Pancreatitis complicating endoscopic biliary sphincterotomy (ES): A prospective multivariate analysis of risk factors including pancreatic sphincter hypertension (PSH)

Risk factors for post-ES acute pancreatitis (AP) have not been comprehensively investigated. Based on data from a large prospective multicenter study of ES complications, we constructed a multivariate model to assess independent effect of multiple risk factors on pancreatitis post-ES. We then addressed the recent hypothesis that untreated PSH is responsible for the high risk of AP after ES for suspected sphincter of Oddi dysfunction (SOD). Methods: Complications were studied prospectively in consecutive pts undergoing ES at 17 institutions. Data were collected at time of procedure, prior to discharge and at 30 days. AP was defined by accepted consensus criteria. Results: Of 2,347 ES, 127 (5.4%) developed AP. Of 13 risk factors significant by univariate analysis, only six retained independent significance by multivariate analysis: suspected SOD (odds ratio 5.01), use of precut ES (o.r. 4.34), acinarization (o.r. 3.97), difficult cannulation (o.r. 3.20), >2 PD injections (o.r. 3.07), and age ≤ 60 (o.r. 1.79). CBD diameter and use of manometry (mano) were not independently significant. 272 pts underwent biliary ES (without pancreatic ES or stent) for suspected SOD, with preceding mano in 134. Empirical ES without mano was associated with an identical risk of AP and a trend toward more severe AP. After pancreatic mano, despite use of aspirating catheters, incidence of AP was higher (20/82[24%]) than for isolated CBD mano (4/52[7.7%j), but this risk disappeared when controlling for difficulty of cannulation. Risk of AP was the same regardless of whether PSH (basal pressure ≥40mmHg) was found. Data for biliary ES for suspected SOD: Mano Pancreatitis Severe AP No 28 of 138 (20.3%) 5 (3.6%) Yes*24 of 134 (17.9%) 1 (0.8%)*CBD, PD or both PSH Pancreatitis Present 15 of 62 (24.2%) Absent 5 of 20 (25.0%) Conclusions: Suspected SOD was the most potent risk factor identified for pancreatitis post-ES. Other independent risk factors included use of precut, difficult cannulation, repeated or acinarized pancreatography, and young age. Pancreatitis following ES for suspected SOD cannot be explained by PSH or use of manometry. The widespread use of empirical ES based solely on clinical suspicion of SOD has to be questioned, given its high risk and uncertain benefit.