Risk factors and nomogram for predicting central lymph node metastasis in papillary thyroid carcinoma based on gasless transaxillary endoscopic thyroidectomy

Introduction

Papillary thyroid carcinoma (PTC) accounts for approximately 80% of all thyroid malignancies and is characterized by its indolent nature and favorable prognosis (1,2). Accurate assessment of lymph node status is crucial for optimizing surgical planning and improving patient outcomes. With the rapid advancement of minimally invasive techniques, endoscopic thyroidectomy is increasingly utilized as an alternative to conventional open surgery (3). Endoscopic thyroidectomy is generally non-inferior to open surgery regarding the efficacy of central lymph node dissection and short-term oncological safety, while significantly enhancing cosmetic outcomes and quality of life (4,5). Among these approaches, gasless transaxillary endoscopic thyroidectomy (GTET) demonstrates excellent performance in terms of complication rates, short-term oncological completeness, and recurrence rates. It represents an attractive surgical option, particularly for patients with unilateral PTC who prioritize superior aesthetic results (6).

Effectively predicting the risk of central lymph node metastasis (CLNM) is essential for selecting the appropriate surgical approach and tailoring postoperative surveillance. Previous studies have established predictive models for CLNM in the context of conventional open surgery (2,7). However, GTET features a unique surgical perspective and operational trajectory. The evaluation of the anatomical relationship between the tumor and adjacent lymph nodes differs from that of open surgery, potentially impacting the accuracy of CLNM risk assessments (6). Currently, there is a distinct lack of predictive models specifically tailored for CLNM in patients undergoing GTET. Consequently, this study aims to construct and validate a CLNM risk prediction model based on the specific clinical characteristics of patients with clinical T1–T2 and cN0 or cNx unilateral PTC undergoing this procedure. This model is intended to provide evidence-based guidance for tailoring individualized surgical strategies, adjusting postoperative follow-up intensity, and managing risk stratification, ultimately promoting the continued development and application of the GTET approach. We present this article in accordance with the TRIPOD reporting checklist (available at https://gs.amegroups.com/article/view/10.21037/gs-2026-0228/rc).

Methods

Study design and population

This study was conducted in accordance with the Declaration of Helsinki and its subsequent amendments. The study protocol was approved by the Institutional Ethics Committee of The Third Hospital of Mianyang (No. 2026S12-1), which waived informed consent from enrolled patients because of the retrospective nature of this study. This single-center, retrospective cohort study enrolled patients who underwent GTET with central lymph node dissection (CLND, levels VI and VII) at The Third Hospital of Mianyang between January 2022 and January 2026, and who were postoperatively confirmed to have PTC via pathological examination. The collected variables included: (I) categorical variables: sex, tumor laterality, multifocality, and the presence of concurrent Hashimoto’s thyroiditis, hyperthyroidism, or subacute thyroiditis; (II) continuous variables: age, tumor diameter, total number of dissected central lymph nodes, and number of metastatic lymph nodes; serological markers within 48 hours preoperatively: free triiodothyronine (FT3), free thyroxine (FT4), thyroid-stimulating hormone (TSH), thyroglobulin (TG), and parathyroid hormone (PTH); serological markers within 48 hours postoperatively: TG, PTH, calcium (Ca), and phosphorus (P). Based on postoperative pathological confirmation of central lymph node status, patients were categorized into two cohorts: the lymph node negative (LNN) group, defined as having no metastatic lymph nodes detected in the central compartment, and the lymph node positive (LNP) group, defined as having at least one metastatic lymph node detected in the central compartment.

