Non-alcoholic fatty liver disease (NAFLD) affects up to a third of the population in many developed countries. Between 10% and 30% of patients with NAFLD have non-alcoholic steatohepatitis (NASH) that can progress to cirrhosis. There are metabolic risk factors common to both NAFLD and cardiovascular disease, so patients with NASH have an increased risk of liver-related and cardiovascular death. Management of patients with NAFLD depends largely on the stage of disease, emphasising the importance of careful risk stratification. There are four main areas to focus on when thinking about management strategies in NAFLD: lifestyle modification, targeting the components of the metabolic syndrome, liver-directed pharmacotherapy for high risk patients and managing the complications of cirrhosis.
Statistics from Altmetric.com
Non-alcoholic fatty liver disease (NAFLD) is now the commonest cause of chronic liver disease in many developed countries.1 Up to a third of the population have evidence of steatosis on imaging,2 ,3 with the majority (70%–90%) having simple steatosis. However, 10%–30% of subjects with NAFLD have non-alcoholic steatohepatitis (NASH) that can progress to cirrhosis, which puts patients at risk of liver-related complications.4–6 Due to the metabolic risk factors that are common to both NAFLD and cardiovascular disease, patients with NASH have an increased risk of cardiovascular death as well as liver-related mortality.7 ,8 In the first part of this review, we discussed the diagnosis and staging of patients with NAFLD and the importance of risk stratification.9 Here we will discuss the management of patients with NAFLD.
Principles of management of NAFLD
Management of patients with NAFLD depends largely on the stage of disease, emphasising the importance of careful risk stratification.10 There are four main areas to focus on when thinking about management strategies in NAFLD: lifestyle modification, targeting the components of the metabolic syndrome, liver-directed pharmacotherapy for high risk patients and managing the complications of cirrhosis (figure 1).
All patients with NAFLD require advice about lifestyle modification aimed at weight loss and increased physical activity, as well as treatment of any associated metabolic risk factors (diabetes, hypertension and dyslipidaemia). For patients with simple steatosis, which carries a relatively benign liver-related prognosis, this can be done in a primary care setting. Patients with steatohepatitis and fibrosis are at highest risk of developing progressive liver disease so require more aggressive lifestyle modification and if this fails can be considered for liver-directed pharmacotherapy with pioglitazone or vitamin E. For patients who have progressed to cirrhosis, surveillance for hepatocellular carcinoma (HCC) is essential and some treatments have been shown to reduce the risk of HCC. Interventions for the treatment of NAFLD are summarised in table 1.
Lifestyle modification aimed at weight loss and increased physical activity is vital in managing all patients with NAFLD irrespective of their underlying liver histology. Patients should be encouraged to avoid a sedentary lifestyle by increasing daily activities, undertaking regular exercise and eating healthily. Overall, lifestyle interventions including behavioural, dietary and exercise changes can be very effective in reducing body weight. In one study, patients who received dietary advice and undertook 200 min moderate physical activity per week for 48 weeks had an overall 9.3% reduction in body weight (versus 0.2% in the control arm) and had improvement in hepatic steatosis and inflammation on liver biopsy.12
The optimum diet to treat NAFLD is not known. Until further evidence is available, a calorie restricted diet (600 Kcal less than a person needs to remain at the same weight) should be recommended aiming to lose 0.5–1 kg per week until the target weight is achieved.11 Patients with NAFLD should be advised to avoid saturated fats, simple carbohydrates and sweetened drinks.25 ,26 A ‘fast food diet’ (high in cholesterol, saturated fats and fructose) is associated with progressive fibrosis in mice models.27
A Mediterranean diet (high in monounsaturated fatty acids), as compared with a diet low in fat and high in carbohydrates, has been shown to reduce hepatic steatosis and improve insulin sensitivity in non-diabetic subjects with NAFLD.28 Dietary supplementation with ω-3 polyunsaturated fatty acids (n-3 PUFAs) has also been shown to decrease liver fat,29 so fish oil supplementation could be a simple therapeutic option but further studies are needed.
