Patterns of pancreatic enzyme replacement therapy (PERT) use among U.S. patients with pancreatic cancer.

Abstract

696 Background: PERT is an important but costly and inconsistently used supportive care medication in PC. We sought to characterize patterns of PERT use among U.S. insured patients with PC. Methods: We conducted a retrospective observational analysis of patients with a primary diagnosis of pancreatic adenocarcinoma in the IBM MarketScan national claims database between 2013-2019. We identified pharmacy claims for PERT (Creon, Pancreaze, Pertzye, Viokace, and Zenpep) and classified patients as PERT-users (≥ 1 PERT prescription [Rx]) and non-PERT users. We used descriptive statistics for clinical, demographics, and costs. Multiple logistic regression identified predictors of PERT use. Kaplan-Meier method assessed time from PC diagnosis to first PERT Rx. Results: Of 19,866 patients with PC (median age 62 years, 49% women), 6,918 (35%) were PERT-users. The median follow-up was 11 months (PERT users: 14 months, non-PERT users: 9 months). Overall, 43% of patients received no cancer-directed treatment. PERT-users were more likely to have a history of pancreatitis (acute, odds ratio [OR] = 1.9; chronic, OR = 3.6), and received chemotherapy (OR = 1.2), surgery (OR = 2.3), and radiation (OR = 1.4). Among all PERT-users, the median time from PC diagnosis to first fill was 1.9 months (95% CI: 1.8,2.0). The median time from surgery to first fill was 1.6 months (95% CI: 1.4,1.7). The median number of PERT Rxs was 4; 22% of PERT-users received 1 Rx. A median of 150 days of PERT was supplied across all Rxs per patient. The median (IQR) cost of each PERT Rx was $25 ($0- $50). PERT accounted for 16% of all drug out-of-pocket costs (OOPC) (excluding systemic outpatient cancer-directed treatment), and 7% of all healthcare OOPC. Patients with PERT had significantly higher monthly OOPC on Rxs but lower on all other services, including inpatient and outpatient services (all p < 0.0001). Overall, PERT-users had significantly lower per-patient per-month (PPPM) OOPC vs non-PERT users ($469, vs $520, p < 0.0001). They also had fewer PPPM outpatient (0.7 vs 1.2, p < 0.0001) and ER (0.4 vs 0.5, p < 0.0001) visits. Conclusions: In this national study of insured U.S. adults with PC, only one-third received PERT. Of those receiving PERT, approximately a fifth received just a single Rx. PERT users, despite higher OOP drug spending had reduced overall OOP costs and fewer healthcare visits. These data highlight the need to explore non cost-related reasons for lack of and ways to standardize PERT use.