Polymorphic Single-Nucleotide Variants in miRNA Genes and the Susceptibility to Colorectal Cancer: Combined Evaluation by Pairwise and Network Meta-Analysis, Thakkinstian's Algorithm and FPRP Criterium

Abstract

Background: Considerable epidemiological studies have examined the correlation between polymorphic single-nucleotide variants (SNPs) in miRNA genes and colorectal carcinoma (CRC) risk, yielding inconsistent results. Herein, we sought to systematically investigate the association between miRNA-SNPs and CRC susceptibility by combined evaluation using pairwise and network meta-analysis, the FPRP analysis (false positive report probability), and the Thakkinstian's algorithm. Methods: The MEDLINE, EMBASE, WOS, and Cochrane Library databases were searched through May 2024 to find relevant association literatures. Pooled odds ratios (ORs) and 95% confidence intervals (CIs) were computed by the pairwise meta-analysis. Network meta-analysis and the Thakkinstian's method were applied for determining the potentially optimal genetic models; additionally, the FPRP was used to identify noteworthy associations. Results: Totally, 39 case–control trials involving 18,028 CRC cases, and 21,816 normal participants were included in the study. Eleven SNPs within nine genes were examined for their predisposition to CRC. miR-27a (rs895819) was found to significantly increase CRC risk among overall population (OR 1.58, 95% CI: 1.32–1.89) and Asians (OR 1.62, 95% CI: 1.31–2.01), with the recessive models identified as the optimal models. Furthermore, miR-196a2 (rs11614913), miR-143/145 (rs41291957), and miR-34b/c (rs4938723) were significantly related to reduced CRC risk among Asian descendants under the optimal dominant (OR 0.75, 95% CI: 0.65–0.86), recessive (OR 0.72, 95% CI: 0.60–0.85), and recessive models (OR 0.69, 95% CI: 0.56–0.85), respectively. The results were also proposed by the network meta-analysis or the Thakkinstian's method and confirmed by the FPRP criterion. Conclusion: The miR-27a (rs895819) is correlated with elevated CRC risk among overall population and Asians, and the recessive model is found to be optimal for predicting CRC risk. Additionally, the miR-196a2 (rs11614913), miR-143/145 (rs41291957), and miR-34b/c (rs4938723), with the dominant, recessive, and recessive models identified as the optimal, might confer protective effects against CRC among Asians.