Intensive 12-month dietician-led lifestyle interventions are more effective than standard care for patients with NAFLD in terms of weight loss (5.6 vs 0.6 kg) and achieving remission of NAFLD (64% vs 20%).30 Previous studies have shown that >7%12 and ≥9%13 loss of body weight was associated with reduced steatosis, hepatocellular injury and hepatic inflammation. The optimum amount of weight loss to reduce fibrosis is not known, but studies in subjects undergoing bariatric surgery where larger weight losses are seen (average reduction 30% of Body Mass Index (BMI) at 5 years) indicate that fibrosis regression occurs in the majority (65%),19 although minor increases in fibrosis were seen 5 years post-bariatric surgery in 20% (>90% of these were worsening from stage 0 to 1).31 Ideally, patients should be encouraged to lose >10% (but more is probably better) of body weight and maintain the weight loss or bring their waist circumference or BMI into their ethnicity specific ‘normal’ range (eg, for Caucasians: waist circumference <94 cm for men and <80 cm for women or BMI <25 m/kg2).
People with NAFLD engage in less physical activity than their healthy counterparts32 ,33 and higher levels of habitual physical activity are associated with lower levels of steatosis.14 Aerobic exercise increases skeletal muscle insulin sensitivity and as a result reverses one of the key pathophysiological mechanisms that causes NAFLD (insulin resistance).15 ,16 Although the optimum exercise to treat NAFLD is not known, studies examining moderate intensity training, high intensity training and resistance exercise have shown improvement in liver enzymes and reduction in liver fat, independent of weight loss, but the effects on histology remain unknown.17 ,18 ,34 Therefore, all patients with NAFLD should be advised to increase physical activity and undertake regular exercise. Until further evidence is available, one approach is to recommend 30 min of moderate exercise five times weekly.11 In reality though, a significant proportion of patients with NAFLD do not comply with these recommendations. In such patients, the use of pedometers can be quite helpful. We recommend that patients increase their daily step count to >10 000 steps/day.
Orlistat as an aid to weight loss
Orlistat is an enteric lipase inhibitor that leads to malabsorption of dietary fat and can aid weight loss in subjects with obesity in conjunction with lifestyle modification. One study demonstrated that orlistat treatment for NAFLD improved alanine transaminase (ALT) and steatosis,35 but a subsequent RCT conducted by Harrison et al showed that orlistat in combination with calorie restriction and vitamin E (800 IU/day) did not enhance weight loss, improve liver enzymes or improve histopathology compared with calorie restriction and vitamin E. In that study, they found that ≥9% weight loss resulted in improved histology irrespective of whether they received orlistat or placebo.13 National Institute of Health and Care Excellence (NICE) guidelines11 for obesity recommend that orlistat can be considered as an aid to weight loss in subjects who have not achieved their target weight by lifestyle intervention and have a BMI>30 kg/m2, so it seems reasonable to consider this treatment in patients with obesity and NAFLD. If orlistat is initiated, only patients who achieve >5% loss of body weight in 3 months should continue the medication. Treatment duration should not be greater than 1 year as there is potential for fat soluble vitamin deficiency if continued for longer.
Bariatric surgery has an increasing role in the management of patients with obesity and the metabolic syndrome. Restrictive procedures such as the gastric band, gastric balloon and sleeve gastrectomy decrease stomach size leading to early satiety. Malabsorptive procedures such as gastric bypass limit the absorption of food. Weight loss after bariatric surgery has beneficial effects on the components of the metabolic syndrome including improving insulin sensitivity, lipid profile as well as reducing long-term mortality.20 ,36 It also has specific effects on liver histology including reduced steatosis, steatohepatitis and fibrosis.19 The majority of the histological benefits occur within the first year post-surgery.31 The optimum type of bariatric surgery for the treatment of NASH is not known. Although gastric bypass surgery produces the largest sustained weight loss compared with other bariatric procedures, Mathurin et al found no significant differences among the gastric band, bilio-intestinal and gastric bypass groups in terms of global NAFLD activity score (NAS), steatosis, inflammation or ballooning.31 As there is a lack of long-term outcome data about bariatric surgery as a specific treatment for NAFLD, bariatric surgery cannot be considered as a primary treatment for NASH.37 However, NICE guidance suggests that surgery should be considered as a treatment for obesity for patients with BMI >40 kg/m2 or between 35 and 40 kg/m2 with other significant disease that could be improved with weight loss. All appropriate non-surgical measures must be tried first and patients must receive intensive specialist management and be prepared to commit to long-term follow-up. Surgery can be considered as a first-line option for the treatment of obesity in adults with a BMI greater than 50 kg/m2.11
Bariatric surgery should be avoided in subjects with advanced cirrhosis with portal hypertension as there is a risk of hepatic decompensation with the rapid weight loss.38 Patients with decompensated cirrhosis have particularly high mortality rates post-bariatric surgery compared with compensated cirrhotics and non-cirrhotic patients (16.3% vs 0.9% and 0.3%, respectively, p=0.0002).39 In carefully selected patients with Child–Pugh A cirrhosis, gastric bypass and sleeve gastrectomy have been shown to achieve weight loss and improve obesity-related comorbidities.40 However, patients with cirrhosis should undergo bariatric surgery at centres that perform large numbers of these procedures.39
For patients with biopsy-proven NASH, where lifestyle intervention has failed, liver-directed pharmacotherapy with pioglitazone or vitamin E can be considered.
Several studies have demonstrated benefit from the thiazolidinediones in patients with NASH both with and without diabetes.21 ,22 ,41 Aithal et al showed that 12 months of treatment with pioglitazone 30 mg/day reduced hepatocellular injury and fibrosis compared with placebo.21 More recently, the PIVENS study showed significantly more patients receiving pioglitazone had resolution of steatohepatitis (a secondary endpoint) than with placebo (47% vs 21% p=0.001), although the treatment arm narrowly failed to meet the very strict primary endpoint (improvement in NAS ≥2 with at least 1-point improvement in hepatocyte ballooning without increase in fibrosis score).22 A recent meta-analysis has demonstrated that pioglitazone treatment in NASH significantly improves steatosis, inflammation and to a lesser degree, fibrosis.23 While pioglitazone appears to be an effective treatment for NASH, there are also some concerns about its long-term safety. Treatment with pioglitazone is associated with weight gain (4.7% in the PIVENS study)22 and there have been reports of increased risk of congestive cardiac failure,42 bladder cancer43 and reduced bone density.44 However, a meta-analysis of 16 390 patients with type 2 diabetes mellitus (T2DM) treated with pioglitazone demonstrated an 18% reduction in death, myocardial infarction and stroke.45 Therefore, after assessing risk and benefit, treatment with pioglitazone can be recommended for patients with more aggressive NASH who have failed lifestyle interventions.46 Caution should clearly be exercised in patients with cardiac disease (a common comorbidity in patients with NAFLD) and echocardiography performed prior to considering treatment with pioglitazone. The optimum dose and duration of treatment are not known.
Vitamin E is an antioxidant that has recently been shown to have beneficial effects on histology in non-diabetic patients with NASH. In the large PIVENS trial, significantly more patients had improvement in steatohepatitis following 96 weeks of vitamin E 800 IU/day compared with placebo (42% vs 19% p<0.001).22 This was confirmed in a study of childhood NASH, where vitamin E was shown to reduce steatohepatitis in a subgroup who had follow-up liver biopsies.24 Although vitamin E looks like a promising treatment for NASH, there are concerns about the long-term effects of vitamin E with a meta-analysis showing a small overall increase in all-cause mortality at doses >400 IU/day.47 It has also been reported that there might be an increased risk of haemorrhagic stroke48 and prostate cancer49 with high dose treatment. Vitamin E has not been evaluated in patients with cirrhosis or diabetes and NASH. Therefore, it should currently be reserved for selected patients with more advanced pre-cirrhotic NASH who have failed lifestyle interventions. The optimum dose and duration of treatment are not known.
