Current role of fenofibrate in the prevention and management of non-alcoholic fatty liver disease

Abstract

Non-alcoholic fatty liver disease (NAFLD) is a common health problem with a high mortality burden due to its liver- and vascular-specific complications. It is associated with obesity, high-fat diet as well as with type 2 diabetes mellitus (T2DM) and metabolic syndrome (MetS). Impaired hepatic fatty acid (FA) turnover together with insulin resistance are key players in NAFLD pathogenesis. Peroxisome proliferator-activated receptors (PPARs) are involved in lipid and glucose metabolic pathways. The novel concept is that the activation of the PPARα subunit may protect from liver steatosis. Fenofibrate, by activating PPARα, effectively improves the atherogenic lipid profile associated with T2DM and MetS. Experimental evidence suggested various protective effects of the drug against liver steatosis. Namely, fenofibrate-related PPARα activation may enhance the expression of genes promoting hepatic FA β-oxidation. Furthermore, fenofibrate reduces hepatic insulin resistance. It also inhibits the expression of inflammatory mediators involved in non-alcoholic steatohepatitis pathogenesis. These include tumor necrosis factor-α, intercellular cell adhesion molecule-1, vascular cell adhesion molecule-1 and monocyte chemoattractant protein-1. Consequently, fenofibrate can limit hepatic macrophage infiltration. Other liver-protective effects include decreased oxidative stress and improved liver microvasculature function. Experimental studies showed that fenofibrate can limit liver steatosis associated with high-fat diet, T2DM and obesity-related insulin resistance. Few studies showed that these benefits are also relevant even in the clinical setting. However, these have certain limitations. Namely, these were uncontrolled, their sample size was small, fenofibrate was used as a part of multifactorial approach, while histological data were absent. In this context, there is a need for large prospective studies, including proper control groups and full assessment of liver histology.

Key Words: Fenofibrate, Non-alcoholic fatty liver disease, Steatohepatitis, Peroxisome proliferator-activated receptors

Core tip: Non-alcoholic fatty liver disease (NAFLD) is a common health problem associated with increased liver- and vascular-specific complications. Dyslipidemia, predominantly hypertriglyceridemia, and insulin resistance play a key role in its pathogenesis. Fenofibrate, by activating peroxisome proliferator-activated receptors appears to decrease liver steatosis in experimental animal studies. This benefit can be attributed to its lipid-lowering potency, together with anti-inflammatory and anti-oxidant actions. Also, fenofibrate increases adiponectin levels and the expression of its liver-active receptor. A potential protective role of fenofibrate against NAFLD has also been implied by few small clinical studies. However, this benefit should be further assessed.

INTRODUCTION

Non-alcoholic fatty liver disease (NAFLD) is a cluster of liver disorders associated with hepatic lipid accumulation (steatosis) in the absence of viral hepatitis or alcohol abuse[1]. These include a histological spectrum ranging from steatosis alone to non-alcoholic steatohepatitis (NASH)[1]. In NASH, beyond lipid accumulation, necroinflammation and fibrosis exist[2,3]. Approximately 29% of NASH patients will develop cirrhosis within 10 years[4]. End-stage liver disease and hepatocellular carcinoma are liver-specific endpoints of NAFLD[3-5].

NAFLD is a common health problem affecting up to 35% of the population in several countries[5]. It is estimated that 19% of the United States adult population, corresponding to 28.8 million individuals, exhibit ultrasonographic findings of NAFLD[6]. Increased body weight considerably increases the risk of this abnormality. Among 257 Italian individuals obesity was associated with a 4.6-fold increased risk of hepatic steatosis[7]. Considering current obesity epidemic it is expected that NAFLD prevalence will rise.

NAFLD is typically asymptomatic[8]. Non-specific complaints include fatigue, malaise and right upper quadrant discomfort[8]. Elevation of aminotransferase activities, especially of alanine aminotransferase (ALT) and γ-glutamyltranspeptidase (γGT), are markers of hepatocellular damage[9,10]. Lipid accumulation in the liver is identified by non-invasive imaging techniques, including ultrasound and magnetic resonance imaging (MRI)[9]. However, these techniques cannot discriminate between simple liver steatosis and NASH. In this context, liver biopsy remains the ‘‘gold standard’’[11].

