Objective To establish a predictive model for central compartment lymph node metastasis (CLNM) in patients with papillary thyroid carcinoma (PTC).Methods We retrospectively evaluated patients with histologically confirmed PTC treated in Chinese PLA General Hospital between January 2018 and January 2019.Totally 855 patients undergoing thyroidectomy with central lymph node dissection were enrolled in this study (577 in the training cohort and 278 in the validation cohort).The clinical manifestations of the patients and the ultrasound features of the tumors were recorded.To predict the probability of CLNM,we developed the clinical-ultrasound model with significant factors selected by the stepwise logistic regression.The Akaike Information Criterion (AIC) was use to select the optimal model,and the area under the curve (AUC) and the calibration curves were used to evaluate the performance of the prediction model.Results Multivariate analysis showed that sex (P<0.001),age 45-55 years (P=0.004) or ≥55 years (P=0.003),tumor size 1-2 cm (P<0.001) or >2 cm (P=0.008),multifocality (P=0.029),microcalcification (P=0.019),diffused distribution of microcalcifications (P=0.001),and vascularity levels 2-3 (P=0.002) were effective for CLNM prediction.The model based on the clinical and ultrasound features showed good performance in both training cohort with AUC of 0.78 (95%CI:0.74-0.82) and validation cohort with AUC of 0.70 (95%CI:0.67-0.76).Conclusions A quantitative model of CLNM in PTC patients was established based on the risk factors.According to the model,central lymph node dissection is recommended for PTC patients with high scores.
Keywords: central lymph node metastasis; papillary thyroid carcinoma; prediction model; ultrasound.