Development and validation of a risk score model for patient stratification and personalized management of papillary thyroid cancer

Introduction

Thyroid cancer (TC), was the most common endocrine malignancy, and its incidence had significantly increased over the past few decades, mainly due to the increasing detection of small, indolent tumors (1,2). These increase in small cancers were considered as low-risk due to the excellent prognosis (3). Papillary thyroid cancer (PTC) is the predominant type of TC, and surgical intervention has consistently been its primary initial treatment modality. However, there is currently no definitive evaluation system for determining the scope of surgery, whether it should be total thyroidectomy, unilateral thyroidectomy, or the necessity of lymph node (LN) clearance. Over the past decade, doctors had focused on identifying low-risk PTC to reduce the extent of LN dissection or even adopted active surveillance (AS) and thermal ablation (TA) to minimize the damage caused by excessive surgical treatment (2). Although there is still some controversy surrounding TA in the treatment of PTC, as a local treatment modality, it offers notable advantages such as minimally invasive procedures, absence of visible scars, and maximal preservation of thyroid function, and shorten the operation times and postoperative hospitalization times (4,5). For selected patients with low-risk characteristics, limited to the thyroid gland without high-risk factors, TA can serve as a prominent treatment option. Accumulating studies confirmed the efficacy and safety of TA (6,7). Some localized PTC (T1N0M0) patients undergoing TA had shown no significant difference in local recurrence and had fewer complications compared to patients with traditional surgical therapy (5,8,9). According to the report of expert consensus workshop, indications for TA therapy were rigorous (10). Pre-operative identification of cervical LN status was a crucial step.

LN metastasis occurs in 20–50% of TC cases and is the most common form of metastasis in PTC (2). LN metastasis, including central lymph node (CLN) metastasis and lateral lymph node (LLN) metastasis, is an independent risk factor associated with recurrence and survival outcomes (11-13). The preoperative assessment of CLN status significantly influences the treatment approach for PTC. Preoperative ultrasound (US) is routinely employed to estimate LN status. LLN can be effectively screened using US with a sensitivity of 93.8% (14). However, for CLN, US and US-guided fine-needle aspiration are limited by the interference of thyroid and surrounding anatomical structures (trachea, esophagus) as well as underlying thyroid disease. The preoperative detection rate of CLN metastasis using US is inadequate, with a sensitivity of only 20–31% (14-16). Therefore, we need to develop a non-destructive and validated method for estimating CLN status.

This study aimed to utilize demographic characteristics and accessible preoperative clinical features to assess the risk of CLN metastasis in unifocal PTC and establish a risk score model to guide the initial treatment. We present this article in accordance with the TRIPOD reporting checklist (available at https://gs.amegroups.com/article/view/10.21037/gs-24-344/rc).

Methods

Patient identification

This retrospective study was performed with approval by Institutional Review Board of the Union Hospital, Tongji Medical College, Huazhong University of Science and Technology (No. 0340-01). Written informed consent for participation was not required for this study in accordance with the national legislation and the institutional requirements. The study was conducted in accordance with the Declaration of Helsinki (as revised in 2013). Patients who underwent unilateral thyroid lobe plus isthmus excision or total thyroidectomy and received a final pathological diagnosis of unifocal PTC at Union Hospital between November 2009 and August 2022 were enrolled in the study. The ethnicity of all participants was Asian. The exclusion criteria were as follows: incomplete medical records and pre-operative examination, a history of neck surgery, a family history of TC, no cervical CLN dissection, and those aged 18 years and younger. Patients from November 2009 to December 2020 were identified as a derivation cohort and from January 2021 to August 2022 were included as a validation cohort. A total of 7,338 patients were diagnosed with unifocal PTC, of which 5,374 patients were finally identified in the analysis, consisting of 3,542 patients in derivation cohort and 1,832 patients in validation cohort.

