html xmlns=”http://www.w3.org/1999/xhtml”>
29
Non-alcoholic fatty liver disease
Introduction
Non-alcoholic fatty liver disease (NAFLD) is increasingly diagnosed worldwide and considered to be the commonest liver disorder in Western countries. It comprises a disease spectrum ranging from simple steatosis (fatty liver), through non-alcoholic steatohepatitis (NASH) to fat with fibrosis and ultimately cirrhosis. Simple steatosis is largely benign and non-progressive, whereas NASH, characterized by hepatocyte injury, inflammation and fibrosis can lead to cirrhosis, liver failure and hepatocellular carcinoma (HCC). NAFLD is strongly associated with obesity, insulin resistance, hypertension and dyslipidemia and is now regarded as the liver manifestation of the metabolic syndrome. Rapid spread of the obesity “pandemic” in adults and children, coupled with the realization that the outcomes of obesity-related liver disease are not entirely benign, has led to rapid growth in clinical and basic studies in NAFLD over the past decade.
Epidemiology
NAFLD is often an asymptomatic illness in which the liver blood tests may be completely normal. This has made studies on prevalence extremely difficult, with most relying on ultrasound, which is known to be sensitive only when more than one-third of the liver is affected by steatosis. With this proviso the prevalence of NAFLD appears to be around 20–30% in Western adults [1, 2], and 15% in Asians [3]. Due to the lack of prospective studies, the true incidence of NAFLD is not well defined, although from the information available, it appears to be low [4]. Since liver biopsy is the only method of accurately diagnosing steatohepatitis, incidence/prevalence studies of NASH are rare. According to available data, NASH is much rarer than NAFLD, affecting 2–3% of the general population [5]. NAFLD and NASH are strongly associated with the presence and severity of obesity. Studies in severely obese patients (BMI > 35kg/m2) undergoing bariatric surgery have reported prevalences of NAFLD and NASH of 91% and 37% respectively [6], while a large post-mortem study reported NASH to be present in 3% of non-obese, 19% of obese and 50% of a morbidly obese individuals [7]. A recent novel observational study in NAFLD patients has demonstrated that while central obesity correlates with the severity of inflammation, dorsocervical lipohypertrophy correlates with hepatocyte injury, inflammation and fibrosis [8]. Type 2 diabetes mellitus (T2DM) is the other major association of NAFLD, with a prevalence of 70% recently reported from an ultrasound survey of almost 3000 unselected Italian T2DM patients [9,10]. Even in the absence of obesity and T2DM, NAFLD is closely associated with other features of the metabolic syndrome, with one study of non-diabetics with NAFLD reporting that 18% of normal weight patients and 67% of obese fulfilled criteria for the metabolic syndrome [11].
There are no accurate data regarding temporal changes in the prevalence of NAFLD; however, the rising prevalence of obesity, diabetes and the metabolic syndrome seems likely to be reflected in an increasing prevalence of NAFLD. This trend is of particular concern in the paediatric population where the reported increase in obesity will undoubtedly result in a higher incidence and prevalence of paediatric and adult NAFLD in the future. To date, studies in children have reported a prevalence of NAFLD of 3% in the general paediatric population and 53% in obese children [12,13]. Reports of toddlers with NAFLD and primary school children with NAFLD- related cirrhosis are clearly a cause for alarm [14].
Natural history of NAFLD
In marked contrast to alcoholic steatohepatitis, the short-term prognosis of NAFLD is good. The largest prospective histological study of the natural history of NAFLD, with a mean follow-up of 13 years, has recently been published [15]. Data from this and other studies suggest that the long-term hepatic prognosis of patients with NAFLD depends on the histological stage of disease at presentation [16]. Among patients with simple steatosis 12–40% will develop NASH with early fibrosis after 8–13 years. For patients presenting with NASH and early fibrosis, around 15% will develop cirrhosis and/or evidence of hepatic decompensation over the same time-period, increasing to 25% of patients with advanced pre-cirrhotic fibrosis at baseline. In the most recent study, weight gain and the presence of portal tract fibrosis on index biopsy were the only significant predictors of fibrosis progression [15]. About 7% of subjects with compensated cirrhosis associated with NAFLD will develop a hepatocellular carcinoma (HCC) within ten years, while 50% will require a transplant or die from a liver-related cause [17]. The risk of HCC in NAFLD-related cirrhosis is comparable to that in cirrhosis associated with alcohol or hepatitis C [18]. This may partly explain the recently reported associations of HCC with high BMI and T2DM [19]. Liver transplantation is increasingly available to those with chronic liver failure and about 10–12% of liver transplants in the USA are for NAFLD cirrhosis [20]. Unfortunately the condition can recur in transplanted organs. The overall and liver-related mortality of patients with NAFLD is higher than in an age and sex-matched population, with adjusted hazard ratios of 1.038 and 9.32 respectively [21, 22].
