Prostate-specific Antigen Density Toward a Better Cutoff to Identify Better Candidates for Active Surveillance

Abstract

To investigate the impact of prostate-specific antigen density (PSAD) on existing prostate cancer (PCa) active surveillance (AS) protocols. Prospectively maintained database on men with PCa who underwent radical prostatectomy was reviewed retrospectively. Demographic data and pathologic characteristics of patients who fulfilled the AS inclusion criteria under the National Comprehensive Cancer Network (NCCN), Prostate Cancer Research International Active Surveillance (PRIAS), and University of California, San Francisco (UCSF) guidelines were examined. Of 930 patients, 231, 280, and 325 fulfilled the NCCN, PRIAS, and UCSF AS criteria, respectively. The frequencies of advanced disease on surgical pathology (upstaging and/or upgrading) were 31.6% (NCCN), 35.4% (PRIAS), and 34.2% (UCSF) of the study cohorts. PSAD was significantly higher in patients with advanced disease compared with that in patients with nonadvanced disease in all 3 AS schemas. Modifying the PRIAS and UCSF criteria using the NCCN's lower PSAD cutoff of 0.15ng/mL(2) decreased the rates of the advanced disease significantly to 33.5% and 31.4%, respectively. Using the receiver operating characteristic curve analysis, the optimal PSAD cutoff level for the prediction of advanced disease was 0.085ng/mL(2) (sensitivity/specificity of 76.7%/50.6% in NCCN and 75.6%/49.7% in PRIAS). Among patients with low-risk PCa who underwent radical prostatectomy, PSAD is a predictor of advanced disease at the time of surgery. Adopting a lower PSAD threshold of 0.085ng/mL(2) decreased the risk of the advanced disease to 17.5%-21.7%. Therefore, PSAD should be part of all AS guidelines.