Can MicroRNAs serve as reliable biomarkers and novel therapeutic targets in age-related macular degeneration? A systematic review and meta-analysis

Age-related macular degeneration (AMD) is a leading cause of irreversible vision loss, affecting 196 million people globally with projections reaching 288 million by 2040. Its pathogenesis involves oxidative stress, chronic inflammation, angiogenesis, and retinal pigment epithelium (RPE) dysfunction. MicroRNAs (miRNAs)—highly stable, non-coding post-transcriptional regulators—have emerged as key modulators of AMD-related pathways, including nuclear factor kappa B mediated inflammation (miR-146a, miR-155), angiogenesis (miR-23a, miR-27a), and RPE homeostasis (miR-204, miR-211). Dysregulation of miRNAs across the retina, serum, plasma, and vitreous may offer diagnostic and therapeutic potential. This systematic review and meta-analysis evaluated the diagnostic accuracy, differential expression, and functional relevance of miRNAs in AMD. A comprehensive literature search was performed across all major databases, including PubMed, Embase, Web of Science, Google Scholar, Scopus, and the Cochrane Library, as well as additional sources identified through websites and citation searching, covering all records from database inception to 15th June 2025, following the Preferred Reporting Items for Systematic Reviews and Meta-Analyses guidelines (PRISMA), with protocol registration in the International Prospective Register of Systematic Reviews (PROSPERO: CRD420251101982). Eligible studies included human case-control studies, in vitro/in vivo AMD models, and miRNA-based therapeutic investigations. Risk of bias was assessed using the Newcastle-Ottawa Scale (NOS), the Office of Health Assessment and Translation tool (OHAT), the SYstematic Review Centre for Laboratory animal Experimentation tool (SYRCLE), and a modified Toxicological data Reliability Assessment Tool (ToxRTool). Meta-analyses were performed using fixed or random-effects models based on heterogeneity. Outcomes included pooled AUC values, standardized mean differences (SMD), pooled mean miRNA expression, and distribution of up- vs. down-regulated miRNAs. Twenty-two studies were included. Differentially expressed miRNAs were consistently reported, with miR-34a (up to 6.3-fold), miR-146a (2.1–6.3-fold), miR-155, miR-19a, miR-410, and miR-126 among the most recurrent. Diagnostic accuracy meta-analysis demonstrated a pooled area under the receiver operating characteristic curve of 0.749 (95% CI: 0.713–0.781). Individual AUCs included: vitreous ratio AUC 0.977, plasma AUC 0.914, miR-27a-3p 0.925, miR-29b-3p 0.757, and miR-195-5p 0.766. Pooled SMD comparing AMD vs. controls was 0.296 (p < 0.001). Mean pooled miRNA expression in AMD was 0.296 (SE 0.0313), with highest levels in miR-486-5p (0.8016), miR-423-3p (0.7329), and miR-885-5p (0.6759). Up-vs-down-regulation analysis showed balanced distribution (pooled effect 0.520, p = 0.505). Functional studies demonstrated restoration of TREM2 after miR-34a inhibition and rescue of RPE phagocytosis via miR-184 modulation. Publication bias was significant in two outcomes but corrected estimates remained directionally consistent. Multiple miRNAs show consistent dysregulation in AMD and display moderate diagnostic accuracy, particularly miR-27a-3p, miR-146a, miR-34a, and vitreous-derived miRNA panels. Several miRNAs regulate key pathogenic pathways—including inflammation, angiogenesis, and impaired RPE function—supporting their potential as biomarkers and therapeutic targets. Large, standardized, multi-center studies are required to validate these findings and advance miRNA-based diagnostics and therapeutics toward clinical translation. This study bridges the gap between molecular biomarker discovery and clinical application in AMD. It highlights the feasibility of incorporating miRNA profiling into AMD diagnostics and supports the future development of miRNA-targeted therapies for personalized treatment strategies.