Importance: Active surveillance is the preferred management strategy for patients with low- or favorable intermediate-risk prostate cancer (PCa); however, frequent health care visits can be costly and burdensome to patients. Identifying patients who may benefit from intensive vs passive surveillance could reduce these burdens.
Objective: To investigate associations between a polygenic risk score (PRS) and risk of upgrading and other prostate tumor features in patients receiving active surveillance.
Design, setting, and participants: This prospective multicenter cohort study across 10 US sites included 1220 patients from the Canary Prostate Active Surveillance Study (PASS) enrolled from July 2008 to November 2017, with follow-up (median, 5.3 years) through August 2022. Participants were those with clinically localized PCa (cT1-T2) receiving active surveillance. Analyses took place from January 2023 to April 2024.
Exposure: Multi-ancestry PRS of 451 PCa risk variants (PRS-451) and 400 PCa risk variants (PRS-400) after excluding prostate-specific antigen (PSA)-associated variants.
Main outcomes and measures: The primary outcome was PCa upgrading (any Gleason grade increase) vs no upgrading. Secondary outcomes included prostate volume, PSA, PSA density, percentage of biopsy cores with cancer, and Gleason grade group at diagnosis.
Results: The median (IQR) age at diagnosis of the 1220 patients receiving active surveillance was 63 (58-67) years. During follow-up, 470 patients upgraded; the 2- and 5-year risks of upgrading were 17.7% (95% CI, 15.5%-19.9%) and 33.3% (95% CI, 30.5%-36.3%), respectively. Each 1-SD unit increase in PRS-451 was associated with 23% increased hazard of upgrading (95% CI, 1.11-1.35; P < .001), whereas PRS-400 was associated with 27% increased hazard (95% CI, 1.15-1.39; P < .001) at any point in time during follow-up. Except for PSA, associations with remaining outcomes were similar or stronger using PRS-400. Higher PRS-400 was associated with smaller prostate volume, a higher percentage of biopsy cores with cancer, and higher PSA density. A model with clinical risk factors had a C-index of 0.64 (95% CI, 0.62-0.67); adding PRS-400 led to a C-index of 0.65 (95% CI, 0.63-0.68).
Conclusions and relevance: In this cohort study, among patients receiving active surveillance, high PRS was associated with risk of upgrading and possibly tumor multifocality. Excluding PSA variants from the PRS revealed an association with smaller prostate size, which has been previously associated with more aggressive tumors. Although PRS may inform active surveillance, it is yet to be seen whether they improve clinical decisions.