• Kazumichi Abe, Atsushi Takahashi, Hiromichi Imaizumi, Manabu Hayashi, Ken Okai, Yukiko Kanno, Naoya Sato, Akira Kenjo, Shigeru Marubashi, and Hiromasa Ohira.
• Department of Gastroenterology, Fukushima Medical University, Fukushima, Japan.
• J. Gastroenterol. Hepatol. 2018 Mar 1; 33 (3): 733-740.

ObjectiveThis study aimed to evaluate the utility of magnetic resonance elastography (MRE) as a non-invasive method for predicting ascites in patients with chronic liver disease (CLD).MethodsA total of 208 CLD patients underwent MRE to measure liver stiffness (LS) at our institution from March 2013 to June 2015. We evaluated the diagnostic performance of MRE for predicting the presence of ascites using receiver-operating characteristic (ROC) curve analysis and compared the performance with that of serum fibrosis markers. Multivariate logistic regression analysis was performed to identify factors associated with the presence of ascites. The cumulative incidence of ascites was examined in patients without ascites at baseline. The pathological stage of liver fibrosis was evaluated in 81 CLD patients using histopathologic diagnosis.ResultsOf the 208 patients, 41 had ascites. The optimal cut-off LS value for the presence of ascites was 6.0 kPa (area under the ROC curve = 0.87). The area under the ROC curve for the presence of ascites was significantly higher for MRE than that for fibrosis markers. Multivariate analysis revealed that LS >6.0 kPa is an independent risk factor for the presence of ascites. The cumulative incidence of ascites was significantly higher among those with LS values >6.0 kPa. There was significantly greater diagnostic accuracy for liver fibrosis stage ≥4 with MRE than that with fibrosis markers.ConclusionsCompared with serum fibrosis markers, MRE has higher diagnostic performance in predicting the presence of ascites. MRE-based LS has the potential to predict the presence of ascites in CLD patients.© 2017 Journal of Gastroenterology and Hepatology Foundation and John Wiley & Sons Australia, Ltd.

31 keywords...

hide…