Effect of Intravenous Gadolinium on Estimation of Liver Stiffness with Mr Elastography

Methods: Liver MRI and MRE were performed in 32 subjects. The indications for liver MRE were suspected fibrosis/cirrhosis (n=24) and liver tumour (n=8). All MRI studies were performed on a single 1.5 Tesla clinical MR scanner (GE, Milwaukee, USA). The MRE was performed before (pre-Gd MRE) and at the end of standard Liver MRI protocol immediately after the 5 minute delayed phase (post-Gd MRE) in the same session. MRE was performed with modified phase-contrast gradient-echo sequences with TR/TE=100/27 ms FOV=30-46cm, 4 x 10 mm slices, gap 5 mm). The sequence automatically generated stiffness maps (elastograms). Mean stiffness values of the liver parenchyma were measured by an experienced reader and by placing at least three large circular regions of interest in each of the four slices. A mean liver stiffness value in kilopascals (kPa) was calculated by averaging the vales from all the four slices. Liver biopsy revealed fibrosis in 25 patients and there was no fibrosis in seven patients. The pre-Gd MRE and post Gd MRE mean liver stiffness values were compared for any significant differences. Receiver operating curve (ROC) analysis was performed and area under curve (AUC) was measured for comparing accuracies for detection of liver fibrosis.