Update – Non-alcoholic fatty liver disease
2nd Oct 2012
NON-ALCOHOLIC fatty liver disease (NAFLD) is frequently encountered by GPs, as well as physicians of many different specialties.
NAFLD is an increasingly prevalent problem and also the most common cause of chronic liver disease in the Western world.
NAFLD is part of the metabolic syndrome. It is known to be associated with central obesity, insulin resistance, type 2 diabetes, hypertension, dyslipidaemia and, more recently, cardiovascular disease (CVD). Unlike chronic liver disease from any other aetiology, CVD is the number-one cause of mortality among NAFLD patients.
NAFLD comprises a pathological spectrum in disease severity ranging from simple steatosis and progressing to non-alcoholic steatohepatitis (NASH), with varying degrees of fibrosis and even cirrhosis. The histology mimics alcoholic liver disease. Therefore, it is essential to take a detailed history of alcohol consumption.
In reality, though, many patients may have a combination of NAFLD and alcoholic liver disease, as they consume excess alcohol and have a degree of obesity and/or type 2 diabetes.
There are other aetiologies such as medications that can potentially cause fatty liver, such as glucocorticoids, amiodarone, synthetic oestrogens and highly active antiretroviral therapy.
To date, there is still no single reliable pharmacological intervention for the prevention of NASH progression to fibrosis. This Update will review the latest information on NAFLD, with a focus on the diagnostic and management aspects.
According to data from the US,1 as much as 30% of the adult population has NAFLD, and about 3% have NASH.
Among morbidly obese patients, the proportions increase to 90% and 37% respectively.2
While no detailed Australian data exist, it is very likely to be similar. Japanese studies quote NAFLD incidence to be 31–86 cases per 1000 person-years.3
NAFLD prevalence has also increased among the paediatric population to 3–10% and 30–40% among obese children.4
Apart from its obvious association with obesity, NAFLD is also very pervasive among patients with type 2 diabetes, with steatosis seen in up to 70% of such patients.5
NATURAL HISTORY AND MORTALITY
NAFLD patients tend to have a better prognosis than the other common cause of fatty liver; namely, alcoholic liver disease. The natural history of NAFLD is a slow process of steatosis leading to NASH and progressive fibrosis occurring over several decades.
In one of the largest NAFLD cohorts followed up for 13 years, it was shown that the histological status at diagnosis determines the long-term prognosis. Up to 40% of patients with steatosis developed NASH with fibrosis. About 15% of NASH patients progressed to cirrhosis.6
The risk of developing hepatocellular carcinoma (HCC) is usually low in NASH patients who are pre-cirrhotic, but is a recognised phenomenon.
Male gender and patients with more severe metabolic syndrome seem to
be at higher risk of developing NASH and HCC.
In NASH with compensated cirrhosis, the risk of developing HCC is similar to that in other cirrhotic patients from chronic liver disease of any aetiology – about 7% over a 10-year period. In the US, up to 12% of liver transplants were because of NASH cirrhosis.4
Cases of recurrent NASH post-transplant have also been reported. More recently, many cohorts of NAFLD patients followed up for a long period highlight the number-one cause of mortality among NAFLD patients to be from CVD.
The pathogenesis of steatosis and progression to steatohepatitis and fibrosis/cirrhosis is not yet fully understood.8
Insulin resistance has been identified as one of the main contributors to steatosis formation.
According to Chitturi,9 insulin resistance is almost universal in NASH patients. This is not surprising, as NAFLD is often linked with other disorders caused by insulin resistance.
He found up to 98% of his cohort, irrespective of obesity, had insulin resistance.
The control group in this study was a chronic hepatitis C population who were fibrosis-, age- and sex-matched. The NASH (non-cirrhotic) subgroup had higher fasting insulin, C-peptide, and insulin resistance.
Prolonged insulin resistance is then thought to lead to hyperinsulinaemia. This is suggested to be due to increased beta cell production and not decreased hepatic extraction of insulin (as seen in all other cirrhosis causes).
Apart from insulin resistance, there is a proposed ‘second hit’ theory of oxidative stress leading to steatohepatitis, whereas the initial steatosis is secondary to accumulation of fatty acid chains in the liver.
Most NAFLD patients are asymptomatic. Occasionally, patients might complain of right upper-quadrant abdominal discomfort, likely from hepatomegaly. Non-specific symptoms such as fatigue and malaise might be found.
Physical examination would also be non-specific in NASH unless the patient is cirrhotic, and thus would feature stigmata of chronic liver disease. Examination may reveal features of metabolic syndrome such as increased adipose tissue deposition, particularly central obesity and hypertension.
One study reported an association between dorsocervical lipohypertrophy and NASH.
NAFLD is associated with many other conditions, particularly those with a similar underlying pathogenesis. For instance, in a small Australian prospective study, polycystic ovary syndrome has been shown to be present in up to 71% of women with NAFLD proven on imaging and/or biopsy regardless of their BMI.
