Magnetic resonance and ultrasound techniques for the evaluation of hepatic fibrosis†

Abstract

In their HEPATOLOGY review, Rockey and Bissell analyze development and applications of noninvasive measures of liver fibrosis,1 stating that “routine imaging modalities — ultrasound, computed tomography, and magnetic resonance imaging — are generally capable of detecting advanced disease from signs of portal hypertension with good sensitivity and specificity. However, they are typically insensitive to mild or moderate fibrosis.”1 We agree this is the case with routine imaging, but would like to highlight emerging protocols that provide discriminating information.2 Imaging biomarkers allow hepatic localisation that is not always available with “direct” serological tests, while assessing multiple hepatic regions to reduce sampling variability, which is a significant problem with liver biopsy. Magnetic resonance (MR) techniques do not rely on signs of portal hypertension to detect fibrosis and cirrhosis. In vivo 31P magnetic resonance spectroscopy (MRS) may be added to imaging protocols on many MR systems and provides metabolic information, which is useful when assessing fibrogenesis.3 The phosphomonoester to phosphodiester ratio (PME/PDE) is an index of cell membrane turnover and correlates closely with disease severity, assessed by histology (Ishak score) in chronic hepatitis C (CHC).3 MRS provided a sensitivity and specificity of 82% and 81% respectively for cirrhosis and showed statistically significant differences between mild hepatitis, moderate hepatitis, and cirrhosis.3 MRI with ultrashort echo time (UTE) pulse sequences has been used to discriminate patients with cirrhosis.4 Chappell and colleagues showed functionally-decompensated liver disease (Child's grade C) was significantly different from functionally-compensated disease (Child's grade A).4 This may be explained, in part, by the fact that protons in water bound to collagen exhibit shorter T2 relaxation times (T2*) than in water bound to other proteins.4 Significant differences in diffusion-weighting (DWI) MRI indices between patients with cirrhosis and normal volunteers have also been reported.5 It is conceivable that structural information and hepatocellular carcinoma surveillance with MRI, metabolic information on ongoing cellular damage and regeneration from 31P MRS, and a measure of fibrotic load from DWI may be gained from a single scanning session. Microbubble-enhanced ultrasound has been used to measure hepatic vein transit times (HVTT), which decrease with severity of liver disease on account of circulatory changes, including arterialization of the hepatic sinusoidal bed, intra- and extrahepatic shunting and the hyperdynamic circulation present with cirrhosis.6 In a cohort of 85 patients with CHC assessed by HVTT, there was 100% sensitivity and 80% specificity for cirrhosis, and 95% sensitivity and 86% specificity for differentiation of mild hepatitis from more severe liver disease.7 The ability of microbubbles to stratify mild and moderate disease suggests that mechanisms other than those directly related to portal hypertension may contribute to effects seen. There has been much excitement over transient elastography in detection of cirrhosis and in prediction of clinical outcomes in patients with cirrhosis,8 but precirrhotic grades are poorly separated.9 Other imaging techniques have potential to provide useful information on fibrosis, often without the need to purchase new hardware. Further research into imaging-related technologies is required to optimize these techniques, to correlate with clinical outcome, and to incorporate them into validated management algorithms. Jeremy Cobbold*, Adrian Lim , Marzena Wylezinska*, Cliona Cunningham*, Mary Crossey*, Howard Thomas*, Nayna Patel*, Jane Cox , Simon Taylor-Robinson*, * Division of Medicine, Imperial College London, London, United Kingdom, Imaging Sciences Department, Imperial College London, London, United Kingdom.