Inclusion and exclusion criteria

The inclusion criteria were: (I) preoperative fine-needle aspiration (FNA) cytology diagnosed PTC or highly suspicious for malignancy, with postoperative pathology confirming PTC; (II) standard GTET procedure, performing systematic and consistent unilateral CLND (including prelaryngeal, pretracheal, and paratracheal lymph nodes); (III) clinical T1–T2 stage (maximum tumor diameter ≤4.0 cm) with no evidence of suspicious extrathyroidal extension on preoperative cervical ultrasound or computed tomography (CT); (IV) preoperative clinical N0 or Nx status (cN0/cNx), confirmed by preoperative imaging showing no evidence of regional lymph node metastasis; (V) age ≥18 years. The exclusion criteria were: (I) missing key data fields, incomplete medical records, or unavailable pathological reports; (II) history of prior thyroidectomy, neck radiotherapy, or other interventions that significantly alter local anatomy; (III) preoperative or intraoperative evidence of distant metastasis or lateral cervical lymph node metastasis (levels II–V); (IV) concurrent malignancies or severe systemic diseases.

Statistical analysis

Statistical analyses were performed using SPSS 27.0 and R 4.5.3. The Kolmogorov-Smirnov test was utilized to evaluate the normality of continuous variables. Normally distributed data are expressed as mean ± standard deviation, and intergroup comparisons were conducted using the independent samples t-test. Non-normally distributed data are presented as medians with interquartile ranges [M (Q1, Q3)]. Categorical variables are summarized as frequencies and percentages [n (%)], and comparisons between groups were performed using the Chi-squared test. Least absolute shrinkage and selection operator (LASSO) regression with 10-fold cross-validation was employed to determine the optimal λ value for feature selection. A nomogram for predicting lymph node metastasis risk was subsequently constructed based on independent predictors identified via multivariate logistic regression analysis. Model discrimination, calibration, and clinical utility were evaluated using the receiver operating characteristic (ROC) curves, the concordance index (C-index), bootstrap calibration curves, and decision curve analysis (DCA). A two-sided P<0.05 was considered statistically significant. The sample size of this study met the criteria for each predictor variable of the 10 outcome events.

Results

Analysis of risk factors for CLNM in PTC patients undergoing GTET

Initially, 384 patients were screened for eligibility. Seven patients were excluded: two due to intraoperative conversion to open surgery and five due to a history of previous thyroid surgery. Ultimately, 377 patients were included in the final analysis, comprising 260 patients (68.97%) in the LNN group and 117 patients (31.03%) in the LNP group. Their baseline clinical characteristics are summarized in Table 1. Three cases (0.80%) required conversion to open surgery, and two cases (0.53%) had recurrent laryngeal nerve injury. Postoperative transient hypoparathyroidism occurred in 10 patients (2.65%), and temporary vocal cord paralysis was observed in 32 patients (8.48%).

LASSO regression analysis revealed that age, operative time, tumor diameter, number of dissected lymph nodes, preoperative FT4, and preoperative FT3 levels were associated with CLNM in these patients (Figure 1). Subsequent multivariate logistic regression analysis demonstrated that age, operative time, tumor diameter, number of dissected lymph nodes, and preoperative FT4 were independent risk factors for CLNM in PTC patients undergoing GTET. Preoperative FT3 (P=0.08) did not reach independent statistical significance. These differences were statistically significant (P<0.05) (Table 2).

Figure 1 Regularization path and cross-validation error for LASSO logistic regression model. (A) Standardized coefficient paths across the −log (λ) sequence. (B) Binomial deviance (misclassification error) with standard error bars plotted against −log (λ). The left dotted vertical line indicates the optimal λ (λ_1se=0.035675) selected by the one-standard-error rule, yielding a parsimonious model with six variables; the right dotted line marks λ_min (0.002637). Variables retained in the final model included tumor diameter, number of dissected lymph nodes, FT3, FT4, age, and operative duration. FT3, free triiodothyronine; FT4, free thyroxine; LASSO, least absolute shrinkage and selection operator.

