Clinical evoked electromyography for recurrent laryngeal nerve preservation: Use of an endotracheal tube electrode and a postcricoid surface electrode

Abstract

Surgery of the thyroid and parathyroid glands is fraught with the possibility of iatrogenic or incidental injury to the recurrent laryngeal nerve (RLN). Kern1 reported that iatrogenic RLN injury was the most common cause of lawsuits in endocrine surgery. The debate over strategies for preservation of the RLN has raged between those surgeons who advocate avoidance of the nerve (“close or intracapsular” dissection) and those who advocate initial nerve location and distal dissection with the nerve in direct view. For otolaryngologists, this argument is very much like the debate over parotidectomy, which has ultimately been resolved in favor of nerve location with the assistance of an electronic nerve stimulator/locator. In the case of thyroid surgery the issue is complicated by the fact that the target muscles in the larynx are not visible during the surgery. In order to use an electronic nerve stimulator in thyroid surgery, some method of identification other than visible movement of the muscle response must be used. Surgeons have advocated the use of a stimulator with endotracheal balloon pressure transducers, flexible intraoperative nasolaryngoscopy, and palpation of the laryngeal “thrill.” Because of the long clinical experience with diagnostic electromyography (EMG), this technique has been studied for measuring the evoked response. The first efforts to do evoked laryngeal EMG have been with needle electrodes, placed either by transoral, transcutaneous, or translaryngeal approach. Problems associated with needle electrodes include difficulty in placement (and corresponding increased skill required of the surgeon or electromyographer), difficulty in retention (perhaps requiring barbs to increase retention), and problems with hematoma or abscess formation. For these reasons, surface electrodes have been developed as a noninvasive method of monitoring for laryngeal evoked EMG. Figure 1 shows the laryngeal surface electrode (LSE),2 (RLN Systems, Inc., Jefferson City, MO), the first device designed for noninvasive evoked laryngeal EMG monitoring. This paddle- or shovel-shaped device rests in the postcricoid space and samples the signal from the larynx, primarily from the posterior cricoarytenoideus muscles. The LSE is anatomically stabilized: inferiorly by the upper esophageal sphincter, laterally by the pharyngeal walls, posteriorly by the spine, and anteriorly by the larynx itself. Superiorly its handle buttresses the electrode. The electrode has two plates, which face the left and right posterior cricoarytenoideus muscle, respectively, thereby monitoring both sides of the larynx concurrently. Insertion is peroral after positioning and entubation of the patient by raising the larynx, “sundowning” the electrode in the postcricoid space, and setting the larynx down on top of the electrode. The electrode has a third plate on its posterior aspect, which can be used as a ground or for monitoring the nerve supply to the inferior constrictor muscle of the pharynx. Position is verified by measuring the electrode impedance, checking for ECG artifact, and by visual verification. The electrode costs $70.00 and is adjustable so that one size fits all. The laryngeal surface electrode. Figure 2 shows the endotracheal tube electrode (ETE) (Xomed-Treace, Jacksonville, FL). Initially described by Eisle,3 it consists of two sets of paired, bare wires placed laterally on either side of a special endotracheal tube, just above the balloon. The variable outer diameter of the endotracheal tube provides for sufficient projection of the electrodes so that the vocalis muscle is stretched over the tube surface, thereby providing contact with at least one of the pair of electrodes. Monitoring of both sides of the larynx is provided. Retention is provided by the balloon and by securing the endotracheal tube at the mouth. Position is verified by measuring the electrode impedance and by visual verification. The electrode costs $175.00 and comes in three endotracheal tube sizes. Two studies have compared the two-surface electrode's performance under controlled conditions. Severtson et al.4 studied four electrode systems concurrently in 10 dogs: 1. peroral needle electrodes in the posterior cricoarytenoid muscles, 2. transglottic needles placed via the thyrohyroid membrane, 3. the endotracheal tube electrode, and 4. the laryngeal surface electrode. The LSE was judged superior on the basis of higher signal quality and one failure to monitor the ETE. Khan et al.5, 6 studied 87 patients who underwent monitoring during thyroidectomy and parathyroidectomy—49 with LSE alone, 16 with ETE alone, and 22 with both LSE and ETE simultaneously. This series contained many very difficult cases with very large goiters and other complicating technical problems. There were 113 nerves at risk in this series, three permanent recurrent nerve palsies, and an additional three temporary palsies. Khan rated monitoring as “crucial to successful accomplishment of surgery and preservation of the RLN” in 21 of the 87 cases but made no definitive judgment of the superiority of one surface electrode over the other. Rea and Khan7 have reported a series of cases of evoked nerve monitoring during thyroidectomy and parathyroidectomy. The cases involving 54 nerves at risk done by Rea in a private-practice setting generally represented a lesser degree of difficulty and risk to the nerve compared with those of Khan. This series was collected during the development of the LSE and there were several failures to identify evoked potentials