Section 5 -

Nonalcoholic Fatty Liver Disease Sanjay Kakar University of California San Francisco, CA

NAFLD is a common liver disease and encompasses a wide spectrum of pathologic changes ranging from simple steatosis at one end to steatohepatitis, fibrosis and cirrhosis at the other. In patients with risk factors for NAFLD like obesity and type 2 diabetes, it is estimated that 50-70% will show steatosis, 20-30% will progress to steatohepatitis and 2-3% will eventually develop cirrhosis [1]. The occurrence of cirrhosis, overall mortality and liver-related mortality in NAFLD patients has been reported as 5-15%, 12-36% and 2-7% respectively, and is significantly higher than the general population [2, 3]. Role of Liver Biopsy: Studies increasingly support a "two-hit" hypothesis, where simple steatosis is the first step and has a very small (<5%) risk of progression to advanced fibrosis. On the other hand, patients with steatohepatitis, which constitutes the "second hit", progress to cirrhosis in around 25% of cases [2]. Hence it is critical to separate steatohepatitis from steatosis. There are no clinical or radiological tests that can reliably diagnose steatohepatitis. Use of liver function tests alone can lead to overdiagnosis of NASH in a significant number of cases. Serum transaminases often correlate poorly with histological activity. Radiological techniques can demonstrate steatosis, but cannot detect steatohepatitis [4]. In several epidemiological and clinical studies, unexplained increase in ALT and AST has been assumed to have fatty liver disease. In a study of 354 patients with clinically unexplained liver test abnormalities, Skelly et al showed that 66% had fatty liver disease on liver biopsy [5]. The remaining cases were either normal, had a different clinical diagnosis or remained cryptogenic. Hence the diagnosis of steatohepatitis can be reliably made only on biopsy. The role of liver biopsy in fatty liver disease is to establish the diagnosis, distinguish steatosis from steatohepatitis, determine the extent of activity and fibrosis, and evaluate for other coexisting disease processes The rest of the discussion in this category is devoted to pathologic features [6, 7, 8, 9, 10] and diagnostic problems encountered in nonalcoholic steatohepatitis (NASH). Pathologic Features: (1) Steatosis: The fat occurs in macrovesicular or mixed patterns; predominantly microvesicular steatosis is not characteristic of steatohepatitis. An upper limit of 5% is widely used as a cutoff for defining significant steatosis [11]. Minor degrees of steatosis have been considered 'normal', but an opposing view holds any degree of steatosis to be pathological. Steatosis is present in nearly 100% of cases, but may not be necessary for diagnosis in certain situations like (a) steatohepatitic cirrhosis; fat often disappears after onset of significant fibrosis or cirrhosis (b) steatohepatitis associated with drugs like amiodarone may show minimal steatosis (2) Inflammation: Mild mixed acinar inflammation is considered the hallmark of steatohepatitis. Neutrophils, usually in small numbers, can surround the ballooned hepatocytes (satellitosis). Mild acinar and portal mononuclear inflammation including lymphocytes and histiocytes is also common. Occasional lipogranulomas, eosinophils, pigmented Kupffer cells and microgranulumoas may be present. (3) Hepatocellular injury (in addition to steatosis) is a prerequisite for the diagnosis of steatohepatitis. Evidence of hepatocellular injury most often occurs in the form of hepatocellular ballooning or pericellular fibrosis. (a) Hepatocellular ballooning, characterized by swelling and rarefaction of cytoplasm of zone 3 hepatocytes is a characteristic finding of steatohepatitis. This can be accompanied by cytoplasmic Mallory hyaline in zone 3 hepatocytes, but is often poorly formed. Scattered acidophil bodies are common but are not a conspicuous finding. (b) Pericellular fibrosis. The pericellular, chickenwire or spider-web pattern of fibrosis results from collagen deposition in the space of Disse and often starts in the zone 3 of the acinus. Portal-based fibrosis can accompany pericellular fibrosis. Fibrosis in the periportal region has been reported in NASH associated with diabetes and in children. Table 1. Histologic features of nonalcoholic steatohepatitis. Adapted from AASLD conference summary on