Clear messages from sonographic shadows? Links between metabolic disorders and liver disease, and what to do about them

See article in J. Gastroenterol. Hepatol. 2003; 18: 588–94. The concept of syndrome X for the clustering of cardiovascular risk factors including hypertension, obesity, high serum triglyceride and low high-density lipoprotein (HDL) was introduced in 1988.1 Since then, several other synonyms including the insulin resistance syndrome, dysmetabolic syndrome, cardiovascular metabolic syndrome, and the deadly quartet, have been coined.2 In 1998, the World Health Organization proposed a unifying definition, using the term ‘metabolic syndrome’.3 This syndrome consists of two of the following conditions: (i) hypertension, (ii) central obesity, (iii) hypertriglyceridemia or low HDL cholesterol, and (iv) microalbuminuria. It occurs in the setting of type 2 diabetes, impaired glucose regulation or insulin resistance. Impaired glucose regulation is a metabolic state intermediate between normal glucose homeostasis and diabetes, and is diagnosed by either impaired fasting glucose (fasting glucose ≥6.1 and <7.0) or impaired glucose tolerance (postprandial hyperglycemia assessed by the oral glucose tolerance test). In those with normal glucose tolerance, subclinical insulin resistance may be identified by the hyperinsulinemic euglycemic clamp study and is defined as the lowest quartile of insulin sensitivity for the background population under investigation. However, there is not yet universal agreement on this definition of the metabolic syndrome and alternative definitions have been proposed.4,5 It is now clear that non-alcoholic fatty liver disease (NAFLD) should be included as one of the associated abnormalities of the metabolic syndrome, as confirmed by Angelico et al. in issue 5 of the Journal.6 The authors used the readily available technique of hepatic ultrasonography to identify subjects attending liver and metabolic outpatient clinics who had increased hepatic echogenicity as a sign of hepatic steatosis. Those drinking more than three standard drinks (>30 g ethanol per day) were excluded. They showed that features of the metabolic syndrome including obesity (body mass index ≥30), hypertension, impaired glucose tolerance or type 2 diabetes and hypertriglyceridemia were com-mon findings in subjects with increased hepatic echogenicity, particularly those with a raised alanine aminotransferase (ALT) level. Of particular interest, the authors were able to demonstrate that an increase in the severity of hepatic echogenicity was associated with the presence of a greater number of features of the metabolic syndrome. It is now generally accepted from multiple studies that insulin resistance is the common link between NAFLD and other features of the metabolic syndrome. In particular, recent studies have demonstrated a high prevalence of insulin resistance and the metabolic syndrome in patients with biopsy-proven non-alcoholic steatohepatitis (NASH).7 Insulin resistance appears to be the result of a complex interplay of genetic and environmental factors, the latter include overnutrition and physical inactivity, with resultant central (visceral) obesity.8 Normally, insulin inhibits the hormone-sensitive lipase, which is an important control point for the release of free fatty acids from adipocytes.9 During states of insulin resistance, increased lipolysis results in enhanced release and transport of free fatty acids from adipose tissue to the liver, where fatty acids provide a substrate for triglyceride synthesis and hepatic steatosis. Hyperinsulinemia, which is the natural corollary of insulin resistance, may directly inhibit mitochondrial β-oxidation of free fatty acids in the liver, leading to further hepatic steatosis.10 A ‘two hit’ hypothesis has been proposed to describe a perpetuating injury mechanism such as oxidative stress/lipid peroxidation, cytokine induction and mitochondrial dysfunction, which converts bland steatosis (‘first hit’) into fibrosing steatohepatitis.11 Diabetes and insulin resistance are risk factors for hepatic fibrosis,12–14 in addition to being pro-steatotic. This suggests that, as well as their undoubted importance to the pathogenesis of hepatic steatosis, they may also be involved in the second hit process. This may be mediated via pro-inflammatory cytokines such as tumor necrosis factor (TNF), which can be released by adipose tissue and is increased in insulin-resistant subjects.15,16 Increased TNF expression also occurs in NASH.17 In the diagnosis of NAFLD, Angelico et al. used ultrasonographic findings in the absence of excessive alcohol intake and viral markers. This has several limitations. First, increased hepatic echogenicity is not specific for steatosis; hepatic fibrosis can give similar sonographic changes, as emphasized by the term ‘fatty-fibrotic