Abstract
Purpose
Extensive dissection of recurrent laryngeal nerve (RLN) is inevitable in some complicated thyroid operations. The study aimed to determine whether extensive dissection of RLN increases the risk of nerve injury.
Method
Three hundred thirty-one patients (506 nerves at risk) who underwent thyroid operations with intraoperative neuromonitoring were included. The study chiefly focused on the 101 RLNs on which extensive nerve dissection from the thoracic inlet to the entry of larynx was performed and for which the nerve exposure was longer than 5 cm. Electromyographic (EMG) signals were obtained from the RLN and vagus nerve before and after complete RLN dissection, and these were defined as R 1 , V 1 and R 2 , V 2 signals, respectively. The RLN palsy rates and the change of EMG signals were evaluated and analyzed.
Results
Among 101 nerves with extensive dissection, 13 nerves were due to the operation for recurrent goiter; 41 nerves, for large goiter with substernal extension; and 47 nerves, for thyroid cancer with paratracheal nodal metastasis. No permanent palsy occurred, but 2 nerves experienced loss of EMG signal after complete RLN dissection from a large recurrent goiter and developed temporary palsy. The palsy rates were 2% (2/101) in the extensive dissection group and 2.5% (10/405) in the nonextensive dissection group ( P = .77). Among 99 nerves with normal vocal function after operation, none experienced weakened signal after complete RLN dissection, and the mean amplitudes of R 2 and V 2 signals were not significantly different from those of R 1 and V 1 signals (R 2 vs R 1 ; 1038 vs 1030 μ V; P = .74; V 2 vs V 1 ; 824 vs 816 μ V; P = .75).
Conclusions
The results of this study suggest that careful surgical dissection is well tolerated by the RLN.
1
Introduction
Many studies have proven that routine identification of recurrent laryngeal nerve (RLN) during thyroid operation is associated with lower rates of postoperative RLN palsy and have recommended it as the criterion standard of nerve treatment . However, RLN palsy can still occur with permanent palsy rates of up to 2% and temporary palsy rates of up to 5% to 6 % . Such RLN injury can result from transecting, clamping, stretching, electrothermal injury, ligature entrapment, or ischemia, but the actual causes are still not well understood, especially in those with visual nerve integrity . Several authors believed that the RLN was extremely vulnerable to surgical dissection and postulated that long exposure of RLN was a risk factor for RLN palsy. Thus, they recommended that the RLN should undergo as minimal dissection as possible during thyroidectomy . However, in some situations such as the operations for recurrent goiter, large goiter, substernal goiter, or thyroid carcinoma for which paratracheal nodal dissection is undertaken, extensive dissection of RLN is inevitable. In this study, we aimed to determine whether extensive dissection of RLN during thyroid operation increases the risk of nerve injury. We compared the RLN palsy rates between the extensive and nonextensive dissection groups. We also evaluated the change of the electromyographic (EMG) amplitudes before and after extensive dissection of RLN.
2
Materials and methods
From April 2006 to October 2009, 331 patients undergoing operations for various thyroid diseases were treated by the same surgeon (F.Y.C.). There were 142 total lobectomies and 189 total thyroidectomies. Fourteen nerves were excluded from this study (11 nerves had preoperative cord palsy; 3 nerves were sacrificed intentionally because of cancer invasion). Thus, in all, 506 nerves at risk were enrolled in this study. Among these nerves, 151 were operated for thyroid cancer; 26, for reoperation; and the other 329, for benign thyroid disease. We chiefly focused the study on the 101 RLNs on which extensive nerve dissection from the thoracic inlet to the entry of larynx was performed and for which the nerve exposure was longer than 5 cm.
All patients were intubated for general anesthesia with a Medtronic Xomed Standard Nerve Integrity Monitor EMG endotracheal tube (Jacksonville, FL). Stimulation of RLN was routinely performed before and after complete nerve dissection. The R 1 signal was defined as the EMG signal that was obtained from the RLN when it was first identified at tracheoesophageal groove. The RLN was touched directly by the nerve stimulator (Prass monopolar probe) with a stimulation level of 1.0 mA. The R 2 signal was defined as the EMG signal that was obtained from the proximal portion of RLN after complete nerve dissection. Because the procedure of vagal stimulation can help to confirm that the monitoring system is working, ensure the normal pathway of RLN, provide the reference data of EMG signal, and ensure that neural testing is not performed distal to the site of RLN injury, vagus nerve stimulation was also routinely performed before identification of RLN and after complete hemostasis of surgical field, and the EMG signals were defined as V 1 and V 2 signals, respectively. This study was approved by the Institutional Review Board of Kaohsiung Medical University Hospital and the ClinicalTrials.gov ( http://www.clinicaltrials.gov . [identifier: NCT00629746 ]). Written informed consent was obtained from each patient. Patients were informed of the intent to use the monitoring system potentially to aid in the localization and identification of the RLNs and assessment of their function during operation.