NASH, 2002 Essential features Steatosis, predominantly macrovesicular, concentrated in zone 3 Mild mixed acinar inflammation Hepatocelluar injury in the form of Hepatocellular ballooning, often most prominent in zone 3, and/or Pericellular fibrosis Often present, not essential for diagnosis Glycogenated nuclei in zone 1 Lipogranulomas in the acinus or portal tracts Occasional acidophil bodies May be present, not essential for diagnosis Mallory hyaline in zone 3, typically inconspicuous Mild iron deposits in hepatocytes or sinusoidal cells Giant mitochondria Unusual features Predominantly microvesicular steatosis Prominent portal and/or acinar inflammation, numerous plasma cells Prominent bile ductular proliferation, cholestasis Perivenular fibrosis, sclerosing hyaline sclerosis Marked lobular inflammation Minimum criteria: The minimum findings necessary for the diagnosis of SH include: (1) Macrovesicular steatosis, (not considered necessary for diagnosis in certain situations described above). (2) Evidence of hepatocellular injury in the form of hepatocellular ballooning (with or without acidophil bodies or Mallory hyaline) or pericellular fibrosis. When present along with steatosis, pericellular fibrosis provides evidence of prior steatohepatitic injury and is sufficient for the diagnosis even in the absence of hepatocellular ballooning or inflammation. (3) Acinar inflammation (often mild and mixed) is required for the diagnosis by some pathologists and is mentioned as a necessary criterion in the AASLD summary conference on NASH [1]. However, many hepatopathologists believe that the first two criteria are enough to make the diagnosis. Steatosis with scattered inflammatory cells in the absence of hepatocellular injury (hepatocyte ballooning or pericellular fibrosis) have low rate of progression to fibrosis, and is not considered enough for the diagnosis of steatohepatitis. Histological Course of NASH There are a limited number of studies that have examined histological features in NASH patients in sequential biopsies. There is a wide variability in progression of disease among different patients. In most cases, disease progression is slow and increase in stage of fibrosis is seen in 30-40% of patients over 3-5 years [12, 13]. Rapid progression to advanced fibrosis and cirrhosis can be seen in a minority of patients. Diabetes, high body mass index and high AST levels have been shown to be predictive of progression [12, 13]. Fibrosis remains stable in 35-50% of patients or may regress in 15-30%. Reduction in fibrosis stage may be due to sampling error or may represent true regression of fibrosis. The severity of steatosis, hepatocellular ballooning and inflammation generally regresses over time independent of progression of fibrosis [12]. Hence typical features of steatohepatitis may be lost by the time cirrhosis develops (see later). Diagnostic Considerations in Steatohepatitis (1) Alcoholic vs. Nonalcoholic In majority of instances, it is not histologically possible to distinguish alcoholic vs. nonalcoholic SH. In general, NASH is characterized by more prominent steatosis and less severe signs of hepatocellular injury like ballooning and inflammation. Glycogenated nuclei in zone 1 have been described in 70-80% of NASH compared to ~10% in alcoholic SH [14]. Abundant Mallory hyaline and prominent neutrophilic infiltrate strongly favor alcoholic etiology for SH [15]. Several features like obliteration of central veins, canalicular cholestasis and extensive microvesicular steatosis can occur in alcoholic SH, but have not been observed in NASH. (2) Risk Factors A vast majority of NASH cases can be attributed to obesity, type 2 diabetes and hyperlipidemia. Type 1 diabetes is not a significant risk factor unless the blood glucose is poorly controlled. The major risk factors for NASH are associated with insulin resistance and are components of the metabolic syndrome. The diagnosis of metabolic syndrome is based on the evaluation of 5 components: obesity, hypertension, hypertryglyceridemia, low HDL and hyperglycemia (presence of 3 or more is generally accepted as metabolic syndrome). Metabolic syndrome is present in 22% of the population and carries increased risk of cardiovascular morbidity and mortality as well as increased risk of progressive fibrosis in NASH. Among the NASH patients, nearly 90% have metabolic syndrome [16, 