Factors affecting the diagnostic accuracy of ultrasonography in assessing the severity of hepatic steatosis

Transcription

1 Journal of the Formosan Medical Association (2014) 113, 249e254 Available online at journal homepage: ORIGINAL ARTICLE Factors affecting the diagnostic accuracy of ultrasonography in assessing the severity of hepatic steatosis Chia-Chi Wang a, Tsung-Cheng Hsieh b, Tai-Chung Tseng a, Pin-Chao Wang a, Ching-Sheng Hsu a, Hans Hsienhong Lin a, Li-Yu Wang c, Jia-Horng Kao d, * a Department of Hepatology, Buddhist Tzu Chi General Hospital, Taipei Branch and School of Medicine, Tzu Chi University, Hualien, Taiwan b Institute of Medical Science, Buddhist Tzu Chi University, Taipei, Taiwan c School of Medicine, Taipei Medical University, Taipei, Taiwan d Graduate Institute of Clinical Medicine and Hepatitis Research Center, National Taiwan University College of Medicine and Hospital, Taipei, Taiwan Received 11 March 2012; received in revised form 28 May 2012; accepted 6 July 2012 KEYWORDS diagnostic accuracy; hepatic steatosis; histology; ultrasonography Background/Purpose: Ultrasonography has long been recognized as a useful tool for detecting hepatic steatosis in clinical practice. However, whether it can assess the severity of hepatic steatosis and which factors affect its diagnostic accuracy remain unclear. Methods: A total of 171 patients with various causes of hepatitis undergoing liver biopsies were retrospectively reviewed. The clinical, serologic data and ultrasonographical findings were recorded. Hepatic steatosis was graded as negative, mild, moderate, or severe by ultrasonography and histology. Histology was used as gold standard and the agreement rates were calculated. Results: Our data showed that the agreement rate of ultrasonography was 61.4% in assessing the severity of hepatic steatosis and 74.3% in diagnosing hepatic steatosis compared with histology (crude kappa Z 0.46 vs. 0.46). Using univariate analyses, body mass index and histology activity index score were associated with the agreement in assessing the severity of hepatic steatosis (p Z and 0.035), whereas Ishak fibrosis score had a trend association (p Z 0.066). Multivariate analyses indicated that age, body mass index, and Ishak fibrosis score could affect the agreement (odds ratio Z 0.72, 0.89, and 1.41; 95% confidence interval Z 0.54e0.97, 0.83e0.97, and 1.1e1.8). Conflicts of interest: The authors have no conflicts of interest relevant to this article. * Corresponding author. Graduate Institute of Clinical Medicine, National Taiwan University College of Medicine, 1 Chang-Te Street, Taipei 100, Taiwan. address: kaojh@ntu.edu.tw (J.-H. Kao) /$ - see front matter Copyright ª 2012, Elsevier Taiwan LLC & Formosan Medical Association. All rights reserved.

2 250 C.-C. Wang et al. Conclusion: Ultrasonography could assess the severity of hepatic steatosis with moderate accuracy. Obese patients are difficult ultrasonographically. In addition, age and hepatic fibrosis could affect the performance of ultrasonography in assessing the severity of hepatic steatosis. Copyright ª 2012, Elsevier Taiwan LLC & Formosan Medical Association. All rights reserved. Introduction Hepatic steatosis is commonly observed in daily clinical practice, and the prevalence of hepatic steatosis is increasing in parallel with the pandemics of obesity and type 2 diabetes mellitus. 1 Hepatic steatosis also frequently coexists with chronic hepatitis C, alcoholic liver disease or nonalcoholic fatty liver disease (NAFLD), which has the potential to develop end-stage liver disease including cirrhosis and liver cancer. 2e5 In addition, patients with NAFLD have a higher prevalence of metabolic syndrome and increased risk of diabetes mellitus as well as cardiovascular disease. 6e14 Therefore, hepatic steatosis has been recognized as an emerging global health threat. The risk of metabolic syndrome has been reported to be positively associated with the severity of hepatic steatosis on ultrasonography. 15 In addition, serum alanine aminotransferase (ALT) abnormalities usually occur in patients with more severe fatty liver. 16 Taking these lines of evidence together, accurate assessment of the severity of hepatic steatosis can help practicing physicians to identify earlier the risk of liver disease progression in patients. Although needle biopsy remains the gold standard for the diagnosis of hepatic steatosis, several disadvantages have been discussed, such as possible biopsy danger, different sampling area, or interobserver variability. 