Needle knife sphincterotomy

Introduction: Bile duct cannulation fails in 5-10% of ERCPs. The use of needle knife sphincterotomy (NKS) in such cases is controversial because of the perceived risk and the experience required. Aim: To assess the efficacy, safety and outcome of NKS in a tertiary referral center. Methods: All cases of NKS from 2/94-10/96 were reviewed for indication, success of access, findings, and outcome. Complications were classified according to established criteria (Cotton, 1991) Results: NKS was performed in 119 of 937 sphincterotomies (12.7%) among 2416 total ERCP's (5%). The mean age was 57 yrs (range 4-95); 55% were female. Most NKS were used only after failure with a variety of accessories. The papilla appeared normal in 55 (47%), small in 12 (10%), bulging 26 (22%), neoplastic 8 (7%), fractured in 6 and edematous in 6. NKS was used for anticipated diagnostic studies in 27 (23%), for cholangiography with anticipated therapy in 59 (49%), and for therapy after cholangiography in 25 (21%). Bile duct access was achieved via NKS in 100 pts (85%), including 94 (80%) at the first ERCP and 6/6 at a second ERCP. Of 19 pts. with failed access 8 underwent diagnostic (2) or therapeutic (6) PTC, and 11 were not further studied. Standard sphincterotomy was completed in 58 pts. Final diagnoses were: bile duct stricture 41(34%) (malignant 20, ampullary ca 8, PSC 7, benign 6), biliary stones 30(25%), non-malignant pancreatic diagnoses 20(20%) and normal or not pursued 28 (24%). Complications occurred in 28 pts. (23.5%). Mild (%) Moderate (%) Seven (%) Total (%) Pancreatitis 6 (5) 6 (5) 1 (0.8) 17(14.2) Bleeding 6 (5)* 1 (0.8)# 0 7 (5.8) Perforation 1(0.8) 1 (0.8) 0 2 (1.6) Cholangitis 1 (0.8) 1 (0.8) 0 2 (1.6) Bleeding(*) was limited to observed oozing with universal response to injection therapy at the initial ERCP. One pt(#) had no evidence of bleeding but Hgb fell 2g and he received 3 U pRBC. Mean hospital stay attributable to NKS complications was 3.8 days (range 0-24). One pt required surgery for complications. There was no procedure-related mortality. Conclusions: 1. NKS is an effective method for gaining bile duct access in pts with failed cannulation. 2. Though not infrequent, most complications were minor, 3. The proportion of complications due to prior failed cannulation is unknown, 4. Decisions to utilize NKS should consider the indication, expected intervention, experience, and availability and risks of alternative modalities.