Routine identification of the recurrent laryngeal nerve (RLN) has been accepted as the gold standard of care during thyroid surgery to decrease RLN palsy rates (1–4). During the RLN identification and dissection phase, the RLN can be injured by the surgical maneuvers of traction, clamping, electrocauterization, mechanical trauma, and transection. However, traction injury is the most common cause of RLN injury, with a rate of up to 70-80% among all maneuvers (5–8). Medial traction of the thyroid lobe is a necessary step in the surgical procedure to identify RLN. The RLN can be overstretched and injured by a dense fibrous band or surrounding blood vessels at the region of Berry’s ligament during medial thyroid traction. Loss of signal (LOS) was found after nerve dissection with the application of intraoperative neuromonitoring (IONM) in thyroid surgery, even though visual anatomical integrity of the RLN was confirmed intraoperatively (8–10).

For stimulating methods, intermittent IONM (I-IONM) applied with a hand-held stimulating probe is still mainstream internationally due to its convenience and inexpensiveness. However, it is challenging to detect early between the intervals of nerve stimulation using the conventional I-IONM, the adverse electromyography (EMG) change, and LOS is consistently recognized after the occurrence of traction injury (5–8). Animal and clinical studies of RLN traction injury with the application of continuous IONM (C-IONM) (11–14) revealed an EMG amplitude decrease during RLN traction and progressive recovery after releasing traction stress. C-IONM is reported to be helpful for the early detection of EMG amplitude decreases caused by traction distress (15–17). Stimulating dissecting instrument (SDI) is a novel and real-time I-IONM stimulating method during surgical procedure. It can maximally reduce the intervals of the nerve stimulation while maintaining the advantages of I-IONM convenience, simplicity and affordability. For recording methods, unstable and variable EMG amplitudes during operation have been reported using traditional EMG tubes for EMG recoding (18, 19). It will be challenging to differentiate an actual decrease in EMG signal caused by traction stress on the RLN from a false decrease caused by EMG tube displacement. The trans-thyroid cartilage EMG recording method can avoid the electrode malpositioning issue and obtain high-quality EMG signals.

IONM facilitates early detection of adverse EMG effect, and the corrective maneuver can reduce severe and repeated nerve injury. However, intraoperative patterns and outcomes of EMG recovery are less studied in the literature. In this study, SDI and the trans-thyroid cartilage EMG recording method were used for nerve stimulation and signal recording. With these two new IONM techniques, we aimed to determine the feasibility of early detection of a significant EMG amplitude decrease during the phase of RLN dissection and the possibility of intraoperative EMG recovery after releasing thyroid traction, which might be helpful for avoiding signal deterioration to complete LOS and postoperative vocal cord (VC) palsy.

Forty-four of 1142 nerves (3.9%) exhibited a significant EMG amplitude decrease (>50% decrease in R₁ signal) during RLN dissection, and all (100%) showed gradual progressive amplitude recovery within a few minutes after releasing thyroid traction (Figure 3). Ten nerves recovered from LOS, and another 34 nerves recovered from a 51-90% amplitude decrease.

After restarting the surgical maneuver, a secondary significant EMG amplitude decrease occurred in 6 (13.6%) of 44 nerves. A substantial R_2p/R_2d ratio reduction (81% and 92%) in 2 nerves and unrecovered LOS in 4 nerves occurred due to inevitable repeated traction, and all 6 nerves developed temporary VC palsy.

Three types of EMG recovery patterns were summarized. Type A recovery pattern showed complete EMG amplitude recovery in 14 (32%) nerves; Type B showed incomplete recovery with an injury point detected on the RLN in 16 (36%) nerves. Type C showed incomplete recovery without an injury point detected in 14 (32%) nerves (Figure 1). Regarding Type A recovery pattern, 2 of 14 nerves had LOS during the surgical maneuver. No (0%) patient had postoperative weak or fixed VC mobility (Table 1). Of the nerves with the Type B recovery pattern, 6 of 16 nerves had LOS during the surgical maneuver. Seven (44%) patients had abnormal postoperative VC mobility, including 2 patients had weak VC mobility and 5 patients had fixed VC mobility. All 16 nerve injury points were located in the upper portion of exposed RLN (at the region of Berry’s ligament or near the laryngeal entry point) (Table 2). Of the nerves with Type C recovery pattern, 2 of 14 nerves had LOS during the surgical maneuver. Two (14%) patients had abnormal postoperative VC mobility, including one patient had weak VC mobility, and one patient had fixed VC mobility (Table 3).

