Depression in patients with advanced prostate cancer in SWOG advanced cancer clinical trials.

Abstract

12090 Background: Depression is common in patients with advanced cancer, but its prevalence has not been well documented. Moreover, depression is likely associated with other patient factors, including sociodemographic and clinical variables. We examined depression at enrollment in clinical trials for patients with advanced prostate cancer. Methods: We pooled clinical trial data from the SWOG Cancer Research Network. We identified phase III treatment trials of advanced prostate cancer patients with baseline mental health symptom measurements. Baseline depression was derived from emotional functioning items from the FACT-G, the SF-36, and the EORTC QLQ-C30 instruments. Using Likert scale distributions, depression was categorized as none, mild or moderate, and severe. We evaluated the prevalence of depression and its association with other baseline variables including age, race, ethnicity, insurance, rural/urban locale, the Area Deprivation Index, measurable disease, and prognostic risk. Generalized estimating equations with binomial logistic regression were used to assess the association of baseline variables with the odds of any depression and severe depression, with study as the clustering variable. A composite risk model was developed by summing the number of baseline risk factors adversely associated with depression. Results: Overall, N = 4,103 patients from four phase III trials were examined, including 69.8% aged 65 or older, 17.5% Black, 3.4% Hispanic, 18.4% rural, and nearly half (46.4%) from socioeconomically deprived areas (ADI score above the national median). At baseline, 50.4% of patients had depression (mild/moderate, 38.3%; severe, 12.1%). Depression was associated with age < 65 years, non-Black race, Hispanic ethnicity, having Medicaid or no insurance, and worse disease clinical characteristics. Patients with >3 risk factors (high risk) vs. 0-2 risk factors (low risk) were more likely to experience depression (54.3% vs. 43.2%, p < .0001) and severe depression (21.0% vs. 9.7%, p < .0001), corresponding to a 50% (OR = 1.50; 95% CI: 1.34-1.67; P < .0001) and 135% (OR = 2.35; 95% CI: 1.84-3.02; P < .0001) increase in risk, respectively. Quartile (Q) level proportions of any depression were 32.8% (Q1), 44.4% (Q2), 53.1% (Q3), and 72.0% (Q4), respectively, with a fivefold higher risk for those in the highest vs. lowest quartiles (Q4 vs. Q1, OR = 4.97; 95% CI, 2.86-8.64, p < .0001). Similar findings were seen with severe depression. Conclusions: Evidence of depression at baseline was reported by one-half of patients with advanced prostate cancer in clinical trials. Moreover, we showed that the number of adverse socioeconomic and clinical variables could strongly predict the prevalence of depression. Screening for depression in patients with cancer could help guide patients to appropriate mitigation resources. Interventions to help screen and treat patients with advanced cancer are warranted.