Resource utilization and use of life-extending therapies and corticosteroids in prostate cancer patients with corticosteroid-sensitive comorbidities

Objectives: Corticosteroids (CSs) are used concomitantly with life-extending therapies (LETs) in patients with castration-resistant prostate cancer (CRPC). This study examined time to LETs, LETs and concomitant CS adherence, and monthly all-cause healthcare utilization and costs in patients with CPRC with and without CS-sensitive comorbidities in the Veterans Health Administration population. Methods: Patients had CRPC if records showed prostate cancer diagnosis, medical/surgical castration and ≥2 prostate-specific antigen increases through 1 June 2007-31 May 2012. CS-sensitive comorbidities were assessed 6 months prior to the index date. Adherence, defined as medication possession ratio (MPR) ≥0.8, among patients initiating LETs (cabazitaxel, docetaxel, or abiraterone acetate) before 30 November 2011, resource utilization and costs among patients with concomitant CS were assessed. Statistical analysis included descriptive, Cox proportional hazards, and logistic regression models. Results: Common CS-sensitive conditions among 12,128 patients with CRPC included hypertension (75.74%) and hyperlipidemia (54.69%). Those with glaucoma (hazard ratio [HR]=0.67), ischemic heart disease (HR=0.78), and peripheral vascular disease (PVD) (HR=0.78) were less likely to be prescribed LETs (all p<0.01). Duration of LET was shorter among patients with CS-sensitive comorbidities (125.02 vs 133.08 days; p=0.04) in the 6 month follow-up period. Among LET-treated patients with and without CS-sensitive comorbidities, less than half had MPR≥0.8 (LET: 48.72% vs 54.05%; concomitant CS: 42.19% vs 40.54%, respectively). Cerebrovascular disease (odds ratio=0.107; 95% confidence interval=0.012 to 0.966) and PVD (odds ratio=0.523; 95% confidence interval=0.276 to 0.991) were associated with reduced CS adherence. Among patients with concomitant CS, those with CS-sensitive comorbidities had more inpatient stays than those without (20.45% vs 12.88%; p=0.033), incurring higher monthly inpatient costs ($1157 vs $342; p<0.0001) and total costs ($5725 vs $4772; p=0.036). Conclusion: CS-sensitive conditions influence initiation and duration of LETs, concomitant CS adherence, inpatient stays, and total costs. Future efforts should focus on specific strategies for treating prostate cancer patients with CS-sensitive comorbidities to ensure that they have appropriate access to LETs and to reduce costs and inpatient stays. Study limitations include the use of retrospective claims data and the relatively restricted subpopulation of older North American males.