基于两样本孟德尔随机化法的不同吸烟状态下COPD与肺癌患病风险的因果关系研究
Causal association between chronic obstructive pulmonary disease and lung cancer risk across smoking statuses: a two-sample Mendelian randomization study
-
摘要:目的 利用两样本孟德尔随机化方法系统评估不同吸烟状态分层的慢性阻塞性肺疾病(chronic obstructive pulmonary disease, COPD)患者与肺癌及其病理亚型间的因果关联, 为高危人群筛查和精准防控提供依据。方法 从公开的全基因组关联研究(genome-wide association studies, GWAS)汇总数据中, 分别纳入非吸烟COPD患者(病例8 631例, 对照120 544例)和当前吸烟COPD患者(病例4 589例, 对照10 001例)的遗传数据, 以及肺癌(病例11 348例, 对照15 861例)、肺腺癌(病例3 442例, 对照14 894例)和肺鳞癌(病例3 275例, 对照15 038例)的遗传数据。筛选全基因组水平显著关联(P < 5×10-8)且无连锁不平衡(r2 < 0.6)的单核苷酸多态性(single nucleotide polymorphism, SNP)作为工具变量。采用逆方差加权法(inverse-variance weighted, IVW)为主分析方法, 加权中位数法(weighted median, WM)和简单众数法(simple mode, SM)为辅, 用比值比(OR)及其95%置信区间(CI)评估COPD对肺癌风险的因果效应。通过孟德尔随机化Egger(mendelian randomization-Egger, MR-Egger)截距法、孟德尔随机化多效性残差和与离群值检验法(mendelian randomization pleiotropy residual sum and outlier, MR-PRESSO)和留一法(leave-one-out)等方法进行敏感性分析。结果 IVW结果显示, 非吸烟COPD患者与肺癌和肺腺癌风险呈显著正相关, OR分别为1.13(95%CI: 1.07~1.20, P < 0.001)和1.18 (95%CI: 1.09~1.28, P < 0.001), 但与肺鳞癌无显著关联。当前吸烟COPD患者与肺腺癌和肺鳞癌患病显著相关, OR分别为1.15 (95%CI: 1.01~1.31, P=0.031)和1.76 (95%CI: 1.55~1.99, P < 0.001)。当前吸烟COPD患者的肺癌风险高于非吸烟COPD患者的风险。敏感性分析显示, 有1组分析纳入的SNP间存在异质性(P < 0.05), 对存在异质性的采用随机效应IVW模型用以克服异质性干扰。MR-Egger法未检测出水平多效性, 留一法未发现有显著影响的SNP。敏感性分析表明结果稳健。结论 研究发现COPD患者与罹患肺癌之间存在因果关联, 且吸烟行为会增加COPD患者的肺癌风险。Abstract:Objective To systematically evaluate the causal associations between chronic obstructive pulmonary disease (COPD) stratified by smoking status and lung cancer and its pathological subtypes using two-sample Mendelian randomization analysis, and to provide evidence for high-risk population screening and precise prevention.Methods Genetic data were extracted from publicly available genome-wide association study (GWAS) summary statistics, including never-smoking COPD patients (8 631 cases, 120 544 controls), current-smoking COPD patients (4 589 cases, 10 001 controls), lung cancer (11 348 cases, 15 861 controls), lung adenocarcinoma (3 442 cases, 14 894 controls), and lung squamous cell carcinoma (3 275 cases, 15 038 controls). Single-nucleotide polymorphisms (SNPs) with genome-wide significant associations (P < 5×10-8) and without linkage disequilibrium (r2 < 0.6) were selected as instrumental variables. The inverse-variance weighted (IVW) method served as the primary analytical approach, complemented by the weighted median and the simple mode methods. Causal effects of COPD on lung cancer risk were estimated using odds ratios (ORs) with 95% confidence intervals (CIs). Sensitivity analyses were performed using the Mendelian randomization-Egger intercept test, the Mendelian randomization pleiotropy residual sum and outlier test, and the leave-one-out test.Results IVW analyses revealed significant positive associations between never-smoking COPD and risks of lung cancer (OR=1.13, 95%CI: 1.07-1.20, P < 0.001) and lung adenocarcinoma (OR=1.18, 95%CI: 1.09-1.28, P < 0.001), but no significant association with lung squamous cell carcinoma. Current-smoking COPD was significantly associated with increased risks of lung adenocarcinoma (OR=1.15, 95%CI: 1.01-1.31, P=0.031) and lung squamous cell carcinoma (OR=1.76, 95%CI: 1.55-1.99, P < 0.001). The risk of lung cancer in current-smoking COPD patients was higher than that in never-smoking counterparts. Sensitivity analyses showed heterogeneity in one set of SNPs (P < 0.05). For the heterogeneous dataset, a random-effects IVW model was applied to mitigate heterogeneity. Mendelian randomization-Egger intercept test indicated no horizontal pleiotropy, and leave-one-out test identified no outlier SNPs. Sensitivity analyses confirmed the robustness of the result.Conclusion This study provides evidence for causal associations between COPD and lung cancer, with smoking behavior amplifying the lung cancer risk in COPD patients.
下载: