Individualized Risk Assessment Of Gastric Cancer Based On Inflammatory Gene Polymorphisms

PurposeMeta-analysis was conducted to summarize recent studies on the association between inflammatory gene polymorphism and gastric cancer(GC).The inflammatory genes polymorphisms related to GC were screened out through combined effect size analysis.Inflammatory gene polymorphism sites which verified experimentally and traditional environmental risk factors were used to build a comprehensive risk prediction model for GC.A risk assessment scoring rule was constructed and initially verified based on the comprehensive risk prediction model of GC.Methods1.A meta-analysis and epidemiological evaluation based on the relationship between inflammatory gene polymorphism and GC(1)Gastric cancer,inflammatory gene and polymorphism were used as the theme words and combined with free words.Literature on Chinese Han population was collected from seven major databases(CNKI,Wanfang,VIP,Embase,Web of Science,Pubmed and Cochrane library).Quality evaluation of studies using NOS score scale.(2)STATA 15.0 software was used to calculate the combined effect-size odds ratio(OR)and 95%confidence interval(95%CI)for each polymorphism site of each inflammatory gene and to detect the publication bias.The false positive reporting probability was used to test the true correlation of the combined effect size results.(3)The effect of inflammatory gene polymorphism sites was epidemiologically evaluated using the percentage of attributable risk and the percentage of population attributable risk(PARP).2.Experimental verification of the association between inflammatory polymorphism sites and susceptibility to GC(1)The case group consisted of 660 GC patients diagnosed by histopathology or histology.The control group consisted of 660 healthy subjects whose gender and age matched the case group.Experimental genotyping was performed using the improved multiplex ligation detection reaction(iMLDR)genotyping technique,and the genotyping results were interpreted by blind method.(1)The goodness-of-fit chi-square test was used to test whether the genotype distribution in the control population conformed to the Hardy-Weinberg equilibrium.The correlation between inflammatory polymorphisms and GC was analyzed by unconditional logistic regression.According to the goodness of fit of the model,the best fitting genetic model of each polymorphism was determined.(2)Mantel-haenszel method was used to analyze and compare the inter-layer differences of environmental factors.The Cochran-Armitage test was used to evaluate the relationship between the number of mutations in inflammatory gene polymorphism sites and the risk of GC.3.Establishment and verification of scoring rules for the risk assessment of GC(1)The subjects determined by the experimental verification were randomly divided into two groups,two thirds of the subjects were the training group and one third were the verification group.In the training group,multi-factor logistic regression analysis was conducted based on inflammatory gene polymorphism sites,traditional environmental risk factors,combined inflammatory gene polymorphism sites and traditional environmental risk factors,and then risk assessment models were established.After comparing the evaluation indexes among the three risk models,the optimal model was selected as the basis for the establishment of the scoring rule for the individualized evaluation of GC risk.(2)Using receiver operating characteristic curve(ROC)and area under curve(AUC),net reclassification index(NRI),Brier score(BS),Hosmer-Lemeshow test Sensitivity(Sen),Specificity(Spe),Accuracy rate(AR),Positive predictive value(PPV),Negative predictive value(NPV),Positive Likelihood Ratio(+LR),Negative Likelihood Ratio(-LR)indicators to evaluate the overall performance of the model,and compare the prediction performance between models.