From a pathophysiological viewpoint, NAFLD is associated with imbalanced influx vs removal of triglycerides (TG) in the liver[1]. In this context, TG accumulation > 55 mg/g measured by MRI or by histological examination is its key diagnostic feature[12,13]. Fatty acids (FA) account for approximately 60% of TG in the liver of NAFLD patients, while approximately 15% originate from dietary fat[14]. De novo production is responsible for the rest 25%[14].

Insulin resistance plays a key role in the pathogenesis of NAFLD[2,15,16]. This can be attributed to enhanced hepatic FA flux and uptake[2,15,16]. Insulin resistance results in hyperinsulinemia and increased circulating levels of free FA, which enter hepatocyte cytoplasm to create TG[2,15,16]. Furthermore, high plasma insulin and glucose levels might stimulate transcription factors associated with enhanced hepatic lipogenesis[17-19]. In the clinical setting, these abnormalities are mirrored by increased circulating concentration of TG-rich lipoproteins and hypertriglyceridemia. NAFLD is commonly noted in insulin resistant states, including type 2 diabetes (T2DM) and metabolic syndrome (MetS)[20-27]. Interestingly, it was suggested that raised activities of ALT or γGT may predict future development of MetS or T2DM[20-24]. In this context, NAFLD has been considered as the ‘‘hepatic component of MetS’’.

It was suggested that NAFLD increases the risk of cardiovascular (CV) events[20,28,29]. In prospective studies, raised serum activity of liver enzymes independently predicted CV events and/or total and CV mortality[20]. For example, γGT activity was raised in 163 patients admitted with an acute ischemic/non-embolic stroke compared with 166 healthy individuals[30]. Interestingly, patients at the highest quartile for γGT activity had a 4.7-fold higher risk of ischemic stroke than those at the lowest quartile[30]. This association was relevant after adjustment for the presence of established CV risk factors[30]. In accordance with these findings, elevated ALT activity was associated with increased CV- and diabetes-related mortality in 37085 Korean subjects[31]. Furthermore, NAFLD patients may exhibit enhanced subclinical atherosclerosis compared with non-steatosic individuals[28]. This may be explained at least in part by the coexistence of NAFLD with an atherogenic risk profile characterized by hyperlipidemia, dysglycemia and hypertension[32]. Also, in NAFLD the liver overproduces various atherogenic factors, including inflammatory cytokines, coagulation factors and molecules that increase blood pressure[32].

To date, an established treatment of NAFLD is gradual weight loss. It was shown that dietary intervention or bariatric surgery improved liver function tests and liver histology in patients with NASH[33-35]. Also, weight reduction by orlistat might be useful[36,37]. Among 50 overweight subjects dietary intervention together with vitamin E and orlistat achieved significant weight loss[37]. Reduction ≥ 5 and 9% was associated with improved insulin resistance and hepatic histological findings, respectively[37]. Also, vitamin E as a potent antioxidant can improve liver histology in non-diabetic NASH. In this context, it is considered a first-line pharmacotherapy for this patient population[38]. Furthermore, pioglitazone can be useful for patients with biopsy-proven NASH[38]. However, it should be acknowledged that most patients on pioglitazone in clinical trials were non-diabetic. Also, long-term safety and efficacy has not been evaluated[38].

