Utility of magnetic resonance cholangiography in the evaluation of biliaryobstruction

Background: Evaluation of suspected biliary obstruction has traditionally i nvolved a variety of imaging modalities including ultrasound, CT, and invas ive cholangiography. These techniques have limitations because of poor visu alization of intraductal stones (ultrasound and CT) and the need for an inv asive procedure (ERCP and percutaneous transhepatic cholangiography). Magne tic resonance cholangiography (MRC) is a noninvasive imaging modality that provides good visualization of the hepatobiliary system. The aim of the pre sent study was to determine the utility of MRC in evaluating patients with suspected biliary obstruction. Study Design: One hundred forty-three patients were identified with suspect ed acute biliary obstruction and underwent MRC. Patient selection was based on clinical criteria including an elevation in serum liver chemistries or evidence of biliary ductal dilatation on conventional imaging. MRC was perf ormed using a half-Fourier acquisition single-shot turbo spin-echo sequence involving single breath-hold rapid image acquisition. A final diagnosis wa s determined in each patient based on invasive cholangiography, findings at surgery, and clinical course. Results: Of the 143 patients, 73 had an obstructing biliary lesion. A malig nant process was identified in 25 patients with final diagnoses of pancreat ic cancer (n = 15), ampullary cancer (n = 4), cholangiocarcinoma (n = 3), a nd hepatic or nodal metastases (n = 3). MRC correctly identified biliary ob struction in all these patients and accurately identified the level of bili ary obstruction in 24 of 25 patients. Based on the MRC images alone, a mali gnant process was suspected in 21 of the 25 patients. Forty patients were f ound to have common bile duct stones and eight patients had a benign distal bile duct stricture. MRC correctly identified common bile duct stones in 3 7 patients with one false-positive exam (sensitivity = 32%; specificity = 9 9%). MRC also correctly identified distal biliary strictures in eight patie nts. In the remaining 70 patients, no definite biliary obstruction was iden tified by MRC, and in all patients the absence of mechanical obstruction wa s confirmed by invasive cholangiography or overall clinical course. Conclusions: This study demonstrates that MRC is able to accurately identif y the level and cause of biliary obstruction in both malignant and benign d isease. MRC may prove to be an important noninvasive tool in preoperative e valuation of patients with suspected biliary obstruction and identification of patients most likely to benefit from an invasive radiologic or surgical procedure. (J Am Cell Surg 1999;189:63-72. (C) 1999 by the American Colleg e of Surgeons).