Hepatic Steatosis and Liver Transplantation Current Clinical and Experimental Perspectives

INTRODUCTION Hepatic steatosis has diverse clinical implications in liver transplantation, hepatology, and hepatic surgery. This review will focus on issues related to liver transplantation; the prevalence and diagnosis of hepatic steatosis in potential donors, the impact of steatosis on graft and recipient survival after transplantation and on donor liver supply. In addition, we will review available animal models to investigate this problem and the data in humans and in experimental animals that help understand how steatosis leads to increased liver ischemia/reperfusion (I/RP) injury. PREVALENCE OF HEPATIC STEATOSIS In an autopsy study addressing liver histology in “normal population” 24% of 503 fatalities from traffic accidents had steatosis. In that series 1% of subjects aged 20 years or younger, 18% between 20–40 years, and 39% older than 60 years had fatty liver (1). Subsequent autopsy studies reported similar occurrence of steatosis (2). The exact prevalence of steatosis will remain unknown because liver biopsy can not be used in the general population for routine screening. Screening studies using ultrasonography show steatosis in 16 to 23% of adults in non-hospital setting (3). In cadaver organ donors, the prevalence of steatosis ranges from 13 to 28% (4–8) approaching 50%, when very sensitive histological techniques are used (9). Moderate to severe steatosis is seen in approximately 20% of potential live donors (10,11). Given the steady increase in the mean age of cadaver donors in the United States and the overall increase in the prevalence of obesity one should expect a further increase in the prevalence of steatosis in both cadaveric and live liver donors. Obesity, age, type II diabetes, female gender, and alcohol ingestion are the conditions most commonly associated with hepatic steatosis (1,2). Nonalcoholic steatohepatitis is recognized increasingly as an important etiological factor in cryptogenic cirrhosis. Obesity and diabetes, two conditions known to be associated with nonalcoholic steatohepatitis are more common in patients with cryptogenic cirrhosis than in primary biliary, alcohol, and hepatitis C cirrhosis (12). DIAGNOSIS OF HEPATIC STEATOSIS Recent studies have addressed the accuracy of predicting the presence of hepatic steatosis by utilizing easily available anthropometric and laboratory measurements such as body mass index (BMI), waist/hip ratio, serum triglycerides levels etc. (11,13,14). Presently, BMI, defined as weight in kilograms/(height in meters) (2) is the most simple and reliable predictor of not only the presence of but also the extent of hepatic steatosis (11,13). In a recent study of living liver donors, BMI correlated well with both the prevalence and severity of steatosis and thus might become useful in determining the need for a liver biopsy among the potential donors (11). Hepatic imaging, although of very limited use in the assessment of cadaveric liver donors, plays a much greater role in the evaluation of living liver donors including the determination of steatosis. Ultrasonography is two to three times less expensive than computed tomography (CT) and magnetic resonance imaging (MRI) but is generally considered to lack specificity. Increased echogenicity, the ultrasonographic characteristic of steatosis can also occur with fibrosis. No data are available comparing the sensitivity and specificity of state of the art ultrasonography with CT or MRI in the imaging of hepatic steatosis (15). Noncontrast CT and MRI are presently the best radiological studies for the diagnosis of steatosis. In an unenhanced CT attenuation of a normal liver is greater than the spleen and when this is reversed and the liver-spleen attenuation difference is greater than −10 Hounsfield units it is considered diagnostic of steatosis (15). In a recent study addressing hepatic steatosis in 33 living liver donors CT and MRI were compared with liver biopsy. Both CT and MRI had 100% specificity but both failed to detect steatosis in approximately a quarter of the patients (11). For the future magnetic resonance spectroscopy offers the potential to quantify the amount of fat, useful for serial follow-up (15). Naked eye inspection at the time of laparotomy appears to be reliable in determining the presence of only severe grades of steatosis with a positive predictive value of 71, 46, and 17% for severe, moderate, and mild degrees of steatosis, respectively (5). Although histological examination of the liver remains the “gold standard” the section thickness and staining techniques have an impact on detection and grading of steatosis (9). In 83 consecutive donor livers, steatosis was detected in 49% of Sudan III-stained sections, 47% of toluidine blue stained sections, 39% of frozen sections and 21% of deparaffinized hematoxylin/eosin (H&E)-stained sections (9). Oil Red O stain also is highly sensitive in detection of steatosis but is technician dependent and also causes spurious staining of sinusoids (6). H&E staining remains the most commonly used stain and due to time constraints, frozen sections are preferred. The