Toxicant-associated Steatohepatitis

Abstract

Hepatotoxicity is the most common organ injury due to occupational and environmental exposures to industrial chemicals. A wide range of liver pathologies ranging from necrosis to cancer have been observed following chemical exposures both in humans and in animal models. Toxicant-associated fatty liver disease (TAFLD) is a recently named form of liver injury pathologically similar to alcoholic liver disease (ALD) and nonalcoholic fatty liver disease (NAFLD). Toxicant-associated steatohepatitis (TASH) is a more severe form of TAFLD characterized by hepatic steatosis, inflammatory infiltrate, and in some cases, fibrosis. While subjects with TASH have exposures to industrial chemicals, such as vinyl chloride, they do not have traditional risk factors for fatty liver such as significant alcohol consumption or obesity. Conventional biomarkers of hepatotoxicity including serum alanine aminotransferase activity may be normal in TASH, making screening problematic. This article examines selected chemical exposures associated with TAFLD in human subjects or animal models and concisely reviews the closely related NAFLD and ALD.

Introduction

The liver is the largest organ in the body, weighing approximately 1.5 kg in adults, and is the most complex organ in terms of metabolism. Hepatocytes make up over 80% of total liver mass and play a critical role in the metabolism of amino acids and ammonia, biochemical oxidation reactions, and detoxification of a variety of drugs, vitamins, hormones, and environmental toxicants. Kupffer cells represent the largest reservoir of fixed macrophages in the body. They play a protective role against gut-derived toxins that have escaped into the portal circulation, and they are a major producer of cytokines, which can influence the toxicity of environmental toxicants. Other specific cell types also have unique functions (e.g., bile duct epithelium in bile flow; sinusoidal endothelial cells in adhesion molecule expression and endocytosis; stellate cells in vitamin A storage and fibrosis). Due to its role as a crucial organ in the first line of defense, the liver also appears to be the most common target organ damaged by industrial chemicals.

As of May 2011, over 60 million unique chemicals were registered with the Chemical Abstracts Service Registry. Notably, only 2 years have passed between the registration of the 50 millionth chemical and the 60 millionth chemical. Due to rapid discovery and commercialization, it is impossible to fully understand the potential impact of these new chemicals on liver disease. However, the problem appears significant because 33% of the 677 most common workplace chemicals reported in the National Institute of Occupational Safety and Health Pocket Guide are associated with hepatotoxicity (Tolman and Sirrine 1998). Occupational and environmental liver diseases may present with a wide clinical spectrum ranging from asymptomatic liver enzyme elevation to acute liver failure, cirrhosis, and cancer. In addition to the exposure level, an individual’s susceptibility to chemical-induced liver disease is determined by polymorphisms in the genes of xenobiotic metabolism, concomitant use of alcohol or prescription medications, nutritional factors, and obesity—as many organic chemicals are lipid soluble (Hsieh et al. 2007; Mastrangelo et al. 2004). High-level vinyl chloride (VC; Cave, Falkner, et al. 2010) and solvent exposures (Brautbar and Williams 2002) have historically been associated with occupational liver disease, while classically described mediators of environmental liver disease include aflatoxins (Kensler et al. 2011) analines (Epping Jaundice; Kopelman, Scheurer, and Williams 1966; Nichols 2004), and toxic oil syndrome (Gelpi et al. 2002)).

Most industrial chemicals, including solvents, exhibit dose-dependent hepatocyte cytotoxicity and typically result in hepatocellular necrosis (Cave et al. 2011). However, both cholestasis (Farrell 1994; Zimmerman 1999) and acute hepatitis with rash and eosinophilia resembling an immunologic drug reaction have been reported (Brautbar and Williams 2002). Examples of histopathologic liver lesions associated with industrial chemicals are provided in Table 1.

Table 1.

Pathology of chemical liver disease in humans or animal models.

Category	Selected examples
Steatohepatitis	Vinyl chloride, aflatoxin, trichloroethylene, tetrachloroethylene, trichloroethane, carbon tetrachloride, petrochemical mixtures,   atrazine, paraquat, chlordecone, polychlorinated biphenyls, nitrobenzene, nitrotoluene, arsenic, methylmercury, thallium,   yellow phosphorus, dioxin, lead, 2-nitropropane, N,N-dimethylformamide, chloroform
Necrosis	Carbon tetrachloride and other halogenated aliphatic hydrocarbons, haloaromatic compounds, nitroaromatic compounds, arsenic,   yellow phosphorus, acetaminophen
Cholestasis	Beryllium, copper, di(2-ethylhexyl) phthalate, methylenedianiline, paraquat, toxic rapeseed oil
Cirrhosis	Arsenic, carbon tetrachloride, polychlorinated biphenyls, trichloroethane, trichloroethylene, trinitrotoluene, vinyl chloride
Peliosis hepatitis	Thorotrast, urethane, vinyl chloride
Autoimmune hepatitis	Trichloroethylene, trinitrotoluene
Granulomas	Beryllium, copper
Pigment deposition	Anthracite, thorotrast, titanium
Cholangiocarcinoma	Thorotrast, polychlorinated biphenyls
Hepatocellular Carcinoma	Arsenic, carbon tetrachloride, polychlorinated biphenyls, tetrachloroethylene, thorotrast, trichloroethylene, trinitrotoluene,   vinyl chloride
Hemangiosarcoma	Vinyl chloride, butoxyethanol, chloronitrobenzene, polyhexamethylene biguanine, urethane, tetrafluoroethylene, thorotrast

Adapted from Cave, M., K. C. Falkner, and C. J. McClain, Occupational and Environmental Liver Disease. In Zakim and Boyer’s Hepatology: A Textbook of Liver Disease. 6th edn. Philadelphia. Elsevier Saunders; 2011

Recently, our group described toxicant-associated steatohepatitis (TASH) occurring in highly exposed chemical workers and associated with increased proinflammatory cytokines, insulin resistance, and antioxidant depletion (Cave, Falkner, and McClain 2011; Cave, Falkner, et al. 2010). Liver biopsies resembled alcoholic (ASH) or nonalcoholic steatohepatitis (NASH), but affected workers were nondrinkers and were not obese. Hepatocyte lipid accumulation has historically been considered to be a benign toxicologic finding and was described in many forms of chemical hepatitis (Cave et al. 2011; Cave, Falkner, et al. 2010). However, given the increase in cirrhosis and hepatocellular carcinoma (HCC) associated with the current NASH epidemic, the significance of steatohepatitis due to industrial chemicals must be reevaluated. Indeed, emerging evidence suggests that steatohepatitis, often occurring with normal serum aminotransferases, appears to be a common, pathophysiological finding, and has been described in many forms of chemical hepaptotoxicity (Brautbar and Williams 2002; Cave, Falkner, and McClain 2011; Cave, Falkner, et al. 2010; Cotrim et al. 2004). Serum intact cytokeratin 18, a biomarker of necrotic hepatocyte death, is generally elevated in these workers (Cave, Appana, et al. 2011; Cave, Falkner, et al. 2010). Recently, persistent organic pollutants (POPs), which accumulate in tissues with age and include polychlorinated biphenyls (PCBs), have been associated with suspected TASH in epidemiologic studies (Cave, Appana, et al. 2010). Nutritional status, co-exposures, and obesity appear to confer increased susceptibility to TASH but more clinical data are needed (Cave et al. 2007; Cave, Falkner, et al. 2010; Mastrangelo et al. 2004). More recently, the term toxicant-associated fatty liver disease (TAFLD) has been proposed to more clearly define the spectrum of fatty liver injury following xenobiotic exposure (Schwingel et al. 2011). The purpose of this article is to review steatohepatitis etiologies and mechanisms with an emphasis on TAFLD/TASH.

Steatohepatitis Overview

Steatohepatitis occurs in susceptible individuals following exposures to overnutrition (NASH), excessive ethanol consumption (ASH, alcoholic steatohepatitis), cancer chemotherapies (chemotherapy-associated steatohepatitis, or CASH; Tannapfel, Reinacher-Schick, and Flott-Rahmel 2011), or industrial chemicals (toxicants—TASH). Regardless of the etiology, the steatohepatitis is pathologically indistinguishable. While many roads lead to steatohepatitis, different exposure-specific mechanisms have been implicated. However, some mechanisms such as insulin resistance and proinflammatory cytokine elevation appear to be conserved. Before addressing the more recently described TASH, it is first appropriate to concisely review the more thoroughly investigated nonalcoholic fatty liver disease (NAFLD) and alcoholic liver disease (ALD). Based on pathologic and in some cases mechanistic similarities, we believe that these extensively studied diseases provide a framework from which to understand the pathologic findings, mechanisms, and potential treatments for TASH.