NAFLD AND LIPID-LOWERING DRUGS

Interest is increasing regarding the effect of lipid-lowering drugs on NAFLD. Long-term statin treatment has been associated with significant decreases or even normalization of serum aminotransferase activities in patients with NAFLD and dyslipidemia[39-41]. Liver steatosis assessed by either imaging techniques or biopsy was also diminished[40,41]. Interestingly, this benefit appears to be associated with greater vascular risk reduction. The Greek Atorvastatin and Coronary Heart Disease Evaluation study included 1600 patients with established coronary heart disease (CHD). These were randomized to atorvastatin or ‘‘usual’’ medical care to achieve low density lipoprotein cholesterol (LDL-C) goal < 100 mg/dL[42]. Atorvastatin was more effective in reducing total and CV mortality as well as coronary morbidity compared with ‘‘usual’’ care[42]. A post hoc analysis of this study included 437 patients with moderately abnormal liver function tests at baseline possibly associated with NAFLD[43]. NAFLD was assumed in patients with moderately elevated aminotransferase activities (< 3x the upper limit of normal) together with relevant ultrasonographic findings, after excluding other causes of abnormal liver function tests[43]. Statin (mainly atorvastatin) treatment was associated with substantial improvements of aminotransferase activities, whereas non-statin use with further increases[43]. Interestingly, among patients with abnormal liver function tests those who received a statin experienced a greater reduction of CV events (69% relative risk reduction, P < 0.0001) compared with those who did not receive a statin[43]. Furthermore, among statin-treated patients, those with abnormal liver function tests had fewer CV events compared with those with normal liver function tests (39% relative risk reduction, P < 0.0001)[43]. However, these promising findings have limitations including the post hoc analysis and small number of patients. Furthermore, liver biopsy was not performed.

Ezetimibe antagonizes cholesterol absorption by inhibiting Niemann-Pick C1 like-1 protein (NPC1L1)[44]. This protein is expressed by both enterocytes and hepatocytes. It was suggested that ezetimibe may be useful for the management of NAFLD by inhibiting hepatic cholesterol accumulation[45]. This can be better achieved by combinations of ezetimibe with drugs facilitating weight loss or enhancing insulin sensitivity[45].

Fibrates are first-line drugs for reducing TG levels. In this context, they are commonly used for correcting lipid abnormalities in obese patients with MetS and T2DM[46-48]. Their hypolipidemic action is attributed to activation of the peroxisome proliferator-activated receptors (PPAR), particularly PPARα[49]. PPARs control the transcription of genes regulating lipid and glucose metabolism[50]. These receptors may also modulate hepatic lipid homeostasis, inflammation and fibrosis, by directing the proliferative and inflammatory response of specific cell types[50].

PPARα isotype is highly expressed in metabolically active tissues, including the liver, muscle, intestine and brown adipose tissue[51]. PPARα is predominantly expressed by hepatocytes and decreases hepatic lipid accumulation[50,51]. This is mostly attributed to a regulation of the FA transport and β-oxidative degradation. PPARα also controls inflammatory responses by inhibiting inflammatory gene expression induced by nuclear factor kappa B (NF-κB)[50,51]. Furthermore, it can limit interleukin (IL)-1-associated C-reactive protein expression[50,52]. In this regard, it has been shown that PPARα deficient mice are susceptible to hepatic steatosis and NASH[53-55].

The current concept is that PPARα activation may prevent these abnormalities[50,53]. Interestingly, PPARα agonism reversed steatohepatitis in mice, suggesting a potential curative role against NASH[56]. This benefit might be attributed to a downregulated expression of inflammatory genes[55]. Also, PPARα activation may reverse fibrosis by reducing the expression of fibrotic markers and the number of stellate cells[56].

FENOFIBRATE AND NAFLD

Fenofibrate is one of the most used fibrates. This drug alone or in combination with statins improves the atherogenic serum lipid profile, by significantly reducing TG, while raising HDL-C levels[57-59]. Also, it appears to exert anti-inflammatory and anti-thrombotic actions, while improving endothelial function, particularly in patients with MetS and T2DM[47,60-63]. In this context, large clinical trials suggest that fenofibrate declines atherosclerosis progression and the risk of vascular events in patients with T2DM and dyslipidemia[64,65]. However, this benefit should be further assessed.

Interestingly, fenofibrate may improve insulin sensitivity by limiting lipid accumulation in several tissues, including the liver and muscles[66-69]. This can also be attributed to increased adiponectin together with reduced expression and plasma levels of several other adipokines, including tumor necrosis factor-α (TNF-α), leptin, resistin and plasminogen activator inhibitor (PAI)-1[66,70,71]. Considering its hypolipidemic and insulin-sensitizing actions it could be assumed that fenofibrate is useful for the prevention and management of NAFLD. Herein, we discuss the role of fenofibrate as a potential treatment option for NAFLD.

