基于流行病学资料及基因单核苷酸多态性的肺癌预测模型的建立

Establishment of the prediction models for lung cancer based on the epidemiology and susceptibility gene sing nucleotide polymorphisms

目的:探讨易感基因单核苷酸多态性(SNP)与肺癌易感性之间的关系,联合SNP与流行病学资料建立适合本地区的肺癌风险预测模型。方法:采用病例-对照研究方法,收集2017年1~10月徐州市肿瘤医院病理确诊的肺癌患者100例作为病例组,选取同期徐州市肿瘤医院体检中心正常人群100例作为对照组。收集病例及对照组流行病学资料,包括一般特征(性别、年龄、居住地)、吸烟史、肺炎史、其他肿瘤史、石棉暴露史、肿瘤家族史、慢性阻塞性肺病(COPD)史、肺结核史、支气管哮喘史等。采用Taqman real-time PCR方法对CYP1A1基因rs4646903位点、CYP2E1基因rs3813867位点、CYP1B1基因rs1056836位点、GSTP1基因rs1695位点、ERCC2基因rs13181位点、XRCC1基因rs25487位点、TP53基因rs1042522位点、NQO1基因rs1800566位点、CCND1基因rs603965位点、MTHFR基因rs1801133位点、EPHX1基因rs1051740位点SNP进行基因分型,利用Logistic回归筛选变量,计算模型中优势比(OR)及其95%可信区间(95%CI),分别构建纳入SNP与未纳入SNP的预测模型,并用受试者工作特征曲线(ROC曲线)对两种模型进行性能评价。结果:多因素Logistic回归模型分析显示,年龄、吸烟史、石棉暴露史、肿瘤家族史是肺癌发生的危险因素;性别、肺炎史、其他肿瘤史、COPD史、肺结核史、支气管哮喘史与肺癌发病无相关性。CYP1B1基因rs1056836位点和GSTP1基因rs1695位点SNP与肺癌易感性相关(OR=3.8,95%CI 1.89~7.63,P<0.05;OR=3.3,95%CI 1.73~6.3,P<0.05)。未纳入SNP的基于环境及机体内在因素的肺癌预警模型ROC曲线下面积为0.839;纳入SNP的基于环境及机体内在因素的肺癌预警模型ROC曲线下面积为0.857。结论:Logistic回归模型提示肺癌危险因素包括年龄、吸烟史、石棉暴露史、肿瘤家族史、CYP1B1基因rs1056836位点和GSTP1基因rs1695位点SNP,本研究所构建的两种肺癌预测模型均能较好预测肺癌的风险。

Objective: To investigate the association between susceptibility gene single nucleotide polymorphisms(SNP)and lung cancer, to establish the prediction models for lung cancer based on the epidemiology and susceptibility gene SNP.Methods: From January to October 2017, a case-control study was performed for 100 lung cancer patients diagnosed by pathology and 100 healthy control individuals who were selected from Center of Health Examination. Epidemiological data mainly included general characteristics(gender, age, the current place of residence), smoking status, previous pneumonia,previous malignant, asbestos exposure,family lung cancer, chronic obstructive pulmonary diseases(COPD), previous pulmonary tuberculosis,bronchial asthma. The genetic polymorphisms of CYP1 A1 rs4646903, CYP2 E1 rs3813867, CYP1 B1 rs1056836, GSTP1 rs1695, ERCC2 rs13181, XRCC1 rs25487, TP53 rs1042522, NQO1 rs1800566, CCND1 rs603965, MTHFR rs1801133, EPHX1 rs1051740 were examined by real-time PCR. Logistic regression model was performed to identify risk factors for lung cancer and calculated odds ratios(OR)and 95% CI. The classification ability of the model was evaluated using the area under the receiver operating characteristic(ROC) curve. Results: Lung cancer risk factors included age, smoking status, asbestos exposure, family lung cancer. No relationship was found between gender, previous pneumonia, previous malignant, COPD, previous pulmonary tuberculosis, bronchial asthma and lung cancer. The individuals who carried with CYP1 B1 rs1056836 and GSTP1 rs1695 mutant heterozygote and homozygote had a high risk of lung cancer(OR=3.8, 95% CI 1.89-7.63, P<0.05;OR=3.3, 95% CI 1.73-6.3, P<0.05). The area under the curve were 0.839 and 0.857 based on environmental factors and inherent factors without SNP and with SNP. Conclusion: Acccording to the multiple logistic regression, age, smoking status,asbestos exposure,family lung cancer, genetic polymorphisms of CYP1 B1 rs1056836 and GSTP1 rs1695 were risk factors for lung cancer. The two prediction models by this thesis can be used to predict the risk of lung cancer preferably.