Advances in the treatment of nonalcoholic fatty liver disease

Abstract

Nonalcoholic fatty liver disease (NAFLD) is the most common cause of chronic liver disease in the Western world, and its prevalence is predicted to rise in the future in parallel with rising levels of obesity and type 2 diabetes mellitus. It is commonly associated with insulin resistance. Many patients have coexisting obesity, hypertension, dyslipidaemia or hyperglycaemia, and are at increased risk of developing cardiovascular disease. Although patients with simple steatosis have a good prognosis, a significant percentage will develop nonalcoholic steatohepatitis which may progress to cirrhosis, end-stage liver failure and hepatocellular carcinoma. Despite promising results from several pilot studies and small to medium randomized controlled trials, there is currently no pharmacological agent that is licensed for the treatment of NAFLD. At present the mainstay of treatment for all patients is lifestyle modification using a combination of diet, exercise and behavioural therapy. With recent advances in the understanding of the pathogenesis of NAFLD, the goal of treatment has shifted from simply trying to clear fat from the liver and prevent progressive liver damage to addressing and treating the metabolic risk factors for the condition. To reduce liver-related and cardiovascular morbidity and mortality, all patients with NAFLD should be invited to enrol in adequately powered, randomized controlled studies testing novel therapies, many of which are targeted at reducing insulin resistance or preventing progressive liver disease. Coexisting obesity, hypertension, dyslipidaemia or hyperglycaemia should be treated aggressively. Orlistat, bariatric surgery, angiotensin receptor blockers, statins, fibrates, metformin and thiazolidinediones should all be considered, but treatments should be carefully tailored to meet the specific requirements of each patient. The efficacy and safety of any new treatment, as well as its cost-effectiveness, will need to be carefully evaluated before it can be advocated for widespread clinical use.

Introduction

Nonalcoholic fatty liver disease (NAFLD) is a common condition characterized by the accumulation of fat, mainly triglyceride (TG), within the liver [Angulo, 2002]. It is histologically identical to alcohol-induced liver injury, but occurs in individuals who do not consume alcohol in excess. The term NAFLD encompasses a wide spectrum of liver disease ranging from simple steatosis (fatty liver), which is generally considered benign, to steatosis with evidence of hepatocellular inflammation and damage (nonalcoholic steatohepatitis [NASH]) which may progress to cirrhosis, liver failure and even hepatocellular carcinoma.

In the Western world, estimates of the prevalence of NAFLD vary between 20% and 30% in the general adult population [Bedogni et al. 2005; Browning et al. 2004]. This prevalence rises to 70% in patients with type 2 diabetes mellitus (T2DM) and to 90% among the morbidly obese [Targher et al. 2007; Machado et al. 2006]. The prevalence of NASH is less common, affecting an estimated 2–3% of the general population and up to a third of the morbidly obese [Machado et al. 2006]. NAFLD is associated with insulin resistance and is now recognized as the hepatic manifestation of the metabolic syndrome [Kotronen and Yki-Jarvinen, 2008; Marchesini et al. 2001a]. Hypertension, hypertriglyceridaemia and mixed hyperlipidae-mia have also been independently linked to NAFLD, and recent studies suggest it is an independent risk factor for cardiovascular disease [Targher et al. 2007; Ekstedt et al. 2006; Donati et al. 2004; Assy et al. 2000].

Whilst fatty liver in the absence of significant fibrosis is generally thought to be a relatively benign condition, the presence of fibrosis on liver biopsy predicts both disease progression and liver-related complications over the next 10 years [Ekstedt et al. 2006; Day, 2005]. NASH also carries an increased risk of hepatocellular carcinoma [Ekstedt et al. 2006].

Aims of treatment

The aims of treatment in NAFLD are to reduce liver-related morbidity/mortality and cardiovascular morbidity/mortality. As our knowledge of the biochemical mechanisms leading to both the development and progression of NAFLD has improved, the goal of treatment has shifted from simply trying to clear fat from the liver to address and treat the metabolic risk factors for fatty liver. Patients with coexisting hypertension, dyslipidaemia and hyperglycaemia should be treated aggressively with a combination of lifestyle modification plus pharmacological therapy if necessary in order to reduce cardiovascular risk. As simple steatosis in the absence of significant fibrosis has a very good prognosis, efforts should be made to identify and treat patients with NASH who are at risk of developing progressive liver disease. At present there are no clinical, biochemical or imaging methods that can reliably identify patients with steatohepatitis, and the gold standard for the diagnosis and staging NAFLD remains liver biopsy [Mehta et al. 2008]. No reliable recommendations are currently in place to indicate which patients with NAFLD should undergo liver biopsy, although persistent elevations in aminotransferases or the presence of the risk factors of age >45 years, aspartate aminotransferase/alanine aminotransferase (AST/ALT) ratio >1, T2DM and body mass index (BMI) >30 kg/m² strengthen the case for liver biopsy [Preiss and Sattar, 2008]. Assessing the potential risks and benefits of liver biopsy remain in the hands of individual clinicians and patients.

