A Simple Model to Predict Moderate to Severe Steatosis in Obese Liver Donors

Francesco Martini, Rodolphe Anty, Anne-Sophie Schneck, Reza Tavana, Stephanie Patouraux, Jean Gugen-heim, Philippe Gual, Albert Tran and Antonio Iannelli


Background: In the selection of marginal liver grafts most transplantation centers use a cut-off of 30% macrovesicular steatosis to define an acceptable risk of non-function. The aim of this prospective cohort study was to create a non-invasive model to predict >30% hepatic steatosis based on simple clinical and biochemical markers available at the time of potential liver-donor evaluation, to avoid futile and expensive procurement proceedings.
Methods: A total of 857 morbidly obese patients underwent liver biopsy during bariatric surgery in our center. Two groups were identified according to the degree of steatosis, classified as absent to mild (0–30%) and moderate to severe (>30%). Univariate and multivariate analysis were performed to identify parameters associated with steatosis of >30%.
Results: Steatosis of >30% was found in 55% of the study population. Age, alanine aminotransferase (ALT), and waist circumference were found to be independently associated with moderate to severe steatosis on multivariate analysis. By combining these three simple parameters, we developed a model to predict hepatic steatosis >30% with an area-under-the-receiver-operating characteristic (AUROC) of 0.78 (95% CI: 0.75–0.81). The best threshold was 0.06, which offered a sensitivity of 72% and a specificity of 70%.
Conclusions: This model combining three simple biological parameters could be useful in the selection of marginal grafts before activating procurement proceedings.

Published on: April 5, 2017
doi: 10.17756/jocd.2017-004
Citation: Martini F, Anty R, Schneck AS, Tavana R, Patouraux S, et al. 2017. A Simple Model to Predict Moderate to Severe Steatosis in Obese Liver Donors. J Obes Chronic Dis 1(1): 15-20.