首页|两样本孟德尔随机化分析炎性因子与前列腺癌因果关系

两样本孟德尔随机化分析炎性因子与前列腺癌因果关系

扫码查看
目的:采用两样本孟德尔随机化方法来评估 91 种炎性因子与前列腺癌的潜在因果关系.方法:选取91 种炎性因子的全基因组关联分析(GWAS)汇总统计数据(n=14 824)结合FinnGen最新第9 版数据库中选取前列腺癌作为结局进行孟德尔随机化分析.通过逆方差加权法(IVW)、MR-Egger回归、简单中位数法(SM)、加权中位数法(WM)、加权中值法(WME)等回归模型的OR值和95%CI评估炎性因子与前列腺癌的因果关系,其中IVW法得出的因果关系相对稳定,作为本研究的主要统计方法.进一步采用贝叶斯分析法对孟德尔分析结果进行验证.使用Cochran's Q检验评估遗传工具变量的异质性,使用MR-Egger截距检验评估多效性,使用留一法评估作为工具变量的单核苷酸多态性(SNPs)对暴露和结局因果关系影响的敏感性.结果:IVW法结果显示,91 种炎性因子中白介素-22 受体 A1(IL-22RA1)、磺基转移酶 1A1(ST1A1)与前列腺癌发病风险呈正向因果关联:IL-22RA1:IVW[OR(95%CI):1.12(1.00~1.25),P=0.04];ST1A1:IVW[OR(95%CI):1.08(1.00~1.16),P=0.03].趋化因子配体 11(CXCL11)、白介素 17A(IL-17A)与前列腺癌发病风险呈反向因果关联;CXCL11:IVW[OR(95%CI):0.88(0.81~0.95),P=0.00];IL-17A:IVW[OR(95%CI):0.91(0.84~0.98),P=0.02].MR-Egger截距分析未检测到潜在的水平多效性(P 均>0.05,IL-22RA1=0.885,ST1A1=0.949,CXCL11=0.391,IL-17A=0.884).MR-PRESSO 未检测到偏倚 SNPs(P 均>0.05,IL-22RA1=0.479,ST1A1=0.629,CXCL11=0.326,IL-17A=0.444),未发现异质性(P均>0.05,IL-22RA1=0.543,ST1A1=0.677,CXCL11=0.336,IL-17A=0.494),留一法敏感性分析结果图提示无个别单核苷酸多态性位点对整体因果关系预测产生明显影响,故分析结果可靠.结论:91 种炎性因子中IL-22RA1、ST1A1 与前列腺癌有正向因果关系,随着这些因子的水平增加,前列腺癌发病的风险升高;CXCL11、IL-17A与前列腺癌有反向因果关系,即随着这些因子的水平增加,前列腺癌发病风险降低.
Inflammatory factors and prostate cancer:Two-sample Mendelian randomization analysis
Objective:To evaluate the potential causal relationship between inflammatory factors and PCa using the two-sample Mendelian randomization(MR)method.Methods:We selected summary statistics of genome-wide association studies(GWAS)(n=14 824)on 91 inflammatory factors,with PCa as the outcome in the latest 9th edition of FinnGen database for MR analysis.We evaluated the causal relationship between inflammatory factors and PCa using the odds ratio(OR)and 95%confidence interval(CI)of such regression models as inverse variance weighting(IVW),MR-Egger regression,simple mode(SM),weighted mode(WM)and weighted median estimator(WME),with IVW as the main statistical method for this study.We further verified the results of MR by Bayesian analysis,and evaluated the heterogeneity of genetic instrumental variables,pleiotropic effects and sensitivity of single nu-cleotide polymorphisms(SNP)as instrumental variables to the exposure-outcome relationship by Cochran's Q test,MR-Egger intercept test and leave-one-out cross validation.Results:IVW showed that among the 91 inflammatory factors,interleukin-22 receptor A1(IL-22RA1)and sulfotransferase 1A1(ST1A1)were correlated positively with the risk of PCa;IL-22RA1:IVW(OR[95%CI]:1.12[1.00-1.25],P=0.04);ST1A1:IVW(OR[95%CI]:1.08(1.00-1.16),P=0.03),while Chemokine ligand 11(CXCL11)and interleukin 17 A(IL-17 A)negatively with the risk of PCa;CXCL11:IVW(OR[95%CI]:0.88[0.81-0.95],P=0.00);IL-17A:IVW(OR[95%CI]:0.91[0.84-0.98],P=0.02).No potential horizontal pleiotropy was detected by MR-Egger intercept analysis(P>0.05,IL-22RA1=0.885,ST1A1=0.949,CXCL11=0.391,IL-17A=0.884),nor biased SNPs in the MR pleiotropy residual sum and outlier(MR-PRESSO)test(P>0.05,IL-22RA1=0.479,ST1A1=0.629,CXCL11=0.326,IL-17A=0.444),or heterogeneity P>0.05,IL-22RA1=0.543,ST1A1=0.677,CXCL11=0.336,IL-17A=0.494).Leave-one-out sensitivity analysis indicated no significant impact of individual SNP sites on the overall causal rela-tionship prediction,suggesting the reliable results of analysis.Conclusion:Among the 91 inflammatory factors,IL-22RA1 and ST1A1 have a positive causal relationship,while CXCL11 and IL-17A have a negative causal relationship with PCa.

inflammatory factorsprostate cancerMendelian randomizationsensitivity analysissingle nucleotide polymor-phismBayesian analysis

曾银、张甘霖、郭军、杨梦萍、韩强、杨国旺

展开 >

首都医科大学附属北京中医医院,北京 100010

中国中医科学院西苑医院男科,北京 100091

北京中医药大学第三临床医学院,北京 100029

炎性因子 前列腺癌 孟德尔随机化分析 敏感性分析 单核苷酸多态性 贝叶斯分析

北京市市属医院科研培育项目中医药传承与创新"百千万"人才工程(岐黄学者)项目

PZ 2023016020450008

2024

中华男科学杂志
南京军区南京总医院

中华男科学杂志

CSTPCD
影响因子:1.052
ISSN:1009-3591
年,卷(期):2024.30(7)
  • 8