Keywords
cerebellopontine angle tumor - neuromonitoring - facial nerve outcome
Around 10% of brain tumors are cerebellopontine angle (CPA) tumors. Of these, vestibular
schwannoma (VS) is the commonest.[1] The incidence of facial paralysis following removal of CPA tumors was as high as
40 to 50% prior to use of microsurgical techniques. Intraoperative neuromonitoring
(IONM) helped to preserve facial nerve (FN) function in nearly 85% of the cases.[2] Among the various structures that are at risk, FN is at high risk due to its anatomical
proximity to these tumors, particularly in tumors greater than 3 cm. The tumor capsule
may be adherent to the nerve or the nerve may be thinned out and splayed apart by
the tumor and thus get damaged during tumor resection because of poor differentiation
from adjacent structures and or difficulty in identification. Free-running and triggered
electromyography (EMG) are commonly used to monitor FN function.[2] House–Brackmann (HB) scoring is widely used to document facial function.[3]
[4] Various factors such as FN proximal and distal absolute EMG amplitude, stimulation
threshold, proximal-to-distal amplitude ratio, and free-run EMG train time have been
used to predict outcomes but not consistently.[5]
[6] In this study, the robustness of a simple technique of anatomical preservation of
FN followed by compliance to physical therapy was analyzed. To the best of our knowledge,
this has not been analyzed before.
Methods
Subjects
IONM data and clinical data of patients undergoing CPA tumor surgery were retrospectively
analyzed from 2015 to 2023, including the 1-year postoperative period. All patients
were assessed for FN outcome immediately post-surgery on postoperative day 1 and at
1 year either in-person or telephonically. Telephonic follow-up consisted of ascertaining
eye closure completeness and smile symmetry. The Jaslok Hospital Ethics Committee
approved this study.
Surgical Procedure
This study was conducted at a tertiary hospital in Mumbai, India, which also has teaching
programs. Short-acting neuromuscular block was used only for induction. Anesthesia
was mainly intravenous.[7] If inhalational agents were used, the mean alveolar concentration was less than
0.5. All surgeries were done using a retro-sigmoid approach. An operating microscope
was used in all surgeries. The bone opening was done using a manual drill and punch.
The intra-dural procedure was done with microsurgical instruments, bipolar cautery
and suction, and IONM.
IONM Set Up
The FN monitoring protocol was standard for all patients and consisted of free-running
EMG and triggered EMG. EMG was recorded at frontalis, orbicularis oculi, orbicularis
oris (upper and lower), and mentalis using subdermal needle electrodes in all patients
to give a wide coverage for all branches of the FN. EMG activity was displayed with
audio for visual and auditory recognition. Settings for sensitivity for free-run and
triggered EMG was 100 µV/division. The time base for free-run EMG was 100 ms/division
and for triggered EMG it was 5 ms/division. A monopolar probe was used for triggered
EMG. Once exposure was complete, the surgeon performed tumor mapping using the monopolar
probe. A stimulation rate of 2 to 3 Hz, a duration of 50 to 300 µs, and an intensity
of 3 mA were used in all cases. As the dissection progressed, triggered EMG was conducted
whenever there was a need to identify neural structures. The stimulation intensity
was gradually reduced as the dissection progressed to avoid false-positive interpretation
due to high intensity. The FN stimulation threshold was always noted. In all cases
it was 0.5 mA or lesser. Free-run EMG was used to alert for neurotonic discharges
lasting more than 1 second and bursts. These could be caused by mechanical stretch
during dissection and thermal injury during cautery. False positive due to patient
anesthesia being light, movement artefact, and electrode artefact was taken cognizance
of prior to alert.[2]
House–Brackmann Grading System
The FN function was scored according to the HB grading system (Grades I–VI: Grade
I, normal function; Grade VI, total palsy).[3] Good FN function was defined as HB I and II, whereas HB III and VI were defined
as poor outcome. The time points for evaluation of FN function outcome were immediately
post-surgery on postoperative day 1 and at 1-year follow up.