Targeting components of the metabolic syndrome
Management of diabetes mellitus in patients with NAFLD
Approximately 40%–50% of patients with NAFLD attending secondary care clinics have T2DM and the majority have evidence of insulin resistance.50 ,51 T2DM is associated with more progressive fibrosis in NASH. It is therefore essential that all patients with NAFLD are regularly screened for diabetes. Once diagnosed, diabetes should be managed initially with dietary intervention. Although metformin has not been shown to improve histology in NASH, it is recommended as the first-line pharmacological treatment for T2DM as it aids weight loss and reduces the risk of any diabetes-related endpoint, microvascular disease, myocardial infarction (large vessel disease) and all-cause mortality.52–54
Treatment of diabetes should be escalated if the glycosylated haemoglobin (HbA1c) rises to >7.5%. NICE guidelines suggest a sulfonylurea (such as gliclazide) as second-line treatment for subjects with T2DM.52 However, sulfonylureas increase insulin secretion and can cause weight gain, which might be detrimental in NAFLD. Therefore, the insulin sensitiser, pioglitazone, should be considered as a second-line agent in subjects with NASH as studies have demonstrated its potential to improve insulin sensitivity as well as reduce hepatic steatosis and inflammation in subjects with NASH and T2DM.22 ,41 However, the potential risks of bladder cancer, increased bone loss and cardiac failure need to be considered.
Glucagon-like peptide-1 (GLP-1) analogues (such as liraglutide or exenatide) should be considered as an alternative to insulin as a third-line agent to control diabetes where the HbA1c is >7.5% in subjects with obesity. GLP-1 is an incretin secreted by ileal L-cells dependent on the presence of food in the small intestine. It increases insulin sensitivity, inhibits gastric emptying and decreases food intake by increasing satiety in the brain.55 On average, they reduce HbA1c by 1% and patients frequently lose weight (overall average 3 kg). In a meta-analysis of the phase III studies of liraglutide for the treatment of type 2 diabetes (n=4442), 26 weeks treatment with liraglutide 1.8 mg was associated with an improvement in liver enzymes as well as a trend towards reduced steatosis measured by CT.56 This effect appeared to be mediated by its effect on weight loss and glycaemic control. Studies are underway to determine the effect of liraglutide on liver histology. However, as with all the antidiabetic drugs, a careful assessment of the risk and benefits should be undertaken as there is recent evidence to suggest that GLP-1 agonists might induce pancreatitis (particularly in those with very high triglycerides) and increase the potential for pancreatic cancer.55
Table 2 summarises the treatment of type 2 diabetes and the metabolic syndrome in NAFLD.
Approximately 70% of patients with NAFLD have hypertension59 ,60 and so it is imperative to screen all patients. If the blood pressure is >140/90 mm Hg, patients should be managed according to NICE hypertension guidelines.61 ACE inhibitors (ACEIs) or angiotensin receptor blockers (ARBs) are recommended first-line for patients less than 55 years old, but calcium channel blockers are recommended in people over 55 years and those of Afro-Caribbean origin. However, targeting the renin–angiotensin system (RAS) might be beneficial in all patients with NAFLD as the RAS plays a role in liver fibrogenesis and blocking the RAS reduces fibrosis in experimental models of hepatic fibrosis.62 ,63 A previous study of 54 hypertensive patients with NASH demonstrated that both telmisartan 20 mg and valsartan 80 mg improved transaminase levels and insulin sensitivity, and telmisartan also showed a significant decrease in NASH activity score and fibrosis.64 Further larger studies are needed to confirm these observations, but ACEI or ARBs seem a logical choice in the management of hypertension in subjects with NAFLD. They may also have the added benefit of reducing the incidence of T2DM. A previous meta-analysis of more than 100 000 patients also showed a 20% reduction in the incidence of new onset diabetes with the use of ACEI or ARBs.65
Dyslipidaemia is very common in patients with NAFLD and the metabolic syndrome. Effective treatment of dyslipidaemias is vital in the management of NAFLD to reduce patients’ cardiovascular risk profile. NICE guidelines66 recommend simvastatin treatment as primary prevention for cardiovascular disease in adults with a ≥20% 10-year risk of developing cardiovascular disease (using a risk calculator such as the Framingham risk calculator). More aggressive lipid control is recommended for secondary prevention of cardiovascular disease, aiming for a total cholesterol <4 mmol/L. A meta-analysis confirmed that statin therapy safely reduces the 5-year incidence of all-cause mortality, major coronary events, coronary revascularisation and stroke by about 20% per mmol/L reduction in LDL cholesterol.67
Raised transaminases are frequently seen in patients receiving statins (usually due to coexistent NAFLD) but they are safe in patients with liver disease, and serious liver injury is rarely seen in clinical practice.46 In a large cardiovascular outcomes study, statins were shown to improve liver enzymes and cardiovascular outcomes in patients with raised liver function tests (LFTs) due to NAFLD.68
Table 3 summarises the management of hypertension and dyslipidaemia in NAFLD.
