| |
Liver Disease, Fatty Liver in HIV
|
| |
| |
NASH, also known as fatty liver, is associated with liver disease. The confluence of insults on the liver can result in serious liver disease. Insults on the liver can include HCV, HBV, fatty liver, concomitant medications.
Non-alcoholic steatohepatitis (NASH): From cell biology to clinical practice
Journal of Hepatology Jan 2006
Helena Cortez-Pintoab, Miguel Carneiro de Mouraa, Christopher Paul Dayc
a Centre of Gastroenterology, Institute of Molecular Medicine (IMM), Faculty of Medicine, University of Lisbon, Lisbon, Portugal
b Centro de Nutrio e Metabolismo, Institute of Molecular Medicine (IMM), Faculty of Medicine, University of Lisbon, Lisbon, Portugal
c School of Clinical Medical Sciences, University of Newcastle upon Tyne, UK
1. Clinicopathological features
1.1. Epidemiology
The reported prevalence of NAFLD ranges from 10 to 51% depending on the population studied and the methodology used (clinical series, imaging or autopsy studies, and general population screening) [1-3] with a consensus figure of 20-30% for Western countries [4]. Importantly, a recent study in the Dionysos population residents in Italy [5] has reported that the prevalence of NAFLD (defined according to current definitions [6,7]) is the same in subjects with suspected liver disease, defined on the basis of an elevated serum ALT or GGT or HBsAg or HCV-RNA positivity, as it is in controls (25 vs. 20%, P=0.203). This explains why studies that have relied on abnormal liver blood tests to determine the prevalence of NAFLD [8,9] have generally come up with figures lower than more recent studies that have used ultrasound or magnetic resonance imaging to diagnose hepatic steatosis [10]. On the basis of estimates obtained from selected series, the prevalence of NASH in the general population of Western countries is considered to be around 2 and 3% [4]. Very little information is available on the incidence of NAFLD, however, results from a 10-year follow-up of a cohort of individuals who took part in the original Dionysos study in 1991-1992 have shown that 20% of them developed liver steatosis, corresponding to an incidence of 2% per year [5].
1.2. Natural history
The natural history of NAFLD/NASH is not well defined [4,11,12]. The few natural-history studies reported to date include small numbers of highly selected populations. The largest study thus far reported included 98 patients referred to a tertiary specialist centre [13]. This study found the occurrence of cirrhosis, overall death, and liver-related death to be 15%, 36, and 7%, respectively, during a mean follow-up of 8.3 years.
In a landmark study, Adams and colleagues conducted a population-based cohort study of 420 patients with NAFLD from Olmsted County, Minnesota, seen over a 20-year period and followed-up for up to 23.5 years. Survival in NAFLD patients was significantly lower than that in the general Minnesota population of the same age and sex. The occurrence of cirrhosis, overall death, and liver-related death was 5, 12.6, and 1.7%, respectively; these figures are lower than those reported for patients seen in specialist centres [14]. Presence of impaired fasting glucose (IFG)/diabetes, older age, or cirrhosis were independent factors associated with a significantly poorer prognosis. Death occurred in 12.6% of patients and was most commonly due to malignancy and ischemic heart disease. Liver disease was also an important contributor of death in NAFLD, being the third most common cause and accounting for 13% of all deaths. This implies that the increased mortality rate among NAFLD patients is partly due to complications of NAFLD. Aggressive management of insulin resistance may potentially reduce the risk of mortality and morbidity among these patients.
From this and other previous studies, it would appear that the natural history of NAFLD depends critically on disease stage [15] Patients with simple steatosis have a relatively benign 'liver' prognosis with a risk of developing clinical evidence of cirrhosis over 15-20 years in the order of 1-2%. Patients with NASH and fibrosis can progress to cirrhosis, defined histologically or clinically, with the risk varying from 0% at 5 years to 12% over 8 years [13,16,17]. The impact of NAFLD on the well established obesity- and diabetes-related risks of malignancy and cardiovascular disease deserves particular attention in view of the implications for screening/surveillance strategies in this growing number of patients.
1.3. Pathology of adult NAFLD and NASH
Hepatocellular steatosis is the hallmark of NAFLD and is more commonly macro vesicular, with a single large fat droplet displacing the nucleus or with smaller well-defined intracytoplasmic droplets. Groups of hepatocytes with micro vesicular steatosis may be observed [18]. The accumulation of fat usually starts in zone 3 and in more severe cases may occupy the whole acinus. Steatosis is usually evaluated semi-quantitatively [13,19]. There are no uniform criteria for the assessment of steatosis, and no exact definition of the smallest amount of fat in the liver to be considered as 'pathologic'. The lack of strict criteria for the diagnosis of NASH [4,18], may account for the significant differences in the reported frequency of histologically diagnosed NASH in NAFLD patients [18]. Recently, Kleiner et al. [20] designed and validated a histological feature scoring system that addresses the full spectrum of lesions of NAFLD and proposed a NAFLD activity score (NAS) for use in clinical trials.
