Correlation between liver histology and novel magnetic resonance imaging in adult patients with non-alcoholic fatty liver disease - MRI accurately quantifies hepatic steatosis in NAFLD.

BACKGROUND: Conventional magnetic resonance imaging (MRI) techniques that measure hepatic steatosis are limited by T1 bias, T(2)* decay and multi-frequency signal-interference effects of protons in fat. Newer MR techniques such as the proton density-fat fraction (PDFF) that correct for these factors have not been specifically compared to liver biopsy in adult patients with non-alcoholic fatty liver disease (NAFLD). AIM: To examine the association between MRI-determined PDFF and histology-determined steatosis grade, and their association with fibrosis.
METHODS: A total of 51 adult patients with biopsy-confirmed NAFLD underwent metabolic-biochemical profiling, MRI-determined PDFF measurement of hepatic steatosis and liver biopsy assessment according to NASH-CRN histological scoring system.
RESULTS: The average MRI-determined PDFF increased significantly with increasing histology-determined steatosis grade: 8.9% at grade-1, 16.3% at grade-2, and 25.0% at grade-3 with P ≤ 0.0001 (correlation: r(2) = 0.56, P < 0.0001). Patients with stage-4 fibrosis, when compared with patients with stage 0-3 fibrosis, had significantly lower hepatic steatosis by both MRI-determined PDFF (7.6% vs. 17.8%, P < 0.005) and histology-determined steatosis grade (1.4 vs. 2.2, P < 0.05). NAFLD patients with grade 1 steatosis were more likely to have characteristics of advanced liver disease including higher average AST:ALT (0.87 vs. 0.60, P < 0.02), GGT (140 vs. 67, P < 0.01), and INR (1.06 vs. 0.99, P < 0.01), higher stage of fibrosis and hepatocellular ballooning.
CONCLUSIONS: MRI-determined proton density-fat fraction correlates with histology-determined steatosis grade in adults with NAFLD. Steatosis is non-linearly related to fibrosis progression. In patients with NAFLD, a low amount of hepatic steatosis on imaging does not necessarily indicate mild disease.