Abstract
BACKGROUND: In men aged above 50 yr, lower urinary tract symptoms (LUTS), benign prostate hyperplasia, and prostate cancer are common urological conditions. Current guidelines for general practitioners frequently recommend prostate-specific antigen (PSA) testing in patients with LUTS for the detection of prostate cancer. OBJECTIVE: To assess the performance of PSA, PSA density, and the Stockholm3 blood test for identification of prostate cancer among men with LUTS. DESIGN SETTING AND PARTICIPANTS: In this post hoc analysis of a population-based diagnostic trial (STHLM3, n = 58 588), 4588 men aged 50-69 yr, without previous prostate cancer, with International Prostate Symptom Score (IPSS) data, and having PSA ≥ 3 ng/mL were identified. Men with at least moderate LUTS, defined as an IPSS score of ≥8, were included. PSA density and Stockholm3 scores were calculated. INTERVENTION: Participants underwent 10-12-core systematic prostate biopsies. OUTCOME MEASUREMENTS AND STATISTICAL ANALYSIS: The primary outcome was significant prostate cancer (sPCa) defined as International Society of Urological Pathology (ISUP) grade ≥2. Logistic regression analysis adjusted for age and previous biopsy status was performed. The area under the receiver operating characteristic curve (AUC) was calculated, and decision curve analysis was performed. RESULTS AND LIMITATIONS: Out of 4588 men, 1544 (34%) reported at least moderate LUTS. The median age was 64 yr, and 11% had undergone a previous prostate biopsy. The Stockholm3 test showed superior discrimination for sPCa to PSA density, which in turn showed superior discrimination to PSA (AUC 0.77 vs 0.70 vs 0.61, p < 0.02). Calibration of the Stockholm3 test was adequate. Performing biopsy only in men with PSA ≥5 ng/mL saved 64% of biopsies, but resulted in missing 52% of detectable sPCa. Recommending biopsy for men with PSA density ≥0.07 resulted in sparing 26% of biopsy procedures and delaying the diagnosis of 12% of sPCa cases, with a 6.1% risk of sPCa among unbiopsied men. Recommending men with Stockholm3 ≥ 0.11 for biopsy resulted in sparing 53% of biopsy procedures and delaying the diagnosis of 20% of sPCa cases, with a 5.1% risk of finding sPCa in unbiopsied men. CONCLUSIONS: PSA density and the Stockholm3 blood test were superior to PSA for the identification of prostate cancer among men with LUTS. PATIENT SUMMARY: In this analysis of a large Swedish study, we find that the use of prostate-specific antigen (PSA) density or the Stockholm3 blood test instead of only PSA might improve the detection of prostate cancer among men with lower urinary tract symptoms.