Logistic回归和ROC曲线评价肿瘤标志物对大肠癌的诊断价值  

Diagnostic value of tumor markers for colorectal cancer analyzed with ROC curve and Logistic regression

作　　者：王越[1] 侴伟平[1] 沈春键[2] 王晓宇[2] 

机构地区：[1]辽宁省肿瘤医院麻醉科,辽宁沈阳110042 [2]沈阳医学院沈洲医院外科,辽宁沈阳110002

出　　处：《中国肿瘤外科杂志》2011年第6期354-356,共3页Chinese Journal of Surgical Oncology

摘　　要：目的探讨Logistic回归和ROC曲线评价CEA、CA19-9和AFP三种肿瘤标志物对大肠癌的诊断价值。方法采用电化学发光免疫分析法对100例大肠癌患者进行血清CEA、CA19-9和AFP的检测,并与100例大肠良性疾病患者进行对比分析。以Logistic回归筛选相关指标,建立回归方程,绘制ROC曲线并计算曲线下面积(AUC)。结果除AFP外,大肠癌患者CA19-9和CEA水平明显高于大肠良性疾病患者,差异有统计学意义(P<0.01)。AUC CEA>AUC CA19-9>AUC AFP。建立的Logistic回归方程为:Y=1/[1+EXP(3.250-0.224X1-0.057X2)],新变量Y的AUC高于三种单一肿瘤标记物的AUC。结论联合检测CEA和CA19-9可提高大肠癌的诊断率,运用Logistic回归和ROC曲线综合分析可提高大肠癌诊断的准确性。Objective To explore the diagnostic value of three tumor marker as analyzed with Logistic regressian and ROC curve in patients with colorectal cancer（CRC）. Methods CEA, CA19-9 and AFP levels were measured in 100 patients with colorectal cancer and 100 patients with benign colorectal disease. Logistic regression and ROC curve were applied to analyze the data. Results The serum concentrations of CEA and CA19-9 in colorectal cancer patients were significantly higher than those in patients with benign coloreetal disease （P 〈 0.01 ） , except AFP. AUC CEA 〉 AUC CA19.9 〉 AUC AFP. According to regression equation： Y = 1/ [ 1 ＋ EXP （ 3. 250 - 0. 224X1 - 0. 057X2 ） ], and the AUC of variable Y was higher than any of the three tumor markers. Conclusions Combined detection CEA and CA19-9 could improve the diagnostic sensitivity and ac- curacy significantly. Applications of Logistic regression and ROC curve increase diagnostic accuracy in eolorectal cancer.

关 键 词：大肠癌 癌胚抗原 甲胎蛋白 CA19-9 ROC曲线 LOGISTIC回归分析 

分 类 号：R735.34[医药卫生—肿瘤]