Nonalcoholic Steatohepatitis

Fatty liver has been recognized to accompany obesity since at least the 1950s. However, it was not until 1980 when Ludwig et al. coined the term nonalcoholic steatohepatitis to describe this previously unnamed condition, often occurring in cirrhotic patients, that its clinical importance became well recognized (Ludwig et al. 1980). NAFLD encompasses a pathologic spectrum of liver disease ranging from steatosis, to steatohepatitis, cirrhosis, and HCC. The pathologic findings in NAFLD including clinical and research grading and staging systems have been recently reviewed (Aly and Kleiner 2011; Kleiner and Brunt 2012). Steatosis is defined as macro- or microvesicular triglyceride accumulation involving at least 5% of hepatocytes on microscopy (Brunt et al. 1999; Kleiner et al. 2005). The typical adult pattern of steatohepatitis involves centrilobular (zone 3) centered injury with lobular inflammation (predominantly lymphocytes but also with neutrophils and activated Kupffer cells), hepatocyte ballooning (often with Mallory-Denk bodies) with or without fibrosis (Kleiner and Brunt 2012; Figure 1). Apoptotic hepatocytes may be noted in liver biopsies from NASH patients and detected noninvasively as serum caspase cleaved cytokeratin 18 fragments (Feldstein et al. 2009; Kleiner and Brunt 2012). The NAFLD activity score (NAS) is a semiquantitative scoring system based on the degree of steatosis, lobular inflammation, and ballooning used for grading activity in serial biopsies in patients enrolled in NASH clinical trials. Thus, the NAS is a research tool that may complement traditional NAFLD grading and staging systems (Aly and Kleiner 2011; Kleiner and Brunt 2012). NAFLD is regarded by most investigators as the hepatic manifestation of obesity and the metabolic syndrome, and its prevalence has risen correspondingly with the obesity epidemic. NAFLD is the most prevalent liver disease in the United States, if not worldwide. The unselected U.S. adult NAFLD prevalence varies from 10 to 35%, depending on diagnostic algorithm and study population (Vernon, Baranova, and Younossi 2011). A frequently cited unselected adult study from Dallas, Texas, determined a 31% NAFLD prevalence by magnetic resonance spectroscopy (Browning et al. 2004). However, the prevalence is much higher in selected subjects with metabolic risk factors and may exceed 90% in extremely obese subjects undergoing bariatric surgery (Vernon, Baranova, and Younossi 2011). Worldwide, NAFLD prevalence is similar to that observed in the United States, although NAFLD frequently occurs in the absence of obesity in Asia (Vernon, Baranova, and Younossi 2011). Unlike isolated hepatic steatosis, NASH may result in progressive liver disease, including cirrhosis with hepatic decompensation and HCC resulting in death or transplantation (Bhala et al. 2011). The U.S. unselected prevalence of NASH is estimated to be 2 to 5%, and progression occurs more commonly in subjects with insulin resistance (Vernon, Baranova, and Younossi 2011). While liver-related deaths are increased by NAFLD, cardiovascular mortality is the leading cause of death in NAFLD (Adams et al. 2005); and NAFLD is an independent risk factor for cardiovascular disease (Bhatia et al. 2012).

Figure 1.

Photomicrograph of a liver biopsy from an obese adult human subject with nonalcoholic steatohepatitis showing macrovesicular steatosis, inflammatory infiltrate, and hepatocyte ballooning with Mallory-Denk bodies (arrows, hematoxylin–eosin stain, 100×).

Hepatic steatosis, the initial manifestation of NAFLD, can lead to inflammatory responses in the liver (steatohepatitis), which then progresses to fibrosis and cirrhosis. Obesity and insulin resistance usually accompany NAFLD and may be brought upon by altered adipocytokine levels. No animal model completely recapitulates obesity, insulin resistance, steatosis, inflammation, and fibrosis as observed in humans with NASH. Most investigators have used dietary, genetic, or combination mouse models (Takahashi, Soejima, and Fukusato 2012). Our group has utilized many of the nutritional approaches (high-fat, high-fructose, methionine-, and choline-deficient diets) in our mechanistic studies of NAFLD (Figure 2; Bergheim et al. 2008; Kirpich et al. 2011; Shi et al. 2012; Song et al. 2012). Nonetheless, animal models and human studies have enabled mechanistic understanding of NAFLD. These mechanisms include insulin resistance with a decreased adiponectin–leptin ratio, proinflammatory cytokine elevation, oxidative stress with reduced antioxidant levels, mitochondrial dysfunction, metabolic endotoxemia, altered transmethylation, and lipotoxicity with hepatocyte apoptosis (Cave et al. 2007; Frazier, DiBaise, and McClain 2011; Ibrahim, Kohli, and Gores 2011). Recently, a genome-wide association study identified a polymorphism (Patatin-like phospholipase domain-containing protein 3) which was identified as an NAFLD susceptibility factor (Rotman et al. 2010). There are no Food and Drug Administration (FDA)-approved medications for NASH, and currently supported treatment options include weight loss, vitamin E supplementation, and insulin sensitizers (Chalasani et al. 2012).

Figure 2.

Liver photomicrograph from a C57/Bl6 mouse fed a Western diet for 12 weeks, demonstrating small and large dropletmacrovesicular steatosis and foci of inflammatory infiltrates (hematoxylin–eosin stain, 40×).

Importantly, NAFLD may increase susceptibility to liver injury from xenobiotic compounds. The weight of evidence indicates that NAFLD is associated with downregulation of CYP1A2 and cellular glutathione (GSH; Merrell and Cherrington 2011). CYP2E1 is upregulated in the majority of human and rat, but not mouse, studies. Limited human data suggest that specific hepatic uptake transporters may also be downregulated (Lake et al. 2011), while adenosine triphosphate (ATP)-binding cassette (efflux) transporters may be upregulated (albeit with abnormal cellular localization in some cases; Hardwick et al. 2011) with increasing severity of NAFLD. Some toxicants which produce TASH, such as VC, are metabolized by CYP2E1 into reactive intermediates, while others, such as PCBs, have been shown to deplete GSH (Shi et al. 2012). Therefore, we hypothesize that NAFLD could predispose to the development of TASH by altering xenobiotic metabolism and antioxidant status.

Alcoholic Hepatitis

The consumption of alcohol is customary in most cultures, and alcohol abuse remains common worldwide. In the United States alone, more than 50% of Americans consume alcohol, with an estimated 23.1% of Americans participating in heavy and/or binge drinking at least once a month (U.S. Department of Health and Human Services 2010). The incidence of ALD parallels average per capita consumption. In 2005, the worldwide per capita consumption of alcoholic beverages was 6.13 L of pure alcohol consumed by an individual aged 15 years or older (World Health Organization [WHO] 2011). The highest alcohol consumption levels are seen in developed countries, mainly in Eastern Europe/Europe and North America (Table 2). Medium alcohol intake is observed in southern Africa as well as certain countries in North and South America. The least alcohol consumption levels can be found in southern Asian and Middle Eastern countries where there are high rates of lifetime alcohol abstinence. Chronic alcoholism is a cause/risk factor in over 60 major types of diseases, most notably liver disease. The cost of excessive alcohol consumption was projected to be about $223.5 billion in the United States in 2006, making it a significant financial burden for the health care system.

Table 2.

Adult (15+) per capita consumption (in liters of pure alcohol) for countries categorized either in the high, moderate, or low alcohol consumption level.

