4671 A randomized clinical trial comparing ercp and mrcp in an effectiveness setting.

Abstract

BACKGROUND: Magnetic Resonance Cholangiopancreatography (MRCP) can accurately image the pancreaticobiliary system without significant risks. Rapid and widespread diffusion of MRCP has occurred in the absence of cost-effectiveness data comparing it to the gold standard of ERCP. AIM: We performed a prospective randomized multicenter clinical trial to study the effectiveness (real-life setting) and costs of MRCP and ERCP in patients with suspected biliary obstruction. METHODS: Over 24 months, 162 patients with a clinical suspicion of obstruction and/or dilated duct on ultrasound (but without a definitive etiologic diagnosis), who were considered for either ERCP or MRCP, were randomized to either undergo MRCP or ERCP using block allocation stratified by level of obstruction on ultrasound. These patients were either at intermediate risk for obstruction or required further imaging because the next step in management was not clear. The main endpoint was the occurrence of a negative outcome attributable to the biliopancreatic disease or the procedure within 12 months of randomization. Secondary outcomes included mortality and accuracy for the detection of given pathologies. An intention to treat analysis was carried out. Standard statistical tests were performed. RESULTS: 82 patients to date have been randomized to MRCP and 80 to ERCP. Mean age in each group was 50.0±25.1 (SD) and 56.6±18.4 yrs, respectively, with 33% men in both groups. 40 (48.8%) patients randomized to MRCP required an ERCP after their MRCP for various reasons, 80% of which were therapeutic. At 12 months, in the MRCP group, 6 had died, 25 were cured, 1 was palliated, and 1 was unchanged vs 5 who had died, 19 cured, 1 palliated, and 3 unchanged in the ERCP group. Excluding one outlier in the MRCP group who suffered multiple complications and had a hospital stay of 157 days, 19 complications occurred in the MRCP group in 15 (18%) patients with a total length of stay of 79 days (0.98 d/patient randomized) vs 14 in the ERCP group in 10 (13%) patients accounting for a total of 63 days in hospital (0.79 d/patient randomized); a difference of 5% of patients (95%CI: -6,16). CONCLUSIONS: For all outcomes, no statistically or clinically significant differences between ERCP and MRCP for patients with suspected biliary obstruction exist to date. As anticipated by our preliminary sample size calculations, further accrual of patients is required to achieve adequate statistical power. Subgroup examinations are required to determine the optimal role of this new diagnostic modality.