Physical Therapy
All patients who had HB II and above in the immediate postoperative period were advised
physical therapy. It was started on postoperative day 1 by a physical therapist. Subsequently,
patients were asked to do the exercises independently at home. Mirror exercises and
facial massages were advised. Mirror exercises help to prevent synkinesis and enhance
symmetry. Facial massages create small recruitments and prevent hyperactivity of adjoining
muscles. Patients were encouraged to practice the exercises more than once every day
and counselled regarding slow gradual improvement to ensure compliance.
Results
Study Population and Tumor Characteristics
The study population consisted of 27 males (31%) and 60 females (69%) with the age
range of 20 to 82 years, with the mean age of 47.82 years. Nine patients (10.3%) were
operated for tumor recurrence. The tumors were categorized as small (less than 2 cm,
3.4%), medium (2–3 cm, 16%), and large (more than 3 cm, 80.5%). The tumor size ranged
from 1.40 to 8.20 cm with the average being 3.78 cm and the median was 3.50 cm. The
final pathology result showed VS in 62%, followed by meningioma in 8%, trigeminal
schwannoma and epidermoid in 9.1%, and others in 11.4%.
Association between Short-Term and Long-Term Facial FN Outcomes
FN functional continuity was preserved in 100% of the cases.
In all cases, FN threshold was 0.5 mA or less, lowest being 0.2 mA.
Short-term facial outcome was good in 75.8% of cases (HB I: 63.2%; HB II: 12.6%),
and poor outcome was obtained in 24.1% (HB III, IV: 21.8; HB V, VI: 2.2%).
Two patients were lost to follow-up at the end of 1 year and one patient expired at
3 months post-surgery due to myocardial infarction.
Long-term facial outcome was poor in 5 (5.7%) cases and good in 94.2% cases. In addition,
100.0% of the cases with immediate post-surgery HB II (mild facial weakness) and 91%
of the cases with immediate post-surgery HB III (moderate facial weakness) had full
recovery (grade HB I). The Chi-square test showed a statistically significant difference
for poor outcome at the end of 1 year for patients who had severe immediate postoperative
weakness ([Table 1]). [Table 2] shows the clinical profile of patients with poor outcome at the end of 1year post-surgery.
Table 1
Association between FN outcomes immediately post-surgery and at 1 year post-surgery
|
Facial weakness immediately post-surgery
|
N
|
1-year outcome (N = 29)
|
p-Value
|
|
HB I
|
HB II or above
|
|
No.
|
%
|
No.
|
%
|
|
HB II (mild)
|
11
|
11
|
100.0
|
–
|
–
|
0.006
|
|
HB III (moderate)
|
11
|
10
|
90.9
|
01
|
09.1
|
|
HB IV (moderate–severe)
|
06
|
03
|
50.0
|
03
|
50.0
|
|
HB V (poor)
|
01
|
–
|
–
|
01
|
100.0
|
Abbreviations: HB, House–Brackmann; FN, facial nerve.
Note: By Chi-square test; significant ≤ 0.05.
Table 2
Profile of patients with poor FN function at the end of 1 year
|
Patient
|
Age in years
|
Tumor size in cm
|
Histopathology
|
MST in mA
|
TT length in seconds
|
Extent of resection
|
Facial function immediately postop
|
Facial function at 1-year postop
|
|
1.
|
43
|
8.2
|
VS
|
0.2
|
>1
|
Near total
|
HB IV
|
HB IV
|
|
2.
|
42
|
2.8
|
VS
|
0.5
|
>1
|
Near total
|
HB III
|
HB III
|
|
3.
|
41
|
4.0
|
VS
|
0.5
|
>1
|
Subtotal
|
HB IV
|
HB IV
|
|
4.
|
41
|
4.1
|
VS recurrent tumor
|
0.5
|
>1
|
Near total
|
HB IV
|
HB IV
|
|
5.
|
59
|
3.4
|
VS
|
0.2
|
>1
|
Near total
|
HB V
|
HB III
|
Abbreviations: HB, House–Brackmann; MST, minimum stimulation threshold; TT, train
time; VS, vestibular schwannoma.
Association between Tumor Size and Severity of Paralysis at 1 year Post-surgery
At 1 year post-surgery, 5.8% of the cases with tumor size less than 3 cm had HB III
and above as against 5.7% of the cases with tumor size more than 3 cm. This difference
was not statistically significant ([Table 3]).
