Zhang Xinxing, Li Zhen
Chengdu New Radiomedicine Technology Co. Ltd, Chengdu, Sichuan, China.
Department of Urology, Hunan Provincial People's Hospital, The First Affiliated Hospital of Hunan Normal University, Changsha, Hunan, China.
Diabetol Metab Syndr. 2024 Jul 12;16(1):159. doi: 10.1186/s13098-024-01397-7.
The relationship between metformin use and prostate cancer (PCa) risk has yet to be clear despite more than a decade of debate on this topic. Hence, we aimed to investigate the causal role of metformin in reducing PCa risk through an up-to-date comprehensive genome-wide analysis.
We employed validated instrument variables of metformin use derived from a prior high-quality study, including five potential targets (AMPK, GCG, GDF15, MCI and MG3). Mendelian randomization (MR) analysis was performed to harmonize genetically predicted metformin use and PCa phenotypes. PCa phenotypes were from two large genome-wide association studies (GWAS), the Prostate Cancer Association Group to Investigate Cancer-Associated Alterations in the Genome (PRACTICAL) and the FinnGen cohort. Seven methods were applied to generate MR results: the inverse variance weighted (IVW), IVW with multiplicative random effects, MR-Egger, MR-Egger (bootstrap), weighted median, simple mode and weighted mode. Strict sensitivity analysis was conducted to satisfy core assumptions of MR design.
We enrolled 32 significant single nucleotide polymorphisms (SNPs) that involved with metformin use. Nearly all targets yielded insignificant primary results (IVW with multiplicative random effects), except that AMPK target posed a positive effect on PCa risk from FinnGen cohort [odds ratio (OR): 6.09, 95% confidence interval (CI): 1.10-33.53, P value: 0.038]. The general effect of metformin use, comprising all 5 targets, also yielded negative results (random-effect meta-analysis with OR: 1.09, 95% CI: 0.76-1.58, P value: 0.637 for PRACTICAL; OR: 2.55, 95% CI: 0.58-11.16, P value: 0.215 for FinnGen). None of the sensitivity analyses provided support for a causal association between metformin use and PCa risk.
This up-to-date study did not support the protective role of metformin in reducing PCa risk, considering each target, overall effect, and sensitivity analysis. It is imperative to reflect on the presumed "almighty medicine" and ongoing phase III trials are anticipated to assess the anti-neoplasm effect of metformin.
尽管关于二甲双胍使用与前列腺癌（PCa）风险之间的关系已争论了十多年，但仍不明确。因此，我们旨在通过最新的全基因组综合分析来研究二甲双胍在降低PCa风险中的因果作用。
我们采用了先前高质量研究中得出的经验证的二甲双胍使用工具变量，包括五个潜在靶点（AMPK、GCG、GDF15、MCI和MG3）。进行孟德尔随机化（MR）分析，以协调基因预测的二甲双胍使用情况和PCa表型。PCa表型来自两项大型全基因组关联研究（GWAS），即前列腺癌基因组关联研究小组（PRACTICAL）和芬兰基因队列。应用七种方法生成MR结果：逆方差加权法（IVW）、带乘性随机效应的IVW、MR-Egger法、MR-Egger法（自助法）、加权中位数法、简单模式法和加权模式法。进行了严格的敏感性分析，以满足MR设计的核心假设。
我们纳入了32个与二甲双胍使用相关的显著单核苷酸多态性（SNP）。几乎所有靶点的主要结果均无统计学意义（带乘性随机效应的IVW），只有AMPK靶点对芬兰基因队列中的PCa风险有正向影响[比值比（OR）：6.09，95%置信区间（CI）：1.10 - 33.53，P值：0.038]。包括所有5个靶点的二甲双胍使用的总体效应也得出了阴性结果（PRACTICAL的随机效应荟萃分析，OR：1.09，95%CI：0.76 - 1.58，P值：0.637；芬兰基因队列的OR：2.55，95%CI：0.58 - 11.16，P值：0.215）。敏感性分析均未支持二甲双胍使用与PCa风险之间的因果关联。
考虑到各个靶点、总体效应和敏感性分析，这项最新研究不支持二甲双胍在降低PCa风险方面的保护作用。有必要反思这种被假定的“万能药物”，预计正在进行的III期试验将评估二甲双胍的抗肿瘤作用。
