Establishment And Application Of Early-warning Models For Lung Cancer In Northeast Of China

Background and objective:Lung cancer is now the leading cause for cancer mortality in the world. Most lung cancerpatients are diagnosed at an advanced stage, and hence are not able to undergo surgicalremoval of tumors. Systemic chemotherapy and radiotherapy are currently the main treatmentoptions for lung cancer, however most patients eventually develop resistance to thesetreatments. As a result, the overall5-year survival rate of lung cancer patients is still low.Therefore, early stage detection is essential for an effective therapy of lung cancer. In thisrespect, a well-established lung cancer risk forecasting model that could identify individualsat high-risk would greatly benefit patients, clinicians and general public.The objective of our study:1. To explore the relationship between single nucleotide polymorphisms (SNP) andsusceptibility to lung cancer. Several environmental factors and lung cancer family historywere analyzed. To screen out effective molecular biomarkers correlated with lung cancer, andfind the most significant index so as to come up with initial value for early warning of lungcancer.2. To explore lung cancer risk prediction models which are suitable for the population inthe Northeast of China, further explore the risk factors of lung cancer, and provide clues forthe research on the etiology of lung cancer.Methods:1. There were1693participants, of which case was705and control was988. Cases andcontrols people lived in the same residence for at least10years. Epidemiological data mainlyincluded: general characteristics (gender, age, culture level, the current place of residence),diet behaviors (history of drinking, different kinds of food), smoking and passive smokinghistory, other exposure history (such as asbestos, cooking emissions, gasoline and so on), family history, medical history.2. There were1000participants, of which were selected randomly. The case was500and control was500, which were matched according to gender and age. After completion ofthe investigation and collecting3-5ml respondent anticoagulant blood, DNA was extractedand SNPs were detected by Sequenom Massarray genotype system.3. All analyses were conducted using SPSS v19.0software. Multiple statistical analysismethods were applied including chi-square test, t test, rank sum test and so on. Logisticregression model was performed to identify independent risk factors for lung cancer. Theclassification ability of the model was evaluated using the area under the receiver operatingcharacteristic (ROC) curve (AUC), and the Sensitivity and Specificitywere given afterwards.Results:1. Epidemiological findings:It indicated that lung cancer occurred more frequently in individuals with the followingcharacteristics: aged, male, low education level, family history of cancer, history of COPD,pulmonary tuberculosis, or tuberculosis, low BMI, consumption of less fish, consumption ofless vegetables, consumption of less fruits, consumption of less dairy products, andconsumption of less soybean products and nuts, consumption of alcohol, smoking;occupational exposure to pesticides, gasoline or diesel, and heavy exposure to cookingemissions. Both inadequate and excessive meat consumption were related to high lung cancerrisk.2. Genetic test results:In the univariate analysis, the genotype AC and CC in rs2736100, AG and GG inrs25487, GG in rs2279744, GG in rs2522137were associated with a higher incidence of lungcancer. The genotype TC and TT in rs1233560were associated with a lower incidence of lungcancer. Other sites of genotype distribution had not been found that there was significantlydifferent between the case and control groups.3. Early warning model for lung cancer:(1) Based on environmental factors and inherent factors（without SNP）Lung cancer risk factors included aged, male, lower education level, family history of cancer, history of COPD, history of pneumonia, history of tuberculosis, lower BMI, smoking,a diet with less fish, vegetables, dairy products, soybean products and nuts, a diet rich in meat,and exposure to pesticides and heavy cooking emissions. The area under the curve was0.8851.In the hierarchical models, lower education level, family history of cancer, history ofCOPD, history of pneumonia, lower BMI, smoking cigarettes, a diet with less fish andvegetables, a diet rich in meat were lung cancer risk factors in smoking population；Lowereducation level, family history of cancer, passive smoking, history of COPD, history oftuberculosis, a diet with less fish and vegetables, exposure to heavy cooking emissions werelung cancer risk factors in non-smoking population. Lower education level, family history ofcancer, history of COPD, history of pneumonia, lower BMI, smoking cigarettes, a diet withless fish and vegetables, a diet rich in meat, and exposure to pesticides were lung cancer riskfactors in male population. Lower education level, history of COPD, smoking cigarettes, adiet with less vegetables, a diet with less soybean products and nuts, exposure to heavycooking emissions, menopause, lactation and high production (more than1time) were lungcancer risk factors, diet deficient in meat was a profective factor of lung cancer in femalepopulation.(2) Based on environmental and inherent factors (include SNP)Lung cancer risk was significantly associated with rs2736100C allele, rs3764340G allele,rs2522137GG and rs2279744GG genotypes in combination with environmental factors,including lower education level, family history of cancer, history of COPD, lower BMI,smoking cigarettes, diet deficient in fish and shrimp, a diet with less vegetable, a diet withless soybean products and nuts, a diet rich in meat, and exposure to heavy cooking emissions.The AUC was0.8703.In the hierarchical models, female, lower education level, family history of cancer,history of COPD, history of pneumonia, smoking cigarettes, a diet with less fish andvegetables, a diet with less soybean products and nuts, a diet rich in meat, rs2736100C allele,rs3764340G allele，rs1136410T allele were lung cancer risk factors in smoking population.Female, lower education level, family history of cancer, history of passive smoking, history of COPD, history of tuberculosis, a diet with less fish and vegetables, lower BMI, rs2736100C allele, rs2522137GG genotype were lung cancer risk factors, rs1052133GG genotype,rs25487G allele were lung cancer protect factors in non-smoking population；Lowereducation level, family history of cancer, history of COPD, history of pneumonia, lower BMI,smoking cigarettes, a diet with less fish and vegetables, a diet rich in meat, rs2736100C allele,rs1052133GG, rs3764340G allele and rs1798802A allele were lung cancer risk factors inmale population. Lower education level, lower BMI, a diet with less vegetables, a diet withless fish, a diet with less soybean products and nuts, family history of cancer, lactation, highproduction (more than1time) were lung cancer risk factors，Menarche age less than15andrs25487G allele were lung cancer protect factors in female population.Conclusions：1. In the general population, lung cancer risk factors include aged, male, lower educationlevel, family history of cancer, history of COPD, history of pneumonia, history oftuberculosis, lower BMI, smoking cigarettes, a diet with less fish, a diet with less vegetables,a diet with less dairy products, a diet with less soybean products and nuts, a diet rich in meat,exposure to pesticides and heavy cooking emissions.2. In smoking population, lower education level, family history of cancer, history ofCOPD, history of pneumonia, lower BMI, smoking cigarettes, a diet with less fish andvegetables, a diet rich in meat are lung cancer risk factors.3. In non-smoking population, lower education level, family history of cancer, passivesmoking, history of COPD, history of tuberculosis, a diet with less fish and vegetables,exposure to heavy cooking emissions are lung cancer risk factors.4. In male population, lower education level, family history of cancer, history of COPD,history of pneumonia, lower BMI, smoking cigarettes, a diet with less fish and vegetables, adiet rich in meat, exposure to pesticides are lung cancer risk factors.5. In female population, lower education level, history of COPD, smoking cigarettes, adiet with less vegetables, a diet with less soybean products and nuts, exposure to heavycooking emissions, menopause, lactation, high production(more than1time), lowermenstrual cycles are lung cancer risk factors 6. In the lung cancer risk models, the genetic polymorphisms of rs2736100, rs3764340,rs2522137, rs2279744, rs1136410, rs1052133, rs25487and rs1798802may contribute to therisk of developing lung caner.7. The warning model of residents in the Northeast district of China can be used topredict the risk of lung cancer.