关键词: 甲状腺乳头状癌, 淋巴结转移, 临床预测模型, 风险评估

Abstract: Objective This study aimed to create a clinical predictive model to assess the probability of lateral lymph node metastasis in patients diagnosed with PTC. Methods A retrospective analysis was conducted on 336 patients with pathologically confirmed PTC with cervical lymph node metastasis at our hospital from January 2022 to January 2024, including 189 cases of central cervical lymph node metastasis and 147 cases of central + lateral lymph node metastasis. Clinical, imaging, and histopathological data were gathered. The cohort was randomly divided into a training group （n=235） and a validation group （n=101）. Independent risk factors were identified through univariate and multivariate logistic regression analyses to develop a nomogram model, with its performance assessed using the receiver operating characteristic （ROC） curve, consistency index （consistency-index, C-index）, and calibration curve. Results The multivariate logistic regression analysis identified significant predictors of LNM, which included the number of lesions （single or multiple）, composition, the presence of fine calcification, invasion of the thyroid envelope, and low or very low echogenicity. A nomogram was constructed based on these five indicators, and the C-index for the training and validation groups was 0.836（95%CI：0.783-0.890）, 0.839（95%CI：0.761-0.917）, respectively. Calibration curves demonstrated good agreement of the model. Conclusion The clinical prediction model developed in this study serves as a valuable instrument for forecasting lateral lymph node metastasis in patients with PTC. This model aids in the effective monitoring of lateral lymph node metastasis in this patient population.

Key words: papillary thyroid carcinoma, lymph node metastasis, clinical predictive model, risk assessment