Diagnostic performance of expression of PCA3, Hepsin and miR biomarkers inejaculate in combination with serum PSA for the detection of prostate cancer

Abstract

BACKGROUND AND METHODS: Here, we report on the evaluation of the diagnostic performance of ejaculate-derived PCA3, Hepsin, and miRNAs to complement serum PSA to detect prostate cancer. cDNA was prepared from 152 candidate specimens following RNA isolation and amplification for PSA, PCA3 and Hepsin qPCR, with 66 having adequate RNA for all three assays. Small RNA sequencing and examination of PCa-associated miRNAs miR-200b, miR-200c, miR-375 and miR-125b was performed on 20 specimens. We compared findings from prostate biopsies using D'Amico and PRIAS classifications and in relation to whole gland histopathology following radical prostatectomy. Multivariate logistic regression modeling and clinical risk (incorporating standard clinicopathological variables) were performed for all ejaculate-based markers. RESULTS: While Hepsin alone was not of predictive value, the Hepsin:PCA3 ratio together with serum PSA, expressed as a univariate composite score based on multivariate logistic regression, was shown to be a better predictor than PSA alone of prostate cancer status (AUC 0.724 vs. 0.676) and risk, using D'Amico (AUC 0.701 vs. 0.680) and PRIAS (AUC 0.679 vs. 0.659) risk stratification criteria as classified using prostate biopsies. It was also possible to analyse a subgroup of patients for miRNA expression with miR-200c (AUC 0.788) and miR-375 (AUC 0.758) showing best single marker performance, while a combination of serum PSA, miR-200c, and miR-125b further improved prediction for prostate cancer status when compared to PSA alone determined by biopsy (AUC 0.869 vs. 0.672; P < 0.05), and risk (D'Amico/PRIAS) as well as by radical prostatectomy histology (AUC 0.809 vs. 0.690). For prostate cancer status by biopsy, at a sensitivity of 90%, the specificity of the test increased from 11% for PSA alone to 67% for a combination of PSA, miR-200c, and miR-125b. CONCLUSIONS: These results show that use of a combination of different types of genetic markers in ejaculate together with serum PSA are at least as sensitive as those reported in DRE urine. Furthermore, a combination of serum PSA and selected miRNAs improved prediction of prostate cancer status. This approach may be helpful in triaging patients for MRI and biopsy, when confirmed by larger studies.