Preoperative multiparametric ultrasound for the prediction of central lymph node metastasis in papillary thyroid carcinoma

Introduction

The incidence of thyroid cancer has increased significantly, with papillary thyroid carcinoma (PTC) accounting for over 80% of cases (1). Cervical lymph node metastasis (LNM) occurs in 30–60% of PTC patients, and can serve as a prognostic marker, as it is associated with an increased risk of local recurrence (2). Current guidelines recommend central lymph node dissection (CLND) only for patients with clinically involved central or lateral lymph nodes (3). However, to date, no consensus has been reached as to the benefits and risks of prophylactic CLND. The American Thyroid Association states that evidence remains insufficient to demonstrate a reduction in recurrence rates with prophylactic CLND, and notes that it increases surgical risks (4). For low-risk papillary thyroid microcarcinoma (PTMC) without LNM, active surveillance, minimally invasive treatment, or surgery may be considered (5). Thus, the preoperative assessment of LNM risk is critical for optimizing the management of PTC.

Current recommendations emphasize the adoption of an individualized approach, requiring the thorough evaluation of each thyroid nodule to estimate both its malignant potential and its likelihood of producing clinical symptoms (6). Ultrasound (US) is the primary imaging modality for evaluating thyroid nodules and cervical lymph nodes preoperatively (4). Only a minority of nodules necessitate extensive investigations, such as fine-needle aspiration biopsy (FNAB) with cytological or molecular testing, and most can be effectively managed through systematic cervical US combined with clinical risk stratification, which forms a solid basis for initial treatment decisions (7). In cases of pathologically confirmed malignancy, the meticulous preoperative assessment of cervical lymph nodes remains essential in guiding the surgical plan.

Central lymph node metastasis (CLNM) is typically the initial and most common site of metastasis (8), but due to anatomical constraints, US has limited sensitivity in detecting central lymph node involvement (9). Recent studies have explored the use of the US features of primary thyroid lesions in predicting the risk of CLNM (10,11). Compared with single-modality US, multiparametric ultrasound (MULTI-US) provides more comprehensive information. Conventional ultrasound (CON-US) (B-mode and color Doppler imaging) serves as the foundation for evaluating the morphology and vascularity of thyroid tumors (4). With the promotion of advanced US techniques in screening and diagnosis, elastography ultrasound (EUS) and contrast-enhanced ultrasound (CEUS) have been recommended by guidelines for characterizing thyroid nodules (12). Tissue stiffness is evaluated by shear-wave elastography (SWE), in which elasticity is quantified, with higher values indicating increased stiffness (13). CEUS involves intravenous contrast agent administration to highlight the macrovascular and microvascular structures (14).

Previous studies have used US, clinical, and postoperative pathological data to develop prediction models for CLNM; however, these models are not suitable for preoperative application, and do not include MULTI-US features (15-17). A previous study combined MULTI-US features of PTC and US features of lateral lymph nodes to predict CLNM. However, as CLNM typically occurs before lateral involvement, the model is not suitable for the early prediction of CLNM (18).

In this study, we combined CON-US, CEUS, and SWE features to explore the potential of primary tumor imaging in predicting the risk of CLNM. We also developed a nomogram of the superior model to quantify CLNM risk preoperatively. We present this article in accordance with the TRIPOD reporting checklist (available at https://qims.amegroups.com/article/view/10.21037/qims-2025-1130/rc).

Methods

Patients

The study was conducted in accordance with the Declaration of Helsinki and its subsequent amendments. The study was approved by the Institutional Review Board of the West China Hospital, Sichuan University, Chengdu, China (No. 20242500), and the requirement of individual consent for this retrospective analysis was waived.

Consecutive patients who underwent surgical treatment and were pathologically diagnosed with PTC at West China Hospital, Sichuan University between January 2022 and January 2023 were enrolled. Patients were included in the study if they met the following inclusion criteria: (I) had undergone lobectomy or total thyroidectomy with central and/or lateral lymph node dissection at our institution; (II) had postoperative histopathology results confirming PTC; (III) had preoperative CON-US, CEUS, and SWE images stored in Digital Imaging and Communications in Medicine (DICOM) format available; and (IV) had complete clinicopathological data. Patients were excluded from the study if they met any of the following exclusion criteria: (I) had received preoperative treatment such as radiofrequency ablation, microwave ablation, radiotherapy, or chemotherapy; (II) had poor-quality US images in which the target lesion could not be clearly delineated due to artifacts; and/or (III) had other concomitant malignancies.

Clinic-pathological features

Data on the patients’ clinical features, including gender, age, Hashimoto’s thyroiditis status, and surgical method, and pathological features, such as LNM status, extrathyroidal extension (ETE), and tumor size (PTMC with a maximum diameter of 1 cm), were collected. Histopathological examination served as the reference standard. Lesion specimens were obtained through surgical excision, and the diagnosis of PTC and the evaluation of lymph node status were confirmed according to the World Health Organization classification of tumors (19).

US image acquisition

All patients underwent US examinations using the Aixplorer system (SuperSonic Imagine, Aix-en-Provence, France) equipped with a 5–14-MHz linear transducer. For PTC cases, CON-US, CEUS, and SWE were performed, acquiring representative images along the maximal longitudinal and transverse axes of the tumor. Dynamic scans covered the entire thyroid gland. CEUS was conducted with a real-time reverse pulse imaging technique at a low mechanical index of 0.06, following the intravenous administration of SonoVue (1.0–2.0 mL, Bracco, Milan, Italy) and a 5-mL saline flush. The imaging timer was triggered at the completion of contrast injection. SWE provided real-time elasticity measurements expressed as Young’s modulus (kPa) and displayed as color-coded stiffness maps. To minimize the influence of arterial pulsation, longitudinal sections were preferred. The probe was applied gently while patients held their breath. After freezing the elasticity map, the stiffest portion of the lesion was visually identified, and a circular region of interest (Q-box^®, SuperSonic Imagine) with a diameter of 1–2 mm was delineated for quantitative analysis.

MULTI-US features

The images were independently reviewed by two radiologists with at least 5 years of experience in US and at least 2 years of experience in EUS and CEUS, who were blinded to the clinical information and pathological results of the patients. Interobserver agreement on the US imaging features was evaluated by the kappa value. If a consensus could not be reached, the issue was resolved by a senior radiologist with 20 years of experience in head and neck US and 10 years of experience in CEUS and EUS. If there were multifocal lesions (≥2) within the same lobe, different lobes, or the thyroid isthmus that exhibited suspicious malignant features on US, the largest lesion was selected for the statistical analysis.

The CON-US features (4) included the nodular size (largest diameter), focality (unifocal, or multifocal), shape (taller-than-wide, or wider-than-tall), echogenicity (hypoechoic, which was defined as an echogenicity lower than that of the anterior cervical muscles, or isoechoic, which was defined as echogenicity comparable to that of the normal thyroid parenchyma), echogenicity pattern (homogeneous, which was defined as a uniform and consistent distribution and intensity, or heterogeneous, which was defined as a mixed, irregular, or variable spatial distribution in the nodule, with marked differences in echogenicity among different regions), margin (smooth or ill-defined, lobulated or irregular), composition (solid, mixed, or cystic), calcification type (20) (none, microcalcifications, which were defined as calcifications less than three tiny punctate bright echoes of 1 mm or less, with or without posterior acoustic shadowing, clustered microcalcifications, which were defined as the presence of three or more microcalcifications within an area of 1 mm², or macrocalcifications, which were defined as those larger than 1 mm), and extra thyroidal extension status (negative, or positive). Adler grade (21,22) was used to classify intralesional vascularity as follows: Grade 0, no detectable flow; Grade 1, one to two small flow signals (<1 mm); Grade 2, three to four signals or the presence of a main vessel; and Grade 3, ≥5 signals or at least two major vessels within the lesion.

Suspicious LNMs were defined as lymph nodes exhibiting at least two of the following features (23): (I) microcalcifications; (II) a cystic appearance; (III) diffuse or focal hyper-echogenicity; (IV) a rounded shape with a long-to-short axis ratio <2; (V) a peripheral vascular pattern, which was defined as flow signal along the periphery or capsular portion of the lymph node; and/or (F) irregular margins with loss of the hilum.

The CEUS features were evaluated in relation to the enhancement pattern (homogeneous or heterogeneous), intensity (hypo-, iso-, or hyper-enhancement), and perfusion dynamics (centripetal, centrifugal, or diffuse distribution). Homogeneous lesions were defined as those in which uniform and diffuse enhancement was observed throughout the entire lesion at peak intensity, regardless of the degree of enhancement. Heterogeneous lesions were defined as those exhibiting only partial enhancement. Centripetal or centrifugal enhancement was characterized by the expansion of microbubbles either from the margins toward the center or from the center toward the margins of the nodule. Nodules were classified as hypo-, iso-, or hyper-enhancing based on internal echogenicity that was decreased, equivalent to, or increased compared to the surrounding thyroid parenchyma (14).

The SWE features were generated by software for each Q-box, and the elasticity values, including the elasticity mean (E-mean), elasticity minimum (E-min), elasticity maximum (E-max), and the elasticity tumor-to-adjacent tissue ratio (E-ratio), were recorded in kPa.

Statistical analysis

The statistical analyses were performed using the Statistical Package for the Social Sciences software (ver. 29.0; IBM Corp., Armonk, NY, USA) and R software (ver. 4.3.2; The R Foundation for Statistical Computing, Vienna, Austria). The categorical variables are summarized as the count with the percentage, while the continuous variables are expressed as the median with the interquartile range. Associations between the categorical variables were examined using the Pearson’s Chi-squared test or Fisher’s exact test, as appropriate. The continuous variables were compared using the independent-sample t-test for normally distributed data and the Wilcoxon rank-sum test for non-normally distributed data. Multivariable logistic regression was applied to identify risk factors and to construct a predictive model for CLNM. A nomogram was subsequently developed based on this model. Model performance was assessed by receiver operating characteristic (ROC) curve analysis, and the area under the curve (AUC) values were calculated and compared using the DeLong test. Calibration was evaluated using calibration plots and the Hosmer-Lemeshow goodness-of-fit test. A decision curve analysis (DCA) was conducted to assess the clinical benefit of the model. Interobserver agreement in image feature interpretation was assessed using Cohen’s κ coefficient, where κ values <0.2 indicated poor agreement, 0.2–0.4 indicated fair agreement, 0.41–0.6 indicated moderate agreement, 0.61–0.8 indicated good agreement, and 0.81–1.0 indicated almost perfect agreement. Statistical significance was set at a two-sided P<0.05.

Results

Analysis of clinic-pathological features

The study flowchart is shown in Figure 1. A total of 764 patients with 764 PTCs were enrolled in the study. The patients had an average age of 41.1±11.0 years, and a male-to-female ratio of 3:1. CLNMs were observed in 410 of the 764 (53.7%) patients. All the data were randomly divided into training (n=534) and validation (n=230) sets at a ratio of 7:3. Of the patients, 55.2% (295/534) in the training set and 50.0% (115/230) in test set had CLNM, respectively. There were no significant differences in the clinical and pathological features between the two groups (all P>0.05; Table 1).

Figure 1 Flowchart of patient enrollment. CLNM, central lymph node metastasis; PTC, papillary thyroid carcinoma.

Analysis of the MULTI-US features

The results of the univariable analyses of the MULTI-US (CON-US, CEUS, and SWE) features for CLNM are set out in Table 2. In relation to the CON-US features, the largest diameter (P<0.001), focality (P<0.001), margin (P=0.002), calcifications (P<0.001), ETE (P<0.001), and Adler grade (P=0.015) differed significantly between the positive and negative CLNM groups. In the training set, the PTCs presenting with a large size, multifocality (51.5%, 152/295), a lobulated or irregular margin (83.4%, 246/295), clustered microcalcifications (23.1%, 68/295) or macrocalcifications (18.0%, 53/295), positive ETE (88.5%, 261/295), and a high Adler grade were mostly CLNMs. In relation to the CEUS features, hypo-enhancement (74.6%, 220/295, P<0.001) and heterogenous enhancement (89.2%, 263/295, P=0.001) were found to be associated with CLNM. In relation to the SWE values, higher E-max (P<0.001), E-mean (P<0.001), and E-ratio (P<0.001) values were significantly correlated with the occurrence of CLNM.

Risk factors of CLNM

The MULTI-US model, developed by multivariate logistic regression, included six significant independent predictors for CLNM. As the forest plot (Figure 2) for the US features shows, compared to those with no calcification, the PTCs presenting with macrocalcifications or clustered microcalcifications were more likely to be CLNM, with odds ratios (ORs) of 5.139 [95% confidence interval (CI): 2.118–12.471; P<0.001] and 6.926 (95% CI: 2.646–18.133; P<0.001) respectively, which were higher than those of the other features. ETE and multifocality were independent risk factors for CLNM, with ORs of 2.175 (95% CI: 1.317–3.583; P=0.002) and 2.040 (95% CI: 1.356–3.068; P<0.001), respectively. Among the CEUS features, enhanced intensity was an independent factor for diagnosing CLNM. Hypo-enhancement was more likely to indicate CLNM (OR: 3.405, 95% CI: 1.202–8.898; P=0.012) than iso-enhancement and hyper-enhancement. In relation to the SWE features, only a large E-max value (OR: 1.097, 95% CI: 1.042–1.153; P=0.006) was identified as an independent risk factor for CLNM.

Figure 2 Forest plot showing the multivariable analyses of the multimodal ultrasound features in the training set. CI, confidence interval; E-mean, elasticity mean value; E-ratio, elasticity ratio value; E-max, elasticity maximum value; ETE, extra thyroidal extension; OR, odds ratio.

Nomogram for predicting CLNM

A nomogram was constructed based on the independent predictors of the MULTI-US model (Figure 3). The MULTI-US model showed a favorable result in predicting CLNM (training set AUC: 0.780, 95% CI: 0.741–0.819, and test set AUC: 0.737, 95% CI: 0.692–0.807). The P values of the Hosmer-Lemeshow test were 0.2726 and 0.8154 for the training and test sets, respectively, indicating good model fit. Examples of the clinical applications of the nomogram are shown in Figures 4,5, respectively. We also compared the diagnostic performance of the MULTI-US model between different sizes in the training set. The results showed that there were no statistically significant differences between the diameter <10 mm group, which had an AUC of 0.761 (95% CI: 0.712–0.805), and the diameter ≥10 mm group, which had an AUC of 0.796 (95% CI: 0.731–0.851) (P=0.410) (Figure S1).

Figure 3 The performance of the MULTI-US model. (A) Nomogram of the MULTI-US model for predicting central lymph node metastasis in the training set. (B) Receiver operating curve and calibration curve of the MULTI-US model in the training set. At the optimal cut-off value of 0.575, the test demonstrated a specificity of 0.774 and a sensitivity of 0.668. (C) Receiver operating curve and calibration curve of the MULTI-US model in the test set. At the optimal cut-off value of 0.591, the test demonstrated a specificity of 0.696 and a sensitivity of 0.661. E-mean, elasticity mean value; E-ratio, elasticity ratio value; E-max, elasticity maximum value; MULTI-US, multiparametric ultrasound.

Figure 4 An example of using the nomogram to evaluate the individual risk of central lymph node metastasis. (A) The total score calculated by the nomogram for a papillary thyroid carcinoma patient, whose pathology confirmed negative central lymph node metastasis, was 372, and the predicted risk probability was low at 0.141. (B) The B-mode showed a tumor with a 9 mm diameter, solidity, hypo-echogenicity, a taller-than-wide shape, an irregular margin, micro-calcification, unifocality, and no extra thyroidal extension. (C) Color Doppler imaging showed Adler grade 1. (D) Shear-wave elastography showed an E-max of 21.3 kPa, an E-min of 7.0 kPa, an E-mean of 12.9 kPa, and an E-ratio of 1.5. (E) Contrast-enhanced ultrasound showed scattered, homogeneous, and iso-enhancement. **, P<0.01; ***, P<0.001. E-mean, elasticity mean value; E-ratio, elasticity ratio value; E-max, elasticity maximum value.

Figure 5 An example of using the nomogram to evaluate the individual risk of central lymph node metastasis. (A) The total score calculated by the nomogram for a papillary thyroid carcinoma patient, whose pathology confirmed positive central lymph node metastasis, was 484, and the predicted risk probability was high at 0.826. (B) The B-mode showed a tumor with a 16 mm diameter, solidity, hypo-echogenicity, a taller-than-wide shape, a lobulated margin, clustered micro-calcifications, unifocality, and extra thyroidal extension. (C) Color Doppler imaging showed Adler grade 1. (D) Shear-wave elastography showed an Emax of 35.7 kPa, an E-min of 13.9 kPa, an E-mean of 23.2 kPa, and an E-ratio of 1.8. (E) Contrast-enhanced ultrasound showed centripetal, heterogeneous and hypo-enhancement. **, P<0.01; ***, P<0.001. E-mean, elasticity mean value; E-ratio, elasticity ratio value; E-max, elasticity maximum value.

Comparison of different models

Further, the diagnostic efficacy and clinical utility of the MULTI-US model were evaluated in comparison with different imaging modalities, including the CON-US, CEUS, and EUS models (Table 3 and Figure 6). The AUC value of the MULTI-US model differed significantly from the AUC values of the CON-US (P=0.008), CEUS (P<0.001), and US (P<0.001) models. The specificity, sensitivity, accuracy, and precision of the MULTI-US model were 0.774, 0.668, 0.715 and 0.784 in the training set, respectively, and were superior and more stable than those of the other single-modal models. A DCA was conducted to assess the utility of the four predictive models by calculating the net benefit at various probability thresholds. Among the four models, the MULTI-US model achieved the highest clinical net benefit across threshold probabilities ranging from 0.2 to 0.9.

Figure 6 Comparisons of different models. (A) Comparison of the decision curves of different models for predicting central lymph node metastasis in the training set. (B) Comparison of the receiver operating curves of different models for predicting central lymph node metastasis in the training set. CEUS, contrast-enhanced ultrasound; CON-US, conventional ultrasound; DCA, decision curve analysis; EUS, elastography ultrasound; MULTI-US, multiparametric ultrasound.

Discussion

The accurate preoperative assessment of central lymph node status is crucial for the optimal management of PTC. While US is the recommended imaging modality for evaluating cervical lymph nodes in PTC, its diagnostic performance for cervical lymph nodes remains limited. In this cohort, CLNM was pathologically confirmed in 410 of the 764 patients (53.7%); however, CLNM was only suspected in 195 patients (25.5%) based on the preoperative cervical lymph node imaging. Thus, an effective noninvasive approach needs to be developed for CLNM assessment. In this retrospective single-center study, we used preoperative MULTI-US features to develop a nomogram for intuitive and quantitative CLNM risk prediction. Multivariate logistic regression identified significant independent predictors of CLNM, including ETE, multifocality, macrocalcifications or clustered microcalcifications, hypo-enhancement, and the E-max value (P<0.05).

Most previous studies have shown that microcalcifications are indicative of PTC; however, the clinical significance of macrocalcifications continues to be debated (4,24). Some studies have investigated the relationship between macrocalcifications and malignancy, suggesting that the risk of malignancy may be influenced by the echogenicity and internal composition of the calcifications (25,26). However, as our study exclusively included malignant cases, we were unable to evaluate the differential implications of various calcification types between the benign and malignant nodules. Therefore, the aforementioned findings can only be referenced theoretically in the context of our study.

The correlation between microcalcifications in PTC and the risk of CLNM has been explored; however, research on macrocalcifications specifically is limited. Liu et al. suggested that the presence of microcalcifications in US imaging was a risk factor for CLNM (27). Tian et al. also suggested that a large number of punctate strong echoes in the thyroid nodule, especially the presence of a diffuse distribution of microcalcifications, has a high predictive value for CLNM (28). Zou et al. found that the presence of calcifications in CT images of PTC also constitute a risk factor for CLNM (29). These are consistent with our findings, but unlike previous studies we further classified the large number of punctate calcifications into clustered microcalcifications based on the distribution pattern.

Our results suggested that both clustered microcalcifications and macrocalcifications were also risk factors of CLNM. One possible reason for the complex clinical significance of macrocalcifications is that their association with malignancy may vary depending on the diverse sonographic morphological characteristics they present. This may be related to the subjective observation of echogenic foci by different sonographers. Overlapping microcalcifications can appear as gross calcifications to the human eye, and no studies or guidelines have addressed how to differentiate between the two. Some studies have reported that there is a considerable risk of underdiagnosis in cases involving calcifications, especially microcalcifications (<2 mm) (25). Wang et al. used deep learning to measure calcifications in thyroid nodules in US images (30). Given the limitations of our retrospective study, we failed to further quantitatively assess the distribution of calcifications in tumors. Future studies should seek to investigate the potential relationship between calcifications and highly aggressive PTCs.

Our study showed that ETE detected by US was significantly associated with CLNM. This may be because tumors infiltrating the thyroid peritumoral region are more likely to metastasize along the lymphatic tissues surrounding the peritumoral region to the lymph nodes in the neck (27). In our study, a bar chart showed the distribution of postoperative pathology-confirmed ETE (none, minimal, and gross) between the CLNM-negative and CLNM-positive groups (Figure S2). The degree of ETE was compared using the two-sided Pearson’s Chi-squared test, and the proportions of minimal and gross ETE were significantly higher in the CLNM-positive group (P=0.001). Previous guidelines have listed ETE as a risk factor for tumor staging and surgical planning (31). Unfortunately, existing imaging examinations cannot provide a reliable diagnosis of ETE (32,33).

Angiogenesis facilitates tumor expansion by supplying oxygen and nutrients while simultaneously creating abnormal, leaky vasculature that enhances tumor cell intravasation (34). This aberrant vascular microenvironment not only accelerates proliferation but also fosters local invasion, thereby increasing tumor aggressiveness and providing routes for dissemination to regional lymph nodes. Chen et al. suggested that CEUS features presenting with hyper- and centripetal enhancement were related to PTC aggressiveness (35). However, our study showed that hypo-enhancement was a risk factor for CLNM. The inconsistency between the studies might be due to the interference of tumors with necrotic liquefaction and coarse calcifications (36). The microvasculature of the primary tumor or lymph nodes evaluated by qualitative and quantitative CEUS could provide further insights into the risk factors for CLNM. In relation to the SWE features, we found that a higher E-max value was associated with a higher risk of CLNM. This is consistent with findings that high SWE values indicate tumor invasion and progression (37,38), which is often accompanied by cellular proliferation and fibrosis, and which causes tumors to appear to have a stiff texture (13).

Our results showed that the MULTI-US model achieved superior diagnostic performance and served as a reliable clinical tool for predicting CLNM in the PTC patients. Incorporating MULTI-US risk factors into a user-friendly nomogram enables individualized preoperative CLNM prediction. Patients with high scores may be suitable candidates for CLND. The DCA indicated that using the MULTI-US nomogram to guide CLND decisions would improve clinical outcomes over current practice patterns, particularly when surgeons consider a 20–90% risk probability as the threshold for intervention. Notably, the threshold range should be considered a suggested reference for clinical decision-making rather than a definitive diagnostic cutoff. In actual clinical practice, the choice of any threshold should take into account the current risk of surgical complications, including recurrent laryngeal nerve injury (2–4%) and hypocalcemia (10–15%) (39). Current guidelines recommend surgical resection or active surveillance for PTMCs (diameter <10 mm) (5). The diagnostic performance of the MULTI-US model did not differ significantly between tumors with diameters <10 mm and those with diameters ≥10 mm. For the active surveillance of low-risk PTMCs, the MULTI-US model may be used to provide management recommendations.

This study had several limitations. First, due to its retrospective nature, the risk of selection bias cannot be excluded, as the enrolled patient cohort may not fully represent the broader PTC population. We did not include preoperative molecular diagnostic information obtained from FNAB. However, mutations in genes such as BRAF, RAS, and TERT have been found to be highly associated with the risk of CLNM (40). Although the interobserver agreement for the MULTI-US features of PTC reviewed by the radiologists with different levels of experience was excellent (κ value =0.84), the performance of the nomogram was influenced by the subjective evaluation of the preoperative MULTI-US images. These factors might have influenced the model’s performance, and could have led to an overestimation of sensitivity or an underestimation of specificity. Second, all the data were collected from a single institution. Variability in patient demographics, imaging equipment, scanning protocols, and operator experience across institutions may limit the generalizability of the model. Therefore, the predictive performance reported in this study primarily reflects our institutional setting and may not be directly applicable to other clinical environments.

Conclusions

We developed a nomogram based on MULTI-US features to preoperatively predict CLNM in PTC. The model showed favorable diagnostic performance in our cohort. However, given the limitations of the retrospective, single-center design, further validation is essential. Future research should focus on conducting prospective, multicenter studies with standardized US acquisition and reporting protocols to enhance model reproducibility and generalizability. Additionally, expanding the analysis to incorporate quantitative and spatial characteristics of features such as calcifications may provide further insights into tumor aggressiveness and metastatic potential. These efforts will help refine predictive tools for broader clinical application.

Acknowledgments

None.

Footnote

Reporting Checklist: The authors have completed the TRIPOD reporting checklist. Available at https://qims.amegroups.com/article/view/10.21037/qims-2025-1130/rc

Data Sharing Statement: Available at https://qims.amegroups.com/article/view/10.21037/qims-2025-1130/dss

Funding: This work was supported by the 1·3·5 Projects for Artificial Intelligence, West China Hospital, Sichuan University (Nos. ZYA124035 and ZYA124056).

Conflicts of Interest: All authors have completed the ICMJE uniform disclosure form (available at https://qims.amegroups.com/article/view/10.21037/qims-2025-1130/coif). All authors report that this work was supported by the 1·3·5 Projects for Artificial Intelligence, West China Hospital, Sichuan University (Nos. ZYA124035 and ZYA124056). The authors have no other 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 and its subsequent amendments. The study was approved by the institutional review board of the West China Hospital, Sichuan University, Chengdu, China (No. 20242500) and individual consent for this retrospective analysis was waived.

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. Miranda-Filho A, Lortet-Tieulent J, Bray F, Cao B, Franceschi S, Vaccarella S, Dal Maso L. Thyroid cancer incidence trends by histology in 25 countries: a population-based study. Lancet Diabetes Endocrinol 2021;9:225-34. [Crossref] [PubMed]
2. Stack BC Jr, Ferris RL, Goldenberg D, Haymart M, Shaha A, Sheth S, Sosa JA, Tufano RPAmerican Thyroid Association Surgical Affairs Committee. American Thyroid Association consensus review and statement regarding the anatomy, terminology, and rationale for lateral neck dissection in differentiated thyroid cancer. Thyroid 2012;22:501-8. [Crossref] [PubMed]
3. Filetti S, Durante C, Hartl D, Leboulleux S, Locati LD, Newbold K, Papotti MG, Berruti AESMO Guidelines Committee. Thyroid cancer: ESMO Clinical Practice Guidelines for diagnosis, treatment and follow-up†. Ann Oncol 2019;30:1856-83. [Crossref] [PubMed]
4. Haugen BR, Alexander EK, Bible KC, Doherty GM, Mandel SJ, Nikiforov YE, Pacini F, Randolph GW, Sawka AM, Schlumberger M, Schuff KG, Sherman SI, Sosa JA, Steward DL, Tuttle RM, Wartofsky L. 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]
5. Liu W, Yan X, Cheng R. The active surveillance management approach for patients with low risk papillary thyroid microcarcinomas: is China ready? Cancer Biol Med 2021;19:619-34. [Crossref] [PubMed]
6. Grani G, Sponziello M, Pecce V, Ramundo V, Durante C. Contemporary Thyroid Nodule Evaluation and Management. J Clin Endocrinol Metab 2020;105:2869-83. [Crossref] [PubMed]
7. Fiorentino V. Dell' Aquila M, Musarra T, Martini M, Capodimonti S, Fadda G, Curatolo M, Traini E, Raffaelli M, Lombardi CP, Pontecorvi A, Larocca LM, Pantanowitz L, Rossi ED. The Role of Cytology in the Diagnosis of Subcentimeter Thyroid Lesions. Diagnostics (Basel) 2021;11:1043. [Crossref] [PubMed]
8. Agarwal S, Chand G, Jaiswal S, Mishra A, Agarwal G, Agarwal A, Verma AK, Mishra SK. Pattern and risk factors of central compartment lymph node metastasis in papillary thyroid cancer: a prospective study from an endocrine surgery centre. J Thyroid Res 2012;2012:436243. [Crossref] [PubMed]
9. O'Connell K, Yen TW, Quiroz F, Evans DB, Wang TS. 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]
10. Qu H, Sun GR, Liu Y, He QS. Clinical risk factors for central lymph node metastasis in papillary thyroid carcinoma: a systematic review and meta-analysis. Clin Endocrinol (Oxf) 2015;83:124-32. [Crossref] [PubMed]
11. Yan B, Hou Y, Chen D, He J, Jiang Y. Risk factors for contralateral central lymph node metastasis in unilateral cN0 papillary thyroid carcinoma: A meta-analysis. Int J Surg 2018;59:90-8. [Crossref] [PubMed]
12. Bernet VJ, Chindris AM. Update on the Evaluation of Thyroid Nodules. J Nucl Med 2021;62:13S-9S. [Crossref] [PubMed]
13. Park AY, Kim JA, Son EJ, Youk JH. Shear-Wave Elastography for Papillary Thyroid Carcinoma can Improve Prediction of Cervical Lymph Node Metastasis. Ann Surg Oncol 2016;23:722-9. [Crossref] [PubMed]
14. Sidhu PS, Cantisani V, Dietrich CF, Gilja OH, Saftoiu A, Bartels E, et al. The EFSUMB Guidelines and Recommendations for the Clinical Practice of Contrast-Enhanced Ultrasound (CEUS) in Non-Hepatic Applications: Update 2017 (Long Version). Ultraschall Med 2018;39:e2-e44. [Crossref] [PubMed]
15. Du J, Yang Q, Sun Y, Shi P, Xu H, Chen X, Dong T, Shi W, Wang Y, Song Z, Shang X, Tian X. Risk factors for central lymph node metastasis in patients with papillary thyroid carcinoma: a retrospective study. Front Endocrinol (Lausanne) 2023;14:1288527. [Crossref] [PubMed]
16. Li WH, Yu WY, Du JR, Teng DK, Lin YQ, Sui GQ, Wang H. Nomogram prediction for cervical lymph node metastasis in multifocal papillary thyroid microcarcinoma. Front Endocrinol (Lausanne) 2023;14:1140360. [Crossref] [PubMed]
17. Wang R, Tang Z, Wu Z, Xiao Y, Li J, Zhu J, Zhang X, Ming J. Construction and validation of nomograms to reduce completion thyroidectomy by predicting lymph node metastasis in low-risk papillary thyroid carcinoma. Eur J Surg Oncol 2023;49:1395-404. [Crossref] [PubMed]
18. Dai Q, Liu D, Tao Y, Ding C, Li S, Zhao C, Wang Z, Tao Y, Tian J, Leng X. Nomograms based on preoperative multimodal ultrasound of papillary thyroid carcinoma for predicting central lymph node metastasis. Eur Radiol 2022;32:4596-608. [Crossref] [PubMed]
19. Jung CK, Bychkov A, Kakudo K. Update from the 2022 World Health Organization Classification of Thyroid Tumors: A Standardized Diagnostic Approach. Endocrinol Metab (Seoul) 2022;37:703-18. [Crossref] [PubMed]
20. Azam S, Eriksson M, Sjölander A, Gabrielson M, Hellgren R, Czene K, Hall P. Mammographic microcalcifications and risk of breast cancer. Br J Cancer 2021;125:759-65. [Crossref] [PubMed]
21. Adler DD, Carson PL, Rubin JM, Quinn-Reid D. Doppler ultrasound color flow imaging in the study of breast cancer: preliminary findings. Ultrasound Med Biol 1990;16:553-9. [Crossref] [PubMed]
22. Ma JJ, Ding H, Xu BH, Xu C, Song LJ, Huang BJ, Wang WP. Diagnostic performances of various gray-scale, color Doppler, and contrast-enhanced ultrasonography findings in predicting malignant thyroid nodules. Thyroid 2014;24:355-63. [Crossref] [PubMed]
23. Ryu KH, Lee KH, Ryu J, Baek HJ, Kim SJ, Jung HK, Kim SM. Cervical Lymph Node Imaging Reporting and Data System for Ultrasound of Cervical Lymphadenopathy: A Pilot Study. AJR Am J Roentgenol 2016;206:1286-91. [Crossref] [PubMed]
24. Kobaly K, Kim CS, Langer JE, Mandel SJ. Macrocalcifications Do Not Alter Malignancy Risk Within the American Thyroid Association Sonographic Pattern System When Present in Non-High Suspicion Thyroid Nodules. Thyroid 2021;31:1542-8. [Crossref] [PubMed]
25. Shin HS, Na DG, Paik W, Yoon SJ, Gwon HY, Noh BJ, Kim WJ. Malignancy Risk Stratification of Thyroid Nodules with Macrocalcification and Rim Calcification Based on Ultrasound Patterns. Korean J Radiol 2021;22:663-71. [Crossref] [PubMed]
26. Ye M, Wu S, Zhou Q, Wang F, Chen X, Gong X, Wu W. Association between macrocalcification and papillary thyroid carcinoma and corresponding valuable diagnostic tool: retrospective study. World J Surg Oncol 2023;21:149. [Crossref] [PubMed]
27. Liu C, Xiao C, Chen J, Li X, Feng Z, Gao Q, Liu Z. Risk factor analysis for predicting cervical lymph node metastasis in papillary thyroid carcinoma: a study of 966 patients. BMC Cancer 2019;19:622. [Crossref] [PubMed]
28. Tian X, Song Q, Xie F, Ren L, Zhang Y, Tang J, Zhang Y, Jin Z, Zhu Y, Zhang M, Luo Y. Papillary thyroid carcinoma: an ultrasound-based nomogram improves the prediction of lymph node metastases in the central compartment. Eur Radiol 2020;30:5881-93. [Crossref] [PubMed]
29. Zou Y, Shi Y, Bi H, Tan J, Guo Q, Qin Y, Lu X, Ma X, Yang S, Liu J. A nomogram for risk stratification of central cervical lymph node metastasis in patients with papillary thyroid carcinoma. Quant Imaging Med Surg 2024;14:5084-98. [Crossref] [PubMed]
30. Wang J, Dong C, Zhang YZ, Wang L, Yuan X, He M, Xu S, Zhou Q, Jiang J. A novel approach to quantify calcifications of thyroid nodules in US images based on deep learning: predicting the risk of cervical lymph node metastasis in papillary thyroid cancer patients. Eur Radiol 2023;33:9347-56. [Crossref] [PubMed]
31. Gulec SA, Ahuja S, Avram AM, Bernet VJ, Bourguet P, Draganescu C, Elisei R, Giovanella L, Grant F, Greenspan B, Hegedüs L, Jonklaas J, Kloos RT, Luster M, Oyen WJG, Smit J, Tuttle RM. A Joint Statement from the American Thyroid Association, the European Association of Nuclear Medicine, the European Thyroid Association, the Society of Nuclear Medicine and Molecular Imaging on Current Diagnostic and Theranostic Approaches in the Management of Thyroid Cancer. Thyroid 2021;31:1009-19. [Crossref] [PubMed]
32. Lamartina L, Bidault S, Hadoux J, Guerlain J, Girard E, Breuskin I, Attard M, Suciu V, Baudin E, Al Ghuzlan A, Leboulleux S, Hartl D. Can preoperative ultrasound predict extrathyroidal extension of differentiated thyroid cancer? Eur J Endocrinol 2021;185:13-22. [Crossref] [PubMed]
33. Chung SR, Baek JH, Choi YJ, Sung TY, Song DE, Kim TY, Lee JH. Sonographic Assessment of the Extent of Extrathyroidal Extension in Thyroid Cancer. Korean J Radiol 2020;21:1187-95. [Crossref] [PubMed]
34. Radzina M, Ratniece M, Putrins DS, Saule L, Cantisani V. Performance of Contrast-Enhanced Ultrasound in Thyroid Nodules: Review of Current State and Future Perspectives. Cancers (Basel) 2021;13:5469. [Crossref] [PubMed]
35. Chen L, Chen L, Liang Z, Shao Y, Sun X, Liu J. Value of Contrast-Enhanced Ultrasound in the Preoperative Evaluation of Papillary Thyroid Carcinoma Invasiveness. Front Oncol 2021;11:795302. [Crossref] [PubMed]
36. Wu Y, Zhou C, Shi B, Zeng Z, Wu X, Liu J. Systematic review and meta-analysis: diagnostic value of different ultrasound for benign and malignant thyroid nodules. Gland Surg 2022;11:1067-77. [Crossref] [PubMed]
37. Jiang M, Li C, Tang S, Lv W, Yi A, Wang B, Yu S, Cui X, Dietrich CF. Nomogram Based on Shear-Wave Elastography Radiomics Can Improve Preoperative Cervical Lymph Node Staging for Papillary Thyroid Carcinoma. Thyroid 2020;30:885-97. [Crossref] [PubMed]
38. Wan F, He W, Zhang W, Zhang H, Zhang Y, Guang Y. Application of decision tree algorithms to predict central lymph node metastasis in well-differentiated papillary thyroid carcinoma based on multimodal ultrasound parameters: a retrospective study. Quant Imaging Med Surg 2023;13:2081-97. [Crossref] [PubMed]
39. Lang BH, Ng SH, Lau LL, Cowling BJ, Wong KP, Wan KY. A systematic review and meta-analysis of prophylactic central neck dissection on short-term locoregional recurrence in papillary thyroid carcinoma after total thyroidectomy. Thyroid 2013;23:1087-98. [Crossref] [PubMed]
40. Patel J, Klopper J, Cottrill EE. Molecular diagnostics in the evaluation of thyroid nodules: Current use and prospective opportunities. Front Endocrinol (Lausanne) 2023;14:1101410. [Crossref] [PubMed]