Intraoperative Neural Monitoring in Neck Dissection

24 Intraoperative Neural Monitoring in Neck Dissection

Bradley R. Lawson, Dipti Kamani, and Gregory W. Randolph


Intraoperative nerve visualization is the accepted gold standard for prevention of neural injury during neck dissection. However, a structurally intact nerve does not always correlate with a postoperatively functioning nerve. Intraoperative neural monitoring has emerged as an adjunctive and additive tool for nerve identification and prognostication of postoperative function. While much of the existing data focus on surgery of the central neck compartment, we present emerging applications for monitoring of the nerves at risk in lateral compartment neck dissection.

Keywords: neural monitoring, thyroidectomy, central neck dissection, lateral neck dissection

24.1 Introduction

Neck dissection is a technically demanding operation that may present risk to numerous nerves during the course of a single procedure. Nerve visualization during surgery has long been considered the gold standard for the prevention of neural injury. However, an intraoperatively visualized and structurally intact nerve does not always correlate with a postoperatively functioning nerve. Neural monitoring has garnered increasing attention from thyroid and parathyroid surgeons worldwide, largely due to its prognostic information regarding postoperative nerve function. This technology has also been applied to lateral neck dissection. Emerging data are beginning to support the hypothesis that neural monitoring can reduce the incidence of neural injury. This chapter focuses on the history of intraoperative neural monitoring (IONM), its impact on surgical practice, standards of IONM, and application of IONM to lateral neck dissection.

24.2 Historical Overview

In 1848, Du Bois-Reymond became the first to demonstrate nerve action potentials and describe electrical activity of muscle with electromyography (EMG).1 IONM allows for localization of neural structures, tests the function of these structures, and provides early detection of neural injury. The goal of IONM is to identify neural injury prior to the onset of irreparable damage, thereby allowing immediate corrective actions to be taken.1 Historically, IONM has been most commonly used by spine surgeons, with neurosurgeons, vascular surgeons, orthopedic surgeons, urologists, and otolaryngologists all utilizing monitoring to some extent. The most common procedures in which IONM is applied include spine surgery, carotid endarterectomy, selected brain surgeries, and ENT procedures such as vestibular schwannoma resection, parotidectomy, thyroidectomy, parathyroidectomy, and neck dissection. Shedd and Durham in 1966 published the first report of electrical stimulation of the recurrent laryngeal nerve (RLN) and superior laryngeal nerve (SLN) in a canine model via endolaryngeal balloon spirography. A pressure recording from a balloon in the larynx consistently demonstrated recognizable changes upon stimulation of the RLN, thereby providing a means for its electrical identification. The authors were then able to confirm in two human patients that the endolaryngeal balloon pressure recording provided a clear signal of RLN and SLN stimulation.2 In 1970, Riddell published a 23-consecutive-year experience of electrical identification of the RLN, with the addition of laryngeal palpation as a confirmatory safety measure.3

Different IONM techniques have been developed over the past four decades, including laryngeal palpation, glottic observation, intramuscular vocal cord electrodes, postcricoid surface electrodes, anterior laryngeal electrodes, and endotracheal tube (ETT)-based surface electrodes.4,5 Due to the ease of setup and use as well as their noninvasive nature, ETT-based surface electrodes have become popular for IONM in central neck surgery.4,5,6,7 While intramuscular electrodes deliver higher amplitudes, they are more complicated to insert, may be placed in the wrong location, may migrate during surgery, and may even break in some cases.

In light of these many different techniques for IONM, standardization became a priority in order to promote uniform application of this emerging technology. The International Neural Monitoring Study Group (INMSG) was founded in 2006 to guide the emerging field of neurophysiologic monitoring, particularly of the vagus and laryngeal nerves in thyroid and parathyroid surgery.8