The difficulties in differentiating between NASH and alcoholic steatohepatitis at the histological level, the recognition that specific underlying disorders are associated with NAFLD and NASH, the acceptance that the criteria used for excluding use of alcohol are not strictly defined and the possible concurrence of steatohepatitis with other forms of chronic liver disease, have led hepatopathologists and clinicians alike to a reconsideration of the term 'non-alcoholic steatohepatitis' [4,18]. The term 'steatohepatitis' followed by the underlying clinical condition if provided (i.e. diabetes, obesity, hyperlipidemia, etc.) may be more appropriate to use in histopathological diagnosis [18,19]. The significance of isolated portal tract fibrosis [21] described in morbidly obese subjects in the absence of NASH needs to be further investigated.
It is of great interest that NASH, once thought as a disease with benign course, is currently considered a significant cause of cryptogenic cirrhosis [22]. The development of hepatocellular carcinoma (HCC) is in some cases the final part in the natural history of NAFLD [23].
1.4. Diagnosis of NAFLD and NASH
Hepatic fat accumulation can be demonstrated by ultrasonography, computerized tomography, or magnetic resonance imaging; however, none of these methods provides information on the fine architecture of liver tissue or on the aetiology of steatosis. Even hepatic fat accumulation, as diagnosed by an imaging technique, along with the finding of elevated aminotransferase levels are not sufficient to distinguish between fatty liver alone and NASH [2]. Thus, for the diagnosis of NASH to be established, a liver biopsy is still required and therefore remains the gold standard. Only information from biopsy allows grading and staging of the disease [18,24]. This latter information is crucial for patient management because simple nonalcoholic fatty livers have a benign prognosis, whereas NASH can be progressive and lead to end-stage liver disease.
Currently, a liver biopsy is still required to confirm the diagnosis of NASH, as was demonstrated in a study, in which 41 of 82 patients with NAFLD showed significant fibrosis on biopsy [25]. A liver biopsy, however, is an expensive invasive procedure associated with a low but important risk of complications and poor patient acceptance. In clinical practice, the presence of clinical features of NAFLD and the absence of other common liver diseases is the usual mode of diagnosis. The issue of liver biopsy for all patients is debated and probably unnecessary. Certain clinical features such as a BMI >30, type 2 diabetes mellitus, age >45 and an AST/ALT ratio >1 are suggestive of higher risk of NASH and a biopsy should be considered in these patients [25]. The results of a liver biopsy in these subjects might lead to a more aggressive treatment strategy, participation in clinical trials of novel therapies, or screening for HCC in the setting of NAFLD-associated fibrosis.
1.4.1. Biomarkers of liver fibrosis
While histologic grading and staging of NASH are established, little is known about the sampling error, which in patients with chronic hepatitis C amounts to 30% for a difference of one stage or grade in the Metavir scale. Ratziu et al. recently assessed the sampling error of liver biopsy in NAFLD and its impact in the diagnosis and staging of NASH [26]. None of the histological features displayed high agreement: substantial agreement was only seen for steatosis grade, moderate agreement for hepatocyte ballooning and perisinusoidal fibrosis and fair agreement for Mallory bodies. Six of 17 patients with bridging fibrosis (35%) on 1 sample had only mild or no fibrosis on the other. Intraobserver variability was lower than sampling variability and did not account for the sampling error. Histological lesions of NASH are unevenly distributed throughout the liver; sampling error can result in substantial misdiagnosis and staging inaccuracies. Serological markers that allow to determine fibrosis stage and in particular the dynamic process of fibrogenesis and fibrolysis (de novo formation/deposition of hepatic connective tissue and its removal, respectively) are urgently needed.
Only few data of non-invasive markers are available for NASH [27], though results from other studies [28-30], are likely transferable. Recently, laboratory indices for the non-invasive classification of fibrosis stage, which are based on liver functional parameters, were developed. Thus the 'fibroscore', an algorithm using bilirubin, GGT, apolipoprotein A1, haptoglobin and a 2-macroglobulin, has been validated for patients with chronic hepatitis C [30]. However, the score only allows the correct classification of approx. 50% of patients into groups without or with relevant fibrosis (stage 0-1 vs. 3-4 according to Metavir). Similar results were obtained by combining other routine laboratory parameters, AST, platelet and AST/ALT ratio [31]. A multicentric European study that investigated the diagnostic and prognostic value of 9 connective tissue parameters in 1.800 patients with liver disease, partly in a follow-up design, has been recently published [32]. Once validated, algorithms based on such fibrosis markers might be invaluable as predictors of fibrosis progression or regression under pharmacological treatment. Novel developments in non-invasive imaging methods may be a valuable adjunct to assess fibrosis stage, as exemplified by measurements of liver stiffness, which correlates with fibrosis [33,34].