In this study, a substantial EMG amplitude decrease was also found in 9 nerves caused by dissection trauma or thermal injury but without showing the feature of progressive recovery. Complete LOS was found in 5 nerves, and incomplete LOS was found in 4 nerves. Temporary VC palsy occurred in 7 nerves, and permanent VC palsy occurred in 2 nerve with thermal injury.

RLN traction injury frequently occurs at the region of Berry’s ligament since the anatomical relationship between RLN, Berry’s ligament and surrounding vessels is highly variable (21–24) The RLN is located not only posterolateral to Berry’s ligament but also posteromedial to it, and the anterior motor branch of the RLN may penetrate the ligament. Additionally, blood vessels often intertwine with the RLN at the region of Berry’s ligament. The RLN can be overstretched during medial thyroid traction by Berry’s ligament or an intertwined vessel (Figures 4A–C).

When applying C-IONM in thyroid surgery, a progressive decrease in EMG amplitude is the characteristic feature of traction distress on the RLN (11–14). Prolonged traction may result in unrecovered LOS and postoperative VC palsy. Therefore, the weakness of conventional I-IONM compared to C-IONM is that early detection of imminent RLN injury and relief of traction distress are less effective to avoid signals deteriorating to an unrecoverable level (13). In this study, SDI was used for nerve stimulation and tissue dissection. It is a simple and effective way to real-time monitor nerve function during the highest risk phase for RLN injury (25, 26). In our experience, SDI is not inferior to C-IONM in early detection of nerve injury, furthermore, C-IONM cannot replace I-IONM in locating the nerve injury point. Additionally, we chose the trans-thyroid cartilage recording method for EMG signal recording. It provides high initial EMG amplitudes during VN and RLN stimulation, and the amplitudes remain stable during the entire course of the operation (27–29). This technique solves the major limitation of the EMG tube recording method. A significant EMG amplitude decrease can also occur due to EMG tube displacement during surgical manipulation of the thyroid lobe or trachea. Using an EMG tube for signal recording make it difficult to tell an actual EMG amplitude decrease from a false one. In this study, applying these two novel IONM techniques, we detected 44 nerves with a significant EMG amplitude decrease during RLN dissection and gradual progressive amplitude recovery after releasing thyroid traction. The EMG signal recovers from LOS in 10 nerves and a 51-90% amplitude decrease in 34 nerves. The results suggest a high chance of signal recovery after a substantial amplitude decrease, even after LOS, if traction strain on the RLN was relieved in time. In cases with large tumors or complicated anatomical relationships between the RLN and Berry’s ligament, a secondary EMG amplitude decrease still occurred after restarting the surgical maneuver with gentle thyroid traction and careful nerve dissection due to unavoidable repeated traction. A persistent substantial amplitude decrease was found in 6 nerves after finishing RLN dissection (2 nerves with 81% and 92% R_2p/R_2d ratio reduction; 4 nerves with unrecovered LOS). All 6 nerves developed temporary VC palsy. The results indicate that the RLN is highly vulnerable to repetitive stretching.

From the results of the five EMG signals (V₁-R₁-R_2p-R_2d-V₂) obtained from VN and RLN stimulation in this study, we found 3 types of EMG recovery patterns. Type A recovery showed complete EMG amplitude recovery, and the amplitudes of the R_2p, R_2d and R₁ signals were similar (Table 1). Type B recovery showed incomplete EMG recovery when comparing R_2p with R₁ signals, and it was combined with an injury point on the RLN and a significant reduction in the R_2p/R_2d ratio (Table 2). Type C recovery showed incomplete EMG recovery, but without an injury point detected on the RLN (Table 3). In Type C recovery pattern, we can see that the amplitudes of the R_2p and R_2d signals are similar, but the amplitudes show a significant decrease compared with the R₁ signal (Figure 3). Case No. 2 in Table 3 did not show a reduction of the R_2p/R_2d ratio. Still, it had a 66% amplitude decrease compared to the R_2p and R₁ signals (509 vs. 1515 µV), and this case developed weak VC mobility postoperatively. This kind of nerve injury was unrecognized when using the EMG tube recording method since the amplitude change would be mistaken for EMG tube displacement. Overall, different recovery patterns have different vocal cord function outcomes, to elucidate the recovery patterns can assist surgeons in the intraoperative decision making and postoperative management (i.e. speech therapy and injection laryngoplasty).

