RESUMO
OBJECTIVE: To confirm the accuracy of serum proPSA (p2PSA) and its derivatives, percentage of p2PSA to free PSA (%p2PSA) and Prostate Health Index (PHI) and to test the value of prostate dimension-adjusted related index p2PSA density (p2PSAD), %p2PSA density (%p2PSAD) and PHI density (PHID) in discriminating between patients with and without prostate cancer (PCa). PATIENTS AND METHODS: This is a prospective cohort study of 275 patients with a total PSA (tPSA) of 2-10 ng/ml who underwent initial prostate biopsy. Multivariate logistic regression models were complemented by predictive accuracy analysis. RESULTS: PCa was diagnosed in 31.2% of subjects. Median tPSA did not differ between groups, while PSA density (PSAD), percent free PSA (%fPSA), p2PSA, %p2PSA, PHI, p2PSAD, %p2PSAD and PHID (all p < 0.05) were different between men with and without PCa. Univariate accuracy analysis showed p2PSAD (area under the receiver-operating characteristic curve [AUC]: 0.71), %p2PSAD (AUC: 0.76) and PHID (AUC: 0.77) to be the most accurate predictors of PCa at biopsy, significantly outperforming tPSA (AUC: 0.54), PSAD (AUC: 0.68) and %fPSA (AUC: 0.59) (p ≤ 0.001). At multivariate logistic regression models, p2PSAD and PHID signiï¬cantly increased the accuracy of the basal multivariate model (all p < 0.01). At 90% specificity, sensitivity for p2PSAD, %p2PSAD and PHID were 33.7, 43 and 40.7%, respectively. Spearman's rho coefficient analysis demonstrated a significant relationship between Gleason score, %p2PSA (r = 0.216, p = 0.046), PHI (r = 0.223, p = 0.039) and %p2PSAD (r = 0.205, p = 0.05). CONCLUSIONS: Considering patients suited for initial prostate biopsy by a tPSA range of 2-10 ng/ml, PSA isoforms were confirmed to be strong predictors of PCa. The prostate dimension-adjusted PSA isoforms have been shown to differentiate between patients with or without PCa, with an AUC of 0.71-0.77, p2PSAD offering a gain in accuracy with respect to tPSA, %fPSA and PSAD.
