Circulating microRNAs: a new era in early detection of oral squamous cell carcinoma (OSCC) — a systematic review and meta-analysis

Abstract

BACKGROUND: The chances of oral squamous cell carcinoma (OSCC) being diagnosed at an early stage are very low, and this leads to a poor prognosis. The presence of circulating miRNAs (miRNAs) in blood or saliva is now a potential non-invasive diagnostic biomarker, although the diagnostic accuracy of the studies reported remains vastly different due to differences in biofluids, miRNA panels, disease spectrum, and platforms. METHODS: We did a meta-analysis of diagnostic accuracy and a systematic review in accordance with PRISMA-DTA 2020 guidelines (PROSPERO: CRD420251274288). Articles that evaluated circulating miRNAs in any biofluid in diagnosing OSCC were taken into consideration. A Bayesian bivariate random-effects model was used to pool together the sensitivity, specificity, likelihood ratios, diagnostic odds ratios and the summary receiver operating characteristic (sROC) curves. Heterogeneity sources were analyzed using pre-constructed subgroup analysis and meta-regression. RESULTS: The number of studies that were evaluated in totality was seventeen and seven were incorporated in the meta-analysis. The pooled sensitivity of the circulating miRNAs towards the detection of OSCC was 0.89 (95% CI: 0.77-0.95) and the pooled specificity was 0.88 (95% CI: 0.75-0.95). It had a pooled positive likelihood ratio (PLR) of 7.25 (95% CI: 3.42-16.08), a negative likelihood ratio (NLR) of 0.13(95% CI: 0.05-0.27), and a diagnostic odds ratio (DOR) of 59.69(95%CI:14.33-209.95). Good performance in discriminating was also noted in the sROC curve with an area under the curve value of approximately 0.90. However, substantial between-study heterogeneity was observed (τ ≈ 0.93 for sensitivity; τ ≈ 0.83 for specificity), and the 95% credible intervals around the DOR were wide (14.33-209.95). Individual studies of multi-miRNA panels reported AUCs up to 0.98, though pooled subgroup estimates were limited by small study numbers and wide uncertainty. CONCLUSIONS: Circulating miRNAs and in particular multi-miRNA saliva-based profiles provide promising but heterogenous diagnostic accuracy for OSCC. While individual panel studies have reported high AUCs, the current pooled evidence-derived from only seven small, mostly case-control studies of advanced-stage disease-does not yet establish utility for true early detection or population screening.