Upper aerodigestive tract (UADT malignancies) encompass cancers of the oral cavity, pharynx, larynx and related anatomical regions and contribute substantially to global morbidity and mortality. Conventional diagnostic approaches for these cancers are often invasive, costly, and commonly identify disease at advanced stages, which limits opportunities for early intervention. MicroRNAs (miRNA) are small non-coding RNAs that regulate gene expression post-transcriptionally and demonstrate notable stability in biological fluids. Dysregulated miRNA expression profiles have been reported across multiple cancers, raising interest in their potential as minimally invasive diagnostic biomarkers for UADT malignancies.
Despite growing literature on miRNA signatures, reported diagnostic performance measures vary across studies. The inconsistency in sensitivity, specificity and other accuracy metrics complicates appraisal of clinical utility. This protocol describes a systematic review and meta-analysis designed to synthesize existing evidence on the diagnostic accuracy of miRNA assays for histopathologically confirmed UADT cancers, quantify pooled performance estimates, and evaluate factors that may explain between-study heterogeneity.
This review will be conducted and reported in accordance with the Preferred Reporting Items for a Systematic Review and Meta-analysis of Diagnostic Test Accuracy Studies (PRISMA-DTA) guidelines. Methodological conduct will follow recommendations from the Cochrane Collaboration Handbook for Diagnostic Test Accuracy Reviews. These frameworks will guide study selection, critical appraisal, data extraction, synthesis, and reporting to ensure transparency and reproducibility.
Eligible studies are those that evaluate the diagnostic performance of miRNA biomarkers for UADT malignancies that are confirmed by histopathology. The review will include primary diagnostic accuracy studies without restriction by publication date. Specific inclusion and exclusion criteria (for example, study designs accepted, minimum sample size thresholds, or language restrictions) were not detailed in the source text and therefore are not reported here.
Electronic databases to be searched systematically from inception are PubMed, Scopus, Web of Science and Embase. The protocol indicates comprehensive database coverage to capture relevant diagnostic accuracy studies of miRNA in UADT cancers. Details of search terms, date limits, and any planned hand-searching or grey literature searches were not provided in the source summary and are not reported here.
Two authors will independently perform study selection, data extraction and quality assessment to minimize bias and errors. Quality assessment of included diagnostic accuracy studies will be undertaken using the Quality Assessment of Diagnostic Accuracy Studies tool (QUADAS-2). Independent dual review will allow resolution of disagreements by consensus or by involving a third reviewer, although the exact adjudication process was not specified in the source text.
Extracted data will include diagnostic accuracy measures and study characteristics necessary for meta-analysis. The source text did not enumerate the specific data fields or forms to be used, so those details are not reported here.
A diagnostic accuracy meta-analysis will be conducted using a hierarchical bivariate random-effects model (Reitsma model) to jointly pool estimates of sensitivity and specificity while accounting for both within-study and between-study variability and threshold effects. Hierarchical summary receiver operating characteristic (HSROC) curves will be constructed. Pooled estimates to be reported include sensitivity, specificity, positive likelihood ratio, negative likelihood ratio and diagnostic odds ratio with 95% confidence intervals.
The use of the Reitsma bivariate model and HSROC approach aligns with best practice for synthesizing diagnostic accuracy data, allowing joint consideration of the trade-off between sensitivity and specificity across diverse studies.
To investigate sources of heterogeneity, meta-regression will be performed followed by subgroup analyses. Potential effect modifiers identified in the protocol include specimen type (for example, tissue versus blood or other fluids), miRNA profiling method, study design and patient characteristics. The source summary does not specify the exact covariates or categories within these domains, nor the statistical thresholds planned for meta-regression, so those methodological specifics are not reported here.
This review involves synthesis of previously published literature and does not require collection of individual participant data or direct participant involvement; as such, no new ethics approval is required according to the protocol description. Findings will be disseminated through publication in a peer-reviewed journal and presentations at national and international scientific forums.
The review is registered in PROSPERO with registration number CRD420261384413, as reported in the source text.