Nonalcoholic fatty liver disease (NAFLD) is considered a major cause of liver-related morbidity and mortality, and the prevalence of the disease is estimated at 20%–30% in Western countries.1–3 NAFLD involves a spectrum of histopathologic changes that range from simple steatosis to steatohepatitis, advanced fibrosis, and cirrhosis.4 Nonalcoholic steatosis (NASH) was initially described by Leevy5 in 1962. NASH was first termed in 1980 by Ludwig and associates6 to describe biopsy findings in patients with steatohepatitis in the absence of significant alcohol consumption. At the American Association for the Study of Liver Diseases Clinical Single Topic Conference on NASH in 2002,1 NASH was subcategorized into “primary” and “secondary” disease. Primary NASH is defined as necroinflammatory steatohepatitis that is associated with the metabolic syndrome, and secondary NASH is defined as steatohepatitis that accompanies other syndromes, such as lipodystrophy, or is caused by drugs, such as amiodarone and tetracycline.1
Nonalcoholic fatty liver disease (NAFLD) is now considered to be the most common liver disease in the United States and involves a spectrum of progressive histopathologic changes. Common risk factors associated with NAFLD include obesity, diabetes, and hyperlipidemia. Although most patients with NAFLD have simple hepatic steatosis, a significant number develop nonalcoholic steatohepatitis , which may progress to fibrosis, cirrhosis, or end-stage liver disease. There is increasing evidence that NAFLD is a common feature in patients with the cardiometabolic syndrome, a constellation of metabolic , cardiovascular, renal, and inflammatory abnormalities in which insulin resistance is thought to play a key role in end-organ pathogenesis. NAFLD is usually diagnosed after abnormal liver chemistry results are found during routine laboratory testing. No therapy has been proven effective for treating NAFLD/nonalcoholic steatohepatitis. Expert opinion emphasizes the importance of exercise, weight loss in obese and overweight individuals, treatment of hyperlipidemia, and glucose control.
NAFLD/NASH and the Metabolic Syndrome
There is increasing evidence that NAFLD/NASH is associated with insulin resistance and abnormal glucose tolerance, along with features of the metabolic syndrome such as obesity, insulin resistance, type 2 diabetes mellitus, dyslipidemia, hypertension, hyperuricemia, and polycystic ovarian syndrome. Consequently, patients with such conditions are also at an increased risk for cardiovascular disease.1,2,7,8 The progression of NAFLD is based on a proposed “two-hit” model. The “first hit,” such as obesity, leads to the development of steatosis, and the “second hit” leads to hepatocyte injury, inflammation, and fibrosis; the best candidates for the second hit are oxidative stress and cytokines, mainly tumor necrosis factor α (TNF-α).9 However, it must be emphasized that the causative and temporal relationships between oxidative stress, insulin resistance, cytokines, and hepatic steatosis remain under investigation and findings may support or challenge the dogma of the two-hit hypothesis. Animal models for NASH have become a great research tool. One such model that simulates the human condition of the metabolic syndrome is leptin-deficient mice, the ob/ob mice, which are genetically obese, insulin-resistant, and develop NASH spontaneously.10 These mice also have elevated TNF-α and, thus, the hepatocytes generate excessive reactive oxygen species, accumulate lipids, and are resistant to insulin. Inhibition of TNF-α activity has shown improvement of NASH in ob/ob mice.
NAFLD/NASH and the CMS
The cardiometabolic syndrome (CMS) is a constellation of metabolic, cardiovascular, renal, and inflammatory abnormalities in which insulin resistance is thought to play a key role in end-organ pathogenesis.11,12 The link between insulin resistance and NASH has been well established. NASH and severe hepatic insulin resistance have been demonstrated in the ob/ob mice model and have been reversed with adiponectin and leptin treatment. It has been postulated that hepatic triglyceride accumulation is a causative factor in hepatic insulin resistance.13 Insulin instigates a cascade of intracellular signaling by the activation of at least nine postreceptor tyrosine kinase-mediated pathways, and a defect in one or more of these pathways is responsible for insulin resistance and a resultant hyperinsulinemic state.12,14 The down-regulation of insulin receptor substrate 1 plays a major role in insulin resistance. Insulin works by phosphorylating the tyrosine residue of the insulin receptor substrate; on the other hand, insulin resistance develops due to impaired tyrosine phosphorylation, accelerated dephosphorylation, and phosphorylation of serine residues.1,13 There is also evidence that increased hepatic fatty acid/triglyceride content, hepatic steatosis, can cause insulin resistance through decreased phosphorylation of phosphoinositol-3 kinase—mediated protein kinase B.13 Furthermore, fatty acid derivatives can block the insulin cascade via the stimulation of protein kinase C as well as agonism of peroxisome proliferator-activated receptors and antagonism of nuclear factor ?B.9,13
Among the peroxisome proliferator-activated receptor family of nuclear receptors, peroxisome proliferator-activated receptor α is mainly expressed in the liver and is involved in the interaction between hepatic mitochondrial, peroxisomal, and microsomal fat oxidation and hepatic insulin resistance.9,13 Also, there is evidence that defects in mitochondrial fatty acid oxidation can lead to oxidative stress via reactive oxygen species formation, hepatic steatosis, and insulin resistance. Ibdah and colleagues15 reported that genetically engineered mice with impaired fatty acid oxidation secondary to a defect in mitochondrial trifunctional protein that catalyzes the first three steps in mitochondrial long-chain fatty acid oxidation develop NAFLD and NAFLD-associated insulin resistance. Figure 1 shows a liver section from a mouse with a mitochondrial trifunctional protein defect compared with a wild-type littermate control, demonstrating abundance of microvesicular and macrovesicular lipid droplets stained with oil red O in association with defective mitochondrial trifunctional protein.15 Cytokine-mediated injury of NAFLD/ NASH, especially TNF-α, has been well established, and several authors have attributed the mitochondrial dysfunction to the mitochondrial effects of TNF-α.9,16 This unifies the second-hit theme of NAFLD progression with oxidative stress and cytokine-mediated hepatocyte damage and further ties NASH/NAFLD to the CMS since they both share insulin resistance, oxidative stress, and cytokine-mediated damage.
