Analysis of the temporal relationship between patient-reported outcomes (PROs) and overall survival (OS) and radiographic progression-free survival (rPFS) in metastatic castration-resistant prostate cancer (mCRPC) patients (pts).

Abstract

240 Background: PROs are used to measure therapeutic impact. The relationship between PROs and clinical outcomes is not well described. Methods: COU-AA-301 (301) and COU-AA-302 (302) were phase 3 trials of abiraterone acetate + prednisone (P) vs P in post-docetaxel and chemotherapy-naïve mCRPC pts, respectively. Using Cox regression models, we explored the association between self-reported fatigue, pain, physical well-being (PWB), functional well-being (FWB), and PC-specific (PCS) signs and symptoms, and OS and rPFS over the first 181 days, regardless of treatment, using data from 301 (N = 1195) and, separately, from 302 (N = 1088). Pts in 301 had more advanced disease and were more symptomatic at entry; pts in 302 were asymptomatic/mildly symptomatic. PRO improvements were assessed for 301 and worsening was assessed for 302. Results: In 301, pts with PRO improvements had reduced risk of death and radiographic progression (p < 0.0001) vs pts with worsening or stable PROs (Table). When all PROs were included in a multivariate (MV) model, all except pain intensity were significantly associated with OS; pain intensity, PWB, and FWB improvements were significantly associated with reduced radiographic progression. Pts in 302 with worsening PROs had greater risk of radiographic progression (p ≤ 0.02) vs pts with improved or stable PROs (Table). When all PRO end points were included in an MV model, worsening PWB remained significantly associated with worse rPFS. There were too few events at cutoff to explore the relationship between PROs and OS in 302. Conclusions: These results reinforce the clinical meaningfulness of self-reported symptoms and suggest that PROs can be used in combination with clinical measures in practice. Clinical trial information: NCT00887198; NCT00638690. [Table: see text]