S2782 Association Between Chronic PPI Therapy and Hepatic Encephalopathy in Cirrhosis: A Retrospective Dose-Stratified Study

Introduction: Despite widespread use, proton pump inhibitors (PPIs) have been implicated as a potential modifiable risk factor for hepatic encephalopathy (HE) in patients with cirrhosis. However, real-world data delineating this relationship remain limited, especially regarding dose and duration effects. We aimed to investigate the association between PPI use and risk of incident hepatic encephalopathy among cirrhotic patients, accounting for exposure duration and dose intensity. Methods: We conducted a retrospective multicenter cohort study using TriNetX, aggregating EHRs from 90+ U.S. health systems. Adults (≥18 years) with cirrhosis (International Classification of Diseases, Tenth Revision [ICD-10]: K70.3, K74.6, K76.9) from Jan 1, 2015–Dec 31, 2023 were included. Exclusions: prior HE, HCC, TIPS, or recent sepsis. Patients were stratified into: non-users (no PPI exposure); low-dose users (≤20 mg omeprazole-equivalent/day); high-dose users (≥40 mg/day); index date was first PPI prescription or matched pseudo-date. All had ≥12 months follow-up. Propensity score matching (1:1:1) used age, sex, MELD-lab equivalents (bilirubin, INR, albumin, creatinine), comorbidities (diabetes, CKD, alcohol use, varices, ascites), and medications (lactulose, rifaximin, diuretics, NSAIDs, beta-blockers). The primary outcome was incident HE within 12 months, defined via ICD-10 codes and initiation of lactulose or rifaximin. Cox models provided adjusted hazard ratios (aHRs), with Kaplan-Meier curves illustrating incidence. Results: 18,991 patients met criteria (6,423 per group). Baseline variables were balanced (SMD < 0.1). Median follow-up: 11.2 months. HE incidence: high-dose: 18.1%; low-dose: 13.6%; non-users: 9.8% (log-rank P < 0.001). Risk comparisons: high-dose vs non-users: aHR 1.62 (95% CI: 1.39–1.87), P < 0.001. High-dose vs low-dose: aHR 1.27 (95% CI: 1.10–1.47), P = 0.002. Low-dose vs non-users: aHR 1.31 (95% CI: 1.14–1.50), P < 0.001. Findings were consistent across subgroups (alcoholic vs non-alcoholic cirrhosis, sex, MELD strata). Sensitivity analyses controlling for lactulose, rifaximin, and diuretic use confirmed robustness. KM curves showed early divergence by month 3. Conclusion: In cirrhotic patients, chronic PPI use—especially high-dose—was independently linked to increased HE risk. These real-world data support cautious PPI use, favoring dose reduction or discontinuation when appropriate. Prospective studies are needed to validate causality and mechanisms.