US

Each US report was performed by two experienced sonographers. Variables obtained from US reports included tumor location, border, calcification, aspect ratio (anteroposterior/transverse), tumor size, capsular abnormalities (including immediately adjacent to capsule, capsular continuity interrupted, and breach of capsule), and LN abnormalities (including enlargement, loss of the fatty hilum, cortical thickening and diminished, microcalcifications, cystic aspect, peripheral vascularity, hyperechogenicity, and round shape).

Surgery method

Each patient underwent unilateral thyroid lobectomy with isthmus or total thyroidectomy and intraoperative frozen section examination. Malignancy, confirmed by intraoperative pathology, would be treated with CLN dissection. CLN (level VI and VII) included pretracheal, paratracheal, prelaryngeal/Delphian, and anterior mediastinum LN.

Pathology

Routine pathological examination was performed by two experienced pathologists according to WHO criteria. The whole specimens were serially sectioned at 3 µm intervals for hematoxylin and eosin staining and immunohistochemistry would be conducted if required.

Statistical analysis

Data were presented as number (percentages) for categorical variables. The χ² or Fisher exact tests were used to compared qualitative data, while the Wilcoxon rank-sum test was used to compared quantitative data. In the derivation cohort, we utilized a penalized multivariable logistic regression technique, least absolute shrinkage and selection operator (LASSO), to identify predictors of CLN metastasis and to develop a risk-scoring system. The validation cohort was used to assess the performance of the risk-scoring system. Risk score weights were assigned by dividing the coefficients of the predictors with the lowest coefficient value in the final model and rounding to the nearest integer. Points were calculated for each patient by adding these weights. Using this risk score on a continuous scale, we then assessed its predictive performance by the area under the curve (AUC) value. Calibration was assessed by plotting the observed versus predicted incidence rate across deciles of the risk score. Multivariable logistic regression analysis was performed to identify risk factors associated with LLN metastasis. The statistical analysis was conducted using R 4.0.5 (R Foundation for Statistical Computing, Vienna, Austria). All the tests were two-sides, and P values less than 0.05 were considered statistically significant.

Results

Patients’ characteristics

The derivation cohort included 3,542 patients with unifocal PTC, while the validation cohort included 1,832 patients with unifocal PTC. Table 1 shows the comparison of baseline characteristics the two cohorts. Patients in the validation cohort exhibited a higher rate of underlying disease (including hypertension and diabetes) and tended to have abnormal ultrasonographic characteristics (unclear border, calcification, and aspect ratio >1) and abnormal thyroid stimulating hormone (TSH) levels compared to patients in derivation cohort. The proportion of US-detected abnormal LN is higher in the training cohort.

Predicts of CLN metastasis

Univariable analysis of risk factor associated with CLN metastasis in the derivation cohort is provided in Table 2. The independent predictors associated with CLN metastasis were derived from a multivariable logistic regression analysis. Predictors in the derivation cohort and their respective weights in the risk-score system are shown in Figure 1. Due to the particularity of isthmic foci, we additionally performed multivariable analyses comparing isthmus with right and left lobes (Table S1), and found no significant difference in CLN metastasis among them (isthmus as reference, right lobe, P=0.26; left lobe, P=0.06). Subsequently, multivariable analysis was performed among isthmus, upper pole, middle and lower pole, and found that the risk of CLN metastasis was higher in the lower pole than in the non-lower pole (including isthmus, upper pole and middle). Ten multivariable predictors were independently associated with CLN metastasis, including sex, age, body mass index (BMI), Hashimoto’s disease, tumor location, calcification, capsular abnormalities, CLN abnormalities, LLN abnormalities and tumor size.

Figure 1 Multivariable predictors of CLN metastasis in the derivation cohort and their respective weights in the risk score. OR, odds ratio; CI, confidence interval; BMI, body mass index; CLN, central lymph node; LLN, lateral lymph node; US, ultrasound.

