Fatty Liver & its impact: HIV, treatment, pathogenesis
Danielle Milano, MD
Boriken Family Health Clinic, New York City
Signs and symptoms
Diagnosis/Role of Biopsy in NAFLD and NASH
Time-line for Progression from NAFLD to NASH
Pathogenesis: 2-Hit Theory
Development of Steatosis
Experimental models: MCD diet
Cytochromes 4A and 2E1
Treatment of NASH
As an HIV-treating provider, my reason for going to the AASLD meeting in October 2003 was to learn more about hepatic steatosis: the spectrum of which ranges from Non-Alcoholic Fatty Liver disease (NAFLD, affectionately pronounced nah-full by the Hepatologists) to Non-Alcoholic Steatohepatitis (NASH) and ultimately cirrhosis. AASLD is similar to Retrovirus in that there is a mix of basic science and clinical trials. What follows is a distillation of information from plenaries and abstracts at AALSD, the Post-Graduate Course which preceded AASLD 1, articles from Hepatology, and the summary of the AASLD single topic conference on Steatohepatitis published in Hepatology, May 2003 2.
We are seeing an explosion of fatty liver in a variety of patient types. Classically, we are accustomed to seeing NAFLD in our non-HIV middle-aged obese patients with Metabolic Syndrome, formerly known as Syndrome X. The nomenclature of this cluster of diseases is changing. The term Metabolic Syndrome as defined by WHO in 1999 includes insulin resistance (IR) or impaired glucose tolerance (IGT) or frank diabetes, plus 2 of the following: hypertension (HTN), dyslipidemia (elevated triglycerides (TG), low HDL-cholesterol), visceral obesity, and/or microalbuminuria 3. Other nomenclature includes the Insulin Resistance Syndrome: IR with 3 or more from the following list: central obesity, dyslipidemia, arterial hypertension, raised serum urate, or microalbuminuria. Yet another is the ATP III (Adult Treatment Panel) definition from the National Cholesterol Education Program 4. They are all quite similar with minor differences in the cut-off's for HDL or blood pressure, or the use of waist-hip ratio versus waist circumference. While important for epidemiologic studies, the differences have no impact on our clinical practice. What is important to note is the addition of visceral obesity to the criteria.
Another group of patients with hepatitic steatosis includes our Hepatitis C (HCV) patients. Steatosis negatively impacts the likelihood of a sustained virologic response (SVR): the more steatosis the less likely to achieve an SVR. The presence of steatosis with HCV also is associated with a more rapid progression of fibrosis. Genotype 3 seems to be a different story from the other genotypes. Genotype 3 is associated with significant hepatic steatosis, and the steatosis will resolve if the patient achieves an SVR when treated with Interferon plus Ribavirin. This is not true for other genotypes: the steatosis does not reverse with treatment of the HCV.
Intriguing and concerning are the HIV patients without HCV whose liver enzymes have been slowly climbing through the years on anti-retroviral medications. Ultrasounds of the liver in these patients show parenchymal changes consistent with fatty liver. The few patients we have biopsied in our clinic have shown significant amount of steatosis, some with rather advanced fibrosis. Elevation of ALT is an insensitive marker for NAFLD and NASH, therefore we may be missing many patients with advancing hepatic steatosis and fibrosis.
So the questions I wanted answered were, 1) Do nucleosides contribute to the steatosis we are seeing in these HIV+/HCV- patients on nucleosides? 2) What is the mechanism of developing steatosis? If you don't understand the mechanism, you won't understand how to treat it. 3) What are the factors associated with progression? Can we identify patients at greatest risk? 4) Are there interventions, such as diet and exercise, or medications such as thiazolidinediones (TZD's) or statins that can slow or reverse steatosis? 5) How many people progress to NASH and cirrhosis. Will we have hundreds of patients on the Liver Transplant list in the future, or is the progression slow enough so we really don't have to worry?
20-30% of adults in America have NAFLD. 10% of those adults with NAFLD meet criteria for NASH and up to 1/3 of those with NASH may progress to cirrhosis. NASH may in fact be the major cause of cryptogenic cirrhosis in which the necroinflammatory and steatotic features are lost: "old burned out NASH". The biopsy criteria for NAFLD and NASH will be discussed later.
Predictors of NAFLD include Type 2 Diabetes, increasing age, severity of obesity, dyslipidemia (especially hypertriglyceridemia) and possible female gender although this is controversial. In women, there may be a more aggressive course of NAFLD to NASH as oppose to an increased incidence, although this is not entirely clear. Genetic and racial factors are also at play with clustering of kindreds and a less aggressive course in African-Americans versus those of European or Hispanic ancestry.
Elevated aspartate aminotransferase (AST) and alanine aminotransferase (ALT) are also predictors of advanced disease. The normal AST/ALT ratio is <1. The AST:ALT ratio is also typically < 1 in NAFLD. In the case where the ratio is reversed, AST:ALT > 1, the disease is advanced with extensive fibrosis, or there another cause of the liver disease.
