Abstract:
Salivary microRNAs (miRNAs) are promising non-invasive biomarkers for the early detection of oral squamous cell carcinoma (OSCC), but variability in normalization strategies limits their reproducibility and clinical translation. This PRISMA-guided systematic review (PROSPERO: CRD420261465738) searched PubMed, Scopus, Cochrane Library, and Google Scholar to evaluate the diagnostic accuracy of salivary miRNAs for OSCC and to assess the impact of normalization strategies on diagnostic performance. Of 602 records screened, 10 studies were included in the systematic review, and 5 studies (7 diagnostic datasets) underwent quantitative synthesis using a bivariate random-effects (Reitsma) diagnostic meta-analysis to estimate pooled sensitivity, specificity, SROC, AUC, PLR, NLR, and DOR. Heterogeneity, meta-regression, influence analysis, and publication bias were assessed using the mada and metafor packages in R. Pooled sensitivity and specificity were 79.9% (95% CI: 68.2–88.0%) and 79.7% (95% CI: 65.2–89.2%), respectively, with an AUC of 0.854. The pooled PLR, NLR, and DOR were 3.73, 0.26, and 16.30, respectively. Meta-regression identified normalization strategy as the main source of heterogeneity (QM = 13.76, p = 0.001), with multiple endogenous reference miRNAs significantly improving diagnostic performance compared with conventional normalization methods. Salivary miRNAs demonstrate good diagnostic accuracy for OSCC, and standardized normalization using multiple endogenous reference miRNAs may improve diagnostic performance and enhance the reproducibility and clinical translation of salivary miRNA-based diagnostics.