All patients received preoperative and postoperative video recording of vocal cord movement with flexible laryngofibroscopy, and all exposed RLNs were routinely measured in length and were photographically documented with a high-resolution camera to show visual nerve integrity after complete nerve dissection. When asymmetric cord movement was found postoperatively, a comparison with the preoperative recording would be performed. When vocal dysfunction was identified, follow-up was per 2 weeks initially and every 4 weeks thereafter until recovery was achieved. Dysfunction was considered permanent if it persisted for 6 months subsequent to operation.
Statistical analysis of continuous variables was compared using independent t test and was presented as mean ± SD. Categorical nominal variables were analyzed by χ 2 test or the Fisher exact test. All analyses were performed using SPSS for Windows version 13.0 (SPSS, Inc, Chicago, IL). All statistical tests were 2-sided, and significant level was set at P < .05.
2
Materials and methods
From April 2006 to October 2009, 331 patients undergoing operations for various thyroid diseases were treated by the same surgeon (F.Y.C.). There were 142 total lobectomies and 189 total thyroidectomies. Fourteen nerves were excluded from this study (11 nerves had preoperative cord palsy; 3 nerves were sacrificed intentionally because of cancer invasion). Thus, in all, 506 nerves at risk were enrolled in this study. Among these nerves, 151 were operated for thyroid cancer; 26, for reoperation; and the other 329, for benign thyroid disease. We chiefly focused the study on the 101 RLNs on which extensive nerve dissection from the thoracic inlet to the entry of larynx was performed and for which the nerve exposure was longer than 5 cm.
All patients were intubated for general anesthesia with a Medtronic Xomed Standard Nerve Integrity Monitor EMG endotracheal tube (Jacksonville, FL). Stimulation of RLN was routinely performed before and after complete nerve dissection. The R 1 signal was defined as the EMG signal that was obtained from the RLN when it was first identified at tracheoesophageal groove. The RLN was touched directly by the nerve stimulator (Prass monopolar probe) with a stimulation level of 1.0 mA. The R 2 signal was defined as the EMG signal that was obtained from the proximal portion of RLN after complete nerve dissection. Because the procedure of vagal stimulation can help to confirm that the monitoring system is working, ensure the normal pathway of RLN, provide the reference data of EMG signal, and ensure that neural testing is not performed distal to the site of RLN injury, vagus nerve stimulation was also routinely performed before identification of RLN and after complete hemostasis of surgical field, and the EMG signals were defined as V 1 and V 2 signals, respectively. This study was approved by the Institutional Review Board of Kaohsiung Medical University Hospital and the ClinicalTrials.gov ( http://www.clinicaltrials.gov . [identifier: NCT00629746 ]). Written informed consent was obtained from each patient. Patients were informed of the intent to use the monitoring system potentially to aid in the localization and identification of the RLNs and assessment of their function during operation.
All patients received preoperative and postoperative video recording of vocal cord movement with flexible laryngofibroscopy, and all exposed RLNs were routinely measured in length and were photographically documented with a high-resolution camera to show visual nerve integrity after complete nerve dissection. When asymmetric cord movement was found postoperatively, a comparison with the preoperative recording would be performed. When vocal dysfunction was identified, follow-up was per 2 weeks initially and every 4 weeks thereafter until recovery was achieved. Dysfunction was considered permanent if it persisted for 6 months subsequent to operation.
Statistical analysis of continuous variables was compared using independent t test and was presented as mean ± SD. Categorical nominal variables were analyzed by χ 2 test or the Fisher exact test. All analyses were performed using SPSS for Windows version 13.0 (SPSS, Inc, Chicago, IL). All statistical tests were 2-sided, and significant level was set at P < .05.
3
Results
Among the 101 RLNs with extensive nerve dissection, 13 nerves were due to the operation for recurrent goiter; 41 nerves, for large goiter with substernal extension; and the other 47 nerves, for thyroid cancer with paratracheal nodes metastasis ( Table 1 ). No permanent palsy occurred, but 2 nerves experienced loss of EMG signal after complete dissection from large recurrent goiter, and they developed temporary palsy. Both nerves regained normal cord function within 6 weeks. The RLN palsy rate in the extensive dissection group was 2% (2/101), which was not significantly different from that in the nonextensive dissection group (2.5%; P = .77) ( Table 1 ). Among 99 nerves with normal vocal function after operation, none experienced a weakened signal after complete RLN dissection, and the mean amplitudes of R 2 and V 2 signals were not significantly different from those of R 1 and V 1 signals (R 2 vs R 1 ; 1038 vs 1030 μ V; P = .74; V 2 vs V 1 ; 824 vs 816 μ V; P = .75) ( Fig. 1 ).