clinical relevance of fat detected by more sensitive stains but not seen with H&E is not known. Macro Versus Microsteatosis and Grading In the majority of steatotic donor livers the fat globules are macrovesicular (larger than and displace the nucleus). Microvesicular steatosis is less common in liver donors. Reye’s syndrome, fatty liver of pregnancy, parenteral nutrition, and drug toxicity, conditions traditionally associated with microsteatosis, are rarely seen in liver donors. Instead, obesity is the most common finding in donors with microsteatosis (16). Steatosis is graded by semiquantitative determination of the amount of fat in the liver. In most series, ≤30% fat is defined as mild, 31–60% moderate, and >60% as severe steatosis. In the literature there is a lack of uniform adoption of this grading, rendering comparison of data sometimes difficult. Impact of Steatosis on Graft and Patient Survival Soon after the initial descriptions of increased injury/primary non-function, the significance of fat in a liver graft as a risk factor in increased liver injury became apparent with reports of additional large series (4–6,17). More recent data show that even mild steatosis penalizes graft and patient survival (8). The significance of steatosis on graft function is underscored by its emergence as the most important variable in multivariate analysis of factors determining graft function after transplantation (7). The decrease in graft survival seen with fatty livers occurs during the early posttransplant period. Although there is a slight but statistically insignificant increase in nonischemic biliary strictures, there is no increased risk of rejection of steatotic grafts (8). In contrast to macrovesiular steatosis, livers with predominantly microvesicular steatosis show less injury and graft survival rates are similar to those in nonsteatotic grafts (16). The adverse effects of steatosis on graft and patient survival are more evident in sicker liver recipients (5). A clinical consensus exists that grafts with severe steatosis should be discarded and grafts with mild steatosis should be used. What about grafts with moderate steatosis? Should such livers be transplanted into patients in no urgent need of a transplant but in whom the probability of graft survival is better? Or into patients in urgent need of a transplant but in whom survival of such grafts is poor? Or should such grafts be discarded? Data from Spain show that approximately 16% of recovered livers are rejected for transplantation and 42% of those rejections are due to hepatic steatosis (18). What Happens to the Fat in Fatty Livers after Transplantation? After transplantation almost all of the fat is cleared by the end of the first week in both humans and experimental animals (8,19). After liver transplantation, de novo occurrence of hepatic steatosis occurs in recurrent hepatitis C and recurrence of steatohepatitis has also been described (20,21). Patients undergoing hepatic resection in the presence of moderate to severe steatosis require longer operating times, more blood transfusions, have higher postoperative bilirubin levels, and most importantly have higher risk of liver failure and mortality (22–24). Data have not emerged yet about the impact of steatosis on the donor and the recipient in live donor liver transplantation. ANIMAL MODELS OF FATTY LIVER Two main approaches have been used in generating animals with fatty liver for ischemia/reperfusion injury research. Genetic abnormalities in glucose and lipid metabolism lead to fatty accumulation in the liver as exemplified by the genetically obese mice and rats. Feeding a choline or choline- and methionine-deficient diet (CMDD) or feeding a fat- enriched diet with or without alcohol also results in production of fatty livers. The characteristics, advantages, and disadvantages of each model are outlined in Table 2. The principal lipid component in fatty livers is triglycerides (increased 5–20 times normal) and there is a slight increase in phospholipids and cholesterol (19,25).Table 2: Currently available animal models of fatty liver used in ischemia/reperfusion researchMECHANISMS OF ISCHEMIA-REPERFUSION INJURY TO FATTY LIVERS Microcirculation Laser flow Doppler studies in humans and in rats with steatosis show approximately a 50% decrease in sinusoidal blood flow, which may induce a state of chronic hypoxia (26,29). The decreased flow is probably secondary to swollen hepatic cords and narrowed sinusoidal lumen and may lead to inadequate perfusion with preservation solutions during liver retrieval (19,29). The sinusoidal lumen is further compromised in fatty livers from blebs or fat globules released during cold preservation and with fibrin microthrombi and cellular elements in the blood after reperfusion (30,31). These changes lead to further decrease in sinusoidal blood flow after both warm I/RP and liver transplantation (26,30). Cellular Changes Sinusoidal endothelial cells (SEC). Electron microscopic (EM) studies show a more severe detachment and rounding of SEC during cold preservation in fatty livers in rats (31,32). However, functional assays comparing clearance of hyaluronate between fatty and normal livers yield conflicting results in different reports (33,34). It is not clear if the more severe morphological changes seen in SEC of fatty livers are due primarily to inherent differences between the SEC of normal and fatty livers or produced by morphological changes in adjacent hepatocytes. Kupffer cells. Colloidal carbon studies show an increase in number and phagocytic activity of Kupffer cells in rat fatty livers with a further increase in phagocytosis after transplantation (30). In contrast, in vitro studies with macrophages from the liver and the peritoneum from obese mice and rats demonstrate decreased phagocytosis but increased release of free radicals and interleukin-6 with or without lipopolysaccharide (LPS) stimulation (35). Hepatocytes. During cold preservation of normal livers SEC changes overshadow the changes in the hepatocytes. Hepatocytes, which are well endowed with antioxidants, are considered relatively resistant to oxidative stress engendered by cold ischemia and reperfusion. However, in fatty livers, EM studies show a marked increase in hepatocyte blebs and the size of the fat vacuoles and a decrease in microvilli during cold preservation (31–33). There is a marked increase in trypan blue uptake in hepatocytes from fatty livers after cold preservation and warm anoxia followed by ex vivo perfusion (34). Neutrophils. Neutrophils play an important role in the reperfusion injury to solid organs through the release of free radicals and multiple proteases. In rat fatty livers in vivo microscopy and histological studies show increased leukocyte/neutrophil accumulation and activation after transplantation or warm I/RP further contributing to the sinusoidal congestion seen in such livers (30,31). The factors mediating the accumulation of neutrophils in fatty liver are not well understood. Antibodies to intercellular adhesion molecule-1 (ICAM-1) decrease neutrophil sequestration, sinusoidal congestion and ischemic injury (31). Subcellular Compartments Plasma membrane. Plasma membrane blebbing of the hepatocytes and SEC detachment referred to above are related to changes in cortical microfilaments and actin cytoskeleton (36). Decreased adenosine triphosphate (ATP) and glutathine levels, more pronounced in fatty livers, increase actin depolymerization (26,27,37). Also, in obese Zucker rats liver plasma membrane cholesterol:phospholipid ratio and the proportion of polyunsaturated fatty acids are decreased (32). Such changes might alter the fluidity of the plasma membranes, especially at the low temperatures of organ preservation. Mitochondria. The mitochondria in fatty livers have decreased ability to generate ATP, most evident after reperfusion after ischemia increasing the risk of graft failure (26,27). Increased mitochondrial oxidative stress is accompanied by decreased F0F1-ATP synthase; fasting worsens this further (38). Uncoupling protein-2, an inner membrane mitochondrial protein, is increased in fatty livers. By decreasing transmembrane potential and altering the normal flow of electrons in the mitochondria, uncoupling proteins decrease ATP synthesis and lead to increased production of reactive oxygen species (39). Peroxisomes. Peroxisomes participate in β-oxidation of fatty acids, especially when mitochondrial oxidation capacity is exceeded. A byproduct of peroxisomal, but not of mitochondrial, β-oxidation is hydrogen peroxide, which is detoxified by catalase (40). Several of the peroxisomal enzymes that metabolize fatty acids are under the control of peroxisome proliferation activation receptor α. Receptor transcriptional up-regulation occurs in conditions where the delivery of fatty acids to the liver is increased such as in rats and mice with obesity, diabetes, and fasting (41). The increase in β-oxidative enzymes results in an increase in the hydrogen peroxide generated, which may overwhelm the detoxifying ability of catalase and induce oxidative stress (41,42). In obese Zucker rats catalase activity is decreased accompanied by chronic oxidative stress (37). Microsomes. Cytochrome P-450 enzymes play an important part in ω-oxidation of fatty acids and in metabolism of drugs and xenobiotics. Obesity and fasting lead to increased P450 II E enzyme activity and similar increase also occurs in rats and mice with CMDD induced steatohepatitis (43). Increased activity of 2E1 leads to nicotinamide dinucleotide phosphate- dependent production of free radicals, lipid peroxidation, and chronic oxidative stress. Cytokines Rodents with fatty livers have increased hepatic injury and mortality upon administration of both LPS and tumor necrosis alpha (TNF-α) (35). Endotoxemia known to occur in human organ donors and during liver transplantation thus may contribute to the poor functioning of fatty livers (44,45). In rats TNF-α, interleukin-1, and chemokine-induced neutrophil chemoattractant are increased after warm hepatic ischemia in an alcohol induced fatty liver model but not in a model of fatty liver induced by cholesterol feeding (46). This raises the question whether the increased cytokine production is related more to injury produced by alcohol rather than lipid per SE. Additional data regarding the