Elhamy Rafat Heba BMBCh, Presenter: Nothing to Disclose

Ajinkya Shrinivas Desai MBBS, Abstract Co-Author: Nothing to Disclose

Tanya Wolfson MS, Abstract Co-Author: Nothing to Disclose

Tanya Chavez, Abstract Co-Author: Nothing to Disclose

Kevin Amir Zand MD, Abstract Co-Author: Nothing to Disclose

Jessica Lam BS, Abstract Co-Author: Nothing to Disclose

Jonathan Charles Hooker BS, Abstract Co-Author: Nothing to Disclose

Lisa Clark MPH, PhD, Abstract Co-Author: Nothing to Disclose

Rohit Loomba MD, MSc, Abstract Co-Author: Nothing to Disclose

Anthony Gamst PhD, Abstract Co-Author: Nothing to Disclose

Claude B. Sirlin MD, Abstract Co-Author: Research Grant, Bayer AG Research Grant, General Electric Company Research Grant, Bracco Group Contract, Isis Pharmaceuticals, Inc Contract, Pfizer Inc Speakers Bureau, Bayer AG Consultant, Bayer AG Medical Advisory Board, Bayer AG Consultant, Merck & Co, Inc Medical Advisory Board, General Electric Company Consultant, sanofi-aventis Group Consultant, Synageva BioPharma Corporation Contract, Profil Consultant, Takeda Pharmaceutical Company Limited Contract, Siemens AG

Michael Simca Middleton MD, PhD, Abstract Co-Author: Research Consultant, Siemens AG Research Grant, General Electric Company Stockholder, General Electric Company Research Grant, Bayer AG Research Consultant, Merck & Co, Inc Contract, Gilead Sciences, Inc Contract, sanofi-aventis Group Contract, Isis Pharmaceuticals, Inc

To evaluate MRI proton density fat fraction (PDFF) accuracy in quantifying hepatic steatosis using MRS PDFF as reference in a cohort of 506 adult subjects with known or suspected non-alcoholic fatty liver disease (NAFLD).

This cross-sectional, IRB-approved, HIPAA-compliant study was conducted from 2009 to 2013 in an independent cohort of adults with known or suspected NAFLD. 2D unenhanced MR images were obtained to estimate MRI PDFF using a spoiled gradient-recalled-echo sequence (low flip angle (10°) and repetition time 120-270 ms to minimize T1 dependence). MRI PDFF maps were generated offline from source images on a pixel-by-pixel basis using a custom open-source software plug-in (Osirix Foundation, Geneva, Switzerland; corrects for fat-fat and fat-water interference, and T2* decay). MR spectra were acquired and analyzed by an experienced MR spectroscopist (8 cm3 MRS voxel, right lobe liver away from artifact and vessels, long TR to avoid T1 dependence, 6 echoes to permit T2 correction, AMARES algorithm and jMRUI platform for analysis). Two trained image analysts reviewed the MR images and manually placed circular regions of interest in the liver on three consecutive slices co-localized to MRS voxel location. MRI PDFF accuracy was evaluated by regression analysis, with MRS PDFF as reference. Histological fibrosis grade was collected for the 72 subjects who had liver biopsy, and the effect of fibrosis on MRI-PDFF estimation was determined for those subjects.

506 adult subjects were enrolled in this study. Regression analysis showed close agreement between MRI PDFF and MRS PDFF with intercept = 0.599%, slope = 1.006, and R2 = 0.979 (p < 0.001; see Figure). There was agreement between MRI and MRS PDFF for subjects with either MRI or MRS PDFF less than 10%, with intercept = 0.401, slope = 1.038, and R2 = 0.838 (p < 0.001). Histology-determined fibrosis grade did not show a significant effect on MRI PDFF estimation in the subset of 72/506 subjects who had liver biopsy.

MRI accurately quantifies hepatic PDFF in adults with known or suspected NAFLD. To our knowledge, this is the largest study to date evaluating MRI PDFF accuracy to assess hepatic steatosis.

This study helps support validation of MRI PDFF as a non-invasive biomarker for hepatic steatosis by showing high accuracy in a large cohort of adults with known or suspected NAFLD.

Heba, E, Desai, A, Wolfson, T, Chavez, T, Zand, K, Lam, J, Hooker, J, Clark, L, Loomba, R, Gamst, A, Sirlin, C, Middleton, M, Correlation between MRI- and MRS-estimated Proton Density Fat Fraction (PDFF) in 506 Adult Subjects with Non-alcoholic Fatty Liver Disease (NAFLD).  Radiological Society of North America 2013 Scientific Assembly and Annual Meeting, December 1 - December 6, 2013 ,Chicago IL. http://archive.rsna.org/2013/13020771.html