Assessing non-alcoholic fatty liver disease with multiparametric MRI

Multiparametric magnetic resonance imaging (MRI) shows the potential to potentially become a single examination that can rule out significant liver disease, according to an article in a March issue of Alimentary Pharmacology and Therapeutics. Researchers from the Centre for Liver Research at the University of Birmingham in the United Kingdom investigated the performance and cost of multiparametric MRI with existing biomarkers in the assessment of non-alcoholic fatty liver disease (NAFLD). They determined that multiparametric MRI accurately identified patients with steatosis, stratified patients with non-alcoholic steatohepatitis (NASH), and reliably excluded clinically significant liver disease. The authors stated that based on their findings, liver biopsy could potentially be avoided for 458 out of every 1,000 patients.

NAFLD (simple steatosis and NASH) is a major cause of liver disease throughout the world, and has been estimated to impact 25% of the global population.¹ It is expected to increase with rising levels of obesity. Therefore, risk stratification of patients diagnosed with or suspected of having NAFLD is of critical importance. Simple steatosis is less serious, with minimal or no progression of fibrosis and no increase in liver related mortality. Patients diagnosed with NASH have up to a 20% risk of developing cirrhosis and a greater hepatocellular carcinoma risk.

A variety of tests are being used, including transient elastography and MRI. The British research team conducted a prospective study of 50 patients with NAFLD and six healthy volunteers to determine how LiverMultiScan™ (Perspectum Diagnostics, Oxford, UK), a new quantitative liver MRI technology, performed in terms of utility and comparative effectiveness with patients having routine liver needle biopsy as a standard of care.

The patient cohort consisted of individuals with a histologically confirmed diagnosis of NAFLD without a secondary cause and without a history of excessive consumption of alcohol. These patients underwent a non-targeted liver biopsy at either the Queen Elizabeth Hospital Birmingham or the Royal Infirmary of Edinburgh between February 2014 and September 2015. The patients and the volunteer group also had MRI scans and transient elastography examinations.

The MRI exams were performed on a 3T scanner (Siemens Verio, Siemens Healthcare), with sequences including T1 mapping, T2* mapping, and modified Dixon sequence and proton MR spectroscopy with Stimulated Echo Acquisition Mode sequence. The T1, T* and PDFF-Dixon maps were analyzed using LiverScan software. The software’s proprietary algorithm produces a liver inflammation and fibrosis (LIF) score, generated from steatosis, iron content and T1 map data. Additional analysis is described in detail in the article.

Joint senior investigator Gideon M. Hirschfield, MD, PhD, senior lecturer and honorary consultant in transplant hepatology, reported that the demographic characteristics and laboratory blood tests showed no significant differences between patients with simple steatosis and NASH. Multiparametric MRI did.

cT1 showed a significant difference between simple steatosis and NASH. Multiparametric MRI was able to grade hepatic steatosis with a high degree of accuracy. However, it was not able to predict the severity of histological liver fibrosis. But negative predictive values, which suggest that biopsy might be postponed or avoided entirely, were substantially higher for cT1 (80.0%-83.3%) compared to liver stiffness (39.1%-42.9%) and Enhanced Liver Fibrosis (ELF) (26.9%-57.1%) test findings.

The cost analysis performed using a “cT1 only” risk stratification pathway was estimated to reduce the number of biopsies required of a hypothetical 1000 patients by almost half. The authors predicted that there could be a cost savings of approximately 151,218 British pounds (approximately $210,400), and that there could be an even greater savings with the combination of a multiparametric MRI and transient elastography examinations.

“A single test to reliably exclude NAFLD would be of considerable value in clinical practice,” the authors wrote. “In this study, multiparametric MRI showed a high degree of accuracy for differentiating between healthy volunteers and those with NAFLD.....Using a cT1 cut-off value of 875 ms gave multiparametric MRI a sensitivity of 88.0% with specificity of 100% for the detection of any liver disease. This was superior to all other non-invasive tests.”

REFERENCE

1. Younossi ZM, Koenig AB, Abdelatif D, et al. Global epidemiology of nonalcoholic fatty liver disease - Meta-analytic assessment of prevalence, incidence and outcomes. Hepatology 2016 64;1:73-84.
2. Eddowes PJ, McDonald N, Davies N, et al. Utility and cost evaluation of multiparametric magnetic resonance imaging for the assessment of non-alcoholic fatty liver disease. Ailment Pharmacol Ther. 201847;5:631-644.