喉返神经术中监测和鉴别

Kaohsiung J Med Sci December 2010 • Vol 26 • No 12 633 © 2010 Elsevier. All rights reserved. Part of this work was presented at the 4th World Congress of the International Federation of Head and Neck Oncologic Societies (IFHNOS) in Seoul, Korea, June 15–19, 2010. Trial Registration: Clinicaltrials.gov. Identifier: NCT00629746 Received: Jul 5, 2010 Accepted: Aug 25, 2010 Address correspondence and reprint requests to: Dr Che-Wei Wu, Department of Otolaryngology, Kaohsiung Municipal Hsiao Kang Hospital, 482 Shanming Road, Siaogang District, Kaohsiung City 812, Taiwan. E-mail: kmuent@yahoo.com.tw INTRAOPERATIVE NEUROMONITORING FOR EARLY LOCALIZATION AND IDENTIFICATION OF RECURRENT LARYNGEAL NERVE DURING THYROID SURGERY Feng-Yu Chiang,1,2 I-Cheng Lu,3 Hui-Chun Chen,4 Hsiu-Ya Chen,3 Cheng-Jing Tsai,3 Ka-Wo Lee,1,2 Pi-Jung Hsiao,2,5 and Che-Wei Wu6,7 Departments of 1Otolaryngology, 3Anesthesiology and 4Nursing, and 5Division of Endocrinology and Metabolism, Department of Internal Medicine, Kaohsiung Medical University Hospital, 6Department of Otolaryngology, Kaohsiung Municipal Hsaio-Kang Hospital, 2Faculty of Medicine, College of Medicine, 7Institute of Clinical Medicine, Kaohsiung Medical University, Kaohsiung, Taiwan. Early and definite identification of the recurrent laryngeal nerve (RLN) is an important step to avoid inadvertent nerve injury during complicated thyroid operations. This study aimed to determine the feasibility of routine use of intraoperative neuromonitoring (IONM) to localize and identify the RLN at an early stage of thyroid surgery. This prospective study enrolled 220 consecutive patients (333 RLNs at risk) who underwent thyroid operations with application of IONM. The RLN was localized and identified routinely with a nerve stimulator after opening the space between the thyroid and carotid sheath. The success rates of early RLN localization and identification were evaluated. The current for localization and the amplitude of evoked laryngeal electromyographic signals were also recorded and analyzed. All RLNs, including 87 (26%) nerves that were regarded as difficult to identify, were successfully localized and identified. The stimu- lation level for RLN localization was 2 mA in 315 nerves (95%) and 3 mA in the other 18 nerves (5%). The signal obtained from RLN localization (amplitude = 932 ± 436 μV) showed a clear and reliable laryngeal electromyographic response that was similar to that from direct vagus (ampli- tude=811±389μV) or RLN stimulation (amplitude=1132±472μV). The palsy rate was 0.6% and no permanent palsy occurred. RLN injury is rare if the nerve is definitely identified early in the thyroid operation. The conclusion of this study is that IONM is a reliable tool for early RLN localization and identification, even in complicated thyroid operations. Key Words: intraoperative neuromonitoring, laryngeal electromyography, recurrent laryngeal nerve, thyroid surgery (Kaohsiung J Med Sci 2010;26:633–9) Kaohsiung J Med Sci December 2010 • Vol 26 • No 12634 F.Y. Chiang, I.C. Lu, H.C. Chen, et al Visual identification of the recurrent laryngeal nerve (RLN) during thyroid surgery has been proven to be associated with lower permanent palsy rates, and it has been recommended in many studies as the gold standard of RLN treatment [1–20]. With regard to iden- tification of the RLN, several nerve approach meth- ods have been reported [17–19]. In most instances, the RLN can easily be identified or even encountered at the tracheoesophageal groove after the thyroid lobe is retracted medially and elevated out of the wound. Nevertheless, the RLN might be difficult to identify intraoperatively because of anatomical variations [21], typically non-recurrent laryngeal nerve (non-RLN), or in complicated thyroid operations, such as those for recurrent goiter, large goiter with substernal extension, or thyroid cancer with significant paratra- cheal node metastasis. The position of the RLN can be significantly abnormal and the nerve can be fixed to the undersurface of a large goiter. In these circum- stances, the RLN can be injured before nerve identifi- cation, due to excessive medial traction on the thyroid, or the nerve can be inadvertently clamped or tran- sected during lateral dissection of the thyroid because of misidentification (the parallel artery can be mis- taken for the RLN, or the posterior sensory branch can be mistaken for the entire RLN) [2,21,22]. There- fore, early and definite identification of the RLN dur- ing thyroid surgery, especially in difficult cases, is an important step to prevent RLN injury. In the present study, we aimed to determine the feasibility of routine use of intraoperative neuro- monitoring (IONM) to localize and identify the RLN at an early stage of thyroid surgery. METHODS Subjects From September 2007 to October 2009, we prospecti- vely enrolled 220 consecutive patients who underwent thyroid operations with application of IONM, and all