MDCT classification of steatotic liver: a multicentric analysis

PURPOSE: Fatty liver disease is the most common cause of chronic liver disease in the western world. The aim of this study was to analyze steatotic liver characteristics using multidetector row computed tomography (CT) to identify reliable criteria to identify the steatosis and quantify its severity. PATIENTS AND METHODS: Multiphasic CT scans of 51 consecutive adults (36 men, mean age 57.1±9.9 years), who underwent ultrasound-guided liver biopsy, were analyzed. In all patients, the Hounsfield units (HU) value was determined for each hepatic segment and for each contrast phase. Also, the splenic attenuation was quantified and the differences in the liver-spleen (DLS). Steatosis was graded according its severity into four grades. Receiver operating characteristic (ROC) curve analysis was carried out to calculate the sensitivity and specificity for the specific HU threshold. Pearson's ρ correlation was also calculated. A P value of 0.05 was considered statistically significant. RESULTS: We found that 14 individuals (10 men, mean age 56±9.8 years) did not have hepatic steatosis. Only the nonenhanced CT scans showed a statistically significant association with liver steatosis (with the only exception of region-of-interest selected in the Couinaud segment VII, where a P value of 0.0513 was obtained). For grades 1, 2, 3, and 4, we identified 50, 45, 35, and 20HU as thresholds. A statistically significant association was found between steatosis and DLS in the nonenhanced and the arterial phase (P=0.0192 and 0.001, respectively). CONCLUSION: The result of our study indicates that the nonenhanced value of the liver can be used to identify steatosis of the liver and to grade its severity. Moreover, the DLS in the arterial phase represents another reliable parameter.