Background: There are few data relating to the role of fatty score (FS) and modified fatty score (MFS) in ultrasonographic (US) examination on the diagnosis of nonalcoholic steatohepatitis (NASH) in patients undergoing bariatric surgery. Methods: We investigated consecutive patients undergoing laparoscopic bariatric surgery with biopsy-proven nonalcoholic fatty liver disease. Patients with other liver diseases and significant alcohol consumption were excluded. Clinico-demographic and anthropometric data were collected before surgery. Each biopsy specimen was assessed by the same pathologist. Liver US examinations were performed by an independent and experienced sonographer before surgery. FS and MFS, determined by the US scoring system based on degrees of parenchymal echogenicity, far gain attenuation, gallbladder wall blurring, portal vein wall blurring and hepatic vein blurring, were used to assess the severity of fatty liver. US findings were correlated with histologic results. Results: Totally 101 patients were enrolled. The mean BMI of the patients was 44.6 ± 5.4 kg/m2. 29 patients (29%) were categorized with simple steatosis and 72 (71%) with NASH. FS and MFS were significantly correlated with the histological steatosis, fibrosis and the presence of NASH (P<0.001). A receiver operating characteristic curve identified the MFS of 2 as the best cut-off point for the prediction of NASH, yielding measures of sensitivity, specificity, positive predictive value, and accuracy for 72%, 86%, 93% and 76%, respectively. The positive likelihood ratio of 5.24 for MFS approximately doubled the post-test probability of NASH from 30% to 70%. Conclusion: FS and MFS on US examination exhibit acceptable sensitivity and high specificity for the detection of the presence of NASH in morbidly obese patients and may aid in the selection of patients for closer follow-up or liver biopsy.
ASJC Scopus subject areas