Comprehensive profiling of urine microRNAs in prostate cancer patients

The diagnosis of early-stage and locally advanced prostate cancer (PCa) continues to present significant clinical challenges. While prostate-specific antigen (PSA) remains the only serum biomarker currently used for PCa detection, its specificity is low, limiting its effectiveness. Although multiparametric magnetic resonance imaging serves as a valuable diagnostic resource, there are many inherent risks, including missing the target lesion during evaluation. This study was designed to identify urine cell-free microRNAs (cfmiRs) as diagnostic markers for patients with PCa of different stages (pT1a-pT3b). Briefly, urine samples were collected from 100 patients diagnosed with PCa and 22 individuals without PCa. Total RNA was isolated from 1 mL of urine using JBS Cell-free RNA & microRNA Isolation Kit and the automated JpurX-S200 isolation system according to manufacturer's protocol. Purified miRNAs were analyzed using an NGS-based approach to quantify 2,083 distinct microRNAs (miRs). Differential expression analysis using DESeq2 revealed 664 differentially expressed miRs in the urine of PCa patients, among which 95 cfmiRs were upregulated. Machine learning analysis considering both upregulated and downregulated cfmiRs led to the development of 10-cfmiR signature based on importance-scores. The 10-cfmiR signature showed strong performance on receiving operating characteristic (ROC) curve analysis [area under the curve (AUC) = 0.88, specificity = 0.91, sensitivity = 0.72, p < 8.31e-09]. When considering only upregulated cfmiRs, we found a 16-cfmiR and a 54-cfmiR signatures, which demonstrated good diagnostic performance for PCa [AUC = 0.77, specificity = 0.64, sensitivity = 0.84, p < 1.84e-06; AUC = 0.82, specificity = 0.82, sensitivity = 0.73, p<1.90e-06]. The diagnostic accuracy observed for the cfmiR signature was further conserved when focusing specifically on let-7a-5p [AUC = 0.79, specificity = 0.73, sensitivity = 0.76, p < 0.001] and let-7b-5p [AUC = 0.80, specificity = 0.73, sensitivity = 0.69, p < 0.001]. Specific cfmiR signatures were associated with higher grade groups and elevated PSA levels. The cfmiR signatures identified in this study were also compared with the cfmiRs identified in a previous study analyzing a cohort of early-stage pT2 PCa patients. Notably, 109 cfmiRs were detected consistently in both groups, with 93.75% (15 out of 16) of the cfmiRs in the 16-cfmiR signature being upregulated and present in both comparisons. In summary, this study successfully quantified and identified potential cfmiR signatures in urine samples from PCa patients at all stages. Specific urine cfmiRs may offer an advantage to distinguish patients with more aggressive disease. The findings suggest that urine is a robust source of cfmiRs with promising potential for the diagnosis of individuals with PCa.