Country	Adult (15+) per capita consumption (in liters of pure alcohol)	Lifetime abstainers	Mortality (age standardized death rates from liver cirrhosis)		Morbidity (Alcohol use disorders)
			Males	Females	Males	Females
Russia	15.7	21.0%	-	-	16.29%	2.58%
Republic of Korea	14.8	12.8%	38.4	7	13.10%	0.41%
France	13.7	2.6%	18.2	6.5	4.54%	1.07%
Germany	12.8	1.7%	24.2	9.5	4.51%	0.88%
New Zealand	9.6	9.3%	4.0	2.0	3.50%	2.20%
South Africa	9.5	65.2%	10.6	3.9	3.64%	0.88%
United States	9.4	17.7%	13.5	6.1	5.48%	1.92%
Brazil	9.2	18.7%	25.0	4.8	7.29%	1.41%
India	2.6	79.2%	-	-	3.47%	0.42%
Egypt	0.37	75.5%	-	-	0.44%	0.00%

The liver is the main site of alcohol metabolism and a major target of alcohol-induced organ injury. The susceptibility of the liver to alcohol-induced toxicity is due to both the high concentrations of alcohol found in the portal blood (versus systemic) and the metabolic consequences of ethanol metabolism. Like NAFLD, ALD is a spectrum of diseases ranging from steatosis, to inflammation and necrosis (ASH), to fibrosis and cirrhosis, and eventually HCC. Steatosis develops in 90% of individuals who drink more than 16 g of alcohol/day (Crabb 1999) but resolves upon cessation of alcohol consumption (Bergheim, McClain, and Arteel 2005). Steatosis sensitizes the liver to injury caused by a second insult (Day and James 1998) and may progress to steatohepatitis. According to this “two-hit hypothesis,” alcohol consumption alone does not cause progression of ALD from steatosis to steatohepatitis, or fibrosis; rather, a second insult is required for the development of later stages of disease. Such a second insult could include an inflammatory response, reactive oxygen species, or hypoxia, among others. For example, ethanol exposure is well known to enhance liver pathology induced by xenobiotics, such as bacterial cell wall products (e.g., lipopolysaccharide [LPS]) or carbon tetrachloride. Furthermore, alcohol consumption itself could provide the second hit; for example, an episode of binge drinking can trigger entry into acute alcoholic hepatitis in a chronic alcoholic (Barrio et al. 2004; Rivara et al. 1993). In later stages of ALD, collagen deposition and regenerative nodules can result in the development of fibrosis and cirrhosis, respectively. Animal models recapitulate human ALD (Figure 3). Abstinence from alcohol is beneficial for patients in all stages of ALD and is necessary to prevent progression of liver injury in those with early stages of disease (Bergheim, McClain, and Arteel 2005).

Figure 3.

Liver photomicrograph from a C57/Bl6 mouse administered ethanol (28 g/kg/day) for 4 weeks, demonstrating macrovesicular steatosis and foci of inflammatory infiltrates (hematoxylin–eosin stain, 40×).

The two-hit hypothesis is mirrored on a molecular level by the concepts of “priming” and “sensitization,” which are now considered to be fundamental to alcohol-induced liver injury (Tsukamoto et al. 2001). Here, priming refers to the ability of ethanol pre-exposure to cause inflammation of the liver cells (e.g., Kupffer cells) to increase the release of proinflammatory cytokines in response to a second stimulus, such as LPS. Alcoholic hepatitis patients have larger amounts of circulating tumor necrosis factor-alpha (TNF-α) both basally and after stimulus (McClain and Cohen 1989). Additionally, ethanol preexposure can prime Kupffer cells to LPS stimulation, resulting in enhanced TNF-α release (Enomoto et al. 1998). Ethanol can also sensitize cells, causing cell populations downstream of inflammatory cytokine signaling (e.g., hepatocytes) to respond more robustly. Specifically, the priming of inflammatory cells by ethanol leads to a more robust cell-killing response that is increased in sensitized hepatocytes, explaining why blocking the activation of Kupffer cells (Adachi et al. 1994) or employing knockouts with an impaired Kupffer cell response (Kono et al. 2000) prevents hepatocyte damage. Indeed, patients with fatty liver are more sensitive to LPS-induced liver injury (Yang et al. 1997). Sensitization to LPS has also been recapitulated in rodent models of acute alcohol exposure (Beier et al. 2009; Bergheim et al. 2006). In addition, there is an imbalance between pro-oxidants and antioxidants in alcoholics leading to oxidative stress (see Beier, Arteel, and McClain 2011; Beier and McClain 2010, for review). Studies in animal models of ALD have established a clear link between oxidative stress and the development of experimental liver damage caused by alcohol (Arteel 2003). More recently, zinc deficiency and ethanol–linoleic acid induced perturbations of the gut–liver axis with subsequent endotoxemia have been reported (Kirpich et al. 2012). Thus, priming, sensitization, oxidative stress, nutrient–ethanol interactions, and perturbations of the gut–liver axis are important mechanisms in the development of ASH. The potential role of these mechanisms should be examined in TASH. Like NAFLD, ALD could also be a susceptibility factor for hepatotoxicity from environmental chemicals by altering metabolism (CYP2E1 upregulation).

Toxicant-associated Steatohepatitis

Based on the nomenclature for the spectrum of pathologies observed in nonalcoholic fatty liver disease (NAFLD, NASH) and alcoholic liver disease (ALD, ASH), we propose that a similar nomenclature be used for fatty liver diseases due to industrial chemicals or toxicants (TAFLD, TASH). Thus, we propose to use TAFLD to describe the spectrum of toxicant-associated fatty liver diseases (steatosis, steatohepatitis, fatty cirrhosis, liver cancer), and TASH to specifically connote only steatohepatitis. The degree of liver injury is likely determined by the specific chemical exposure and dose; although other factors including interactions with genes and nutritional status likely modulate the severity of TAFLD.

While some groups have utilized the term, TAFLD, to describe fatty liver disease due to prescription medications (Schwingel et al. 2011), we propose to limit the use of TAFLD to industrial chemicals, as there appear to be mechanistic differences between hepatotoxicity due to prescription medications (drug-induced liver injury [DILI]) and industrial chemicals. For example, DILI is generally idiosyncratic, while most industrial chemicals exert dose-dependent hepatotoxicity (Cave et al. 2011). However, overlap occurs and important exceptions do exist. For example, cirrhosis/fatty liver due to methotrexate has been associated with high cumulative exposures to this medication. Likewise, in our recently published biomarker study, suspected TASH in elastomer workers did not seem to occur in an exposure-dependent manner, although all workers had relatively high cumulative exposures (Cave, Falkner, and McClain 2011).

Many classes of industrial chemicals have been associated with TAFLD (Table 1). These include (but are not limited to) solvents and other halogenated hydrocarbons, volatile organic mixtures, POPs, pesticides, and some nitro-organic compounds. In contrast to ASH and NASH, which are typically associated with increased serum aspartate aminotransferase (AST) and alanine aminotransferase (ALT) activities, respectively, TASH may frequently be associated with normal liver enzymes (Table 3; Brautbar and Williams 2002; Cave, Falkner, and McClain 2011; Cave, Falkner, et al. 2010). Thus, TASH is likely to be more prevalent than is currently recognized because current screening methodologies may not be sensitive enough for its detection. Our work to date has implicated several mechanisms in TAFLD/TASH including hepatocyte necrosis (rather than apoptosis seen in ASH/NASH), insulin resistance with altered adipokines, proinflammatory cytokine elevation, decreased antioxidants, and mitochondrial dysfunction (Cave, Falkner, and McClain 2011; Cave, Falkner, et al. 2010; Falkner et al. 2010; Shi et al. 2012). Based on the ASH/NASH literature, other mechanisms can be anticipated to occur in TASH and are likely to be exposure specific (Table 4). Both alcohol and obesity modify the response to xenobiotics, so ALD/NAFLD/TAFLD can occur within the same individual. For example, despite the “healthy worker” effect, the prevalence of obesity has increased in many chemical plants. Thus, dietary practices must be incorporated into risk assessment and mechanistic studies of TAFLD (Hennig et al. 2012). The remainder of this article reviews TAFLD associated with specific chemical exposures.

Table 3.

Selected chemicals associated with steatohepatitis in humans or animal models with either normal or elevated serum transaminases.

Steatohepatitis with normal   transaminases	Steatohepatitis with elevated (or unknown)   transaminases
Vinyl chloride	Carbon tetrachloride
Tetrachloroethylene	Dimethylformamide
Solvents (occasionally   including VOCs)	Methylmercury
Nitrobenzene	Pesticides: chlordecone, atrazine, paraquat
Nitromethane	Polychlorinated biphenyls
	Yellow phosphorus
	1,1,2-Trichloroethane
	VOCs (occasionally)
	Arsenic
	Lead

Adapted from Cave, M., K. C. Falkner, and C. J. McClain, Occupational and Environmental Liver Disease. In Zakim and Boyer’s Hepatology: A Textbook of Liver Disease. 6th ed. Philadelphia. Elsevier Saunders, 2011.

Table 4.