Pathogenesis and identification of targets for treatment

A full discussion of the pathogenesis of NAFLD is beyond the scope of this review and has been recently covered elsewhere [Dowman et al. 2010]. However, a brief summary is presented below to make the reader aware of areas potentially amenable to intervention. Table 1 lists potential treatment modalities.

Table 1.

Potential therapeutic targets for nonalcoholic fatty liver disease.

Target	Treatment
Obesity	Weight loss Lifestyle modification Medications Orlistat, sibutramine Bariatric surgery
Insulin resistance	Insulin sensitizers Metformin, Thiazolidinediones (TZDs)
Dyslipidaemia	Lipid lowering agents Statins, Fibrates, Polyunsaturated fatty acids
Oxidative stress	Antioxidants Vitamins E and C, Betaine, N-acetylcysteine
	Probiotics VSL#3, Oligofructose
	Venesection
Cellular apoptosis	Cytoprotective agents Ursodeoxycholic acid (UDCA), Lecithin
Pro-inflammatory cytokines	Anti-tumour necrosis factor alpha agents Pentoxifylline
End-stage liver failure	Transplantation
Other	Novel treatments Angiotensin receptor blockers (ARBs) Incretin hormones Sulfonylureas

The pathogenesis of NAFLD is complex and incompletely understood. The liver synthesizes TG from the esterification of free fatty acids (FFA) and glycerol within the hepatocyte. FFA is derived from the lipolysis of TG within adipose tissue, diet, uptake of chylomicron remnants or de novo lipogenesis. Once taken up by the liver, FFA can either be oxidized in the mitochondria to form adenosine triphosphate (ATP) or esterified to produce TG for storage, or incorporation into very low-density lipoprotein (VLDL) particles. Triglycerides accumulate in the liver when their synthesis exceeds their export via VLDL [Adams et al. 2005].

Such hepatic TG accumulation is often referred to as the ‘first hit’ in the pathogenesis of NASH, as it increases the susceptibility of the liver to injury mediated by a number of ‘second hits’, such as pro-inflammatory cytokines, mitochondrial dysfunction, oxidative stress and gut-derived bacterial endotoxin, which in turn lead to the development of steatohepatitis and/or fibrosis [Day, 2006; Day and James, 1998]. In chronic liver injury, the development of fibrosis and cirrhosis is dependent on the efficacy of hepatocyte regeneration. In the healthy liver cell death stimulates replication of mature hepatocytes, but in the presence of oxidative stress the replication of mature hepatocytes is inhibited. Cell death with impaired proliferation of hepatocyte progenitors has been proposed as the ‘third hit’ in the pathogenesis and progression of NASH [Jou et al. 2008].

Treatment

Potential therapeutic targets for NAFLD are listed in Table 1. As insulin resistance is central to the pathogenesis of NAFLD, therapies targeting obesity and insulin resistance are reviewed first, followed by a discussion of lipid-lowering therapies, antioxidants, cytoprotective agents, anti-tumour necrosis factor (TNF) agents and finally new and emerging treatments.

Weight loss through lifestyle changes

Overweight or obese patients with NAFLD should be encouraged to adopt a program of dietary self-management and moderate daily physical activity. Adherence to combined dietary restriction and increased physical activity has been shown to result in larger and progressive weight loss that can be maintained over time [Saris et al. 2003]. Many studies have examined the effects of weight loss achieved by diet with or without exercise. However, the majority of these have had no control group and have not included histological data from paired liver biopsies before and after the intervention. Promrat and coworkers recently performed a small randomized controlled trial testing the effects of weight loss on NASH [Promrat et al. 2010]. They demonstrated that a 12-month intensive lifestyle intervention using a combination of diet, exercise and behaviour modification significantly promoted weight loss and improved steatosis and lobular inflammation without improving fibrosis.

Dietary composition also has a role to play in both the pathogenesis and treatment of NAFLD. Patients with hepatic steatosis have been reported to consume a diet rich in refined sugars and saturated fats and poor in fibre and antioxidants [Solga et al. 2004; Musso et al. 2003]. Browning and coworkers and Westerbacka and colleagues have shown that either reducing dietary carbohydrate intake or fat content may be useful in reducing intrahepatic TG content [Browning et al. 2006; Westerbacka et al. 2005]. Zelber-Sagi and colleagues recently demonstrated that patients with NAFLD consume almost double the amount of soft drinks compared with the general population, and the daily intake of these is associated with an increased risk for NAFLD, independently of age, gender, BMI and total daily calorie intake [Zelber-Sagi et al. 2007]. Furthermore, in the last 2 years, a series of studies have shown that large amounts of fructose in the form of high fructose corn syrup present in carbonated drinks may have a direct role in liver fat accumulation [Kechagias et al. 2008; Marchesini et al. 2008]. Thus, in individual patients, specific dietary habits may constitute specific targets for behavioural changes, although the aim should be to achieve a healthy balanced diet.

