基于炎症指标和肿瘤标志物的肺癌治疗前远处转移预测模型的构建和验证  

Construction and verification of distant metastasis prediction model of lung cancer basedon inflammatory markers and tumor markers before treatment *

作　　者：颜晰[1] 牛英豪[2] 郭秀娟[1] 冯军花[1] 李燕[1] 张金艳[1] YAN Xi;NIU Yinghao;GUO Xiujuan;FENG Junhua;LI Yan;ZHANG Jinyan(Department of Clinical Laboratory,the Fourth Hospital of Hebei Medical University,Shijiazhuang,Hebei 050011,China;Department of Clinical Biobank,the First Hospital of Hebei Medical University,Shijiazhuang,Hebei 050031,China)

机构地区：[1]河北医科大学第四医院检验科,河北石家庄050011 [2]河北医科大学第一医院临床生物样本库,河北石家庄050031

出　　处：《检验医学与临床》2023年第4期470-474,480,共6页Laboratory Medicine and Clinic

基　　金：河北省医学科学研究重点课题计划项目(20170757)。

摘　　要：目的探讨炎症指标及肿瘤标志物联合检测对肺癌远处转移的预测价值。方法选取河北医科大学第四医院肺癌患者133例作为研究组,根据转移情况分为转移组、未转移组,检测外周血中性粒细胞/淋巴细胞比值(NLR)、细胞因子及肿瘤标志物,筛选肺癌远处转移的危险因素,以Logistic回归分析构建模型,采用受试者工作特征(ROC)曲线分析评估该模型对肺癌远处转移的预测效能,并以河北医科大学第一医院98例肺癌患者作为验证组进行验证。结果转移组NLR、白细胞介素(IL)-8、癌胚抗原(CEA)、细胞角蛋白19片段(CYFRA21-1)、神经元特异性烯醇化酶(NSE)水平均明显高于未转移组,差异有统计学意义(P<0.05);多因素Logistic回归分析结果显示,IL-8(OR=1.008,P=0.014)、CYFRA21-1(OR=1.067,P=0.035)水平升高是预测肺癌远处转移的独立危险因素,联合二者建立的风险预测模型为Logit(P)=-2.217+0.008 X IL-8+0.081 X CYFRA21-1,对预测肺癌远处转移有良好的效能,在研究组、验证组中曲线下面积分别为0.820和0.744,灵敏度分别为91.2%和80.0%,准确率分别为78.9%和72.4%。结论IL-8、CYFRA21-1水平升高是肺癌远处转移的独立危险因素,二者联合构建的风险预测模型预测肺癌远处转移具有良好的性能。Objective To explore the value of combined detection of inflammatory markers and tumor markers in predicting distant metastasis of lung cancer.Methods A total of 133 patients with lung cancer in the Fourth Hospital of Hebei Medical University were selected as the study group,and divided into metastatic group and non-metastatic group.Peripheral blood neutrophil/lymphocyte ratio(NLR),cytokines and tumor markers were detected to screen the high risk factors of distant metastasis of lung cancer.The model was constructed by Logistic regression analysis,and the predictive efficiency of the model was evaluated by receiver operation characteristic(ROC)curve analysis.A total of 98 patients with lung cancer in the First Hospital of Hebei Medical University were used as verification group.Results The levels of NLR,interleukin(IL)-8,carcinoembryonic antigen(CEA),cytokeratin 19 fragment(CYFRA21-1)and neuron-specific enolase(NSE)in metastatic group were significantly higher than those in non-metastatic group,the differences were statistical significance(P<0.05).The results of multivariate Logistic regression analysis showed that increased IL-8(OR=1.008,P=0.014)and CYFRA21-1(OR=1.067,P=0.035)were independent risk factors for identifying distant metastasis of lung cancer.The risk prediction model Logit(P)=-2.217+0.008 X IL-8+0.081 X CYFRA21-1 was effective in predicting distant metastasis of lung cancer.The areas under the curve of the study group and verification group were 0.820 and 0.744 respectively,the sensitivities were 91.2%and 80.0%,the accuracies were 78.9%and 72.4%respectively.Conclusion IL-8 and CYFRA21-1 are independent risk factors of distant metastasis of lung cancer,the risk prediction model of them has good performance for distant metastasis of lung cancer.

关 键 词：肺癌 远处转移 白细胞介素-8 细胞角蛋白19片段 

分 类 号：R734.2[医药卫生—肿瘤] R446.6[医药卫生—临床医学]