Vagal and RLN monitoring have gained widespread acceptance within the global surgical community. A recent study from Pennsylvania State University indicates that IONM is used by 80% of otolaryngologists and 48% of general surgeons who perform thyroid and parathyroid surgery at academic centers in the United States.9 Neural monitoring has become the standard of care in Germany; 93% of thyroidectomies were performed with RLN monitoring according to a national survey in 2010.10 Exposure to IONM during training is associated with a 3.1 times greater likelihood of using it in practice.11 Of note, use of IONM is actually more common among high-volume thyroid surgeons (> 100 cases per year). This suggests that IONM is not being used as a substitute for anatomical knowledge and surgical skill, but rather as a useful adjunctive tool to those who are most surgically experienced.11,12

Data continue to be discordant regarding the impact of neural monitoring on rates of neural injury in thyroid and parathyroid surgery. In a study of more than 850 patients undergoing revision surgery, Barczyński et al found transient and permanent RLN injuries in 2.6 and 1.4% of nerves with IONM versus 6.3 and 2.4% of nerves without IONM, respectively. The rate of transient paralysis was statistically significantly reduced when neural monitoring was used.13 Thomusch et al compared visual RLN identification with neural monitoring in over 5,000 procedures; they found rates of transient and permanent paralysis of 1.4 and 0.4% of nerves at risk with IONM versus 2.1 and 0.8% of nerves at risk with visual identification alone. A multivariate logistic regression confirmed that use of neural monitoring decreased the rate of postoperative transient (p < 0.008) and permanent (p < 0.004) RLN palsy as an independent factor by 0.58 and 0.30, respectively.14 However, Pisanu et al performed a meta-analysis of over 35,500 nerves at risk and found no statistically significant difference in the rates of transient and permanent RLN paralysis with and without neural monitoring.15

Dralle et al looked into this issue and found that an adequately powered study would require 9 million patients per arm for benign goiter and 40,000 patients per arm for thyroid malignancy surgery to detect statistical differences in the rate of RLN palsy with or without IONM.16 The argument for selective versus routine use of IONM continues to evolve among the community of surgeons who use it. We perform IONM in all cases in order to distribute the benefit of this additional information to all patients. Dionigi et al have also articulated the point that difficult cases may not always be apparent preoperatively.7 Routine use of IONM provides greater experience in interpreting the data, along with improved troubleshooting skills with the device itself should difficulties arise during a particularly challenging case. Current INMSG guidelines recommend the routine use of IONM for thyroid and parathyroid surgery, including central neck dissection.4 Currently, no guidelines exist for the use of IONM in lateral neck dissection.

24.3 Benefits of Intraoperative Neural Monitoring Application

The discussion relating to rates of RLN paralysis represents a single and rather limited perspective for evaluation of the use of IONM and its benefits to patients. Overall benefits offered by IONM include (1) nerve identification/neural mapping, (2) aid in dissection following nerve identification, and (3) injury identification/postoperative neural prognostication. When one appreciates the electrical information provided by IONM as additive and confirmatory to visual information, these benefits become rather apparent. Neural monitoring does not replace but adds to anatomic knowledge and surgical skill, and it provides a new functional dynamic. The need for visual identification of the nerve is not replaced by IONM; this technology is not intended to be used as a “divining rod.”

24.3.1 Neural Identification and Mapping

The RLN can be mapped out in the paratracheal region through linear stimulation with the neural probe. Visual identification then follows using directed dissection in the process referred to as neural mapping. Multiple studies report nerve identification rates of 98 to 100% using such neural mapping.5 Chiang et al reported a 100% identification rate, including nerves (25% of the total) that were classified as difficult to identify visually because of their complex anatomy.17 The previsualization neural mapping allows for rational and directed dissection, which may be a substantial advantage in scarred operative fields and cases with complex anatomy (such as ramified nerves, large goiters, revision surgeries, etc.).18

24.3.2 Aid in Neural Dissection

After the nerve has been visually identified, intermittent stimulation of the nerve versus adjacent nonnerve tissue can be helpful in tracing the nerve and its branches through the surgical field. This is analogous to intermittent facial nerve stimulation during parotidectomy. Accurate delineation of the medial border of the RLN can be very useful during ligament of Berry’s dissection.

