Abstract
Aim: The goal of this study was to examine if determining [-2]proPSA levels and calculating the Prostate Health Index (PHI) could improve overall sensitivity and specificity of this marker in the diagnosis of prostate cancer compared to standard markers (PSA and %freePSA), and propose an optimal PHI cut-off. Methods: A group of 76 patients with suspected prostate cancer, who were scheduled to undergo prostate biopsies, was tested to determine the total PSA, freePSA and [-2]proPSA levels, calculated %freePSA and Prostate Health Index (PHI).
Biomarkers were determined using chemiluminescent technology on a DxI 800 (Beckman Coulter, USA). Statistical analysis was performed using SAS version 9.2.
Results: We found a statistically significant improvement in the area under the ROC (AUC) for both [-2]proPSA (0.77) and especially for PHI (0.88) levels compared with total PSA (0.59) and % freePSA (0.61) levels. None of the patients included in this study, with histological diagnosis of prostate cancer on biopsy, had a PHI level under 40.
Conclusion: Determining the [-2]proPSA and derived PHI values contributes significantly to accuracy in the process of the differential diagnosis between BPH and prostate cancer. Based on out experience, a PHI cut-off of 40, with a gray area between 30 and 40, is optimal for use in routine clinical practice.