Background: Prostate cancer (PCa) remains a major clinical challenge due to the limited ability of current diagnostic tools to distinguish indolent from aggressive tumors. MicroRNAs (miRNAs) have emerged as promising biomarkers given their stability, tissue specificity, and regulatory roles in carcinogenesis. This study analyzed miRNA expression and clinical data from The Cancer Genome Atlas–Prostate Adenocarcinoma (TCGA-PRAD) to identify differentially expressed miRNAs and evaluate their association with tumor aggressiveness. Methods: miRNA expression profiles and clinical variables from TCGA-PRAD were examined. A total of 550 samples (498 tumors, 52 adjacent normal) and 1,881 miRNAs were analyzed using the differential expression sequencing 2 workflow. Differential expression was assessed using adjusted significance thresholds. Highly deregulated miRNAs were identified based on log2 fold change ≥2 and false discovery rate (FDR) <0.01. Associations with the Gleason score were evaluated using Spearman correlation with multiple testing correction. Results: We identified 137 significantly deregulated miRNAs (FDR < 0.05), with no global trend toward over-or underexpression. Applying stringent filters yielded 24 miRNAs with high-magnitude changes. Correlation analysis revealed 25 miRNAs significantly associated with the Gleason score. The miR-449a, miR-184, and miR-891a showed both strong deregulation and significant correlation with aggressiveness. miR-449a exhibited the greatest tumor overexpression, whereas miR-184 and miR-891a displayed expression patterns consistent with pathways previously linked to proliferation, invasion, and tumor progression. Conclusion: miR-449a, miR-184, and miR-891a emerge as strong candidates for PCa progression biomarkers. Their combined deregulation and association with the Gleason score support their potential as prognostic indicators and therapeutic targets. Validation in independent cohorts and functional studies will be essential to confirm their clinical applicability.
Bergez-Hernández et al. (2026) studied this question.