Multimodal intraoperative neurophysiological monitoring with special emphasis on facial MEPs for facial nerve preservation in vestibular schwannoma surgeries: Surgical nuances and outcome predictability

Inclusion and exclusion criteria

All VSs larger than 2 cm who were operated on for the first time, irrespective of NF2 status and radiological subtypes, were included in the study. In recurrent cases, those who were not willing to participate and those who had received previous radiotherapy were excluded from the study.

Anaesthesia and positioning

Anaesthetic considerations were taken, aiming at the use of drugs to not interfere with IONM techniques. Total intravenous anaesthesia was used with propofol and fentanyl combination. Muscle relaxant (short-acting) was used only during intubation. Field blocks using local anaesthetics around the FN or muscle were avoided as they could induce temporary paresis and confound the electrophysiological changes in IONM. Local anaesthetics near the stylomastoid foramen were also avoided. All of the patients were operated on in a supine position with their necks turned toward the opposite side of the tumour.

IONM

IONM was done using free-running electromyography (EMG), triggered EMG (using both monopolar flush tip stimulator and suction stimulator), FMEPs, electroencephalography (EEG) and its spectral derivatives and train-of-four testing. The terminal branches of the FN were monitored using subdermal paired-pin braided electrodes placed in the muscles, namely, frontalis, orbicularis oculi, nasalis, orbicularis oris and mentalis. Mandibular division of the trigeminal nerve was also monitored using a subdermal electrode placed in the masseter.

We used corkscrew electrodes for corticobulbar FMEP stimulation in Cz (cathode) – C3/C4 (anode) or Mz (cathode) – M3/M4 (anode) for hemispheric stimulation. The stimulation parameters were titrated in the following range: intensity was between 100V and 400V or 40 and 150 mA; a train of 3–8 rectangular pulses was used with a pulse width ranging from 75 us to 500 us, and the interpulse interval was in the range of 1.5–3 ms. This train stimulus was followed by a single pulse of the same characteristic delivered 50–90 ms later to differentiate true CoMEP from a peripheral response.

Direct stimulation using a flush-tip monopolar stimulator probe (Medtronic, USA) helped in the identification of FN and guiding tumour resection. A suction stimulator probe (Inomed, Germany) was used for quasi-continuous stimulus delivery and FN mapping. A single pulse of cathodal polarity with a width ranging from 200 to 500 us was used for triggered EMG, and it was delivered at a rate of 3 Hz for both monopolar flush tip and suction probe stimulation with a maximal intensity not exceeding 4 mA.

FMEPs helped in the assessment of the integrity of FN at any given instance. All consistent changes in FMEP and persistent neurotonic discharges were conveyed to the surgical team. A drop in FMEP amplitude of >50% constituted an alert.

Train-of-four was used to assess the residual neuromuscular blockade effect following the initial dose of relaxant administration during intubation. EEG and spectral derivatives of EEG were used to assess the depth of anaesthesia. Representative graphs of an illustrative case is demonstrated in Figures 1-4.

Close

Figure 1:

(a) The monopolar stimulator was not used concomitantly, as can be seen on the left side panel. (b) The surgeon is alerted to the proximity of the facial nerve by suction stimulator response at a threshold of 0.3 mA (right panel). (Channels from top to bottom: frontalis, oculi_1, oculi_2, nasalis, oris_upperlip, oris_lowerlip, masseter_1, masseter_2).

Export to PPT

Close

Figure 2:

Double-train technique for establishing true corticobulbar facial motor-evoked potentials – responses noted only following train stimulation and not after a single pulse stimulation (Channels from top to bottom: frontalis, oculi_1, oculi_2, oris_upperlip, oris_lowerlip, masseter).

Export to PPT

Close

Figure 3:

Contaminated facial responses due to direct activation of the peripheral facial nerve bypassing the corticobulbar pathway – as responses are noted following single pulse stimulation (Channels from top to bottom: frontalis, oculi, oris, masseter).

