Endoscopic versus surgical treatment for infected necrotizing pancreatitis: a systematic review and meta-analysis of randomized controlled trials

Objective To compare outcomes of endoscopic and surgical treatment for infected necrotizing pancreatitis (INP) based on results of randomized controlled trials (RCT). Background Treatment of INP has changed in the last two decades with adoption of interventional, endoscopic and minimally invasive surgical procedures for drainage and necrosectomy. However, this relies mostly on observational studies. Methods We performed a systematic review following Cochrane and PRISMA guidelines and AMSTAR-2 criteria and searched CENTRAL, Medline and Web of Science. Randomized controlled trails that compared an endoscopic treatment to a surgical treatment for patients with infected walled-off necrosis and included one of the main outcomes were eligible for inclusion. The main outcomes were mortality and new onset multiple organ failure. Prospero registration ID: CRD42019126033 Results Three RCTs with 190 patients were included. Intention to treat analysis showed no difference in mortality. However, patients in the endoscopic group had statistically significant lower odds of experiencing new onset multiple organ failure (odds ratio (OR) confidence interval [CI] 0.31 [0.10, 0.98]) and were statistically less likely to suffer from perforations of visceral organs or enterocutaneous fistulae (OR [CI] 0.31 [0.10, 0.93]), and pancreatic fistulae (OR [CI] 0.09 [0.03, 0.28]). Patients with endoscopic treatment had a statistically significant lower mean hospital stay (Mean difference [CI] - 7.86 days [- 14.49, - 1.22]). No differences in bleeding requiring intervention, incisional hernia, exocrine or endocrine insufficiency or ICU stay were apparent. Overall certainty of evidence was moderate. Conclusion There seem to be possible benefits of endoscopic treatment procedure. Given the heterogenous procedures in the surgical group as well as the low amount of randomized evidence, further studies are needed to evaluate the combination of different approaches and appropriate timepoints for interventions.