17,18 Furthermore, it is invasive and often declined by patients. Therefore, it is important for clinicians to develop reliable and specific techniques for noninvasive detection and quantification of hepatic steatosis. On the basis of increased echogenicity of liver and the presence of contrast as compared to adjacent organs such as kidney, vascular wall, or diaphragm, ultrasonography is the most commonly used modality in diagnosing hepatic steatosis. 19 However, the ability to make an accurate grading or quantification is necessary for longitudinal follow-up of disease progression and assessing the response to treatment. 20 Computed tomography and magnetic resonance imaging have been evaluated for this purpose. 21,22 Because ultrasonography is cheaper and adds no increased exposure to radiation, it is the best candidate to be used in longitudinal monitoring, if the diagnostic accuracy could be confirmed. In this study, we thus retrospectively investigated whether ultrasonography could assess the severity of hepatic steatosis and factors influencing its diagnostic accuracy. Materials and methods Patients and methods A total of 171 patients with various causes of hepatitis receiving liver biopsy between 2007 and 2009 in Tzu Chi General Hospital were included in this study. The causes of chronic liver disease were chronic hepatitis B in 61, chronic hepatitis C in 88, hepatitis B and C coinfection in 9, and other causes in 13 patients including 1 acute hepatitis C infection, 1 drug induced hepatitis, 1 autoimmune hepatitis, and 10 NAFLD. Demographic, clinical, serologic, and biochemical data were obtained from each participant. The ultrasonographical findings in terms of the severity of hepatic steatosis (negative, mild, moderate, and severe) were retrieved from medical records. Hepatic steatosis on histology was categorized as negative (5%), mild (6e33%), moderate (34e66%), or severe (>67%). Agreement was defined as having the same results between ultrasonographical and histological findings. The agreement rates in the presence and severity of hepatic steatosis were separately calculated. According to the presence or absence of agreement, the participants were divided into agreement and non-agreement groups. Variables including age, sex, causes of hepatitis, ultrasound machines, operators, body mass index (BMI), aspartate aminotransferase (AST), ALT, creatinine, platelet count, white blood cell (WBC) count, hemoglobin, histology activity index (HAI) score, and Ishak fibrosis score were included in univariate and multivariate analyses for their associations with the diagnostic accuracy of ultrasonography. Clinical features and biochemical examinations We retrospectively collected information on age, sex, BMI, creatinine, hemoglobin, WBC count, platelet count, and ALT and AST levels. BMI was calculated as weight in kilograms divided by height squared. The biochemical data were measured by an autoanalyzer (Roche Analytics; Roche Professional Diagnostics, Penzberg, Germany). Ultrasound examination of the liver All ultrasonograms were obtained with one of two units: one (LOGIQ-5; GE, Medical System Ltd, Seoul, Korea) with a 4-MHz electronic probe and the other (LOGIQ-7; GE, Medical System Ltd, Seoul, Korea) with a 5-MHz electronic probe. One of the ten hepatologists who were qualified and well experienced in abdominal ultrasonography performed the examinations. The technical parameters, including gain adjustment, use of tissue harmonics, and use of harmonics, were optimized on a case-by-case basis. Routine images were recorded and saved on a PACS system. The severity of hepatic steatosis was recorded as negative, mild, moderate, or severe. If the echogenicity of the liver is the same as that of the renal cortex, it is defined as negative steatosis. A slight increase in liver echogenicity with clear vascular wall and diaphragm defines mild steatosis. In moderate steatosis, visualization of vascular wall and