Although the optimum amount of exercise to support long-term weight loss is unclear, weight reduction should be achieved by aiming for a calorie deficit (with combined dietary restriction and exercise) of ~500 kcal/day, and should continue at a rate of ~0.5 kg per week. More aggressive weight loss (> 1.6 kg/week) may result in the rapid mobilization of fatty acids from visceral fat depots which can be taken up by the liver and exacerbate pre-existing hepatic inflammation and fibrosis [Andersen et al. 1991]. In patients with T2DM, moderate weight reduction (8%) has been shown to improve hepatic steatosis [Petersen et al. 2005]. The reduction in liver fat was accompanied by a dramatic improvement in hepatic insulin sensitivity, such that suppression of hepatic glucose production by insulin returned to normal.

However, because lifestyle changes associated with dietary restriction and exercise are so difficult to sustain for most patients, attention has turned to other means of achieving sustainable weight loss.

Weight loss by pharmacological measures

Pharmacological treatment for obesity may be offered to patients who have failed to lose body weight by lifestyle interventions alone.

Orlistat, an enteric lipase inhibitor which reduces dietary fat absorption, has been evaluated in NAFLD in several studies. In a pilot study of 10 obese patients with NASH treated with orlistat for 6 months, Harrison and colleagues demonstrated a reduction in aminotransferase levels with improved liver histology in 9 out of the 10 patients [Harrison et al. 2004]. They noted that improvements in steatosis and fibrosis were generally associated with a weight loss of 10% or more. In a small, double-blind, randomized, placebo-controlled trial of patients with NAFLD, Zelber-Sagi and coworkers showed that treatment with orlistat for 6 months led to a twofold reduction in serum ALT levels and a statistically significant reversal of liver steatosis detected by ultrasound [Zelber-Sagi et al. 2006]. More recently, Hussein and colleagues conducted an open-label study in which 14 patients underwent liver biopsy before and after 6 months treatment with orlistat [Hussein et al. 2007]. At the end of the treatment period, 10 patients (70%) had reduced fatty infiltration. Inflammation and fibrosis improved by two grades in 22% and by one grade in 50% of patients. There was a reduction in serum aminotransferase levels, total cholesterol, TG and low-density lipoprotein cholesterol (LDL-C), and an improvement in insulin sensitivity. Although the size of the cohort was small, the findings included histological data from paired liver biopsies.

Sibutramine is a serotonin and norepinephrine reuptake inhibitor that acts by enhancing satiety, thus helping to reduce food intake. Its use has been compared with orlistat in obese patients with NASH [Sabuncu et al. 2003].

Both treatment groups lost weight and improvements were demonstrated in both serum aminotransferase levels and in the extent of steatosis visible on ultrasound, but there were no data on histological outcomes.

Despite these encouraging findings, questions have been raised about the long-term safety profile of both these agents and whether sustained weight loss can be achieved. Orlistat causes gastrointestinal side effects and malabsorption of fat soluble vitamins in up to 30% of patients, and sibutramine can elevate heart rate and blood pressure [Padwal et al. 2004; Kim et al. 2003]. In fact, sibutramine has recently been withdrawn from use by the European Medicines Agency (EMA) after an interim analysis of the Sibutramine Cardiovascular Outcome (SCOUT) Study found the drug increased morbidity from cardiovascular disease [Williams, 2010].

In summary, there is very little controlled clinical trial evidence to support the hypothesis that either orlistat or sibutramine improve NAFLD in the short term. There is currently no available long-term data on the effect of these medications on liver-related outcomes such as cirrhosis or its complications.

Weight loss through bariatric surgery

In patients with NAFLD who are severely obese (BMI > 40 kg/m²) or have a BMI > 35 kg/m² with obesity-associated comorbidities, lifestyle modification alone may not be enough to achieve sustained weight loss, and so bariatric surgery can be considered. Gastric restrictive procedures (such as vertical banded gastroplasty, adjustable gastric banding and Roux-en-Y gastric bypass) reduce gastric volume and thereby create a mechanical barrier to the ingestion of food. They promote gradual weight reduction and are currently the bariatric techniques of choice. At least five small studies have examined the effect of Roux-en-Y gastric bypass on patients with NASH [Furuya et al. 2007; Liu et al. 2007; de Almeida et al. 2006; Barker et al. 2006; Clark et al. 2005]. Paired liver biopsies during and after Roux-en-Y gastric bypass were performed in a total of 108 patients. All five studies reported varied measures of histological improvement, with NASH resolving in up to 89% [Barker et al. 2006]. At least four studies have examined the effect of gastroplasty on NAFLD [Jaskiewicz et al. 2006; Stratopoulos et al. 2005; Luyckx et al. 1998; Ranlov and Hardt, 1990]. Improvement in hepatic steatosis was reported in all four studies, but the reports regarding inflammation and fibrosis are mixed. Reports on patients with paired liver biopsies with laparoscopic adjustable gastric banding (LAGB) are limited. Two published studies by Dixon and colleagues showed improvement in steatosis and regression of fibrosis after LAGB [Dixon et al. 2006, 2004].

