Intraoperative Monitoring of the Recurrent Laryngeal Nerve

Laryngeal Nerve Monitoring

Intraoperative Monitoring of the Recurrent Laryngeal Nerve in 151 Consecutive Patients Undergoing Thyroid Surgery
1. Thomas M. Hemmerling, MD, DEAA*, 2. Joachim Schmidt, MD*, 3. Christian Bosert, MD*, 4. Klaus E. Jacobi, MD* and 5. Peter Klein, MD†
+Author Affiliations

1. Departments of *Anesthesiology and †Surgery, University Erlangen-Nuremberg, Erlangen, Germany
1. Address correspondence and reprint requests to Thomas Hemmerling, MD, DEAA, Centre Hospitalier de l’Université de Montréal, Hôtel-Dieu, Départment d‘Anesthésie, 3840, Rue Saint-Urbain, Montreal (Quebec), H 2W 1T8, Canada. Address e-mail

Next Section

IMPLICATIONS: We present a technique of intraoperative monitoring of the recurrent laryngeal nerve using a surface electrode attached to a routine endotracheal tube. The technique proved noninvasive, easy to use, and reliable in 151 prospective consecutive patients for preventing permanent laryngeal nerve damage in thyroid surgery.

Intraoperative identification and preservation of the recurrent laryngeal nerve is essential to thyroid surgery. The risk to the recurrent laryngeal nerve increases with the extent and difficulty of thyroidal surgery (1). The incidence of either a temporary or permanent paresis may be as frequent as 20% in thyroid cancers or recurrent goiter (2,3). The latest methods of intraoperative monitoring involve the use of special disposable endotracheal tubes with integrated electrodes (4–6). These electrodes, however, are expensive and cannot be positioned when other specialty endotracheal tubes are required.

We report a prospective study in which intraoperative monitoring of the laryngeal nerve (IRM) was performed by using a surface electrode attached to a routine endotracheal tube.

Previous SectionNext Section

After approval of the local ethics committee and informed consent from all patients, we performed intraoperative IRM on 151 consecutive patients (46 men and 105 women) undergoing thyroid surgery.

Anesthesia was induced and maintained with an infusion of remifentanil at 0.5 μg · kg−1 · min−1 and a target-controlled infusion of propofol (initial and maintenance target concentration, 4 and 3 μg/mL, respectively). Intubation was performed without aid of neuromuscular blockade by using a routine Woodbridge® tube (Mallinckrodt, Oxfordshire, UK) with the recording surface laryngeal electrode attached circularly 2 cm above the cuff and placed between the vocal cords (Fig. 1). The electrode (Magstim Company, Whitland, Wales, UK) used was a disposable long, plastic sheet consisting of eight incorporated electrodes with a distal self-adhesive recording part. Care must be taken not to create folds in the recording part to avoid interferences.

View larger version: In this page In a new window Download as PowerPoint Slide
Figure 1. Disposable electrode system used for intraoperative recurrent laryngeal nerve monitoring attached to the endotracheal tube 2 cm above the cuff. The picture shows the distal recording part of the electrode, glued circularly around the endotracheal tube. It is important to attach the distal part in such a way as to avoid folds, which make monitoring impossible.

Recurrent laryngeal nerves were first identified after the induction of anesthesia before skin incision by percutaneous stimulation. It was thought that such preoperative identification might help the surgeon in intraoperative localization. The nerve stimulator (Multiliner®; Toennies Co., Wuerzburg, Germany) (Fig. 2) was placed close and just lateral to the notch of the thyroid cartilage by stimulating the recurrent laryngeal nerve where it penetrates the cricothyroid muscle. Single-twitch stimulation (frequency, 1 Hz; pulse width, 0.2 ms) was performed on the left and right recurrent nerve. The best stimulation site was determined and recorded. The current was increased from 0 mA to a current that gave the maximal electromyography (EMG) response (<70 mA) and then increased by a further 10 mA to ensure supramaximal stimulation. The amplitude of the compound action potentials, the latency after stimulation, and their wave form were measured and recorded.

View larger version: In this page In a new window Download as PowerPoint Slide
Figure 2. Preoperative localization of the recurrent laryngeal nerve by using percutaneous stimulation. The area of primary stimulation attempt is marked, just lateral of the thyroid notch. The percutaneous probe is then moved in all directions to determine the maximum signal response. The left graph shows a typical triphasic compound action potential after percutaneous stimulation.

During the operation, the recurrent laryngeal nerve was monitored with a Neurosign® 100 recorder (Magstim Company). The Neurosign® 100 was connected to a surgical stimulation probe (Inomed Company, Teningen, Germany) 10 cm long and delivered 3-Hz impulses and a current between 0 and 5 mA, allowing determination of impulse thresholds.

During surgical exposure, the thyroid gland is elevated and retracted medially, exposing the region of the inferior thyroid artery and the recurrent laryngeal nerve. The inferior thyroid artery is under run cleanly and precisely with a vessel loop. The operator confirms the laryngeal nerve or the region where he or she suspects it by direct electrical stimulation with the handheld bipolar stimulation probe (Fig. 3). Both an audible signal and a recorded visual response confirm that the nerve is isolated. When the nerve is not evident, an aberrant course should be suspected. Here the bipolar electrode is helpful in actively searching the nerve. With the stimulating current threshold set at 1 mA, the audio signal increases as the electrode approaches the location of the nerve.

View larger version: In this page In a new window Download as PowerPoint Slide
Figure 3. Intraoperative monitoring of the laryngeal nerve using a hand-held bipolar stimulating probe. The bipolar stimulating probe points toward the recurrent laryngeal nerve just above the inferior thyroid artery.