24.3.3 Injury Identification and Prognostication of Postoperative Function

While monitoring is helpful in neural identification and is an excellent adjunct during nerve dissection, the key utility of IONM is the intraoperative prediction of postoperative function. A structurally intact nerve is not necessarily equivalent to a functional nerve. Blunt and stretch injury to the nerve may not always be visually detectable. Several studies have suggested that the visualization of the nerve by the surgeon is a poor judgment of RLN injury intraoperatively. Only 10 to 14% of injured nerves are identified as being such during the course of the surgery.19,20 Bergenfelz et al, when reviewing the Scandinavian endocrine quality register in over 3,660 cases, noted that they were able to intraoperatively identify nerve injury only in 11.3% of injured nerves. In addition, bilateral RLN injury was only identified during the operation in 16% of cases when it occurred.21 Snyder et al have also recently concluded that the majority of injured nerves are judged to be visually intact during surgery.22 Thus, visual examination is vastly insufficient to prognosticate postoperative RLN function.

In contrast, postoperative neural function prediction with IONM has been associated with uniformly high negative predictive values for injury ranging from 92 to 100%.23 Criteria supported by the INMSG for neurapraxic injury include a 50% or greater decline in amplitude and a 10% or greater increase in latency. In the absence of these conditions, the nerve is judged to retain normal physiologic function. This predictive ability of IONM is of particular importance in bilateral thyroid surgery, because both nerves governing the laryngeal airway introitus are at risk with one surgery. If neurapraxic injury identified on the initial side of surgery does not quickly recover, operative strategy may be rationally changed. Goretzki et al found that when loss of signal was identified on the first side, surgery could be terminated and a staged contralateral procedure performed at a later date with complete avoidance of bilateral nerve paralysis. However, when the surgeon continued to contralateral side surgery after loss of signal on the first side, 19% of patients developed bilateral vocal cord paralysis.24 The prognostic ability of IONM in the avoidance of the major morbidity of bilateral vocal cord paralysis is evident. This advantage is not amenable to statistical analysis but is likely one of the main reasons for IONM application.

24.3.4 Prognostic Testing Errors and Their Avoidance

As discussed previously, multiple recent studies report very high negative predictive value, making IONM vastly superior to visual identification of nerve injury. However, the following categories of errors may occur and should be considered by all monitoring surgeons. In these terms, we define positive (+) test as EMG loss of signal at the end of surgery (i.e., the test for the disease of postoperative vocal cord paralysis is positive), and we define negative (-) test as maintained EMG at the end of surgery (i.e., the test for vocal cord paralysis is negative).

1. False positives (i.e., loss of signal with intact neural function postoperatively). The causes of false positives include the following:

a) Various equipment problems on both the stimulation (i.e., faulty probe, inaccurate probe connection to monitor) and recording (i.e., ETT malposition or displacement) sides.

b) Neuromuscular blockade.

c) Blood or fascia obscuring the stimulated nerve segment.

d) Early-response elimination due to the monitor’s latency cutoff period for recording artifact suppression.

e) Vocal cord paralysis with early neural recovery.

Note that the majority of prognostic testing false positives relate to tube malpositioning.

2. False negatives (i.e., a good EMG with postoperative vocal cord paralysis).

a) Stimulation distal to the injured nerve segment (canine models suggest distal segments maintain electrical stimulability for up to 3 days). This is the primary rationale for vagal stimulation at the end of surgery.

b) Injuries subsequent to the final neural stimulation such as during wound irrigation, suctioning, and closure.

c) Delayed neurapraxia. One hypothesis is that progressive edema may affect the RLN at an intralaryngeal location such as the cricothyroid joint articulation.

d) Posterior RLN branch injury. Robust ETT electrode waveform confirms anterior branch RLN integrity. Posterior branches may be disrupted despite strong amplitude with stimulation, and such patients may have an abduction deficit postoperatively.25,26

e) Vocal cord immobility secondary to nonneural issues such as arytenoid cartilage dislocation or laryngeal edema.