1.5. NAFLD and NASH in children
v
Childhood NAFLD has become an important childhood liver disease [35-37]. It is likely to reach epidemic proportions in children in the next decade. The full spectrum of NAFLD has been identified in children. NAFLD may occur in young (pre-school) children, and there is no female predominance in the paediatric age bracket. If symptomatic, children often present with vague abdominal pain. Laboratory studies show that serum aminotransferase elevations are mild to moderate, with serum ALT higher than AST. Hypertriglyceridemia is the typical blood lipid abnormality, although hypercholesterolemia may occur. NASH may be more severe in certain ethnic groups including Hispanics and Asians or in association with certain metabolic disorders characterized by abnormalities in insulin receptor/signalling, such as lipodystrophy syndromes [38,39]. It is clear that childhood obesity (and overweight) should be identified as early as possible. A dietary strategy to minimize postprandial hyperinsulinemia and overall fat intake, such as a low glycemic index diet, may be the best dietary strategy [37]. The real efficacy of drug treatments in children requires further investigation. Studies examining the natural history of childhood NAFLD and NASH are urgently needed.
2. Mechanisms of disease progression in NAFLD
Understanding the mechanisms that lead to the progression from simple steatosis to advanced disease is clearly important to inform the rational design of treatment strategies directed at those who have developed progressive disease. Based on the clinical and experimental data available at the time, the so called 'two hit' model of progressive NAFLD was proposed in 1998 [40]. This model considered the development of steatosis to be the 'first hit' increasing the sensitivity of the liver to the putative 'second hits' leading to hepatocyte injury, inflammation and fibrosis. The best candidates for these second hits were considered to be oxidative stress and associated lipid peroxidation and cytokines, principally, TNFa. Studies published over the subsequent seven years have, perhaps not unsurprisingly, led to revisions in this model of pathogenesis, although oxidative stress and cytokines retain a central role. The most important modifications to the model have come from an increased understanding of the sources of oxidative stress and cytokines, in particular the prominent role of insulin resistance, free fatty acids (FFA) and adipose tissue inflammation. Endoplasmic reticulum (ER) stress has emerged as a third potential second hit leading to liver injury and several studies have demonstrated an important role for apoptosis as a mechanism of hepatocyte death in NASH [41]. Moreover, it has become increasingly clear that the well-established correlation observed between the severity of steatosis and the severity of liver injury and fibrosis may be explained by these classical histological features of NAFLD being due to the same pathological mechanisms. Steatosis may therefore be an epiphenomenon of the injurious mechanisms rather than a true 'first hit' involved in the development of progressive liver damage.
2.1. Oxidative stress and lipid peroxidation
Reactive oxygen species (ROS)-mediated lipid peroxidation is an attractive candidate for a central role in the pathogenesis of NASH since, as reviewed by Pessayre et al [42], it potentially explains all of the typical histological features. Peroxidation of plasma/mitochondrial membranes may cause direct cell necrosis/apoptosis and megamitochondria. ROS-induced expression of Fas-ligand on hepatocytes may induce fratricidal apoptotic cell death given that Fas is expressed at increased levels on hepatocytes isolated from liver biopsies from NASH patients [41]. The aldehyde end-products of lipid peroxidation, 4-hydroxynonenal (HNE) and malondialdehyde (MDA) can covalently bind to hepatic proteins forming adducts that are capable of initiating a potentially injurious immune response [43]. MDA and HNE can also stimulate the synthesis of extracellular matrix (ECM) proteins by hepatic stellate cells (HSC) [44], cross link cytokeratins to form Mallory bodies and stimulate neutrophil chemotaxis. Lipid peroxidation products may also initiate a cascade that leads to the activation of the transcription factor NF-KB which will lead to the increased transcription of various inflammatory cytokines, adhesion molecules, chemokines and death ligands by hepatocytes and non-parenchymal cells [45,46].
Studies demonstrating evidence of oxidised proteins (tyrosine nitration), DNA or lipids have provided persuasive evidence that oxidative stress occurs in animal models of NAFLD [47,48] and in patients with NASH [43,49-51], with the most recent human studies demonstrating that the magnitude of oxidative stress correlates with disease severity [43,51]. Additional evidence for oxidative stress playing a key role in NAFLD has come from animal models demonstrating the development of spontaneous steatohepatitis in antioxidant (methionine adenosyltransferase [MAT]1A-/-) deficient mice [52] and increased ROS production and an adaptive increase in antioxidant defences in the Ob/Ob, leptin deficient model of NAFLD [53]. Further support for a role for oxidative stress in human NAFLD has come from a dietary study demonstrating that, compared to 'healthy' obese individuals, patients with NASH consume a diet lower in antioxidants [54].