Potential TAFLD/TASH mechanisms.

Hepatocellular necrosis	Altered hepatic lipid metabolism
Lipotoxicity	Proinflammatory cytokines
Insulin resistance with altered   adipokines	Oxidative stress with decreased   antioxidant levels
Mitochondrial dysfunction	Metabolic endotoxemia
Endoplasmic reticulum stress	Proteasomal dysfunction
Altered transmethylation/   transulfuration	Decreased intestinal barrier function
Autophagy	Altered intestinal microflora
Nuclear receptor activation	Nutrient-toxicant interactions
Kupffer cell activation	Liver sinusoidal endothelial cell damage
Carbonyl stress	Priming and Sensitization
Epigenetic modifications	Hypoxia

Note: TAFLD = toxicant-associated fatty liver disease; TASH = toxicant-associated steatohepatitis.

Chemical Exposures Associated with TASH

Chloroalkenes

Chloroalkenes including tetrachloroethylene (PCE) and trichloroethylene (TCE) were introduced as less hepatotoxic solvent alternatives to chloroalkanes. These high production volume chemicals have found widespread use as degreasers and dry-cleaning fluids. Unfortunately, hepatotoxicity has been documented, and TASH does appear to occur, particularly with PCE exposures, while TCE hepatotoxicity more closely resembles autoimmune hepatitis. VC is most commonly used as a chemical intermediate, and VC exposures have been associated with TASH, HCC, and hemangiosarcoma. Chloroalkenes are among the most commonly found contaminants at Superfund sites.

VC

VC monomer is a colorless gas and a classically described mediator of occupational liver diseases, most notably hemangiosarcoma and TASH. VC is a high production volume chemical, and its worldwide capacity exceeds 35 million pounds. VC is used for polyvinyl chloride (PVC) production and for the synthesis of chlorinated solvents. To date, more than 80,000 American chemical workers have been exposed to VC (Kielhorn et al. 2000). At production facilites such as Louisville, high-level exposures estimated to exceed 1,000 ppm occurred in “poly cleaners,” who entered and manually cleaned PVC batch reactor vessels prior to 1975. VC exposures were subsequently limited by the Occupational Safety and Health Administration’s Vinyl Chloride Standard (29 CFR 1910.1017) to less than 1 ppm averaged over an 8-hr workday. In addition to chemical workers, high-level occupational exposures occurred in hair dressers, as VC was used as an aerosol propellant in hair spray from approximately 1962 through 1974 (Infante et al. 2009). VC has been identified as a solvent degradation product, and it is present in landfill leachate where it potentially places surrounding populations at risk.

The identification of three cases of VC-related hemangiosarcoma at a single B.F. Goodrich chemical plant in Louisville, Kentucky, in the 1970s is perhaps the most important sentinel event in occupational hepatology (Creech and Johnson 1974). In response to this outbreak, a liver cancer screening and surveillance program was instituted by the company in collaboration with the University of Louisville, and a large database and biorepository was established (Dannaher, Tamburro, and Yam 1981). Importantly, due to a long lag time, hemangiosarcomas still occur in workers with high-level VC exposures prior to implementation of the VC Standard (Figure 4). Elsewhere, studies have since documented associations between VC exposures and a variety of other liver pathologies including HCC (Mastrangelo et al. 2004), cirrhosis, peliosis hepatitis, focal hepatocytic hyperplasia (FHH), and focal mixed hyperplasia (FMH; Tamburro, Makk, and Popper 1984). VC is classified as a group 1 (definite) human carcinogen by the International Agency for Research on Cancer (IARC).

Figure 4.

Computed tomography scan from a VC worker with a 20 cm hypodense liver mass which was subsequently confirmed to be hemangiosarcoma on biopsy. VC = vinyl chloride.

However, as early as 1975, it was recognized that “fatty changes” similar to those observed in ALD were present in about half of biopsied chemical workers with high cumulative exposures (Gedigk, Muller, and Bechtelsheimer 1975). In fact, in his final article addressing FHH and FMH, Hans Popper noted that these lesions were often obscured by “fatty infiltration and/or chronic disease (hepatitis/granuloma)” (Tamburro, Makk, and Popper 1984). Our group became interested in this topic following the publication of reports documenting steatohepatitis in lean Brazilian petrochemical workers exposed to chemical mixtures including VC (Cotrim et al. 2004) and ultrasound studies observing hepatomegaly, steatosis, and fibrosis in VC workers (Hsiao et al. 2004; Maroni et al. 2003). Subsequently, 25 liver biopsies from highly exposed VC workers at the Louisville plant were reread, yielding an 80% steatohepatitis prevalence (Figure 5; Cave, Falkner, et al. 2010). Fibrosis was present in 55% of TASH cases. Remarkably, serum transaminases were normal in most cases. TASH was associated with insulin resistance, reduced serum adiponectin levels, marked elevation of proinflammatory cytokines, reduced serum antioxidants, and necrotic, rather than apoptotic, hepatocellular death.

Figure 5.

Photomicrograph of a liver biopsy (100×, hematoxylin–eosin [H&E] stain) of a VC worker with TASH and cirrhosis. VC = vinyl chloride; TASH = toxicant-associated steatohepatitis.

It is perhaps not surprising that VC exposures result in steatohepatitis, because VC is metabolized in a strikingly similar fashion to ethanol. At concentrations up to approximately 220 ppm, VC is metabolized by CYP2E1 forming the highly reactive genotoxic epoxide, chloroethylene oxide. CYP2E1 polymorphisms were associated with fibrosis, liver injury, and chromosomal damage in Asian VC workers and possibly hemangiosarcoma at the Louisville plant (Antonino-Green et al. 2000; Hsieh et al. 2007; Huang et al. 1997; Ji et al. 2011). Chloroethylene oxide is either spontaneously or enzymatically converted into chloroacetaldehyde (CAA). CAA may be metabolized by aldehyde dehydrogenase 2 or scavenged by sulfhydryl groups/GSH. Indeed, glutathione S-transferase (GST) polymorphisms have been associated with increased biomarkers of genotoxicity and liver disease in VC workers (Huang et al. 1997; Ji et al. 2011). Thus, VC is metabolized in a similar fashion to ethanol and an exceptionally high prevalence of steatohepatitis occurred in highly exposed workers in association with insulin resistance, altered adipocytokines, and antioxidant depletion. VC-gene (CYP2E1, GST) interactions appear to be important for hepatotoxicity. Because fatty liver has been reported with occupational VC exposures at multiple sites, VC appears to be a model chemical for TASH.

TCE

TCE, a high production volume chemical and important environmental contaminant, is a stable, colorless liquid at room temperature. Annual U.S. TCE production exceeds 300 million pounds, primarily for use as a degreasing agent and chemical intermediate; and an estimated 400,000 American workers are routinely exposed. Common household products that contain TCE include typewriter correction fluid, paint removers, adhesives, glues, and spot removers (ATSDR 1997b). Thus, home indoor air contamination occurs. The largest source of environmental contamination stems from TCE evaporation from factories that utilize it for the removal of grease from metal. TCE is also a groundwater contaminant, and it has been documented at over 1,500 hazardous waste sites and military installations where it was extensively used for degreasing aircraft and weapons (EPA 2009).

Hepatotoxicity has long been documented following TCE exposures. HCCs occur in mice but not rats; and TCE is an IARC Group 2A (probable) human carcinogen. However, fatal hepatic necrosis with steatosis was documented following occupational TCE exposure as early as 1955. (Joron, Cameron, and Halpenny 1955). More recently, fatty liver accompanied by rash was reported in Chinese workers (Liu 2009); although other reports suggested that liver dysfunction more closely resembled a hypersensitivity reaction or autoimmune hepatitis, rather than TASH (Cooper et al. 2009). Oxidative metabolites of TCE include trichloroacetic acid (TCA) and to a lesser degree, dichloroacetic acid (DCA). TCA is a ligand of the peroxisome proliferator–activated receptor alpha (PPARα), which is involved in hepatic lipid metabolism. Fatty liver occurred with TCE exposure in PPARα-null, and humanized PPARα (hPPARα) mice (Ramdhan et al. 2010) and rats (Kumar et al. 2001). Similar to human reports, exposed guinea pigs developed fatty liver with dermal hypersensitivity (Tang et al. 2008).