Risk score building and validation

The risk-score system was constructed using these multivariable predictors including sex, age, BMI, Hashimoto’s disease, tumor location, calcification, capsular abnormalities, CLN abnormalities, LLN abnormalities, and tumor size. The scores theoretically ranged from 0 to 38. Figure 2 shows the relationship between the risk score value and the observed incidence of CLN metastasis. The scores ranged from 0 to 34 in the derivation and from 0 to 36 in validation cohorts. The odds ratio (OR) associated with a one-point increase in score was 1.25 [95% confidence interval (CI): 1.23–1.27] and 1.22 (95% CI: 1.19–1.25) in the derivation and validation cohorts, respectively. Calibration plots showing the predicted probability vs. observed incidence of CLN metastasis in the derivation and validation cohorts are depicted in Figure 3. In derivation and validation cohorts, the AUC value of the final model was 0.764 and 0.72, respectively. The risk score demonstrated good predictability and calibration.

Figure 2 Relationship between the risk score value and the observed incidence of CLN metastasis in the derivation and validation cohorts. CLN, central lymph node.

Figure 3 Calibration plots. Calibration plots showing the predicted probability vs. observed incidence of CLN metastasis in the derivation and validation cohorts. The diagonal grey dotted line represents the perfect calibration (y=x). CLN, central lymph node.

A risk stratification was developed to better guide clinical practice (Figure 4). The optimal cut-off value was determined to be 13.5 (AUC 0.77), and the nearest integer, 13, was selected as the cut-off. Moreover, quartiles 1 and 3 were utilized to stratify the risk levels. The risk score was divided into four levels based on the incidence of CLN metastasis. A score of 0 to 9 was classified as low risk, with an observed incidence of 20.34% (895 patients, 25.27% of the derivation cohort). A score of 10 to 13 was classified as low to intermediate risk, with an observed incidence of 37.42% (1,045 patients, 29.5% of the derivation cohort). A score of 14 to 17 was classified as intermediate to high risk, with an observed incidence of 59.65% (922 patients, 26.03% of the derivation cohort). A score exceeding 17 was classified as high risk, with an observed incidence of 83.82% (680 patients, 19.2% of the derivation cohort). When the scores higher than 25 (108 patients, 3.05% of derivation cohort), the incidence of CLN metastasis number ≥5 reached 57% in derivation cohort and 66% in validation cohort (Figure S1). Analysis included 425 patients who underwent LLN dissection to elucidate risk factors correlated with LLN metastasis. Number of metastatic CLN was significantly correlated with LLN metastasis in unifocal PTC patients (0 as reference, 1–4: OR =6.155, P<0.001; ≥5: OR =17.561, P<0.001; Table S2).

Figure 4 Observed incidence of CLN metastasis in different risk group. Observed incidence of CLN metastasis according to categories of the risk score in the derivation and validation cohorts. CLN, central lymph node.

Discussion

With the increasing incidence of TC, the therapeutic strategies for TC are continuously evolving. In particular, the management of low-risk PTC is being refined, and minimally invasive alternatives to surgery are being explored, such as AS and TA therapy (17,18). In addition to the size and invasiveness of the tumor itself, the presence of LN metastasis is a crucial factor influencing the selection of initial treatment modalities. Currently, there is a lack of a definitive measurement standard for assessing LN status in clinical practice. US, as the preferred examination for TC, is effective in evaluating LLN but falls short in assessing CLN (14). Misjudgment of CLN metastasis can lead to erroneous selection of initial treatment approaches, missing the opportunity for minimally invasive or conservative treatments. Insufficient LN clearance can result in subsequent recurrence or even more extensive metastasis, while excessive clearance poses risks of surgical complications. For tumors with high metastasis risk, more aggressive treatment measures may be required. However, for tumors with low metastatic risk, we considered that TA was a novel treatment method to manage PTC that effectively eliminated the lesion and avoided over-treatment like surgery (19). Some researchers suggested that AS could be an effective treatment (20-22). The 2015 American Thyroid Association Management Guidelines also supported AS as an alternative to immediate surgery (2). Unifocal PTC patients with clinically negative LN status were major selected population for AS or TA therapy (10,17). Few studies focused on unifocal PTC to build clinical predictive models of CLN metastasis. In our research, demographic characteristics, past history and routine pre-operative examinations results were collected to build a useable and efficient risk score to determine the appropriate treatment strategy.