17]. Drug-related SH is uncommon and only a few drugs are associated with this pattern of liver injury [18]. Amiodarone is the best known example and is a very effective antiarrhythmic drug. Majority of patients have asymptomatic abnormalities of liver function tests, but 1-3% of patients with long-term amiodarone use develop more serious toxicity with typical histological features of NASH [19]. The biopsy often shows prominent Mallory hyaline (occasionally in zone 1} and neutrophilic satellitosis, while steatosis is less conspicuous. Reversal of liver injury often occurs with discontinuation of the drug, but may be delayed by weeks or months. In addition, amiodarone is also associated with a different type of lipid accumulation called phospholipidosis characterized by accumulation of drug in the lysosomes. On liver biopsy, these cells appear as 'foamy hepatocytes'. Similar changes can occur in Kupffer cells that appear as foamy cells in the sinusoids. Perhexiline maleate (Pexid), an antianginal drug, and diethylaminoethoxyhexestrol (Coralgil), a vasodilator, were extensively used in Europe and Japan respectively. Both drugs can cause NASH and phospholipidoses similar to amiodarone. Table 2. Risk factors for NASH Metabolic diseases (acquired) Obesity Diabetes, type 2 Hypertriglyceridemia Rapid weight loss Malnutrition Metabolic diseases (genetic) Wilson disease Tyrosinemia Abetalipoproteinemia Other Lipodystrophy Jejunoileal bypass Drugs Definite association Amiodarone Perhexiline maleate Questionable etiologic association; may exacerbate NASH Tamoxifen Steroids Estrogens Diethylstibestrol Methotrexate Calcium channel blockers (like nifedipine, verapamil and diltiazem) The evidence linking other drugs to NASH is tenuous. Drugs like tamoxifen, steroids, estrogen and diethylstilbesterol often lead to hepatic steatosis, but steatohepatitis is rare. These drugs may exacerbate or precipitate steatohepatitis in patients with risk factors for NASH rather than play an etiological role. The evidence linking NASH and calcium channel blockers like nifedipine is also anecdotal. Risk factors for NASH were present in many reported cases creating uncertainty about the association of these drugs with steatohepatitic injury. Methotrexate is well known to cause steatosis, portal-based fibrosis and eventually cirrhosis in some patients with long term use. However, typical features of NASH like ballooned hepatocytes and pericellular fibrosis are not seen with methotrexate use. It has been suggested that methotrexate can exacerbate SH. Three metabolic disorders that can mimic NASH warrant special mention. Diabetic patients with poor glycemic control can develop glycogenic hepatopathy, characterized by swollen hepatocytes and glycogenated nuclei similar to NASH [20]. However, it occurs more commonly in type 1 diabetes as opposed to type 2 diabetes in NASH. Fat, Mallory hyaline, significant inflammation and pericellular fibrosis are usually absent or inconspicuous in glycogen hepatopathy. The term diabetic hepatosclerosis has been suggested to describe dense perisinusoidal fibrosis and basement membrane deposition that occur in patients with long standing insulin-dependent diabetes [21]. It is associated with severe microvascular disease in other organs and may represent a form of hepatic microangiopathy. In addition to sinusoidal fibrosis, perivenular fibrosis and hyaline thickening of small hepatic artery branches may be present. Features typical of NASH such as steatosis or hepatocellular ballooning are not seen. These patients often have elevated alkaline phosphatase, a finding unusual for NASH [21]. Wilson disease is more often associated with steatosis, but histologic features of steatohepatitis with Mallory hyaline can occur. The non-zonal distribution of fat and Mallory hyaline in Wilson disease, 'chronic hepatitis-like' pattern of inflammation and portal-based fibrosis are helpful in the differential diagnosis. (3) Pediatric NASH NAFLD is reportedly increasing in pediatric patients and affects more than 1-2% of adolescents. A strong association with male gender, Hispanic race and insulin resistance has been reported. Majority of patients are obese and metabolic abnormalities like hypertryglyceridemia and hypercholesterolemia are common [22, 23]. Some of the diagnostic features for adult NASH are often not seen in children (7,8,9,10). Steatosis