Which Clinical and Sonographic Parameters May Be Useful to Discriminate NASH from Steatosis? Enrico Maria Zardi, MD,* Ilario De Sio, MD,w Giorgia Ghittoni, MD,z Beniamino Sadun, MD,* Bruno Palmentieri, MD,w Paola Roselli, MD,y Marcello Persico, MD,w and Eugenio Caturelli, MDy Background: The natural history of nonalcoholic steatohepatitis (NASH) includes the passage through steatosis. Goal: To retrospectively evaluate the usefulness of sonographic parameters compared to histological diagnosis when differentiating steatosis from NASH. Study: This retrospective study reviewed records of patients with steatosis from databases of our Departments, selecting only those who had been diagnosed by sonography and liver biopsy [64 males (63.82%); 30 females (36.18%)]. Results: Attenuation of the echo amplitude (P<0.05; odds ratio (OR): 3.43; confidence interval (CI): 1.02-11.57), focal fat sparing (P<0.05; OR: 3; CI: 1.02-11.88) and splenic diameter (P<0.05; OR: 1.66; CI: 1.04-3.26) were independent predictors of NASH. A significantly higher association of attenuation of the echo amplitude, enlarged splenic diameter, and presence of focal fat sparing was observed in NASH patients (P<0.01). Conclusions: It is very difficult to build a predictive system to distinguish NASH from steatosis based on sonographic scores. However, it is already possible to differentiate NASH from steatosis by combining 3 simple sonographic parameters: attenua- tion of the echo amplitude, enlarged splenic diameter, and presence of focal fat sparing. Key Words: liver biopsy, fatty liver, nonalcoholic steatohepatitis, ultrasonography (J Clin Gastroenterol 2011;45:59–63) S teatosis and nonalcoholic steatohepatitis (NASH), classified as nonalcoholic fatty liver diseases (NAFLD) 1,2 are 2 hepatic disorders of fat metabolism in the liver parenchyma. Data from the Third National Health and Nutrition Examination raise the suspicion that 9.1 million individuals in the United States may have NAFLD 3 and similar numbers can be found in Asia. 1 Steatosis, a reversible form of NAFLD, may progress to definitive steatosis when fat with inflammation without ballooning and fibrosis are present in the liver. 1,2,4 NASH, a more severe form of NAFLD, is the principal etiology of cryptogenic cirrhosis in 30% to 70% of cases. NASH affects 28% and 34% of Japanese and British patients with unexplained abnormal liver function tests, respectively, 5,6 but its true prevalence is underestimated. Mechanisms by which steatosis evolve to NASH and NASH evolves to cirrhosis are uncertain. 1,2,4 Diagnosis of steatosis is made when fat is present in more than 5% of hepatocytes, whereas NASH is diagnosed when there is a combination of lesions including an excess of fat, hepatocyte ballooning, lobular inflammation, and perisinusoidal fibrosis. 1,6,7 To better classify the entire spectrum of NAFLD, a system for grading and staging NASH was proposed—the ‘‘Brunt’’ fibrosis score. 8 Sonography is a useful, inexpensive, sure, and noninvasive method to diagnose steatosis and to distinguish between the presence of fatty liver infiltration and other causes of chronic liver disease. However, once other liver diseases have been excluded, the diagnosis of steatosis and NASH must be confirmed by liver biopsy. 9,10 In fact, radiological methods are incapable of determining the presence of steatosis when the liver fatty infiltration is <33% 4 and sonography is not a standard method for diagnosing NASH in patients with mild steatosis. 6 There- fore, we conducted a retrospective study by reviewing, from databases of our Departments in which patients, admitted in the past, were sonographically diagnosed with steatosis, based on their clinical, biochemical, sonographic, and histo- logic parameters; we compared the data of liver histology with biochemical and sonographic parameters to deter- mine whether the latter could be helpful in distinguishing between steatosis and NASH. We aimed at building a synthetic score by using only those sonographic parameters that showed a statistically significant correlation with the histological result. This score was based on the summation of the individual score assigned to each one of these sonographic parameters. Furthermore, we searched for a cutoff number able to differentiate steatosis from NASH. MATERIALS AND METHODS Study Population From the data banks of our Departments, after approval of the Institutional Research Committee, we conducted a retrospective review of all patients (620) who underwent hepatic ultrasonography and liver biopsy over a 3-year period; we selected clinical, sonographic, and Copyright r 2010 by Lippincott Williams & Wilkins Received for publication December 2, 2009; accepted March 3, 2010. From the *Department of Clinical Medicine, ‘‘Campus Bio-Medico’’ University, Rome; wDepartment of Internal Medicine and Gastro- enterology, Second University of Naples; zVI Department of Internal Medicine and Interventional Ultrasonography, Policlinico ‘‘S. Matteo’’ Foundation, IRCCS, Pavia; and yGastroenterology Unit, ‘‘Belcolle’’ Hospital, Viterbo, Italy. Funding Source: none. Conflict of Interest Statement: The other authors do not have any commercial associations that pose a conflict of interest in connection with the submitted manuscript. Reprints: Enrico Maria Zardi, MD, Universita` ‘‘Campus Bio-Medico’’, Via A ` lvaro del Portillo 200-00128 Roma, Italy (e-mail: e.zardi@ unicampus.it). ORIGINAL ARTICLE J Clin Gastroenterol  Volume 45, Number 1, January 2011 www.jcge.com | 59