2.1.1. FFA oxidation, insulin resistance and the generation of ROS in NAFLD
In addition to the expected generation of ROS from inflammatory cells once steatohepatitis has become established, a considerable body of evidence suggests that hepatocytes are a prominent source of ROS in NAFLD, arising as a consequence of the oxidation of an increased supply of FFA by mitochondria, peroxisomes and microsomal cytochrome (CYP) P450 enzymes [55]. Peripheral insulin resistance, particularly in splanchnic adipose tissue [56], leads to unopposed lipolysis and a subsequent increased supply of FFA to the liver. The resulting accumulation of fat within the liver is enough to induce hepatic insulin resistance via mechanisms involving activation of PKC-E, JNK, I-KB kinase B (IKK-B) and NF-KB [57,58]. Hepatic insulin resistance will favour the entry of FFA into the mitochondria for oxidation. In addition, FFAs and their metabolites are ligands for the transcription factor, peroxisomal proliferator-activated receptor a (PPARa) that regulates the transcription of a variety of genes encoding enzymes involved in mitochondrial, peroxisomal and microsomal fat oxidation. Fat-induced hepatic insulin resistance and up-regulation of PPARa-regulated genes will therefore result in increased FFA oxidation by at least three different pathways, all of which are capable of generating ROS that can contribute to the development of oxidative stress [59]. Interestingly, PPARa-/- mice appear to be particularly susceptible to methionine-choline deficient diet-induced NASH [60]. This may appear to be counter-intuitive, but may be explained by the massive lipid accumulation that occurs in these mice and the apparent ability of PPARa-induced enzymes to clear potentially toxic lipid peroxides.
2.1.2. Intracellular source of hepatocyte ROS in NAFLD
Most of the available data from animal models and patients with NAFLD suggest that mitochondria are the most important intracellular source of ROS in NASH [61]. With respect to animal studies, mice lacking the gene encoding fatty acid acyl-CoA oxidase, the initial enzyme in the peroxisomal B-oxidation cascade, develop severe NASH, presumably reflecting increased FFA oxidation via the mitochondrial and microsomal pathways [62]. Moreover, mitochondria isolated from the livers of Ob/Ob mice have an increased generation of hydrogen peroxide and superoxide anions compared to mitochondria isolated from wild type mice [53,63]. With respect to human data, there is increasing evidence that mitochondrial dysfunction is present in patients with NAFLD [49,64-66]. This dysfunction, which would be expected to increase ROS generation by mitochondria, has been attributed by several authors to the mitochondrial effects of TNFa, providing a further example of the 'cross-talk' that exits between oxidative stress and cytokine-related mechanisms of injury in NAFLD.
2.1.3. Extra-hepatic sources of ROS
Recent evidence suggests that ROS and subsequent oxidative stress may arise in the expanded adipose tissue mass in obesity due to an augmented expression of NADPH oxidase and decreased expression of antioxidant genes [67]. As in hepatocytes, this may also arise in response to elevated levels of FFA with subsequent activation of NADPH oxidase and ROS production. The degree to which this oxidative stress arises in adipocytes themselves or in the macrophages infiltrating fat depots in obesity [68] is at present unclear, however, this adipose tissue-derived ROS could contribute to the systemic oxidative stress observed in patients with the metabolic syndrome and therefore to hepatic oxidative stress.
2.2. Endoplasmic reticulum stress
The endoplasmic reticulum (ER) is the organelle in which membrane and secretory proteins achieve correct folding and oligomerisation due to the presence of specialised chaperones. Because this is an essential function in protein management, the ER is exquisitely sensitive to alterations in homeostasis. Stress (glucose or nutrient deprivation, viral infections, lipids), increased synthesis of secretory proteins or an increased expression of mutant proteins, elicits the so called ER stress response which results in the activation of a number of transcription factors and kinases [69]. These transcription factors lead to up-regulated lipid synthesis (via sterol regulatory element-binding proteins [SREBPs]) and to the transcription of pro-apoptopic genes. Activation of ER resident kinases (IRE1a, PERK) leads to the activation of stress kinases (including JNK1), which can contribute to the induction of ER stress-related apoptosis, while release of the pro-apoptotic caspase 12 from its inhibitor TRAF2 is also typically part of the ER stress response. The ER stress response is therefore characterised by increased lipid synthesis, apoptosis and JNK-induced insulin resistance, all typical features of NAFLD. Evidence that this response may play a role in NAFLD is at present indirect, but is derived from at least two independent observations. First, ER stress is increasingly recognised as an important mechanism of alcohol induced liver disease [70,71]. Second, ER stress has recently been shown to be an important feature of obesity, linking obesity with insulin resistance and the development of diabetes [72]. The precise cause of ER stress in obesity, and whether this occurs in the liver is unclear at present, but this emerging disease mechanism clearly provides a potential link between obesity, steatosis and apoptotic hepatocyte death worthy of further study.