Construction and validation of the prediction model for lymph node metastasis

A nomogram model predicting the probability of CLNM in PTC patients undergoing GTET was developed based on the five identified independent predictors (Figure 2). Each predictor corresponds to a specific point value on the nomogram; the sum of these points maps to a total score, which reflects the estimated probability of CLNM for an individual patient. ROC curve analysis demonstrated a C-index of 0.797 [95% confidence interval (CI): 0.758–0.847] for the model (Figure 3). Following 1,000 bootstrap resampling validations, the C-index remained robust at 0.787 (95% CI: 0.740–0.834). The calibration curve indicated excellent concordance between the predicted probabilities and actual observed incidences (Figure 4). Furthermore, the DCA curve showed that the model provides a substantial clinical net benefit across a broad threshold probability range of 0.05 to 0.8 (Figure 5).

Figure 2 Nomogram for predicting CLNM in patients with PTC based on GTET using five predictive factors. Adding up the total scores of each predictive factor, finding the point corresponding to the horizontal axis of the score, and drawing a perpendicular line will yield the corresponding probability of CLNM. Surg_time: surgical operative time; tumor_diam: tumor diameter; CLN_dissected: number of central lymph nodes dissected; Pre_FT4: preoperative free thyroxine. CLNM, central lymph node metastasis; GTET, gasless transaxillary endoscopic thyroidectomy; PTC, papillary thyroid carcinoma.

Figure 3 ROC curve analysis demonstrated a C-index of 0.797 (95% CI: 0.758–0.847) for the model. AUC, area under the curve; C-index, concordance index; CI, confidence interval; ROC, receiver operating characteristic.

Figure 4 Bootstrap calibration curve after 1,000 bootstrap resamples. Apparent: apparent calibration curve, the curve fitted directly from the original dataset used for modeling without any correction. Bias-corrected: bias-corrected calibration curve, the core curve after 1,000 bootstrap resampling corrections. The closer this curve is to the ideal line, the higher the model’s predictive accuracy and generalization ability. Ideal: ideal calibration curve, the curve when the predicted CLNM probability by the model is exactly equal to the actual occurrence probability of the patients. All actual curves are compared against this line as the gold standard. The smaller the deviation, the better the model performance; C.L.: confidence limit, the gray shaded area around the bias-corrected curve, representing the uncertainty range of the model’s prediction results. The narrower the confidence band, the more stable and reliable the model’s prediction results are. CLNM, central lymph node metastasis; PTC, papillary thyroid carcinoma.

Figure 5 DCA of the nomogram. The red curve depicts the net benefit assuming all patients undergo CLND; the green curve reflects the net benefit assuming no patients undergo CLND; and the blue curve shows the net benefit achieved by applying our prediction model to guide individualized treatment decisions. As indicated by the decision curve, across a clinically relevant threshold probability range of 5% to 80%, the prediction model conferred greater net benefit than either the treat-all or treat-none strategies. CLND, central lymph node dissection; DCA, decision curve analysis.

Discussion

Our clinical prediction model indicates that age, operative time, total number of dissected lymph nodes, and preoperative FT4 levels are independent influencing factors for lymph node metastasis. Specifically, advanced age acts as a protective factor, while prolonged operative time, a greater number of dissected lymph nodes, larger tumor diameter, and elevated preoperative FT4 serve as risk factors. Although postoperative parameters (such as PTH, serum calcium, and serum phosphorus) were collected, they were intentionally excluded from the predictive model. First, the primary objective of the model is perioperative risk assessment; incorporating postoperative variables would fundamentally compromise its preoperative and intraoperative clinical utility. Second, postoperative metrics primarily characterize the patient’s perioperative physiological fluctuations and serve to interpret research findings, indirectly substantiating the biological plausibility of the model’s selected variables.

