Evaluation of the efficacy and reproducibility of cholangiopancreatography by magnetic resonance for detecting biliary complications following orthotopic liver transplantation

Abstract

To evaluate the accuracy and reproducibility of magnetic resonance cholangiopancreatography (MRCP) in the detection of biliary complications in liver transplanted patients. A study was conducted, with blinded review of 28 MRCP exams of 24 patients submitted to liver transplantation. The images were reviewed by two independent observers, at two different moments, regarding the degree of biliary tree visualization and the presence or absence of biliary complications. The MRCP results were compared, when negative, to at least 3 months of clinical and biochemical follow-up, and when positive, to the findings at surgery or endoscopic retrograde cholangiopancreatography (ERCP). The degree of intrahepatic biliary tree visualization was considered good or excellent in 78.6% and 82.1% of the exams by the two observers and visualization of the donor duct, recipient duct and biliary anastomosis was considered good or excellent in 100% of the exams, by both observers. Six biliary complications were detected (21.4%), all of them anastomotic strictures. Intra and interobserver agreement were substantial or almost perfect (kappa k values of 0.611 to 0.804) for the visualization of the biliary tree and almost perfect (k values of 0.900 to 1.000) for the detection of biliary complications. MRCP achieved 100% sensitivity, 95.45% specificity, 85.7% positive predictive value and 100% negative predictive value for the detection of biliary complications. MRCP is an accurate examination for the detection of biliary complications after orthotopic liver transplantation and it is a highly reproducible method in the evaluation of the biliary tree of liver transplanted patients.