Obstructive sleep apnoea
Due to the high prevalence of obesity in subjects with NAFLD, obstructive sleep apnoea (OSA) is very common and often unrecognised. OSA is associated with an increased risk of NAFLD as well as more advanced fibrosis due to NASH.70 Clinicians should have a high index of suspicion for OSA and can screen patients in the clinic using the STOP BANG questionnaire that encompasses eight factors: Snoring, Tiredness, Observed apnoeas, Elevated blood pressure, BMI (>35 kg/m2), Age (>50 years), Neck circumference (>40 cm) and Gender (male). Patients who score ≥3 points (a point is given for each positive risk factor) have a high likelihood of having OSA. This questionnaire has 83.6%, 92.9% and 100% sensitivity for mild, moderate and severe sleep apnoea, respectively.71 It is not known whether treatment of OSA slows or halts progression of fibrosis in NASH.
Managing the complications of cirrhosis
Patients with NASH cirrhosis are at risk of the same complications of cirrhosis as with any other aetiology of liver disease. The yearly cumulative incidence of HCC is 2.6% per year in patients with NASH cirrhosis.72 Therefore, surveillance with abdominal ultrasonography (USS) should be performed every 6 months.73 ,74 In unselected cirrhotic patients, USS has a sensitivity ranging from 58% to 89% with a specificity >90% when used as a surveillance test.75 ,76 Ultrasound can be technically difficult in obese patients so it is likely that ultrasound is less sensitive as a surveillance modality in patients with NAFLD, but currently there are no data to support the routine use of CT or MRI for surveillance.74 CT or MRI could be considered in subjects where ultrasound has failed to produce adequate examination of the liver, but these modalities have their own limitations in obese patients.
α-Fetoprotein is a tumour biomarker and has recently been removed from the European guidelines but is still widely used in clinical practise alongside 6-monthly ultrasound. Although rare, HCC can occur on a background of non-cirrhotic NAFLD.77
There is a clear association between obesity and malignancy78 and hyperinsulinaemia is well recognised as an independent risk factor for many malignancies.79 ,80 Aberrant genes involved in metabolic pathways, such as the AMP-activated protein kinase (AMPK)–LKB1 pathway, are emerging as therapeutic targets in cancer treatment.81 Metformin acts by inhibiting hepatic glucose production through an LKB1–AMPK-mediated mechanism. There is emerging evidence that metformin reduces the risk of cancer (including HCC) in patients with diabetes in a dose-dependent manner.57 The mechanisms of the anticancer effects of metformin may include triggering of the LKB1-mediated AMPK pathway, resulting in cancer cell death under conditions of nutrient deprivation,82 ,83 and reducing proliferation of cancer cell lines.83
Donadon et al found that metformin treatment was associated with a statistically significant reduction of HCC risk (OR 0.33; CI 0.1 to 0.7, p=0.006) in diabetic patients.84 The benefit of metformin was confirmed by Hassan et al who found the use of metformin or thiazolidinediones in diabetic patients was associated with a 70% risk reduction of HCC compared with the use of insulin or sulfonylureas.85 A recent large Taiwanese nationwide case-control study demonstrated a 7% risk reduction per year for HCC in diabetics treated with metformin (adjusted OR=0.93, 95% CI 0.91 to 0.94, p<0.0001).58