While lymphatic metastasis is prevalent in PTC, there remains a notable deficiency in effective and precise methods for evaluating lymph node status both preoperatively and intraoperatively (8,9). During transaxillary endoscopic procedures, the restricted operating workspace significantly complicates CLND, particularly regarding the clearance of paratracheal lymph nodes and those situated deep to the right recurrent laryngeal nerve (10,11). Surgeons are required to meticulously identify and preserve the recurrent laryngeal nerve, parathyroid glands, and their respective vascular supplies within a confined, two-dimensional surgical field while simultaneously ensuring comprehensive lymph node clearance (10,12). Compared to conventional open surgery, this technique is associated with a longer operative time, an increased risk of conversion to an open approach, and a potentially higher incidence of transient postoperative hypoparathyroidism and recurrent laryngeal nerve injury (6). Some GTET studies have considered reducing the extent of lymph node dissection, but this may increase the risk of recurrence due to residual metastatic lymph node disease (13). The CLNM prediction model developed in this study, based on clinical T1–T2 stage, cN0 or cNx unilateral papillary thyroid cancer patients undergoing surgery, demonstrates strong discriminatory ability, outperforming single-variable predictions and significantly enhancing the risk assessment paradigm for CLNM.

The female-to-male ratio in our study cohort was approximately 4.39:1, slightly exceeding the epidemiological proportions commonly reported for PTC (14). This disparity may be attributed to female patients placing a higher priority on cervical aesthetics, making them more predisposed to selecting endoscopic surgery to avoid visible anterior neck scarring (15,16). Although female patients constitute the vast majority of the cohort, current literature indicates that male PTC patients inherently carry a higher risk of CLNM and exhibit greater tumor aggressiveness (13). Consistent with this, the incidence of CLNM in our study was 35.7% (25/70) in males compared to 30.0% (92/307) in females. Given their intrinsically elevated risk profile, male patients necessitate precise risk stratification. When deemed necessary, surgeons should proactively consider prophylactic or therapeutic CLND, or potentially alter the surgical approach, to effectively mitigate the risk of local recurrence.

Endoscopic surgery is particularly appealing to younger demographics as it maximizes cosmetic outcomes without compromising oncological safety. Existing research indicates that while younger patients with thyroid cancer (particularly those under 45 years of age) generally experience a more favorable overall prognosis, their incidence of lymph node metastasis is markedly higher compared to middle-aged and elderly patients (17,18). The risk of CLNM exhibits an inverse relationship with age; this trend is largely attributed to the more aggressive tumor biology, higher occult lesion rates, and increased challenges in preoperative staging typical of younger patient cohorts (19). Age represents a pivotal variable that must be comprehensively analyzed in conjunction with other clinical factors to holistically evaluate the patient and facilitate optimal clinical decision-making.

In this analysis, FT3 was selected during the LASSO regression phase but failed to achieve independent statistical significance in the multivariate model. Previous investigations have highlighted that FT4 is more robustly correlated with the severity of lymph node metastasis, whereas FT3 is more commonly associated with the mere presence or absence of metastasis (20). When other predictors are included, FT3’s effect is diluted or confounded in multivariable models, whereas FT4 retains independent predictive value (20,21). Consequently, while FT3 may appear more sensitive during univariate evaluations, FT4 likely provides a more stable and reliable predictive value for CLNM (22). GTET patients are mostly diagnosed with T1/T2 stage thyroid cancer, and their thyroid function is generally normal at diagnosis, without the use of thyroid-related medications, allowing serum FT3 and FT4 levels to accurately reflect endogenous thyroid hormone homeostasis. Preoperative FT4 is a non-invasive and easily accessible serum biomarker. Incorporating preoperative FT4 levels into a multivariate risk prediction model can more accurately identify patients at high risk for CLNM.