Table 3
Association between tumor size and severity of paralysis at 1-year post-surgery
|
Tumor size
|
N
|
No. of cases = 87
|
p-Value
|
|
HB II
|
HB III–VI
|
|
No.
|
%
|
No.
|
%
|
|
Tumor size less than 3 cm
|
17
|
12
|
70.6
|
1
|
5.8
|
0.482
|
|
Tumor size greater than 3 cm
|
70
|
43
|
61.4
|
4
|
5.7
|
Note: By Chi-square test; significant ≤ 0.05.
Association between FN Outcome and Physical Therapy
All patients with HB II and above were advised physical therapy. In addition, 95.5%
of the cases with mild and moderate immediate postoperative weakness and 57.1% with
severe and poor immediate postop weakness who were compliant with physical therapy
showed improvement, thus implying a statistically significant relation with compliance
to physical therapy ([Table 4]). One patient with tumor size less than 3 cm had poor outcome. This patient was
uncooperative for physical therapy. An EMG nerve conduction study (EMG NCS) was done
only for this patient from this series. The test was done at 10 months post-surgery.
It showed attenuated distal facial motor compound muscle action potential amplitudes
from the orbicularis oculi, nasalis, and orbicularis oris suggestive of distal continuity.
EMG showed fibrillations and positive sharp waves at rest. No recruitment was noted
on maximal volitional effort from single-site needle EMG for frontalis, orbicularis
oculi, and orbicularis oris.
Table 4
Association between compliance with physical therapy and outcome at 1-year post-surgery
|
Immediately postop weakness
|
N
|
Advised physical therapy (N = 29)
|
p-Value
|
|
Compliant and improved
|
Noncompliant, incomplete recovery
|
|
No.
|
%
|
No.
|
%
|
|
Mild and moderate
|
22
|
21
|
95.5
|
01
|
04.5
|
0.001
|
|
Severe and poor
|
07
|
03
|
42.9
|
04
|
57.1
|
Note: By Chi-square test; significant ≤ 0.05.
Profile of Patients Operated for Tumor Recurrence
Nine patients were operated for tumor recurrence. All had large tumors, i.e., more
than 3 cm. Furthermore, 62.5% of the patients had HB II and above immediate post-surgery
weakness, ranging from HB II to HB IV. But at the end of 1 year, eight patients (88.8%)
had no facial weakness. Only one patient had HB IV at the end of 1 year post-surgery.
This patient was noncompliant for physical therapy.
Discussion
With the advent of IONM, FN preservation after CPA tumor surgeries has improved significantly.
Using both free-run and triggered EMG is the accepted paradigm for IONM for FN monitoring.
However, because the tumor is in close proximity and often adherent to the FN, there
is always a concern for FN injury in spite of IONM.
The incidence of long-term facial dysfunction after VS surgery ranges from 4.8 to
41%, the mean being approximately 19%.[8] In the present study, 80% tumors were large but not associated with poor outcome.
On the other hand, interestingly, only one patient with tumor less than 3 cm who was
not compliant with physical therapy had poor outcome. No correlation for tumor size
has been reported previously in 100 patients of VS.[5]
Several studies have shown that immediate postoperative FN function has a direct correlation
with the final FN outcome. In our study, 100.0% of the cases with immediate post-surgery
HB II (mild facial weakness) and 91% of the cases with immediate post-surgery HB III
(moderate facial weakness) had full recovery (grade HB I). An immediate postoperative
HB IV–VI correlated with poor outcome, in accordance with published literature.[6]
[7] A higher immediate postoperative HB grade was associated with a poorer outcome at
the end of 1 year in this study. This finding has been noted in other studies too.[4]
[9]
Various monitoring techniques have been described for monitoring FN function intraoperatively,
such as free-running EMG, triggered EMG, A-train time, FN threshold, facial motor-evoked
potential, and blink reflex (BR) study. However, there is an overall lack of standardization
even in electrode montage. Usually, a montage of one to three muscles has been used
in previous studies.[5]
[9] In this study, four muscles were used to provide a wider and complete coverage.
Because the nerve might be splayed by the tumor into fascicles, sometimes the responses
are noted in just one muscle. Hence using more muscles for monitoring reduces the
chances of missing the nerve. This four-muscle montage serves well for both free-run
EMG and triggered EMG. As a neuromonitoring paradigm, employing more target muscles
for motor monitoring provides very high sensitivity.[10] A study comparing two-muscle montage versus four-muscle montage may objectively
corroborate the effectiveness.