3 Sonography in grading hepatic steatosis 251 diaphragm was impaired and blurred. Severe steatosis was recognized as marked increase brightness, far-field beam attenuation of the posterior segment of the right lobe of liver, and no visualization of vascular wall and diaphragm. 23,24 Histological evaluations of liver biopsy specimens After we obtained written informed consent for the use of clinical data and residual tissue, each patient without contraindications for liver biopsies received an echo-guided percutaneous liver biopsy from the right hepatic lobe using an 18-gauge biopsy needle (Temno REF T1820; CardinalHealth, Dublin, Ireland). The sampling tissues were fixed with formalin, embedded with paraffin, and stained with hematoxylin and eosin. One experienced pathologist, who was blinded to the clinical status of patients and results of ultrasonography readings, evaluated the samples. Hepatic steatosis on histology was categorized as negative (5%), mild (6e33%), moderate (34e66%), or severe (>67%). The staging of hepatic fibrosis was scored by Ishak fibrosis score, which ranged from F0 to F6. Ethical considerations The study was performed in accordance with the principles of the 1975 Declaration of Helsinki and approved by the Ethical Committee of Buddhist Tzu-Chi General Hospital, which waived the study design-specific informed consent. Statistical analysis The agreement rates were calculated by dividing the number of occasions of complete agreement by the total number of occasions. Weighted kappa statistics were used to determine the degree of agreement after correction for the agreement expected by chance. 25 The kappa statistic was interpreted as follows: less than 0, poor agreement; 0e0.2, slight agreement; 0.2e0.4, fair agreement; 0.4e0.6, moderate agreement; 0.6e0.8, substantial agreement; 0.8e1, almost perfect agreement. Chi-square and Student s t-tests were used to compare whether there was a significant difference in the frequency distributions and the means between agreement and nonagreement groups. Univariate and multivariate logistic regression analyses were used to measure the strength of association between all parameters and agreement in determining the presence of hepatic steatosis or assessing the severity of hepatic steatosis, signified as odds ratio (OR). Statistical analyses were performed using SAS 9.1 (SAS Institute Inc., Cary, NC, USA). A p value of < 0.05 was considered statistically significant. Results Study population and demographic data A total of 171 patients (100 men and 71 women with a mean age of years) were included for analysis. The median interval between performing ultrasonography and liver biopsy was 1 month (range: 0e10 months). Of the total, the severity of hepatic steatosis on histological findings was negative in 111, mild in 38, moderate in 10, and severe in 12. Using Ishak fibrosis score, the severity of fibrosis on histological findings was F0 in 2, F1 in 21, F2 in 21, F3 in 57, F4 in 23, F5 in 24, and F6 in 23, respectively. Diagnostic accuracy of ultrasonography in the presence and severity of hepatic steatosis There were two patients with pathological severe steatosis considered as negative on ultrasonographydthey were chronic hepatitis C patients with severe fibrosis (Ishak fibrosis score 6) and mild inflammation (HAI score of 5 and 6, respectively) on liver histology. The possible reason was that hepatic fibrosis or inflammation could cause echogenic abnormalities of liver on ultrasonography. Compared to histological findings, the agreement rates of ultrasonography were 74.3% (127/171, crude kappa Z 0.46) in assessing the presence of hepatic steatosis and 61.4% (105/ 171, crude kappa Z 0.46) in the severity of hepatic steatosis (Table 1). Because the different cutoff points chosen in four groups for hepatic steatosis may affect the agreement, especially the three patients with severe steatosis in ultrasonography, we combined moderate to severe fatty Table 1 Agreement of the severity of hepatic steatosis between ultrasonographical and histological findings. Results of N Results of histological findings ultrasonographical findings Negative Mild Moderate Severe N (%) N (%) N (%) N (%) Negative (82.3) 14 (14.3) 1 (1.0) 2 (2.0) Mild (57.1) 14 (33.3) 3 (7.1) 1 (2.4) Moderate 25 3 (12.0) 10 (40.0) 5 (20.0) 7 (28.0) Severe 3 0 (0.0) 0 (0.0) 1 (33.3) 2 (66.7) (1) Agreements for the presence of steatosis between ultrasonographical and histological findings. Crude kappa Z 0.46 (0.32e0.59). Adjusted (for hepatologists) kappa Z 0.46 (95% CI, 0.33e0.58; test for equal kappa: c 2 Z 5.97, p Z 0.54). (2) Agreements for the severity of steatosis between ultrasonographical and histological findings. Crude kappa Z 0.46 (0.35e0.57). Adjusted (for hepatologists) kappa Z 0.44 (95% CI, 0.35e0.54; test for equal kappa: c 2 Z 4.13, p Z 0.76).