We conducted a multivariate analysis on several preoperative factors and found that age, gender, BMI, Hashimoto’s disease, tumor location, calcification, capsule abnormalities, CLN abnormalities, LLN abnormalities and tumor size were independent risk factors for CLN metastasis. These factors were included in the predictive model for further investigation. The risk score, established in the derivation cohort, showed excellent predictability and calibration and were well validated in different populations.

We found that age was an extremely important factor that affects CLN metastasis, with a higher risk of CLN metastasis observed in younger patients. Therefore, we further subdivided age into four groups: ≤29, 30–44, 45–59, and ≥60 years, in order to more accurately predict the risk of CLN metastasis and more suitably apply in clinical practice. This might be related to the difference in molecular features, hormone levels, increased psychological stress or irregular lifestyle between young patients and elderly (23). The specific reasons, however, required further investigation. Although guidelines suggested that patients under the age of 55 with PTC had a lower stage and a better prognosis, younger patients tended to have more varied prognostic outcomes due to the longer survival time (24). Moreover, younger patients often had higher demands for aesthetic outcomes and thyroid function, and a greater expectation of quality of life. Therefore, early and appropriate treatment was particularly important for young patients who required greater attention. In addition, for the elderly population, most of whom had underlying diseases such as heart disease, hypertension, and diabetes, the risks of anesthesia and surgery were high. Therefore, it was very important and meaningful to identify elderly patients who were suitable for TA treatment, avoid unnecessary surgery, benefit from appropriate treatment, reduce unnecessary risks, and shorten the duration of surgery. The gender bias that males were associated with worse behavior and outcomes compared to females and that females had a higher incidence of TC, existed in TC (25,26). The explanation for this bias, which remained controversial, might be related to genetic forms, estrogen and estrogen receptor levels or androgen levels and required further researches (27).

Six US parameters of cancer foci were calculated to estimate the risk in our model. We discovered that nodules in the left lobe, right lobe, or isthmus did not increase the risk of CLN metastasis, while nodules in the lower pole were a high-risk factor for CLN metastasis. This might be related to the anatomical location, as lower pole nodules are closer to the position of CLN, which might be the preferred site for CLN metastasis. A recent study had yielded similar findings (28). However, further investigations were needed to determine the specific reasons behind these results. Tumor size had always played a significant role on CLN metastasis (29). Compared with previous models, our study provided a more detailed classification of tumor size, especially for tumors smaller than 10 mm, to accurately target TA candidates. In clinical practice, certain abnormalities in CLN can be attributed to autoimmune thyroid disorders, which can mislead physicians into performing unnecessary procedures, such as CLN clearance (30). By incorporating these factors into the model, we can assist physicians in better recognizing such anomalies.

Most researches suggested that Hashimoto’s disease was a detrimental factor in TC, promoting its development, and that TC with the background of Hashimoto’s disease had more aggressive biological behavior (31). Our analysis showed that Hashimoto’s disease was more likely to play a protective role on CLN metastasis, possibly due to the immune response caused by high circulating antibodies level. The patient’s immune system might generate a response against TC cells, inhibiting the spread of TC (32). Similar findings had been reported in previous study (33). Further researches were needed to investigate the role and mechanism of Hashimoto’s disease in TC.

Based on a large population of unifocal PTC, our model considered each node as an individual and assessed the metastatic risk for single nodes. The predictors in the model were clinically easily accessible. The risk score could be quickly calculated in clinical practice as it did not require complex transformation of examination results. In previous clinical practice, the size of the tumor and the US findings of neck LN had been the primary factors for clinicians to assess the risk of CLN metastasis. However, for other risk factors related to CLN metastasis, such as age, gender, past history, and US findings of the tumor, clinicians might consider them but how to assign risk weights was unclear, leading to the potential neglect of this information. Our model integrated demographic features, Hashimoto’s disease, and US parameters to provide clinicians with a quantitative and more accurate assessment of the risk of CLN metastasis, enabling more appropriate treatment decisions. For patients with low risk of CLN metastasis, TA treatment was recommended, while for high-risk patients, surgical treatment was more advisable. For patients categorized as low to intermediate risk, the option of therapeutic alternatives combined with regular follow-up examinations may be considered in cases where the patient expresses a strong preference or when their baseline condition precludes surgical intervention. For patients with a weighted score of 26 or more, the rate of CLN metastatic number ≥5 was high, significantly increasing risk of LLN metastasis. Timely removal of metastatic LN was essential to reduce recurrence. More aggressive treatment, total thyroidectomy plus post-operative radioactive iodine therapy and comprehensive assessment of cervical LN, such as computed tomography or magnetic resonance imaging, should be considered for those patients. By implementing a stratified treatment approach, excessive surgical treatment and arbitrary observation could be avoided.

