Hepatic Function Imaging Using Dynamic Gd-EOB-DTPA Enhanced MRI and Pharmacokinetic Modeling

Jia Ning,Zhiying Yang,Sheng Xie,Yongliang Sun,Chun Yuan,Huijun Chen
DOI: https://doi.org/10.1002/mrm.26520
2016-01-01
Magnetic Resonance in Medicine
Abstract:PurposeTo determine whether pharmacokinetic modeling parameters with different output assumptions of dynamic contrast-enhanced MRI (DCE-MRI) using Gd-EOB-DTPA correlate with serum-based liver function tests, and compare the goodness of fit of the different output assumptions. MethodsA 6-min DCE-MRI protocol was performed in 38 patients. Four dual-input two-compartment models with different output assumptions and a published one-compartment model were used to calculate hepatic function parameters. The Akaike information criterion fitting error was used to evaluate the goodness of fit. Imaging-based hepatic function parameters were compared with blood chemistry using correlation with multiple comparison correction. ResultsThe dual-input two-compartment model assuming venous flow equals arterial flow plus portal venous flow and no bile duct output better described the liver tissue enhancement with low fitting error and high correlation with blood chemistry. The relative uptake rate K-i(r) derived from this model was found to be significantly correlated with direct bilirubin (r= -0.52, P=0.015), prealbumin concentration (r=0.58, P= 0.015), and prothrombin time (r=-0.51, P=0.026). ConclusionIt is feasible to evaluate hepatic function by proper output assumptions. The relative uptake rate has the potential to serve as a biomarker of function. Magn Reson Med 78:1488-1495, 2017. (c) 2016 International Society for Magnetic Resonance in Medicine.
What problem does this paper attempt to address?