The purpose of this study is to prospectively compare the efficacy of endoscopic ultrasound-guided biliary interventions (EUS-rendezvous or direct EUS-guided therapy) to that of advanced ERCP cannulation techniques in the setting of difficult cannulation. We hypothesize that EUS-guided interventions will be more successful and may be associated with the same complications as using advanced ERCP techniques in cases of difficult bile duct access.

Background:

Biliary tract obstruction from stones or strictures are usually treated using endoscopic retrograde cholangiopancreatography (ERCP) techniques. However, biliary access (cannulation) during ERCP can be difficult. Overall cannulation success is about 90-95% at expert, centers. Difficult cannulation is associated with increased complication rates, as prolonged and repeated attempts can cause injury to the ampulla. When standard biliary cannulation fails advanced cannulation techniques, such as precut access sphincterotomy, are often indicated. These advanced techniques, however, increase the complication rate of ERCP, and still fail in a certain percentage of cases.

Endoscopic ultrasound-guided biliary interventions are relatively newer techniques that have emerged as a useful option to achieve biliary access or therapy when standard and advanced ERCP techniques for biliary access have failed. Many centers, including ours, are currently using EUS-guided interventions to facilitate ERCP access and therapy, when needed. EUS-guided interventions have emerged as a safe and effective procedure for gaining access to the biliary tree. A recent retrospective study comparing the safety and efficacy of EUS rendezvous to precut sphincterotomy (advanced ERCP technique for access), found that EUS-rendezvous was significantly more successful with no significant difference in the rate of procedural complications.

Currently, there are no accepted standards for deciding which biliary access technique to use in cases of difficult biliary cannulation. We hypothesize that EUS-guided biliary interventions will be more successful and may be associated with the same (or less) complications than using advanced ERCP techniques in cases of difficult bile duct access.

Procedures:

Patients referred for ERCP for biliary interventions at participating centers will be approached for consent. Patients who agree to participate will undergo a standard ERCP. The attending endoscopist will use standard ERCP techniques (i.e. cannulation with a standard cannula or sphincterotome with or without a wire) to gain biliary access for a minimum of ten minutes. If biliary access fails, the patient will be randomized to either the EUS-guided intervention arm or the advanced cannulation techniques arm of the study. The randomization ratio will be 50-50.

Methods of Data Analysis:

Based on current available data as well as our own experience, we estimated the success for immediate biliary access for EUS-guided intervention to be 90%, and for advanced ERCP cannulation techniques to be 75%. To detect this difference with power of 0.8 and a p-value of <0.05, a sample size of approximately 230 patients will be needed. Assuming that only 10% of patients that are consented will be randomized (based on the rationale that 90% of patients will either be successfully cannulated within 10 or will be excluded for other reasons) we estimate that approximately 2,300 patients will be consented for this study.

As this is the first study of its kind and it is possible that we will achieve statistical significance at an enrollment number earlier than our anticipated number of 230 participants two interim analyses are planned for this study. The first interim analysis is scheduled to occur after the enrollment of 70 patients (35 in each arm) and the second after enrollment of 140 patients (70 in each arm). The block randomization should result in equal numbers of patients in each arm at these enrollment levels. Adjustment for significance level at each interim analysis was calculated using the Pocock Boundary method (9). For either of the interim analyses, the level of significance required for termination of the trial for efficacy is p = 0.0221. An independent statistician will review the results and make recommendations regarding study termination based on results of interim analyses.

Inclusion Criteria:

• Patients able to give informed consent
• Patients referred to participating centers for ERCP with the intent to cannulate the bile duct for therapy

Exclusion Criteria:

• Patients with a bleeding diathesis (INR >1.4, platelets <80, current use of anticoagulant medications)
• Patients whose bile ducts are cannulated using standard methods
• Patients with prior biliary sphincterotomy or endoscopic papilla dilation
• Patients with previous surgical biliary-intestinal operations
• Patients with pancreas divisum
• Patients with indwelling pancreatic or biliary stent placement
• Patients who are pregnant or breast-feeding