is often severe but may lack zone 3 predominance. Ballooned hepatocytes, Mallory hyaline, acinar inflammation and neutrophilic infiltrate may be mild or absent, while portal-based chronic inflammation may be prominent. Portal-based fibrosis in the absence of pericellular fibrosis can occur. Advanced fibrosis can occur in pediatric NASH; cirrhosis is less common than in adults but has been observed. A recent study has proposed that pediatric steatohepatitis can be classified as types 1 and 2 based on histological features [24]. Type 1 NASH shows the classic histological features with zone 3 predominance of steatosis and fibrosis while type 2 NASH shows variant features like portal-based inflammation and fibrosis. Half of the pediatric patients with NASH in this series showed type 2 features. Type 2 NASH was more frequent in boys and in children of Asian, Native American and Hispanic ethnicity. Children with type 2 NASH have a high incidence of severe obesity. Review of NASH studies in adults suggests that type 2 NASH may also occur in adults, especially in patients with extreme obesity. (4) NASH in Conjunction with Other Liver Diseases In view of the increasing incidence of fatty liver disease, it is evident that NASH can coexist with other liver diseases. The features of NASH are sufficiently distinctive to allow diagnosis in the presence of another liver disease. Histologic features of steatohepatitis are present in around 5% of patients with another liver disease even in the absence of significant alcohol consumption [25]. This overlap is often encountered in hepatitis C, perhaps because both are common diseases. Macrovesiclular steatosis occurs in approximately 40% of patients with chronic hepatitis C, but is generally mild (except genotype 3a) and is either non-zonal or predominiantly in zone 1. Ballooned hepatocytes in zone 3 and pericellular fibrosis are not seen hepatitis C. On the other hand, lymphoid aggregates, periportal activity, prominent acidophil bodies and portal-based fibrosis are not typically features of NASH. Coexisting NASH may accelerate the progression of fibrosis in patients with hepatitis C [26]. (5) Unusual Histologic Features (a) Extensive microvesicular steatosis. Some degree of microvesicular steatosis is often observed in NASH. However, exclusive or predominant microvesicular steatosis should prompt consideration of other etiologies like alcohol foamy liver degeneration, Reyes's syndrome (in children), acute fatty liver of pregnancy and genetic diseases like carnitine deficiency. Adverse effect of drugs/toxins like cocaine, tetracycline, valproic acid and zidovudine can also lead to predominantly microvesicular steatosis. (b) Prominent portal inflammation. Mild portal inflammation can be observed in NASH. However, portal inflammation exceeding acinar changes is unusual and should trigger the consideration of other hepatitic processes like viral and autoimmune hepatitis. The presence of lymphoid aggregates (as in hepatitis C) and numerous plasma cells (as in autoimmune hepatitis) are unusual in NASH. Serum autoantibodies have been detected in 23-36% of NASH patients and can lead to clinical confusion with autoimmune hepatitis. Antinuclear antibodies (ANA) and anti-smooth muscle antibodies (SMA) have been reported in 20% and 3-5% of cases respectively [27, 28]. Both antibodies are present in rare cases [28]. These patients generally show typical histological features of NASH and do not benefit from steroid therapy. (c) Portal-based fibrosis. Mild portal fibrosis has been commonly described in NASH. However, portal-based fibrosis in the absence of or out of proportion to pericellular fibrosis is not a typical feature and should invoke another etiology as the cause of fibrosis Exceptions include NASH in children, which is often associated with portal-based fibrosis in the absence of pericellular fibrosis. In some patients with type 2 diabetes, the pericellular fibrosis may first appear in the periportal region (zone 1). (d) Bile ductular reaction. Mild degree of ductular reaction occurs commonly as a nonspecific finding. However, prominent bile ductular reaction, bile duct injury and duct loss are not part of the spectrum of NASH. (e) Increased hepatic iron. A mild increase of iron in hepatocytes (generally nonzonal) or sinusoidal cells has been described in 15-55% of biopsies showing