dynamics of various cytokines in response to cold and warm ischemia are not available and merit further investigation. Lipid Peroxidation Several studies show increased amounts of lipid peroxidation products in fatty livers (25,38,47). Initially, the peroxidation process may be initiated by physiologically produced oxidants in combination with an excess of easily oxidizable lipid substrate (47). Under basal conditions the antioxidant defenses of the liver appear to handle this level of peroxidation without deleterious consequences. Lipid peroxidation is further increased in fatty livers after cold and warm I/RP and may overwhelm the antioxidant systems (25). Phospholipid depletion is one serious consequence of lipid peroxidation and may result in membrane dysfunction (48). In addition, lipid peroxide products such as malondialdehyde and 4-hydroxynonenal are known to cross-link various proteins including cytokeratins may act as neutrophil chemoattractants and induce cell necrosis (40,47,49). Fatty livers show more tissue death after ischemia/reperfusion and cell death is preponderantly by necrosis rather than apoptosis (50). Platelet-activating factor, prostaglandins, and leukotrienes are some of the other lipid derived products and their role in I/RP injury of fatty livers is not known. A TWO HIT HYPOTHESIS IN ISCHEMIA/REPERFUSION INJURY TO FATTY LIVERS First Hit Although the existence of chronic oxidative stress has been accepted in alcoholic fatty livers, there is mounting evidence for such a phenomenon in fatty livers of various nonalcoholic etiologies (37,38,43). The common denominator of persistent and increased delivery of fatty acids and ketones to the liver appears to result in increased production of prooxidants in various subcellular compartments, most notably mitochondria, peroxisomes, and microsomes (38,41,43). In addition the availability in fatty livers of an easily oxidizable substrate such as lipid appears to further promote this problem (47). This chronic oxidative stress and the associated lipid peroxidation lead to depletion of hepatic antioxidants and, in certain instances, of antioxidant enzymes (37,38). In addition a potential state of chronic hypoxia may exist secondary to reduced sinusoidal blood flow (26). Hence although in most donors with steatosis there is no evidence of histological (except for steatosis itself) or biochemical evidence of liver injury, the liver is vulnerable to additional pathophysiological processes. Second Hit The role of reactive oxygen species and TNF-α in the causation of I/RP injury to the liver is well established. There is increased production of free radicals in fatty livers after liver transplantation during early reperfusion with a further decline in antioxidants and cellular energy stores (26–28,31). The improved survival of fatty livers seen when antioxidant enzymes are made available intracellularly rather than administered into the circulation suggests the oxidative stress is mostly intracellular (51,52). Increased accumulation of neutrophils during later stages of reperfusion imposes additional oxidative stress—mostly extracellular—to an already antioxidant-depleted liver (31). Thus, the “second hit” of I/RP superimposed on the “first hit” of oxidative stress is poorly tolerated by fatty livers. Treatment Strategies Other than minimization of cold and warm ischemia there are no treatment modalities proven to help in decreasing ischemia reperfusion injury in fatty livers in humans. The potential options, based on the available experimental data, need to be examined in clinical situations. Fasting worsens the oxidative stress and thus enteric feeding of the donor may be beneficial (38). It is unclear if increasing the perfusion volume and pressure at the time of circulatory arrest would improve preservation of steatotic livers (29). In living donors weight reduction regimen over a period of few months may improve steatosis but is dependent on donor motivation. Pharmacological therapy aimed at improving the antioxidant status of the donor liver before organ retrieval and at the time of organ recovery and reperfusion, and antibodies to decrease neutrophil accumulation (anti-ICAM, need to be Although antioxidant enzyme therapy has additional of the oxidative stress clinical will have to further in clinical therapy (51,52). Steatosis and This is important in the of hepatic resection and in living liver donors and their Recent data show of the liver and increased mortality in obese Zucker rats after of at the level of and in addition to have been as potential In hepatic steatosis is in approximately 20% of liver donors, both cadaveric and and of cadaveric livers are discarded due to severe steatosis. of the of donor livers with steatosis leads to a decrease in both graft and patient models are to increase of the various pathophysiological for the increased I/RP injury seen in fatty livers. It is that this will lead to treatment which will not only increase the number of available donor livers but also decrease the and mortality after liver transplantation and hepatic surgery. Graft and patient survival data in months comparing liver grafts with and without steatosis