胃食管反流病与特定部位食管外癌症之间关联的遗传学证据：一项孟德尔随机化研究。

文献信息

中文摘要

背景： 观察性研究提示胃食管反流病（GERD）患者食管外癌症风险增加，但这些关联是否具有因果性仍不确定。本研究系统评估了遗传预测的GERD与22种食管外癌症之间的潜在关联。方法：GERD的遗传工具来自一项大型全基因组关联研究，包括129,080例病例和473,524例对照。22种特定部位食管外癌症的结局数据来自多个全基因组关联研究数据集，包括英国生物银行、FinnGen和国际癌症联盟。进行了两样本孟德尔随机化（MR）分析，随后使用荟萃分析合并数据集特异性MR估计值。对于阳性发现，进一步进行了反向MR和多变量孟德尔随机化分析，以评估因果方向以及独立于体重指数、吸烟强度和饮酒的情况。结果：两样本MR的荟萃分析显示，遗传预测的GERD与肺癌（比值比[OR]=1.220；95%置信区间[CI]：1.100-1.354；P=.0002）和胰腺癌（OR=1.517；95%CI：1.214-1.895；P=.0002）存在Bonferroni显著的正向关联。喉癌（OR=1.774；95%CI：1.091-2.884；P=.021）和膀胱癌（OR=1.254；95%CI：1.047-1.501；P=.014）也观察到提示性正向关联，但均未通过Bonferroni校正。在考虑体重指数、吸烟强度和饮酒的多变量孟德尔随机化分析中，与肺癌的关联仍具有统计学显著性（OR=1.190；95%CI：1.040-1.362；P=.012），而与胰腺癌、喉癌和膀胱癌的关联不再具有统计学显著性。结论：这些发现提供了遗传学证据，支持遗传预测的GERD与肺癌之间的关联。胰腺癌、喉癌和膀胱癌的证据不太稳健，需要在更大的独立数据集中进一步验证。这些发现可能有助于优先开展针对GERD个体癌症风险的未来研究，并需要进一步研究来验证观察到的关联并阐明其潜在机制。

英文摘要

Observational studies have suggested an increased risk of extraesophageal cancers in patients with gastroesophageal reflux disease (GERD), but whether these associations are causal remains uncertain. This study systematically evaluated the potential associations between genetically predicted GERD and 22 extraesophageal cancers. Genetic instruments for GERD were obtained from a large genome-wide association study including 129,080 cases and 473,524 controls. Outcome data for 22 site-specific extraesophageal cancers were obtained from multiple genome-wide association study datasets, including the UK Biobank, FinnGen, and international cancer consortia. Two-sample Mendelian randomization (MR) analyses were performed, and dataset-specific MR estimates were subsequently combined using meta-analysis. For positive findings, reverse MR and Multivariable Mendelian randomization analyses were further conducted to assess causal direction and independence from body mass index, smoking intensity, and alcohol consumption. Meta-analysis of two-sample MR showed Bonferroni-significant positive associations of genetically predicted GERD with lung cancer (odds ratio [OR] = 1.220; 95% confidence interval [CI]: 1.100-1.354; P = .0002) and pancreatic cancer (OR = 1.517; 95% CI: 1.214-1.895; P = .0002). Suggestive positive associations were also observed for laryngeal cancer (OR = 1.774; 95% CI: 1.091-2.884; P = .021) and bladder cancer (OR = 1.254; 95% CI: 1.047-1.501; P = .014), although neither survived Bonferroni correction. In Multivariable Mendelian randomization analyses accounting for body mass index, smoking intensity, and alcohol consumption, the association with lung cancer remained statistically significant (OR = 1.190; 95% CI: 1.040-1.362; P = .012), whereas the associations with pancreatic, laryngeal, and bladder cancers were no longer statistically significant. These findings provide genetic evidence supporting an association between genetically predicted GERD and lung cancer. The evidence for pancreatic, laryngeal, and bladder cancers was less robust and requires further validation in larger independent datasets. These findings may help prioritize future research on cancer risk among individuals with GERD, with further studies warranted to validate the observed associations and clarify their underlying mechanisms.