An algorithm for prostate cancer detection in a patient population using prostate-specific antigen and prostate-specific antigen density

Abstract

Prostate-specific antigen (PSA) is the most accurate serum marker for cancer of the prostate (CaP). However, its sensitivity and specificity are suboptimal, especially at values ranging between 4.1 and 10.0 ng/ml (monoclonal), because benign prostatic hypertrophy and hyperplasia (BPH) and CaP frequently coexist in this range. This study was undertaken to determine the value of incorporating prostate volume measurements with serum PSA levels in a quotient (PSA/volume) entitled PSA density (PSAD). A total of 3140 patients were analyzed and stratified by serum PSA, digital rectal examination (DRE), transrectal prostate ultrasound (TRUS), TRUS volume determination and PSAD. All patients were referred for evaluation and therefore do not represent a screened population. Patients underwent prostate biopsies when abnormalities in TRUS or DRE were detected. Although both PSA and PSAD have statistical significance when the serum PSA value is < or = 4.0 ng/ml, neither has clinical significance in differentiating BPH from CaP. At serum levels ranging between 4.1 and 10.0 ng/ml, PSA has no ability to differentiate BPH from CaP, whereas PSAD does so with statistical and clinical significance. When the PSA value is between 10.1 and 20.0 ng/ml, only PSAD is statistically significant. When PSA exceeds 20 ng/ml, PSAD is redundant. We conclude that all patients with an abnormality on DRE or TRUS should undergo prostate biopsy. If the PSA value is < or = 4.0 ng/ml, TRUS and PSAD are not warranted and routine biopsy is not recommended. For intermediate PSA levels, 4.1-10.0 ng/ml, TRUS, TRUS prostate volume, and PSAD are important.2_