Hepatic steatosis: quantification by proton density fat fraction with MR imaging versus liver biopsy

Abstract

Purpose: To determine utility of proton density fat fraction (PDFF) measurements for quantifying the liver fat content in patients with nonalcoholic fatty liver disease (NAFLD), and compare these results with liver biopsy findings.

Materials and methods: This retrospective study was approved by the institutional review board with waivers of informed consent. Between June 2010 and April 2011, 86 patients received a diagnosis of NAFLD. Ten patients did not accept liver biopsy and six patients had contraindications for magnetic resonance (MR) imaging. Seventy patients were included in this study. Seventy patients with NAFLD (40 men, 30 women; mean age, 44.7 years; range, 16-69 years) underwent T1-independent volumetric multiecho gradient-echo imaging with T2* correction and spectral fat modeling. Median time interval between MR imaging and liver biopsy was 14.5 days (range, 0-259 days). MR examinations were performed with a 1.5-T MR imaging system. Complex-based PDFF measurements were performed by placing regions of interest in Couinaud system segments V-VI and all liver segments from I to VIII. All liver biopsy specimens were retrieved from archives and evaluated by one pathologist for hepatic steatosis according to criteria from a previous study. Pearson correlation coefficient, receiver operating characteristics, and linear regression analyses were used for statistical analyses.

Results: Mean PDFF calculated with MR imaging was 18.1% ± 9.5 (standard deviation). Close correlation for quantification of hepatic steatosis was observed between PDFF and liver biopsy (r = 0.82). PDFF was effective in discriminating moderate or severe hepatic steatosis from mild or no hepatic steatosis, with area under the curve of 0.95. The correlation between biopsy and PDFF-determined steatosis was less pronounced when fibrosis was present (r = 0.60) than when fibrosis was absent (r = 0.86; P = .02).

Conclusion: PDFF measurement by MR imaging provided a noninvasive, accurate estimation of the presence and grading of hepatic steatosis in patients with NAFLD. Hepatic fibrosis reduced the correlation between biopsy results and PDFF.