A Personalized, Risk-Based Approach to Active Surveillance for Prostate Cancer with Takeaways from Broader Oncology Practices: A Mixed Methods Review.

Abstract

Background/Objectives: To summarize the current state of knowledge regarding personalized, risk-based approaches in active surveillance (AS) for prostate cancer (PCa) and to explore the lessons learned from AS practices in other types of cancer. Methods: This mixed methods review combined a systematic review and a narrative review. The systematic review was conducted according to the Preferred Reporting Items for Systematic rviews and Meta-Analyses (PRISMA) guidelines, with searches performed in the Medline, Embase, Web of Science, Cochrane Central Register of Controlled Trials, and Google Scholar databases. Only studies evaluating personalized, risk-based AS programs for PCa were included. The narrative review focused on AS approaches in other solid tumors (thyroid, breast, kidney, and bladder cancer) to contextualize the findings and highlight lessons learned. Results: After screening 3137 articles, 9 were suitable for inclusion, describing the following four unique risk-based AS tools: PRIAS, Johns Hopkins, Canary PASS, and STRATCANS. These models were developed using data from men with low-risk (Grade Group 1) disease, with little to no magnetic resonance imaging (MRI) data. They used patient information such as (repeated) prostate-specific antigen (PSA) measurements and biopsy results to predict the risk of upgrading at the next biopsy or at radical prostatectomy, or to assign a patient to a pre-defined risk category with a corresponding pre-defined follow-up (FU) regimen. Performance was moderate across models, with the area under the curve/concordance index values ranging from 0.58 to 0.85 and calibration was generally good. The PRIAS, Canary PASS, and STRATCANS models demonstrated the benefits of less burdensome biopsies, clinic visits, and MRIs during FU when used, compared to current one-size-fits-all practices. Although little is known about risk-based AS in thyroid, breast, kidney, and bladder cancer, learning from their current practices could further refine patient selection, streamline monitoring protocols, and address adoption barriers, improving AS's overall effectiveness in PCa management. Conclusions: Personalized, risk-based AS models allow for a reduction in the FU burden for men at low risk of progression while maintaining sensitive FU visits for those at higher risk. The comparatively limited evidence and practice of risk-based AS in other cancer types highlight the advanced state of AS in PCa.