Abstract
Objective Biliary anatomy is of paramount importance for hepatobiliary pancreatic surgeons for operative planning. Preoperative assessment with magnetic resonance cholangiopancreatography (MRCP) to evaluate the biliary anatomy plays a vital role, especially for prospective liver donors in living donor liver transplantation (LDLT). Our objective was to evaluate the diagnostic accuracy of MRCP in assessing the anatomical variations of the biliary system and the frequency of biliary variation in the donors of LDLT. Materials and Methods Sixty-five donors of living donor liver transplantation in the age range of 20 to 51 years were studied retrospectively to evaluate the anatomical variations of the biliary tree. As a part of the pre-transplantation donor workup, MRI with MRCP was performed in a 1.5T machine for all these candidates. MRCP source data sets were processed with maximum intensity projections, surface shading, and multi-planar reconstructions. Images were reviewed by two radiologists, and the classification system of Huang et al. was utilized to evaluate the biliary anatomy. The results were compared with the intraoperative cholangiogram, considered the gold standard. Results We identified standard biliary anatomy in 34 candidates (52.3%), and variant biliary anatomy was observed in 31 candidates (47.7%) on MRCP. An intraoperative cholangiogram showed standard anatomy in 36 candidates (55.4%) and biliary variation in 29 candidates (44.6%). Our study showed a sensitivity of 100% and a specificity of 94.5% for identifying biliary variant anatomy on MRCP in comparison with the gold standard intraoperative cholangiogram. The accuracy of MRCP in detecting the variant biliary anatomy in our study was 96.9%. The most common biliary variation was the right posterior sectoral duct draining into the left hepatic duct, Huang type A3. Conclusion The frequency of biliary variations is high in potential liver donors. MRCP is sensitive and highly accurate in identifying the biliary variations of surgical significance.