Navigating Unanticipated Non-recurrent Laryngeal Nerves in Thyroid Surgery: Strategies for Preservation and Anticipation

Abstract

Background/Aim

This study aimed to investigate the incidence, anatomical characteristics, and clinical implications of non-recurrent laryngeal nerves (NRLNs) discovered during thyroid surgeries and autopsies.

Patients and Methods

A total of 2,215 thyroid surgeries and 194 autopsies were reviewed, identifying 17 and 1 case of NRLN, respectively. Data regarding nerve anatomy, associated vascular anomalies, and patient medical history were collected and analyzed. Neck ultrasound examinations were subsequently performed on the 17 living patients, 5 months to 19 years post-surgery, by three radiologists specializing in head and neck soft-tissue imaging. Two radiologists were blinded to the specific nerve anatomy, while one was unblinded.

Results

Vascular anomalies of the aortic arch were described only by the unblinded radiologist. Among the 17 NRLN cases, three (16.7%) exhibited normal vascular anatomy. According to Toniato et al.’s classification, one case each was categorized as 2a, 2b, and one case involved a coexisting right NRLN and right RLN. No definitive recurrent paresis was observed, with only one patient experiencing transient palsy lasting 5 weeks. Histological evaluation revealed no structural differences between NRLN and RLN variants, although their epineurium and adventitia were notably thinner than those of the vagal nerve.

Conclusion

Achieving blood-free conditions for meticulous dissection and preserving sympathetic nerve branches are essential for optimal functional outcomes in thyroid surgeries involving NRLNs.

Introduction

In 1823, Stedman was the first to discover that the laryngeal nerve originated from the cervical vagus rather than looping around the subclavian artery. Several subsequent autopsy findings confirmed this observation (1). Nearly 100 years later, Pemberton and Beaver drew attention to the surgical risks associated with this abnormality after accidentally dissecting a non-recurrent laryngeal nerve (NRLN) that had been mistaken for a branch of the inferior thyroid artery (ITA) (2). The incidence of NRLNs ranges from 0.3-2.0% on the right side and approximately 0.04% on the left. While on the left side this anomaly is exclusively associated with situs viscerum inversus, the situation on the right side is far more complex, as a right-sided NRLN may occur even without an anomalous right subclavian artery (3-5). Injury to the RLN, resulting in vocal cord paresis or paralysis, is one of the most serious complications of thyroid surgery, and the risk is significantly higher in cases involving a non-recurrent variant (6). The high rate of injury is largely due to the difficulty of suspecting this anatomical anomaly preoperatively. Identifying an NRLN before surgery might substantially reduce the risk of nerve damage (7). This raises the question: Is there a method that allows us to detect the presence of an atypical nerve pre- or intraoperatively?

To explore this, we conducted a retrospective review of the records of 2,215 patients who underwent thyroid resection, along with 194 autopsies, and postoperative examinations of 17 patients with previously unsuspected NRLN.

Patients and Methods

From 2002 to 2022, a total of 2,215 thyroid operations were performed at the Department of Surgery at St. George University Teaching Hospital. The mean patient age was 54.1 years, and the male-to-female ratio was 379/1,836 (17.1%/82.9%). Thyroid lobectomy (uni- or bilateral), subtotal resection, and near-total resection were performed in 1,908 (86.1%), 34 (1.5%), and 273 (12.3%) of cases, respectively. Among unilateral thyroid lobectomies, 55.7% (374/672) were right-sided and 44.3% (298/672) were left-sided. We routinely attempted to identify both the superior and RLNs, achieving a 36.7% success rate for the superior laryngeal nerve and a 100% success rate for the RLN.

A total of 17 NRLNs were identified among the 1,908 right-sided unilateral or bilateral thyroid lobectomies (0.89%), and one additional NRLN was found during 194 autopsies (0.52%). Of the 17 surgical cases, four patients underwent near-total resection, 11 underwent total thyroidectomy, and two underwent right-sided lobectomy. The first operation involving an NRLN was performed in 2004, and the most recent case was in 2025.

