Surgical versus endoscopic treatment of bile duct stones

Abstract

Background: 10% to 18% of patients undergoing cholecystectomy for gallstones have common bile duct (CBD) stones. Treatment options for these stones include pre- or post-operative endoscopic retrograde cholangiopancreatography (ERCP) or open or laparoscopic surgery.

Objectives: To systematically review the management of CBD stones by four approaches: (1) ERCP versus open surgical bile duct clearance. (2) Pre-operative ERCP versus laparoscopic bile duct clearance. (3) Post-operative ERCP versus laparoscopic bile duct clearance. (4) ERCP versus laparoscopic bile duct clearance in patients with previous cholecystectomy.

Search strategy: We systematically searched key relevant electronic databases, bibliographies of relevant papers, and abstracts of relevant subspecialty meetings until November 2005.

Selection criteria: The quality of included trials was assessed by adequacy of allocation sequence generation, allocation concealment, blinding, and follow-up.

Data collection and analysis: Published and unpublished data relevant to 12 predefined outcome measures were used to conduct fixed- and random-effects models meta-analyses, with exploration of heterogeneity and use of sensitivity and subgroup analysis where required.

Main results: Thirteen trials randomised 1351 patients. Eight trials (n = 760) compared ERCP with open surgical clearance, three (n = 425) compared pre-operative ERCP with laparoscopic clearance, and two (n = 166) compared post-operative ERCP with laparoscopic clearance. There were no trials of ERCP versus laparoscopic clearance in patients without an intact gallbladder. Methodology was considered adequate in at least two of three assessable fields in ten trials. A significantly increased number of total procedures (including for complications) per patient was seen in the ERCP arms in all three comparisons with weighted mean differences of 0.62 (95% CI 0.15 to 1.09), 0.96 (95% CI 0.96 to 0.96), and 1.09 (95% CI 0.93 to 1.24), respectively. ERCP was less successful than open surgery in CBD stone clearance (Peto OR 2.89, 95% CI 1.81 to 4.61) with a tendency towards higher mortality (risk difference 1%, 95% CI -1% to 4%). Laparoscopic CBD stone clearance was as efficient as pre- (Peto OR 1.00, CI 0.53 to 1.80) and post-operative ERCP (OR 2.27, 95% CI 0.37 to 13.9) and with no significant difference in morbidity and mortality. Laparoscopic trials universally reported shorter hospital stays in surgical arms. Insufficient data were reported for cost analysis.

Authors' conclusions: In the era of open cholecystectomy, open bile duct surgery was superior to ERCP in achieving CBD stone clearance. In the laparoscopic era, data are close to excluding a significant difference between laparoscopic and ERCP clearance of CBD stones. The use of ERCP necessitates increased number of procedures per patient.