4 252 C.-C. Wang et al. liver both in ultrasonography and histology to a group of advanced fatty liver. Thus, there were 28 cases with advanced fatty liver in ultrasonography and 22 cases in histology. If the severity of hepatic steatosis was categorized to three groups, the agreement rates of ultrasonography in assessing the severity of hepatic steatosis increased to 66.1% (113/171, crude kappa Z 0.59) (data not shown). Relationship among ultrasonographical steatosis, HAI score, and Ishak fibrosis score In this special population of acute or chronic liver diseases, there was a trend toward lower HAI score or Ishak fibrosis score in histology with increasing ultrasonographic steatosis (Table 2). Factors affecting the diagnostic accuracy of ultrasonography in assessing the severity of hepatic steatosis By using univariate analyses, BMI and HAI score were associated with agreement (p Z and 0.035). In addition, Ishak fibrosis scores had a trend association with agreement (p Z 0.066). Age, sex, ultrasound machine, and the status of viral hepatitis did not affect the agreement (Table 3). Using multivariate analyses after adjustment with age, sex, BMI, WBC, HAI score, and Ishak fibrosis score, we found that Ishak fibrosis score, age, and BMI were associated with agreement between ultrasonography and histology (OR Z 0.72, 0.89 and 1.41; 95% confidence interval Z 0.54e0.97, 0.83e0.97, and 1.1e1.8) (Table 4). Discussion Hepatic steatosis is an emerging clinical problem worldwide and is the most common liver disease in Western countries. 26 Ultrasonography is one of the most popular methods to diagnose hepatic steatosis in daily practice. 27,28 A recent meta-analysis indicated that ultrasonography is a reliable and accurate method in detecting moderate-severe fatty liver comparable to histology. 29 In this study, the diagnostic accuracy of ultrasonography was 74.3% in the detection of hepatic steatosis and 61.4% in assessing the severity of hepatic steatosis compared to histology in patients with different causes of hepatitis. In addition, age, BMI, and Ishak fibrosis score could affect the diagnostic accuracy of ultrasonography in assessing the severity of hepatic steatosis. Because ultrasonography is operator-dependent and interobserver variability dose exists, the accuracy of diagnosing hepatic steatosis varied in different populations or studies. Previous studies found that several factors might affect the ability of ultrasonography in diagnosing hepatic steatosis. For example, morbidly obese patients had the lowest accuracy, and advanced fibrosis could reduce the sensitivity of ultrasonography. 30 To the best of our knowledge, this is the first study to elucidate the ability of ultrasonography in assessing the severity of hepatic steatosis. Age, BMI, and Ishak fibrosis score were identified to be associated with the diagnostic accuracy in assessing the severity of hepatic steatosis. Our data consistently confirmed that a higher BMI could influence the diagnostic ability of ultrasonography. Regarding the relationship between age and diagnostic accuracy of ultrasonography, the possible explanation was that the aging process could change the echo texture of liver and/or kidney, and thus lead to a poor performance of ultrasonography in older populations. A previous study had a similar population composed of 131 patients with chronic liver diseases. 31 The authors found that ultrasonographic diagnosis of fatty liver correctly identified steatosis on biopsy in only 47.8% of the patients, but 66% had significant fibrosis or significant inflammation. They concluded that hepatic fibrosis or inflammation was likely to cause echogenic abnormalities of the liver. Our study also found 57.1% (24/42) of patients with mild steatosis in ultrasonography without pathological steatosis. Taking these lines of evidence, hepatic fibrosis or inflammation could affect the diagnostic accuracy of ultrasonography. Another study with 118 biopsy-proven NAFLD patients found that the sensitivity of ultrasonography was 100% for detecting moderate to severe histological steatosis in patients with mild histological fibrosis. However, it reduced to 77.8% in those with advanced histological fibrosis. 32 Thus, advanced fibrosis could decrease the sensitivity of ultrasonography for detecting moderate to severe histological steatosis. In this study, a higher agreement in advanced fibrosis was inconsistent with previous studies. The advanced fibrosis could cause echogenic abnormalities in ultrasonography. In addition, a higher agreement found in the group without ultrasonographical hepatic steatosis, which had higher Ishak fibrosis score, may partly explain the findings of a higher Table 2 Relationship among ultrasonographical steatosis, histologically inflammatory, and fibrotic score. Ultrasonographical steatosis Histological Findings Negative Mild Moderate Severe p HAI score a Median SD <0.01 b Ishak fibrosis score < a Median SD <0.01 b HAI Z histology activity index; SD Z standard deviation. a Analysis of variance (ANOVA). b Test for trend across ordered groups.

