Ultrasonic Characteristics Improve Prediction of Central Lymph Node Metastasis in cN0 Unifocal Papillary Thyroid Cancer

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.

Central lymph nodes in frozen sections can effectively guide extended lymph node resection for papillary thyroid carcinoma

OBJECTIVES

The scope of lateral neck lymph node dissection (LND) in papillary thyroid carcinoma (PTC) remains controversial. Our research aimed to explore the value of central lymph node metastasis (CLNM) in frozen sections for predicting neck lateral lymph node metastasis (NLLNM) and to guide clinical surgeons in performing surgical lymph node dissection.

PATIENTS

A total of 275 patients with PTC with suspected 'Cervical lymph node metastasis (LNM, including CLNM and NLLNM)' underwent unilateral or bilateral thyroidectomy and an intraoperative frozen diagnosis of central lymph nodes (LNs), as well as central and neck lateral LND. Validity indices and consistency of central LNs in frozen sections were calculated. In total, 216 patients then met the inclusion criteria and were enrolled in the follow-up study. The clinical and pathological data of the patients were retrospectively analyzed. The relationship between the number, metastatic diameter, and the ratio of CLNM to NLLNM was investigated.

RESULTS

CLNM in frozen and paraffin-embedded sections was associated with NLLNM. Univariate and multivariate analyses revealed the following risk factors for NLLNM metastasis: maximum diameter, total number, and ratio of metastatic LNs. A significant result was obtained when a cut-off value of 2.050 mm for the maximum metastatic diameter, 5.5 in the total number, and 0.5342 for the CLNM ratio level was used. Interaction term analyses showed that the association between the number of CLNM and NLLNM differed according to maximum diameter.

CONCLUSION

Central LNs in frozen sections accurately predicted NLLNM. In patients with PTC with >5 CLNMs, ≥2 and ≤5 CLNMs and maximum metastatic diameter > 2 mm, neck lateral LND should be considered. Our findings will facilitate the identification of patients who are likely to benefit from extended lateral neck LND.

A model for predicting lymph node metastasis of thyroid carcinoma: a multimodality convolutional neural network study

Background

Early preoperative evaluation of cervical lymph node metastasis (LNM) in papillary thyroid carcinoma (PTC) is critical for further surgical treatment. However, insufficient accuracy in predicting LNM status for PTC based on ultrasound images is a problem that needs to be urgently resolved. This study aimed to clarify the role of convolutional neural networks (CNNs) in predicting LNM for PTC based on multimodality ultrasound.

Methods

In this study, the data of 308 patients who were clinically diagnosed with PTC and had confirmed LNM status via postoperative pathology at Beijing Tiantan Hospital, Capital Medical University, from August 2018 to April 2022 were incorporated into CNN algorithm development and evaluation. Of these patients, 80% were randomly included into the training set and 20% into the test set. The ultrasound examination of cervical LNM was performed to assess possible metastasis. Residual network 50 (Resnet50) was employed for feature extraction from the B-mode and contrast-enhanced ultrasound (CEUS) images. For each case, all of features were extracted from B-mode ultrasound images and CEUS images separately, and the ultrasound examination data of cervical LNM information were concatenated together to produce a final multimodality LNM prediction. Sensitivity, specificity, accuracy, and area under the receiver operating characteristic curve (AUC) were used to evaluate the performance of the predictive model. Heatmaps were further developed for visualizing the attention region of the images of the best-working model.

Results

Of the 308 patients with PTC included in the analysis, 158 (51.3%) were diagnosed as LNM and 150 (48.7%) as non-LNM. In the test set, when a triple-modality method (i.e., B-mode image, CEUS image, and ultrasound examination of cervical LNM) was used, accuracy was maximized at 80.65% (AUC =0.831; sensitivity =80.65%; specificity =82.26%), which showed an expected increased performance over B-mode alone (accuracy =69.00%; AUC =0.720; sensitivity =70.00%; specificity =73.00%) and a dual-modality method (B-mode image plus CEUS image: accuracy =75.81%; AUC =0.742; sensitivity =74.19%; specificity =77.42%). The heatmaps of our triple-modality model demonstrated a possible focus area and revealed the model's flaws.

Conclusions

The PTC lymph node prediction model based on the triple-modality features significantly outperformed all the other feature configurations. This deep learning model mimics the workflow of a human expert and leverages multimodal data from patients with PTC, thus further supporting clinical decision-making.

Constructing a nomogram based on the distribution of thyroid nodules and suspicious lateral cervical lymph nodes in fine-needle aspiration biopsies to predict metastasis in papillary thyroid carcinoma

Purpose

Elevated concentrations of thyroglobulin eluent is a risk factor for lateral cervical lymph node metastasis (LLNM) in patients with papillary thyroid cancer (PTC). We aimed to develop a practical nomogram based on the distribution of thyroid nodules and the presence of suspicious lateral cervical lymph nodes in fine-needle aspiration biopsies (LN-FNABs), including the cytopathology and the suspicious lateral cervical lymph node (LLN) thyroglobulin eluent (Tg), to predict the possibility of LLNM preoperatively in patients with PTC.

Methods

The clinical data of PTC patients who were admitted to the Third Affiliated Hospital of Soochow University from January 2022 to May 2023 to undergo fine-needle aspiration biopsy (FNAB) were included in this study. A total of 208 patients in 2022 served as the training set (70%), and 89 patients in 2023 served as the validation set (30%). The clinical characteristics and LN-FNAB results were collected to determine the risk factors of LLNM. A preoperative nomogram was developed for predicting LLNM based on the results of the univariate and multivariate analyses. Internal calibration, external calibration, and decision curve analysis (DCA) were performed for these models.

Results

The multivariate logistic regression analysis showed that the maximum thyroid nodule diameter (Odds Ratio (OR) 2.323, 95% CI 1.383 to 3.904; p = 0.001), Tg level (OR 1.007, 95% CI 1.005 to 1.009; p = 0.000), Tg divided by serum thyroglobulin, (Tg/sTg) [odds ratio (OR) 1.005, 95% CI 1.001 to 1.008; p = 0.009], and cytopathology (OR 9.738, 95% CI 3.678 to 25.783; p = 0.000) (all p <  0.05) had a significant impact on the LLNM of patients with suspicious LLNs. The nomogram showed a better predictive value in both the training cohort [area under the curve, (AUC) 0.937, 95% CI 0.895 to 0.966] and the validation cohort (AUC 0.957, 95% CI 0.892 to 0.989). The nomogram also showed excellent internal and external calibration in predicting LLNM. According to the DCA, the diagnostic performance of this model was dependent on the following variables: maximum thyroid nodule diameter, Tg level, Tg/sTg, and cytopathology.

Conclusion

Based on the aforementioned risk factors, we believe that it is necessary to establish a personalized LLNM model for patients with PTC. Using this practical nomogram, which combines clinical and Tg risk factors, surgeons could accurately predict the possibility of LLNM preoperatively. The nomogram will also help surgeons to establish personalized treatment plans before surgery.