevented by maintaining an adequate level of

anesthesia. Fortunately, the ABR and EMG are

not aﬀected by routine concentrations of common

anesthetics, such as nitrous oxide, opiates, or

halogenated agents, so no other constraints on an-

esthetic technique are necessary. Short-acting

agents such as succinylcholine may be given to

facilitate intubation, but it must be veriﬁed that

such agents have cleared before any manipulations

that might aﬀect the facial nerve are undertaken.

For a suboccipital approach, this would be the

time of opening the dura and retraction of the cer-

ebellum; in a translabyrinthine approach, the facial

nerve is ﬁrst at risk during skeletonization of the

horizontal portion in the temporal bone. Fortu-

nately these events typically occur far enough

into the procedure that any relaxants given at intu-

bation will have cleared in time.

Instrumentation

Electromyography instrumentation

The essential requirements for facial EMG

monitoring are a stimulator that can be precisely

controlled at low levels, one or more low noise

ampliﬁers capable of amplifying microvolt level

signals, an oscilloscope, and an audio monitor

with a squelch circuit to mute the output during

electrocautery. The NIM-2 (Nerve Integrity Mon-

itor), manufactured by Xomed-Treace (Jackson-

ville, Florida), is a commercial device oﬀering two

channels (only one of which is displayed at a time)

and appropriate stimulation and squelch circuits.

It is relatively easy, however, to put together

a system from oﬀ-the-shelf components that can

provide more channels at a substantially lower

cost. At the University of California, San Fran-

cisco (UCSF), we use a four-channel system with

Grass ampliﬁers and stimulator (Quincy, Massa-

chusetts) and a Tektronix oscilloscope (Beaver-

ton, Oregon), with a custom audio monitor.

Another possibility, although generally more

expensive, is to use a commercial multichannel

EMG machine, provided that low enough levels

of stimulation are available. Although several

multichannel machines are available, most are

designed for percutaneous stimulation at higher

levels (ie, 1 to 300 V or 1 to 50 mA), whereas the

levels needed for safe intracranial stimulation are

less than 1 V or 1 mA. A qualiﬁed biomedical

engineer can usually modify such systems to lower

the stimulation range, although care must be

Anesthetic considerations

Unlike cortical evoked potentials, which are

notoriously sensitive tomany anesthetic agents, the

ABR and EMG responses that are monitored

during acoustic neuroma surgery are essentially

unaﬀected by any commonly used anesthetic reg-

imens. The one exception to this is a contraindica-

tion to the use of any muscle relaxants because

blockade of the neuromuscular junction is incom-

patible with meaningful monitoring of EMG

activity. A recent report [34] has suggested that par-

tial blockade can be used to prevent patient move-

ment while still retaining the ability to elicit EMG

responses with facial nerve stimulation. Our experi-

ence has veriﬁed this observation but indicates that

although electrically evoked EMG is relatively pre-

served, both spontaneous EMG and mechanically

elicited activity appear to be obliterated by these

agents. This compromises two of the more impor-

tant indicators of facial nerve injury.

We therefore recommend that no paralytic

agents be used during acoustic neuroma surgery.

This, of course, creates its own problems for

292

YINGLING & GARDI

taken not to compromise patient safety features.

The availability of more channels allows simulta-

neous monitoring of multiple divisions of the

facial nerve independently as well as other cranial

motor nerves such as V and XI, which are often

involved in acoustic tumor surgery (see later).

such a protocol has been implemented on a custom

system.

Degree of automation and size are also impor-

tant design issues. In general, the more compact

and portable the system, the more likely it will be

accommodated without major complaints from

surgical personnel, especially if it is transported

between various operating rooms. Automated

data collection protocols, with simultaneous

display of baseline traces and recent trends as

well as the current trace, facilitate continuous

monitoring and assessment of intraoperative

changes, although the capability for manual over-

ride of automated protocols is desirable.

