Use of magnetic resonance cholangiopancreatography in clinical practice: Not as good as we once thought

Background Magnetic resonance cholangiopancreatography (MRCP) is believed to be a useful tool to evaluate the biliary tree and pancreas for stones, tumors, or injuries to the ductile system. The purpose of this study was to compare the accuracy of MRCP to the gold standard, endoscopic retrograde cholangiopancreatography (ERCP), in our institution. Study Design We performed a retrospective review of all MRCP followed by ERCP (follow-on ERCP) at a single institution over a 6-year period. Exam findings from MRCP were compared with findings on the follow-on ERCP and compared. Studies were grouped into 2 main classifications: tests being performed for patients with suspected choledocholithiasis (stone disease) and tests being performed for concerns of malignant strictures or duct injuries (non-stone disease). Results A total of 81 patients had MRCPs and follow-on ERCPs in this time period. Thirty-six patients had positive findings on MRCP and ERCP for stones in the common duct system, and 14 patients had positive findings on MRCP and subsequent ERCP for masses and strictures of the common duct. Three patients had positive MRCP and ERCP findings for pancreatic duct abnormalities. The specificity and positive predictive value of MRCP were 94% and 98%, respectively. However, 13 of 28 patients had lesions identified on ERCP after a normal MRCP. The sensitivity and negative predictive value were 80% and 54%, respectively. Conclusions Magnetic resonance cholangiopancreatography was not useful in the management algorithm of either stone or non-stone disease of the biliary tree or pancreas. It should be abandoned as a diagnostic tool for work-up of biliary duct pathology.