There is evidence to suggest that rapid weight loss can lead to progression of hepatic inflammation and fibrosis [Andersen et al. 1991], and some authors remain concerned that the risk of liver disease progression due to rapid weight loss during the first few months postsurgery makes the role of bariatric surgery in the treatment of NAFLD unclear. Nevertheless, the available data suggest that in the setting of appropriate surgical expertise, these procedures can be safe and may reverse both the biochemical and histological abnormalities associated with NASH.

Insulin-sensitizing agents

The biguanide metformin is widely used worldwide either alone or in combination with sulfo-nylureas, thiazolidinediones (TZDs) or insulin for the treatment of T2DM. The primary anti-hyperglycaemic action of metformin results from improved insulin sensitivity, primarily in the liver and secondarily in skeletal muscle [Scarpello and Howlett, 2008]. Within the liver, the principle action of metformin is to reduce hepatic glucose production, largely by inhibiting gluconeogenesis but also by inhibiting glycogenolysis [Scarpello and Howlett, 2008; Natali and Ferrannini, 2006]. The increase in peripheral glucose ulitization (between 10% and 30%) arises largely through nonoxidative glucose ulitization in skeletal muscle [Natali and Ferrannini, 2006].

The most compelling evidence for a potential widespread use of metformin comes from the large United States Diabetes Prevention Program, in which 3234 subjects with impaired glucose tolerance were randomized to placebo, metformin (850 mg twice daily) or a lifestyle-modification program, with a follow-up of 2.8 years [Knowler et al. 2002]. Lifestyle intervention reduced the incidence of T2DM by 58% and metformin by 31% compared with placebo [Knowler et al. 2002].

Several studies have examined the effect of metformin in patients with NAFLD, showing it to consistently improve insulin resistance and serum liver enzymes without weight gain, but with more variable improvements in liver histology [Bugianesi et al. 2005; Schwimmer et al. 2005; Duseja et al. 2004; Nair et al. 2004; Uygun et al. 2004; Marchesini et al. 2001b]. However, only two of these have been randomized controlled trials [Bugianesi et al. 2005; Uygun et al. 2004].

Uygun and coworkers randomized 36 patients to calorie restriction alone or calorie restriction plus metformin [Uygun et al. 2004]. Significant improvement in serum aminotransferases and plasma insulin and C-peptide levels were noted in the metformin group. Although greater improvement in necroinflammatory activity was noted in the metformin group, the difference was not statistically significant.

Bugianesi and colleagues randomized 55 patients to receive metformin 2000 mg daily for 12 months, 28 patients to receive vitamin E 800 IU/day, and 27 patients to a prescriptive diet [Bugianesi et al. 2005]. Owing to concerns raised by the ethics committee, only 17 patients treated with metformin underwent biopsy at the end of treatment, but significant reductions in steatosis, necroinflammation and fibrosis were reported.

TZDs are a novel class of drugs that directly reduce insulin resistance by enhancing insulin action in adipose tissue, skeletal muscle and liver [Yki-Jarvinen, 2004]. Currently, two TZDs, rosiglitazone and pioglitazone, are approved for the treatment of hyperglycaemia in T2DM, as monotherapy or in combination with other oral hypoglycaemic agents.

TZDs act as agonists of peroxisome proliferator activator receptor gamma (PPAR-γ) receptor, a nuclear ligand-activated transcription factor that regulates multiple target genes. PPAR-γ is most abundantly expressed in adipose tissue, but also in pancreatic β-cells, vascular endothelium, macrophages and skeletal muscle cells. TZDs improve hepatic and peripheral insulin sensitivity by promoting fatty acid uptake into adipose tissue, decreasing serum FFA concentrations, and increasing hepatic fatty acid oxidation. They also increase the production of the insulin sensitizing cytokine adiponectin and reduce circulating levels of several pro-inflammatory mediators. They are thought to do this by acting as selective ligands of the PPAR-γ receptors, predominantly in adipose tissue [Yki-Jarvinen, 2004].

TZDs have been shown to prevent the progression to T2DM in high-risk individuals. In the large randomized DREAM (Diabetes Reduction Assessment with ramipril and rosiglitazone Medication) study, rosiglitazone, compared with placebo, significantly reduced the incidence of T2DM and increased the likelihood of regression to normoglycaemia in participants with glucose intolerance and no prior history of cardiovascular disease [Gerstein et al. 2006].