RLNs were routinely identified in all patients using our standard procedure. After division of the superior thyroid artery, the thyroid lobe was grasped with an Ellis retractor and retracted cranially and medially. Palpation was used to assist in locating the RLN. The nerve was exposed caudally and posterior to the inferior pole of the thyroid gland within the esophago-tracheal groove, following exclusion or ligation of the ITA. When an ultrasonic dissector or bipolar scissors were used, the surgical field was irrigated with physiologic saline at body temperature to prevent lateral thermal spread. Once identified, the RLN was traced to its laryngeal entry point.

According to Toniato et al.’s classification (8), 3 out of the 17 NRLNs originated at the upper pole of the thyroid gland (type 1), 12 coursed parallel above the inferior ITA (type 2a), and 2 passed beneath the ITA trunk or between its branches (type 2b).

Ultrasound examinations were performed using a GE Logiq S7 device (GE HealthCare Hungary, Budapest, Hungary). Each patient was examined independently on the same day by three radiologists specializing in head and neck soft-tissue imaging. All 17 patients’ medical histories were known to the radiologists; however, two radiologists were not informed of the specific nerve anomaly (blind assessment). The radiology report was expected to include a description of the operative region, the size of the residual thyroid gland, cervical lymph nodes and vessels, and the submandibular salivary gland. Routine imaging, however, does not typically include evaluation of the bifurcation of the brachiocephalic artery, a key anatomical feature associated with NRLN. Furthermore, visualization of the brachiocephalic trunk may be limited in patients with obesity or large goiters. Three-dimensional reconstructions of the aortic arch were generated using a GE Revolution 128 device (GE Healthcare Hungary, Budapest, Hungary).

Indirect laryngoscopy was performed in all patients by an ENT specialist, with particular attention to vocal cord mobility.

The study was conducted in accordance with the Declaration of Helsinki (revised 2013). Institutional Review Board approval was obtained (document number: S25C 924020615030). Informed consent was obtained from all patients.

Results

The mean size of the residual thyroid gland following near total resection was 11×6×9±3.3 (SD) mm on the left side and 15×7×8±2.9 (SD) mm on the right side. In one exceptional case, two radiologists did not identify any residual tissue, whereas the third radiologist detected a 5.6×7 mm remnant. Five patients had atherosclerotic plaques in the common carotid artery without significant hemodynamic alteration. No pathological lymph nodes were detected along the major cervical vessels. Three patients had cysts in their submandibular salivary glands, measuring 3×6 mm, 11×6 mm, and 13×6 mm, respectively.

Only the unblinded radiologist described bifurcation of the brachiocephalic artery. Three patients exhibited a typical Y-shaped bifurcation (Figure 1A), whereas in 14 patients, the subclavian artery was slightly separated from the common carotid artery (Figure 1B), a finding confirmed by computed tomographic angiography (Figure 1C and D).

Figure 1.

Imaging diagnostics. A: Ultrasonographic view of normal arterial anatomy, Y sign. acc: Right common carotid artery; trbrch: brachiocephalic artery; ascl: aberrant right subclavian artery; B: Color Doppler image with aberrant right subclavian artery. C: Transversal computed tomographic angiography image. The arrow indicates the origin of the aberrant right subclavian artery. D: Three-dimensional reconstruction with independent course of the aberrant right subclavian artery (arrow).

Among the patients with a typical Y sign on ultrasound, during surgery one was found to have a type 2a (Figure 2A) anatomical variation and another a type 2b (Figure 2B). In the third case, coexisting right NRLN and right RLN were identified. Histological evaluation revealed that the epineurium and adventitia were significantly thinner in the NRLN compared with the vagal nerve (Figure 2C and D).

Figure 2.

Surgical views with classification of non-recurrent laryngeal nerve (NRLN) according to Toniato et al. (8). Single asterisks indicate NRLN, double asterisks denote bifid NRLN. A: Type 2a NRLN accompanying the trunk of the inferior thyroid artery (ITA). B: Type 2b under the ITA trunk. C: The tissue block of the pharyngo-esophageal junction with type 2b. The arrow indicates the transection of the nerves for histological examination. D: Histological section of type 2b NRLN with pharyngo-esophageal junction. E: Type 1 on the upper thyroid/superior pole. F: Bifid NRLN with sympathetic nerve.

Indirect laryngoscopy was performed in all 17 patients between 5 months and 19 years after thyroid lobectomy, and no definitive vocal cord dysfunction was detected. Based on retrospective chart review, only one patient experienced transient vocal cord palsy, which resolved within 5 weeks.