2.3. Cytokines
Cytokines are obvious candidates for a role in the pathogenesis of progressive NAFLD for reasons similar to those outlined for ROS and lipid peroxidation. First, cytokines are capable of producing all of the classical histological features of NASH including hepatocyte death/apoptosis (TNFa/TGFB), neutrophil chemotaxis, (IL8) HSC activation (TNFa/TGFB) and Mallory bodies (TGFB) [42]. Second, cytokines (TNFa, IL-6 and IL-1B) may play a role in the hepatic and systemic insulin resistance associated with NASH [73]. Third, studies in patients with obesity-related NASH have shown that compared to obese controls, patients have an increased expression of TNFa mRNA and its receptors in both their liver and adipose tissue, with the increased expression correlating with histological severity [74].
2.3.1. TNFa as a mediator of apoptosis in NAFLD
Recent reports that apoptosis is an important mode of cell death in NASH [41,46] makes TNFa a particularly attractive candidate for a role in mediating liver injury given its ability to induce apoptosis in hepatocytes under conditions of oxidative stress [75]. The interaction of TNFa with its receptor TNF-R1 initiates a number of apoptotic cascades [76]. First, acting via the death-inducing signalling complex (DISC) TNF-R1 activates caspase 8 which cuts Bid. In combination with Bax this leads to increased permeability of the outer mitochondrial membrane and the release of cytochrome c from the intermembrane space. This leads to a partial block in the flow of electrons in the respiratory chain and increasing mitochondrial ROS formation [77] which induce the opening of the inner membrane mitochondrial permeability transition (MPT) pore. This leads to the leakage of apoptosis inducing factors (predominantly cytochrome c) into the cytosol where they can activate caspase 9 and initiate the apoptotic cascade. Second, TNFa-TNF-R1 binding also activates acid sphingomyelinase (ASMase), which generates ceramides from the major membrane sphingolipid, sphingomyelin [76]. Ceramide can induce apoptosis by opening the MPT pore directly and, via inhibition of MAT1A, lead to the depletion of glutathione (GSH) and oxidant stress-mediated MPT pore opening [78]. These mechanisms of apoptosis provide ample opportunities for TNFa to act synergistically with an increased FFA supply and the associated oxidative and ER stress to induce apoptosis in NASH. An increased supply of FFA would be expected to increase the synthesis of sphingomyelin [79] while depletion of mitochondrial GSH by oxidant and ER stress [71] is known to increase the sensitivity of hepatocytes to the mitochondrial effects of TNFa [75,80].
2.3.2. Intra-hepatic sources of TNFa in NAFLD
2.3.2.1. Hepatocytes
A direct link between an increased supply of FFA to the liver and the increased expression of TNFa observed in the livers of NASH patients [72] has been provided by a recent study demonstrating that FFA lead to NF-kB-dependent TNFa expression in hepatocytes [81]. This pathway involves translocation of Bax to lysosomes with subsequent lysosomal destabilisation and release of the lysosomal cystine protease, cathepsin B, into the cytosol, which activates NF-KB. This finding presumably explains the activation of NF-KB and the increased expression of its target genes (TNFa, IL-6, IL-1B) observed in the livers of animals fed a high fat diet [58].
2.3.2.2. Kupffer cells
The recent observation that resident hepatic macrophages (Kupffer cells) are activated in the livers of mice fed a high fat diet suggests that these 'professional' cytokine-producing cells may be an important source of pro-inflammatory cytokines in NAFLD [58]. The activating stimulus is at present unknown but hepatocyte-derived cytokines appear to be important given evidence that hepatocyte specific inhibition of NF-KB reduces macrophage activation in mice fed a high fat diet [58]. Clearance of oxidised lipid deposits via scavenger receptors may be a further mechanism of macrophage activation in NAFLD [73], and evidence from animal models [82] and a study reporting a high frequency of small intestinal bacterial overgrowth in patients with NASH [83] suggest that gut-derived portal endotoxin may be further stimulus for Kupffer cell activation in NAFLD.
2.3.3. Adipocyte-derived cytokines in NAFLD
Recent evidence that adipose tissue in obesity is characterised by macrophage infiltration and associated chronic inflammation [68] suggests a further potential source of TNFa and other inflammatory cytokines in obesity [84]. As in the liver, the precise stimulus to this adipose tissue-related macrophage activation in adipose tissue is unknown, but cytokine release from metabolically 'stressed' adipocytes, macrophage scavenging of oxidised lipids [73] and systemic endotoxemia [85] may all be important. Whether the production of cytokines by these adipose tissue macrophages is sufficient to exert endocrine effects on the liver is uncertain, however, recent data from myeloid-specific conditional IKK-B loss suggest that the IKK-B/NF-KB dependent production of pro-inflammatory cytokines by myeloid cells is an important mediator of systemic and hepatic insulin resistance [73].