Mechanistic implications

Animal studies suggested a protective role of fenofibrate against NAFLD providing explanatory mechanisms. Fenofibrate prevented from high-fat diet-induced hepatic TG accumulation[72-75]. Consequently, all histological findings of NAFLD, including hepatic steatosis, necroinflammation and collagen deposition, were reversed[72-74]. These benefits were associated with its lipid-lowering together with anti-inflammatory properties. Namely, fenofibrate prevented from diet-associated weight gain and increases in circulating TG and free FA[72,73]. It was suggested that PPARα activation by fenofibrate enhances hepatic FA turnover. Namely, it increased mRNA expression of FA β-oxidation enzymes in obese rats with T2DM[76]. These include FA transport protein, FA binding protein, carnitine palmitoyltransferase II, as well as medium- and long-chain acyl-CoA dehydrogenase and acyl-CoA oxidase[76,77].

A high-fat diet-associated increase in the liver inflammatory gene expression may be ameliorated by fenofibrate[74]. Importantly, this effect was relevant immediately after treatment initiation, before liver steatosis occur. This finding implies a potential protective role of fenofibrate against liver inflammation resulting in NASH[74]. TNF-α plays a key role in NASH pathogenesis. Its plasma levels can be decreased by fenofibrate[72]. It appears that TNF-α hepatic expression is reduced by PPARα activation[73]. Macrophage infiltration of the liver may be also limited[72,74]. Inhibition of the liver expression of monocyte chemoattractant protein (MCP)-1, intercellular adhesion molecule (ICAM)-1 and vascular adhesion molecule (VCAM)-1 may help explain this benefit[75]. This action is PPARα-dependent[75]. Anti-inflammatory properties of fenofibrate imply a protective effect against NASH.

Anti-oxidant actions may account for anti-steatosic effects of fenofibrate on the liver[73,78]. Fenofibrate reduced hepatic steatosis in mice developing hereditary NAFLD without obesity[78]. Increased expression of genes facilitating FA turnover, while reducing hepatic lipid peroxidation, were mechanisms explaining this benefit[78].

Reducing hepatic insulin resistance may also account for protective effects of fenofibrate against NAFLD[72,73,76,79]. This may be mediated by enhanced FA β-oxidation together with eliminated accumulation of diacylglycerols, which have an impact on insulin signaling[79]. In this context, fenofibrate improved liver steatosis in animal models of obesity-related T2DM and hepatic insulin resistance[76,79]. Likewise, in a NASH animal model with obesity, dyslipidemia and insulin resistance fenofibrate improved insulin sensitivity and hepatic morphology, while decreasing ALT activity[77]. In this regard fenofibrate was more effective than rosiglitazone[77]. Therefore, PPARα might be preferred over PPARγ activation for treating insulin resistance-associated NASH.

High-fat diet may adversely affect hepatic microvasculature by narrowing sinusoids and reducing hepatic microcirculatory perfusion[80]. Consequently, oxygenation of portal venules may be disturbed. It was suggested that these abnormalities promote the development of NAFLD[80]. It was suggested that PPARα agonists exert beneficial effects on the microcirculation of several tissues, including the retina, kidney and nerves[80]. It appears that PPARα activation inhibits various mediators of vascular damage, including lipotoxicity, inflammation, reactive oxygen species generation, endothelial dysfunction and thrombosis[80]. Also, it can influence intracellular signalling pathways associated with microvascular complications[80]. In this context, fenofibrate was associated with a slower progression of retinopathy and albuminuria in the clinical setting of T2DM[64,80-82]. It was shown that fenofibrate exerts beneficial effects on the liver microvascular environment and oxygen metabolism. Namely, it remarkably improved microvascular patency and tissue oxygenation in high-fat diet-induced NAFLD mice[83]. These findings imply a potential protective effect of the drug against T2DM-related hepatic steatosis.

