Background & aims: The gold standard in assessing liver fibrosis is biopsy despite limitations like invasiveness and sampling error and complications including morbidity and mortality. Therefore, there is a major unmet medical need to quantify fibrosis non-invasively to facilitate early diagnosis of chronic liver disease and provide a means to monitor disease progression. The goal of this study was to evaluate the ability of several magnetic resonance imaging (MRI) techniques to stage liver fibrosis.
Methods: A gadolinium (Gd)-based MRI probe targeted to type I collagen (termed EP-3533) was utilized to non-invasively stage liver fibrosis in a carbon tetrachloride (CCl4) mouse model and the results were compared to other MRI techniques including relaxation times, diffusion, and magnetization transfer measurements.
Results: The most sensitive MR biomarker was the change in liver:muscle contrast to noise ratio (ΔCNR) after EP-3533 injection. We observed a strong positive linear correlation between ΔCNR and liver hydroxyproline (i.e. collagen) levels (r=0.89) as well as ΔCNR and conventional Ishak fibrosis scoring. In addition, the area under the receiver operating curve (AUR0C) for distinguishing early (Ishak ≤ 3) from late (Ishak ≥ 4) fibrosis was 0.942 ± 0.052 (p<0.001). By comparison, other MRI techniques were not as sensitive to changes in fibrosis in this model.
Conclusions: We have developed an MRI technique using a collagen-specific probe for diagnosing and staging liver fibrosis, and validated it in the CCl4 mouse model. This approach should provide a better means to monitor disease progression in patients.
Keywords: ADC; AUROC; CCl(4); CNR; ELF; EP-3533; Fibrosis; Gadolinium; Gd; HCC; MR elastography; MRE; MRI; MTR; Molecular imaging; Non-invasive; ROC; SD; SI; SL; TE; TI; Type 1 collagen; apparent diffusion coefficient; area under the ROC; carbon tetrachloride; contrast to noise ratio; echo time; enhanced liver fibrosis; gadolinium; hepatocellular carcinoma; intravenous; inversion time; iv; magnetic resonance imaging; magnetization transfer ratio; receiver operating curve; signal intensity; spin lock time; standard deviation.
Copyright © 2013 European Association for the Study of the Liver. Published by Elsevier B.V. All rights reserved.