The Research Of Prognostic Factors Between TERT, CLPTM1L, CHRNA3Gene Polymorphism And Non-small Cell Lung Cancer In Never-smokers

[Objective]Investigate the prognostic factors of lung cancer patients; Research therelationship between TERT, CLPTM1L susceptibility locus rs2736100and rs401681and prognosis of non-smoking lung cancer patients; By constructing prognosticsubgroups survival tree method to explore the interaction between multiple factorsand predict the survival in lung cancer patients.[Methods]A prospective follow-up study of field epidemiology was conducted with451non-smoking lung cancer patients. Patients who diagnosed with NSCLC wererecruited from the First Clinical Medical College and the Affiliated Union Hospital ofFujian Medical University and Fuzhou General Hospital (survey period fromDecember2006-December2013). The inclusion criteria for the study cases:①obtaintissue samples by surgery or endoscopy, confirmed by pathology;②new primary lungcancer of non-smokers, exclude secondary critically ill or those who can’t clearlyanswer the question. The study using the Kaplan-Meier survival analysis and mappingthe distribution of the survival curve, COX multivariate analysis model.Methods of molecular epidemiology. Taking blood samples, extracting DNA anddetecting gene sequence. The study included160cases. Using χ2test theHardy-Weinberg genetic equilibrium of genotype. COX model analyse therelationship between lung cancer prognosis and site in different genetic model, COXmultivariate analysis model.The use of S-PLUS2000statistical analysis. By using the survival tree model toanalyse the interaction of different factors. Kaplan-Meier calculate differentprognostic group of1,3,5-years and observe median survival time, plot the survival curves.[Results]Prognostic factors of lung cancer in non-smokers: gender (HR=0.882,95%CI=0.715-0.962), histological type (adenocarcinoma: HR=1.439,95%CI=1.063-1.950;undifferentiated carcinoma: HR=1.596,95%CI=1.039-2.451), tumor N stage (N1:HR=2.448,95%CI=1.652-3.626; N2: HR=4.222,95%CI=2.657-6.708; N3: HR=5.857,95%CI=3.601-9.528), tumor size (3~7CM: HR=1.459,95%CI=1.082-1.967;≥7CM: HR=1.788,95%CI=1.002-3.199), surgical treatment (HR=0.427,95%CI=0.301-0.618), Karnofsky score (70points: HR=0.389,95%CI=0.173-0.878;90points: HR=0.345,95%CI=0.141-0.842).COX univariate analysis showed that in the recessive model of rs2736100, GGhomozygous type significantly increased risk of death (HR=1.645,95%CI=1.129-2.397) than GT/TT. In the recessive model of rs401681, AA homozygoustype significantly lower risk of death (HR=0.350,95%CI=0.177-0.692) than GA/GG.The combined effects of adverse genotype analysis showed that the more adversegene carried, the higher risk of death. Multivariate analysis showed that gender, Nstage, tumor size, whether surgery, Karnofsky score, rs2736100recessive model(GT/TT, GG) and rs401681recessive model (GA/GG, AA) were independentprognostic factors of lung cancer in non-smokers.By using the survival tree model to build prognostic subgroups. Prognosticsubgroups found that low-risk group (N0or N1stage) have the longest mediansurvival time (1283days,95%CI=1165.474-1500.526) and high-risk groups (N2orN3stage) have the highest risk of death (HR=4.943,95%CI=3.391-7.207).[Conclusion]COX univariate analysis showed that genotype GT/TT, GG of rs2736100andGA/GG, AA of rs401681in lung cancer non-smokers were significantly associatedprognosis. The GG of rs2736100and GA/GG of rs401681were the adversegenotypes. The combined effects of adverse genotype analysis showed that the moreadverse gene carried, the higher risk of death and the survival time distribution weresignificant difference (log-rank P=0.001). COX multivariate analysis showed that gender, tumor N stage, tumor size,whether surgery, Karnofsky score, rs2736100and rs401681were independentprognostic factors of lung cancer non-smokers.Survival tree model showed that N stage was the most important factor whichaffected the prognosis of lung cancer. Regression tree constructed by the method ofsurvival prognosis subgroup grouping found N0or N1stage had the best prognosis oflung cancer in non-smokers, and N2or N3stage had the worst prognosis.By using the survival tree model to build prognostic subgroups. The resultsshowed that N0or N1stage had the best prognosis of lung cancer in non-smokers,and N2or N3stage had the worst prognosis.