However, a single risk score could not replace comprehensive clinical assessment. Due to the independent evaluation of each nodule in our model, we were able to better understand the individual CLN metastasis risk of each nodule compared to existing predictive models. Previous models considered multifocality as an independent risk factor, making it difficult to accurately assess the risk of individual nodules and further evaluate the cumulative risk of multiple nodules, thus making them unsuitable as a predictive model for TA selection (34,35). Our model demonstrated greater accuracy and advantages in its use.

There were some limitations in our study. First, as a retrospective study, there was the possibility of selection bias. Secondly, our study lacked external data validation. Thirdly, as increasing biomarkers for TC were discovered and investigated, the management of TC had become more individualized (36). Moreover, the availability of liquid biopsy, a less invasive method, had led to the widespread use of biomarkers in clinical practice. However, biomarkers were not analyzed in our study as few patients were examined. In the future, we would collect data on biomarkers to further refine our predictive model, improve its accuracy and determine optimal treatment choices for PTC patients. Furthermore, we would further validate our model on external data and drive the construction of a model in multifocal PTC.

Conclusions

Compared to relying solely on tumor size and LN US findings, our risk score, incorporating demographic characteristics and routine pre-operative examinations, serves as a more practical and effective tool for risk stratification of CLN metastasis in unifocal PTC patients. This facilitates in clinical decision-making by identifying candidates suitable for treatment alternatives to surgery and discerning those who should contemplate more aggressive therapeutic strategies.

Acknowledgments

The authors sincerely appreciate the invaluable support of our department members.

Funding: This work was supported by Key Program of Natural Science Foundation of Hubei Province (Grant No. 2021BCA142, to T.H.)

Footnote

Reporting Checklist: The authors have completed the TRIPOD reporting checklist. Available at https://gs.amegroups.com/article/view/10.21037/gs-24-344/rc

Data Sharing Statement: Available at https://gs.amegroups.com/article/view/10.21037/gs-24-344/dss

Peer Review File: Available at https://gs.amegroups.com/article/view/10.21037/gs-24-344/prf

Conflicts of Interest: All authors have completed the ICMJE uniform disclosure form (available at https://gs.amegroups.com/article/view/10.21037/gs-24-344/coif). T.H. reports that this work was supported by Key Program of Natural Science Foundation of Hubei Province (No. 2021BCA142). The other 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. The study was conducted in accordance with the Declaration of Helsinki (as revised in 2013). The study was reviewed and approved by Institutional Review Board of the Union Hospital, Tongji Medical College, Huazhong University of Science and Technology (No. 0340-01). Written informed consent for participation was not required for this study in accordance with the national legislation and the institutional requirements.