Direct laryngoscopy was performed to confirm recurrent laryngeal nerve function on the third postoperative day and again on the fourth week after the operation in those patients in which the first laryngoscopy identified a paresis. Patient data are presented as mean ±SD. Latency, amplitude, and form of the EMG signal (biphasic or monophasic) were compared between groups of different thyroid diseases by using analysis of variance and Pearson’s test for possible correlation; P < 0.05 was regarded as showing significant differences. Stimulation thresholds before and after resection were compared by using paired Student’s t-test (P < 0.05).

Previous SectionNext Section

A total of 151 patients underwent thyroid surgery during the study period. The characteristics of these patients are summarized in Table 1.

View this table: In this window In a new window
Table 1.
Demographic Data of 151 Patients Who Underwent Thyroid Resection

Two-hundred-sixty-six nerves were identified during surgery (34 nerves needed no identification because of unilateral thyroid surgery). In 62 of these 266 nerves (23%), IRM was not only used to confirm an already surgically exposed structure, but also to locate and identify the nerve before surgical exposure of the nerve (Table 2).

View this table: In this window In a new window
Table 2.
Indication for Thyroid Resection and Percentage of Nerves Where IRM was Used to Locate and Identify the Nerve

Triphasic action potentials were most often obtained at percutaneous stimulation (70% vs 30%, bi- or monophasic, P < 0.001) (Fig. 2). Of the 24 nerves identified in patients with thyroid cancer, however, 10 had a biphasic action potential and 14 a monophasic action potential, being statistically significantly different from the other groups (P < 0.001). No correlation or significant difference could be demonstrated between supramaximal current of the percutaneous stimulation, amplitude, and latency of the compound action potential and the indication for surgery or the size of the thyroid gland. The mean current was 45 ± 7 mA, the mean amplitude was 1.2 ± 0.4 mV, and the mean latency was 3.5 ± 0.5 ms. There was no difference between the threshold of the stimulating current amplitude before or after thyroid resection (mean amplitude, 0.4 ± 0.2 mA before and 0.4 ± 0.3 mA after resection).

A preoperative unilateral paresis was present in three patients, as demonstrated by direct laryngoscopy before surgery. A recurrent laryngeal paresis was identified in seven patients on direct laryngoscopy after surgery (Table 3). Six of these pareses were unilateral with complete glottis closure. One bilateral nerve paresis was demonstrated. Six of these pareses were temporary, as confirmed by direct laryngoscopy 4 wk after surgery, when a complete return to function was noted. Two of these patients required speech therapy. In one patient with a thyroidal cancer, the nerve was within the tumor substance. A decision was made during surgery to perform a curative R0 resection that involved a deliberate division of the nerve.

View this table: In this window In a new window
Table 3.
Details of Patients with Paresis and Outcome

Previous SectionNext Section

This study shows that IRM with a surface electrode attached to a routine tube is a reliable method for monitoring and identifying the recurrent laryngeal nerve. In 23% of the nerves, IRM was not only used to confirm the recurrent laryngeal nerve, but also to locate and identify the nerve before surgical exposure. There was a very small incidence of temporary paresis (2%). One nerve was deliberately dissected to allow complete removal of thyroid cancer. Preoperative recording of nerve response via percutaneous stimulation revealed a prevalence of biphasic and monophasic compound action potentials in thyroid cancer in comparison to predominant triphasic potentials in other patients, suggesting subclinical nerve impairment (7). Stimulation current, signal latency, amplitude, or intraoperative stimulation thresholds did not correlate with disease or operative procedure. There was no incidence of unintentional permanent nerve damage, in comparison to a permanent palsy rate of 0.5% even in specialized centers (8) and a smaller incidence of temporary recurrent laryngeal nerve pareses (2% vs 3.6%).

We observed in one patient (in whom the recurrent nerve was deliberately transected for radical tumor removal) that stimulation of the recurrent laryngeal nerve immediately after and proximal to the dissection resection still created a recordable response. One study of deliberate transection of the ulnar nerve in dogs showed that a palpable muscle response is weakened at 24 hours but can be detected up to four to five days after injury (9). This might explain why evoked EMG responses, regardless of the electrodes used (10), cannot be used to detect temporary, incomplete nerve damage during surgery. Although some studies describing alternative methods IRM have identified nerve irritation, presumably by the electrode resulting in reversible pareses (4,11,12), we have not observed such a problem with our technique.

The success of this method relies on selection of tubes of sufficient diameter (more than size 6-mm diameter) because the “flaps” of the electrode otherwise overlap, causing a short circuit. The correct positioning of all the connecting cables is also important, and we have noted that the presence of a kink in the cable connecting the recording electrode to the monitor can create signal interference.

Intubation without aid of neuromuscular blockade and placement of the surface electrode between the vocal cords was successful at the first attempt in all patients. Neuromuscular blockade diminishes the response traced by the surface electrode. When neuromuscular blockade is required for intubation, a short-acting nondepolarizing muscle relaxant can be used, and percutaneous stimulation of the recurrent laryngeal nerve can be performed after the return of normal neuromuscular transmission.

The question of whether any kind of monitoring system should be used for thyroid surgery was not the purpose of this study, nor was a detailed cost-benefit calculation. Apart from the purchase of the monitoring device (approximately US$5000), there is the cost of US$25 per single-use electrode.

Previous SectionNext Section
In conclusion, this EMG technique is a safe, noninvasive, and reliable method of IRM. When we compare our results with those of even specialized centers, where a traditional, visual technique of recurrent laryngeal nerve identification is used, a reduction in transient recurrent laryngeal nerve paresis can be demonstrated with no case of unintentional permanent nerve damage.

* Accepted April 3, 2001

Changed in Medtronic NIM ET EMG Tube