Aim: This study aimed to critically appraise the evidence of the diagnostic effectiveness of miRNAs for the detection of cervical cancer. Methods & materials: A systematic review and meta-analysis was performed, searching PubMed, EMBASE and Web of Science. An umbrella meta-analysis of meta-analyses of individual biomarkers was performed. A Grading of Recommendations, Assessment, Development and Evaluations (GRADE) assessment of evidence was also performed. Results: A total of 52 miRNAs were included. Umbrella meta-analysis revealed significant heterogeneity in terms of sensitivity, specificity, receiver operating characteristic (ROC), positive predictive value and/or negative predictive value. Umbrella effects were 0.76 (95% CI: 0.73-0.78), 0.78 (95% CI: 0.75-0.81), 0.77 (95% CI: 0.75-0.80), 0.75 (95% CI: 0.71-0.79) and 0.76 (95% CI: 0.74-0.79), respectively. Conclusion: Moderate quality evidence suggested miR199a-5p, miR21-5p and miR-141a had excellent diagnostic performance.
Keywords: ROC; biomarker; cervical cancer; cervical intraepithelial neoplasia; meta-analysis; miRNA sensitivity; negative predictive value; positive predictive value; specificity.
miRNAs are small chemical messengers that play a role in the regulation of protein produced inside the cytoplasm of cells, including cancer cells. In cervical cancer cells, miRNAs become dysregulated – that is, their levels become increased or decreased – and therefore detecting their relative abundance or absence in test samples may enable identification of cervical cancer. This study aimed to systematically collect and appraise the evidence for the diagnostic ability of miRNAs for detection of cervical cancer. A systematic appraisal of the evidence was conducted by searching three research databases (PubMed, EMBASE and Web of Science) to collect evidence published up to 13 November 2022. Results for diagnostic performance of 52 miRNAs were extracted from 20 relevant studies. An assessment of risk of bias for each study was performed using a standardized checklist, which identified one high-quality study, 18 moderate-quality studies and one low-quality study. Results for each individual biomarker were assessed by meta-analytic methods, which generated weighted averages for 38 of 52 miRNAs. All 52 miRNAs were then compared using an umbrella meta-analysis (a weighted average of all miRNA biomarkers), which identified significant differences in diagnostic ability between miRNAs. Sensitivity analyses suggested that these differences were partly explained by differences in grades of cervical cancer and differences in types of sample used for testing. A GRADE assessment of the overall evidence quality suggested that moderate-quality evidence supported further investigation of three miRNA biomarkers (miR-199a-5p, miR-21-5p and miR-141a), which performed excellently (i.e., better than the umbrella weighted average) across five performance parameters, including sensitivity, specificity, receiver operator characteristic, positive predictive value and negative predictive value. In summary, this study suggested miR-199a-5p, miR-21-5p and miR-141a had excellent diagnostic performance for detection of cervical cancer and recommends further investigation of these miRNAs in randomized controlled trials.