A Model Based on Computed Tomography Imaging Parameters to Predict Poor Survival in Patients With Liver Cirrhosis

Overview

Cirrhosis, as the end stage of most chronic liver diseases, is an important clinical landmark portending high risk of death. Early identification and accurate prognostic scores is critical issue to improve survival rate. Loss of muscle mass and other body features, which can be determined from CT, have been associated with mortality in cirrhosis or hepatic carcinoma. In this study, we sought to investigate serial changes of CT imaging parameters, such as the skeletal muscle index (SMI), liver volume, adiposity density and so forth, to develop a new prognostic model for long-term motality in patients with liver cirrhosis. The final predictive model was developed under the Cox regression framework with MELD, Child-Pugh score, baseline and serial changes of CT imaging parameters. The discrimination of the new risk score was assessed by the overall C index.

Cirrosis, which is always manifesting as ascites, hepatic encephalopathy, upper gastrointestinal bleeding, severe jaundice and coagulopathy, deteriorates rapidly with a high mortality. Early identification and accurate predicted model may be the key to make duly clinical decision and improve prognosis. When hepatic disorders accur, cirrhosis is often accompanied by varying degrees of malnutrition, which is accurately reflexed by the cross-sectioned areas of the muscles at the third lumbar vertebra level normalized by the square of the height （L3-SMI）. In addition, previous studies have shown that lower density of adiposity and altered liver volume were often associated with adverse clinical events. These changes in body composition can be accurately assessed and measured by using CT scans. However, there is no ideas whether these imaging patameters are reliable indicators to predict the mortality of cirrhosis. In this study, we aimed to evaluate the difference of serial changes of L3-SMI, the density of adiposity and liver volume between 5-year survival and non-survival group and develop a new prognostic model based on these CT imaging data. This study highlighted the clinical significance of serial changes of CT imaging data in predicting the long-term outcome of patients with cirrhosis for the first time.

Conditions