RESUMO
BACKGROUND/PURPOSE: Prostate specific antigen (PSA) with low specificity that causes unnecessary prostate biopsies increases clinical morbidities, psychological stress, and medical expenses. We aimed to test the accuracy and cutoff value of Prostate Health Index (PHI) in men for prostate cancer detection. METHODS: We prospectively enrolled 213 men who underwent prostate biopsy with PSAâ¦10 ng/ml or abnormal findings on digital rectal examination. Total PSA (tPSA), free PSA (fPSA) and p2PSA levels were measured by serum samples before prostate biopsy. PHI was calculated as (p2PSA/fPSA) × âtPSA. Multivariable logistic regression analyses were used to predict the risk of cancer and detect clinically significant prostate cancer. RESULTS: 33 (27.0%) patients were confirmed with the diagnoses of prostate cancer by prostate biopsy. The levels of p2PSA, %p2PSA, and PHI showed statistically significant differences between prostate cancer patients and non-cancer patients. %p2PSA and PHI had the highest area under the receiver operating characteristic curve (AUC) of 0.723 and 0.772 (both p < 0.001), respectively, predicting cancer detection at biopsy than other predictors (tPSA, fPSA, %fPSA, and PSA density (AUC: 0.544, 0.538, 0.593, and 0.664, respectively). In multivariable logistic regression, %p2PSA had a statistical significant odds ratio 8.51 (p = 0.003) and PHI had an odds ratio with marginal significance 4.18 (p = 0.06). CONCLUSION: %p2PSA and PHI increased the diagnostic accuracy with significantly greater sensitivity and specificity than tPSA. We determined an optimal cut-off value of PHI among Taiwanese population. These findings support the usefulness in the decisional process of prostate biopsy.