Efficacy and safety of primary, early and late needle-knife fistulotomy for biliary access

European Society of Gastrointestinal Endoscopy recommends needle-knife fistulotomy (NKF) as the preferred precut technique. However, there is little information on whether NKF performed at different times is associated with different success and adverse event rates. We compared the outcomes of 3 different timings of NKF. This was an observational study conducted at 4 institutions and this was a retrospective analysis of prospectively collected data. We included 330 consecutive patients submitted to NKF attempt for biliary access. Patients were divided into three groups: NKF as an initial procedure for biliary access (group A, n = 121); early NKF defined as after 5 min, 5 attempts, or 2 pancreatic passages (group B, n = 99); and late NKF: after at least 10 min of unsuccessful standard biliary cannulation (group C, n = 110). We assessed the success rate of biliary cannulation at initial ERCP, time to perform NKF until biliary cannulation, overall biliary cannulation rate (second ERCP when initial failure), adverse event rate, and predictors of post-ERCP pancreatitis (PEP). The initial cannulation rate was 98%, 91% and 94% for groups A, B and C respectively, p = 0.08, whereas overall biliary cannulation rate was 100%, 95% and 98%, p = 0.115. The adverse event rate/PEP was 4.1%/2.5%, 7.1%/4% and 10.9%/8.2%, for groups A, B and C respectively, (p = 0.197 and p = 0.190). Median time for creating the fistula was A = 4.0 min, B = 3.2 min, and C = 5.6 min, p < 000.1. Each additional minute spent attempting cannulation increased the odds ratio (OR) for PEP by 1.072, and patients with 3 or more risk factors for pancreatitis had a higher chance of PEP. In conclusion, the timing of NFK does not appear to influence success rates but late NFK is associated with a higher time to create a fistula and an increased risk of pancreatitis. Primary NFK is associated with a high rate of success and a low rate of PEP and deserves additional investigation.