Abstract
Background: A significant rise is observed in the incidence of early-onset colorectal cancer (EOCRC) worldwide, demanding discovery of promising non-invasive diagnostic biomarkers. This systematic review and meta-analysis evaluated the diagnostic accuracy of single circulating microRNAs (miRNAs) for EOCRC detection. Methods: The protocol for this systematic review was prospectively registered with PROSPERO (registration number: CRD420251252155). We systematically searched PubMed, Embase, Web of Science and Scopus through September 2025 following PRISMA 2020 guidelines. Studies stating diagnostic accuracy of single circulating miRNAs in blood samples for histologically confirmed CRC were included. The quality assessment of included studies was done by using QUADAS-2 and bivariate random-effect meta-analysis was performed to calculate pooled diagnostic metrics. Results: Sixteen studies comprising 909 CRC cases and 1214 controls evaluating 22 distinct miRNAs were included. In the primary meta-analysis restricted to early-onset colorectal cancer (EOCRC, <50 years), pooled sensitivity was 84.4% and specificity was 85.7%. Analyses including mixed-age or all CRC populations were conducted as secondary analyses and showed comparable diagnostic performance. Subgroup analysis showed EOCRC patients (<50 years, n = 15) demonstrated sensitivity of 84.4% and specificity of 85.7%. Substantial heterogeneity existed (I(2) > 85%). Quality assessment revealed high risk of bias in patient selection (87.5%) and index test domains (87.5%). Mechanistic analysis identified key pathways including Wnt/β-catenin, PI3K/AKT and EMT regulation. Sensitivity analyses confirmed robust estimates and Deeks' test (p = 0.99) indicated no publication bias. Conclusion: This study has shown that individual circulating miRNAs provide consistent diagnostic accuracy for early-onset colorectal cancer (EOCRC); however, these findings require prospective validation in true screening settings before clinical implementation.