In patients with Type-2 Diabetes, 75% of patients have evidence of some amount of fatty liver. Conversely, almost all patients with NAFLD or NASH have Insulin Resistance (IR). IR seems to be central to the development of NAFLD and NASH. There are patients who are not overweight but nevertheless develop NAFLD. These patients have central adiposity and some level of IR, despite having a normal BMI. Some thin Asian patients develop IR, which may be due to a defect in CD36, a fatty acid transporter.
There are small groups of other types of patients who develop NAFLD. Certain drugs such as amioderone, tetracycline and valproic acid cause liver steatosis. Certain inherited diseases of mitochondria and lipodystropies also have associated hepatic steatosis.
Signs and symptoms
Typically asymptomatic, the following list of signs and symptoms has been adapted directly from the article in Hepatology summarizing the AASLD Single Topic Conference on steatohepatitis.
Symptoms and physical findings:
Right upper quadrant pain (from pressure on the capsule as the liver enlarges)
Hepatomegaly (abnormal enlargement of the liver)
Bowel dysmotility and small bowel bacterial overgrowth
Waist circumference consistent with central adiposity
Acanthosis nigricans (an eruption of velvety wartlike growths accompanied by hyperpigmentation on skin in various areas of body
Lipomatosis (excessive accumulation of fat in the body)
Lipoatrophy/lipodystrophy (may be under recognized in its focal or partial form)
--Mild elevations of AST and ALT
Seldom exceeding10x upper limit of normal (ULN).
More typically <1.5x ULN
--ALT > AST (AST>ALT suggests advanced fibrosis)
--GGT and Alk Phos elevations
--Hyperglycemia (in 1/3 of patients)
--Hyperlipidemia (usually TG's) in 20-25%
Elevated serum IgA in 25%
Elevated Anti-nuclear antibody (ANA) in 1/3
Abnormal iron indices (not to the levels seen in genetic hemochromatosis)
Diagnosis/Role of Biopsy in NAFLD and NASH
The role of biopsy is controversial in NAFLD. Ultrasound, CT and MRI are insensitive for the diagnosis of steatosis and fibrosis. The blood tests now available for the diagnosis of fibrosis can differentiate absence of fibrosis from advanced fibrosis, but are unreliable for differentiating the middle stages of fibrosis. The Fibrotest uses a-2-macroglobulin, Apolipoprotein-A1, haptoglobulin, gGGT, and total bilirubin, plus ALT for activity. Biopsy should be considered in order to make the diagnosis of NAFLD when LFT's are elevated and other causes of liver disease must be ruled out. Recall, however, that in the list referenced above, ANA can be elevated in NAFLD patients. In women with NAFLD and NASH, markers of autoimmune hepatitis may be elevated when if fact the disease entity is NAFLD. Another cautionary note is that LFT's are insensitive for the diagnosis of NAFLD. In fact, a small series showed that there was no difference in the range of steatosis and fibrosis in patients with normal LFT's when compared to those with elevated LFT's.5
The spectrum of NAFLD on biopsy ranges from pure steatosis to steatosis with lobular inflammation, through to NASH with perisinusoidal fibrosis and cirrhosis with bridging fibrosis. The actual definition of steatosis is fat accumulation of more than 5-10% by weight. (The normal liver is 5% fat by weight.) That is practically estimated by the number of fat laden hepatocytes. The steatosis should be mainly macrovesicular, which is accumulation of a large triglyceride droplet in the hepatocyte. If the steatosis is primarily microvesicular, other causes such as alcoholic liver disease should be pursued.
There have been proposed grading systems, using a range of 0-3 or 0-4 to grade the amount and type of steatosis. There are a number of other factors taken into account in the grading of NAFLD, such as ballooning degeneration of hepatocytes, lobular inflammation, presence of Mallory bodies, etc. There is not yet a consensus of opinion among the expert pathologists as to which histologic abnormalities are necessary for the diagnosis of NAFLD. A widely accepted system is the one by Elizabeth Brundt. This system grades the severity of steatosis on a scale of 1 to 3 (< 33% to >66% steatosis), and fibrosis on a scale of 0 (no fibrosis) to 4 (cirrhosis).6 There is also a proposed classification system which attempts to predicts clinical outcomes. In this system, Class I is simple steatosis, through to Class IV which encompasses NASH. Class IV is further subcategorized according to the stage of fibrosis.
There have been proposed alcohol ingestion cut-offs for when damage is no longer NASH but is actually alcoholic liver damage. This cut-off is set at 20 gm/d of ethanol for men, and 10 gm/d for women. This is a clinical differentiation since the pathologist may not have the patient's alcohol history when examining the slides.