Susceptibility
While the vast majority of individuals with obesity, insulin resistance and the metabolic syndrome will have steatosis, only a minority will ever develop steatohepatitis, fibrosis and cirrhosis. Potential environmental determinants of NAFLD are dietary factors and small bowel bacterial overgrowth [23, 24]. Recent studies have shown that diets high in saturated fat, fructose-containing soft drinks and meat, and low in antioxidants and omega-3 containing fish are associated with an increased risk of NAFLD/NASH [25, 26, 27]. With respect to alcohol intake, while there is no doubt that obesity increases risk of cirrhosis in heavy drinkers [28], emerging evidence suggests that “sensible” light alcohol intake may be protective versus NAFLD/ NASH [29, 30], an effect that appears likely to be due to the beneficial effect of light alcohol intake on insulin sensitivity. Family studies and inter-ethnic variations in susceptibility suggest that genetic factors may be important in determining disease risk. Thus far, only one genetic association with advanced NAFLD has been replicated in an independent population – an association between a functional polymorphism in the gene encoding the tumor suppressor gene Kruppel-like factor 6 and fibrosis [31]. More recently the first genome-wide association study has been performed in patients with proton magnetic resonance spectroscopy (1H-MRS) diagnosed NAFLD and found an association between hepatic fat content and genetic variation in the PNPLA3 gene encoding a protein involved in lipid metabolism [32].
Disease associations with NAFLD
Cardiovascular disease
Given the close association between NAFLD and classical cardiovascular risk factors it is perhaps not surprising that, when compared to controls, patients with NAFLD have a higher prevalence of atherosclerosis, as shown by increased carotid wall intimal thickness, increased numbers of atherorosclerotic plaques and increased plasma markers of endothelial dysfunction [9, 33, 34]. This association also extends to children, with the prevalence of coronary and aortic atheroma higher in children with fatty liver compared to controls in an autopsy based report [35]. Consistent with these observations three natural history studies have reported that the increased age-related mortality observed in patients with NAFLD is attributable to cardiovascular as well as liver-related deaths [15, 17, 22]. Although an indirect association between NAFLD and cardiovascular disease is expected, a growing body of evidence supports a direct role for NAFLD in the pathogenesis of atheromatous cardiovascular disease. A recent study of unselected patients with T2DM reported that the prevalence of cardiovascular, cerebrovascular and peripheral vascular disease was significantly greater in those with NAFLD than in those without, independent of the individual components of the metabolic syndrome [10]. A similar finding has been observed for microvascular diseases, nephropathy and retinopathy [36]. The mechanism of any direct effect of NAFLD on cardiovascular risk remains unclear; possibilities include the release of atherogenic inflammatory cytokines and pro-coagulant factors from the steatotic liver [37].
Polycystic ovary syndrome [PCOS]
As with the association between NAFLD and the metabolic syndrome, the now well-established association between NAFLD and the PCOS seems likely to be indirect as a result of both conditions being characterized by insulin resistance. Up to 30% of females with PCOS have elevated alanine transaminase (ALT) levels [38], and NAFLD prevalence of 42% has been reported in a series of PCOS patients with a mean age of 25 years [39]. More recently, advanced fibrotic liver disease has been reported in patients with PCOS, suggesting that women with this syndrome require careful hepatic evaluation [40].
Obstructive sleep apnoea
Chronic intermittent hypoxia, as seen in obstructive sleep apnoea (OSA), has been associated with cardiovascular disease, the metabolic syndrome and insulin resistance [41]. As might be expected, therefore, a proportion of patients with OSA have elevated liver enzymes and histological features of NASH independent of body weight [42]. The severity of histology and the associated insulin resistance both correlate with the severity of OSA, strongly implicating insulin resistance as the pathogenic mechanism linking OSA to NASH although not entirely excluding a role for hypoxic liver injury. As with PCOS, this and other similar reports suggest that patients with OSA require hepatic evaluation, and that the diagnosis of OSA should be considered in NAFLD patients reporting daytime somnolence, sleep disturbances or any other symptoms suggesting a diagnosis of OSA.