The increased secretion of 'offensive' cytokines, such as TNFa, and the generation of ROS in obesity may also lead to a reduction in the secretion of the 'defensive' adipocytokine, adiponectin. Adiponectin is anti-steatotic in both muscle and hepatocytes, probably by activating PPARa and AMP-dependent kinase (AMPK) [86]. Adiponectin also has profound anti-inflammatory effects probably via suppressing the production and action of TNFa [87]. Circulating adiponectin levels are reduced in obesity, insulin resistance and T2DM [88] due, at least in part, to suppression of its synthesis by TNFa [87] and possibly to increased oxidative stress [67]. Recent studies in humans showing reduced serum levels of adiponectin and reduced hepatic expression of its receptor RII in patients with NASH compared to those with steatosis [89,90], provide direct evidence that TNFa- and ROS-mediated suppression of adiponectin play an important role in the pathogenesis of progressive NAFLD.
2.4. Mechanisms of fibrosis in NAFLD
Any of the above mechanisms of hepatocyte injury, cell death and associated inflammation will lead to the activation of HSC and deposition of ECM proteins as part of the normal 'healing' response. Recent evidence has suggested a prominent role for hepatocyte apoptosis in hepatic fibrogenesis which is pertinent given the importance of apoptotic cell death in NASH [41]. Hepatocyte apoptosis leads to fibrogenesis following the ingestion of apoptosing hepatocytes by Kupffer cells and HSC themselves which subsequently release TGFB capable of activating HSC [91,92]. In addition to fibrosis arising as a result of hepatocyte injury and inflammation, an increasing body of evidence supports a role for non-necro-inflammatory mediators related to obesity and insulin resistance in the pathogenesis of liver fibrosis in NAFLD. The adipocytokine leptin directly stimulates the production of ECM proteins in HSC [92] and has been shown to play a prominent role in hepatic fibrosis in animal models of NASH [93]. Circulating leptin levels are related to adipose tissue mass and are raised in patients with NAFLD [94] suggesting a potential role in obesity-related fibrosis. Direct HSC-activating pro-fibrogenic roles have also been demonstrated for angiotensin II [95] and norepinephrine [96], both of which are secreted by adipose tissue and are raised in the serum of obese patients. A direct fibrogenic role for insulin resistance-associated hyperglycaemia and hyperinsulinaemia has been suggested by studies showing that the synthesis of the fibrogenic growth factor, connective tissue growth factor (CTGF), by HSC is upregulated by glucose and insulin [97]. Furthermore, CTGF is overexpressed in the livers of patients with NASH and correlates with the degree of fibrosis [97]. The reduced production of adiponectin associated with obesity may also contribute to the development of liver fibrosis since it appears to exert potent anti-fibrotic effects [97].
2.5. Susceptibility
The reason why only a minority of patients with classical risk factors for NAFLD ever develop more than simple steatosis remains largely unclear Dietary and lifestyle factors contributing to the development of obesity and insulin resistance seem likely to be important, however, family studies and ethnic variations also suggests a role for genetic determinants of disease susceptibility. As yet, studies looking for genetic factors predisposing to the development of advanced NAFLD have been limited by small sample sizes with no reproducible, robust associations reported thus far. However, increased collaborative efforts between groups, coupled with the growing information on genetic variation emerging from the human genome project seem likely to lead to important advances in this field in the next 4-5 years. Not only will this type of information provide further evidence for particular pathogenic mechanisms and thus provide further rationale for the design of therapies, but it should also allow the identification of particular high risk individuals and allow the opportunity for targeted prevention strategies [98].
3. Management of NAFLD/NASH
The most important aim in the management of NAFLD/NASH is to prevent the progression to end-stage liver disease associated with the development of cirrhosis and liver failure [99]. When NAFLD/NASH is secondary to a well defined cause, including certain drugs, exposure to hepatotoxins, or a specific associated medical disease, the most important step is to remove or treat the cause. Accordingly causes of secondary NASH must be actively sought in all patients with NAFLD.
In the primary forms of NAFLD/NASH, resulting from a multi-factorial process in which insulin-resistance promotes the storage of fat into the liver, patients often present with various components of the metabolic syndrome [100] associated with an increased risk of cardiovascular morbidity and mortality. Treatment should have the twin aims of decreasing cardiovascular risk and preventing liver disease progression. Fortunately, an increasing body of evidence suggests that most of the treatments proven to reduce cardiovascular mortality in patients with the metabolic syndrome may be beneficial for NAFLD. The first step should be life-style changes aimed primarily at reducing obesity, followed, if indicated by pharmacological treatment of insulin resistance. At present there is no firm evidence base to recommend any specific liver-directed therapies in patients with NAFLD.
3.1. Change in lifestyle and treatment of obesity
NAFLD is strongly associated with obesity and the metabolic syndrome. Attempts should be made to change lifestyle principally by dietary restriction and implementing physical activity aimed at weight loss. Cost-effective and safe lifestyle modifications are difficult to implement in the long term as they call for changes in individuals' often long entrenched habits and require intensive counselling for any chance of success. This explains the relative lack of published trial evidence available to confirm the widely held belief that gradual and sustained weight loss will cure a sizable fraction of patients affected by NASH.