Another experimental study investigated a cross-link between anti-steatosic and anti-atherosclerotic effects of fenofibrate[84]. Microparticles are small membrane vesicles produced by activated and apoptotic cells, being not only biomarkers, but also functional actors in NAFLD and atherosclerosis. In mice with atherosclerosis and NAFLD, fed with Western diet, the expression of microparticles was increased in atherosclerotic lesions and the liver[84]. Fenofibrate was associated with reduced expression of microparticles in atherosclerotic lesions, but not in the liver[84]. Therefore, limiting microparticle expression might not help explain the protective role of fenofibrate against NAFLD.

Adiponectin is an adipokine with various functions associated with insulin sensitivity and inflammation[85-87]. Reduced adiponectin levels have been associated with increased insulin resistance and risk of vascular events[85]. Patients with MetS and/or T2DM have low circulating adiponectin levels[86,87]. These are also decreased in patients with NAFLD, possibly due to enhanced hepatic insulin resistance together with declined FA β-oxidation[88,89]. In contrast, recombinant adiponectin administration exerted protective effects against NAFLD in mice[90]. These were attributed to increased β-oxidation and limited hepatic synthesis of FA[90]. Also, liver production and plasma levels of TNF-α may be reduced[90]. Adiponectin downregulated aldehyde oxidase 1 in vivo[91]. This enzyme produces reactive oxygen species that promote cell damage and fibrogenesis[91]. Its activity is high in obesity-related hepatic steatosis[91]. Adiponectin also inhibited hepatic fibrosis by downregulating connective tissue growth factor in vitro[92]. This molecule promotes liver fibrosis by activating transforming growth factor (TGF) β[92]. PPARα activation appears to play a key role in these benefits of adiponectin; thus fenofibrate exhibited the same properties[91,92]. Also, clinical and experimental studies showed that fenofibrate increases adiponectin plasma levels[93,94]. This was associated with vascular benefits of the drug and the rise in HDL-C levels[93,94].

Except for adiponectin, its liver-specific R2 receptor (AdipoR2) appears to play a role in steatosis and inflammation. At the cellular/molecular level AdipoR2 mRNA was reduced in liver samples of patients with NASH[95,96]. It was suggested that AdipoR2 may be protective against steatosis-related hepatic insulin resistance[97]. In contrast, impairment of its expression was associated with decreases PPARα signaling pathways[97]. FA load and endoplasmic reticulum stress decreased AdipoR2 levels in vitro[98]. Fenofibrate preserved AdipoR2 levels, while preventing from TG accumulation and endoplasmic reticulum stress[98].

Clinical evidence

Few clinical studies assessed the effect of fenofibrate on biochemical and imaging surrogates of NAFLD. A study included 186 patients with MetS and both biochemical and ultrasonographic evidence of NAFLD[99]. These received lifestyle advice and treatment for hypertension, impaired fasting glucose (metformin) and obesity (orlistat)[99]. For the management of dyslipidemia study participants were randomized to atorvastatin 20 mg/d or micronized fenofibrate 200 mg/d monotherapy or their combination. After 54 wk the percentage of patients having no longer biochemical or ultrasonographic evidence of NAFLD was 67%, 42% and 70% for atorvastatin, fenofibrate and combination, respectively[99]. Interestingly, each treatment option was independently associated with this benefit. Other variables independently predicting hepatic steatosis elimination included corrections of low-grade inflammation (high-sensitivity C-reactive protein), anthropometric variables (waist circumference and body weight), the serum lipid profile (TG, LDL-C and total cholesterol levels), systolic blood pressure and glucose levels[99]. This study highlighted a potential role of multifactorial treatment, including fenofibrate, in reducing hepatic steatosis associated with MetS. However, these results should be considered under certain limitations. For example, no placebo group was included. Furthermore, no biopsy, which is the ‘‘gold standard’’ for NAFLD diagnosis and staging, was performed[99]. Another small study included 15 patients with T2DM randomized to fenofibrate or pioglitazone[100]. Pioglitazone significantly improved glucose homeostasis and reduced fasting TG and free fatty acid (FFA) concentrations. These changes were associated with decreased hepatic fat content assessed by MRI[100]. However, these changes were not relevant in fenofibrate-treated patients. Adding pioglitazone on fenofibrate decreased insulin resistance as well as FFA and TG levels, thereby reducing hepatic fat content[100]. This finding implies the efficacy of multifactorial treatment in restricting hepatic steatosis in T2DM, as well. In contrast, adding fenofibrate on top of pioglitazone failed to significantly decrease insulin resistance, while lowering circulating TG and FFA levels. These changes were not associated with significantly reduced hepatic steatosis[100]. Considering these, improving insulin resistance by PPARγ activation might be preferred over reducing TG levels for the management of T2DM-related NAFLD.