PCE

PCE, also known as perchloroethylene (PCE), is a colorless, nonflammable liquid with a sweet smell. U.S. PCE demand exceeds 300 million pounds, largely for use as a chemical intermediate, dry cleaning fluid, and degreaser. It has been estimated that over 27,000 dry cleaners still use PCE, and ground water contamination remains a significant concern. It has been shown that 1 to 3% of absorbed PCE metabolizes into TCA and is sequentially eliminated in the urine. The remaining 97 to 99% is exhaled as PCE (ATSDR 1997a).

PCE exposures have been associated with liver disease in human and animal studies. TASH was noted in a study of 29 dry cleaner workers exposed to PCE at 16 ppm (Brodkin et al. 1995). A greater percentage of hepatic parenchymal changes, including fatty infiltration, were noted in dry cleaning operators than in nonexposed laundry workers (67% vs. 39%). Most PCE-exposed workers with fatty liver had normal serum aminotransferases. More recently, fatty liver by magnetic resonance imaging (MRI) was reported in an elderly woman with daily PCE exposures who worked in the dry cleaning industry (Pezzini et al. 2008). The majority of animal studies focus on HCC, and PCE is a probable human carcinogen (Lash and Parker 2001). Indeed, a significantly elevated risk for primary liver cancer was demonstrated in female launderers and dry cleaners (Lynge and Thygesen 1990). However, fatty liver was noted in mice following acute PCE inhalation exposures as early as 1962 (Kylin et al. 1962). Notably, PCE produced a similar degree of steatosis as chloroform but significantly more steatosis than TCE. Hepatic steatosis was more recently noted in mice treated with high-dose PCE (Philip et al. 2007).

Chloroalkanes

Chlorinated alkanes, including chloroform, carbon tetrachloride, dichloroethane, trichloroethane, and tetrachloroethane, have been used as industrial solvents, chemical intermediates, and medicine (anesthetic agents) since the 1850s. However, exposure to these compounds has been associated with steatohepatitis and acute liver failure. With the adoption of less hepatotoxic anesthetic agents, reports of acute chloroalkane intoxication have also dwindled, and now case studies of acute intoxication with chlorinated alkanes tend to describe workplace accidents and incidents of drug abuse (Choi et al. 2006; Kim 2008; Lionte 2010). The demonstrated toxicity of chloroalkanes coupled with their ozone-depleting effects has resulted in strong efforts to replace them with less toxic alternatives. However, low-level chronic occupational exposures have led to some reports of clinical liver injury that may hint at a much larger subclinical population (Halevy et al. 1980). Chlorinated aliphatic compounds remain in use in industry as degreasers because of their superior properties as solvents of organic molecules and are particularly common in the dry cleaning industry. Occupational exposure is typically by inhalation or skin absorption. Most absorbed trichloroethane is excreted in exhaled breath or biotransformed via CYP2E1 to trichloroethanol or TCA and excreted in urine.

Chloroform

Chloroform was one of the first anesthetic agents to be used in surgery, and its use continued into the second half of the twentieth century, despite a strong history of hepatotoxicity. Chloroform is a classic liver toxicant, causing steatosis with mid-zonal (zone 2) or centrilobular (zone 3) necrosis (Thorpe and Spence 1997). Although no longer in use as an anesthetic agent, chloroform is a by-product of water chlorination, and humans are exposed to it in drinking water, or breathing vapor that has been volatilized by hot water in the shower. However, the potential impact of chronic low-level environmental chloroform exposures on fatty liver disease is unknown.

Carbon Tetrachloride

Carbon tetrachloride (CCl₄) was in widespread use as a solvent, vermicide (kills worms), refrigerant, and in fire extinguishers. However, its use for these applications was limited after several hundred cases of its toxicity were reported in the early to mid-twentieth century (Zimmerman 1999). It has continued to be used as a feedstock for the synthesis for chlorofluorocarbon (CFC) gases. However, following the adoption of the Clean Air Act and the Montreal Protocol banning CFCs, U.S. carbon tetrachloride production has been reduced to 130 million pounds per year (ATSDR 2005). CCl₄ is a stable molecule and a persistent environmental pollutant but does not appear to bioconcentrate in animals. Nonetheless, CCl₄ is a ubiquitous ambient air pollutant and may also contaminate groundwater supplies. The estimated daily intake by the U.S. general population from air and water ranges from 12 to 511 µg/day and from 0.2 to 60 µg/day, respectively.

The primary route of acute carbon tetrachloride poisoning is via inhalation, leading to multiorgan failure and death in 25%of the cases (Zimmerman 1999). A prodromal phase consisting of dizziness, headache, confusion, nausea, vomiting, and diarrhea subsided for 1 to 2 days before the onset of acute liver failure, typically including extremely high transaminase levels. The AST levels were usually greater than ALT levels, and in one case it reached 27,000 U/L. Jaundice developed in half of the cases. Hepatic coma and ascites developed in severe cases. Oliguric renal failure ensued and was the usual cause of death, as most cases occurred before the advent of hemodialyisis. When death occurred, it was typically within 10 days of exposure. Liver histopathology consisted of centrilobular (zone 3) steatosis and necrosis was also observed in the renal tubular epithelium in most fatal cases. Similar to acetaminophen overdose, recovery was typically rapid in those who survived. Treatment was supportive, although some reports attributed benefit to intravenous N-acetylcysteine or hyperbaric oxygen (Zimmerman 1999). Cirrhosis has been reported in cases of chronic exposure (ATSDR 2005).

Although human cases of acute carbon tetrachloride poisoning seldom occur today, it remains the classic experimental model of occupational hepatotoxicity. Carbon tetrachloride is a well-studied hepatotoxicant, which has been shown multiple times to induce liver injury (steatosis, necrosis, fibrosis, and HCC) similar to that observed in humans. As with other chlorinated aliphatic compounds, bioactivation of CCl₄ by CYP2E1 to reactive metabolites is critical for its toxicity (Manibusan, Odin, and Eastmond 2007). The mechanisms of acute CCl₄ hepatotoxicity involve immediate cleavage of CCl₄ by CYP2E1 in hepatocytes (Johansson and Ingelmann-Sundberg 1985), which generates the trichloromethyl radical, leading to lipid peroxidation and membrane damage (Recknagel et al. 1989). Subsequently, activated hepatic macrophages (Kupffer cells) produce toxic mediators (e.g., inflammatory cytokines, reactive oxygen intermediates, and eicosanoids), resulting in the injury of parenchymal cells (Edwards et al. 1993).

It was shown that chronic alcoholics had increased susceptibility to CCl₄ poisoning (Zimmerman 1999). It is well known that CYP2E1 is robustly induced by alcohol and can contribute to a far greater amount of total alcohol metabolism in alcohol-dependent individuals (Beier and McClain 2010; Lieber 1997). Therefore, the induction of this enzyme by chronic abuse of ethanol may increase the risk of liver damage by other agents (e.g., CCl₄). More recently, higher hepatic CYP2E1 expression and activity have also been associated with obesity and NAFLD (Aubert et al. 2011). Acute CCl₄ also enhanced liver damage in mice fed the obesogenic Western diet (Allman, Gaskin, and Rivera 2010). Therefore, although it has not been directly studied in humans, it is distinctly possible that obesity and related sequelae (e.g., NAFLD) will also increase the susceptibility to CCl₄.

Chloroethanes

1,1,1-trichloroethane was considered to be the least hepatotoxic halogenated hydrocarbon solvent available based on studies undertaken in the late 1950s and 1960s (Hodgson et al. 1989). Consequently, it was widely used as an industrial solvent and was present in many household cleaners and adhesives until its use was phased out due to ozone depletion. Fatalities associated with exposure to 1,1,1-trichloroethane are generally related to abuse, and like 1,1,2-trichloroethane, are most often due to its central nervous system (CNS) depressant effects. Hepatic effects have been described as transient (Halevy et al. 1980).

In contrast, 1,1,2-trichloroethane is a relatively potent hepatotoxin and its industrial use is therefore restricted. There are a few reports of adverse effects from occupational exposures, but animal experiments have suggested the potential for steatohepatitis as well as HCC (NIOSH 1978). A mouse study (90-day) demonstrated liver enzyme elevation with GSH depletion following 1,1,2-trichloroethane exposure in drinking water (White et al. 1985).