3.1.1. Diet combined with physical activity
Current recommendations call for at least 30min of physical activity such as brisk walking five days of the week. When used alone to treat obesity, physical activity usually achieves modest weight reduction. However, this should not be regarded as the only measure of the positive effects of exercising. Exercising has numerous physiological benefits such as reduction of blood pressure, correction of atherogenic lipid profile and improvement of glucose tolerance, among others. In a case series of 39 obese patients published in 1990, Palmer and Schaffner found that a weight loss of 10% body weight associated with physical activity suffices to improve liver biochemical tests [101]. In a prospective controlled study by Ueno et al. a restricted diet associated with physical activity for three months lead to a decrease of the BMI from 31 to 28kg/m2 in the 15 treated patients that also showed a normalization of their metabolic abnormalities and a resolution of the hepatic steatosis; these changes were not observed in the 10 matched patients of the control group [102]. These benefits can also be expected in children, since weight loss was found to correct the hepatic biochemical tests and the ultrasonographic image of bright liver (interpreted as steatosis) [103]. McClain and co-workers enrolled 16 patients in a 3-months protocol-combining step 1 of the American Heart Association diet with encouragement to walk or jog daily [104]. A mean weight loss of 1kg/m2 combined with exercise improved liver enzyme levels in overweight patients with NASH.
3.1.2. Pharmacological control of body weight
Orlistat inhibits pancreatic and gastric lipase and interferes with the intestinal hydrolysis of ingested triglycerides. Patients treated with orlistat avoid fat-containing foods, since their consumption results in diarrhoea. In ten obese patients with biopsy proven NASH Harrison et al. demonstrated that Orlistat given for 6 months together with dietary counselling was associated with significant decrease in body weight, haemoglobin A1c, ALT and AST. Patients achieving a 10% or greater reduction in body weight also had an improvement in steatosis and fibrosis [105]. Sabuncu et al. compared 13 patients treated with sibutramine with 12 patients treated with orlistat for 6 months and found comparable improvements in terms of weight loss, liver tests, hyperglycaemia and hypertriglyceridaemia [106]. Whether such strategies are truly beneficial for patients with NASH is at present unclear and requires further controlled trials of longer duration and with histological follow-up.
3.1.3. Bariatric surgery
Three procedures are currently in use for the surgical treatment of obesity: biliopancreatic diversion, gastric bypass and gastric restriction (banding). In a series of 104 patients biliopancreatic diversion decreased steatosis in all patients, but led to an increase in fibrosis in 40%, with an improvement in only 27% [107]. Subacute hepatic failure after biliopancreatic diversion has been reported [108]. In a series of 91 patients Silverman et al reported that gastric bypass surgery led to an improvement in the degree of steatosis in 83 patients, which correlated with the length of time elapsed since the surgery. They also observed an amelioration of the perisinusoidal fibrosis in 11 out of 13 patients with this finding [109]. More recently Dixon et al. analyzed the effect of the adjustable gastric banding surgery on 36 patients with NASH. This procedure, which acts by inducing satiety and restricting food intake without affecting intestinal absorption, led to a decrease in BMI from 47 to 34kg/m2 in 26 months and a significant improvement in steatosis, necroinflammation and fibrosis [110].
3.2. Medical treatment of insulin resistance
Considering the importance of insulin resistance in disease pathogenesis it is not surprising that insulin sensitizing agents, metformin and more recently the thiazolidinediones, have been tested as treatments for patients with NAFLD.
