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.

Abstract

Background and objective: Tailored treatment for prostate cancer (PCa) requires accurate risk stratification. This study examines the effectiveness of the European Association of Urology (EAU) classification in predicting long-term PCa-specific mortality (PCSM) and assesses whether an alternative system can improve the identification of patients with low-risk disease.

Methods: This study included two cohorts of patients with localized PCa: one with screen-detected PCa (n = 1563; S-cohort) and the other with clinically detected PCa (n = 755; C-cohort), all from a population-based, randomized screening study, who underwent primary radical prostatectomy or radiation monotherapy. Patients were stratified according to the traditional EAU risk classification and an alternative risk classification where low-risk disease is adjusted according to contemporary active surveillance (AS) eligibility criteria. The 15-yr time-dependent area under the curve (AUC) and the cumulative incidence of PCSM at 15 yr after diagnosis were assessed for each risk classification and cohort.

Key findings and limitations: With a median follow-up of 20 yr in the S-cohort and 12 yr in the C-cohort, the EAU classification demonstrated 15-yr AUCs of 0.76 (95% confidence interval [CI]: 0.71-0.80) and 0.72 (95% CI: 0.65-0.79), respectively, for predicting PCSM. The alternative classification showed a 15-yr AUC of 0.74 (95% CI: 0.69-0.79) in the S-cohort and 0.75 (95% CI: 0.68-0.81) in the C-cohort. The alternative classification identified 45% more men having a low risk in the S-cohort and 83% more in the C-cohort than the EAU classification, with no statistically significant increase in the 15-yr PCSM incidence (S-cohort subhazard ratio: 1.33 [95% CI: 0.66-2.68]; C-cohort subhazard ratio: 0.99 [95% CI: 0.29-3.38]).

Conclusions and clinical implications: The EAU classification predicts PCSM accurately, but an alternative classification, adjusted for AS eligibility, identifies substantially more men as having a low risk. This could enhance AS acceptance and utilization in clinical practice, reducing overtreatment.

Patient summary: This study shows that while a commonly used pretreatment risk classification for prostate cancer predict disease prognosis accurately, an alternative system based on active surveillance eligibility criteria identifies many more men as having a low risk. Adopting this classification could enhance the acceptance and use of active surveillance, reducing unnecessary treatments.