1,2-dichloroethane has been shown to cause elevations in ALT in mice associated with hepatomegaly and steatosis (Storer, Jackson, and Conolly 1984). In humans, acute toxicity caused elevated liver enzymes, hepatomegaly, and centrilobular (zone 3) necrosis (ATSDR 2001). The hepatotoxic effects of 1,1-dichloroethane are considerably milder than those of 1,2-dichloroethane, possibly due to the differences in metabolism and the generation of toxic intermediates (Mitoma et al. 1985; Wilbur and McClure 2004).

Volatile Organic Compounds

Volatile organic compounds (VOCs) such as benzene, toluene, styrene, and xylene are colorless, flammable liquids with a sweet odor that evaporate quickly into air. VOCs may be used as solvents or chemical intermediates. Reported U.S. production for benzene was 2.775 billion gallons in 2008 (Kirschner 2009), toluene with 1.905 million gallons in 2005 (Kirschner 2006), and styrene with 9.605 million pounds per year (Kirschner 2004).

Petrochemical workers, painters, and printers are frequently exposed to VOC mixtures. VOC exposures have been associated with TASH with both normal liver enzymes (Brautbar and Williams 2002; Cave, Falkner, and McClain 2011) and abnormal liver enzymes (Cotrim et al. 1999). Several studies document steatohepatitis by pathology or ultrasound in exposed petrochemical workers, painters, or printers with abnormal liver enzymes. A study from Brazil showed that 20 nonobese, nondrinking petrochemical workers with abnormal liver enzymes had hepatic steatosis on biopsy following exposure to 18 industrial chemicals including benzene, toluene, xylene, styrene, and VC. Repeat liver biopsies 8 to 14 mo following removal from the workplace showed improvement in the severity of steatosis in 9 of the 10 subjects (Cotrim et al. 1999). Steatohepatitis was confirmed in subsequent reports of Brazilian petrochemical workers (Barberino et al. 2005; Cotrim et al. 2004). Likewise, at a petrochemical plant in Argentina, 27 of 92 workers exposed to VOCs had elevated transaminase levels. Fatty liver was subsequently detected by ultrasound in 14 of the subjects with aminotransferase elevation (Perez et al. 2006). Eleven of 13 household painters with VOC exposures and abnormal liver enzymes had fatty liver on biopsy (Dossing et al. 1983). Likewise, all 8 toluene-exposed printers with persistent mild liver enzyme elevation had hepatic steatosis (Guzelian, Mills, and Fallon 1988).

Several additional serum biomarker studies have been performed to evaluate VOC hepatotoxicity and/or TASH. Although pathology was not provided, based on the aforementioned liver biopsy and ultrasound studies, we suspect that a significant proportion of subjects with elevated hepatotoxicity biomarkers had TASH. For example, a 45.55% prevalence of liver enzyme elevation occurred in shoe repairmen exposed to toluene (Tomei et al. 1999). Serum bile acids have been proposed as an alternate biomarker for solvent hepatotoxicity and were elevated in toluene- and xylene-exposed workers (Franco et al. 1986) as well as workers exposed to styrene when processing sewage pipes (Edling and Tagesson 1984). An increase in γ-glutamyltransferase was associated with exposure to a mixture of 6 VOCs including benzene, toluene, and xylene (Liu et al. 2009) in the National Health and Nutrition Examination Survey (NHANES). Our group reported elevated adipocytokines in elastomer workers exposed to styrene, butadiene, and acrylonitrile mixtures in conjunction with elevated serum cytokeratin 18 consistent with TASH (Cave, Falkner, and McClain 2011).

Dioxins and PCBs

Dioxins and PCBs are structurally similar polychlorinated aromatic POPs. Dioxins and some PCBs bind and activate the aryl hydrocarbon receptor (AhR). Although tumors and a wasting syndrome are the most classically described toxicologic effects of these chemicals, more recent data suggest that TASH may also occur.

Dioxins

Polychlorinated dibenzo-para-dioxins (PCDDs), popularly known as dioxins, are persistent environmental contaminants and unwanted by-products of industrial processes such as incineration, metal processing, and pesticide production. Dioxins contain two benzene rings connected by two oxygen atoms and contain four to eight chlorines with a total of up to 75 congeners. 2,3,7,8-Tetrachlorodibenzo-p-dioxin (TCDD), the most potent congener of the series, was a contaminant of Agent Orange, a defoliant used during the Vietnam War. TCDD was classified by the IARC as a Group 1 carcinogen in 1997 (IARC 1997). Dioxins are found in air, sediments, and soil. Most human exposures to dioxins occur through food consumption such as poultry, milk, fish, and meat since these persistent compounds accumulate in animal fat. Occupational exposures to dioxins, particularly TCDD, have been reported since the 1940s with the Seveso disaster, an industrial explosion in Italy in 1976, contributing to the highest known exposures to TCDD (Sweeney and Mocarelli 2000). Another poisoning episode occurred in 1977 in Missouri, where salvage oil was contaminated with TCDD, due to improper disposal of industrial toxic waste (Kimbrough et al. 1977). Dioxin levels in humans have decreased over time.

In addition to tumor development and a wasting syndrome, hepatotoxicity is a major defining feature of TCDD toxicity in laboratory animals as seen in rat, mouse, chicken, zebra fish, guinea pig, and rabbit, which was the most sensitive species to liver toxicity (El-Sabeawy, Enan, and Lasley 2001; Mann 1997; Zodrow, Stegeman, and Tanguay 2004). Apart from hepatocellular tumors, other toxic manifestations in the liver were steatosis, hepatic necrosis, inflammation, multinucleate hepatocytes, and cytoplasmic vacuolation. TCDD administration in immature, ovariectomized C57B/l/6 mice either alone or in combination with another environmental pollutant, PCB 153, resulted in hepatic histological changes including lipid accumulation, hepatocellular hypertrophy inflammatory cell infiltration, hyperplasia, and necrosis (Kopec et al. 2011). Notably, another study demonstrated an increase in total hepatic fatty acids and triglycerides and serum ALT levels in TCDD-treated mice (Boverhof et al. 2006). A recent study showed that TCDD increased fatty acid transport by utilizing dietary fat rather than carbohydrate sources to elicit fatty liver disease (Angrish et al. 2012).

In contrast to animal studies, it is unclear whether TASH occurs in TCDD-exposed humans. Nonetheless, some human studies demonstrate alteration in liver enzymes in both adults and children following TCDD exposures. A follow-up study on an Austrian cohort of chemical workers involved in herbicide production suggested that high TCDD exposure at a younger age led to chronic liver disease (Neuberger et al. 1999). Clinical chemistry on TCDD-exposed subjects showed elevated γ-glutamyl transferase (γ-GT), AST, and ALT levels as compared to controls. Some subjects exhibited weight loss with increased TCDD plasma concentrations while in other subjects; weight gain was reported as TCDD plasma concentrations decreased. A trend for an increase in chronic liver disease was also reported in another long-term follow-up study with a TCDD-exposed subcohort (Zober et al. 1994).

Dioxins exert their effects by high-affinity binding to the AhR, an intracellular ligand-activated receptor that belongs to the PAS domain protein family of transcription factors (Denison et al. 2002). Although more traditionally linked to hepatocarcinogenesis, AhR activation by TCDD has more recently been shown to cause hepatic steatosis in rodents. AhR activation interferes with peripheral fat mobilization, increases fatty acid uptake in the liver by upregulating CD36 and other fatty acid transporters, and suppresses peroxisomal β-oxidation of fatty acids leading to hepatic triglyceride accumulation (Lee et al. 2010). TCDD also induces hepatic dyslipidemia by AhR-mediated induction of stearoyl-CoA desaturase 1, the enzyme that catalyzes the rate-limiting step in monounsaturated fatty acid biosynthesis and hence alters hepatic lipid composition (Angrish et al. 2012). However, the effects of AhR on lipid homeostasis in humans must be investigated further due to species-specific differences in gene batteries activated by AhR.

PCBs

PCBs are halogenated compounds consisting of up to 10 chlorine atoms attached to a biphenyl group. They were manufactured during 1930s to 1970s and were used as dielectric and heat transfer fluids in electric capacitors, wax extenders, and flame retardants. A total of 1.3 million tons of PCB mixtures containing 130 congeners were manufactured worldwide prior to 1977 (Breivik et al. 2002). Although PCB production was subsequently banned by the Stockholm Convention, their high thermodynamic stability makes PCBs resistant to environmental degradation, and thus, they are POPs. PCBs are present in the ecosystem, including the atmospheric air, lakes, rivers, fish, human adipose tissue, and serum, and breast milk. PCBs still contaminate the food supply, and daily intake in the American diet is estimated to be approximately 30 ng/day based on a recent study from a Dallas supermarket (Schecter et al. 2010).

