Abstract
Single nucleotide polymorphisms in genes involved in microRNA processing/maturation and release may deregulate the microRNAome expression levels. We aimed to assess the relationship between miRNA machinery genetic variants and human cancer risk using integrative bioinformatics analyses to identify the role of these genes in cancer aggressiveness. Mutations of 8176 pan-cancer samples were retrieved from 33 studies in "TCGA" database, and a Cox regression model for survival was performed. Next, 22 computationally identified variants within 11 genes were selected based on their high citation rate and MAF. Relevant articles through March 2020 were included. Pooled estimates under the five genetic association models were calculated. Publication bias and heterogeneity between articles were evaluated. Trial Sequential Analysis (TSA) was applied to assess the power and reliability of the draw conclusions. TCGA patients with different cancer types revealed significant alterations in miRNA machinery genes, with mutation frequency ranging from 0.6-13% of samples. RAN was associated with LN metastasis, while TARBP2 and PIWIL1 gene mutations exhibited better overall survival. In the meta-analysis, 45 articles (74,593 cases and 89,198 controls) met the eligibility criteria. Pooled analysis revealed an increased cancer risk with DROSHArs10719*G, RANrs3803012*G, DGCR8rs417309*A, and GEMIN3rs197414*A. In contrast, both DICER1rs1057035*T and GEMIN4rs2743048*G conferred protection against developing cancer. TSA showed the cumulative evidence is inadequate, and the addition of further primary studies is necessary. This study suggests a potential role of miRNA biogenesis genes in cancer development/prognosis. Further functional studies may reveal biological explanations for the differential risks of the machinery variants in different cancer types.