typical NASH. Serum ferritin is an acute phase reactant and can be elevated in these patients raising the clinical suspicion for genetic hemochromatosis. However, genetic hemochromatosis is characterized by a more intense iron deposition that predominantly occurs in hepatocytes and is more pronounced in the periportal region; the typical histological features of NASH are not observed in hemochromatosis. The role of iron in progression of fibrosis is uncertain, but the present evidence suggests that it is unlikely to be significant [29, 30]. (6) Grading and Staging Although several different grading and staging schemes have been proposed, modified Brunt classification is the most widely used [1, 31]. The combination of steatosis, hepatocellular ballooning and inflammation in the grading scheme can create problems since these three features do not change in conjunction. It is not uncommon for NASH biopsies to show marked (>66%) steatosis but minimal hepatocellular ballooning and inflammation. It has been advocated that steatosis can be graded separately (mild, moderate, severe) and features of hepatocellular injury (like ballooned hepatocytes and inflammation) be used for the grading of steatohepatitis. Since pediatric NASH has features different from adult NASH and often shows portal-based inflammation and fibrosis, different grading and staging strategies have to be devised. Table 3. Grading of steatohepatitis Grade 1 (mild) Steatosis: up to 66%, predominantly macrovesicular Ballooning: occasional, zone 3 Lobular inflammation: Scattered, mixed Portal inflammation: none or mild Grade 2 (moderate) Steatosis: any degree, macrovesicular or mixed Ballooning: present, zone 3 Lobular inflammation: Neutrophils often present, +/- chronic inflammation Portal inflammation: none, mild or moderate Grade 3 (severe) Steatosis: usually >66%, often mixed Ballooning: marked, predominantly in zone 3 Lobular inflammation: Scattered mixed inflammation; neutrophils often prominent in zone 3 Portal inflammation: mild or moderate; not predominant or marked Staging of steatohepatitis Stage 1 Zone 3 perivenular or pericellular fibrosis, focal or extensive Stage 2 As in stage 1 plus portal fibrosis, focal or extensive Stage 3 Bridging fibrosis, focal or extensive Stage 4 Cirrhosis (+/- residual pericellular fibrosis) (7) NASH-Related Cirrhosis Cirrhosis in NAFLD may retain the characteristic histological features of NASH, but the phenomenon of 'burnt-out' cirrhosis is well documented. Biopsy-proven cases of NASH may lose the diagnostic features with progression to cirrhosis [32]. Based on correlation with clinical risk factors for NASH, it is believed that a large proportion of otherwise cryptogenic cirrhosis may be related to NASH [33]. A high incidence of obesity and type 2 diabetes has been found in cryptogenic cirrhosis. It has been speculated that remodeling of hepatic architecture in cirrhosis leads to shunting of gut-derived lipoproteins from the portal to systemic circulation, diverting blood-borne lipids away from the liver. Patients with steatohepatitic cirrhosis are at risk for hepatocellular carcinoma [34, 35]. The incidence of HCC in cryptogenic cirrhosis ranges from 18-27%. The precise risk in NASH-related cirrhosis is not known, but is speculated to be similar to hepatitis C –related cirrhosis. References: Neuschwander-Tetri BA, Caldwell SH. Nonalcoholic steatohepatitis: summary of an AASLD Single Topic Conference. Hepatology. 2003; 37(5): 1202-19. 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(6): 1413-9. Adams LA, Lymp JF, St Sauver J, et al. The natural history of nonalcoholic fatty liver disease: a population-based cohort study. Gastroenterology. 2005; 129(1): 113-21. Saadeh S, Younossi ZM, Remer EM, Gramlich T, et al. The utility of radiological imaging in nonalcoholic fatty liver disease. Gastroenterology. 2002; 123(3): 745-50. Skelly MM, James PD, Ryder SD. Findings on liver biopsy to investigate abnormal liver function tests in the absence of diagnostic serology. J Hepatol. 2001; 35(2): 195-9. Ludwig J, Viggiano TR, McGill DB, Oh BJ. Nonalcoholic steatohepatitis: Mayo Clinic experiences with a hitherto unnamed disease. Mayo Clin Proc. 1980; 55(7): 434-8. Brunt EM. Nonalcoholic steatohepatitis: definition and pathology. Semin Liver Dis. 2001; 21(1): 3-16. Brunt EM, Tiniakos DG. Pathology of steatohepatitis. Best Pract Res Clin Gastroenterol. 