Acute effects due to PCB toxicity such as fatigue, anorexia, nausea, and jaundice were known since the 1930s (Flinn and Jarvik 1938), although evidence of cancer due to occupational exposures to PCBs was not observed in chemical workers until the 1960s. The “Yusho disease” that occurred in Japan in 1968 due to contamination of edible rice oil with a PCB mixture clearly pointed out the carcinogenic effects of PCBs. A mortality report on the Yusho patients showed an increased risk for all malignancies upon PCB exposure with an elevated risk for primary liver cancer (Kuratsune et al. 1987). Subsequent cohort studies of exposed electrical capacitor workers also demonstrated increased hepatobiliary malignancy rates; and PCBs are IARC Group 2A carcinogens. Occupational PCB exposure was associated with elevated liver enzymes and hepatomegaly (Maroni et al. 1981). We became interested in PCBs after our exposome-wide study demonstrated that 20 congeners were dose-dependently associated with suspected fatty liver disease in adult NHANES participants (Cave, Appana, et al. 2010). In subjects involved in the “Yu-cheng” incident in Taiwan where cooking oil had been highly contaminated by PCBs, the mortality rate due to cirrhosis was 2.7-fold higher than expected (Yu et al. 1997).

In contrast to human studies, pathologic data document hepatic steatosis following PCB treatment in animal studies. Long-term rat studies conducted by the National Toxicology Program document “toxic hepatopathy” which is characterized by prominent steatosis, inflammation, and fibrosis and could be considered to represent a TASH variant (NTP 2006). Moreover, we recently demonstrated that PCB 153 exposure causes TASH with hepatic antioxidant depletion (Shi et al. 2012). PCB mode of action in TASH is uncertain but appears to be congener specific and critically dependent on nutrient interactions, at least in mice (Hennig et al. 2005; Shi et al. 2012). Coplanar, or dioxin-like PCBs, bind and activate the AhR and are expected to have similar modes of action to dioxin. In contrast, some non-coplanar PCBs may activate the pregnane × receptor (P×R) or the constitutive androstane receptor (CAR) in a tissue- and substituent-dependent manner (Al-Salman and Plant 2012). However, the potential importance of the activation of these receptors in PCB-mediated TASH is unknown

Pesticides

Pesticides are compounds or mixtures intended to kill, inactivate, repel, or mitigate pests and are primarily used for agricultural and antimalarial purposes. According to the Stockholm Convention on Persistent Environmental Pollutants in 2001, 9 of the 12 most harmful chemicals, the so-called “dirty dozen,” were pesticides. Pesticides are not limited to commercial use and 74% of U.S. households use pesticides. Pesticides are associated with over 3 million acute poisonings and 250,000 deaths per year. In 2007, the world expenditure on pesticides was $39.4 billion, with the U.S. expenditure alone totaling $12.5 billion (Grube et al. 2011). Limited data suggest that some pesticides including organochlorine insecticides and triazine herbicides may be associated with TASH.

Organochlorine Insecticides

Organochlorine insecticides include but are not limited to dichlorodiphenyltrichloroethane (DDT); cyclodienes, namely chlordane, dieldrin, aldrin, endrin, and heptachlor; and caged structures such as mirex and chlordecone. DDT is the most abundant parent compound of its class. DDT was marketed as an insecticide beginning in 1944, and a total of 1.8 million tonnes have been produced globally since then. Organochlorine pesticides have been banned from most agricultural use for over three decades, after adverse effects on wildlife were observed. However, organochlorine insecticides are thermodynamically stable POPs that continue to contaminate living organisms and the human food supply. Although DDT is a known carcinogen in rodents, occupational human exposures have only been occasionally linked to liver cancer, among others. Thus, DDT is an IARC Class 2B (possible) human carcinogen.

Hepatic effects, including steatosis, were reported in a chlordecone intoxication incident involving 32 plant workers in Virginia (Guzelian et al. 1980). Many of the cases had hepatomegaly, and 12 workers underwent liver biopsy, revealing mild steatosis, portal inflammation, fibrosis, glycogenated nuclei, and lipofuscin accumulation. Interestingly, liver enzymes were repeatedly normal in all subjects, a characteristic commonly observed in TASH. Chlordecone undergoes enterohepatic circulation and cholestyramine was successfully used to promote chlordecone elimination. Using the NHANES 2003–2004 database, we investigated the effects of chronic, low-level exposures to organochlorine insecticides on liver enzymes in American adults (Patel et al. 2009). We reported multiple pesticides including dieldrin, trans-nonachlor (component of chlordane), and heptachlor epoxide (metabolite) that were dose dependently associated with increased odds ratios for ALT elevation and suspected NAFLD. The hepatotoxic modes of action for organochlorine insecticides are unknown, but they are known to induce cytohrome P450 enzymes and appear to activate nuclear receptors. More data are needed to determine whether TASH is a significant problem resulting from exposures to these legacy POPs.

Triazine Herbicides

Herbicides are widely used in the U.S. agricultural sector, accounting for about 70% of total pesticide usage. The triazine family of herbicides, including atrazine and simazine, were first introduced in the 1950s. Triazine herbicides were banned in the European Union, but widespread use continues in the United States. Triazines may contaminate groundwater, particularly in the summer months following application. Liver steatosis was reported in B6C3F1 mice fed with simazine for 35 weeks. Hepatic triglycerides increased and mitochondrial oxidative phosphorylation was perturbed after long-term simazine feeding in these mice (Vancova et al. 2000). Simazine toxicosis in sheep also revealed fatty liver change in these animals (Allender and Glastonbury 1992). Chronic exposure to atrazine led to steatosis, obesity, insulin resistance, and mitochondrial dysfunction in Sprague-Dawley rats (Lim et al. 2009). Triazine herbicides mediate their actions by inhibiting the electron transport chain in chloroplasts and hence are photosynthesis inhibitors. The similarity between plant chloroplasts and mammalian mitochondria may be responsible for the mitochondrial toxicity induced by triazine exposures in animal studies. However, there are currently insufficient data to correlate triazine human exposures to TASH.

Nitroaliphatic Chemicals

2-Nitropropane, a high production volume chemical, has been used in numerous applications since 1940, including printing inks and dyes, adhesives, resins, waxes, waterproof coatings, varnish remover, and as a fuel additive (Harrison, Letz, et al. 1987; Zimmerman 1999). 2-nitropropane stands out as the most toxic of the nitroaliphatic compounds, leading to severe hepatic injury in humans including fatty change, centrilobular (zone 3) necrosis, bile duct proliferation with accompanying cholestasis, fulminant hepatic failure, and death (Harrison, Letz, et al. 1987; Farrell 1994). In experimental animals, hepatotoxic effects are species dependent but generally include steatosis, necrosis, and liver tumors (Zitting, Savolainen, et al. 1981). Proposed mechanisms include oxidative stress and lipid peroxidation via catalase inhibition and mitochondrial dysfunction.

N-Substituted Amide Solvents

N,N-Dimethylformamide (DMF) is a solvent commonly used in the synthetic leather and polyurethane industry and has been implicated in case reports of hepatic injury, including steatohepatitis, fibrosis, cirrhosis, and cancer (Nakasone et al. 2011; Nomiyama et al. 2001; Redlich et al. 1988). In a 1990 polyurethane worker study, workers exposed to DMF for less than 3 months showed marked elevations in AST and ALT, with the ratio of ALT/AST always greater than 1, indicative of nonalcoholic liver injury. Workers exposed for longer than 1 year showed much more modest elevations in liver enzymes. On biopsy, microvesicular and/or macrovesicular steatosis was apparent in short-term exposed workers, along with enlarged Kupffer cells in zones 1 and 2 and evidence of hepatocyte injury. Steatosis persisted over time and progressed to moderate and/or severe fatty change. Hepatocytes in these workers were notable for prominent smooth endoplasmic reticulum, small fat droplets, and pleomorphic mitochondria (Redlich et al. 1990). Dimethylacetamide (DMAC), a related solvent used in the manufacture of synthetic fibers and acrylic resins, has also been investigated as a cause of occupational exposure-related liver injury. Experiments with dogs showed fatty change after dermal application for 6 weeks (Kim 1988). Thus, exposures to N-substituted amide solvents appear to be associated with TASH.

