IMPACT OF EXOCRINE PANCREATIC INSUFFICIENCY AND PANCREATIC ENZYME REPLACEMENT THERAPY ON ALL-CAUSE MORTALITY AMONG US VETERANS WITH CHRONIC PANCREATITIS: A NATIONWIDE RETROSPECTIVE COHORT STUDY

Background: Exocrine pancreatic insufficiency (EPI) is highly prevalent in chronic pancreatitis (CP). EPI can cause bothersome gastrointestinal symptoms and serious adverse clinical outcomes including weight loss, sarcopenia, malabsorption of fat and fat-soluble vitamins, and osteoporotic fractures. However, the impact of EPI and pancreatic enzyme replacement therapy (PERT) on mortality has yet to be evaluated in a population-based CP cohort. Our aims were:1) to determine the association between EPI status and all-cause mortality in CP patients; 2) to determine the association between PERT treatment status and all-cause mortality in CP patients with EPI.

Methods: This retrospective cohort study used the nationwide Veterans Administration (VA) electronic medical records data 2000-2023. Included were VA patients with a diagnosis of CP by relevant ICD-9 and ICD-10 codes (sensitivity 87% and specificity 86%) and >1 year of follow-up. Excluded were patients with cystic fibrosis or pancreatic cancer at the start of follow-up. The exposure of interest was EPI status for Aim 1 and PERT treatment status (i.e., untreated EPI, under-dosed PERT, adequately-dosed PERT) for Aim 2, both as time-varying exposure. EPI was determined by any one of the following: 1) a recorded ICD-10 diagnosis of EPI (K86.81); 2) fecal elastase <200 mcg/gm stool; or 3) a prescription for PERT. Adequate PERT dose contained >120,000 USP units of lipase daily. Mortality information was extracted from the VA vital status file. Multivariable Cox proportional hazards models were built to compute hazard ratios (HRs) and 95% confidence intervals (CIs) for the association between EPI status and mortality and the association between PERT treatment status and mortality. We accounted for clustering by VA facility and controlled for the following covariates: age, sex, body mass index (BMI), smoking status, alcohol use disorder (AUD), diabetes mellitus, Charlson comorbidity score and calendar year. Complete-case analysis was performed given the limited missing data.

Results: Among 128,983 Veterans with CP, those with EPI ever (n=84,223) were slightly older, less likely to have AUD and more likely to be White and have normal BMI than those without EPI (n=44,760) (Table 1). Among patients with CP, having EPI was associated with a 31% higher risk of death than without a diagnosis of EPI (adjusted HR 1.31, 95% CI [1.26-1.35]) (Table 2). Compared to EPI patients not receiving PERT, EPI patients receiving adequately dosed PERT had a 24% lower risk of death (adjusted HR 0.76, 95% CI [0.72-0.80]), while EPI patients receiving inadequately-dosed PERT did not have a lower risk of death (adjusted HR 1.05, 95% CI [1.01-1.09]) (Table 2).

Conclusions: EPI is associated with increased risk of death among CP patients. Adequately-dosed PERT is associated with reduced risk of death among CP patients with EPI.