The Importance of Including NAFLD/NASH in the Metabolic Syndrome and the CMS
CMS includes the metabolic syndrome along with its increased risk of type 2 diabetes mellitus, cardiovascular disease, and chronic kidney disease.11 Both CMS and NASH/NAFLD share insulin resistance as a common factor and, since insulin resistance is a key mediator in the end-organ pathogenesis of CMS, NAFLD/NASH should be included in the clinical picture of CMS.
NAFLD is defined as fat accumulation in the liver exceeding 5%–10% by weight and has been divided into four classes.1 Class 1 constitutes simple steatosis, class 2 includes steatosis with lobular inflammation, class 3 has the addition of ballooned hepatocytes, and class 4 requires the presence of fibrosis. Macrovesicular and microvesicular accumulation of triglycerides is the initial cellular remodeling that occurs during hepatocellular steatosis, with a single large fat droplet that displaces the nucleus (macrovesicular steatosis) or with small well defined intracytoplasmic droplets (microvesicular steatosis). The accumulation of fat starts in zone 3 (a poorly oxygenated area near the central vein) and, in more severe cases, may occupy the whole acinus. NASH has lobular inflammation, mainly consisting of scattered polymorphonuclear leukocytes and mononuclear cells and, in zone 3, peridinusoidal and portal fibrosis. Furthermore, hepatocellular ballooning is present near steatotic hepatocytes, typically in zone 3.
Hepatocyte injury, cell death, and associated inflammation activate the hepatic stellate cell. This cell is responsible for excess extracellular matrix deposition (types I and III collagen) and fibrosis. Hepatocyte apoptosis via TNF-α and transforming growth factor-β may also lead to fibrogenesis. The apoptosing hepatocyte is ingested by the Kupffer and hepatic stellate cells. These, in turn, release transforming growth factor-β, which is capable of activating the hepatic stellate cell. Furthermore, angiotensin II and norepinephrine have both been shown to activate the hepatic stellate cell's pro-fibrogenic role. Also, connective tissue growth factor is overexpressed in the livers of NASH patients and correlates with fibrosis. Once the necroinflammatory and steatotic features of NASH are lost, cirrhosis develops, which ultimately leads to end-stage liver disease.1,4,9,10
NASH is generally considered a clinically stable disorder and has a markedly better prognosis than alcoholic hepatitis. A population-based study in the United States found that patients with NAFLD had slightly lower overall survival than expected for the general population (standardized mortality ratio, 1.34; 95% confidence interval, 1.003–1.76).17 Higher mortality was associated with advancing age, impaired fasting glucose, and cirrhosis. NASH may be an important underlying cause of cryptogenic cirrhosis, particularly among older, diabetic women.18
In most patients, there is little change in liver function tests throughout the course of the disease. In a sizable minority, however, histologic progression occurs and a small fraction of patients progress to end-stage liver disease. A study of 103 patients who underwent serial liver biopsies (mean interval between biopsies, 3.2 years) found that fibrosis stage progressed in 37%, remained stable in 34%, and regressed in 29%.19 Independent predictors of fibrosis progression included diabetes mellitus, a low initial fibrosis stage, and a higher body mass index. The degree to which these changes may have reflected sampling variation of liver biopsy specimens is uncertain.
