2 Hydrogen Peroxide-Assisted Endoscopic Necrosectomy vs. Standard Endoscopic Necrosectomy for Walled-Off Pancreatic Necrosis: A Multi-Center Retrospective Comparative Study

Abstract

INTRODUCTION: Endoscopic necrosectomy has emerged as the preferred treatment modality for walled off pancreatic necrosis (WON). Standard endoscopic necrosectomy can be tedious and time consuming, employs sharp instruments in vascular cavities, and does not provide for microdebridement. Hydrogen peroxide (H2O2) lavage has been reported in small series; however, there have been no comparative studies on its safety and efficacy in the management of WON. METHODS: We performed a retrospective chart review of all cases undergoing endoscopic transmural management of WON at 9 major medical centers in the US from November 2011 to August 2018. Cystogastrostomies were performed using lumen apposing metal stents (LAMS). Patients who underwent cystogastrostomy without necrosectomy were excluded. Clinical success was defined as resolution of the WON by imaging within 6 months, without requiring non-endoscopic drainage procedures or surgery. RESULTS: A total of 296 patients underwent cystgastrostomy and 206 patients underwent one or more necrosectomies. Clinical success was achieved in 171/192 (89.1%) of those with follow-up data at 6 months. In terms of patients, 122 (59.2%) patients had at least one H2O2 necrosectomy (H2O2 group), and 84 (40.8%) patients had standard endoscopic necrosectomy. Clinical success was higher in the H2O2 group (n = 107 (95.5%) vs 63 (79.7%), P < 0.001). On multivariate logistic regression, the use of hydrogen peroxide (OR 17.4, 95% CI: 2.9–103, P = 0.002) was associated with higher clinical success rates (Table 2). During a mean follow up of 278 days, 29 complications occurred: bleeding (n = 18), perforation (n = 5), respiratory complication (n = 3), and sepsis (n = 3). One death was presumed to be procedure related (splenic pseudoaneurysm rupture day 8; standard necrosectomy group). A total of 394 necrosectomies were performed. Comparing necrosectomy procedures with and without H2O2 (n = 209 vs 185), there was no difference in procedure time (72.5 ± 39.6 vs 72 ± 44.2, P = 0.936), post-procedural bleeding (8 vs 10, P = 0.45), perforation (3 vs 2, P = 1), infection (1 vs 2, P = 0.6) or overall complication rate (n = 15 (7.1%) vs 14 (7.6%), P = 1) (Table 3). CONCLUSION: In this multicenter retrospective cohort study, H2O2 necrosectomy appeared to be safe and effective with a higher clinical success rate and equivalent rates of bleeding, perforation, death, and overall complications relative to standard necrosectomy.