On exposure, tumor mapping at 3 mA is highly recommended. Direct electrical stimulation
(mapping) of FN was the first neurophysiological method, due to the high incidence
of FN palsy during VS removal.[11]
[12]
[13] Nerve tissue can be identified from the surrounding tumor tissue so that a nerve-sparing
dissection can be attempted. Direct stimulation of a normal nerve root using less
than 2 mA can ensure an electrical activity in the appropriate muscle group.[14] However, the lowest “safe” threshold is not described. The lowest threshold is a
measure of nerve excitability. It is not an unequivocally reliable prognostic indicator
of FN function.[15] Also, the lowest possible stimulation may vary between IONM equipment. Hence, in
this study, no attempt was made to correlate the lowest threshold for FN stimulation,
although in all patients, the FN threshold was obtained at 0.5 mA or lower. Since
the lowest threshold is not standardized, the latency and amplitude can vary and thus
their absolute values and ratios may be rendered obsolete for prognostication. Thus,
the mainstay of functional outcome would rely on preventing neurotmesis. Gazia et
al[9] noted in a series of 157 patients that at 1 year postoperative assessment, a minimum
stimulation threshold of 0.1 mA showed a sensitivity and a specificity of 62 and 73%,
respectively. With a CMAP cut-off < 200 µV, for long-term prediction, sensitivity
was 73% and specificity was 73%. Turel et al[5] documented that latency and amplitude ratios were not reliable predictors of FN
outcome. A sustained neurotonic discharge may be produced due to mechanical injury
during dissection or thermal injury during electrocautery. But the correlation of
discharge duration, also called A-train based on such software; for FN outcome is
not consistent.[16]
[17] Although longer train times would be associated with severity of postoperative FN
function, it is difficult to quantify the duration of train time because sometimes
these may occur as repetitive bursts rather than continuous trains. They may also
be produced due to cold saline, patient's anesthesia being light, or a combination
of these factors. Liu et al[18] proposed that BR and facial corticobulbar motor evoked potentials (FCoMEP) may be
used in combination to complement each other for predicting FN outcome. The study
focused on eye closure function as the main measure of FN outcome. Valid BR was recorded
in 93.6% of the patients and FCoMEP was recorded in 39% of the patients. BR had better
predictive value compared to FCoMEP and needs further evaluation for feasibility.
In this series, one patient with a tumor size less than 3 cm and noncompliance to
physical therapy had poor outcome. His EMG NCS at 10 months showed preserved distal
continuity of FN, ongoing axonal degeneration, and poor reinnervation. The role of
physical therapy in improving outcomes after nerve injury is well known and appears
to be a major confounding variable in this case, considering that this was the only
patient with poor outcome in the tumor group less than 3 cm. Hence, from this series
we may conclude that noncompliance to physical therapy leads to poor outcome irrespective
of tumor size. FN is at risk during parotid surgeries and the role of IONM is gaining
wider acceptance for FN preservation in parotid surgeries. The study by Molinari et
al showed that in the patients with FN preservation, 2-year FN rehabilitation program
led to sustained significant improvement.[19]
Limitations
This study can be improved by correlating FN outcomes to extent of resection, cystic/solid
nature of the tumor, extent of tumor into internal auditory meatus, neurofibromatosis
status, and surgeon's experience. The main limitation of this study is that this protocol
relies mainly on FN identification, which appears rather simplistic without the absolute
and relative quantifications. Also, a comparison of this protocol with individual
techniques like train length, minimum stimulation threshold, BR study, and facial
corticobulbar motor-evoked potentials may be particularly useful in large tumors causing
brainstem compression. Finally, the outcome assessment relies on self-reporting. A
video consultation may be a good alternative whenever feasible. Follow-up may be extended
to 2 years because late recovery is known.
Conclusion
Thus, in conclusion, this study uses a simple IONM paradigm using dense muscle montage,
emphasis on FN anatomical preservation followed by physical therapy for at least 1
year in the postoperative period. This kind of protocol may be particularly useful
to implement in settings with limited technical expertise or advanced software which
preclude prognostication based on ratios.
