Year: 2007
Type: article
Abstract: Nonalcoholic fatty liver disease (NAFLD) is now recognized as the commonest liver disorder in most industrialized countries, and affects 15–25% of many Asian populations.1 It is a metabolic liver disorder characterized by histological similarity to alcoholic liver disease in persons without excessive alcohol intake, and the pathological spectrum includes hepatic steatosis, nonalcoholic steatohepatitis (NASH) (steatosis plus hepatic inflammation, hepatocyte injury and fibrosis) and cirrhosis.2 Hepatic steatosis alone is a non-progressive condition,3 but up to one-third of individuals with NASH show progression in hepatic fibrosis.4 Individuals with NASH-related cirrhosis face a worrying outcome,5 with approximately one in 10 dying from liver failure over 7 years. In population-based studies of all types of NAFLD, rates of liver-related mortality are much lower (∼2%),6 but liver disease may rank as highly as the third most common cause of death.6 The pathogenesis of NAFLD has been conceptualized as hepatic fat accumulation as the first step, followed by superimposed secondary insult(s), the "two hit" hypothesis.7 Among the latter, diverse processes ranging from oxidative stress, impaired mitochondrial function through to genetically determined differences in inflammatory recruitment and metabolic disposal pathways have been implicated. However, with the understanding that many of these processes (e.g. oxidative stress) could qualify equally as a "primary" or a "secondary" hit, NAFLD/NASH is now viewed as the culmination of multiple interactive rather than sequential pathways.2 Current treatment strategies against NAFLD and NASH are directed at individual pathways considered critical in promoting liver injury. Apart from attempts to reverse steatosis by lifestyle and weight-reducing measures, these include attempts at reducing oxidative stress (vitamin E, glutathione precursors such as betaine, N-acetylcysteine), improving insulin sensitivity (metformin, thiazolidinediones [TZD]) or by promoting intrahepatic fatty acid β-oxidation by using peroxisomal proliferator-activated receptor (PPAR-α) agonists (bezafibrate). Others have directed their efforts towards reducing bacterial endotoxemia (antimicrobials), blocking triglyceride absorption (Orlistat, a gastric lipase inhibitor), inhibiting renin-angiotensin activation (Losartan) or by simply focusing on management of the other components of the associated metabolic syndrome (hyperlipidemia and type 2 diabetes). Unfortunately, the studies of NASH treatment to date have been plagued with methodological problems: small sample sizes, non-randomized or uncontrolled observations, absence of post-treatment histology and short follow up.8 A particularly crucial issue is the interpretation of uncontrolled "positive" studies because of the impressive responses in placebo recipients9 and also the lack of correlation between changes in serum enzyme levels and liver histology. To date, level I evidence is available only for bariatric surgery (which is effective)10 and for ursodeoxycholic acid, which is not different from placebo.9 Of the drugs available, insulin-sensitizing agents (TZD, metformin) are regarded as leading contenders in the field.8,11 However, the overall experience with metformin is yet to justify its routine use. In a systematic review of nine studies with metformin, most involved small sample sizes (median, n = 18).12 Although insulin resistance indices improved, serum transaminases decreased in only a third and reduced hepatic inflammatory activity was noted in only one trial.12 In another recent study reported in abstract form, only those patients taking metformin who lost weight had improvement in steatosis.13 Trials with TZD have shown more impressive reduction in hepatic steatosis, necroinflammatory activity and fibrosis. However, weight gain (up to 4% of baseline body weight), lack of sustained response after drug withdrawal, adverse effects and high cost are problems that need addressing. The fact that none of the drugs tested so far satisfy strict levels of evidence to validate their use mandates a search for alternative approaches to treatment. One such strategy may be targeting tumor necrosis factor (TNF)-α. Cytokine mobilization, particularly TNF-α, is one of many processes implicated in hepatic inflammatory cell recruitment in NASH,2 the key step in initiation and perpetuation of liver injury. Further, by impairing insulin receptor signaling, TNF-α can also reduce insulin sensitivity, thereby causing or worsening insulin resistance, a key part of the pathophysiology of NAFLD/NASH. Some investigators have invoked a pivotal role for this cytokine in NASH because serum TNF-α is increased in human and experimental NAFLD,14 although at least one study found no difference between steatosis and NASH.14 Further, there are now reports of experimental steatohepatitis occurring in mice with metabolic syndrome in the absence of TNF-α receptor signaling.15,16 These observations cast doubt on the essential relevance of TNF-α as a primary mediator of inflammatory cell recruitment in NASH. However TNF-α, like many other cytokines, has multiple effects and its influence on other pathophysiological processes such as insulin signaling cannot be discounted. Pentoxifylline, a phosphodiesterase inhibitor, is an agent that has been shown to attenuate the production of TNF-α and other proinflammatory cytokines.17 Treatment with pentoxifylline reduced mortality in alcoholic hepatitis,18 and it also has antioxidant and antifibrogenic properties19 that could be of value against NASH. Further, pentoxifylline attenuated the hepatic inflammatory response in the methionine-choline deficient murine model of steatohepatitis.20 Data on pentoxifylline in human NASH are limited to three studies.21–23 In all studies, pentoxifylline treatment reduced serum transaminases but histological follow up was not reported. In this issue of the Journal, Satpathy et al. have extended their earlier work to evaluate the histological impact of a 12-month course of pentoxifylline in nine patients with NASH.21 There were no placebo-treated or untreated controls. Significant improvement in scores for hepatic steatosis necroinflammatory activity (five of nine patients) and reduction in the stage of fibrosis (four of six patients with baseline fibrosis) were observed in repeat liver biopsies performed after 6–12 months of treatment. Possible explanations for these results could be a reduction in TNF-α levels or improved insulin sensitivity. TNF-α levels were not measured by the authors in the present study but a decline in TNF-α levels had been observed in their earlier report.21 A reduction in HOMA-IR, an index of insulin resistance, was documented in three of four patients. There was a small non-significant reduction in body mass index but there was no correlation between serum transaminase decline and weight change or waist circumference. Should we be initiating pentoxifylline therapy for patients with NAFLD? Not outside clinical trials. Both this and the study by Adams et al. were non-randomized, open-labeled trials involving small samples (9 and 20 subjects, respectively).21,22 The third study was a small randomized trial involving a total of 20 patients and is available only in abstract form.23 Nine of 20 (45%) patients in one of these studies withdrew due to nausea; a higher dose was used in this study (400 mg t.d.s.). The safety of the drug should be clearly established in patients with cirrhosis because there is an eightfold increase in serum levels of pentoxifylline and its metabolites.17 Therefore, at the present time, these should be regarded as proof-of-concept rather than definitive studies, but the findings could pave the way for larger trials should less toxic approaches to NASH therapy prove ineffective. These approaches, which are based on lifestyle adjustments to effect modest weight reduction through diet and exercise, and monitoring and treatment of associated metabolic disorders (obesity, type 2 diabetes, hyperlipidemia, hypertension) have been reviewed by the Asia–Pacific Working Party on NAFLD, and will be published in a forthcoming issue of the Journal.24
Authors: Shivakumar Chitturi, Geoffrey C Farrell
Institutions: Canberra Hospital, ACT Government
Language: English
Subfield: Epidemiology
Field: Medicine
Domain: Health Sciences
Open Access status: bronze
OpenAlex ID: W2130365746