From February 2005 to February 2006, a total of 22 patients were enrolled. All enrolled patients met the following selection criteria: a) referral for ERCP because of symptoms of bile duct stones, b) over 18 years of age, c) informed consent obtained before ERCP, d) bile duct stones visualized at ERCP, and e) deep cannulation of the bile duct achieved without sphincterotomy. Patients were not eligible if they had signs of acute pancreatitis (severe epigastric pain combined with serum amylase more than three times the upper normal limit), acute cholecystitis (localized pain in the right upper abdomen, fever, and a thickened gallbladder wall on ultrasonography), or a history of previous sphincterotomy, choledochoduodenal fistula, hemostatic disorders, intrahepatic stone diseases, concomitant pancreatic or biliary malignant disorders, co-existing bile leakage or choledochoduodenal fistula.

The endoscopic procedure for the study was performed by single expert in ERCP. ERCP was done with Olympus TJF-240 side-viewing endoscopes (Olympus Optical Corporation, Tokyo, Japan). Patients were sedated with midazolam (3-4 mg) supplemented by an intramuscular injection of 20 mg meperidine. Prophylactic antibiotics were not routinely given. To stop duodenal peristalsis, 40 mg butylscopolamine was administered intravenously just before the start of ERCP. The major papilla was located, the bile duct was deeply cannulated with 0.035 Fr catheter (ERCP-Katheter MTW® endoskopie, Wesel, Germany), and a diagnostic cholangiogram was obtained. For further procedure, a guide wire (Hydra Jagwire™ guidewire, Boston Scientific Corp. Natick, MA, USA) was passed through a diagnostic catheter into the bile duct and then the catheter was removed. Limited EST with a pull-type sphincterotome (Papillotome, MTW® endoskopie, Germany) was done for easy insertion of the balloon catheter, control of the direction of balloon dilation during EBD with large balloon. A 5.5 cm, 12 mm or a 5.5 cm, 15 mm balloon catheter for pyloric dilation (CRE wire-guided dilator, Boston Scientific Corp. Natick, MA, USA) was then passed over the guide wire and positioned across the papilla. Each balloon was gradually expanded to 10-15 mm, depending on the maximal diameter of the CBD, measured by cholangiography. The sphincter was considered adequately dilated if the waist in the balloon had disappeared completely. The fully expanded balloon was maintained in position for 40 seconds and then deflated and removed (Fig. 1). After EBD, the stones were retrieved using a Dormia basket (Web™ extraction basket, Wilson-Cook Medical Inc. Winston-Salem, North Carolina, USA) and/or a retrieval balloon catheter (Two lumen retrieval balloon catheter, Boston Scientific Corp. Natick, MA, USA). If necessary, mechanical lithotripsy (Lithotriptoren, MTW® endoskopie, Germany) was performed to fragment the stones prior to extraction from the bile duct. When the bile juice drained to the major papilla was so sticky or mixed with sludges, a 7 Fr, nasobiliary drainage catheter (ENBD-7-Liguory, Wilson-Cook Medical Inc., Winston-Salem, North Carolina, USA) was inserted. And a follow-up cholangiogram was obtained 3 days after the initial procedure. If remnant stones were found, a second ERCP with or without repeated EBD was performed for retrieval of bile duct stones.

Before ERCP and 1 and 3 days after ERCP, blood samples were taken for a complete blood count, liver-function tests (bilirubin, alanine aminotransferase, aspartate aminotransferase, alkaline phosphatase, and γ-glutamyl transpeptidase), measurement of serum amylase, lipase and coagulation profiles. During the initial ERCP, the stone size and number were documented on cholangiograms, taken during the initial filling of the bile duct, after optimum opacification of the biliary tree. Stone size was assessed by comparing the diameter of the stone with the tip of the endoscope, as measured on the cholangiogram. Complete stone removal was verified either by the final cholangiogram or by the combination of initial fluoroscopic findings and the follow-up cholangiogram obtained 3 days after the initial ERCP through a nasobiliary drainage catheter.

The primary endpoint was the success rate of complete removal of stones with the initial ERCP session. Secondary outcomes measured were the number of ERCP sessions required for complete stone removal, the frequency of use of mechanical lithotripsy, and associated complications. All complications were classified and graded according to the 1991 consensus guidelines.13

Statistical analyses were performed using statistical software (SPSS 11.5 for Windows; SPSS Inc., Chicago, Ill, USA). Categorical and binary variables were tested by the chi-squared test with Yates' correction or Fisher's exact test if more than 20% of the cells in the frequency tables had an expected frequency below 5.

In all patients, EBD of the biliary sphincter was successful and complete retrieval of bile duct stones was achieved. Mean duration of the whole procedure was 32.9 ± 13.7 minutes (Table 2). Successful stone removal in the initial session of ERCP with EBD was accomplished in 16 patients (72.7%). It was not possible to achieve complete stone removal during the initial procedure in six patients. In four of these 6 cases, small-sized remnant stones, not found at the end of the initial procedure, were detected in the follow-up cholangiography through the nasobiliary drainage catheter 3 days later. In the remaining two cases, the failure to remove stones was due to severe pain and poor cooperation of patients during the initial procedure. In all 6 cases, complete stone removal was ultimately accomplished by an additional session of ERCP without further EBD or EST. Mechanical lithotripsy was necessary in only two patients.

When the patients were sub-grouped according to stone diameter (less or greater than 10 mm) or number of stones (single or multiple), the success rate of complete stone removal after one session of ERCP for a stone < 10 mm in diameter tends to be higher than with a stone > 10 mm in diameter (87.5% vs. 64.3%, respectively). And the success rate of complete stone retrieval in a single session of ERCP for patients with a single stone seemed to be also better than in patients with multiple stones (83.3% vs. 60%, respectively) (Table 3). However, the results did not show statistical significances due to small size of subgroups.

With respect to complications, EBD with large balloon after minor EST proved to be very safe. In our study group, only one patient developed mild-grade post-ERCP pancreatitis. Further, hemorrhage did not occur in any patients. In 3 cases, minor oozing that spontaneously stopped during the procedure was noted. Fatal complications such as perforation or severe pancreatitis did not occur. However, asymptomatic elevation of serum amylase/lipase was noted in 68.2% (15/22) of the patients. The elevated serum amylase/lipase usually normalized within 3 days after the procedure and did not affect the clinical course of the patients. During follow-up of at least 3 months after complete stone retrieval, any complications including recurrence of bile duct stones, cholangitis, bile duct stenosis or acute cholecystitis were documented.