Regarding LOS after traction injury, Chiang (7) reported that nerve injuries caused by overstretching of Berry’s ligament could be divided into type 1 and type 2 stretch injuries. A type 1 stretch injury was caused by direct distress on the nerve and featured an injury point on the nerve. A type 2 stretch injury could be caused by pulling down the distal part of the RLN when the RLN was stretched excessively at the region of Berry’s ligament, and the nerve might be injured at a higher position above the laryngeal entry point. From the evidence of type B and type C EMG recovery patterns, we find that type 1 or type 2 nerve injury occurs not only in LOS (7, 9, 30) but also in incomplete LOS. In this study, there were 3 nerves with type 1 LOS and 1 nerve with type 2 LOS. In addition, there were 13 nerves with a type 1 incomplete LOS with a weak point of nerve conduction detected on the RLN (case No. 4-16 in Table 2) and 13 nerves with a type 2 incomplete LOS with no weak point detected (case No. 2-14 in Table 3).

In our clinical experience, some patients with very late recovery of muscle tone under neuromuscular blockade agent and the accuracy of the amplitudes of V1 and R1 may be questionable. In this study, the time from the administration of rocuronium (0.3 mg/kg), neck positioning, skin preparation to skin incision is around 20 minutes. And the time from skin incision, skin flap elevation, pyramidal dissection, superior thyroid pole dissection to V1 and R1 signals takes another 30 minutes, which is even longer. From the results of our patients without nerve injury, the data of V1-R1-R2p-R2d-V2 showed stable. It indicates that neuromuscular blockade has been almost completely recovered through 50 minutes. Therefore, the use of R1 signal as basic reference baseline should be reliable, although more studies are necessary to prove it.

Several limitations of this study were noted. First, this was an observational study and lacked a control group. Second, the comparisons between conventional I-IONM, SDI, and C-IONM were not studied. Furthermore, the combined SDI and C-IONM technique is an optimal procedure (least missed EMG changes) with distinct advantages that merit evaluation by a specific design study. Third, the recovery time might differ between type A, B, and C EMG recovery patterns in this study. When the nerve recovers quickly, the surgeon tends to wait and observe until the signal reaches a stable amplitude; when the nerve recovers slowly, the surgeon tends to perform surgical dissection of the contralateral or other regions first. Therefore, no exact recovery time can be provided in the current study, but a gentle thyroid traction and nerve dissection is strongly recommended after adverse EMG event. Fourth, the patient number of each recovery type is relatively small; however, the characteristics of each recovery type can be clearly shown in this study. Further large-scale studies in multiple centers are helpful for the analysis of the clinical presentation in the different recovery types.

A significant EMG amplitude decrease and progressive recovery after releasing thyroid traction is the characteristic feature of traction strain on the RLN during thyroid surgery. In this study, early detection and release of thyroid traction results in intraoperative EMG recovery and helps to avoid a signal deterioration to an unrecoverable LOS. Gentle traction of the thyroid lobe is recommended during the entire course of the operation, and frequent stimulation of the RLN is necessary for the early detection of EMG amplitude changes, particularly when using a stimulating probe for nerve stimulation. Overall, different recovery patterns result in different vocal cord function outcomes which elucidate that the recovery patterns can assist surgeons in the intraoperative decision making and postoperative management.

The original contributions presented in the study are included in the article. Further inquiries can be directed to the corresponding authors.

Ethical approval of this study was obtained from the Kaohsiung Medical University Hospital Institutional Review Board (KMUHIRB-E(I)-20210358). Written informed consent for participation was not required for this study in accordance with the national legislation and the institutional requirements.

Supervision – T-ZH, W-HY, C-WW, GD, and F-YC; Materials – C-FL, T-ZH, Y-CS, T-YH, and F-YC; Data Collection and Processing – K-LC, ChihCW, ChieCW , and Y-CS; Analysis and Interpretation – K-LC, C-FL, W-HVY, T-YH, and F-YC; Literature Search – K-LC, ChihCW, ChieCW , T-YH, and F-YC; Writing Manuscript – All authors. All authors have read and agreed to the published version of the manuscript.

This study was supported by grants from Kaohsiung Medical University Hospital, Kaohsiung Medical University (KMUH110-0R51), and Ministry of Science and Technology (MOST 110-2314-B-037-104-MY2, MOST 110-2314-B-037-120, MOST 111-2628-B-037-007), Taiwan.

The authors would like to thank all the patients included in this study.

The authors declare that the research was conducted in the absence of any commercial or financial relationships that could be construed as a potential conflict of interest.

All claims expressed in this article are solely those of the authors and do not necessarily represent those of their affiliated organizations, or those of the publisher, the editors and the reviewers. Any product that may be evaluated in this article, or claim that may be made by its manufacturer, is not guaranteed or endorsed by the publisher.