Quantitative estimation of progression of chronic liver disease using gadoxetate disodium-enhanced magnetic resonance imaging.

Abstract

The purpose of this study was to determine whether the liver stiffness (LS) measured on magnetic resonance (MR) elastography can be estimated by a combination of gadoxetate disodium-enhanced MR imaging (EOB-MRI) and ordinary blood tests. We evaluated 33 consecutive patients with suspected liver disease who underwent EOB-MRI using a Differential Subsampling with Cartesian Ordering MR sequence and MR elastography using a 1.5-T MR system in this prospective study. A stepwise multiple linear regression model analysis of LS was performed using various predictive values obtained from two-in-one-uptake, two-compartment model analysis of EOB-MRI (velocity constants of arterial inflow [K1a ], portal venous inflow [K1p ], hepatocellular uptake [Ki ]), and ordinary blood test results (blood platelet count, serum albumin level [ALB], total serum bilirubin level [T-BIL], and prothrombin time [PT%]). Multiple linear regression model analysis revealed that hepatic perfusion-uptake index (HPUI = -K1a + K1p + Ki ) (P < 0.0001), albumin-bilirubin linear predictor (ALBI-LP = 0.66 × log10 T-BIL-0.085 × ALB) (P = 0.034), and blood platelet count (P = 0.046) were significant independent predictors of LS (r = 0.863). The area under receiver operator characteristics curve of multiple linear regression model in prediction of the liver stiffness corresponding to higher (LS > 5.0kPa) and lower (LS < 4.2kPa) risk for developing hepatocellular carcinoma were 0.956 and 0.938, respectively. LS can be estimated quantitatively with the use of HPUI obtained from compartment model analysis of EOB-MRI combined with ALBI-LP and blood platelet count.