Abstract
Background: The Prostate Health Index (PHI) and the Prostate CAncer gene 3 (PCA3) have shown to have predictive values for early detection of PCa. The aim of this prospective observational study was to compare the values of the both markers in patients consulting for a suspicion of PCa Methods: Samples of 573 consecutive patients presenting a suspicion of PCa were included in a biobank between 2010 and 2012. Total PSA (tPSA),Free PSA (fPSA) and -2proPSA (Beckman Coulter) were performed on serum to calculate the PHI ([-2proPSA/fPSA] X [tPSA]1/2). Urine samples were collected after digital rectal examination (DRE) by an urologist in order to calculate the PCA3 score ([PCA3 mRNA]/[PSA mRNA] X 1,000). Prostate biopsies were performed for 235 men according usual clinical practice. Correlation between PSA, PHI, PCA3 score and clinico-pathological factors were studied with the Kruskall-Wallis tests. ROC–derived area under the curve (AUC) was used to quantify the predictive accuracy of the tests predicting pathologic findings in biopsies specimens. Results: PCA3 score was only correlated with age (Spearman’s rho = 0.21; p=0.001) and PHI was only correlated with tPSA (Spearman’s rho = 0.35; p<0.0001). PHI and PCA3 were slightly correlated (Spearman’s rho = 0.16; p=0.01). Comparing biomarkers values in patients with positive (P+; n=60, 25.5%) and negative biopsy (P-; n=175, 74.5%), PHI and PCA 3 were statistically different according to patients status. Conclusions: In this large series of patients consulting for a suspicion of PCa, tPSA is hardly more effective than a coin toss to predict PCa. PHI, an independent biological marker, presented a higher predictive value for positive biopsy prediction than PCA3, a marker slightly correlated with age.