Diagnostic accuracy of preoperative magnetic resonance cholangiopancreatography in delineating biliary anatomy in live liver donors: An ambispective observational study.

BACKGROUNDS/AIMS: Accurate preoperative delineation of biliary anatomy is essential for living donor liver transplantation (LDLT), as unrecognised variants contribute to biliary complications. This study evaluated the diagnostic accuracy of magnetic resonance cholangiopancreatography (MRCP) for characterising biliary anatomy in live donors, using intraoperative cholangiography (IOC) as the reference standard.

METHODS

An ambispective observational study included 89 live liver donors (65 retrospective, 24 prospective) at a tertiary LDLT centre. Biliary anatomy was classified on MRCP and IOC using the Huang classification. Variant anatomy (Huang types A2-A5) was defined as the positive test condition and standard anatomy (A1) as the negative condition. Diagnostic performance was assessed using sensitivity, specificity, positive and negative predictive values, and Cohen's kappa.

RESULTS

The cohort comprised 58 females (65.2%) and 31 males (34.8%), with a mean age of 34.34 ± 9.53 years. Huang type A1 was most common on MRCP (65.2%) and IOC (59.6%), while variants (A2-A5) were detected in 34.8% and 40.4%, respectively. Against IOC, MRCP showed 77.78% sensitivity (95% CI, 61.92%-88.28%), 94.34% specificity (95% CI, 84.63%-98.06%), 90.32% positive predictive value (95% CI, 75.10%-96.65%), 86.21% negative predictive value (95% CI, 75.07%-92.84%), and 87.64% overall accuracy, with substantial agreement (κ = 0.737, p < 0.001).

CONCLUSIONS

MRCP is highly specific for preoperative biliary mapping in LDLT donors but has limited sensitivity for variants, supporting routine IOC before ductal transection.