5 Sonography in grading hepatic steatosis 253 Table 3 Comparison of variables between agreement and non-agreement groups by chi-square test and Student s t-test. Severity of steatosis between ultrasonographical and histological findings Non-agreement Agreement p N Z 66 N Z 105 Age (y), mean (SD) 51.0 (12.9) 47.8 (13.2) 0.12 Male sex (n, %) 34 (51.5) 66 (62.9) 0.14 Body height (m), mean (SD) a 1.62 (0.11) 1.64 (0.09) 0.23 Body weight (kg), mean (SD) 69.3 (14.2) 67.0 (14.3) 0.30 BMI (kg/m 2 ), mean (SD) a 26.6 (4.8) 24.6 (4.7) 0.008** HBV positivity (n, %) b 27 (40.9) 43 (41.4) 0.96 HCV positivity (n, %) c 35 (53.9) 62 (59.6) 0.46 AST (IU/L), mean (SD) 80.3 (63.2) 92.6 (80.8) 0.30 ALT (IU/L) mean (SD) (100.4) (99.3) 0.59 Creatinine (mg/dl), mean (SD) 0.84 (0.65) 0.81 (0.38) 0.71 Platelet (K/mL), mean (SD) (63.8) (63.6) 0.52 HAI score, mean (SD) 4.53 (1.95) 5.11 (1.62) 0.035* Ishak fibrosis score, mean (SD) 3.14 (1.52) 3.59 (1.59) þ þ 0.05 < p < 0.1. *0.01 < p < **p < ALT Z alanine aminotransferase; AST Z aspartate aminotransferase; BMI Z body mass index; HAI Z histology activity index; HBV Z hepatitis B virus; HCV Z hepatitis C virus. a Missing data for seven patients. b Missing data for one patient. c Missing data for two patients. agreement in advanced fibrosis. Therefore, further studies are needed to confirm the impact of hepatic fibrosis on the diagnostic accuracy of ultrasonography. This study has several notable features. Although liver biopsy is the gold standard in assessing the severity of hepatic steatosis, interobserver variability dose exists. In this study, one experienced pathologist, who was blinded to the clinical status of the patients and results of ultrasonographical findings, evaluated the samples. Therefore, the possible bias due to interobserver variability could be avoided. Furthermore, the quality of static images could affect the decision of radiologists in the impression of echogenicity. In our institute, ultrasonography was performed by well experienced hepatologists. Therefore, the variability between real-time and static images could be reduced. In addition, our retrospective study could truly Table 4 Multivariate-adjusted OR for the agreement of the severity of steatosis between ultrasonographical and histological findings. Variables included in each model OR (95% CI) Age (every 10-y increment) 0.72* (0.54e0.97) BMI (every 1 kg/m 2 increment) 0.89** (0.83e0.97) Ishak fibrosis score (each point increment) 1.41* (1.10e1.80) Sex, age, BMI, HAI score, and Ishak fibrosis score were regarded as potential determinants. *0.01 < p < **p < BMI Z body mass index; CI Z confidence interval; OR Z odds ratio. reflect the ability of ultrasonography in assessing the severity of hepatic steatosis in real-world clinical settings. To the best of our knowledge, this is the first report to investigate the factors affecting the ability of ultrasonography in assessing the severity of hepatic steatosis after adjustment with clinical parameters. Several limitations should be acknowledged. First, ultrasonographical parameters including bright liver echo, hepatorenal contrast, far-field beam attenuation, and blurred vascular wall are quite subjective. Second, although the ultrasound examination was not performed at the same time with liver biopsy, the mean interval of 2.16 months might affect the result of agreement minimally. Third, the ultrasonographic examinations were not performed by the same hepatologist, but one of ten hepatologists. Interobserver variability could exist among these hepatologists. However, since the agreement was not different among the 10 hepatologists (Table 1), the results would not undermine our major conclusions. In summary, ultrasonography could assess the severity of hepatic steatosis with moderate accuracy compared to histology. Obese patients are difficult to deal with ultrasonographically. In addition, age and hepatic fibrosis could affect the performance of ultrasonography in assessing the severity of hepatic steatosis. Acknowledgments This work was supported by grants from the Buddhist Tzu Chi General Hospital, Taipei Branch, Taiwan (TCRD-TPE-97- C1-2), Liver Disease Prevention & Treatment Research Foundation, the Department of Heath, and the National Science Council, Executive Yuan, Taiwan.