Clinical presentation
NAFLD is a largely asymptomatic condition that may reach an advanced stage before it is suspected or diagnosed. Symptoms such as right upper quadrant discomfort, fatigue and lethargy have been reported in up to 50% of patients, but are uncommon modes of presentation. Fatigue is a significant problem in NAFLD patients and associates with objectively measured inactivity, but not histological disease severity [43]. Most patients with NAFLD are diagnosed after they are found to have hepatomegaly, or, more commonly, unexplained abnormalities of liver blood tests performed as part of routine health checks or during drug monitoring (e.g. statin therapy). NAFLD is the commonest cause of incidental abnormal liver blood tests, accounting for between 60–90% of such cases [44, 45]. Importantly, the vast majority (around 80%) of patients with NAFLD have normal liver blood tests [2], and there is no difference in histological severity between those with and without abnormal tests [46]. Accordingly, NAFLD should be suspected and sought in all patients with established risk factors, including PCOS and OSA, regardless of liver blood tests. The history should concentrate on determining the presence/absence of conditions commonly associated with “primary” NAFLD – metabolic syndrome components, cardiovascular disease and OSA – and on excluding alternative causes of steatosis, including excessive alcohol intake, previous abdominal surgery (leading to bacterial overgrowth) and drugs causing NAFLD, such as amiodarone and tamoxifen. On examination, most patients are centrally obese and dorsocervical lipohypertrophy (a “buffalo hump”) appears to be a particular feature of the fat distribution in patients with advanced NAFLD [8]. Features of PCOS (hyperandrogenism) should be sought in young women with suspected NAFLD [39], and clinical evidence of lipodystrophy should be sought in young, non-obese patients in view of its association with NAFLD [47].
Investigation
In the absence of advanced disease routine liver blood tests are either normal or typically show mild elevations of transaminases, alkaline phosphatase and gamma glutamyl transpeptidase (GGT) 1.5–3 times the upper limit of normal. The ALT/AST ratio is greater than one unless there is advanced fibrotic NAFLD or the patient is a covert heavy drinker. Other blood tests are aimed at detecting associated conditions, such as dyslipidemia, and excluding alternative causes of abnormal liver blood tests. Regarding lipids, it is worth measuring serum levels of apolipoprotein B (apoB) in patients either with no obvious risk factors for NAFLD or with low levels of LDL and HDL cholesterol, looking for evidence of hypobet alipoproteinemia a rare, familial cause of NAFLD [48]. Serum ferritin is often raised in NAFLD patients [49], and has been associated with advanced fibrosis [50]. HFE genotyping should be carried out when hyperferritemia is accompanied by raised trans-ferrin saturation. Autoantibodies associated with autoimmune hepatitis (AIH), including ANA SMA, are often present at low titers in patients with NAFLD and have been associated with more advanced disease in some, but not all studies [51, 52]. Around one in ten of these patients have histological features of autoimmune hepatitis on biopsy and fulfil diagnostic criteria for probable/definite AIH [51]. Currently available imaging modalities including ultrasound, CT and routine MR imaging are all excellent at detecting steatosis (once more than around one-third of the liver is affected) but none can reliably detect NASH or fibrosis [53]. Newer imaging techniques, including proton magnetic resonance spectroscopy [54], and transient elastography [55] show promise (particularly in children [56]), but require further study prior to routine use for disease staging.
The role of liver biopsy
Undoubtedly the most important and controversial issue to consider in the investigation of patients with suspected NAFLD is whether or not to perform a liver biopsy. For diagnosis, biopsy is not required in a “typical” patient with abnormal liver blood tests, classical risk factors for NAFLD (obesity, T2DM, dyslipidemia) and an ultrasound showing steatosis; however, a high ferritin with HFE mutations, positive autoantibodies (ANA, SMA) or the use of medications associated with drug-induced liver injury all may justify a biopsy to exclude alternative/additional diagnoses. The main indication to perform a biopsy is, however, the accurate staging of the disease since (1) different stages have different prognoses and therefore require different management strategies, and (2) no currently availably imaging techniques can perform this role [53].
Non-invasive markers for staging NAFLD
The current reliance on liver biopsy for disease staging has prompted many studies aimed at defining clinical or laboratory-based variables capable of acting as surrogate markers of disease stage [57]. Various clinical and laboratory markers have been shown to be associated with advanced fibrosis (bridging fibrosis or cirrhosis) in patients with NAFLD, notably advanced age (>45 years), BMI > 30kg/m2, central obesity, T2DM (or raised fasting blood glucose), the severity of OSA [42] an AST:ALT ratio greater than one, hyperferritinemia [50] and positive autoantibodies [45]. In patients with normal ALT increased insulin resistance is the best predictor of advanced fibrotic disease [58]. At present, it would seem reasonable to restrict liver biopsy to patients with at least some, if not all, of these risk factors. Some of these markers (age, BMI, T2DM, AST/ ALT ratio) have recently been combined together with platelet count and serum albumin concentration, into a NAFLD fibrosis “score” that accurately predicts the presence or absence of advanced fibrosis in the majority of patients with NAFLD [59]. This score has recently been combined with the European Liver Fibrosis (ELF) panel of serum fibrosis markers [60] and shown to have an accuracy of over 90% in differentiating the fibrosis stages in NAFLD [61]. An even simpler score – the BARD score – combining BMI, AST/ALT Ratio and D
Related posts:
Stay updated, free articles. Join our Telegram channel
Full access? Get Clinical Tree