had been treated by the same surgeon (Dr FY Chiang). There were 97 total lobectomies and 123 total thy- roidectomies. Ten nerves were excluded from this study (8 with preoperative palsy and 2 with intentional transection due to cancer invasion). Finally, 333 nerves were at risk in this study. Among these 333 nerves, 98 nerves (29%) were dissected for thyroid malignancy and 19 (6%) for revision operations. Anesthesia and IONM system For general anesthesia, all patients were intubated with Nerve Integrity Monitor (NIM) Standard Reinforced Electromyography (EMG) Endotracheal Tube (6.0mm internal diameter for women and 7.0mm for men) (Metronic Xomed, Jacksonville, FL, USA). The position of the electrodes was routinely checked by laryngofibroscopic examination after the neck was placed at full extension to ensure that the middle of the blue marked region (3 cm of the exposed elec- trodes) was in good contact with the true vocal cords. A Prass monopolar stimulation probe (Medtronic Xomed) was used for nerve stimulation during thy- roidectomy. EMG activity was recorded on an NIM- response 2.0 monitor (Medtronic Xomed). Procedure for early RLN localization and identification Vagal stimulation After mobilization of the superior pole, the thyroid lobe and laryngotracheal complex were retracted medially as one unit to open the space between the thyroid and carotid sheath. The vagus nerve was rou- tinely tested to ensure the monitoring system was working and the normal pathway of the RLN. The vagus nerve was typically stimulated in the mid- neck region with a current of 2 mA, and the evoked EMG signals were defined as V1 signals. RLN localization After opening the space between the thyroid and carotid sheath, we localized the position of the RLN with a stimulating probe at the tracheoesophageal groove. Initially, we tried to localize the RLN near the level of the inferior thyroid artery. In some cases, such as recurrent goiter, large goiter with substernal extension, or thyroid cancer with significant paratra- cheal node metastasis, neural localization near the level of the inferior thyroid artery might be impeded; thus, RLN localization was performed at the upper or lower portion of the tracheoesophageal groove. The stimulation current used was 2 mA, and was increased to 3 mA if RLN localization failed with 2 mA. The EMG signals obtained from RLN localization were defined as L signals. RLN identification and dissection After the RLN was localized and identified, we tested it with a stimulation current of 1 mA for definite confirmation; then the nerve was dissected meti- culously to the entry of the larynx. In the case of large goiter with substernal extension, or thyroid cancer with paratracheal node metastasis, the RLN was dissected to the thoracic inlet until the nerve was well separated from the thyroid. The EMG sig- nals obtained from direct RLN stimulation were defined as R1 signals. The success rates of early RLN localization and identification were evaluated. The localization current, the position, and the evoked EMG signals (V1, L, and R1) were recorded and analyzed. Pre- and postoperative work-up All patients received pre- and postoperative exami- nation of cord mobility with flexible laryngofibroscopy. When asymmetric cord movement was found post- operatively, a comparison with the preoperative re- cording was performed. When vocal dysfunction was identified, a follow-up of 2 weeks initially and then every 4 weeks thereafter until recovery was achieved. Dysfunction was considered as permanent if it per- sisted for 6 months after surgery. This study was approved by the Institutional Review Board of Kaohsiung Medical University Hospital and ClinicalTrials.gov (NCT00629746). Written informed consent was obtained from each patient. Patients were informed of the intent to use this monitoring system potentially to aid in the localization and iden- tification of the RLN, and in the assessment of its function during surgery. There was no financial or professional association between the authors and the commercial company whose nerve-monitoring prod- uct was utilized. RESULTS As shown in the Table, all nerves were successfully localized and identified early at the tracheoesophageal groove, with the application of IONM. Among 333 RLNs, 315 (95%) were localized with a stimulation current of 2 mA and 18 (5%) with 3 mA. Eighty-seven (26%) nerves were regarded as difficult to identify (10 were dissected from the capsule of large recurrent goiters, 40 were dissected from thyroid cancer with significant paratracheal node metastasis, 35 were dis- sected from large goiters with substernal extension, and 2 were non-RLN). The signal obtained from RLN localization (L sig- nal) showed a clear and reliable EMG response that was similar to that from direct vagus (V1) or RLN (R1) stimulation; the mean amplitudes of each signal (V1, L and R1 signals) were 811 ± 389 μV, 932 ± 436 μV and 1,132±472μV, respectively (Figure). Two nerves (0.6%) developed temporary palsy and both were associated with inevitable stretch injury (one nerve was dissected from the large recurrent goiter which underwent the 3rd operation; the other nerve was meticulously dis- sected from thyroid cancer to which the nerve was adherent). No permanent RLN palsy occurred. Neuromonitoring for RLN localization Kaohsiung J Med Sci December 2010 • Vol 26 • No 12 635 Table. Characteristics of 333 recurrent laryngeal nerves with early localization and identification by intraoper- ative neuromonitoring during thyroid operation RLN (NAR = 333) n (%) Successful localization and identification 333 (100) Stimulation current 2 mA 315 (95) 3 mA 18 (5) Difficult identification 87 (26) Etiology Recurrent goiter with nerve adherent 10 (3) on thyroid capsule Cancer needs paratracheal LN dissection 40 (12) Large or substernal goiter needs 35 (11) extensive dissection Non-RLN 2 (0.6) RLN = recurrent laryngeal nerve; NAR = nerve at risk; LN = lymph node. Figure. Mean amplitude of electromyography signals obtained from vagus stimulation (V1 signal), recurrent laryngeal nerve localization (L signal) and direct recurrent laryngeal nerve stimulation (R1 signal). 0 V1 L R1 E M G a m pl it ud e (m V ) 811 932 1,132 200 400 600 800 1,000 1,200 1,400 1,600 1,800 DISCUSSION For identification of the RLN during thyroid surgery, some landmarks, such as inferior thyroid artery, tracheoesophageal groove, Berry’s ligament and the inferior cornu of thyroid cartilage, are used. Several authors [12,23–26] have reported that the laryngeal entry point represents the most constant position of the RLN in the neck, and that the inferior cornu of the thyroid cartilage is a consistent and reliable land- mark that indicates the laryngeal entry point. The laryngeal entry point of the RLN is approximately 1 cm below and just anterior to the inferior horn of the thyroid cartilage, which can be easily palpated, and the nerve is always found here. Therefore, the operative procedure for RLN identification, described by Harness et al in 1986 [19], is commonly used. They emphasized that the RLN is not identified early in the operative procedure and the thyroid lobe is dis- sected along its capsule. When dissection proceeds to the area of Berry’s ligament, the RLN is identified or even encountered where it runs through the ligament or close to it. It has been suggested that this proce- dure has several potential advantages, including less chance of disrupting the blood supply to the inferior parathyroid, and dissection of a shorter segment of the RLN with less risk of contusion injury. However, using this RLN approach, we experi- enced higher palsy rates in our previous study [5], and found several pitfalls that could cause nerve injury. (1) The thyroid lobe must be retracted medially and upwards, and excessive retraction of the thyroid can put the RLN at high risk of stretch injury in the region of Berry’s ligament. We have experienced eight nerves with stretch injury [5], and Snyder et al [2] also reported seven nerves with injury caused by over-traction near the region of Berry’s ligament. (2) The posterior branch of the RLN can be mistaken for the entire nerve under circumstances of partial nerve exposure. We have experienced one clamping injury of the anterior motor branch for this reason [5]. (3) The RLN in a large goiter can be displaced in any direc- tion and might even be adherent to the thyroid cap- sule at a lateral position, especially in a large recurrent goiter [21]. We have experienced one transecting injury during lateral dissection of a large goiter. The distorted RLN appeared to run vertical to the trachea; the RLN was mistaken for the inferior thyroid artery and it was transected inadvertently. (4) It is difficult to approach the RLN from the distal segment in the case of recurrent goiter, large goiter with substernal extension, or thyroid cancer with significant paratra- cheal node metastasis. After ascertaining the mechanism of RLN injury with the application of IONM, we changed our RLN approach to early identification. In this study, all RLNs were localized and identified early with a nerve stimulator after opening the space between the thy- roid and the carotid sheath. Most nerves were local- ized and identified at the tracheoesophageal groove near the level of the inferior thyroid artery. We found that there were several advantages with this RLN approach: (1) stretch injury caused by excessive trac- tion of the thyroid can be avoided; (2) extralaryngeal branches of the RLN, typically anterior and posterior branches, can be identified if they are present; and (3) the inferior parathyroid gland and its feeding vessels can be visualized and preserved. The operations for recurrent goiter, large goiter with substernal extension, or thyroid cancer with significant paratracheal node metastasis have been reported to be associated with higher RLN palsy rates [27–29]. In