基于列线图评估分化型甲状腺癌颈部淋巴结转移

Nomogram-Based Prediction of Risk of Cervical Lymph Node Metastasis in Differentiated Thyroid Carcinoma

目的 应用列线图模型预测分化型甲状腺癌(DTC)颈部淋巴结转移的风险。方法 收集2019年1月至2021年12月手术病理确诊的DTC患者的临床及颈部超声检查资料,分为训练组(n=444)和验证组(n=125)。采用Lasso回归筛选组间差异的数据,采用多因素Logistic回归分析结合Lasso回归分析中选择的特征,建立预测模型。采用C指数、校准图来评估预测模型的诊断性能。结果 淋巴结短径≥0.5 cm、长短径比值<2、淋巴结门消失、囊性变、高回声、钙化及异常血流(P均<0.001)是建立模型的预测因素。该模型具有良好的鉴别能力,训练组C指数为0.938(95%CI=0.926~0.961)。结论 基于DTC颈部淋巴结超声声像图特征建立了列线图模型,可以准确评估DTC颈部淋巴结转移风险。

Objective To establish a nomogram for predicting the risk of cervical lymph node metastasis in differentiated thyroid carcinoma(DTC).Methods The patients with complete clinical data of DTC and cervical lymph node ultrasound and diagnosed based on pathological evidence from January 2019 to December 2021 were assigned into a training group(n=444)and a validation group(n=125).Lasso regression was performed to screen the data with differences between groups,and multivariate Logistic regression to establish a prediction model with the factors screened out by Lasso regression.C-index and calibration chart were employed to evaluate the prediction performance of the established model.Results The predictive factors for establishing the model were lymph node short diameter≥0.5 cm,long-to-short-axis ratio<2,disappearance of lymph node hilum,cystic transformation,hyperechogenicity,calcification,and abnormal blood flow(all P<0.001).The established model demonstrated a good discriminative ability,with the C index of 0.938(95%CI=0.926-0.961)in the training group.Conclusion The nomogram established based on the ultrasound image features of cervical lymph nodes in DTC can accurately predict the risk of cervical lymph node metastasis in DTC.