ACCURATE risk stratification is fundamental in the management of prostate cancer (PCa), guiding clinicians in recommending either radical treatment or active surveillance (AS). The European Association of Urology (EAU) currently categorises patients into low, intermediate, or high risk, based on prostate-specific antigen (PSA), tumour stage, and biopsy grade group (GG). However, this classification primarily predicts biochemical recurrence, a short-term endpoint, rather than prostate cancer-specific mortality (PCSM), which is more clinically meaningful.
This study assessed the long-term discriminative ability of the EAU risk classification using data from the European Randomized Study of Screening for Prostate Cancer (ERSPC) Rotterdam. Importantly, the researchers explored whether reclassifying favourable intermediate-risk patients, specifically those with GG 1 or GG 2 disease, low PSA, and low PSA density, into the low-risk category could improve patient allocation to AS. Two cohorts were analysed: one with screen-detected disease (20 years of follow-up) and one with clinically detected disease (12 years of follow-up).
The EAU classification demonstrated strong performance, with time-dependent AUC values of 0.72–0.76 at 15 years. However, the alternative model that incorporated PSA density (≤0.20 ng/ml²) and reclassified certain intermediate-risk men as low risk achieved similar discriminative ability (AUC 0.74–0.75). Crucially, this reclassification expanded the proportion of men categorised as low risk by 45% in the screen-detected cohort and 83% in the clinically detected cohort, without increasing PCSM. The 15-year cumulative incidence of PCSM remained low and comparable across groups (around 2%).
These findings have important clinical implications. By aligning pretreatment risk stratification more closely with AS eligibility, clinicians may feel more confident recommending surveillance, potentially reducing overtreatment and associated healthcare costs. The incorporation of PSA density strengthens risk assessment, particularly for clinically detected cases, where it was shown to lower PCSM risk estimates further.
While validation in contemporary cohorts using MRI and additional pathological criteria is needed, this pragmatic adjustment to the EAU system offers a practical way to refine clinical decision-making. By safely expanding the low-risk category, the proposed model may encourage broader adoption of AS, striking a better balance between tumour control and quality of life.
Reference
de Vos II et al. Reducing overtreatment of prostate cancer patients: revisiting the European Association of Urology pretreatment risk group classification using long-term follow-up data from the European randomized study of screening for prostate cancer Rotterdam. Eur Urol Oncol. 2025;8(3):747-754. doi:10.1016/j.euo.2024.11.004
