Abstract
The main advantage of urinary biomarkers is their noninvasive character and the ability to detect multifocal prostate cancer (CaP). We have previously implemented a quadruplex assay of urinary markers into clinical practice (PCA3, AMACR, TRPM8 and MSMB with KLK3 normalization). In this study, we aimed to validate it in a larger cohort with serum PSA 2.5-10 ng/mL and test other selected transcripts and clinical parameters, including the percentage of free prostate-specific antigen (PSA) (% free PSA) and inflammation. In the main cohort of 299 men, we tested the quadruplex transcripts. In a subset of 146 men, we analyzed additional transcripts (CD45, EPCAM, EZH2, Ki67, PA2G4, PSGR, RHOA and TBP). After a prostate massage, the urine was collected, RNA isolated from a cell sediment and qRT-PCR performed. Ct values of KLK3 (i.e., PSA) were strongly correlated with Ct values of other genes which play a role in CaP (i.e., PCA3, AMACR, TRPM8, MSMB and PSGR). AMACR, PCA3, TRPM8 and EZH2 mRNA expression, as well as % free PSA, were significantly different for BPH and CaP. The best combined model (% free PSA plus PCA3 and AMACR) achieved an AUC of 0.728 in the main cohort. In the subset of patients, the best AUC 0.753 was achieved for the combination of PCA3, % free PSA, EPCAM and PSGR. PCA3 mRNA was increased in patients with inflammation, however, this did not affect the stratification of patients indicated for prostate biopsy. In conclusion, the percentage of free PSA and urinary markers contribute to a more accurate indication for prostate biopsy.