Ramkumar Krishnamurthy, Abstract Co-Author: Nothing to Disclose

Debra Dees RN, Abstract Co-Author: Nothing to Disclose

Medhavi Jogi MD, Abstract Co-Author: Nothing to Disclose

Vivek C. Yagnik MD, Abstract Co-Author: Nothing to Disclose

Raja Muthupillai PhD, Presenter: Nothing to Disclose

Mandeep Bajaj MD, Abstract Co-Author: Nothing to Disclose

To test if MR imaging and MR Spectroscopy can non-invasively quantify hepatosteatosis in patients with Type-II Diabetes undergoing drug treatment.

Patients: 36 patients (21 male; Age: 52.8 ± 9.9 yrs) who underwent baseline MRI and MRS to quantify liver fat were assigned to three different drug treatment arms. In this ongoing study, 21/36 subjects had completed follow-up MRS and MRI in 6-12 months after the initiation of drug treatment. Imaging Protocols: A single-voxel MR spectra was obtained from a 30 x 30 x 30 mm voxel localized at three locations on the liver (right anterior, right posterior, and left lobes) using a Position Resolved Spectroscopy (PRESS) sequence. Liver fat fraction was spectroscopically determined using vendor software. For the MRI, MR liver dual-echo (TE = 2.3, 4.6 ms) images were obtained at α = 20º (Proton Density weighted) and α = 70º (T1-weighted). Two in-phase images (TE = 4.6, 9.2 ms) were obtained at α = 90º for T2* calculation. Pixel-by-pixel estimate of fat fraction was calculated using a custom written MATLABTM code, and fat fraction was calculated[Ref1]. Fat fraction was compared pre and post drug treatment, and between MRS and MRI. 1 Hussain et al., Radiology 2005; 237:1048 - 1055  

i) MRI and MRS both yielded similar liver fat fractions with good correlation. (y = 0.91x + 2.02). The image based method overestimated hepatic fat-fraction by about 1-2% compared to MRS. ii) Reduction in liver fat from baseline to follow-up in the drug treatment were: from 12.1 ± 1.3% to 3.7 ± 0.4% for arm 1, 14.9 ± 1% to 7.9 ± 0.8% for arm 2, and 11.2 ± 1.6%; to 11.6 ± 1.2 for arm 3. iii) There was no statistically significant difference between the treatment responses estimated using MRI and MRS. iv) There was little heterogeneity in the fat-fraction across the liver.  

MRI and MRS can be used to non-invasively quantify hepatosteatosis, and they may play crucial role in quantitatively evaluating treatment response in patients with diabetes and/or non-alcoholic fatty liver disease.  

• Non-alcoholic fatty liver disease is a significant and a growing global problem. A simple non-invasive MR imaging test could provide a quantitative estimate of hepatosteatosis.

Krishnamurthy, R, Dees, D, Jogi, M, Yagnik, V, Muthupillai, R, Bajaj, M, Noninvasive Quantitative Evaluation of Hepatosteatosis in Response to Drug Treatment in Diabetic Population Using Magnetic Resonance Spectroscopy (MRS) and Multi-Echo MR Imaging (MRI).  Radiological Society of North America 2009 Scientific Assembly and Annual Meeting, November 29 - December 4, 2009 ,Chicago IL. http://archive.rsna.org/2009/8016660.html