Export to PPT

Close

Figure 4:

Stack view of facial motor-evoked potentials (FMEP) drop followed by recovery following bipolar cautery to arrest bleeding: illustrated channels are oculi (left) and oris (right). There were no concomitant neurotonic discharges. Note from the timestamp that it took about 12 min for the FMEP drop to recover. During this time, surgical pause was initiated, blood pressure was raised, and papaverine was irrigated. This is a classical illustration of the identification of an ongoing facial nerve injury at an early reversible stage using FMEPs and initiation of corrective measures resulting in near-total recovery. Surgery was resumed following recovery.

Export to PPT

Surgical procedure

Classic retromastoid suboccipital craniotomy was done. A durotomy was done, and CSF was let out from the cerebellopontine (CP) angle cistern. The tumour was identified, baseline FMEPs were established, and the tumour capsule was stimulated with a monopolar probe to map the FN. Monopolar stimulation threshold intensity of <0.1 mA would indicate that the probe is on the nerve; a higher intensity of up to 1 mA would be required to stimulate an injured FN; higher stimulation intensities would also be required when testing through the bone (0.5–2 mA) or soft/tumour tissue (0.3–1 mA).^[12] The presence of fluid causes shunting of current, and it can produce a false-negative response. Hence, the use of stimulation threshold to guide resection is highly contextual and requires a good understanding between the surgical and the neurophysiologist team.

A simple heuristic that we practice is as follows: stimulation at currents lower than 0.2 mA indicates close proximity of the nerve to the probe, and it may be covered by a thin layer of tumour, which is at risk of injury if manipulated.^[13,14] If there is a thick tumour covering the nerve, stimulation thresholds shall be at currents >0.5 mA.^[14] Highly adherent tumours also need higher stimulation thresholds as they create thicker barriers.^[15] However, as has already been explained, this current-distance relationship is not so simple and depends on multiple factors.

The surgery proceeded with central debulking using an ultrasonic aspirator. The capsule of the tumour is dissected from the arachnoid plane and rolled inwards. Every time the part to be decompressed is swept with the suction stimulator; hence, the cycle of stimulation, decompression and mobilisation is repeated many times to thin out the tumour.

After achieving approximately more than half of the tumour resection, internal auditory canal (IAC) drilling was done, and tumour decompression in IAC was done while intermittently mapping the nerve with the monopolar stimulator. FN is identified at its proximal end at the brainstem and distal end at the IAC. If FN dissection at any point along its course was found difficult due to dense adhesion, part of the tumour capsule was left behind after adequately thinning it with the ultrasonic aspirator.

FMEPs were recorded at regular intervals and every time before changing the direction of dissection and before and after any coagulation when approaching the FN response zone. Any fall in the amplitude was notified, and, in such cases, the surgical field was irrigated with saline and diluted papaverine; blood pressure was raised, and surgery was paused to facilitate recovery of FMEPs, usually for a maximum of 15–30 min.

Whenever the electrophysiology team alerts the surgeon about high frequency neurotonic discharges, tumour manipulation was briefly paused to let the discharges subside, and papaverine and saline irrigation was done, if persistent.

An anatomical and functional preservation of the FN was confirmed at the end of the surgery by checking FMEPs and stimulating the nerve at the brainstem end. If the FMEP at closure was at least 50% of baseline response, we expect the functional integrity of the FN, while an FMEP amplitude drop of > 50% would predict an immediate mild-to-moderate post-operative facial weakness and if there was no response, a severe injury, which was predicted with dense facial palsy. Proximal stimulation threshold 0.05–0.1 mA would indicate good functional preservation^[16-18], whereas 0.2–1 mA would indicate some damage, at least in the immediate postoperative period, which may be due to temporary stunning (neuropraxia), which does not reflect the long-term outcome.^[14,19-23]

Inadvertent injury causing a breach in the anatomical integrity of FN was identified using monopolar stimulation and inspection of the course of the nerve. In such cases, it was sutured using 10-0 nylon and the anastomosis was stabilised with the use of thrombin glue Tisseel (Baxter International Inc., Westlake Village, CA, USA).