Surgical monitoring is done in an electrically

hostile environment. Every eﬀort must be mar-

shalled to eliminate or reduce 50 or 60 Hz power

line interference as well as the frequently broad-

band noise originating in other operating room

equipment (eg, electrocautery, lasers, ultrasonic

aspirators, microscopes, anesthesia machines,

electriﬁed beds, light dimmers, patient warmers,

compression stockings). The 60 Hz notch ﬁlters

found on most equipment are of limited utility

because they remove only 60 Hz sinusoidal

activity; more common is noise that recurs at the

line frequency but consists of complex spikes with

a high fundamental frequency that is not aﬀected

by notch ﬁlters. Therefore every eﬀort should be

made to identify such sources and eliminate their

interference if possible. Frequently this can be

done by grounding these items, plugging them

into a diﬀerent AC outlet, rerouting cables away

from monitoring equipment, or even disconnect-

ing them during crucial periods for monitoring.

Unfortunately it is not always possible to elimi-

nate or even identify some sources of interference

(at UCSF, one particularly noisy operating room

turned out to be upstairs over amagnetic resonance

imaging scanner, which generated large pulsatile

magnetic ﬁelds that were of sucient strength to

cause problems a ﬂoor away). Techniques for

distinguishing residual artifact from physiologic

activity are discussed later under ‘‘Monitoring the

VIIth and Other Cranial Motor Nerves.’’

Another important technique is to ensure that

the patient is adequately grounded to the re-

cording apparatus and that no alternate ground

paths exist. The patient ground should be placed

close to the recording electrodes and care taken to

obtain a low impedance ground by removing

surface oils with alcohol, then rubbing conductive

paste into the skin before applying a ground pad.

All equipment should be grounded to the same

Auditory brain stem response

The primary requirements for ABR monitor-

ing are an averaging computer with appropriate

high gain, low noise electroencephalogram (EEG)

ampliﬁers, and an acoustic stimulus generator

capable of delivering clicks of calibrated intensity,

with control of polarity (condensation, rarefac-

tion, or alternating) and repetition rate. Most

commercial evoked potential systems meet these

essential speciﬁcations and can be adapted to use

in the operating room. Typical clinical systems

include modules that accomplish two- to four-

channel, high-gain (100 to 500 K) diﬀerential

ampliﬁcation with multipole, band-pass ﬁltering

capabilities; acoustic stimulus generation with

a stimulus intensity range from threshold to at

least 70 to 80 dB normal hearing level (NHL);

response averaging with real time display of the

evolving averages as well as the raw trace; and

permanent record keeping on a disk medium with

hard copy printout. Several additional features,

however, are desirable for optimum monitoring

performance. Key design features of the ideal

monitoring system are versatility, portability (and

size), and degree of automation.

General technical considerations

Ideally, systems for use during acoustic neu-

roma surgery would be capable of simultaneous

EMG and ABR monitoring. This would require

independent control of the time base, stimulation,

and averaging parameters for the EMG and ABR

channels, features that are not generally available

in clinical EMG/evoked potential (EP) machines,

which are designed to perform a single test at

a time. The only exception that we are aware of at

this writing is the Nicolet Viking II (Nicolet

Instrument Corporation, Madison, Wisconsin),

which has a recently released software package

for intraoperative monitoring that allows for such

simultaneous protocols. Simultaneous collection

of ABRs from left and right ears is also desirable

to control for nonspeciﬁc eﬀects, such as anesthe-

sia, acoustic artifact, and patient temperature.

Again, this feature is not typically available in

commercial systems; see under ‘‘Monitoring the

VIIIth Cranial Nerve’’ later for details on how

CRANIAL NERVE MONITORING DURING ACOUSTIC NEUROMA SURGERY

293

spot with heavy-duty cables to avoid ground

loops. A detailed analysis of these issues is beyond

the scope of this article, but an excellent tutorial is

provided by Møller [35] .