24.4 Intraoperative Neural Monitoring Standards Guidelines

Despite the increasingly broad use of IONM, a review of the literature and clinical experience demonstrates there is significant variability in the application of neural monitoring across different centers. Variation exists in the use of pre- and postoperative laryngeal examination, a variety of stimulation probes and recording electrodes, and in monitor output with some providing only audio tone and others generating quantitative laryngeal EMG waveforms. Heterogeneity also exists regarding technique of ETT placement and the troubleshooting algorithm enacted when loss of signal occurs. The literature suggests that this nonstandard application of monitoring techniques leads to a significant rate of monitoring inaccuracies, most notably due to equipment-related problems such as ETT malposition in 3.8 to 23% of patients.27,28,29,30,31

24.4.1 Basic System Setup

Recording ground and nerve stimulator anode electrodes are placed on the patient’s shoulder and are interfaced with the monitor through a connector box (image Fig. 24.1). The recording electrodes from the right and left vocal cords exit the ETT proximally and are also interfaced with the connector box. Finally, the nerve stimulator cathode electrode (i.e., the probe) is placed on the sterile surgical field and connected to the box underneath the sterile drapes.

24.4.2 Anesthesia

Close partnership with anesthesiology is absolutely essential in a successful neural monitoring program.32,33 Anesthetic needs must be discussed prior to the case. Since IONM requires accurate and robust EMG response, neuromuscular blockade must be avoided without exception. Any neuromuscular blockade after induction could interfere with EMG activity; therefore, it is advisable to utilize short-acting neuromuscular blockade and to allow this to wear off following intubation.34,35

The ETT should be inserted without the use of lidocaine or any lubricant jelly. Pooled saliva may obscure the EMG signals, and using suction and possibly a drying agent may be helpful. The electrodes should abut very closely to the vocal cords; hence, the largest possible size tube for intubation should be used. Appropriate ETT electrode contact with the vocal cords must be confirmed after the patient is fully positioned. The INMSG has suggested two options for confirmation of optimal tube position prior to the start of surgical dissection. The first is to observe spontaneous respiratory variation, which is defined as the spontaneous bilateral EMG waveforms between 30 and 70 microvolts which is observed after the paralytic induction agent has worn off but before the inhalational plane of anesthesia becomes deep. Respiratory variation is typically present as the patient resumes spontaneous breathing and may be observed in conjunction with spontaneous movement or “bucking” (image Fig. 24.2). When respiratory variation is present unilaterally, this is an indication that the electrode has lost contact with the vocal cord due to rotation. Tube rotation may then be performed until bilateral waveforms are observed. The second option is to perform repeat laryngoscopy in order to visually ensure adequate ETT positioning. The video-laryngoscope may again be helpful for this post-positioning examination. Repeat laryngoscopy is recommended in all cases when respiratory variation cannot be identified. A recently published study by our unit found that identification of respiratory variation was possible in 91% of their patients, whereas the remaining 9% required a repeat laryngoscopy.33,34,35

After successful positioning of the ETT, the monitor setting should be assessed; low impedance values suggest good electrode–patient contact. The impedance of the electrodes should be less than 5 Ω and the imbalance between the two sides should be less than 1 Ω; monitor event threshold should be at 100 μV and the stimulator probe should be set to a pulsatile output of 4 per second with stimulating current set between 1 and 2 mA. At the onset of surgery, the stimulation of strap muscles resulting into a gross muscle twitch can be performed to confirm the absence of neuromuscular blockade as well as an intact stimulatory pathway. Vagal stimulation (V1) is performed prior to formal surgical dissection of the central neck compartment. It is only when the vagus nerve is stimulated and provides robust EMG activity that the surgeon may be assured that the system is fully functional and that the RLN can be safely sought after through the neural mapping technique. Only once the vagus nerve has been positively stimulated (true positive) can a subsequent negative response be regarded as a true negative. For each patient, IONM data must essentially include preoperative laryngeal exam (L1), an initial intraoperative suprathreshold vagal nerve stimulation (V1), an initial intraoperative RLN stimulation (R1), and also a similar set of events (R2 and V2) should be recorded at the end of the surgery followed by a postoperative laryngeal exam (L2).

Feb 14, 2020 | Posted by in OTOLARYNGOLOGY | Comments Off on Intraoperative Neural Monitoring in Neck Dissection

Full access? Get Clinical Tree

Get Clinical Tree app for offline access