Operative time and the total number of dissected lymph nodes emerged as independent risk factors in our study. These variables are rarely highlighted in the context of traditional open surgery, underscoring the unique influence of surgical complexity on metastasis risk intrinsic to the GTET approach. In endoscopic settings, a prolonged operative duration frequently serves as a proxy for increased procedural complexity, suggesting a more extensive range of tumor invasion or a higher locoregional metastatic burden (10). Similarly, the intraoperative lymph node yield not only reflects an expanded surgical scope but also inherently quantifies the technical difficulty of performing CLND via GTET (6,23). Given the spatial constraints of the endoscopic view and the unique operational angles required, harvesting a larger quantity of lymph nodes invariably demands more intricate anatomical dissection and broader tissue exposure. This requirement intrinsically indicates the complex nature of the lesion and a high-risk metastatic profile (3,12).

Tumor size is defined by the maximum diameter, irrespective of the total number of lesions in multifocal disease (24). Extensive research confirms that a tumor diameter ≥1 cm is a universally recognized, potent predictor of CLNM. Furthermore, even in cases of papillary thyroid microcarcinoma (PTMC), a diameter exceeding 0.5 cm significantly amplifies the risk of occult metastasis (25,26). Because transaxillary endoscopic surgery involves a relatively restricted operative workspace, larger tumors can exponentially increase the difficulty of anatomical dissection, prolong the operation, and indirectly heighten the clinical imperative for thorough lymph node clearance (25,27). Tumor diameter is a readily accessible metric routinely obtained via preoperative imaging and pathological examination. For patients presenting with small, low-risk nodules, surgeons might appropriately limit the extent of dissection to mitigate postoperative complications such as hoarseness and hypoparathyroidism, thereby maximizing the cosmetic and rapid-recovery benefits inherent to this minimally invasive technique.

The proposed model relies entirely on routine clinical and laboratory parameters, conferring distinct advantages in terms of broad accessibility and cost-effectiveness, positioning it as a highly practical adjunctive tool for patient risk stratification. Nonetheless, this study presents certain limitations. First, its single-center, retrospective design inherently introduces the potential for selection and information biases. Second, the overall sample size remains relatively limited. Furthermore, the lack of external validation means the model’s generalizability and robustness across diverse populations require further substantiation. Future large-scale, multicenter, prospective cohort studies are imperative to rigorously validate the stability and broader applicability of this nomogram. Subsequent research endeavors could integrate advanced imaging features and novel molecular biomarkers to construct a multi-dimensional, comprehensive predictive framework. Enhancing the predictive accuracy for lymph node metastasis will significantly facilitate the translation of this model into routine clinical practice, propel the advancement and adoption of the GTET technique, and furnish a more robust, evidence-based foundation for the individualized management of patients with thyroid cancer.

Conclusions

This study successfully establishes a feasible and highly effective predictive model for estimating CLNM risk in PTC patients undergoing transaxillary endoscopic thyroidectomy. The implementation of this nomogram has the potential to actively guide clinical decision-making, optimize individualized treatment strategies, and ultimately assist clinicians in mitigating the risks of both over-treatment and under-treatment.

Acknowledgments

The authors thank the patients and their families for their assistance, as well as all clinical departments.

Footnote

Reporting Checklist: The authors have completed the TRIPOD reporting checklist. Available at https://gs.amegroups.com/article/view/10.21037/gs-2026-0228/rc

Data Sharing Statement: Available at https://gs.amegroups.com/article/view/10.21037/gs-2026-0228/dss

Peer Review File: Available at https://gs.amegroups.com/article/view/10.21037/gs-2026-0228/prf

Funding: None.

Conflicts of Interest: All authors have completed the ICMJE uniform disclosure form (available at https://gs.amegroups.com/article/view/10.21037/gs-2026-0228/coif). The authors have no conflicts of interest to declare.

Ethical Statement: The authors are accountable for all aspects of the work in ensuring that questions related to the accuracy or integrity of any part of the work are appropriately investigated and resolved. This study was conducted in accordance with the Declaration of Helsinki and its subsequent amendments. The study protocol was approved by the Institutional Ethics Committee of The Third Hospital of Mianyang (No. 2026S12-1), which waived informed consent from enrolled patients because of the retrospective nature of this study.

