Prevalence of exocrine pancreatic insufficiency fi ciency at 12 months after acute pancreatitis: a prospective, multicentre, longitudinal cohort study

Background Exocrine Pancreatic insufficiency (EPI) occurs following acute pancreatitis (AP) at variably reported rates and with unclear recovery timeline. The aim of this study was to establish the prevalence and predictors of EPI at 12 months after AP in a prospective cohort. Methods In this prospective, multicentre, longitudinal cohort study, adult participants (>= 18 years) admitted to the hospital with an AP attack (defined by Revised Atlanta Classification) were enrolled in a United States multi-centre longitudinal cohort (Sites: The Ohio State University, University of Pittsburgh, and Johns Hopkins University). Patients were excluded if they had pancreatic cancer, chronic pancreatitis, or malabsorptive disease (including previously diagnosed EPI). Participant data was obtained by interview and by review of the electronic medical record. EPI was assessed by stool fecal elastase (FE-1) levels collected at baseline, 3 months, and 12 months (primary endpoint). EPI was defined by FE-1 <200 mu g/g; severe FE-1 level <= 100 mu g/g; mild FE-1 101-200 mu g/g. Multivariable logistic regression was used to identify predictors of EPI at 12 months. This study is registered with ClinicalTrials.gov, NCT03063398. Findings EPI was observed in 29 (34.1%) of the 85 participants [44 (51.8%) male, mean age 54.7 +/- 14.1 years] who provided stool samples at 12 months. For the study overall, participants were recruited between June 22, 2017 and October 18, 2021. A total of 5794 individuals were screened, 311 of whom were eligible for the study. 112 participants provided stool samples at baseline, 79 completed stool samples at 3 months, and 85 completed samples at 12 months. 64 participants included samples at all 3 timepoints. In univariable analysis, factors significantly associated with EPI at 12 months included recurrent (versus index) AP, pre-existing diabetes, alcohol, and idiopathic etiologies, and increasing severity of AP. In multivariable analysis, the odds of having EPI at 12 months increased 4-fold with idiopathic AP etiology (Odds Ratio 4.095, 95% Confidence Interval [CI] 1.418, 11.826), and 3-fold with moderately severe or severe AP (Odds Ratio 3.166, 95% CI 1.156, 8.670), and baseline diabetes mellitus (Odds Ratio 3.217, 95% CI 1.113, 9.298). Even individuals with an index mild attack of AP (n = 39) developed severe EPI at 12 months (prevalence 12.8%). Interpretation EPI as diagnosed by FE-1 is present in over one third of prospectively assessed patients at 12 months post-AP. Since EPI develops in patients with mild AP, investigations are needed to understand the mechanisms of injury and identify methods for tailored screening.