Despite the frequency of histologic progression and the histologic similarity between NASH and alcoholic hepatitis, the outcome is quite different in the two disorders. Approximately 38%–50% of patients with alcoholic hepatitis progress to cirrhosis over a 7-year period20; comparable values for NASH are much lower at 8%–26%.21 NASH is also associated with higher 5- and 10-year survival rates than alcoholic hepatitis (67% vs. 38% and 59% vs. 15%, respectively).
Most patients with NASH are asymptomatic, although fatigue, malaise, and vague right upper abdominal discomfort can be present.22 The most common presentation is liver function abnormalities detected on routine laboratory testing. Hepatomegaly is a frequent finding.22
Serum aspartate aminotransferase and alanine aminotransferase are elevated in almost 90% of patients.22 The aspartate aminotransferase-alanine aminotransferase ratio is usually <1; this is much lower than the ratio in alcoholic hepatitis, which is usually above 2 and averaged 2.85 in one report23 and 2.6 in another.24 Alkaline phosphatase is less frequently elevated and hyperbilirubinemia is uncommon.22 However, normal serum aminotransferases do not exclude the presence of advanced histologic features. This was illustrated in a study that included 51 patients with a fatty liver and normal alanine aminotransferase who had undergone a liver biopsy (either for unexplained hepatomegaly or as part of an evaluation for live-donor liver transplantation).25 A total of 12 subjects had bridging fibrosis while six had cirrhosis. The only independent predictor of advanced fibrosis was the presence of diabetes (relative risk, 2.3).
Ultrasonography often reveals a hyperechoic texture or a bright liver because of diffuse fatty infiltration (Figure 2).26 However, this is a nonspecific finding and should not be used to make the diagnosis of NASH. Both computed tomography and magnetic resonance imaging can identify steatosis but are not sufficiently sensitive to detect inflammation or fibrosis (Figure 3).27
Liver biopsy is the only way to confirm or exclude the diagnosis of NASH.27 Liver biopsy also permits determination of disease severity and may provide insight into prognosis. A histologic scoring system has been proposed that can assist in diagnosis of NAFLD and may be useful for assessing the response to therapy. However, sampling variability continues to be a limitation of liver biopsy in staging of NASH.
There is no proven effective therapy for NASH, although modification of risk factors, such as obesity, hyperlipidemia, and poor diabetic control is generally recommended.
Weight reduction should be gradual, since rapid weight loss has been associated with worsening of liver disease.28,29 One report suggested that weight loss should not exceed approximately 1.6 kg/wk in adults.30 Weight loss and increased physical activity can lead to sustained improvement in liver enzymes, histology, serum insulin levels, and quality of life.31–33 Several other potential treatments have been described, although none have been used routinely in clinical practice.
The observation that vitamin E decreases oxidative stress provides a rationale for its use in patients with NASH. A placebo-controlled trial involving 45 patients concluded that 6 months of treatment with a combination of vitamin E and C (1000 IU and 1000 mg, respectively) was associated with significant improvement in liver fibrosis.34 There was no benefit in necroinflammatory activity.
Several hypoglycemic agents continue to be studied. For example, one controlled trial included 36 patients with NASH who were randomly assigned to a lipid- and calorie-restricted diet with or without metformin (850 mg twice daily) for 6 months.35 Mean serum aminotransferase levels, insulin, and C-peptide levels decreased significantly in both groups but to a significantly greater extent in the metformin group. Improvement in necroinflammation was observed more frequently in patients in the metformin group, but results did not achieve statistical significance.
Another example is shown in an uncontrolled trial that included 30 adults with histologically confirmed NASH who were treated with rosiglitazone (4 mg twice daily) for 48 weeks.36 All patients were overweight (body mass index, >25 kg/m2), including 23% who were severely obese. Paired biopsies before and after treatment were available in 26 patients. The mean global necroinflammatory score improved significantly; in 10 patients (45%), the biopsies improved to an extent where they no longer met published criteria for NASH. There was also significant improvement in perisinusoidal fibrosis. Mean serum alanine aminotransferase levels showed corresponding improvement. Ursodeoxycholic acid was suggested in a pilot study of 40 patients37; however, a subsequent larger controlled trial showed no benefit.38
Angiotensin II is involved in the pathogenesis of hepatic fibrosis and enhances iron deposition and insulin resistance. A pilot study of the angiotensin II receptor antagonist losartan in seven patients with NASH suggested a benefit in blood markers of hepatic fibrosis and serum aminotransferase levels.39 Further studies are needed.
NAFLD/NASH is a very important component of the CMS, with evidence suggestive of common pathogenesis. There is no proven effective therapy for NASH. Attempts should be made to modify potential risk factors such as obesity, hyperlipidemia, and poor diabetic control.
Some hepatologists are already using insulin-sensitizing drugs (metformin or rosiglitazone) based on the preliminary data presented above. Further controlled data concerning the efficacy of these approaches should be available in the next few years. Given the slow rate of progression in most patients with NASH, we currently emphasize control of risk factors rather than medical therapy in most patients.