Abstracts
1997 Digestive Disease Week

Magnetic resonance cholangiography compared to ultrasound and other clinical criteria in the preoperative prediction of choledocholithiasis.

TH Magnuson, F Regan, SA Ahrendt, DC Schaefer, JS Bender. Departments of Surgery and Radiology, The Johns Hopkins Medical Institutions, Baltimore, MD.

The ability to preoperatively detect common bile duct stones (CBDS) has assumed increasing importance with the widespread acceptance of laparoscopic cholecystectomy. Clinical criteria such as elevated serum bilirubin, dilated bile duct on ultrasound, and history of jaundice or pancreatitis have been used to identify patients at risk for CBDS who would benefit from pre-op ERCP. Unfortunately, many patients with abnormalities in these criteria will prove to have a normal bile duct and undergo an unnecessary invasive ERCP. We hypothesized that magnetic resonance cholangiography (MRC) might be a better noninvasive predictor of CBDS prior to laparoscopic cholecystectomy. Twenty-nine patients were identified prospectively who were felt to be at risk for CBDS based on clinical criteria. Each patient underwent either preoperative ERCP or intraop cholangiography to definitively determine the presence or absence of CBDS. MRC was performed <24 hrs. prior to invasive cholangiography (T2 weighted HASTE technique). Each MRC was evaluated for the presence or absence of common bile duct stones and blinded to the results of definitive invasive cholangiography. Overall 11 patients were found to have CBDS on ERCP/IOC, while 18 patients had no CBDS present. The sensitivity and specificity of MRC and ultrasound as predictors of CBDS was:

                               Sensitivity       Specificity
MRC (+ CBDS)                        91%               94%
Ultrasound (+ CBDS)                 18%               100%
Bile Duct > 9 mm (ultrasound)       36%               67%

Clinical predictors such as serum bilirubin >2.0 (sensit =73%; specif=33%), alk phos >180 (sensit=55%; specif=56%) and Amylase >500 (sensit=18%; specif=61%) were all inferior to MRC. In addition, MRC correctly demonstrated acute cholecystitis in 6 of these patients, later confirmed by gallbladder histology. These data suggest that MRC is a better predictor of CBDS than traditional clinical criteria. In patients with suspected CBDS, MRC may prove to be an important noninvasive tool to select patients most likely to benefit from preop ERCP.