The Role of EUS-Assisted Biliary Drainage after Failed ERCP

Abstract

Purpose: To evaluate technical success and outcomes of EUS-assisted biliary drainage after ERCP attempts at a tertiary center have failed. Methods: Patients included all those with failed ERCP for distal biliary obstruction at a tertiary center over 9 years, in whom repeat ERCP was felt unlikely to succeed. EUS-assisted biliary drainage was performed under general anesthesia with fluoroscopy, and categorized into 2 methods: 1) EUS rendezvous for transpapillary access followed by ERCP and 2) direct EUS guided transmural biliary drainage. If ampulla was accessible at initial ERCP, EUS rendezvous was attempted first. In cases of inaccessible ampulla and inoperable malignancy, direct EUS guided transmural drainage was performed. 1) For rendezvous, EUS-assisted transduodenal (N = 12) or transhepatic (N = 1) bile duct puncture was performed via a diagnostic linear EUS scope with a 19 or 22 gauge needle; a guidewire was advanced through papilla by fluoroscopy; the guidewire left in place, and ERCP performed immediately afterwards with stent insertion ± sphincterotomy. 2) Direct EUS-guided biliary drainage was performed through a transduodenal approach, the fistula tract dilated, and metallic stents placed. If bile duct access failed by all methods, patients were immediately converted to PTC. Results: EUS-BD was attempted in 15 patients (mean age 68, malignant 12/benign 3, CBD diameter 4–20 mm). EUS rendezvous was attempted in 13 cases, and direct EUS guided transmural biliary drainage in 2. Reasons for initial ERCP failure included tumor distorting papilla (N = 8), duodenal stents (N = 2), intradiverticular papilla (N = 1), or other anatomic anomalies (N = 4). EUS-assisted biliary drainage was performed at the same session as initial ERCP attempt in 12/15 patients. EUS-guided bile duct puncture was achieved in all 15 (100%) patients, with drainage successfully completed in 12/15 (80%). Failures occurred in 3 attempted rendezvous cases because of inability to traverse biliary stricture (N = 2) or dissecting a choledochocele with guidewire (N = 1); all were successfully drained via PTC. Stents placed were metallic in 9 and plastic in 3. Complications occurred in 2/15 patients (13%); 1 moderate pancreatitis after difficult ERCP attempt in papillary stenosis, and 1 bacteremia after PTC, with no perforations. Mean hospital stay was 5.4 (0–33) days, mostly for preexisting medical problems. Conclusion: EUS-assisted biliary drainage using a preferentially transduodenal and rendezvous approach demonstated a high success rate without any complications attributable to the EUS access. Advantages over PTC include performance under the same anesthesia as initial ERCP attempt, and internal drainage with access achieved by a very small caliber needle puncture similar to EUS/FNA.Table: Overtube Assisted Endoscopic Retrograde Cholangiography in Roux-en-Y Anatomy