Abstract
PURPOSE: We investigated the extended use of Prostate Health Index (PHI) and percentage of [-2]pro-prostate-specific antigen (%p2PSA) in Chinese men with prostate-specific antigen (PSA) 10-20 ng/mL and normal digital rectal examination (DRE). MATERIALS AND METHODS: All consecutive Chinese men with PSA 10-20 ng/mL and normal DRE who agreed for transrectal ultrasound (TRUS)-guided 10-core prostate biopsy were recruited. Blood samples were taken immediately before TRUS-guided prostate biopsy. The performances of total PSA (tPSA), %free-to-total PSA (%fPSA), %p2PSA, and PHI were compared using logistic regression, receiver operating characteristic, and decision curve analyses (DCA). RESULTS: From 2008 to 2015, 312 consecutive Chinese men were included. Among them, 53 out of 312 (17.0%) men were diagnosed to have prostate cancer on biopsy. The proportions of men with positive biopsies were 6.7% in PHI<35, 22.8% in PHI 35-55, and 54.5% in PHI>55 (chi-square test, p<0.001). The area under curves (AUC) of the base model including age, tPSA and status of initial/repeated biopsy was 0.64. Adding %p2PSA and PHI to the base model improved the AUC to 0.79 (p<0.001) and 0.78 (p<0.001), respectively, and provided net clinical benefit in DCA. The positive biopsy rates of Gleason 7 or above prostate cancers were 2.2% for PHI<35, 7.9% for PHI 35-55, and 36.4% for PHI>55 (chi-square test, p<0.001). By utilizing the PHI cutoff of 35 to men with PSA 10-20 ng/mL and normal DRE, 57.1% (178 of 312) biopsies could be avoided. CONCLUSIONS: Both PHI and %p2PSA performed well in predicting prostate cancer and high grade prostate cancer. The use of PHI and %p2PSA should be extended to Chinese men with PSA 10-20 ng/mL and normal DRE.