6 254 C.-C. Wang et al. Each author involved in Chia-Chi Wang: study concept and design; Tai-Chung Tseng: acquisition of data; Pin-Chao Wang: acquisition of data; Ching-Sheng Hsu: acquisition of data; Hans Hsienhong Lin: acquisition of data; Li-Yu Wang: analysis and interpretation of data; Tsung-Cheng Hsieh: analysis and interpretation of data; Jia-Horng Kao: critical revision of the manuscript. References 1. Chitturi S, Farrell GC, George J. Non-alcoholic steatohepatitis in the Asia-Pacific region: future shock? J Gastroenterol Hepatol 2004;19:368e Liu CJ, Jeng YM, Chen PJ, Lai MY, Yang HC, Huang WL, et al. Influence of metabolic syndrome, viral genotype and antiviral therapy on superimposed fatty liver disease in chronic hepatitis C. Antivir Ther 2005;10:504e Williams R. The pervading influence of alcoholic liver disease in hepatology. Alcohol Alcohol 2008;43:393e7. 4. Matteoni CA, Younossi ZM, Gramlich T, Boparai N, Liu YC, McCullough AJ. Nonalcoholic fatty liver disease: a spectrum of clinical and pathological severity. Gastroenterology 1999;116: 1413e9. 5. Caldwell SH, Oelsner DH, Iezzoni JC, Hespenheide EE, Battle EH, Driscoll CJ. Cryptogenic cirrhosis: clinical characterization and risk factors for underlying disease. Hepatology 1999;29:664e9. 6. Marchesini G, Brizi M, Bianchi G, Tomassetti S, Bugianesi E, Lenzi M, et al. Nonalcoholic fatty liver disease: a feature of the metabolic syndrome. Diabetes 2001;50:1844e Marchesini G, Bugianesi E, Forlani G, Cerrelli F, Lenzi M, Manini R, et al. Nonalcoholic fatty liver, steatohepatitis and the metabolic syndrome. Hepatology 2003;37:917e Clark JM, Diehl AM. Hepatic steatosis and type 2 diabetes mellitus. Curr Diab Rep 2002;2:210e5. 9. Santoliquido A, Di Campli C, Miele L, Gabrieli ML, Forgione A, Zocco MA, et al. Hepatic steatosis and vascular disease. Eur Rev Med Pharmacol Sci 2005;9:269e Wang CC, Lin SK, Tseng YF, Hsu CS, Tseng TC, Lin HH, et al. Elevation of serum aminotransferase activity increases risk of carotid atherosclerosis in patients with non-alcoholic fatty liver disease. J Gastroenterol Hepatol 2009;24:1411e Targher G, Arcaro G. Non-alcoholic fatty liver disease and increased risk of cardiovascular disease. Atherosclerosis 2007; 191:235e Brea A, Mosquera D, Martin E, Arizti A, Cordero JL, Ros E. Nonalcoholic fatty liver disease is associated with carotid atherosclerosis: a case-control study. Arterioscler Thromb Vasc Biol 2005;25:1045e Targher G, Bertolini L, Padovani R, Rodella S, Zoppini G, Zenari L, et al. Relations between carotid artery wall thickness and liver histology in subjects with nonalcoholic fatty liver disease. Diabetes Care 2006;29:1325e Sookoian S, Pirola CJ. Non-alcoholic fatty liver disease is strongly associated with carotid atherosclerosis: a systemic review. J Hepatol 2008;49:600e Hamaguchi M, Kojima T, Itoh Y, Harano Y, Fujii K, Nakajima T, et al. The severity of ultrasound findings in nonalcoholic fatty liver disease reflects the metabolic syndrome and visceral fat accumulation. Am J Gastroenterol 