基于多层螺旋CT双能量参数联合临床特征预测甲状腺乳头状癌颈部淋巴结转移

Prediction of Cervical Lymph Node Metastasis in Papillary Thyroid Carcinoma Based on Multisliecshelieal CT Dual Energy Parameters Combined with Clinical Features

目的探讨多层螺旋CT双能量参数联合临床特征预测甲状腺乳头状癌(PTC)患者颈部淋巴结转移(CLNM)的价值。方法回顾性分析280例经病理证实为PTC患者的临床及影像资料。其中训练组196例,验证组84例。采用单因素Logistic回归分析PTC患者CLNM的危险因素,采用多因素Logistic回归分析建立预测模型,最终建立列线图。使用ROC曲线、校准曲线及决策曲线对预测模型进行评估。结果单因素Logistic回归分析显示钙化、静脉期能谱曲线斜率(VK)和动脉期标准化CT值(ANCT)可能是影响CLNM的因素(P<0.05)。多因素Logistic回归分析显示钙化(P=0.03)和ANCT(P=0.02)是CLNM的独立预测因素,VK(P=0.41)是CLNM的危险因素并纳入模型,建立列线图。训练组ROC曲线下面积为0.841(0.727~0.954);验证组曲线下面积为0.854(0.668~1.000)。列线图对训练组和验证组都表现出良好的校准和鉴别能力。结论钙化、ANCT及VK是CLNM的3个独立预测因素,能够较好地预测PTC患者是否发生CLNM,帮助临床医师对PTC患者术前CLNM进行准确评估,为实现PTC患者的个性化治疗提供更多依据。

Objective To investigate the value of combining multisliecshelieal CT dual energy parameters with clinical features to predict cervical lymph node metastasis(CLNM)in patients with papillary thyroid carcinoma(PTC).Methods A retrospective analysis was performed for 280 patients with pathologically confirmed papillary thyroid carcinoma.Among them,196 cases were in the training group and 84 cases were in the validation group.Univariate logistic regression was used to analyze the risk factors of CLNM in PTC patients,multivariate logistic regression was used to establish a predictive mod-el,and finally a nomogram was established.The predictive model was evaluated using ROC curves,calibration curves and decision curve analysis.Results Univariate logistic regression analysis showed that calcification,VK and ANCT may be factors affecting CLNM(P<0.05).Multivariate logistic regression analysis showed that calcification(P=0.03)and AN-CT(P=0.02)were independent predictors of CLNM,and VK(P=0.41)as a risk factor for CLNM and were included in the model to establish a nomogram.The area under the ROC curve(AUC)in the training group was 0.841(95%CI,0.727-0.954).The AUC for the validation group was 0.854(95%CI,0.668-1.000).The nomogram showed good cali-bration and discrimination ability for both the training group and the validation group.Conclusion Calcification,ANCT and VK are three independent predictors of CLNM,which can better predict whether cervical lymph node metastasis occurs in PTC patients,helping clinicians to accurately evaluate preoperative CLNM in PTC patients and providing more basis for a-chieving personalized treatment for PTC patients.