Epidemiology of High-risk Biochemical Recurrence After Primary Prostate Cancer Treatment

Abstract

Background and objective

Biochemical recurrence (BCR) risk stratification guides treatment decisions after primary prostate cancer (PCa) treatment. We evaluated high-risk BCR (HR-BCR) definitions after radical prostatectomy (RP) or radiotherapy (RT) and their association with PCa-specific mortality (PCSM).

Methods

A population-based cohort study including 17 753 men treated with RP (n = 12 010) or RT (n = 5743) for localized PCa in Stockholm County between 2003 and 2021 was conducted. We assessed the cumulative incidence of any BCR (RP: prostate-specific antigen [PSA] ≥0.2; RT: PSA ≥nadir + 2), European Association of Urology (EAU) HR-BCR (PSA doubling time ≤1 yr or pathological International Society of Urological Pathology (ISUP) grade group 4–5 after RP; time to BCR ≤18 mo or biopsy ISUP grade group 4–5 after RT), and EMBARK HR-BCR (PSA doubling time ≤9 mo and PSA >1 ng/ml after RP or PSA ≥nadir + 2 ng/ml after RT). PCSM after HR-BCR was estimated using the competing risk method.

Key findings and limitations

The 10-yr incidence of HR-BCR was 10% (95% confidence interval [CI]: 9–11) for EAU HR-BCR and 4% (95% CI: 3–4) for EMBARK HR-BCR after RP, and 10% (95% CI: 9–11) for both definitions after RT. Patients meeting the EMBARK criteria had the highest PCSM (RP: 30%, 95% CI: 24–37; RT: 50%, 95% CI: 45–56). Up to 50% of RP and 31% of RT patients with BCR did not progress to HR-BCR and had lower PCSM.

Conclusions and clinical implications

HR-BCR incidence varies by definition and treatment. The EMBARK criteria identify a smaller subset with the highest PCSM risk. Many patients with BCR never develop HR-BCR. Refining BCR definitions with PSA kinetics and imaging may optimize risk stratification, balancing therapeutic efficacy and overtreatment.