our experience, abnormal RLN position plays an important role in the occurrence of nerve injury. Randolph [30] also emphasized that goitrous enlargement can be associated with fixation and splay- ing of the RLN to the undersurface of the enlarged thyroid lobe, and RLN identification might be more difficult if there is significant paratracheal RLN chain nodal disease. In these situations, blunt dissection without nerve identification puts the nerve at high risk of injury. Therefore, identification of the RLN in such cases is a necessary initial step to prevent nerve injury. In the present study, among 19 nerves dissected for recurrent goiter, 10 were found to be adherent to the thyroid capsule, and these nerves could have been injured or even transected inadvertently if early localization and identification of RLN with IONM had not been performed. Temporary palsy occurred in one nerve that was undergoing its third operation. The nerve injury was caused by over-traction during dissection of the nerve from the capsule of a large recurrent goiter. Among 98 nerves operated upon for thyroid cancer, 39 were dissected from thyroid cancer with significant nodal metastasis. After localization and identification of the RLN, the nerve was dis- sected completely to the entry of the larynx, and then dissected downwards to the thoracic inlet until the Kaohsiung J Med Sci December 2010 • Vol 26 • No 12636 F.Y. Chiang, I.C. Lu, H.C. Chen, et al nerve was well separated from the metastatic nodes. We experienced one nerve with temporary palsy that was also caused by over-traction of the thyroid. The nerve was found to be adherent to the thyroid cancer, and the EMG signal was lost after the nerve was meticulously dissected from the tumor. Another 35 nerves were dissected for large goiter with substernal extension, and all of these were localized and identi- fied before tumor dissection. No nerve palsy occurred in these 35 nerves, although long exposure of the RLN was needed. Vagal stimulation before RLN localization is a nec- essary step to confirm that the monitoring system is working, to ensure the normal pathway of the RLN, and to provide reference data for the EMG signal [5]. Two non-RLNs were recognized as the signal from vagal stimulation could not be elicited, and both were localized and identified at the upper tracheoesoph- ageal groove. High EMG amplitude is the key to suc- cessful RLN localization. In this study, we routinely checked that the electrodes were in the optimal posi- tion by laryngofibroscopic examination after the neck was placed at full extension. In our previous studies, we have found that suboptimal EMG tube position (too deep, too shallow, or rotating) can result in equipment failure or reduce amplitude response [31,32]. In the present study, no equipment failure was encountered, and the signal obtained from RLN localization showed a clear and reliable EMG response that was similar to that from direct vagus or RLN stimulation. Several authors have reported that most nerve injuries are unexpected, and the actual causes of nerve injury are unknown in those with visual integrity of the nerves [33,34]. In our experience, anatomical vari- ations of the RLN, such as extralaryngeal branches, distorted RLN, non-RLN, or intertwining between branches of the RLN and inferior thyroid artery, can cause visual misidentification, and can result in unrec- ognized clamping or transecting injury of the RLN or the motor branch [21]. In this study, we found that early and definite identification of the RLN with IONM was a reliable way to prevent nerve injury caused by visual misidentification. Furthermore, when facing a patient with preoperative cord palsy, or when undertaking an operation in which one RLN has been invaded by thyroid cancer (10 patients in the present study), the surgeon can be more confident about avoiding the risk of postoperative bilateral cord palsy with this procedure. In the present study, all RLNs, including 87 (26%) that were regarded as difficult to identify, were suc- cessfully localized and identified at an early stage of surgery. Furthermore, similar to direct vagus or RLN stimulation, a clear and reliable EMG response was obtained from RLN localization. No permanent palsy occurred. Only two nerves (0.6%) that were associ- ated with inevitable stretch injury developed tempo- rary palsy. These results confirmed that IONM is a reliable tool to localize the RLN and facilitate its early identification during thyroid operations, especially in difficult cases. Using this procedure during thy- roid operation, RLN injury rarely occurs. ACKNOWLEDGMENTS This study was supported by grants from Kaohsiung Medical University Research Foundation (KMU- M099001) and the National Science Council (NSC 99- 2314-B-037-015-MY2). We thank the study participants for their contribution to this study. REFERENCES 1. Barczynski M, Konturek A, Cichon S. Randomized clinical trial of visualizati