Can echocardiography be used in the diagnosis of fatty liver disease?

Abstract

Listen

Translate article

Abstract Background Liver steatosis is a common, usually asymptomatic hepatic disorder caused most often by insulin resistance and hepatotoxic factors. It is currently regarded as a sensitive imaging biomarker of the metabolic syndrome and an independent risk factor of future cardiovascular events. Diagnosis of fatty liver in the cardiologists office may be of potential clinical value, revealing the presence of previously unknown metabolic disorder and influencing the prognostic assessment, as well as treatment of cardiovascular disease. The assessment of liver steatosis during transthoracic echocardiography has not yet been validated. Purpose To investigate the accuracy of echocardiography in the diagnosis of fatty liver disease in severely obese individuals. Methods We conducted a retrospective analysis of liver and right kidney images acquired during preoperative echocardiography (GE, Vivid 9, 1.7/3.3MHz, sector probe, abdominal preset) in 220 severely obese patients (mean age 41.5 years, BMI 43.6 kg/m2, 67.3% of females) admitted for bariatric surgery with liver biopsy. Echocardiographic diagnosis of fatty liver was based on computer-aided (HRI – hepatorenal index) and visual analysis of the liver-kidney echogenicity contrast. Images were randomly divided into training (n=111) and validation (n=109) sets with similar proportion of patients with steatosis in each set. The evaluation was performed by a cardiologists not trained in abdominal ultrasound. Increased brightness of liver tissue relative to adjacent right kidney cortex was classified as steatosis, while similar brightness as lack of steatosis. Sensitivity, specificity, positive and negative predictive values were calculated using histopathologic results as reference. Results The prevalence of liver steatosis in the studied group was 68%. The sensitivity, specificity, positive and negative predictive values of the echocardiographic diagnosis of liver steatosis were as follows: 1) for HRI method: 90%, 81%, 91% and 81%, respectively; 2) for visual method: 88%, 77%, 89% and 75% - by an observer experienced in echocardiography and 96%, 57%, 82%, and 87% by the first unexperienced observer and 87%, 81%, 92% and 72% by the second unexperienced observer, respectively. ROC curve analysis of the HRI method showed area under the curve of 0.879, with the optimal cut-off value of HRI 1.34 for the presence of steatosis. Spearman correlation coefficient between HRI and degree of steatosis was 0.68 (p<0.001). Conclusions Visual or computer aided echocardiographic screening for fatty liver based on comparison of liver-kidney echogenicity contrast seems feasible, even for unexperienced observers. Its implementation into everyday clinical practice may improve prognostic assessment and help to optimize treatment in patients with cardiovascular diseases.