摘要
目的 探讨基于CT征象和血清巨噬细胞集落刺激因子(M-CSF)、β-连环蛋白(β-Catenin)构建的联合模型对孤立性肺结节的评估价值.方法 选取2020年2月至2023年2月萍乡市人民医院病理结果显示孤立性肺结节的105例患者作为研究对象,以孤立性肺结节性质为结局指标,软件计算建模组所需样本量为75例,故随机抽取75例作为建模组、其余30例作为验证组,分析建模组孤立性肺结节不同性质患者一般资料、CT征象、生化指标,多因素logistic回归分析法筛选影响孤立性肺结节性质的独立影响因子,并建立预测模型.用验证组临床资料进行验证并将验证组临床资料代入预测模型,根据预测概率绘制ROC曲线评估该预测模型的效能.结果 建模组75例病例中,良性肺结节43例(57.33％)、恶性肺结节32例(42.67).单因素分析结果显示,良性、恶性肺结节患者毛刺征、分叶征、血管集束征、胸膜凹陷征、磨玻璃密度、白细胞介素-6(IL-6)、白细胞介素-8(IL-8)、癌胚抗原(CEA)、神经特异性烯醇化酶(NSE)、M-CSF、β-Catenin水平与建模组孤立性肺结节性质有关,差异有统计学意义(P＜0.05).多因素 logistic 回归分析显示:毛刺征(β=1.572,OR=1.612,95％CI=1.475～2.159)、分叶征(β=1.484,OR=1.650,95％CI=1.408～2.184)、血管集束征(β=1.292,OR=1.674,95％CI=1.466～2.198)、磨玻璃密度(β=1.307,OR=1.785,95％CI=1.501～2.263)、CEA(β=0.915,OR=1.797,95％CI=1.505～2.199)、NSE(β=0.920,OR=1.824,95％CI=1.618～2.218)、M-CSF(β=0.859,OR=1.765,95％CI=1.506～2.301)、β-Catenin(β=0.867,OR=1.806,95％CI=1.542～2.325)是影响建模组孤立性肺结节性质的独立危险因素(P＜0.05),据此建立的联合模型ROC曲线结果显示:该模型评估孤立性肺结节良恶性的灵敏度为90.50％、特异度为75.26％、AUC为0.880(95％CI=O.802～0.935)、阳性似然比为4.05、阴性似然比为0.12、阳性预测值为82.80％、阴性预测值为89.90％.结论 毛刺征、分叶征、血管集束征、磨玻璃密度及CEA、NSE、M-CSF、β-Catenin水平升高是影响孤立性肺结节良恶性的独立影响因素,据此构建的联合模型评估孤立性肺结节良恶性的效能较高.
Abstract
Objective To investigate the value of the combined model based on CT signs,serum macrophage colony-stimu-lating factor(M-CSF)and β-Catenin in the evaluation of isolated pulmonary nodules.Methods A total of 105 patients with isolated pulmonary nodules were selected as research objects from Pingxiang People's Hospital from February 2020 to February 2023,with the properties of isolated pulmonary nodules as the outcome index,the software calculated that the sample size required for the modeling group was 75 cases,so 75 cases were randomly selected as the modeling group and the remaining 30 cases as the verification group.General data,CT signs and biochemical indicators of patients with different characteristics of solitary pulmonary nodules in the modeling group were analyzed,and independent factors affecting the na-ture of solitary pulmonary nodules were screened by multivariate logistic regression analysis,and prediction models were es-tablished.The clinical data of the validation group were validated and substituted into the regression model of this study,and the ROC curve was drawn according to the prediction probability to evaluate the effectiveness of the model.Re-sults Among the 75 cases in the modeling group,43(57.33％)were benign pulmonary nodules and 32(42.67％)were malig-nant pulmonary nodules.The results of single factor analysis showed that burr sign,lobed sign,vascular cluster sign,pleural depression sign,ground glass density,levels of interleukin-6(IL-6),interleukin-8(IL-8),carcinoembryonic antigen(CEA),neurospecific enolase(NSE),M-CSF,β-Catenin in benign and malignant pulmonary nodules were correlated with the properties of solitary pulmonary nodules in the modeling group,the differences were statistically significant(P＜0.05).Mul-tivariate logistic regression analysis showed that burr sign(β=1.572,OR=1.612,95％CI=1.475-2.159),lobular sign(β=1.484,OR=1.650,95％CI=1.408-2.184),vascular cluster sign(β=1.292,OR=1.674,95％CI=1.466-2.198),ground glass density(β=1.307,OR=1.785,95％CI=1.501-2.263),CEA(β=0.915,OR=1.797,95％CI=1.505-2.199),NSE(β=0.920,OR=1.824,95％CI=1.618-2.218),M-CSF(β=0.859,OR=1.765,95％CI=1.506-2.301),β-catenin(β=0.867,OR=1.806,95％CI=1.542-2.325)was an independent risk factor affecting the nature of solitary pulmonary nodules in the modeling group(P＜0.05).ROC curve results of the combined model showed that the sensitivity,specificity and AUC of the model in evaluating benign and malignant solitary pulmonary nodules were 90.50％,75.26％and 0.880(95％CI=0.802-0.935),positive likelihood ratio was 4.05,negative likelihood ratio was 0.12,positive predictive value was 82.80％,negative predictive value was 89.90％.Conclusion Burr sign,lobed sign,vascular cluster sign,ground glass den-sity and CEA,NSE,M-CSF and β-Catenin levels are independent factors affecting the benign and malignant of solitary pulmonary nodules,and the combined model built based on this is effective in evaluating the benign and malignant of soli-tary pulmonary nodules.
基金项目
江西省卫生健康委科技计划项目(SKJP220200901)
维普
万方数据