Discussion

Our retrospective study investigated the rare anatomical variation of the inferior laryngeal nerve, known as NRLN, and examined whether routine ultrasound examinations performed by specialists is able to provide indirect indications of its presence. Such preoperative identification might help surgeons take appropriate precautions during thyroid surgery.

We selected ultrasound as our imaging modality because, although several techniques −such as computed tomography, magnetic resonance imaging, barium esophagram, and intraoperative neuromonitoring (9, 10)− have been proposed for detecting NRLN, their routine use is limited due to variable diagnostic accuracy (11), invasiveness, and high cost (12, 13). In contrast, ultrasound is widely accessible, cost-effective, and non-invasive, making it a potentially valuable tool for identifying vascular anomalies associated with NRLN.

Our results confirmed that NRLN is indeed rare, with an incidence of 0.89%, consistent with previous reports (14).

In the blinded assessment, the radiologists did not describe any epi-aortic vessel abnormalities, whereas the unblinded examiner identified slight vascular variations in 14 cases and normal anatomy in 3. These findings suggest that extended neck ultrasound examinations may assist in detecting vascular anomalies associated with NRLN, although they are not definitive (15). Therefore, both surgeons and radiologists should remain aware of epi-aortic vascular abnormalities and the potential presence of NRLN.

Our previous histological investigations revealed considerable variability in the diameter of RLNs and in the thickness of their epineurium. A similar pattern was observed in NRLNs, except that the epineurium and adventitia were notably thinner than those of the vagal nerve. The presence of fragile, blood-filled adventitial tissue, variable epineural structure, and inflammatory changes extending beyond the thyroid capsule to the perineural fascia are significant factors contributing to nerve vulnerability. Surgical manipulation in such an environment may induce inflammatory edema through mediator release or cause diffuse microhemorrhages by damaging delicate capillaries, potentially leading to neurapraxia or axonotmesis (16).

Given these findings, the surgical approach to both RLN and NRLN requires exceptional precision and care. When bipolar scissors or an ultrasonic dissector are used, continuous irrigation with physiological saline at body temperature is essential to minimize thermal injury.

Finally, the answer to the most important emerging question –whether an exclusive method exists for anticipating and preserving the integrity of an unsuspected NRLN– is a definitive no! Only precautionary measures can be implemented, including:

– educating radiologists to include upper mediastinal evaluation during thyroid ultrasound;

– increasing surgeons’ awareness of this anatomical variation;

– carefully dissecting the upper pole of the gland and avoiding en bloc division of the superior thyroid artery, instead, dividing its branches separately to prevent injury to the superior laryngeal nerve and any potential type 1 NRLN (Figure 2E);

– identifying and excluding the ITA and irrigating the surgical field throughout nerve preparation;

– performing meticulous thyroid dissection;

– preserving not only the RLN and NRLN but also any sympathetic nerve branches in this region.

However, even this strategy has limitations, as NRLNs with two separate branches (bifid NRLNs) originating at different anatomical levels may also occur (Figure 2F). Nevertheless, adherence to the above precautions may enable surgeons to significantly reduce the rate of permanent paresis.

Conclusion

The NRLN represents a rare but clinically significant anatomical variation that poses substantial challenges during thyroid surgery. Although routine preoperative imaging techniques, such as neck ultrasound, can provide valuable information, particularly in identifying epi-aortic vascular abnormalities, their diagnostic accuracy is not absolute. Consequently, reliance on imaging alone is insufficient, underscoring the need for meticulous intraoperative dissection and heightened surgical vigilance.

Our findings emphasize that preserving the integrity of both RLNs and NRLNs is best achieved through a combination of careful operative strategies.

Conflicts of Interest

The Authors declare no conflicts of interest concerning the publication of this article.

Authors’ Contributions

ZF wrote the article. RJ, VR, MSZ and ZSP collected the data and participated in the operations. AÁG edited the manuscript and collected the data. EK carried out radiological examinations. MR performed the otorhinolaryngological tests. Pathological analysis was completed by ASZ. AV and ÁA reviewed and edited the article.

Funding

This research did not receive any specific grant from funding agencies in the public, commercial, or not-for-profit sectors.

Artificial Intelligence (AI) Disclosure

No artificial intelligence (AI) tools, including large language models or machine-learning software, were used in the preparation, analysis, or presentation of this manuscript.