Preliminary results of a prospective randomized trial of pre-cut papillotomy (PC) vs persistence(Pers)to achieve biliary cannulation

Background: There is controversy as to the safety and efficacy of pre-cut techniques to achieve biliary cannulation. The recognized higher complication rate has been attributed to patient selection (eg.S.O.D.) and difficulty of cannulation rather than the technique itself. To date, no randomized trials have addressed this issue. Study Aim: To compare the success rates and complications of PC vs Pers for difficult biliary cannulation. Methods: Pts in whom cannulation was deemed necessary and with no prior sphincterotomy were considered for entry into the study. Exclusions included coagulopathy, Billroth II anatomy and a papilla within a diverticulum. Bile duct cannulation was attempted in the usual fashion for 7 min by the fellow and 5 min by staff (GH, PK). If unsuccessful, pts were randomized to PC or Pers, for an additional 15 min. of cannulation time. At expiry of this time limit, farther cannulation attempts were left to the discretion of the staff endoscopist. Study success was defined as cannulation within 15 min. of randomization. Complications leading to blood transfusions, hospitalization, repeat ERCP or surgery were compared. Telephone follow-up was obtained at 1 and 4 wks post ERCP to assess complication rate; hospital charts were reviewed and physicians contacted when indicated. Results: To date (06/95-11/96), 237 pts have been enrolled. Forty percent of the study grp and 19% of the controls were referred after prior failed ERCP. 90% (212/237) of CBDs were cannulated within 12 min; these pts comprise the control(ctl) group. Of the remaining 25 pts (study grp) 14 were randomized to PC and 11 to Pers. Success rates were: 8/11pts (73%) for Pers and 9/14 (64%) for PC grp with mean cannulation times of 18.3 min and 20.3 min respectively. Study failures; 213 Pers and 5 PC, were cannulated after 33.3 min and 46.1 min respectively at the initial ERCP. One other (Pers grp) was successfully cannulated 1 week later, using the same method. Sphincterotomies were performed in 140/212 ctls (66%), 8/14 PC pts, and 6/11 Pers pts. Study grp diagnoses included: 3/25 CBD stones, 4/25 S.O.D, 7/25 normal exams, 11/25 malignant strictures. Bleeding (controlled with local injection of adrenalin) occurred in 9% of the Pers grp, 14.2% of the PC grp, and 6.6% of the ctl grp. Only 1 delayed bleed required transfusion (ctl grp). Pancreatitis, all cases mild /moderate (avg length of hospital stay = 3.5 days), occurred in 9% of the Pers grp, 7% of the PC grp and 2.2% of the ctl group. Conclusion: Cannulation was achieved in 100% of enrolled pts, 99.5%of them at same (index) session. Preliminary results suggest that there are no significant differences in either success or complication rates when using PC or Pers in difficult CBD cannulations. Pancreatitis, but not bleeding, is more frequent than in controls in this high risk population. There is a trend towards longer time to cannulation in cases of failed PC. This study is ongoing.