Metals

Metals including arsenic, mercury, and lead are potentially associated with TASH. While arsenic is a classically described hepatotoxicant, the liver is not traditionally considered to be a principal target organ for lead and mercury toxicities, despite the fact that these metals concentrate within the liver. However, limited data suggest that steatosis occurs following exposure to these metals, and more research is needed to determine the significance of these findings to human populations.

Arsenic

Inorganic arsenic is a ubiquitous element and a drinking water contaminant. Owing to its toxic potential to humans, Arsenic is a high-priority hazardous substance in the United States. Chronic exposure to arsenic has been linked with a myriad of possible health effects, including skin lesions, hypertension, cardiovascular disease, respiratory disease, and malignancies of the skin and internal organs (Waalkes et al. 2004). The liver is a well-known target organ of arsenic exposure. Hepatic abnormalities caused by arsenic exposure include hepatomegaly, noncirrhotic portal fibrosis, portal hypertension, HCC, and hemangiosarcoma (Mazumder 2005; Santra et al. 1999, 2000; Smith et al. 1992; Waalkes et al. 2006).

The major route of arsenic exposure is via ingestion of well water. High levels of arsenic can be found naturally in some water sources such as in West Bengal, an area with an unexpectedly high level of NAFLD in nonobese individuals, or as a consequence of mining activity where arsenic frequently contaminates mine tailings and then leaches into the water supply (Das et al. 2010). The hepatotoxic action of arsenicals is often attributed to the ability of either arsenite or its metabolites to induce oxidative stress. Arsenite binds lipoic acid in the mitochondria leading to inhibition of pyruvate dehydrogenase with resultant mitochondrial uncoupling and an increase in hydrogen peroxide production (Patrick 2003). Increases in free radical production, lipid peroxidation products, and oxidized DNA damage have been reported in multiple studies.

Although arsenic can be directly hepatotoxic, the concentrations/doses required are generally higher than that present in the U.S. water supply. However, the risk for developing a human disease derived from environmental exposure is not based solely on that environmental exposure but is rather modified by other mitigating conditions, such as genetic (e.g., polymorphisms in key genes) or other environmental (e.g., diet, lifestyle, etc.) factors. Indeed, it was recently shown that arsenic exposure, at concentrations that are not overtly hepatotoxic, enhances LPS-induced liver injury in mice (Arteel et al. 2008). Straub et al. (2007) demonstrated that mouse liver is also sensitive to more subtle hepatic changes (e.g., hepatic endothelial cell capillarization and vessel remodeling) at lower arsenic exposure levels (250 ppb) without any gross pathologic effects. It is nevertheless unclear at this time whether environmental arsenic exposure at the levels observed in the United States causes fatty liver disease. However, a striking feature of arsenic exposure and NAFLD is the significant overlap between areas of risk in the United States (Mokdad et al. 2003; Welch et al. 2000). For example, states with clusters of municipal wells with high levels of arsenic (e.g., Michigan, Texas, West Virginia, and Oklahoma; U.S. Geological Survey 2007) also have high incidences of obesity and diabetes. Furthermore, high arsenic in the drinking water is generally localized to private artesian water supplies (not regulated by the Environmental Protection Agency [EPA]) in rural communities, where the incidence of obesity tends to be even higher than in most areas of the country (National Center for Health 2001). It is therefore possible that arsenic exposure is an unidentified environmental risk factor in the development of NAFLD. In support of this hypothesis, it was recently demonstrated that liver injury during experimental NAFLD is enhanced by concomitant arsenic exposure in mice (Tan et al. 2011), suggesting that the relative risk of hepatic damage caused by arsenic exposure may have to be modified to take into account other mitigating factors, such as underlying NAFLD.

Mercury and Lead

The liver is not traditionally considered to be a target organ for mercury and lead toxicities. However, mercury and lead exposures were dose dependently associated with suspected NAFLD in NHANES participants (Cave, Appana, et al. 2010). Case studies and anecdotal reports involving fulminant hepatic failure or hepatitis following mercury exposure have been reported (Al-Sinani, Al-Rawas, and Dhawan, 2011; Wiwnaitkit and Chaiyasit 2011). Previous human epidemiological studies have inconsistently linked mercury contamination in Japanese fishing villages to increased liver-related mortality in villagers (Futatsuka et al. 1992; Futatsuka et al. 2005; Futatsuka, Shibata, and Kinjo 1987).

Despite the fact that methyl mercury (MeHg) concentrates considerably within the liver due to enterohepatic recirculation, few animal studies have examined the potential role of MeHg in liver disease. However, acute and chronic toxicity studies conducted in rats and cats demonstrated that mercury exposure resulted in depletion of body fat with development of centrilobular (zone 3) hepatic steatosis with increased lipid peroxidation products, proliferation of the endoplasmic reticulum, and floccular degeneration of the mitochondria with extrusion of diseased organelles into the sinusoidal space (Chang and Yamaguchi 1974; Desnoyers and Chang 1975; Klein et al. 1972; Lin, Huang, and Huang 1996). Many of these changes were irreversible even after discontinuation of MeHg exposure. The primary mechanism of MeHg hepatotoxicity may be related to its high affinity for sulfhydryl residues and consequent poisoning of cysteine-containing proteins and GSH depletion (Lin, Huang, and Huang 1996).

Most clinical cases of lead intoxication are related to its neurological effects. Although hepatotoxicity has been reported in animal studies, there are only a few case studies in humans describing reversible microvesicular and macrovesicular steatosis, and hepatitis (Beattie et al. 1979; Verheij et al. 2009). Although “lead-induced hepatic hyperplasia” is a classic pathologic liver lesion, a recent mouse study of inhaled lead acetate documented centrilobular (zone 3) hepatic steatosis, hepatocyte proliferation, apoptosis, inflammatory infiltrate, and fibrosis (Rendon et al. 2012). More work is required to document the potential role and mechanisms of arsenic, lead, and mercury in TASH.

Conclusions

Although the terms TASH and TAFLD were only recently coined, fatty liver has long been known to occur following industrial chemical exposures both in humans and in animal models. However, prior to the description of NASH in the 1980s, hepatic lipid accumulation was believed to have no clinical consequence, as lipids were believed to be inert. Therefore, while earlier pathologic studies frequently note the development of hepatic steatosis following chemical exposures, this finding was typically not emphasized as its clinical significance was considered doubtful. However, the NASH epidemic has clearly demonstrated that steatohepatitis may lead to meaningful clinical outcomes including decompensated cirrhosis, HCC, death, liver transplantation, and cardiovascular events. We suspect that TASH may likewise result in similar outcomes or even synergize to worsen liver disease of other etiologies. TASH is likely to be underrecognized clinically, because it may occur despite normal aminotransferses, emphasizing the need for effective noninvasive biomarkers. While TASH appears to share many disease mechanisms with both ASH and NASH, key differences appear to exist including divergent hepatocyte cell death mechanisms in some cases. Thus, more research is needed to understand TAFLD, so that more effective diagnostics and therapeutics may be developed.

Abbreviations

ALD

alcoholic liver disease

AhR

aryl hydrocarbon receptor

ASH

alcoholic steatohepatitis

CAA

chloroacetaldehyde

CASH

chemotherapy-associated steatohepatitis

CFC

chlorofluorocarbon gases

DCA

dichloroacetic acid

DILI

drug induced liver injury

DDT

dichlorodiphenyltrichloroethane

DMF

N,N-Dimethylformamide

FHH

focal hepatocytic hyperplasia

FMH

focal mixed hyperplasia

GSH

glutathione

GST

glutathione S-transferase

HCC

hepatocellular carcinoma

IARC

International Agency for Research on Cancer

MeHg

methylmercury

NAFLD

nonalcoholic fatty liver disease

NAS

NAFLD activity score

NASH

nonalcoholic steatohepatitis

NHANES

National Health and Nutrition Examination Survey

PCBs

polychlorinated biphenyls

PCDDs

polychlorinated dibenzo-para-dioxins

PCE

tetrachloroethylene

POPs

persistent organic pollutants

PPARα

peroxisome proliferator–activated receptor alpha

TAFLD

toxicant-associated fatty liver disease

TASH

toxicant-associated steatohepatitis

TCA

trichloroacetic acid

TCDD

2,3,7,8-Tetrachlorodibenzo-p-dioxin

TCE

trichloroethylene

VC

vinyl chloride

VOCs

volatile organic compounds