Post-operative care

After recovery, the FN was examined and compared with the opposite side and pre-operative status. Post-operative MRI was done three months following the surgery. The resection extent was considered gross-total resection (GTR) when no tumour was left, near-total resection when a thin layer of the tumour was left on the FN to prevent its injury, and subtotal resection when a significant tumour was left in the CP angle cistern. All the patients were followed up for a minimum of one year.

Statistical analysis

IBM Statistical Package for the Social Sciences Statistics 26 was used for analysing the data. Mean values were used to describe continuous variables, and percentages were used for categorical variables. A paired t-test was applied for continuous data. Pearson correlation test was used for the analysis of correlation in continuous data, and Spearman’s rho test was used for analysing the correlation when at least one parameter was ordinal. P ≤ 0.05 was considered statistically significant.

Prediction of post-operative FN outcome using IONM

The role of IONM has grown from an identification tool for mapping the nerve during surgery to an indicator of FN functional preservation and predictor of outcome.^[24,25] Several studies have correlated the intraoperative parameters such as A-train neurotonic discharges, stimulus threshold, response amplitude and proximal to distal amplitude ratio with FN outcome.

Post-resection stimulation thresholds at the root entry zone and response amplitudes have proven to predict good functional outcomes and have been used as prognostic indicators.^[24,25] In our series, seven cases who had stimulus threshold >1 mA (or nerve was not excitable) at the end of the surgery had higher grade (HB Grade III or IV) weakness at 3 months.

Tawfik et al.^[30] defined degradation of FMEP response as a final-to-baseline amplitude ratio of 0.5 or less and found that FMEP has high specificity (88.9%) and moderate sensitivity (54.5%) for predicting immediate post-operative function. Ling et al.^[31] also found that the FMEP amplitude ratio significantly correlated with short-term and long-term postoperative FN functions. Machetanz et al. have noted different sensitivity and specificity of FMEP and EMG in predicting post-operative FN palsy, and their combined use was found to be more accurate.^[32]

FMEPs are of major use in large VS, where the proximal part of FN will be identified at a much later stage of the surgery.^[33] Response amplitude of the FMEPs at the end of resection correlated with its function postoperatively. In our series, more than 50% fall in FMEPs correlated well with temporary higher grade facial palsy.

In our series, five cases had absence of electrophysiological responses (flat MEPs) at the end of the surgery and HB grade IV weakness at the 1^st week; they improved to Grade III in 3 months, hence in an anatomically intact FN, the absence of electrophysiological responses or spontaneous tonic/train activity cannot be considered as definite indicator of permanent FN paralysis as it may indicate temporary stunning of the FN conduction due to neuropraxia.^[34]

Drugs in outcome improvement

Perioperative steroids were used to control oedema and nerve damage. Papaverine, a vasodilator used during surgery, when we have fall in FMEPs, helps in relieving vasospasm. Nimodipine, a calcium channel blocker, helps in resprouting, growth of axons and remyelination.^[35]

Neurotisation in outcome improvement

With the use of IONM, we can identify the FN injury and do the neurotisation. The presence of response distal to the site of injury and its absence on proximal stimulation helps to localise the site of transection of FN. Following neurotisation, there may not be immediate recovery of the FMEPs, and there can be temporary high-grade palsy, but there will be a long-term improvement in FN function.

Another observation made in our study is that the use of thrombin glue (Tisseal) for stabilising neurotisation helps in the early recovery of FN function. This could be attributed to fibrin, which may help to align and stabilise the anastomosis, or calcium chloride, which could help in nerve conduction. However, this needs further evaluation to prove its efficacy, as it was based on only two case observations, and we have not done statistical analysis.