Predictive model for cervical lymph node metastasis of papillary thyroid carcinoma based on ultrasound and thyroglobulin infine-needle aspirate fluid detection

Objective To establish a predictive model of lateral lymph node metastasis in patients with papillary thyroid carcinoma(PTC), and further to compare the diagnostic efficiency of this model with the suspected abnormal lymph node thyroglobulin in fine-needle aspirate fluid (FNA-Tg) for lateral lymph node metastasis. Methods The preoperative clinical and ultrasonographic data of 110 patients (257 lymph nodes) who underwent PTC cervical lymph node dissection were retrospectively analyzed. According to the postoperative pathological results, they were divided into lateral lymph node metastasis and non-metastasis group. Regression analysis was used to screen out independent risk factors affecting lateral lymph node metastasis and establish a predictive model. The ROC curve was used to evaluate the diagnostic efficacy and the best diagnostic cut-off point. Results Prediction model: Logit(P)=-2.987+ 2.189(S/L ratio of lymph nodes)+ 1.748(hilum absent)+ 2.030(hyperechoic)+ 1.849(vascular abnormalities). The sensitivity, specificity, accuracy and AUC of the prediction model in the diagnosis of lateral lymph node metastasis were 92.1%, 83.9%, 87.9% and 0.929, respectively. The Homser-Lemeshow goodness of fit test showed that the Logistic model has a good fitting effect. The sensitivity, specificity, accuracy, and AUC of FNA-Tg in the diagnosis of lateral lymph node metastasis were 87.4%, 95.4%, 90.3% and 0.968, respectively. The sensitivity, specificity, accuracy, and AUC of the combined diagnosis of the predictive model and FNA-Tg were 92.9%, 96.9%, 94.2% and 0.989, respectively. Conclusions The model has a good predictive value for PTC cervical lymph node metastasis. Combined with FNA-Tg, it can improve its diagnostic efficiency and provide more valuable information for the decision-making of clinical surgical procedures. Key words: Ultrasonography; Thyroid neoplasms; Lymphatic metastasis; Logistic models