RESUMO
OBJECTIVE: To measure a combination of novel molecular biomarkers in urine/blood samples of consecutive patients referring lower urinary tract symptoms (LUTS) not previously diagnosed, to improve prostate cancer diagnosis. METHODS: Serum and urine samples from 113 men who went consecutively to the Department of Urology of our Institution. Biomarkers analyzed were AMACR and MMP-2 levels, and GSTP1/RASSF1A methylation status, in addition to PSA levels. Sensitivity, specificity, area under the ROC (AUROC) curves, and discriminant function analysis were assessed to determine the diagnostic potential of each variable alone or in combination. RESULTS: Of the patients, 30.08% had PCa and the remaining ones were tumor free. Areas under the ROC (AUROC) curves were as follows: 0.476 for PSA, 0.532 for AMACR, and 0.706 for MMP-2. Sensitivity and specificity for methylation status were 53.3 and 45.9%, respectively. The combination of these biomarkers resulted in an AUROC curve of 0.788, which significantly outperformed AUROC curves for PSA (P = 0.0033) and AMACR (P = 0.0375). Sensitivity, specificity, positive and negative predictive values for the combination of biomarkers were 57.1, 96.6, 88.9, and 82.4%, respectively. CONCLUSION: We conclude that analysis of this biomarker combination in body fluids improves very significantly the diagnosis of PCa compared to the PSA test.
