Needle Knife Sphincterotomy for Biliary Access at ERCP: Experience At Two United States Academic Tertiary Referral Centers

Abstract

Introduction: Endoscopic retrograde cholangiopancreatography (ERCP) is a highly effective modality for the management of various panctreaticobiliary disorders. Cannulation of the desired duct is an essential first step towards a successful ERCP. Standard methods of biliary cannulation have been reported to fail in 15-20% of cases. When conventional methods fail, a needle knife or access sphincterotomy (NKS) can be used to achieve bile duct access. Our goal was to evaluate the overall success and complication rates associated with NKS at our institutions, including those cases that were referred after primary biliary cannulation failure at other institutions. Methods: Retrospective review of all patient records where NKS was performed as part of the ERCP procedure at the University of Rochester Medical Center (URMC) and at Geisinger Health System between 4/30/2003 and 5/15/2014 (n=611). Billing records and ERCP databases were used to acquire relevant information. Initial biliary cannulation was attempted using standard techniques; if unsuccessful, a NKS was performed to achieve deep biliary cannulation. In some cases, a pancreatic duct stent was placed before the use of NKS. Results: In 611 patients where NKS was performed during ERCP, the initial success rate for deep biliary cannulation was 85.1%. Repeat ERCP attempt within 7 days resulted in an overall success rate of 93%. There were a total of 48 complications (7.9%): 33 post-ERCP pancreatitis (5.4%), 9 post-sphincterotomy bleeds (1.5%), 4 perforations (0.7%), 1 hepatic abscess (0.2%), and 1 pneumothorax (0.2%). All complications were managed non-surgically. There were no deaths directly associated with the ERCP or NKS procedures. For those patients who were referred after initial failed ERCP at an outside institution and who underwent NKS (n=55), the overall success rate of biliary cannulation was 92.7%. The complication rate in this group was 7.3% (2 post-ERCP pancreatitis, 1 bleed, 1 perforation). In those patients where a pancreatic duct stent was placed prior to NKS (n=159) compared to those with no stent placed (n=452), there was higher rate of overall success rate (96.9% vs. 91.6 %; p=0.03). The overall complication rate for the pancreatic stent group was 10.7% and, for those without pancreatic stent placement was 6.9% (p=0.12). Conclusion: In patients who undergo ERCP at a tertiary referral center, NKS is highly effective with an acceptable complication profile. This is true even in patients where biliary access has previously failed at an outside institution. The use of pancreatic duct stent placement prior to NKS does improve overall biliary cannulation rates without significantly impacting the risk of complications.