A Prospective Evaluation of Transient Elastography in Non-Alcoholic Fatty Liver Disease - BMI and Subcutaneous Fat Thickness Do Not Affect Accuracy

Abstract

Purpose: Transient elastography (TE) has emerged as an important non-invasive technique of evaluating hepatic fibrosis. However, recent data suggest that greater hepatic steatosis may downgrade TE values relative to the degree of hepatic fibrosis on biopsy in nonalcoholic fatty liver disease (NAFLD) patients. This prospective study evaluates the relationship between TE values and hepatic steatosis as measured on magnetic resonance imaging (MRI) of the liver. Other variables which may affect TE accuracy in NAFLD were also investigated, including higher patient body mass index (BMI) and greater subcutaneous fat overlying the liver. Methods: All adult patients at a tertiary care centre in London, Canada, who met diagnostic criteria for NAFLD were invited to participate in this study. All patients underwent liver biopsy, TE and had anthropomorphic measurements, liver transaminases and measures of metabolic dysfunction measured. Liver biopsies were read by a single, blinded pathologist and scored according to the validated nonalcoholic steatohepatitis activity score (NAS). The FibroScan 502 system (Echosens) was used to conduct TE measurements for all study patients. Subcutaneous fat and hepatic steatosis measured as hepatic fat fraction was measured using Magnetic Resonance Imaging with Iterative Decomposition of water and fat with Echo Asymmetry and Least-squares estimation (IDEAL-MRI), a computer-based quantification method separating fat and water signals in MRI images allowing accurate, validated and reproducible measurement of subcutaneous fat thickness. Linear regression analysis was used to evaluate the correlation between BMI and TE values; subcutaneous fat thickness and TE values; and degree of fibrosis on liver biopsy and TE values. Results: Linear regression demonstrated a positive correlation between TE values and the degree of fibrosis on liver biopsy (P<0.001). There was no correlation between subcutaneous fat thickness and TE values (P=0.699), between BMI and TE values (P=0.883), between subcutaneous fat thickness and liver fibrosis score (p=0.183), nor between BMI and liver fibrosis score (p=0.057). There was no correlation between hepatic fat fraction and TE values (p=0.152), nor hepatic fat fraction and liver fibrosis score (p=0.0118). Conclusion: This study demonstrates that TE correlates well with the degree of hepatic fibrosis on liver biopsy in NAFLD patients, and suggests that common anthropomorphic features of such patients such as increased hepatic steatosis, increased BMI, and peri-hepatic subcutaneous fat does not adversely affect the reliability of TE. This finding supports the continued use of this important diagnostic modality in a population that is becoming increasingly prevalent.