WASHINGTON — Imaging modalities, such as magnetic resonance imaging and magnetic resonance elastography, provided significantly confirmative results of non-alcoholic fatty liver disease and non-alcoholic steatohepatitis in an adult patient cohort, according to study results presented at Emerging Trends in Non-Alcoholic Fatty Liver Disease.
“I think [these tools are] really important, because when we start to look at 100 million Americans with fatty liver, and we know that approximately 75 to 80 million of those are going to grow old with their disease ... how can we quickly adjudicate these patients so that further workup is not unnecessarily performed?” Stephen A. Harrison, MD, FAASLD, of Pinnacle Clinical Research in San Antonio, said. “We can examine them, we can rule them out as having a disease that’s likely to progress, and this is where I think it’s important to really use a noninvasive assessment tool.”
Harrison reported recent updates from an ongoing study at the Brooke Army Medical Center. Researchers developed the study to assess the prevalence and severity of both NAFLD and NASH in San Antonio and to evaluate the performance of several imaging modalities, including FibroScan (Echosens), MRI using LiverMultiscan (Perspectum Diagnostics), MRE, controlled attenuated parameter (CAP) and proton density fat fraction (PDFF).
At the time of the update, the study comprised 673 patients, 461 of whom had a full analysis with biomarkers. Of the 461, 315 had normal results, 120 were diagnosed with NAFLD and 26 were diagnosed with NASH.
According to Harrison, the results also showed that, compared with patients with normal results, patients with NAFLD and NASH had increased BMI, were more often Hispanic, more often had diabetes and hypertension, had increased fast food and non-diet soda intake and exercised less.
Results from the FibroScan showed a mean score of 4.6 for normal patients, 5.35 for patients with NAFLD and 8.35 for patients with NASH. MRI using LiverMultiscan showed a mean score of 1 for normal, 2 for NAFLD and 2.81 for NASH. MRE showed a mean score of 2 for normal, 2.2 for NAFLD and 2.65 for NASH. CAP showed a mean score of 2.68 for normal, 3.1 for NAFLD and 3.63 for NASH. PDFF showed a mean score of 2 for normal, 7.4 for NAFLD and nearly 15 for NASH.
Scatter plot results comparing FibroScan liver stiffness scores to MRE liver stiffness scores showed a correlation coefficient of 0.56. Comparing imaging modalities to liver biopsy in patients with stage 1 fibrosis, receiver operating characteristic (ROC) analysis for FibroScan was 0.73 and for MRE was 0.78. For patients with stage 2 fibrosis, ROC analysis for FibroScan was 0.83 and for MRE was 0.67.
“FibroScan liver stiffness and MRE liver stiffness appear to be better at predicting the stage of fibrosis compared to LiverMultiscan, particularly in the more fibrotic phenotypes,” Harrison said. “When you compare CAP and PDFF, CAP is very good based on the ROC, PDFF is excellent at predicting the grade of steatosis in NAFLD patients. For both imaging modalities, sensitivity negative predictive value remains high regardless of the cut off chosen, whereas specificity positive predictive decreases quickly with increasing liver stiffness measurement cut point that are chosen.” – by Talitha Bennett
Harrison SA. Predicting the degree of liver-biopsy-confirms steatosis and fibrosis using transient elastography and magnetic resonance imaging-based techniques in adult patients with suspected non-alcoholic fatty liver disease. Presented at: Emerging Trends in Non-Alcoholic Fatty Liver Disease; March 18-19, 2017; Washington.
Disclosure: Harrison reports he is an advisor for Merck, Intercept, CLDF, Zafgen, Nimbus Discovery, Fibrogen, Pfizer and Gilead; is a consultant for Medivation, NGM Biopharmaceuticals and Alexion; is a speaker for Gilead, CLDF, AbbVie, Merck and Alexion; and receives grants or research support from Genfit.
Editor's note: This has been updated with clarifications from the presenter.