PSA density does not improve predictive accuracy of the UCSF-CAPRA score.

Abstract

The University of California, San Francisco Cancer of the Prostate Risk Assessment (CAPRA) score is a validated tool using factors at diagnosis to predict prostate cancer outcomes after radical prostatectomy (RP). This study evaluates whether substitution of prostate-specific antigen (PSA) density for serum PSA improves predictive performance of the clinical CAPRA model. Participants were diagnosed in 2000-2019 with stage T1/T2 cancer, underwent RP, with at least a 6-month follow-up. We computed standard CAPRA score using diagnostic age, Gleason grade, percent positive cores, clinical T stage, and serum PSA, and an alternate score using similar variables but substituting PSA density for PSA. We reported CAPRA categories as low (0-2), intermediate (3-5), and high (6-10) risk. Recurrence was defined as two consecutive PSA ≥ 0.2 ng/mL or receipt of salvage treatment. Life table and Kaplan-Meier analysis evaluated recurrence-free survival after prostatectomy. Cox proportional hazards regression models tested associations of standard or alternate CAPRA variables with recurrence risk. Additional models tested associations between standard or alternate CAPRA score with recurrence risk. Cox log-likelihood ratio test (-2 LOG L) assessed model accuracy. A total of 2880 patients had median age 62 years, GG1 30% and GG2 31%, median PSA 6.5, and median PSA density 0.19. Median postoperative follow-up was 45 months. Alternate CAPRA model was associated with shifts in risk scores, with 16% of patients increasing and 7% decreasing (p < 0.01). Recurrence-free survival after RP was 75% at 5 years and 62% at 10 years. Both CAPRA component models were associated with recurrence risk after RP on Cox regression. Covariate fit statistics showed better fit for standard CAPRA model versus alternate (p < 0.01). Standard (hazard ratio [HR]: 1.55; 95% confidence interval [CI]: 1.50-1.61) and alternate (HR: 1.50; 95% CI: 1.44-1.55) CAPRA scores were associated with recurrence risk, with better fit for standard model (p < 0.01). In a 2880 patient cohort followed for median 45 months after RP, alternate CAPRA model using PSA density was associated with higher biochemical recurrence (BCR) risk, but performed inferior to standard CAPRA at predicting BCR. While PSA density is an established prognostic variable in prediagnostic settings and sub-stratifying low-risk disease, it does not improve BCR model predictive accuracy when applied across a range of cancer risk.