The above data are the rationale for the clinical use of TZDs in the treatment of NAFLD. Several pilot studies examining the effect of TZDs on NAFLD and NASH have reported favourable results, with improvement in both liver function tests and liver histology [Belfort et al. 2006; Promrat et al. 2004; Sanyal et al. 2004; Neuschwander-Tetri et al. 2003]. In a large, recently published trial, Aithal and colleagues randomized 74 biopsy-proven NASH patients without diabetes to receive either pioglitazone or placebo with standard diet and exercise for 12 months [Aithal et al. 2008]. Sixty one patients (30 in the placebo group and 31 in the pioglitazone treated group) had follow-up biopsies, which showed significant improvements in liver injury and fibrosis in the pioglitazone treated group, despite a weight gain of ~3 kg. In the recent FLIRT (Fatty Liver Improvement with Rosiglitazone Therapy) trial, Ratziu and colleagues showed rosiglitazone to significantly improve serum aminotransferase levels and the histological score of steatosis but not necroinflammation or fibrosis [Ratziu et al. 2008]. An important large, multicentre, randomized, double-blinded, placebo-controlled trial that has very recently been published is the ‘Pioglitazone or Vitamin E for Nonalcoholic Steatohepatitis’ (PIVENS) study, comparing pioglitazone with vitamin E or placebo [Sanyal et al. 2010]. In this study, which was undertaken by investigators from the NASH Clinical Research Network, 247 adults with NASH without diabetes were randomly assigned to receive either pioglitazone at a dose of 30 mg/day (80 subjects), Vitamin E at a dose of 800 IU/day (84 subjects) or placebo (83 subjects), for 96 weeks. Paired liver biopsies were performed before and after treatment. The primary outcome was an improvement in the histological features of NASH, as assessed by the use of a composite of standardized scores for steatosis, lobular inflammation, hepatocellular ballooning and fibrosis. Although there was no benefit of pioglitazone over placebo for the primary outcome, pioglitazone was associated with highly significant reductions in steatosis, inflammation and hepatocellular ballooning, in addition to improvements in insulin resistance and liver enzyme levels. However, subjects who received pioglitazone gained more weight than those who received vitamin E or placebo.

On a comparative basis, TZDs seem to be more effective than metformin [Tiikkainen et al. 2004]. However, they need to be used continuously, as their benefits appear to be reversed on discontinuation of treatment [Lutchman et al. 2007].

Despite these favourable results, further, adequately powered, randomized, placebo- controlled trials of longer duration with histological data from paired biopsies are required to confirm the histological benefits of TZDs and metformin in NAFLD. In addition, the side-effect profile of both classes of drug should be borne in mind. TZDs can cause weight gain and congestive cardiac failure [Aithal et al. 2008; Lago et al. 2007]. In addition, rosiglitazone has deleterious effects on bone mineral density and may increase cardiovascular morbidity and mortality [Grey et al. 2007; Nissen and Wolski, 2007]. Metformin is generally well tolerated with the most commonest adverse effects being gastrointestinal [Scarpello and Howlett, 2008]. Neither metformin nor TZDs have so far been approved by United States or European agencies for the treatment of NAFLD [Targher et al. 2010].

Lipid-lowering agents

Interest in the use of lipid lowering agents for the treatment of NAFLD has stemmed from the close association between NAFLD and dyslipidaemia [Kotronen and Yki-Jarvinen, 2008; Assy et al. 2000].

Statins reduce cholesterol production and hence serum cholesterol by competitively inhibiting hepatic 3-hydroxyl-3-methylglutaryl coenzyme A (HMG CoA) reductase. They are commonly used in patients with vascular disease and T2DM to prevent further vascular events. The use of statins in patients with chronic liver disease has raised concerns about their potential to cause hepatotoxicity. However, it is now established that statin use in patients with compensated liver disease is essentially safe, with the incidence of reported hepatotoxicity being exceedingly rare [Browning, 2006].

Only a handful of studies have examined the efficacy of statins for NAFLD treatment. Rallidis and colleagues performed a very small pilot study in which they examined pravastatin use in four patients with NASH for 6 months [Rallidis et al. 2004]. They reported improvement in inflammation in three patients and improvement in steatosis in one patient. A beneficial effect of atorvastatin has so far been proven only on aminotransferases and ultrasonographically detected steatosis [Gomez-Dominguez et al. 2006]. In view of the above, at present no conclusions can be drawn about the efficacy of statins in the treatment of NAFLD.

There is some suggestion that fibrates, such as clofibrate, fenofibrate and gemfibrozil may be of benefit in the treatment of NAFLD. A 12-month pilot study comparing 16 biopsy-proven NASH patients treated with clofibrate and 24 NASH patients treated with ursodeoxycholic acid (UDCA) reported significant improvement in serum ALT and gamma glutamyl transpeptidase (GGT) levels and liver histology with UDCA, but no significant changes in levels of amino-transferases or liver histology in the clofibrate group [Laurin et al. 1996].

A 4-week randomized controlled trial of gemfibrozil 600 mg/day showed significant improvements in serum ALT levels, but histological data were not obtained [Basaranoglu et al. 1999].

Fenofibrate has been shown to reduce total cardiovascular events in the FIELD study, mainly due to fewer nonfatal myocardial infarctions and revascularizations [Keech et al. 2005]. Pioglitazone, a TZD with weak PPAR-α activity, has been shown to improve both liver enzyme levels and liver histology in patients with NAFLD [Aithal et al. 2008; Belfort et al. 2006]. Thus, it is possible that fenofibrate may also be of benefit in the treatment of NAFLD, although this is yet to be evaluated in clinical trials [Zambon and Cusi, 2007].