ABR recording procedures.) The positioning

of the recording electrodes for a suboccipital

approach with an eﬀort to preserve hearing are

shown in Fig. 1 . For translabyrinthine approaches,

the same conﬁguration is used, with the exception

of the earphone and electrodes for ABR

recording.

Type and placement of recording electrodes

Either surface or needle electrodes can be used.

Surface electrodes are less speciﬁc, more prone to

artifact, and more timeconsuming to apply, so

their use has largely been supplanted by needle

electrodes, which can be quickly inserted and

taped into place. The most commonly used are

platinum needle electrodes designed for EEG

recording (Grass E2), which have a larger un-

insulated surface than electrodes designed for

single-ﬁber EMG recording and thus are more

likely to detect EMG activity arising anywhere

in the desired muscle. Prass and Liiders [36]

recommend the use of intramuscular hook wire

electrodes, which are inserted with the aid of

a hypodermic needle; in our experience, these

are more traumatic and oﬀer no major practical

advantage, so we routinely employ the simpler

needle electrodes.

The ﬁrst uses of facial EMG primarily

employed a single recording channel, typically

with a bipolar conﬁguration with one electrode in

orbicularis oculi and another in orbicularis oris

[37] . This montage provides coverage of muscles

innervated from both superior and inferior

branches of the facial nerve. It has several disad-

vantages, however, which have led to increasing

use of multiple channels. First, the wider the

spacing between two electrodes, the greater is

the sensitivity to artifact pickup, which in the elec-

trically hostile environment of the operating room

can lead to dicult or erroneous interpretations.

Second, mechanical trauma to the VIIth cranial

nerve frequently causes sustained EMG activity

that can make the identiﬁcation of responses to

electrical stimulation dicult. With two or more

independent channels, there is a greater likelihood

that at least one will be quiet enough to allow

stimulation to be used even during high ongoing

EMG activity.

For these reasons, we advocate the use of at

least two channels of facial EMG as well as

recordings from muscles innervated by other

cranial nerves. For ABR recording in hearing

conservation, one electrode is placed in the ear

canal and another on the forehead or vertex; the

placement of this electrode is not critical as long

as it is near the midline. (See under ‘‘Monitoring

the VIIIth Cranial Nerve’’ for further details on

Monitoring VIIth and other cranial

motor nerves

Three main techniques for monitoring cranial

motor nerve activity can be distinguished: (1)

monitoring ongoing EMG activity for increased

activity or changes in activity patterns related to

irritation of the nerves by intraoperative events,

such as retraction, tumor dissection, use of

electrocautery, lasers, and ultrasonic aspiration;

(2) identifying and mapping the course of the

nerves with activity evoked by intracranial

electrical stimulation; and (3) determining nerve

functional

integrity using evoked EMG

methods.

Activity evoked by electrical stimulation

Until the late 1970s, the typical method for

facial nerve identiﬁcation involved someone

(usually the anesthesiologist) observing the

patient’s face for evidence of movement related

to intraoperative events or electrical stimulation.

Unfortunately in many cases a complete facial

palsy resulted even though the face was observed

to move with stimulation. It is likely that the high

level of stimulation necessary to produce gross

movement from a nerve both chronically

stretched by the tumor and acutely traumatized

during surgery was itself damaging to the

nerve and thus contributed to this apparently

contradictory outcome. As a result, considerable

eﬀort has gone into developing more sensitive

measures of facial activity that can be elicited with

lower and safer levels of stimulation.

Modalities for monitoring

Several early eﬀorts focused on the use of more

sensitive detectors of facial motion, using photo-

electric devices, strain gauges, or accelerometers

mounted on the face [38,39] . A commercial device

is available that uses this technique [40,41] . A low-

tech version of this method has been described in

a whimsically titled paper ‘‘Bells against palsy,’’

[42] which uses small ‘‘jingle bells’’ sutured at

the points of maximum excursion of the facial

musculature. A technique has also been described

Plik z chomika:

Inne pliki z tego folderu:

Inne foldery tego chomika: