Value of Ultrasound Combined with Immunohistochemistry Evaluation of Central Lymph Node Metastasis for the Prognosis of Papillary Thyroid Carcinoma

Interpretable machine learning models for predicting skip metastasis in cN0 papillary thyroid cancer based on clinicopathological and elastography radiomics features

Background

Skip lymph node metastasis (SLNM) in papillary thyroid cancer (PTC) involves cancer cells bypassing central nodes to directly metastasize to lateral nodes, often undetected by standard preoperative ultrasonography. Although multiple models exist to identify SLNM, they are inadequate for clinically node-negative (cN0) patients, resulting in underestimated metastatic risks and compromised treatment effectiveness. Our study aims to develop and validate a machine learning (ML) model that combines elastography radiomics with clinicopathological data to predict pre-surgical SLNM risk in cN0 PTC patients with increased risk of lymph node metastasis (LNM), improving their treatment strategies.

Methods

Our study conducted a retrospective analysis of 485 newly diagnosed primary PTC patients, divided into training and external validation cohorts. Patients were categorized into SLNM and non-SLNM groups based on follow-up outcomes and postoperative pathology. We collected preoperative clinicopathological data and extracted, standardized radiomics features from elastography imaging to develop various ML models. These models were internally validated using radiomics and clinicopathological data, with the optimal model's feature importance analyzed through the Shapley Additive Explanations (SHAP) approach and subsequently externally validated.

Results

In our study of 485 patients, 67 (13.8%) exhibited SLNM. The extreme gradient boosting (XGBoost) model, integrating elastography radiomics with clinicopathological data, demonstrated superior performance in both internal and external validations. SHAP analysis identified five key determinants of SLNM: three radiomics features from elastography images, one clinical variable, and one pathological variable.

Conclusion

Our evaluation highlights the XGBoost model, which integrates elastography radiomics and clinicopathological data, as the most effective ML approach for the prediction of SLNM in cN0 PTC patients with increased risk of LNM. This innovative model significantly enhances the accuracy of risk assessments for SLNM, enabling personalized treatments that could reduce postoperative metastases in these patients.

Predictive Values of Clinical Features and Multimodal Ultrasound for Central Lymph Node Metastases in Papillary Thyroid Carcinoma

Papillary thyroid carcinoma (PTC), the predominant pathological type among thyroid malignancies, is responsible for the sharp increase in thyroid cancer. Although PTC is an indolent tumor with good prognosis, 60-70% of patients still have early cervical lymph node metastasis, typically in the central compartment. Whether there is central lymph node metastasis (CLNM) or not directly affects the formulation of preoperative surgical procedures, given that such metastases have been tied to compromised overall survival and local recurrence. However, detecting CLNM before operation can be challenging due to the limited sensitivity of preoperative approaches. Prophylactic central lymph node dissection (PCLND) in the absence of clinical evidence of CLNM poses additional surgical risks. This study aims to provide a comprehensive review of the risk factors related to CLNM in PTC patients. A key focus is on utilizing multimodal ultrasound (US) for accurate prognosis of preoperative CLNM and to highlight the distinctive role of US-based characteristics for predicting CLNM.

Quantitative parameters of contrast-enhanced ultrasound effectively promote the prediction of cervical lymph node metastasis in papillary thyroid carcinoma

Papillary thyroid carcinoma (PTC), the most common endocrine tumor, often spreads to cervical lymph nodes metastasis (CLNM). Preoperative diagnosis of CLNM is important when selecting surgical strategies. Therefore, we aimed to explore the effectiveness of quantitative parameters of contrast-enhanced ultrasound (CEUS) in predicting CLNM in PTC. We retrospectively analyzed 193 patients with PTC undergoing conventional ultrasound (CUS) and CEUS. The CUS features and quantitative parameters of CEUS were evaluated according to PTC size ≤ 10 or > 10 mm, using pathology as the gold standard. For the PTC ≤ 10 mm, microcalcification and multifocality were significantly different between the CLNM (+) and CLNM (-) groups (both P < 0.05). For the PTC > 10 mm, statistical significance was noted between the two groups with respect to the margin, capsule contact, and multifocality (all P < 0.05). For PTC ≤ 10 mm, there was no significant difference between the CLNM (+) and CLNM (-) groups in all quantitative parameters of CEUS (all P > 0.05). However, for PTC > 10 mm, the peak intensity (PI), mean transit time, and slope were significantly associated with CLNM (all P < 0.05). Multivariate analysis showed that PI > 5.8 dB was an independent risk factor for predicting CLNM in patients with PTC > 10 mm (P < 0.05). The area under the curve of PI combined with CUS (0.831) was significantly higher than that of CUS (0.707) or PI (0.703) alone in the receiver operator characteristic curve analysis (P < 0.05). In conclusion, PI has significance in predicting CLNM for PTC > 10 mm; however, it is not helpful for PTC ≤ 10 mm.

Cervical lymph node metastasis prediction of postoperative papillary thyroid carcinoma before 131I therapy based on clinical and ultrasound characteristics

BACKGROUND

The status of lymph nodes is crucial to determine the dose of radioiodine-131(¹³¹I) for postoperative papillary thyroid carcinoma (PTC). We aimed to develop a nomogram for predicting residual and recurrent cervical lymph node metastasis (CLNM) in postoperative PTC before ¹³¹I therapy.

METHOD

Data from 612 postoperative PTC patients who underwent ¹³¹I therapy from May 2019 to December 2020 were retrospectively analyzed. Clinical and ultrasound features were collected. Univariate and multivariate logistic regression analyses were performed to determine the risk factors of CLNM. Receiver operating characteristic (ROC) analysis was used to weigh the discrimination of prediction models. To generate nomograms, models with high area under the curves (AUC) were selected. Bootstrap internal validation, calibration curves and decision curves were used to assess the prediction model's discrimination, calibration, and clinical usefulness.

RESULTS

A total of 18.79% (115/612) of postoperative PTC patients had CLNM. Univariate logistic regression analysis found serum thyroglobulin (Tg), serum thyroglobulin antibodies (TgAb), overall ultrasound diagnosis and seven ultrasound features (aspect transverse ratio, cystic change, microcalcification, mass hyperecho, echogenicity, lymphatic hilum structure and vascularity) were significantly associated with CLNM. Multivariate analysis revealed higher Tg, higher TgAb, positive overall ultrasound and ultrasound features such as aspect transverse ratio ≥ 2, microcalcification, heterogeneous echogenicity, absence of lymphatic hilum structure and abundant vascularity were independent risk factors for CLNM. ROC analysis showed the use of Tg and TgAb combined with ultrasound (AUC = 0.903 for "Tg+TgAb+Overall ultrasound" model, AUC = 0.921 for "Tg+TgAb+Seven ultrasound features" model) was superior to any single variant. Nomograms constructed for the above two models were validated internally and the C-index were 0.899 and 0.914, respectively. Calibration curves showed satisfied discrimination and calibration of the two nomograms. DCA also proved that the two nomograms were clinically useful.

CONCLUSION

Through the two accurate and easy-to-use nomograms, the possibility of CLNM can be objectively quantified before ¹³¹I therapy. Clinicians can use the nomograms to evaluate the status of lymph nodes in postoperative PTC patients and consider a higher dose of ¹³¹I for those with high scores.

Clinical Value of Ultrasonography and Serum Markers in Preoperative N Staging of Thyroid Cancer

We aimed to determine factors influencing lymph node metastasis (LNM) and develop a more effective method to assess preoperative N staging. Overall, data of 2130 patients who underwent thyroidectomy for thyroid cancer between 2018 and 2021 were retrospectively analysed. Patients were divided into groups according to pN0, pN1a, and pN1b stages. Pathology was used to analyse the correlation between preoperative serum marker indicators and LNM. Receiver operating characteristic curves were used to compare the diagnostic value of ultrasound (US) examination alone, serum thyroglobulin, age, and combined method for LNM. A significant moderate agreement was observed between preoperative US and postoperative pathology for N staging. Between the pN0 and pN1 (pN1a + pN1b) groups, the differences in free triiodothyronine, anti-thyroid peroxidase antibody, and serum thyroglobulin levels were statistically significant. Among the indicators, serum thyroglobulin was an independent predictor of LNM. The area under the receiver operating characteristic curve was 0.610 for serum thyroglobulin level for predicting LNM, 0.689 for US alone, and 0.742 for the combined method. Both preoperative US and serum thyroglobulin level provide a specific value when evaluating the N staging of thyroid cancer, and the combined method is more valuable in the diagnosis of LNM than US alone.

Contrast-enhanced ultrasound characteristics of preoperative central cervical lymph node metastasis in papillary thyroid carcinoma

This study evaluated the preoperative diagnostic value of lymph node ultrasonography in distinguishing between benign and malignant central cervical lymph nodes (CCLNs) in patients with papillary thyroid carcinoma (PTC). A total of 176 patients who had PTC with 216 CCLNs (49 benign and 155 malignant) were enrolled in this study and preoperatively imaged by ultrasonography, including conventional ultrasound (US) and contrast-enhanced ultrasound (CEUS). We evaluated the ultrasonography parameters for each lymph node. Binary logistic regression analysis indicated that multifocality of PTC and the absence of Hashimoto's thyroiditis are independent clinical features related to patients with PTC who also have malignant CCLNs. For preoperative ultrasonography features, heterogeneous enhancement and centripetal perfusion are independent ultrasonographic features to identify malignant and benign CCLNs. This study demonstrated that preoperative CEUS characteristics help to distinguish malignant CCLNs from benign CCLNs.

Analysis of the Relevance of the Ultrasonographic Features of Papillary Thyroid Carcinoma and Cervical Lymph Node Metastasis on Conventional and Contrast-Enhanced Ultrasonography

Background

Preoperative prediction of lymph node metastases has a major impact on prognosis and recurrence for patients with papillary thyroid carcinoma (PTC). Thyroid ultrasonography is the preferred inspection to guide the appropriate diagnostic procedure.

Purpose

To investigate the relationship between PTC and cervical lymph node metastasis (CLNM, including central and lateral LNM) using both conventional ultrasound (US) and contrast-enhanced ultrasound (CEUS).

Material and Methods

Our study retrospectively analyzed 379 patients diagnosed with PTC confirmed by surgical pathology at our hospital who underwent US and CEUS examinations from October 2016 to March 2021. Individuals were divided into two groups: the lymph node metastasis group and the nonmetastasis group. The relationship between US and CEUS characteristics of PTC and CLNM was analyzed. Univariate and multivariable logistic regression methods were used to identify the high-risk factors and established a nomogram to predict CLNM in PTC. Furthermore, we explore the frequency of CLNM at each nodal level in PTC patients.

Results

Univariate analysis indicated that there were significant differences in gender, age, tumor size, microcalcification, contact with the adjacent capsule, multifocality, capsule integrity and enhancement patterns in CEUS between the lymph node metastasis group and the nonmetastasis group (all P<0.05). Multivariate regression analysis showed that tumor size ≥1 cm, age ≤45 years, multifocality, and contact range of the adjacent capsule >50% were independent risk factors for CLNM in PTC, which determined the nomogram. The diagnostic model had an area under the curve (AUC) of 0.756 (95% confidence interval, 0.707-0.805). And calibration plot analysis shown that clinical utility of the nomogram. In 162 PTC patients, the metastatic rates of cervical lymph nodes at levels I-VI were 1.9%, 15.4%, 35.2%, 34.6%, 15.4%, 82.1%, and the difference was statistically significant (P<0.001).

Conclusion

Our study indicated that the characteristics of PTC on ultrasonography and CEUS can be used to predict CLNM as a useful tool. Preoperative analysis of ultrasonographical features has important value for predicting CLNM in PTCs. The risk of CLNM is greater when tumor size ≥1 cm, age ≤45 years, multifocality, contact range of the adjacent capsule >50% are present.

How to identify indolent thyroid tumors unlikely to recur and cause cancer death immediately after surgery-Risk stratification of papillary thyroid carcinoma in young patients

Current histopathological diagnosis methods cannot distinguish the two types of thyroid carcinoma: clinically significant carcinomas with a potential risk of recurrence, metastasis, and cancer death, and clinically insignificant carcinomas with a slow growth rate. Both thyroid tumors are diagnosed as "carcinoma" in current pathology practice. The clinician usually recommends surgery to the patient and the patient often accepts it because of cancer terminology. The treatment for these clinically insignificant carcinomas does not benefit the patient and negatively impacts society. The author proposed risk stratification of thyroid tumors using the growth rate (Ki-67 labeling index), which accurately differentiates four prognostically relevant risk groups based on the Ki-67 labeling index, ≥30%, ≥10 and <30%, >5 and <10%, and ≤5%. Indolent thyroid tumors with an excellent prognosis have the following four features: young age, early-stage (T1-2 M0), curatively treated, and low proliferation index (Ki-67 labeling index of ≤5%), and are unlikely to recur, metastasize, or cause cancer death. Accurate identification of these indolent tumors helps clinicians select more conservative treatments to avoid unnecessary aggressive (total thyroidectomy followed by radio-active iodine) treatments. Clinicians can alleviate the fears of patients by confirming these four features, including the low proliferation rate, in a pathology report immediately after surgery when patients are most concerned.

Identification of biomarkers associated with cervical lymph node metastasis in papillary thyroid carcinoma: Evidence from an integrated bioinformatic analysis

Integrated analysis of accumulated data is an effective way to obtain reliable potential diagnostic molecular of cervical lymph node metastases (LNM) in papillary thyroid carcinoma (PTC). The benefits of prophylactic lymph node dissection (PLND) for these clinically node-negative (cN0) patients remained considerable controversies. Hence, elucidation of the mechanisms of LNM and exploration of potential biomarkers and prognostic indicators are essential for accurate diagnosis of LNM in PTC patients. Up to date, advanced microarray and bioinformatics analysis have advanced an understanding of the molecular mechanisms of disease occurrence and development, which are necessary to explore genetic changes and identify potential diagnostic biomarkers. In present study, we performed a comprehensive analysis of the differential expression, biological functions, and interactions of LNM-related genes. Two publicly available microarray datasets GSE60542 and GSE129562 were available from Gene Expression Omnibus (GEO) database. Differentially expressed genes between clinically node-positive (cN1) and cN0 PTC samples were screened by an integrated analysis of multiple gene expression profile after gene reannotation and batch normalization. Our results identified 48 differentially expressed genes (DEGs) genetically associated with LNM in PTC patients. Gene ontology (GO) analyses revealed the changes in the modules were mostly enriched in the regulation of MHC class II receptor activity, the immune receptor activity, and the peptide antigen binding. Kyoto encyclopedia of genes and genomes (KEGG) enrichment analysis of DEGs displayed the intestinal immune network for IgA production, staphylococcus aureus infection, and cell adhesion molecules (CAMs). To screen core genes related to LNM of PTC from the protein-protein interaction network, top 10 hub genes were identified with highest scores. Our results help us understand the exact mechanisms underlying the metastasis of cervical LNM in PTC tissues and pave an avenue for the progress of precise medicine for individual patients.

Value of Combining Clinical Factors, Conventional Ultrasound, and Contrast-Enhanced Ultrasound Features in Preoperative Prediction of Central Lymph Node Metastases of Different Sized Papillary Thyroid Carcinomas

Purpose

Early and accurate preoperative diagnosis of central lymph node metastasis (CLNM) is crucial to improve surgical management of patients with clinical lymph node-negative papillary thyroid carcinoma (PTC). Towards improving diagnosis of CLNM, we assessed the value of combining preoperative clinical characteristics, conventional ultrasound, and contrast-enhanced ultrasound (CEUS) in preoperative prediction of CLNM of different sized PTCs.

Patients and Methods

Patients were divided according to tumor size: a PTC group (>10 mm) and a papillary thyroid microcarcinoma (PTMC) group (≤10 mm). We retrospectively analyzed the clinical and ultrasonographic features of 120 PTC patients and 165 PTMC patients. Multivariate logistic regression analysis was used to screen independent risk factors and establish prediction models. Receiver operating characteristic curves were used to determine the best cut-off values for continuous variables and assess the performance of prediction models.

Results

Independent risk predictors of CLNM for the PTC group were extrathyroidal extension in CEUS (OR=7.923), tumor size >14 mm (OR=5.491), and multifocality (OR=3.235). For the PTMC group, the independent risk factors were the distance from the thyroid capsule =0 mm (OR=4.629), male (OR=3.315), tumor size >5 mm (OR=3.304), and microcalcification (OR=2.560). The predictive model of combined method had better performance in predicting CLNM of PTC compared with models based on CEUS and conventional ultrasound alone (area under the curve: 0.832 vs 0.739, P=0.0011; 0.832 vs 0.678, P=0.0012). For PTMC, comparing with CEUS, the combined method and conventional ultrasound performed better than CEUS alone in predicting CLNM (area under the curve: 0.783 vs 0.636, P=0.0016; 0.738 vs 0.636, P=0.0196).

Conclusion

The predictive models of combined method obtained from significant preoperative clinical and ultrasonographic features can potentially improve the preoperative diagnosis and individual treatment of CLNM in patients with PTC and PTMC. CEUS may be helpful in predicting CLNM of PTC, but CEUS would be ineffective in predicting CLNM of PTMC.