NASH patients also have more cardiovascular events than patients without steatohepatitis. As previously alluded to, coronary artery disease (CAD) is the number-one mortality cause among NAFLD patients. Some recent studies suggest NAFLD as an independent risk factor for CAD, independent of traditional risk factors that are a part of the metabolic syndrome.
Another study suggests a linear relationship between increasing fibrosis stages and carotid intimal thickness, which is a known marker for atherosclerosis.
Most cases of fatty liver are found incidentally on ultrasound. Almost two-thirds of NAFLD patients have normal liver function tests, even the cirrhotic ones.
Derangement is usually mild transaminases rise with alanine aminotransferase (ALT) more than aspartate aminotransferase (AST), but rarely more than three times the upper limit of normal. In 90% of cases, AST:ALT is less than one. Alkaline phosphatase is less often elevated, but GGT is frequently elevated. Hypoalbuminaemia and hyperbillirubinaemia would indicate cirrhosis.
There are other biomarkers used mostly in clinical research that have been proposed to diagnose or predict progression of NAFLD to NASH such as leptin, adiponectin, high-sensitivity CRP and cytokeratin.10
Ultrasound is the most commonly used imaging modality with suspected NAFLD. Steatosis appears as hyperechogenic liver. The typical criteria for fatty liver are liver brightness, deep attenuation, vascular blurring and hepatorenal echo contrast.
A recent study examined the accuracy of ultrasound scan compared with liver biopsy. Ultrasound sensitivity was 64% and specificity was 97% for steatosis detection. The sensitivity rises to 91% in patients with at least 30% steatosis.
However, morbid obesity considerably reduces sensitivity and specificity.
The drawback from ultrasound scan is that it is operator-dependent.
Computed tomography (CT)
CT is similar to ultrasound scan in its diagnostic yield for steatosis.
Un-enhanced CT is accurate in predicting steatosis of greater than 30%. Non-contrast CT cannot exclude steatosis.
Steatosis can be seen, but CT cannot identify inflammation or fibrosis. Therefore, it cannot differentiate between benign NAFLD and NASH, which has the risk to progress to cirrhosis.
Magnetic resonance imaging (MRI)
Normal MRI can confidently exclude steatosis due to the signal differences between fat and water. MR spectroscopy measures triglyceride content with reproducible result.
However, MRI is difficult to access and has a higher expense.
FibroScan is an ultrasound-based technology measuring liver stiffness. There is a good correlation between liver biopsy-determined fibrosis stages I–IV and liver stiffness based on FibroScan. The positive predictive values range from 64% to 93% for the varying stages.11
A meta-analysis of nine studies found that the sensitivity and specificity for FibroScan in determining stage IV fibrosis were 81% and 93% respectively, and 70% and 84% for stages II–IV. However, steatosis and BMI > 25 reduce reproducibility. It can also be technically difficult to do for obese patients.
Liver stiffness measurements tend to be overestimated in patients with higher ALT (stiffer liver during inflammation).
Liver biopsy is the gold standard in terms of its accuracy in diagnosing fatty liver disease, but it carries morbidity risks. The current recommendation is to perform biopsy on patients at risk of progression to NASH/fibrosis, such as those with metabolic syndrome.3
The pathological spectrum of NAFLD can be classified from simple steatosis through to steatohepatitis, fibrosis and cirrhosis. Liver biopsy features of NAFLD/NASH would show steatosis, hepatocyte balloon degeneration, mixed acute and chronic inflammation, pericellular fibrosis and Mallory-Denk bodies. Mitochondrial abnormalities can be seen in NASH with a special liver stain.
There are no pathological features to definitively distinguish NAFLD from alcoholic liver disease.
Furthermore, features of steatosis and inflammation often disappear in advanced disease, possibly explaining ‘cryptogenic’ cirrhosis diagnosis.
Thus far we know that progressive disease is limited to the NASH subgroup. The histological findings in this subgroup tend to show portal or periportal fibrosis, which then progress to linking fibrous septa and, later, cirrhosis. This is not seen in simple steatosis. Hepatocyte ballooning, portal inflammation and ductal reaction are associated independently with fibrosis formation.
Fatty liver disease: steatosis, fibrosis and inflammation.
Treatment for NASH will likely involve a multidisciplinary approach, with the main focus being to reverse the metabolic syndrome components and the mechanism behind it such as insulin resistance, oxidative stress and obesity/diabetes.
Currently there is no single effective pharmacological therapy to slow disease progression to NASH or cirrhosis. Diet and exercise remain the first-line therapy for NASH.
Diet and exercise
Patients with NAFLD/NASH are often overweight. Dietary kilojoule restriction has been shown to improve insulin resistance and intrahepatic lipid content. Studies show that kilojoule restriction improves ALT, AST and the rest of hepatic serology. However, the degree of restriction is still under question. One study quotes that restriction down to 6000kJ a day for 12 months leads to 60% histological improvement, with mean weight loss of 7%.12
However, starvation and too-rapid weight loss lead to worsening liver histology (fibrosis) and should be avoided. Alcohol consumption should be minimal, as high intake can increase oxidative stress and steatosis formation.