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. Megwalu UC, Moon PK. Thyroid Cancer Incidence and Mortality Trends in the United States: 2000-2018. Thyroid 2022;32:560-70. [Crossref] [PubMed]
2. Haugen BR, Alexander EK, Bible KC, et al. 2015 American Thyroid Association Management Guidelines for Adult Patients with Thyroid Nodules and Differentiated Thyroid Cancer: The American Thyroid Association Guidelines Task Force on Thyroid Nodules and Differentiated Thyroid Cancer. Thyroid 2016;26:1-133. [Crossref] [PubMed]
3. Wang TS, Sosa JA. Thyroid surgery for differentiated thyroid cancer - recent advances and future directions. Nat Rev Endocrinol 2018;14:670-83. [Crossref] [PubMed]
4. Lan Y, Luo Y, Zhang M, et al. Quality of Life in Papillary Thyroid Microcarcinoma Patients Undergoing Radiofrequency Ablation or Surgery: A Comparative Study. Front Endocrinol (Lausanne) 2020;11:249. [Crossref] [PubMed]
5. Shen K, Xue S, Xie Y, et al. Comparison of thermal ablation and routine surgery for the treatment of papillary thyroid microcarcinoma: a systematic review and Meta-analysis. Int J Hyperthermia 2020;37:913-24. [Crossref] [PubMed]
6. Choi Y, Jung SL. Efficacy and Safety of Thermal Ablation Techniques for the Treatment of Primary Papillary Thyroid Microcarcinoma: A Systematic Review and Meta-Analysis. Thyroid 2020;30:720-31. [Crossref] [PubMed]
7. Cao XJ, Wang SR, Che Y, et al. Efficacy and Safety of Thermal Ablation for Treatment of Solitary T1N0M0 Papillary Thyroid Carcinoma: A Multicenter Retrospective Study. Radiology 2021;300:209-16. [Crossref] [PubMed]
8. Kim HJ, Cho SJ, Baek JH. Comparison of Thermal Ablation and Surgery for Low-Risk Papillary Thyroid Microcarcinoma: A Systematic Review and Meta-Analysis. Korean J Radiol 2021;22:1730-41. [Crossref] [PubMed]
9. Yan L, Li X, Li Y, et al. Comparison of ultrasound-guided radiofrequency ablation versus thyroid lobectomy for T1bN0M0 papillary thyroid carcinoma. Eur Radiol 2023;33:730-40. [Crossref] [PubMed]
10. Xu D, Ge M, Yang A, et al. Expert consensus workshop report: Guidelines for thermal ablation of thyroid tumors (2019 edition). J Cancer Res Ther 2020;16:960-6. [Crossref] [PubMed]
11. Podnos YD, Smith D, Wagman LD, et al. The implication of lymph node metastasis on survival in patients with well-differentiated thyroid cancer. Am Surg 2005;71:731-4. [Crossref] [PubMed]
12. Scheumann GF, Gimm O, Wegener G, et al. Prognostic significance and surgical management of locoregional lymph node metastases in papillary thyroid cancer. World J Surg 1994;18:559-67; discussion 567-8. [Crossref] [PubMed]
13. Zaydfudim V, Feurer ID, Griffin MR, et al. The impact of lymph node involvement on survival in patients with papillary and follicular thyroid carcinoma. Surgery 2008;144:1070-7; discussion 1077-8. [Crossref] [PubMed]
14. Hwang HS, Orloff LA. Efficacy of preoperative neck ultrasound in the detection of cervical lymph node metastasis from thyroid cancer. Laryngoscope 2011;121:487-91. [Crossref] [PubMed]
15. Stulak JM, Grant CS, Farley DR, et al. Value of preoperative ultrasonography in the surgical management of initial and reoperative papillary thyroid cancer. Arch Surg 2006;141:489-94; discussion 494-6. [Crossref] [PubMed]
16. O'Connell K, Yen TW, Quiroz F, et al. The utility of routine preoperative cervical ultrasonography in patients undergoing thyroidectomy for differentiated thyroid cancer. Surgery 2013;154:697-701; discussion 701-3. [Crossref] [PubMed]
17. Chen DW, Lang BHH, McLeod DSA, et al. Thyroid cancer. Lancet 2023;401:1531-44. [Crossref] [PubMed]
18. Lim H, Devesa SS, Sosa JA, et al. Trends in Thyroid Cancer Incidence and Mortality in the United States, 1974-2013. JAMA 2017;317:1338-48. [Crossref] [PubMed]
19. Tufano RP, Mohamed Ali K. The Year in Surgical Thyroidology: Recent Technological Developments and Future Challenges. Thyroid 2022;32:14-8. [Crossref] [PubMed]
20. Ito Y, Miyauchi A, Inoue H, et al. An observational trial for papillary thyroid microcarcinoma in Japanese patients. World J Surg 2010;34:28-35. [Crossref] [PubMed]
21. Cho SJ, Suh CH, Baek JH, et al. Active Surveillance for Small Papillary Thyroid Cancer: A Systematic Review and Meta-Analysis. Thyroid 2019;29:1399-408. [Crossref] [PubMed]
22. Sakai T, Sugitani I, Ebina A, et al. Active Surveillance for T1bN0M0 Papillary Thyroid Carcinoma. Thyroid 2019;29:59-63. [Crossref] [PubMed]
23. Vriens MR, Moses W, Weng J, et al. Clinical and molecular features of papillary thyroid cancer in adolescents and young adults. Cancer 2011;117:259-67. [Crossref] [PubMed]
24. Perrier ND, Brierley JD, Tuttle RM. Differentiated and anaplastic thyroid carcinoma: Major changes in the American Joint Committee on Cancer eighth edition cancer staging manual. CA Cancer J Clin 2018;68:55-63.
25. Nilubol N, Zhang L, Kebebew E. Multivariate analysis of the relationship between male sex, disease-specific survival, and features of tumor aggressiveness in thyroid cancer of follicular cell origin. Thyroid 2013;23:695-702. [Crossref] [PubMed]
26. Zahedi A, Bondaz L, Rajaraman M, et al. Risk for Thyroid Cancer Recurrence Is Higher in Men Than in Women Independent of Disease Stage at Presentation. Thyroid 2020;30:871-7. [Crossref] [PubMed]
27. Suteau V, Munier M, Briet C, et al. Sex Bias in Differentiated Thyroid Cancer. Int J Mol Sci 2021;22:12992. [Crossref] [PubMed]
28. Liu Y, Huang J, Zhang Z, et al. Ultrasonic Characteristics Improve Prediction of Central Lymph Node Metastasis in cN0 Unifocal Papillary Thyroid Cancer. Front Endocrinol (Lausanne) 2022;13:870813. [Crossref] [PubMed]
29. Sun W, Lan X, Zhang H, et al. Risk Factors for Central Lymph Node Metastasis in CN0 Papillary Thyroid Carcinoma: A Systematic Review and Meta-Analysis. PLoS One 2015;10:e0139021. [Crossref] [PubMed]
30. Ralli M, Angeletti D, Fiore M, et al. Hashimoto's thyroiditis: An update on pathogenic mechanisms, diagnostic protocols, therapeutic strategies, and potential malignant transformation. Autoimmun Rev 2020;19:102649. [Crossref] [PubMed]
31. Ferrari SM, Fallahi P, Elia G, et al. Thyroid autoimmune disorders and cancer. Semin Cancer Biol 2020;64:135-46. [Crossref] [PubMed]
32. Sulaieva O, Selezniov O, Shapochka D, et al. Hashimoto's thyroiditis attenuates progression of papillary thyroid carcinoma: deciphering immunological links. Heliyon 2020;6:e03077. [Crossref] [PubMed]
33. Paulson LM, Shindo ML, Schuff KG. Role of chronic lymphocytic thyroiditis in central node metastasis of papillary thyroid carcinoma. Otolaryngol Head Neck Surg 2012;147:444-9. [Crossref] [PubMed]
34. Feng JW, Hong LZ, Wang F, et al. A Nomogram Based on Clinical and Ultrasound Characteristics to Predict Central Lymph Node Metastasis of Papillary Thyroid Carcinoma. Front Endocrinol (Lausanne) 2021;12:666315. [Crossref] [PubMed]
35. Wang Z, Chang Q, Zhang H, et al. A Clinical Predictive Model of Central Lymph Node Metastases in Papillary Thyroid Carcinoma. Front Endocrinol (Lausanne) 2022;13:856278. [Crossref] [PubMed]
36. Agarwal S, Bychkov A, Jung CK. Emerging Biomarkers in Thyroid Practice and Research. Cancers (Basel) 2021;14:204. [Crossref] [PubMed]