S2216 Impact of Proton Pump Inhibitor Use on Post-Drainage Outcomes Following Axios Stent Placement for Pancreatic Fluid Collections: A Multi-Institutional Real-World Data Analysis

Introduction: Proton pump inhibitors (PPIs) are frequently continued after endoscopic drainage of pancreatic fluid collections (PFCs) using lumen-apposing metal stents (LAMS), often due to prior prescriptions or perceived benefits. However, current guidelines do not recommend routine PPI use in this context, and its impact on clinical outcomes remains unclear. Observational studies have raised concerns that PPI use may delay fistula closure and increase the risk of complications such as infection, bleeding, or the need for necrosectomy. This study evaluates the association between post-procedural PPI use and adverse outcomes including infection, gastrointestinal (GI) bleeding, and pancreatic necrosectomy. Methods: We conducted a retrospective cohort study using the TriNetX US Collaborative Network, comprising 69 healthcare organizations. Adult patients (≥18 years) with PFCs (ICD-10: K86.2/K86.3) who underwent endoscopic transmural drainage with a LAMS (CPT: 43240) were included. Two cohorts were defined based on PPI use within one month post-procedure: Cohort 1 (PPI users, n = 2,805) and Cohort 2 (non-users, n = 3,000). Propensity score matching was performed (1:1), resulting in 2 balanced cohorts of 2,184 patients each. Outcomes were assessed over a one-year period using risk analysis and Kaplan-Meier survival estimates. Results: Post-procedural PPI use was associated with significantly higher risks of adverse outcomes. Infections occurred in 23.6% of PPI users vs 12.7% of non-users (Risk Ratio [RR] 1.85, P < 0.001). GI bleeding occurred in 14.1% vs 5.7% (RR 2.46, P < 0.001), and the need for pancreatic necrosectomy was 3.8% vs 1.4% (RR 2.77, P < 0.001). Kaplan-Meier analysis showed lower survival probabilities for each outcome in the PPI cohort. Corresponding hazard ratios (HR) were significantly elevated: infection (HR 2.01), GI bleeding (HR 2.63), and necrosectomy (HR 2.82), all with P < 0.001. Conclusion: PPI administration following Axios stent placement for PFC drainage is associated with increased risks of infection, GI bleeding, and the need for necrosectomy. These findings call for a critical evaluation of routine PPI use in this setting and suggest a need for prospective studies to guide post-drainage management.