Open Access Statement: This is an Open Access article distributed in accordance with the Creative Commons Attribution-NonCommercial-NoDerivs 4.0 International License (CC BY-NC-ND 4.0), which permits the non-commercial replication and distribution of the article with the strict proviso that no changes or edits are made and the original work is properly cited (including links to both the formal publication through the relevant DOI and the license). See: https://creativecommons.org/licenses/by-nc-nd/4.0/.

References

1. Ringel MD, Sosa JA, Baloch Z, et al. 2025 American Thyroid Association Management Guidelines for Adult Patients with Differentiated Thyroid Cancer. Thyroid 2025;35:841-985. [Crossref] [PubMed]
2. Jiang LH, Yin KX, Wen QL, et al. Predictive Risk-scoring Model For Central Lymph Node Metastasis and Predictors of Recurrence in Papillary Thyroid Carcinoma. Sci Rep 2020;10:710. [Crossref] [PubMed]
3. Li X, Ding W, Zhang H. Surgical outcomes of endoscopic thyroidectomy approaches for thyroid cancer: a systematic review and network meta-analysis. Front Endocrinol (Lausanne) 2023;14:1256209. [Crossref] [PubMed]
4. Xu W, Teng C, Ding G, et al. Oncologic safety and surgical outcomes of the different surgical approaches of endoscopic thyroidectomy for papillary thyroid carcinoma. Surg Today 2023;53:554-61. [Crossref] [PubMed]
5. Jiang WJ, Yan PJ, Zhao CL, et al. Comparison of total endoscopic thyroidectomy with conventional open thyroidectomy for treatment of papillary thyroid cancer: a systematic review and meta-analysis. Surg Endosc 2020;34:1891-903. [Crossref] [PubMed]
6. Ding Y, Qiu C, Zhu C, et al. Comparison of gasless transaxillary endoscopic thyroidectomy, endoscopic thyroidectomy via areola approach and conventional open thyroidectomy in patients with unilateral papillary thyroid carcinoma. World J Surg Oncol 2024;22:148. [Crossref] [PubMed]
7. Gao X, Luo W, He L, et al. Predictors and a Prediction Model for Central Cervical Lymph Node Metastasis in Papillary Thyroid Carcinoma (cN0). Front Endocrinol (Lausanne) 2022;12:789310.
8. Luo X, Wang J, Xu M, et al. Risk model and risk stratification to preoperatively predict central lymph node metastasis in papillary thyroid carcinoma. Gland Surg 2020;9:300-10. [Crossref] [PubMed]
9. Dolidze DD, Shabunin AV, Mumladze RB, et al. A Narrative Review of Preventive Central Lymph Node Dissection in Patients With Papillary Thyroid Cancer - A Necessity or an Excess. Front Oncol 2022;12:906695. [Crossref] [PubMed]
10. Ge JN, Yu ST, Sun BH, et al. A novel anatomy-based five-settlement method for endoscopic thyroid lobectomy and ipsilateral central compartment neck dissection via gasless unilateral axillary approach: a preliminary report. Front Endocrinol (Lausanne) 2023;14:1147313. [Crossref] [PubMed]
11. Sun P, Chen W, Mak TK, et al. Right Central Lymph Node Dissection in Thyroidectomy: Can Endoscopic Chest-Breast Approach Be Used? J Laparoendosc Adv Surg Tech A 2020;30:308-14. [Crossref] [PubMed]
12. Xu JJ, Zhang LZ, Zhang QH, et al. Clinical application of the gasless unilateral axillary approach in endoscopic thyroid surgery. Zhonghua Er Bi Yan Hou Tou Jing Wai Ke Za Zhi 2020;55:913-20. [Crossref] [PubMed]
