基于双能量CT的列线图在预测甲状腺乳头状癌腺外侵犯的价值

The value of dual energy CT-based nomogram in predicting extraglandular invasion of papillary thyroid

目的:探讨基于双能量CT建立的列线图在术前预测甲状腺乳头状癌(PTC)腺外侵犯(ETE)的应用价值。方法:回顾性收集2017年3月至2021年12月温州医科大学附属第五医院经手术病理证实的252例PTC患者,根据病理诊断结果分为ETE组(n=150)和非ETE组(n=102)。采用7:3比例将所有患者随机分为训练集(n=176)和验证集(n=76)。所有患者术前均接受双能量CT颈部增强扫描,测量并计算病灶动脉期和静脉期标准化碘浓度(NIC)、能谱曲线斜率(λ_(HU))及有效原子序数(Z_(eff));分析常规CT特征包括病灶位置、最长径、有无钙化及有无突破包膜。经单因素分析和多因素Logistic回归分析筛选发生ETE的独立危险因素,构建联合模型并绘制列线图。使用受试者工作特征(ROC)曲线下面积(AUC)、校准曲线及决策曲线分析(DCA)评估模型的效能和临床价值。结果:训练集中,肿瘤位置及有无突破包膜在两组间的差异均有统计学意义(均P<0.05),ETE组患者的动脉期NIC、λ_(HU)、Z_(eff)值及静脉期Z_(eff)值均显著高于非ETE组(均P<0.05)。经多因素Logistic回归分析显示病灶位于腹侧、CT下突破包膜、动脉期NIC、动脉期Z_(eff)及静脉期Z_(eff)是预测PTC发生ETE的独立危险因素(均P<0.05)。构建的列线图预测模型在训练集和验证集中的AUC分别为0.902(0.859~0.946)和0.892(0.810~0.964)。校正曲线显示样本的实际值和模型预测概率之间具有较好的一致性,DCA表明模型的临床应用价值较高。结论:基于双能量CT的列线图可以为PTC患者术前无创化预测ETE提供一定的指导价值。

Objective:To investigate the value of nomograms based on dual-energy CT in predicting extraglandular invasion(ETE)of papillary thyroid carcinoma(PTC)before surgery.Methods:A total of 252 patients with PTC confirmed by surgery and pathology in the Fifth Affiliated Hospital of Wenzhou Medical University from March 2017 to December 2021 were retrospectively collected,and were divided into ETE group(n=150)and non-ETE group(n=102)according to the results of pathological diagnosis.All patients were randomly divided into training group(n=176)and validation group(n=76)using a ratio of 7:3.All patients underwent preoperative dual-energy CT neck enhanced scan,and the standardized iodine concentration(NIC),slope of the spectral curve(λHU),and effective atomic number(Z eff)of the lesion in the arterial and venous phases were measured and calculated;conventional CT features including lesion location,longest diameter,presence or absence of calcification,and presence or absence of breakthrough capsule were analyzed.Independent risk factors for ETE were screened by univariate analysis and multivariate logistic regression analysis,and a joint model was constructed and a nomogram was drawn.Model performance and clinical value were assessed using area under receiver operating characteristic(ROC)curve(AUC),calibration curve,and decision curve analysis(DCA).Results:In the training set,there’s significant difference between the two groups(all P<0.05)in the tumor location and whether or not the capsule had broken through the capsule.The NIC,λHU and Zeff in the arterial phase and the Zeff in the venous phase of PTC in the ETE group were significantly higher than those in the non-ETE group(all P<0.05).Multivariate logistic regression analysis showed that ventral location of the lesion,breakthrough of the capsule under CT,NIC in the arterial phase,Zeff in the arterial phase,and Zeff in the venous phase were independent risk factors for predicting ETE in PTC.The AUCs of the constructed nomogram prediction model in the training and validation sets were 0.902(0.859-0.946)and 0.892(0.810-0.964),respectively.The calibration curve showed a good agreement between the actual value of the sample and the predicted probability of the model,and the DCA indicated that the clinical application value of the model was high.Conclusion:The nomogram based on dual-energy CT can provide a certain guiding value for non-invasive preoperative prediction of ETE in patients with PTC.