Introduction: Radiographic imaging is used to monitor disease progression for men with metastatic castrate-resistant prostate cancer (mCRPC). The optimal frequency of imaging, a costly and limited resource, is not known. Our objective was to identify predictors of radiographic progression to inform the frequency of imaging for men with mCRPC.
Methods: We accessed data for men with chemotherapy-naive mCRPC in the abiraterone acetate plus prednisone (AA-P) group of a randomized trial (COU-AA-302) (n=546). We used Cox proportional hazards modelling to identify predictors of time to progression. We divided patients into groups based on the most important predictors and estimated the probability of radiographic progression-free survival (RPFS) at six and 12 months.
Results: Baseline disease and change in prostate-specific antigen (PSA) at eight weeks were the strongest determinants of RPFS. The probability of RPFS for men with bone-only disease and a ≥50% fall in PSA was 93% (95% confidence interval [CI] 87-96) at six months and 80% (95% CI 72-86) at 12 months. In contrast, the probability of RPFS for men with bone and soft tissue metastasis and <50% fall in PSA was 55% (95% CI 41-67) at six months and 34% (95% CI 22-47) at 12 months. These findings should be externally validated.
Conclusions: Patients with chemotherapy-naive mCRPC treated with first-line AA-P can be divided into groups with significantly different risks of radiographic progression based on a few clinically available variables, suggesting that imaging schedules could be individualized.