Thyroid cancer, the most common endocrine-related cancer, has been increasing rapidly in recent years and is particularly prevalent in young women (1). Conventional open thyroidectomy (COT) is the standard surgical treatment for thyroid cancer; however, the conspicuous scarring and the associated emotional burden pose significant challenges, especially for young female patients. Endoscopic thyroidectomies were first performed in 1997 (2). This minimally invasive surgery not only provides better cosmetic results, but also contribute to reduces surgical complications. The introduction of endoscopic thyroidectomy has attracted the interest of a wide range of surgeons. Over the past 20 years, lumpectomies have developed rapidly. The most common surgical approaches are the gasless unilateral transaxillary approach (GUA), bilateral axillo-breast approach (BABA), axillo-bilateral breast approach (ABBA), unilateral axillo-breast approach (UABA), chest-breast approach (CBA), anterior chest approach (ACA) and transoral approach (OA).

It is important to acknowledge that endoscopic surgeries are not uniformly “minimally invasive” and do not universally reduce invasiveness. While they offer advantages in terms of cosmesis and surgical complication reduction, these advantages may not be consistently realized in every case. Endoscopic surgeries, while typically associated with reduced scarring and fewer complications, can, in certain circumstances, present unique challenges. For example, the use of unconventional entry sites can increase the risk of injury to critical nerves, such as brachial plexus injury with the axillary approach (3), or central nerve palsy with the transoral approach can lead to central nerve palsy (4). In light of these nuances and the potential variability in outcomes, a comprehensive evaluation of the efficacy and safety of endoscopic thyroid surgery is imperative. Previous meta-analyses have often compared different surgical approaches solely with COT, leaving critical gaps in our understanding. To address these limitations and provide a more nuanced assessment, we employ network meta-analysis (NMA). This approach integrates a wealth of clinical experiences, allowing us to offer a more balanced perspective on the advantages and disadvantages of endoscopic thyroid surgery across various surgical techniques. Network meta-analysis can showcase direct and indirect comparison results (5). Through NMA, we aim to gain a more comprehensive understanding of when and how endoscopic procedures can truly be minimally invasive and when they might not offer significant advantages over conventional approaches.

In summary, this network meta-analysis seeks to provide a comprehensive evaluation of endoscopic thyroid surgery, considering the variable nature of “minimally invasive” outcomes. By doing so, we aim to offer a more detailed perspective on the true merits and limitations of endoscopic surgery in the treatment of thyroid cancer.

Our systematic review was designed based on the Preferred Reporting Items for Systematic Reviews and Meta-Analyses(PROSPERO) guidelines (6) and registered with PROSPERO (CRD42023430413). Ethical approval or registration was not required as this study involved an analysis of existing published data.

This meta-analysis found that endoscopic surgery not only has a lower incidence of complications than COT but may even be superior to COT in terms of operative bleeding volume, hospital days, and retrieved LNs. Previous meta-analyses have explored the efficacy of endoscopic thyroidectomy for thyroid neoplasms, but only compared different endoscopic surgeries or conventional surgeries (51, 52). Our study adopted an NMA approach to comprehensively and meticulously discuss the advantages of seven surgical approaches. Among the various surgical approaches, ABBA stands out for its notably shorter operative time and proximity to COT. ACA excels in effectively controlling intraoperative bleeding. UABA demonstrates superior outcomes in terms of hospital stay and drainage, while OA prove to be more conducive to straightforward lymph node acquisition. The patients included in the study were all patients with thyroid cancer, and studies only focused on different endoscopic surgeries, which greatly reduced research heterogeneity and made our results more reliable.

COT required the least operative time, which is consistent with previous research results (15, 16, 22). This may be because of the need for additional surgical channels and workspaces for endoscopic surgery. For endoscopic surgery, the ABBA and UABA require shorter operative times. It needs to be emphasized that OA serves as the “shortest surgical pathway” for endoscopic surgery, but the results show that it has the longest operative time. The main reason for this phenomenon may be: 1) To improve safety and feasibility, OA usually involves three oral vestibular incisions (53), and during the surgical process, the three trocars often encounter unavoidable collisions in the limited workspace, subsequently increasing the difficulty of the surgery; 2) OA is a top-down operation. Owing to the presence of the thyroid cartilage, treatment of the superior thyroid requires more time than other methods. Many of the operation time data included in this study did not specify whether it was a total thyroidectomy or lobectomy, and roughly categorizing them into the same category would, to some extent, have affected our conclusion. ACA and UABA show obvious advantages in terms of intraoperative bleeding volume and hospital stay, while OA and GUA often have poor control of intraoperative bleeding volume due to the need for more extensive free platysma muscle to create surgical space. In the transaxillary approach, the UABA is relatively better than the GUA in bleeding control; however, because it does not require CO₂ insufflation, it avoids the occurrence of mediastinal emphysema, subcutaneous emphysema, and other related gas complications.

Thyroid cancer has a high rate of central lymph node metastasis (54), which is associated with high patient mortality (55). This study reveals that ACA outperforms COT in terms of retrieving LNs; however, it’s important to note that the limited inclusion of only two articles involving ACA may introduce potential bias into the interpretation of this result. OA exhibits superior results in the retrieval of both total LNs and metastatic LNs compared to other endoscopic procedures. This can be attributed to the fact that, from top to bottom, the field of view in OA can seamlessly encompass the central region, thereby circumventing sternum and clavicle limitations, which uniquely facilitates the removal of central LNs.

TG is an important indicator for the postoperative follow-up of thyroid cancer, and lower stimulated TG levels are considered to be associated with total thyroidectomy and low recurrence rate (56). We found no difference between the endoscopic surgery and COT groups for TG, with CBA being significantly lower than COT. Regarding postoperative pain, CBA was not significantly better than COT. The reasons for this may be as follows: (1) fewer articles were included and (2) inconsistent measurement time points after surgery. We found that many studies used different methods to measure postoperative pain, which made it difficult to merge the results of the studies. However, this meta-analysis included studies that used VAS scores. It is well known that endoscopic thyroid surgery avoids scarring on the neck, and our results fully demonstrate this advantage. However, we also encountered a difficulty, there is no common scoring criterion for objective evaluation, and it is necessary for us to find a more objective standard in future research. The incidence of postoperative complications was not lower in the endoscopic surgery group than that in the COT group. Each surgical approach has its own drawbacks; for example, OA is prone to incision infection (8), CBA causes subcutaneous effusion (57), and GUA causes brachial plexus nerve injury (58).

Our NMA had some limitations. First, most of the included articles were retrospective studies, and further large-scale randomized controlled trials are needed. Second, the articles we included were all from Asia, which may limit the generalisability of our results. Third, the scope of surgery included in the article was not specified, and there was a certain degree of heterogeneity.

Based on comprehensive NMA, endoscopic thyroidectomy is safe and feasible. Endoscopic surgical approaches have unique advantages. ABBA and UABA are the closest to COT in terms of operative time and are superior to other endoscopic surgeries. The UABA and ACA are superior to other surgical approaches in controlling intraoperative bleeding and hospital stay. UABA and BABA showed significant advantages in terms of postoperative drainage volume. The OA is more adept at retrieving lymph nodes. Endoscopic thyroidectomy did not increase the incidence of postoperative complications. However, it is clear that careful selection and consideration of the surgical approach are essential. In light of these findings, it is apparent that prospective, randomized studies are needed to validate and further refine the findings of our NMA, providing a more robust foundation for surgeons to make informed decisions when selecting appropriate endoscopic surgical approaches.

XL: Data curation, Writing – original draft. HZ: Writing – review & editing. WD: Data curation, Writing – original draft.