Incidence and predictors of post-ERCP pancreatitis in patients with suspected sphincter of Oddi dysfunction undergoing biliary or dual sphincterotomy: results from the EPISOD prospective multicenter randomized sham-controlled study

Background and study aim: Pancreatitis following endoscopic retrograde cholangiopancreatography (ERCP) is a significant and potentially life-threatening adverse event and is common in patients with suspected sphincter of Oddi dysfunction (SOD). Here we aimed to identify predictors of the risk in this population. Patients and methods: The Evaluating Predictors and Interventions in SOD (EPISOD) study prospectively enrolled 214 post-cholecystectomy patients with SOD type III in seven US centers. Patients were randomized, using a 2: 1 allocation, to sphincterotomy or sham procedure, irrespective of the results of sphincter of Oddi manometry. Patients in the sphincterotomy armwho had elevated pancreatic sphincter pressure were randomized to biliary only or to dual (biliary and pancreatic) sphincterotomy. All but one patient received prophylactic pancreatic stents, but none received pharmacological prophylaxis. Post ERCP pancreatitis (PEP) was defined as acute pancreatitis within the subsequent 7 days. Blinded research coordinators at each site called patients at 1 week post-procedure. Results: PEP occurred in 26 patients, in 10.6% (15/141) in the sphincterotomy arm and 15.1% (11/73) in the sham arm; unadjusted relative risk 0.71 (95% confidence interval [95% CI] 0.34-1.46). PEP rate was not significantly different in patients who received sphincterotomy compared with those undergoing sham treatment. In addition, the proportion was not statistically different in those who received biliary sphincterotomy alone (12/94; 12.8% [95 % CI 6.0%-19.5 %]) compared with dual sphincterotomy (3/47; 6.4% [95 % CI 0.0%-13.4 %]). Multivariate analysis identified an interaction between duration of ERCP and sedation type (P< 0.02). Conclusion: The performance of biliary or dual sphincterotomy does not increase the risk of PEP in patients suspected of SOD. However, the high rate of PEP in patients with suspected SOD, despite pancreatic stenting in expert centers, is confirmed in this prospective study. The combined effect of duration of ERCP and sedation type on the development of PEP should be further explored.