There is also evidence that statins might reduce the incidence of HCC. An American case-control study examining HCC in diabetic patients found an adjusted OR for statin use of 0.74 (95% CI 0.64 to 0.87). There were no significant associations with non-statin cholesterol- or triglyceride-lowering medications.86 A Taiwanese population-based case-control study found an adjusted OR of 0.62 (95% CI 0.42 to 0.91) for HCC in patients prescribed statins.87 Singh et al conducted a systematic review and meta-analysis and found that statin users were less likely to develop HCC than non-users (adjusted OR, 0.63; 95% CI 0.52 to 0.76). The risk reduction was most marked in Asian populations but was also observed in Western populations.69
Statins appear to have anticancer effects through both 3-hydroxy-3-methyl-glutaryl-CoA (HMG-CoA) reductase dependent and HMG-CoA reductase independent pathways. They have a variety of actions including: anti-inflammatory and immunomodulatory effects, inhibiting downstream production of mediators of cell growth88 and promoting programmed cell death.89 Studies have shown that inactivation of Myc can induce remission of HCC and atorvastatin blocks Myc activation resulting in suppression of tumour growth.90 ,91
Cirrhotic patients with NASH are at risk of varices as with other liver diseases and their presence correlates with the severity of liver disease (40%–44% if Child–Pugh grade A, 75%–85% if Child–Pugh grade C).92 Therefore, patients with NASH cirrhosis should undergo endoscopic screening for oesophageal and gastric varices at the time of diagnosis and at regular intervals according to British Society of Gastroenterology or American Association for the Study of Liver Diseases guidelines93 ,94 (see table 4).
Osteoporosis and fracture risk
Patients with cirrhosis are at increased risk of osteoporosis and fractures. Cirrhotic patients should have a dual x-ray absorptiometry scan to assess bone mineral density and have their fracture risk assessed using the FRAX tool (http://www.shef.ac.uk/FRAX/tool). The FRAX tool has an easy online link to the National Osteoporosis Guideline Group website which provides management advice according to the patient's risk of fracture. Table 4 summarises the management of the complications of NASH cirrhosis.
NASH cirrhosis is now the third commonest indication for liver transplantation in the USA and accounted for 12% of patients listed for transplantation in the UK in 2009.96 Patient and graft survival in liver transplantation for NASH are comparable with other indications.96 ,97 Although recurrence of NASH is common post-transplant, occurring in 4%–25% of patients, it does not appear to impact on graft survival.98 ,99 Patients who are transplanted for NASH cirrhosis frequently have multiple cardiovascular risk factors that should be managed aggressively post-transplant to prevent cardiovascular-related mortality.100 UK guidelines on liver transplantation suggest that bariatric surgery can be considered at the time of liver transplantation for patients with NASH and morbid obesity.100
Lifestyle interventions aimed at weight loss and increased activity are essential for all patients with NAFLD and if sustained are effective in the treatment of NAFLD. For patients who fail lifestyle intervention, liver-directed pharmacotherapy with pioglitazone or vitamin E can be considered for those with advanced, but pre-cirrhotic, NASH. Features of the metabolic syndrome and cardiovascular risk factors are very common in NAFLD, so all patients should be screened for these and have them managed aggressively. For patients who develop advanced disease, they require surveillance for and management of the complications of cirrhosis (HCC, varices, osteoporosis).
Contributors JD, QA and SM have all contributed to this review.
Competing interests None.
Provenance and peer review Not commissioned; externally peer reviewed.
If you wish to reuse any or all of this article please use the link below which will take you to the Copyright Clearance Center’s RightsLink service. You will be able to get a quick price and instant permission to reuse the content in many different ways.