13. Chereau N, Christou N, Caiazzo R, et al. Impact of prophylactic central lymph node dissection on the complications and recurrence rates in papillary thyroid carcinoma-An AFCE (French-speaking Association of Endocrine Surgery) multicentre study based on the EUROCRINE® national data. Cancer Rep (Hoboken) 2024;7:e1993. [Crossref] [PubMed]
14. Zhai M, Zhang D, Long J, et al. The global burden of thyroid cancer and its attributable risk factor in 195 countries and territories: A systematic analysis for the Global Burden of Disease Study. Cancer Med 2021;10:4542-54. [Crossref] [PubMed]
15. Zhou D, Zhang Z, Dou X, et al. Advances in the assessment of cosmetic outcomes, sensory alteration in surgical areas, and health-related quality of life of endoscopic thyroidectomy. World J Surg Oncol 2024;22:52. [Crossref] [PubMed]
16. Liu Y, Wang J, Chen S, et al. Preliminary report on a novel technique for endoscopic transaxillary thyroidectomy: a case-control study. Int J Surg 2024;110:654-9. [Crossref] [PubMed]
17. Liu H, Zhang S, Liu X, et al. Thyroid cancer survival and prognostic factors in Yantai, China (2012-2022): a population-based study. BMC Cancer 2024;24:1587. [Crossref] [PubMed]
18. Shteinshnaider M, Muallem Kalmovich L, Koren S, et al. Reassessment of Differentiated Thyroid Cancer Patients Using the Eighth TNM/AJCC Classification System: A Comparative Study. Thyroid 2018;28:201-9. [Crossref] [PubMed]
19. van Velsen EFS, Visser WE, Stegenga MT, et al. Finding the Optimal Age Cutoff for the UICC/AJCC TNM Staging System in Patients with Papillary or Follicular Thyroid Cancer. Thyroid 2021;31:1041-9. [Crossref] [PubMed]
20. Wang L, Cheng P, Zhu L, et al. Predicting lymph node metastasis in papillary thyroid carcinoma with Hashimoto’s thyroiditis using regression and network analysis. Sci Rep 2024;14:27585. [Crossref] [PubMed]
21. Duan S, Yang Z, Wei G, et al. Nomogram for predicting the risk of central lymph node metastasis in papillary thyroid microcarcinoma: a combination of sonographic findings and clinical factors. Gland Surg 2024;13:1016-30. [Crossref] [PubMed]
22. Liu Y, Huang Y, Mo G, et al. Combined prognostic value of preoperative serum thyrotrophin and thyroid hormone concentration in papillary thyroid cancer. J Clin Lab Anal 2022;36:e24503. [Crossref] [PubMed]
23. Si L, Mei H, Wang Q, et al. Surgical outcomes of different approaches to dissection of lymph nodes posterior to right recurrent laryngeal nerve: a retrospective comparative cohort study of endoscopic thyroidectomy via the areolar approach and via the axillo-breast approach. Gland Surg 2022;11:1936-45. [Crossref] [PubMed]
24. Kayhan Y, Azizova L, Yılmaz M, et al. Prognostic factors for aggressiveness in subcentimeter papillary thyroid carcinoma: impact of tumor size and lymph node metastases. Arch Endocrinol Metab 2024;68:e230422. [Crossref] [PubMed]
25. Can N, Bulbul BY, Ozyilmaz F, et al. The Impact of Total Tumor Diameter on Lymph Node Metastasis and Tumor Recurrence in Papillary Thyroid Carcinomas. Diagnostics (Basel) 2024;14:272. [Crossref] [PubMed]
26. Zhou W, Li L, Hao X, et al. Predicting central lymph node metastasis in papillary thyroid microcarcinoma: a breakthrough with interpretable machine learning. Front Endocrinol (Lausanne) 2025;16:1537386. [Crossref] [PubMed]
27. Tang L, Qu RW, Park J, et al. Prevalence of Occult Central Lymph Node Metastasis by Tumor Size in Papillary Thyroid Carcinoma: A Systematic Review and Meta-Analysis. Curr Oncol 2023;30:7335-50. [Crossref] [PubMed]