either because of misplacement of the electrode or problems with associated EMG equipment, which was also under development. There were no accidental nerve injuries in this series, although one nerve was intentionally sectioned. The endotracheal tube electrode. Evoked laryngeal EMG monitoring for preservation of the RLN requires complete understanding of the process and a high degree of motivation on the part of the surgeon to be successful. Additional surgical time must be allotted to placing and securing the electrodes, verifying positioning, and evaluating responses. The choice of EMG unit is influenced by whether the surgeon intends to have a dedicated electrophysiologic monitoring assistant in the operating room. Units, which have been used in the above studies, include the Xomed NIM (Xomed-Treace, Jacksonville, FL), Nicolet Spirit (Nicolet Biomedical Inc., Madison, WI), Axon Sentinel (Axon Systems, Inc., Hauppage, NY), and RLN Systems Neurovision (RLN Systems, Jefferson City, MO). Placement of the electrode is critical to performance. In the case of the LSE, the surgeon places the electrode after initiation of anesthesia, intubation, and positioning. The larynx and endotracheal tube are elevated with a laryngoscope and the electrode is positioned in the pharynx and slid along the posterior pharyngeal wall until the electrode can be “sundowned” behind the arytenoids. The larynx is then placed on top of the electrode and the endotracheal tube and LSE secured with tape. The initial positioning of the LSE is critical for proper monitoring. The anesthetist commonly places the ETE. The tube is placed with the exposed wires even with the vocalis muscles and secured by inflating the balloon and taping at the mouth. Electrode position is verified in the LSE by ECG artifact, impedance of less than 1 ohm between the separate channels, and by direct visual observation. The ETE is readjusted until positioning can be verified by impedance and visually. Electrode impedance measurements are critical for the ETE to verify that the wire pair being monitored is touching the vocalis. Analysis of EMG with surface electrodes is somewhat different compared with needle EMG. Because larger areas of the muscle are being sampled, responses tend to be lower in frequency, higher in amplitude, and more characteristic of a compound motor action potential. An evoked response is identified by recognizing the stimulation artifact and by identifying the characteristic multiphasic compound motor action potential, a damped oscillation that occurs with a delay of about 1 ms after the stimulation event and lasts 5 to 10 ms. Figure 3 shows representative evoked EMG signals from the LSE and ETE taken during concurrent monitoring. Trace 1 is the ESE and trace 2 is the LTE from the same side. CAP peak levels were 392 mV for the ETE and 3058 mV for the LSE. Screen picture comparing the evoked EMG signal from the ETE (trace 1) and the LSE (trace 2) during concurrent monitoring of a hemithyroidectomy with a NIM-2 unit. (Courtesy of Dr. Gregory Randolph.) Inability to evoke an EMG signal is the most serious complication reported to date with both these electrodes. In the case of the LSE, this usually reflects malpositioning of the device on initial insertion. In patients with short or stiff necks, high larynges, or very small mouths, proper visualization and positioning of the LSE can be difficult. A strict adherence to protocol and direct visualization of the postcricoid larynx usually prevent this problem. The ETE, although easier to insert, is a less stable platform for extended monitoring. Softening of the endotracheal tube due to warming and movement of the endotracheal tube have been reported as possible reasons for loss of contact with the vocal cords. Frequent recourse to impedance measurements will reassure that contact is being maintained. This potential complication could affect both electrodes' performance and must be kept in mind. Progressive neuropraxia of the nerve can mimic electrode malposition. In this case all indicators will be positive for electrode position and impedance but no signal or a reduced signal will be obtained. Movement artifact is a problem with both surface electrode systems, the LSE being sensitive to anterior-posterior movement of the larynx and the ETE being sensitive to movement of the endotracheal tube. Movement artifact is worsened by the fact that the thyroid gland is generally adherent to the larynx/trachea and normal dissection causes some movement. The false alarms of movement artifact distract the surgeon from the surgical task. Two of the nerve palsies reported by Khan were large goiters with stretching of the nerve over its bulk and leading to disruption of the integrity of the nerve, commonly referred to as splaying. Analogous with acoustic tumor surgery, this condition often leads to a rapidly neuropraxic nerve during dissection with loss of monitoring and high risk of permanent nerve injury even if continuity is preserved. Evoked laryngeal monitoring to preserve the RLN is an evolving clinical tool. The present techniques are not refined to the point where monitoring is a seamless, unobtrusive process. The perceived need for monitoring will vary with the difficulty of the case. Clear superiority of one of the two surface electrodes evaluated is not apparent. The LSE seems superior in signal quality, reliability, and cost. The ETE is superior in ease of insertion, a significant factor when the surgeon is a nonotolaryngologist. Given the nonreconstructable nature of the RLN and the serious morbidity associated with its injury, it would seem that monitoring this nerve will become a routine clinical practice.