基于肺癌相关肿瘤标志物的肺癌列线图诊断模型的构建  

Construction of nomogram diagnostic model for lung cancer based on lung cancer-related tumor markers

作　　者：王永峰[1] 曾浈浈 娄若林 姚明解 吕全军 WANG Yongfeng;ZENG Zhenzhen;LOU Ruolin;YAO Mingjie;LYU Quanjun(Department of Laboratory Medicine,the First Affiliated Hospital,Zhengzhou University,Key Clinical Laboratory of Henan Province,Zhengzhou 450052;Department of Nuclear Medicine,the First Affiliated Hospital,Zhengzhou University,Zhengzhou 450052;Department of Laboratory Medicine,Zhumadian Central Hospital,Zhumadian,Henan 463000;Department of Anatomy and Embryology,School of Basic Medical Sciences,Peking University,Beijing 100191;Department of Nutrition,the First Affiliated Hospital,Zhengzhou University,Zhengzhou 450052)

机构地区：[1]郑州大学第一附属医院检验科,河南省检验医学重点实验室,郑州450052 [2]郑州大学第一附属医院核医学科,郑州450052 [3]驻马店市中心医院检验科,河南驻马店463000 [4]北京大学基础医学院人体解剖学与组织胚胎学系,北京100191 [5]郑州大学第一附属医院营养科,郑州450052

出　　处：《郑州大学学报（医学版）》2024年第4期513-518,共6页Journal of Zhengzhou University(Medical Sciences)

基　　金：国家自然科学基金青年项目(81802429,82103817);河南省医学科技攻关省部共建项目(SBGJ2018020091)。

摘　　要：目的:基于肺癌相关肿瘤标志物构建肺癌诊断模型。方法:选取2020年1月至2022年9月郑州大学第一附属医院就诊的258例原发性肺癌患者、109例肺部良性疾病患者及321名健康体检者。随机抽取70%(n=483)作为训练集,纳入性别、年龄和肺癌相关肿瘤标志物[癌胚抗原(CEA)、细胞角蛋白19片段抗原21-1(Cyfra21-1)、神经元特异性烯醇化酶(NSE)、糖类抗原125(CA125)、甲胎蛋白(AFP)、肿瘤异常糖链糖蛋白(TAP)]进行Logistic回归分析后构建诊断肺癌的列线图模型,剩余30%(n=205)作为验证集,采用ROC曲线评估列线图模型对肺癌的诊断价值。结果:列线图模型最终纳入了年龄、性别、CEA、Cyfra21-1、NSE和TAP共6个因素。训练集中,以非肺癌组(包括肺部良性疾病患者和健康体检者)为对照时,该模型诊断肺癌的AUC(95%CI)为0.915(0.886~0.938),灵敏度为0.801,特异度为0.894;验证集中,以非肺癌组为对照时,模型诊断肺癌的AUC(95%CI)为0.924(0.879~0.957),灵敏度为0.909,特异度为0.828。结论:基于肺癌相关肿瘤标志物构建的列线图模型对肺癌有较高的诊断价值,可用于肺癌的辅助诊断。Aim:To construct a diagnostic model of lung cancer based on lung cancer-related tumor markers.Methods:A total of 258 patients with primary lung cancer,109 patients with benign lung diseases,and 321 healthy individuals accepting physical examination,who were admitted to the First Affiliated Hospital of Zhengzhou University from January 2020 to September 2022 were enrolled.70%(n=483)was randomly selected as the training set,and the nomogram model for the diagnosis of lung cancer was constructed after Logistic regression analysis of gender,age and lung cancer-related tumor markers such as carcinoembryonic antigen(CEA),cytokeratin 19 fragment antigen 21-1(Cyfra21-1),neuron-specific enolase(NSE),carbohydrate antigen 125(CA125),alpha fetoprotein(AFP),and tumor abnormal protein(TAP);the remaining 30%(n=205)was used as the validation set,and the ROC curve was used to evaluate the value of nomogram model for the diagnosis of lung cancer.Results:Six factors were included in the nomogram model,including age,gender,TAP,CEA,Cyfra21-1,and NSE.In the training set,the AUC of the model in the diagnosis of lung cancer was 0.915(95%CI:0.886-0.938),the sensitivity was 0.801,and the specificity was 0.894.In the validation set,the AUC of the model for the diagnosis of lung cancer was 0.924(95%CI:0.879-0.957),the sensitivity was 0.909,and the specificity was 0.828.Conclusion:The constructed lung cancer nomogram diagnostic model has high diagnostic value and could be used for the auxiliary diagnosis of lung cancer.

关 键 词：肺癌 肿瘤标志物 诊断 列线图 

分 类 号：R734.2[医药卫生—肿瘤]