Changes in Patient-Reported Outcomes Linked to Better Survival in Metastatic Prostate Cancer
Patient-reported outcomes had significant associations with survival in metastatic castration-resistant prostate cancer (mCRPC), suggesting potential implications for clinical management, reported Tomasz M. Beer, MD, Oregon Health & Science University, Knight Cancer Institute, Portland, and colleagues, at the 2016 International Society for Pharmacoeconomics and Outcomes Research annual meeting.
Multiple measures of patient-assessed health-related quality of life (QOL) proved to be prognostic for both progression-free survival (PFS) and overall survival (OS). Baseline scores for physical well-being, functional well-being, prostate cancer subscale, and the Functional Assessment of Cancer Therapy–Prostate (FACT-P) health questionnaire all had significant associations with OS.
Time-dependent longitudinal changes in physical well-being and FACT-P total scores had significant associations with radiographic PFS, and changes in physical well-being, functional well-being, prostate cancer subscale, and FACT-P all had prognostic relationships with OS.
“This exploratory study provided preliminary evidence indicating that patients with mCRPC whose PWB [physical well-being], FWB [functional well-being], PCS [prostate cancer subscale], and FACT-P scores improve during treatment have a significantly increased probability of survival,” noted Dr Beer and colleagues.
The PREVAIL Clinical Trial
The findings came from a retrospective analysis of the randomized, placebo-controlled PREVAIL clinical trial of enzalutamide (Xtandi) in men with untreated mCRPC. The primary finding was that treatment with the androgen receptor inhibitor enzalutamide significantly delayed radiographic progression or death, delayed deterioration of QOL, and improved OS.
In this new analysis, Dr Beer and colleagues investigated potential associations of QOL (at baseline and changes over time) with radiographic PFS and OS in the PREVAIL study. The analysis included 1717 randomized patients—872 patients received enzalutamide 160 mg daily, and 845 patients received placebo. Data collected during the clinical trial included a 39-item FACT-P questionnaire validated for mCRPC, including the FACT-P total score and scores for physical well-being, functional well-being, and prostate cancer subscales of the FACT-P questionnaire.
Associations Between QOL and Survival
A review of baseline clinical and demographic variables showed that radiographic PFS had significant associations with the treatment group (enzalutamide vs placebo; hazard ratio [HR], 0.31; P <.001), presence of visceral metastases (yes vs no; HR, 1.47; P <.001), number of bone metastases (≤5 vs >5; HR, 0.72; P <.001), and prostate-specific antigen (PSA) progression at baseline (PSA projection only vs radiographic or both; HR, 1.57; P <.001).
The same variables had significant associations with OS (P <.001 for all), as did the time since diagnosis, baseline performance status, and baseline pain (asymptomatic vs symptomatic).
The associations between QOL and survival were assessed in 2 ways—baseline values and longitudinal change over time. A univariate analysis of baseline values showed that the FACT-P total score and each subdomain score had significant associations with radiographic PFS and OS, including physical well-being, functional well-being, and prostate cancer subscale.
In a multivariate analysis of baseline patient-reported outcomes, none of the baseline QOL factors retained significant associations with radiographic PFS. Overall, 3 baseline factors retained significance for OS—physical well-being (HR, 0.85; P = .001), prostate cancer subscale (HR, 0.89; P <.001), and the FACT-P total score (HR, 0.90; P <.001).
In a multivariate model of longitudinal patient-reported outcomes, 2 factors retained significant associations with radiographic PFS—physical well-being and FACT-P total score. The multivariate model of OS yielded significant associations for physical well-being, functional well-being, prostate cancer subscale, and the FACT-P total score.
“In general, point estimates for the hazard ratios were smaller in the time-dependent model, revealing a strong association between changes in QOL throughout the study and survival outcomes compared with the simpler model, including baseline QOL score only,” the researchers reported.
“The observed associations between QoL scores and clinical outcomes may have the potential to influence clinicians’ treatment approaches for mCRPC in the future,” concluded Dr Beer and colleagues.