Another study included 16 patients with biopsy-confirmed NAFLD. These were treated with fenofibrate (200 mg/d) for 48 wk[101]. Fenofibrate was associated with significant improvement of the serum lipid profile and insulin sensitivity. Alkaline phosphatase and γGT activities were significantly reduced[101]. The proportion of patients with abnormal aminotransferase activities (> 45 IU/L) was also decreased: 93.7% at baseline vs 62.5% following treatment for ALT; 50% at baseline vs 18.7% following treatment for AST[101]. Interestingly, a control liver biopsy at the end of the study showed a significantly decreased grade of hepatocellular ballooning degeneration compared with baseline. However, the grade of steatosis, lobular inflammation, fibrosis and NAFLD activity score did not change significantly[101].

Data from large clinical trials are lacking. This may be attributed to the exclusion of patients with chronic liver disease due to transient increases in aminotransferase activities associated with fibrate treatment[102]. Although these elevations occur without any other signal of hepatotoxicity, moderate increases in liver function tests are an indication of adverse effect reaction in large clinical trials. This was also relevant for fenofibrate trials, including the Fenofibrate Intervention and Event Lowering in Diabetes study[64]. Also, it was suggested that aminotransferases may be direct PPARα target genes[103,104]. This implies that fibrate-induced elevations in plasma aminotransferase activities are a mechanism-related treatment effect. In this context, aminotransferase activities might be problematic as surrogates of NAFLD in studies assessing drug effects on liver steatosis.

CONCLUSION

NAFLD is a common health problem associated with increased liver-specific as well as CV morbidity and mortality. Impaired FA turnover, often associated with insulin resistance, is its pathophysiological hallmark. In the presence of inflammation hepatic steatosis can progress to NASH and eventually cirrhosis. Weight loss, vitamin E and pioglitazone can be useful for the management of this abnormality, while the role of lipid-lowering drugs is being investigated.

PPARα plays a role in the pathogenesis of NAFLD by regulating lipid and glucose metabolic pathways. The novel concept is that PPARα activation may be protective and therapeutic against NAFLD. Experimental data suggested such a role of fenofibrate in the setting of high-fat diet, obesity, insulin resistance and T2DM. An improved FA turnover can help explain this benefit. Indeed, fenofibrate appears to enhance the expression of genes promoting FA β-oxidation. Also, its anti-inflammatory together with anti-oxidant actions may prevent from NASH-related necroinflammation, apoptosis and fibrosis. These are attributed to inhibited expression of inflammatory mediators, including TNF-α, MCP-1,VCAM-1 and ICAM-1, together with reduced lipid peroxidation and reactive oxygen species formation. Also, it was suggested that insulin resistance is also improved by fenofibrate. All these effects appear to be PPARα-dependent. Furthermore, fenofibrate increases the expression and plasma levels of adiponectin, while preserving its liver-active receptor. Beyond insulin-sensitizing effects this adipokine enhances FA hepatic β-oxidation and exerts various anti-inflammatory and anti-fibrotic effects on the liver.

Data are inconclusive regarding the effect of fenofibrate on NAFLD surrogates in the clinical setting. Fenofibrate treatment as a part of multifactorial approach may be useful in MetS and T2DM. In this regard, insulin-sensitizing may be more important than lipid-lowering effects of the drug. However, this benefit should be further established by histological studies. Other limitations of available clinical studies include small sample size and the use of fenofibrate in combination with other strategies. In this context, there is a need for large prospective studies, including proper control groups and full assessment of liver histology.