Polyunsaturated fatty acids (PUFA) are ligands of PPAR-α [Reddy, 2001]. Deficiency of the n-3 series long-chain polyunsaturated fatty acids (LCPUFA) with subsequent increased n-6/n-3 fatty acid ratio leads to impairment of PPAR-α activity in hepatocytes, followed by higher hepatic uptake of FFA, reduction in hepatic β-oxidation and an up-regulation of the lipogenic transcription factor SREBP-1c. Studies in rats and mice showed that a diet enriched in n-3 PUFA increased insulin sensitivity, reduced liver TG and improved steatohepatitis [Levy et al. 2004]. Araya and colleagues demonstrated an increased n-6/n-3 fatty acid ratio in patients with NAFLD compared with controls [Araya et al. 2004]. In the first pilot study in humans, Capanni and colleagues treated 42 NAFLD patients with n-3 PUFA (1g/day) for 12 months and demonstrated a reduction in serum ALT levels and ultrasound features of hepatic steatosis [Capanni et al. 2006]. In a larger study of 144 patients with NAFLD treated with n-3 PUFA (2 g/day) for 24 weeks, Zhu and colleagues reported improvement in serum ALT, lipid levels and normalization of ultrasonographic features [Zhu et al. 2008]. A disadvantage of both studies was the lack of liver histology and description of their ultrasound scoring system.

Antioxidants

The theory that NASH develops as a result of oxidative stress on the steatotic liver has generated interest in the therapeutic role of several antioxidants in the treatment of NAFLD.

α-tocopherol, the form of vitamin E that is preferentially metabolized in humans, inhibits transforming growth factor β1(TGF-β1), which is thought to contribute to fibrosis progression. Several pilot studies and three randomized controlled studies have examined the effect of vitamin E in patients with NAFLD, with inconsistent findings. In the first randomized placebo-controlled trial, Harrison and colleagues randomized 45 patients with biopsy-proven NASH to 6 months treatment with vitamin E (1000 IU/day) plus vitamin C (1000 mg/day) versus placebo [Harrison et al. 2003]. Repeat liver biopsy showed a small improvement in fibrosis score in the vitamin E group but no significant improvement in inflammation or necrosis score. In a second, controlled study, Kugelmas and coworkers randomized 16 adults with biopsy-proven NASH to vitamin E (800 IU/day) plus diet and exercise versus diet and exercise alone for 12 weeks [Kugelmas et al. 2003]. All patients demonstrated an improvement in weight and serum aminotransferases, but the authors did not detect an added benefit of vitamin E over and above lifestyle modification alone. The recently published PIVENS study, showed that vitamin E, compared with placebo, was associated with a significantly higher rate of improvement in histological features of NASH, as assessed by the use of a composite of standardized scores for steatohepatitis, lobular inflammation, hepatocyte ballooning and fibrosis. Vitamin E was associated with a reduction in hepatic steatosis and lobular inflammation, but not with an improvement in fibrosis scores [Sanyal et al. 2010].

Betaine (trimethylglycine) is a naturally occurring metabolite of choline that serves as an alternative methyl donor in the conversion of homocysteine to methionine and in the formation of phosphatidylcholine from phosphatidylethanolamine [Neuschwander-Tetri, 2001]. Phosphatidylcholine (lecithin) is a component of cell membranes and VLDL, the vehicle through which TG are exported from the liver. Theoretical benefits of betaine include antioxidant effects (by increasing levels of S-adenosyl-methionine (SAMe) as well as cell membrane stabilization (through the production of phosphatidylcholine). Abdelmalek and colleagues conducted an uncontrolled pilot study of the pharmaceutical form of betaine—betaine anhydrous (20g/day) in 10 patients with biopsy-proven NASH for 1 year [Abdelmalek et al. 2001]. Although in the seven patients who completed the study there was a statistically significant improvement in serum aminotransferase levels, the improvement in the amount of steatosis, histological inflammation and fibrosis were not statistically significant. Larger, controlled studies of betaine are required before any recommendations regarding its use in NASH can be made.

N-acetylcysteine has been studied in animal models of hepatic steatosis [Nakano et al. 1997]. In a small human study of 11 patients with NASH managed with diet regulation followed by N-acetylcysteine therapy, Gulbahar and colleagues reported improvements in amino-tranferases, although no liver biopsies were performed [Chang et al. 2006].

Some researchers have hypothesised that endotoxins produced by gut flora may also contribute to oxidative stress in the liver. Alteration of these flora with the probiotic VSL#3 has been shown to limit oxidative and inflammatory liver damage in animal models of NAFLD [Velayudham et al. 2009]. Loguercio and coworkers have studied the same agent in 22 patients with NAFLD and shown it to be well tolerated [Loguercio et al. 2005]. They reported beneficial affects on aminotransferases and markers of lipid peroxidation, but no biopsies were performed. Oligofructose, an indigestible fructan, reduces hepatic uptake of triacylglycerol in rats [Daubioul et al. 2000]. An 8-week double-blind crossover pilot study by Daubioul and coworkers randomized seven patients with NASH to receive oligofructose or placebo [Daubioul et al. 2005]. Insulin levels improved after 4 weeks of therapy and aminotransferases improved after 8 weeks.