Exercise could lead to positive modification of metabolic syndrome and reduce the incidence of insulin resistance. Physical exercise has been shown to reduce enzyme abnormalities, and improve histological abnormality, insulin level and quality of life.
Patient should aim for at least 6–7% weight loss from their original starting weight. Weight loss should be gradual, aiming for a loss of 1.6kg per week. The Gastroenterological Society of Australia recommends aerobic exercise for 140 minutes/day, spread over at least four days, aiming for no more than 0.5kg per week of weight loss.
Conservative management with diet and exercise to lose weight is often difficult in the morbidly obese. In some studies, bariatric surgery has been shown to improve and even resolve metabolic syndrome.
A systematic review of the effect of bariatric surgery on NAFLD patients13concludes that after a two-year follow-up, 766 patients (with paired liver biopsy) who had bariatric surgery were found to have improved steatosis, inflammation and fibrosis.
In 69% of cases, resolution of NASH was observed. This is thought to be due to a decrease in hepatic factors available that are involved in regulation of inflammation and fibrosis.
Bariatric surgery is not contraindicated in eligible patients with NASH; however, based on the available data, it is still premature to recommend bariatric surgery as a part of definitive NASH management.3
A few medications have been studied to varying degrees for treatment of NASH, such as insulin sensitiser, antioxidants and weight-loss inducers, among others.
There is some evidence for metformin use in NAFLD, but it is not currently recommended as a specific treatment for NASH.3 It decreases hepatic gluconeogenesis.
In a study looking at the effect of kilojoule-restricted diet with or without metformin for six months, mean serum aminotransferase levels, insulin and C-peptide decreased significantly in both groups, but more so in the metformin group.
In an Italian randomised controlled trial involving 110 patients, metformin was more effective in terms of reductions in necroinflammation and steatosis than a combination of weight loss and vitamin E.
Pioglitazone and vitamin E
PIVENS is a randomised controlled study14 looking at the efficacy of vitamin E, pioglitazone versus placebo among 247 (non-diabetic) NASH patients over
There was histological improvement in 34% of subjects in the pioglitazone group and 43% in the vitamin E group. However, there was no improvement in fibrosis score, nor was there reduction in portal inflammation (known risk of progression).
The study concludes that vitamin E is superior to placebo in non-diabetic adults. Based on current recommendations, pioglitazone and vitamin E can be used in treatment of biopsy-proven NASH patients.
As with other insulin sensitisers, there is continuing concern about possible increased risk of heart disease associated with thiazolidinediones.
Orlistat is a reversible inhibitor of gastric and pancreatic lipase. It is the most studied weight-loss medication in NASH.
There were two randomised controlled trials spanning 6–9 months, which resulted in 8% and 6% weight loss respectively after taking orlistat. The histology from those studies showed improved steatosis and inflammation, but not fibrosis.
Since weight loss is a common pathway of benefit in NAFLD patients, improvements in biochemistry and histology were seen. The study suggests 5–10% weight loss to achieve end points.
A few pilot studies have suggested that pentoxifylline can reduce liver injury among NASH patients.
In one small, recently published randomised controlled trial from the US, pentoxifylline reduced aminotransferase levels, steatosis and cellular ballooning.
However the improvements were non-significant compared to placebo.
The study suggests that pentoxifylline is safe and well-tolerated and may be used in the future for treatment of NASH.14
There is often associated high cholesterol (mainly dyslipidaemia) among NAFLD patients as a part of their metabolic syndrome.
Statin therapy is beneficial among NAFLD patients to reduce steatosis on imaging as well as their cholesterol level.10
Evidence is pointing towards the safety of statins for NAFLD patients, and adverse reaction is thought to be idiosyncratic.
Statins should be used to treat dyslipidaemia among NAFLD patients, but there is no evidence currently for its use to specifically treat NAFLD/NASH.3
NAFLD is a common problem encountered in general practice. Thus far, predictors of more aggressive disease are not readily available and liver biopsy is invasive.
The presence of NAFLD should be a clear indication for diabetes screening.15 More aggressive follow-up and management of metabolic syndrome are also warranted due to the increased risk of IHD among the NAFLD population.
There are no released guidelines on how to follow up these NAFLD patients.
Fatty liver can be caused by obesity, type 2 diabetes, alcohol, glucocorticoids, amiodarone, synthetic oestrogens and antiretroviral therapy.
Cardiovascular disease is the most common cause of death in non-alcoholic fatty liver disease (NAFLD).
NAFLD slowly progresses to steatosis, leading to non-alcoholic steatohepatitis (NASH), then to fibrosis.
Insulin resistance has been identified as one of the main contributors to steatosis formation.
Normal MRI can confidently exclude steatosis.
Liver biopsy is currently the gold standard in terms of its accuracy in diagnosing fatty liver disease, but it carries morbidity risks.
Diet and exercise remain the first-line therapy for NASH.