2007;102:2708e Fishbein MH, Miner M, Mogren C, Chalekson J. The spectrum of fatty liver in obese children and the relationship of serum aminotransferases to severity of steatosis. J Pediatr Gastroenterol Nutr 2003;36:54e Maharaj B, Maharaj RJ, Leary WP, Cooppan RM, Naran AD, Pirie D, et al. Sampling Variabilty and its influence on the diagnostic yield of percutaneous needle biopsy of the liver. Lancet 1986;1:523e Poynard T, Ratziu V, Bedossa P. Appropriateness of liver biopsy. Can J Gastroenterol 2000;14:543e Mathiesen UL, Franzen LE, Aselius H, Resjo M, Jacobsson L, Foberg U, et al. Increased liver echogenicity at ultrasound examination reflects degree of steatosis but not of fibrosis in asymptomatic patients with mild/moderate abnormalities of liver transaminases. Dig Liver Dis 2002;34:516e Mazhar SM, Shiehmorteza M, Sirlin CB. Noninvasive assessment of hepatic steatosis. Clin Gastroenterol Hepatol 2009;7: 135e Lupsor M, Badea R. Imaging diagnosis and quantification of hepatic steatosis: is it an accepted alternative to needle biopsy? Romanian J Gastroenterol 2005;14:419e Fishbein M, Castro F, Cheruku S, Jain S, Webb B, Gleason T, et al. Hepatic MRI for fat quantification: its relationship to fat morphology, diagnosis and ultrasound. J Clin Gastroenterol 2005;39:619e Rumack CM, Wilson SR, Charboneau JW. Diagnostic ultrasound. 2nd ed. St. Louis, MO: Mosby; p. 110e Strauss S, Gavish E, Gottlieb P, Katsnelson L. Interobserver and intraobserver variability in the sonographic assessment of fatty liver. AJR 2007;189:W320e Landis JR, Koch GG. The measurement of observer agreement for categorical data. Biometrics 1977;33:159e Palmentieri B, de Sio I, La Mura V, Msarone M, Vecchione R, Bruno S, et al. The role of bright liver echo pattern on ultrasound B-mode examination in the diagnosis of liver steatosis. Dig Liver Dis 2006;38:485e Celle G, Savarino V, Picciotto A, Magnolia MR, Scalabrini P, Dodero M. Is hepatic ultrasonography a valid alternative tool to liver biopsy? Report on 507 cases studied with both techniques. Dig Dis Soc 1988;33:467e Saadeh S, Younossi ZM, Remer EM, Gramlich T, Ong JP, Hurley M, et al. The utility of radiographic imaging in nonalcoholic fatty liver disease. Gastroenterology 2002;123: 745e Hernaez R, Lazo M, Bonekamp S, Kamel I, Brancati FL, Guallar E, et al. Diagnostic accuracy and reliability of ultrasonography in the detection of fatty liver: a meta-analysis. Hepatology 2011;54:1082e Mottin CC, Moretto M, Padoin AV, Swarowsky AM, Toneto MG, Glock L, et al. The role of ultrasound in the diagnosis of hepatic steatosis in morbidly obese patients. Obes Surg 2004;14: 635e Perez NE, Siddiqui FA, Mutchnick MG, Dhar R, Tobi M, Ullah N, et al. Ultrasound diagnosis of fatty liver in patients with chronic liver disease: a retrospective observational study. J Clin Gastroenterol 2007;41:624e Tobari M, Hashimoto E, Yatsuji S, Torii N, Shiratori K. Imaging of nonalcoholic steatohepatitis: advantages and pitfalls of ultrasonography and computed tomography. Intern Med 2009; 48:739e46.