3.2.1. Metformin
Metformin down-regulates hepatic glucose production and diverts fatty acids from triglyceride production to mitochondrial B-oxidation, reducing hyperinsulinemia and improving hepatic insulin resistance, without any risk of hypoglycemia. A recent randomized, trial compared the effect of metformin (850mg bid) together with a lipid and calorie-restricted diet vs. diet alone for 6 months in patients with NAFLD[111]. Patients on metformin showed a greater improvement in ALT levels and insulin resistance compared to those on diet alone. Although more patients in the metformin group showed an improvement in necroinflammatory activity (46 vs. 10%), there were no significant differences in the severity of necroinflammation or fibrosis at the end of treatment between the two groups. During a post-treatment follow-up of 6-12 months, ALT levels did not return to pre-treatment values. In another recent open label study 15 patients with NAFLD completed 1 year of treatment with metformin [112]. Although there was an improvement in transaminanses and insulin sensitivity after 3 months, there was no further improvement in insulin sensitivity thereafter and there was a gradual rise in transaminases back to pre-treatment levels. Among the 10 patients undergoing post-treatment biopsy, three showed an improvement in steatosis, two an improvement in inflammation and one an improvement in fibrosis. In the latest open label, randomized trial [113], 55 nondiabetic NAFLD patients were given metformin (2g/day) for 12 months, and control NAFLD patients were given either vitamin E (800IU/day; n=28) or were treated by a prescriptive, weight-reducing diet (n=27). Metformin treatment was associated with higher rates of transaminase normalization, and the frequency of positive criteria for the metabolic syndrome was reduced only in the metformin arm (P=0.001). A repeat biopsy in 17 of the metformin-treated cases (14 non-biochemical responders) showed a significant decrease in liver fat (P=0.0004), necroinflammation, and fibrosis (P=0.012 for both). With respect to children with NAFLD, Schwimmer et al [114], in a single-arm, open-label pilot study of metformin 500mg twice daily for 24 weeks in 10 non-diabetic children with biopsy-proven NASH, reported a normalization of ALT in 40% and of AST in 50% of subjects. In addition these children demonstrated significant improvements in liver fat, measured by magnetic resonance spectroscopy (30-23%, P<0.01); insulin sensitivity, measured by quantitative insulin sensitivity check index (0.294-0.310, P<0.05); and quality of life measured by pediatric quality of life inventory 4.0 (69-81, P<0.01).
3.2.2. Thiazolidinediones
The thiazolidinediones, which are ligands for peroxisome proliferator-activated receptor-r (PPAR-r), dramatically improve insulin sensitivity and may also have anti-inflammatory and anti-fibrotic effects. The second-generation thiazolidinediones pioglitazone and rosiglitazone have recently been studied in patients with NASH with encouraging results. Pioglitazone, at a dose of 30mg/day for 48 weeks, produced a significant decrease in serum ALT, which normalized in 72% of the patients [115]. Liver histology showed improvements in steatosis, cell injury, inflammation, Mallory bodies and fibrosis. Magnetic Resonance Imaging (MRI) of the liver confirmed a marked decrease in liver fat and liver volume. Most patients also had improvements in insulin sensitivity, as measured by oral and intravenous glucose tolerance tests. In a larger trial [116], rosiglitazone at a dose of 4 mg bid for 48 weeks was tested in obese NAFLD patients (50% with impaired glucose tolerance or diabetes). Treatment improved insulin sensitivity and serum ALT, in the absence of side effects. Biopsies of 10 patients (45%) no longer met the published criteria for NASH after treatment.
In summary, insulin-sensitizing agents seem promising drugs for NAFLD. However, all of the studies thus far are uncontrolled pilot studies in a disease where any intervention usually results in an improvement in liver biochemistry and natural history studies demonstrate an improvement in NASH in significant proportion of untreated patients. Thiazolidinediones tend to induce weight gain, which may be an important drawback of prolonged treatment. Clearly, the results of current ongoing randomized controlled clinical trials are awaited with interest.
3.3. Other medications for NASH
Several medications have been used to treat NASH, mostly in small non-controlled clinical studies. Among them, the cytoprotective agent, ursodeoxycholic acid, and the anti-oxidant agent vitamin E have been the most frequently studied although encouraging trials of fibrates, angiotensin blockers and pentoxifylline have recently been reported.
3.3.1. Ursodeoxycholic acid
Although an early non-randomized not controlled pilot study with ursodeoxycholic acid suggested a beneficial effect [117], this was not confirmed in a subsequent paediatric series [118]. In a recent trial, 166 adults with biopsy-proven NASH were enrolled in a randomized double-blind placebo-controlled study comparing UDCA 13-15mg/kg/d vs. placebo for a period of 2 years [119]. Although a significant improvement in biochemistry and histology was observed in the treated group, a similar degree of improvement occurred in the placebo group.
3.3.2. Vitamin E
Lavine et al treated 11 children with NASH with vitamin E 400-1200IU/d [120]. Liver biochemistry improved significantly, however, in some patients they became abnormal after withdrawal of vitamin E. In a randomized double-blind, placebo-controlled study [121], vitamin E 1000IU/d in combination with vitamin C, 1g/day was tested in 23 patients with NASH, and although after 6 months a significant improvement of the fibrosis score was found in the treated group in comparison with baseline values, a comparison between the groups revealed no differences.
3.4. Conclusion
At present there is no definite treatment for NAFLD/NASH. The first step should be a change in life style and, if indicated, pharmacological management of the associated metabolic conditions. Other treatments, including pharmacological treatment of insulin resistance in non-diabetic patients require further studies in randomised, controlled trials of at least one year duration with follow-up biopsies. Patients with NASH-related cirrhosis should be managed conventionally including surveillance for HCC, which appears to a frequent occurrence in these patients. Patients with evidence of decompensation should be considered for liver transplantation, although a high rate of recurrence has been found [122-126] suggesting that attention should be paid to the underlying metabolic syndrome post-transplantation.
|
|
| |
| |
|
|
|