Abnormal serum iron concentrations have been described in patients with NAFLD [Ruhl and Everhart, 2003]. The role of iron in the pathophysiology of NAFLD is unclear, but it may contribute to disease progression because high hepatic iron levels are a source of oxidative stress and are associated with insulin resistance [McCullough, 2006]. With this in mind, phlebotomy to achieve ‘near iron deficiency’ has been tested in nonrandomized studies and has shown improvements in serum aminotransferase levels and insulin resistance [Valenti et al. 2003; Facchini et al. 2002]. However, this intervention also needs histological validation in larger randomized controlled studies.

Cytoprotective agents

UDCA is a hydrophilic dihydroxy bile acid found in small quantities in humans. It is thought to act by competitively displacing hepatotoxic hydrophobic endogenous bile acids in the bile acid pool, and has cytoprotective actions through stabilization of cell membranes and inhibition of apoptosis [Chang et al. 2006]. An initial pilot study showed improvements in aminotransferases and steatosis [Laurin et al. 1996]. These results, however, were not confirmed in a large, multicentre, randomized placebo-controlled trial conducted by Lindor and colleagues [Lindor et al. 2004]. One hundred and sixty six patients with biopsy-proven NASH were randomized to receive either UDCA (13–15 mg/kg/day) or placebo for 2 years. One hundred and twenty six patients completed the study and 107 patients underwent posttreatment biopsies. Both the treatment group and the placebo group demonstrated improvements in transaminases and steatosis over the study duration without any significant difference between the two groups. The authors concluded that UDCA as a single agent and at the dose given was not associated with greater improvement in liver enzymes or histology compared with placebo. Durfour and colleagues, however, reported improved biochemistry and histology, mostly due to regression of hepatic steatosis, in biopsy-proven NASH patients treated with UDCA and vitamin E compared with UDCA alone or placebo [Dufour et al. 2006]. The favourable safety profile of UDCA and its proven benefit in other forms of liver disease suggest that its potential role in the treatment of NASH should be further studied, either at higher doses or in combination with other agents.

Lecithin, which has antioxidant and cytoprotective properties, increases levels of plasma free choline and reduces hepatic steatosis in long-term total parenteral nutrition patients [Buchman et al. 1992; Demetriou, 1992]. Future pilot studies using this agent should be considered.

Anti-TNF agents

Pentoxifylline, a xanthine derivative that reduces blood viscosity, has raised interest because of its ability to nonspecifically inhibit TNF-α. Two small, open-label studies have examined its safety and efficacy in patients with NASH [Adams et al. 2004; Satapathy et al. 2004]. Adams and coworkers and Satapathy and colleagues both demonstrated improvement of aminotransferases after 12 months of open-label treatment with pentoxifylline. In a further study, Satapathy and colleagues treated nine patients with biopsy proven NASH with pentoxifylline for 12 months and demonstrated, on follow-up liver biopsy, reduction in steatosis, lobular inflammation and fibrosis stage [Satapathy et al. 2007]. These histological changes were mirrored by improvements in aminotransferases. Further investigation of these preliminary findings in larger, randomized controlled studies would be helpful.

Liver transplantation

Liver transplantation may be required if cirrhosis develops and is complicated by end-stage liver failure or hepatocellular carcinoma. Currently, ~3% of all transplants in North America are performed for end-stage NAFLD, although this figure excludes patients with cryptogenic cirrhosis [Charlton et al. 2001]. Recurrence of steatosis after transplantation is common, with progression to steatohepatitis reported in one-third of cases [Charlton et al. 2001; Contos et al. 2001].

New and emerging therapies

A close relationship exists between essential hypertension and insulin resistance, with approximately half of all patients with essential hypertension being insulin resistant [Ferrannini et al. 1987]. The renin—angiotensin system (RAS) plays a crucial role in blood pressure regulation, and also has an important role in insulin resistance. Suppression of the RAS improves intracellular insulin signalling through activation of PPAR-γ, and prevents hepatic stellate cell activation, thus reducing the potential for hepatic inflammation and fibrogenesis [Georgescu, 2008]. Yokohama and colleagues treated seven patients with NASH and hypertension with the angiotensin receptor blocker (ARB) losartan at a dose of 50 mg/day for 48 weeks [Yokohama et al. 2004]. Significant improvements were noted in the levels of TGF-β1, serum markers of fibrosis and ferritin. Five patients showed reduced necroinflammation on follow-up biopsy and four patients showed reduced fibrosis. More recently, Enjoji and coworkers reported that telmisartan and olmesartan improve hepatic insulin sensitivity (HOMA-IR) and ALT levels in patients with NAFLD [Enjoji et al. 2008]. Considering the widespread use of angiotensin receptor blockers in the treatment of hypertension, further randomized controlled studies in patients with NAFLD should be considered.