pattern’.18 Second, it is now quite clear that degree of steatosis does not correlate with the extent of fibrosis in NAFLD.12 Indeed, steatosis can disappear as NASH progresses to cirrhosis.19 Hence, the finding of mild hepatic steatosis on ultrasonography does not provide any reliable evidence for a pathologically early stage of NAFLD.20 Further, ultrasound and other forms of hepatic imaging (computerized tomography, magnetic resonance imaging) cannot distinguish bland steatosis from steatohepatitis and fibrosis.20 This distinction is clinically important because bland steatosis has a benign prognosis, whereas steatohepatitis can progress to cirrhosis and liver failure.21 Liver biopsy remains the only means to distinguish the different stages of NAFLD. The influence of steatohepatitis on a patient's prognosis was demonstrated in a large population study of diabetic patients that found the standardized mortality rates for liver disease to be even greater than that for cardiovascular disease.22 Another definitional aspect highlighted by the study of Angelico et al. is the inclusion of subjects with an average daily alcohol consumption of up to 30 g per day. There is no consensus regarding the amount of alcohol intake that may contribute to liver injury.23 Alcohol intake as low as 20 g per day can cause hepatic steatosis.24 The risk of alcoholic liver disease and alcohol-induced cirrhosis increases in women with an alcohol intake as low as 10–20 g per day and in men of 20–30 g per day.25 Although clinical practice guidelines or consensus definitions are lacking in the area of NASH/NAFLD, a current USA convention is to include only those who drink <20 g ethanol per day, while others (including ourselves) prefer even more stringent guidelines (<40 g per week) in order to avoid inclusion of possible cases of alcohol/metabolic factor interactions. The study by Angelico et al. used the homeostasis model (HOMA) index, a surrogate marker of insulin resistance that has been validated against the euglycemic hyperinsulinemic clamp study. An increased HOMA index is a common frequent finding in NASH and is associated with an increased risk of cardiovascular events.7,26,27 Because NASH is often the first clinical manifestation of insulin resistance,28 the identification of hepatic steatosis on ultrasonography should prompt clinicians to search for associations with diabetes mellitus, high blood pressure and other features of the metabolic syndrome. This is particularly important as large population studies have shown that up to 50% of those with diabetes are undiagnosed.29 Of those with undiagnosed diabetes, half have fasting plasma glucose concentrations in the non-diabetic range, so it is important that the oral glucose tolerance test be routinely performed in subjects suspected of being insulin resistant. How else should the clinician intervene in patients with hepatic steatosis diagnosed on ultrasound? Short-term studies with lifestyle intervention have shown that exercise (aerobic activity, like walking) and dietary modification reduce hepatic steatosis and correct abnormal biochemical results in NASH.30,31 Importantly, these measures are now proven to reduce the long-term risk of diabetes mellitus levels,32 but any long-term benefit of preventing the progression of NAFLD is yet to be proven. In diabetic patients, weight loss as little as 5% has the benefit of reducing fasting blood glucose, insulin and hemoglobin A1c levels.33 In contrast, rapid and profound weight loss can exacerbate inflammation and fibrosis in NASH and should be avoided.34 Non-alcoholic fatty liver disease affects a large part of the population, in Europe, North America and in the Asia–Pacific region. It is commonly identified on hepatic ultrasound, and should now be considered as a manifestation of the metabolic syndrome. Morbidity and mortality from NASH are increasingly being recognized. An argument can be made for submitting selected patients suspected to have NASH to liver biopsy; the findings can be used to reinforce attempts at correcting the underlying metabolic abnormalities that likely cause progressive liver disease in the metabolic syndrome. Others would argue that, in the absence of evidence that treatment (lifestyle adjustments or pharmacological measures) improves the long-term outcome of the liver condition, it can be debated that performing a liver biopsy in those who appear to be affected by NAFLD is an unwarranted intrusion. The evidence is mounting that steatosis associated with raised ALT, and particularly in persons with risk factors for fibrotic severity such as obesity and diabetes (as reinforced by the Angelico et al. study), are harbingers of a significant form of liver disease (often cirrhosis). These are disorders for which closer monitoring by gastroenterologists may potentially offer improved outcomes.