2002; 16(5): 691-707. Hubscher SG. Role of liver biopsy in the assessment of non-alcoholic fatty liver disease. Eur J Gastroenterol Hepatol. 2004; 16(11): 1107-15. Brunt EM. Nonalcoholic steatohepatitis. Semin Liver Dis. 2004; 24(1): 3-20. Burt AD, Mutton A, Day CP. Diagnosis and interpretation of steatosis and steatohepatitis. Semin Diagn Pathol. 1998; 15(4): 246-58. Harrison SA, Torgerson S, Hayashi PH. The natural history of nonalcoholic fatty liver disease: a clinical histopathological study. Am J Gastroenterol. 2003; 98(9): 2042-7. Adams LA, Sanderson S, Lindor KD, Angulo P. The histological course of nonalcoholic fatty liver disease: a longitudinal study of 103 patients with sequential liver biopsies. J Hepatol. 2005; 42(1): 132-8. Cortez-Pinto H, Baptista A, Camilo ME, De Moura MC. Nonalcoholic steatohepatitis--a long-term follow-up study: comparison with alcoholic hepatitis in ambulatory and hospitalized patients. Dig Dis Sci. 2003; 48(10): 1909-13. Ludwig J, McGill DB, Lindor KD. Review: nonalcoholic steatohepatitis. J Gastroenterol Hepatol. 1997; 12(5): 398-403. Chitturi S, Abeygunasekera S, Farrell GC, et al. NASH and insulin resistance: Insulin hypersecretion and specific association with the insulin resistance syndrome. Hepatology. 2002; 35(2): 373-9. Marchesini G, Bugianesi E, Forlani G, et al. Nonalcoholic fatty liver, steatohepatitis, and the metabolic syndrome. Hepatology. 2003; 37(4): 917-23. Erratum in: Hepatology. 2003; 38(2): 536. Stravitz RT, Sanyal AJ. Drug-induced steatohepatitis. Clin Liver Dis. 2003; 7(2): 435-51. Lewis JH, Ranard RC, Caruso A, et al. Amiodarone hepatotoxicity: prevalence and clinicopathologic correlations among 104 patients. Hepatology. 1989; 9(5): 679-85. Torbenson M, Chen YY, Brunt E, Glycogenic hepatopathy: an underrecognized hepatic complication of diabetes mellitus. Am J Surg Pathol. 2006; 30(4): 508-13. Harrison SA, Brunt EM, Goodman ZD, Di Bisceglie AM. Diabetic hepatosclerosis: diabetic microangiopathy of the liver. Arch Pathol Lab Med. 2006; 130(1): 27-32. Baldridge AD, Perez-Atayde AR, Graeme-Cook F, Higgins L, Lavine JE. Idiopathic steatohepatitis in childhood: a multicenter retrospective study. J Pediatr. 1995; 127(5): 700-4. Rashid M, Roberts EA. Nonalcoholic steatohepatitis in children. J Pediatr Gastroenterol Nutr. 2000; 30(1): 48-53. Schwimmer JB, Behling C, Newbury R, Deutsch R, Nievergelt C, Schork NJ, Lavine JE. Histopathology of pediatric nonalcoholic fatty liver disease. Hepatology. 2005; 42(3): 641-9. Brunt EM, Ramrakhiani S, Cordes BG, et al. Concurrence of histologic features of steatohepatitis with other forms of chronic liver disease. Mod Pathol. 2003; 16(1): 49-56. Castera L, Hezode C, Roudot-Thoraval F, Bastie A, Zafrani E-S, Pawlotsky J-M, et al. Worsening of steatosis is an independent factor of fibrosis progression in untreated patients with chronic hepatitis C and paired liver biopsies. Gut 2003; 52: 288-92. Loria P, Lonardo A, Leonardi F, et al. Non-organ-specific autoantibodies in nonalcoholic fatty liver disease: prevalence and correlates. Dig Dis Sci. 2003; 48(11): 2173-81. Adams LA, Lindor KD, Angulo P. The prevalence of autoantibodies and autoimmune hepatitis in patients with nonalcoholic fatty liver disease. Am J Gastroenterol. 2004; 99(7): 1316-20. George DK, Powell LW, Losowsky MS. The haemochromatosis gene: a co-factor for chronic liver diseases? J Gastroenterol Hepatol. 1999; 14(8):745-9. Chitturi S, Weltman M, Farrell GC, et al. HFE mutations, hepatic iron, and fibrosis: ethnic-specific association of NASH with C282Y but not with fibrotic severity. Hepatology. 2002; 36(1): 142-9. Brunt EM, Janney CG, Di Bisceglie AM, Neuschwander-Tetri BA, Bacon BR. Nonalcoholic steatohepatitis: a proposal for grading and staging the histological lesions. Am J Gastroenterol. 1999; 94(9): 2467-74. Abdelmalek M, Ludwig J, Lindor KD. Two cases from the spectrum of nonalcoholic steatohepatitis. J Clin Gastroenterol. 1995; 20(2): 127-30. 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(3): 664-9. Marrero JA, Fontana RJ, Su GL, Conjeevaram HS, Emick DM, Lok AS. NAFLD may be a common underlying liver disease in patients with hepatocellular carcinoma in the United States. Hepatology. 2002; 36(6):1349-54. Caldwell SH, Crespo DM, Kang HS, Al-Osaimi AM. Obesity and hepatocellular carcinoma. Gastroenterology. 2004; 127(5 Suppl 1): S97-103.