(3)According to the value of regression coefficient in the optimal risk model,the risk assessment scoring rules were established and the internal validation was carried out in the validation group.ROC and AUC were used to evaluate and compare the predictive efficacy of the scoring rules in the training group and the control group.Sen,Spe,AR,PPV,NPV,+LR and-LR were used to evaluate the performance of the scoring rules.Results1.A total of 87 articles and 151 studies of inflammatory polymorphism were included in the quantitative comprehensive analysis of the combined effect size.High quality studies(NOS>7)accounted for 43.71%(66/151).There were 7 sites associated with GC in two or more genetic models,including Cox-2 rs20417,IL-10 rs1800896,IL-17F rs763780,IL-17A rs2275913,IL-17A rs8193036,TGF-BR2 rs3773651,VDR rs731236.The epidemiological evaluation results showed that the top three PARP in the general population were IL-17A rs2275913 GG vs AA(23.17%),IL-10 rs1800896 AA vs GG(12.13%),and TGF-BR2 rs3773651 GG vs AA(8.63%).2.The genotyping results showed that the co-dominant model of Cox-2 rs20417,IL-17A rs2275913,TGF-BR2 rs3773651,IL-17F rs7637 80,and IL-17A rs8193036 were statistically associated with GC.The risk allele of Cox-2 rs20417 was G,and CG genotype increased the risk of GC compared with wild-type CC genotype(OR=1.55,95%CI:1.10-2.19,P=0.012).The risk allele of IL-17A rs2275913 was A,and AA genotype increased the risk of GC compared with wild-type GG genotype(OR=1.43,95%C7:1.05-1.93,P=0.022).The risk allele of TGF-BR2 rs3773651 is A.The AG genotype reduces the risk of GCcompared with AA genotype(OR=0.69,95%CI:0.51-0.93,P=0.015).Risk allele of IL-17F rs763780 was C,and the TC genotype increased the risk of GC compared with the TT genotype of wild type(OR=1.42,95%CI:1.09-1.85,P=0.010).The risk allele of IL-17A rs8193036 was T,and the CT genotype increased the risk of GC compared with the wild-type CC genotype 1.29,95%CI:1.03-1.63,P=0.027).3.The recessive model is the best fitting genetic model for IL-17A rs2275913 and IL-23R rs1884444.The dominant model is the best fitting genetic model of Cox-2 rs20417,TGF-BR2 rs3773651,IL-17A rs8193036 and IL-17F rs763780.The incidence of GC increased with the increase of the number of mutation sites after adjusting for gender,age,family history of GC,smoking and alcohol consumption(Ptrend=3.75 ×10-5).4.The personalized risk prediction model for GC constructed by combining the optimal fitting genetic model of inflammatory gene polymorphism and traditional environmental risk factors had the best evaluation effect(AUC=0.724,Calibration=0.986,BS=0.021,NRI=16.2%,Sen=77.06%,Spe=62.56%,AR=67.95%,PPV=67.6%,NPV=72.9%,+LR=2.06,-LR=0.37).5.The score range of the risk assessment rule for GC was[0,25],and the AUC of the risk assessment rule for GC was 0.713(95%CI:0.682-0.743),P<0.0001,with medium differentiation.The optimal critical value of the total score was 10 points,and the evaluation indexes for predicting the occurrence of GC were:Spe=77.29%,Spe=61.40%,AR=68.30%,+LR=2.00,-LR=0.37,PPV=67.00%,and NPV=72.70%,respectively.The proportion of GC patients in the high risk group of 11-25 points was 67.0%,which was significantly higher than that in the low risk group of 0-10 points(27.3%),P<0.0001,indicating good differentiation.The validation group is consistent with the training group.Conclusions1.The results of meta-analysis showed that COX-2 rs20417,IL-10 rs1800896,IL-17F rs763780,IL-17A rs2275913,IL-17A rs8193036,TGF-BR2 rs3773651 and VDR rs731236 may have real associations with GC.2.Genotyping experiments confirmed that COX-2 rs20417,IL-17F rs763780,IL-17A rs2275913,IL-17A rs8193036,TGF-BR2 rs3773651 and IL-23R rs1884444 were associated with the risk of GC.3.A comprehensive GC risk prediction model constructed by incorporating inflammatory-related gene polymorphisms into a traditional environmental risk factor GC risk prediction model has optimized the predictive power and calibration of the traditional environmental risk factor model.4.The risk assessment scoring rules established based on the comprehensive GC risk prediction model have good discrimination.The internal verification results of the scoring rules are consistent.This scoring rule is of great significance for the early screening of high-risk groups of GC and the detection and control of individual GC risk.