Multiparametric magnetic resonance imaging/magnetic resonance elastography assesses progression and regression of steatosis, inflammation, and fibrosis in alcohol-associated liver disease

Autor: Ziying Yin, Jenifer Siegelman, Jin Wang, Taofic Mounajjed, Richard L. Ehman, Jiahui Li, Jingbiao Chen, Vijay H. Shah, Jie Chen, Meng Yin, Rosa Martin-Mateos, Kevin J. Glaser, Xin Lu, Christopher T Winkelmann, Hiroaki Yashiro
Rok vydání: 2021
Předmět:
Zdroj: Alcohol Clin Exp Res
ISSN: 1530-0277
Popis: BACKGROUND Magnetic resonance imaging (MRI) and MRI-based elastography (MRE) are the most promising noninvasive techniques in assessing liver diseases. The purpose of this study was to evaluate an advanced multiparametric imaging method for staging disease and assessing treatment response in realistic preclinical alcohol-associated liver disease (ALD). METHODS We utilized four different preclinical mouse models in our study: Model 1-mice were fed a fast-food diet and fructose water for 48 weeks to induce nonalcoholic fatty liver disease; Model 2-mice were fed chronic-binge ethanol (EtOH) for 10 days or 8 weeks to induce liver steatosis/inflammation. Two groups of mice were treated with interleukin-22 at different time points to induce disease regression; Model 3-mice were administered CCl4 for 2 to 4 weeks to establish liver fibrosis followed by 2 or 4 weeks of recovery; and Model 4-mice were administered EtOH plus CCl4 for 12 weeks. Mouse liver imaging biomarkers including proton density fat fraction (PDFF), liver stiffness (LS), loss modulus (LM), and damping ratio (DR) were assessed. Liver and serum samples were obtained for histologic and biochemical analyses. Ordinal logistic regression and generalized linear regression analyses were used to model the severity of steatosis, inflammation, and fibrosis, and to assess the regression of these conditions. RESULTS Multiparametric models with combinations of biomarkers (LS, LM, DR, and PDFF) used noninvasively to predict the histologic severity and regression of steatosis, inflammation, and fibrosis were highly accurate (area under the curve > 0.84 for all). A three-parameter model that incorporates LS, DR, and ALT predicted histologic fibrosis progression (r = 0.84, p
Databáze: OpenAIRE
načítá se...