Independent Validation of a Diagnostic Noninvasive 3-MicroRNA Ratio Model () for Prostate Cancer in Cell-Free Urine.

BACKGROUND

Detection of prostate cancer (PC) based on serum prostate-specific antigen (PSA) testing leads to many unnecessary prostate biopsies, overdiagnosis, and overtreatment of clinically insignificant tumors. Thus, novel and more accurate molecular biomarkers are required.

METHODS

Using reverse transcription quantitative PCR, we measured the concentrations of 45 preselected microRNAs (miRNAs) in extracellular vesicle-enriched cell-free urine samples from 4 independent patient cohorts from Spain and Denmark, including 758 patients with clinically localized PC, 289 noncancer controls with benign prostatic hyperplasia (BPH), and 233 patients undergoing initial transrectal ultrasound (TRUS)-guided prostate biopsy owing to PC suspicion (101 with benign and 132 with malignant outcome). Diagnostic potential was assessed by ROC and decision curve analysis.

RESULTS

We identified and successfully validated 8 upregulated and 21 downregulated miRNAs in urine from PC patients. Furthermore, we validated a previously identified 3-miRNA diagnostic ratio model, (miR-222-3p*miR-24-3p/miR-30c-5p). High scores were distinctive of PC in urine samples from BPH vs PC patients in 3 independent cohorts [area under the curve (AUC) = 0.84, 0.71, 0.72]. Additionally, predicted TRUS biopsy results with greater accuracy than PSA (AUC = 0.644; AUC PSA = 0.527) for patients within the diagnostic gray zone (PSA ≤ 10 ng/mL).

CONCLUSIONS

We successfully validated a urine-based diagnostic 3-miRNA signature for PC () in 3 independent patient cohorts from 2 countries. In the future, the simple and noninvasive test may be used to help more accurately select patients for prostate biopsy. Prospective clinical validation is warranted.