In patients with T2DM glucagon-like peptide-1 (GLP-1) analogues such as exenatide and liraglutide have been shown to stimulate insulin secretion and suppress glucagon secretion in a glucose-dependent manner, delay gastric emptying and increase satiety in association with modest weight loss [Drucker and Nauck, 2006]. In animal models the GLP-1 agonist exendin-4 reduced insulin resistance, markers of oxidative stress and hepatic steatosis, suggesting that incretin based therapies may represent a novel therapeutic option for human NAFLD [Ding et al. 2006].

The short-acting insulin secretagogues repaglinide and nateglinide have also been considered as possible treatment options for NAFLD. Morita and colleagues randomized 10 patients with T2DM and biopsy-proven NASH to receive nateglinide 270 mg/day with diet and exercise or diet and exercise alone for 20 weeks [Morita et al. 2005]. In the nateglinide group improvements were noted in postprandial glucose, glycosylated hemoglobin, liver function tests and imaging and histological findings of NAFLD.

Suggested treatment algorithm for patients with nonalcoholic fatty liver disease

Despite nearly two decades of research and clinical trials, data on the efficacy and safety of pharmacotherapy in the treatment of NAFLD remains inconclusive. The majority of studies conducted have had methodological limitations such as a lack of randomization, small sample size, short follow-up duration and lack of histological data from paired liver biopsies, making it difficult to draw definitive conclusions from them.

At present, no single pharmacological agent has a license for the treatment of NAFLD. Thus, for clinicians treating patients with NAFLD a pragmatic approach should be used, tailoring therapy to meet the specific needs of the individual. A suggested algorithm for treatment is shown in Figure 1.

Figure 1.

Suggested approach to the treatment of nonalcoholic fatty liver disease.

The cornerstone of treatment for all patients with the condition remains lifestyle modification, which should comprise healthy eating, regular physical activity and reduction of alcohol intake to less than 14 units per week with at least 1–2 alcohol free days. All patients should be offered the opportunity to enrol into clinical trials. For patients with NAFLD that are overweight or obese, weight loss should be encouraged through lifestyle changes in the first instance. If these fail, and the patient has a BMI > 28 kg/m², pharmacological therapy with orlistat may be tried. For patients who have a BMI > 40 kg/m² (35 kg/m² if they also have comorbidities such as hypertension or T2DM), referral for bariatric surgery should be considered. Careful pharmacological control of cardiovascular risk factors (such as hypertension and dyslipidaemia) is mandatory in order to decrease the burden of cardiovascular disease accounting for considerable morbidity and mortality in NAFLD [Targher et al. 2008; Ekstedt et al. 2006]. More than one antihypertensive agent may be required in order to achieve a target blood pressure of less than 140/90 mmHg, and ARBs should be considered, particularly in patients with T2DM in whom they have renoprotective effects. Statins and fibrates are both now recognized to be safe in patients with NAFLD, and can be used in the primary prevention of cardiovascular disease in high-risk patients with hypercholesterolemia and hypertriglyceridemia respectively [Browning, 2006]. In patients who also have T2DM, hyper-glycaemia should be treated aggressively with dietary and lifestyle modification followed by metformin as first line pharmacological therapy. If a target glycosylated hemoglobin (HbA1c) of between 6.5% and 7.0% cannot be achieved with these measures alone, then addition of either a TZD, sulphonylurea, or incretin-based therapy should be performed. The advantage of incretin-based therapy over TZDs and sulphonylureas is that it is unlikely to promote further weight gain, and in many cases is either weight neutral or promotes weight loss.

Conclusions

NAFLD is the most common cause of chronic liver disease in the Western world today. With rising levels of obesity and T2DM, its prevalence will increase in the future, and cause considerable morbidity and mortality. Although simple steatosis carries a relatively benign prognosis, a significant percentage of patients with develop NASH which may progress to cirrhosis, end-stage liver failure and hepatocellular carcinoma.

Despite considerable research and multiple clinical trials, data on the efficacy and safety of pharmacotherapy in the treatment of NAFLD remains inconclusive, and at present no single pharmacological agent has a license for the treatment of NAFLD. The cornerstone of treatment for the condition remains lifestyle modification using a combination of diet, exercise and behavioural therapy. If pharmacological therapy is being considered, it should be tailored to the needs of the patient. In all cases, careful pharmacological control of cardiovascular risk factors (hyperglycaemia, dyslipidaemia and hypertension) is mandatory in order to decrease the burden of cardiovascular disease accounting for considerable morbidity and mortality in NAFLD [Targher et al. 2008; Ekstedt et al. 2006].

Further advances in our understanding of the pathogenesis of NAFLD, will, in the future, help develop novel therapeutic strategies to prevent and treat this increasingly common condition.