Validation of the STAR-CAP clinical prognostic system for predicting biochemical recurrence after radical prostatectomy in a Japanese cohort.

Abstract

253 Background: An international staging collaboration for cancer of the prostate (STAR-CAP) was recently established and used for predicting oncological outcomes for patients with prostate cancer. However, the validation of STAR-CAP has not yet been performed in the Japanese cohort. Our study aimed to assess the STAR-CAP prognostic model for Japanese patients with prostate cancer and compare it to the CAPRA score for the prediction of biochemical recurrence (BCR) after radical prostatectomy. Methods: We retrospectively analyzed data from a multi-institution database of 1,645 patients who underwent radical prostatectomy between 2010 and 2020. We excluded patients with inadequate follow-up duration ( < 1 year), missing or incomplete perioperative clinical parameters relevant for STAR-CAP, and patients who received neoadjuvant therapy. BCR was defined as two consecutive prostate-specific antigen values ≥ 0.2 ng/mL. The 3-year BCR-free survival was assessed by Kaplan-Meier analyses and log-rank test. The performance of both models to predict 3-year BCR-free survival was assessed using the area under the curve (AUC). Results: The 3-year BCR-free survival ranged from 47.3% (high risk) to 93.5% (low risk) in the STAR-CAP classification ( p< 0.001) and from 62.5% (high risk) to 94.6% (low risk) in the CAPRA classification ( p< 0.001). AUCs for the 3-year BCR prediction were 0.756 and 0.784 for STAR-CAP and CAPRA, respectively. Conclusions: In Japanese patients, both STAR-CAP and CAPRA prognostic systems provide sufficient stratification and their predictive ability for 3-year BCR-free survival is comparable, with a small advantage for CAPRA (2.8%).