Nomograms Combining Ultrasonic Features With Clinical and Pathological Features for Estimation of Delphian Lymph Node Metastasis Risk in Papillary Thyroid Carcinoma

Abstract

Background: This work explores the clinical significance of Delphian lymph nodes (DLN) in thyroid papillary carcinoma (PTC). At the same time, a nomogram is constructed based on clinical, pathological, and ultrasonic (US) features to evaluate the possibility of DLN metastasis (DLNM) in PTC patients. This is the first study to predict DLNM using US characteristics.

Methods: A total of 485 patients, surgically diagnosed with PTC between February 2017 and June 2021, all of whom underwent thyroidectomy, were included in the study. Using the clinical, pathological, and US information of patients, the related factors of DLNM were retrospectively analyzed. The risk factors associated with DLNM were identified through univariate and multivariate analyses. According to clinical + pathology, clinical + US, and clinical + US + pathology, the predictive nomogram for DLNM was established and validated.

Results: Of the 485 patients with DLN, 98 (20.2%) exhibited DLNM. The DLNM positive group had higher positive rates of central lymph node metastasis (CLNM), lateral lymph node metastasis (LLNM), and T3b-T4b thyroid tumors than the negative rates. The number of CLNM and LLNM lymph nodes in the DLNM+ group was higher as compared to that in the DLNM- group. Multivariate analysis demonstrated that the common independent risk factors of the three prediction models were male, bilaterality, and located in the isthmus. Age ≥45 years, located in the lower pole, and nodural goiter were protective factors. In addition, the independent risk factors were classified as follows: (I) P-extrathyroidal extension (ETE) and CLNM based on clinical + pathological characteristics; (II) US-ETE and US-CLNM based on clinical + US characteristics; and (III) US-ETE and CLNM based on clinical +US + pathological features. Better diagnostic efficacy was reported with clinical + pathology + US diagnostic model than that of clinical + pathology diagnostic model (AUC 0.872 vs. 0.821, p = 0.039). However, there was no significant difference between clinical + pathology + US diagnostic model and clinical + US diagnostic model (AUC 0.872 vs. 0.821, p = 0.724).

Conclusions: This study found that DLNM may be a sign that PTC is more invasive and has extensive lymph node metastasis. By exploring the clinical, pathology, and US characteristics of PTC progression to DLNM, three prediction nomograms, established according to different combinations of features, can be used in different situations to evaluate the transfer risk of DLN.

Keywords: Delphian lymph nodes metastasis; nomograms; papillary thyroid carcinoma; risk factors; ultrasound.

Figure 1

The ultrasound features for Delphian lymph node metastasis prediction. (A) A solid and hypoechoic thyroid nodule with an irregular shape (burr visible at the edge of the nodule) and US-ETE (yellow arrow: interrupted envelope echogenicity at the site of contact between the nodule and the envelope). (B) A thyroid gland located in the isthmus, with solid hypoechoic, irregular shape (lobulated), clear envelope lines, and US-ETE (the nodule is in contact with the envelope for more than 20% of its circumference). (C) A US-CLNM with internal microcalcification (indicated by the red arrow). (D) A US-CLNM with hyperechoic change and loss of hilar echogenicity. (E) Two US-CLNMs lost hilar echogenicity with liquefaction necrosis (indicated by the red arrow). (F) A US-CLNM with abundant and disordered blood flow signals.

Figure 2

(A) Interactive nomogram for predicting DLNM based on clinical + pathological features. (B) Interactive nomogram for predicting DLNM based on clinical +US features. (C) Interactive nomogram for predicting DLNM based on clinical + US + pathological features. (The scores corresponding to the features in each nomogram are shown in the Supplementary Material . USETE, ultrasonic based extrathyroidal extension; PETE, Pathology-based extrathyroidal extension; USCLNM, ultrasonic-based central lymph node metastasis; CLNM, central lymph node metastasis; NG, Nodular goiter.

Figure 3

(A) AUC comparison of the ROC curves of the three models for predicting DLNM in the training group. (B) AUC comparison of the ROC curves of the three models for predicting DLNM in the validation group.

Figure 4

(A) Calibration plots of recalibrated prognostic model using the clinical + pathological features to predict the risk of DLN metastasis. (B) Calibration plots of recalibrated prognostic model using the clinical + US features to predict the risk of DLN metastasis. (C) Calibration plots of recalibrated prognostic model using the clinical + US + pathological features to predict the risk of DLN metastasis.

Figure 5

Decision curve analysis (DCA) curves of the three models.