Time-line for Progression from NAFLD to NASH
The progression of liver fibrosis seems to be slow, with progression to cirrhosis rare in mild NAFLD (4% at 10 years). In later classes, III and IV, risk of developing cirrhosis is higher (21% to 28% at 10 years).7 This is based on a small number of patients: there are few prospective or longitudinal studies. Death is frequently from co-morbid conditions, although a study from Japan showed a death rate higher for cirrhosis than for heart disease in a cohort of diabetic patients. 8
A study which looked at paired biopsies in 100 patients over 10 years was presented at AALSD. The results showed approximately 1/3 of patients progressing and 1/5 regressing over time, and half remaining stable. The rate of change was 0.08 fibrosis stages per year (range --1.71 to 2.30) for the cohort. If one looked only at the patients who progressed, the rate was higher at 0.65 fibrosis stages per year. A synopsis of this study was posted on one of the commercially-sponsored hepatitis Websites, therefore it may be on the road to being a popularly quoted study. Beware, because there were several drawbacks to this study. As a matter of fact, I attended the presentation and thought to myself that the data was misleading unless one realized all the caveats attached. The synopsis on the website did not address the entire context of the study. This was a retrospective chart review that pulled biopsy results of NAFLD patients who had paired liver biopsies over a ten year period. They did not tell us the number of charts that needed to be reviewed to find those 100 that fit into the study criteria. 17 of the 100 patients were cirrhotic at baseline and were excluded from several of the analyses. This obviously confounds the data. Although the study can show us that NAFLD can either progress or regress, which is an important point, we can not point to the numbers and predict with accuracy the rate of fibrosis progression/year from this small retrospective study. This is not to imply that this study is not of extreme importance, it is. 9
Pathogenesis: 2-Hit Theory
The hepatologists and basic science researchers are proposing a "2-Hit model" for the development of NASH. The 1st hit is steatosis. For some unknown reason, the presence of hepatic steatosis from excess intracellular fatty acid accumulation predisposes the liver to further damage. The second hit is hepato-cellular injury. There are a number of postulates for the source of the 2nd hit. Reactive Oxygen Species (ROS) causing oxidative stress and lipid peroxidation of lipid membranes resulting in aldehyde formation is one of the proposed mechanisms for the 2nd hit. Proinflammatory cytokines may also be involved and many are know to be elevated in diabetics, such as TNF-a. Connective Tissue Growth Factor is another cytokine elevated in diabetics, and is involved in lying down of connective tissue, and is therefore postulated to be a possible mechanism.10 The Cytochrome P450 system has also been postulated as the mechanism of the 2nd hit, specifically the cytochrome P450 lipid oxidases CYP4A and CYP2E1.
Whatever the origin of the second hit, the end result is stellate cell activation, which causes extra-cellular matrix to be laid down in the liver, and thus ultimately fibrosis.
Interestingly, many of these postulates of the second hit have the mitochondria as a central player. If ROS are the cause of the 2nd hit, mitochondria are the cell's main source of ROS during the production of energy. Outright mitochondria damage or mitochondria respiratory chain defects with resultant ATP depletion have both been postulated as causes of NASH. (Below there is an in depth review of the mitochondria respiratory chain enzymes).
The hepatologists seem to be very excited about stellate cells and their contribution to cirrhosis. If one could avoid turning them on, fibrosis would not occur and in the case of NAFLD, one would be left with a fatty liver but not necessarily a cirrhotic one. Stellate cells are normally quiescent. Activation of stellate cells causes them to migrate to the site of injury, lay down collagen, and behave in manner similar to myofibroblasts. They contract in a similar fashion to muscle cells, which then causes distortion of the hepatic parenchyma. A variety of cytokines are involved in the activation of stellate cells. The researchers feel that cytokines do not occur as isolated chemicals but occur in a milieu, therefore it has been hard to pinpoint one or another as the activating factor because there is no one factor.
After hepatic cellular insult, stellate cells are activated by TGF-b, and IL-13. Extracellular matrix production increases. TGF and HGF cause proliferation and migration of stellate cells. In a normal system without ongoing insult, proteases would be then be produced to induce fibrinolysis, and the activated stellate cells would undergo apoptosis. Thus, the balance between fibrinolysis and fibrosis would be maintained, and the liver would not become cirrhotic. Therefore, researchers are looking for ways to either down-regulate the activators of fibrinogenesis, or up-regulate the mediators of fibrinolysis. Thus the reports you may have heard about studies looking at IL-10 or IL-12 in cirrhotic patients to reverse fibrosis. There had been some buzz about IL-10: in a pilot study, 2/3 of patients had decreases in their fibrosis scores but there have been delays in the publication of the large definitive trial. Other cytokines being studied to reverse fibrosis include IL-13, angiotensin II antagonists, Transforming Growth Factor b antagonists, and IFN-g. Also being studied are pentoxifylline (Trental), isotretinoin (Accutane), Prostaglandins and Phosphatidylcholine (Lechitin). In my opinion, the stellate cells are among the last links in the chain of events leading to cirrhosis and will not be discussed further here, except for one fascinating piece of evidence. It seems as though stellate cells are extremely similar to neural astrocytes, express several neural proteins and may be activated by autonomic neurotransmission. Autonomic nerve fibers do terminate on stellate cells, therefore an area of study is the role of neural control of fibrinogenesis.
Apoptosis, i.e., regulated cell death, will not be addressed in this article either: It is beyond the scope, however, know that apoptosis is a critical component in the development of cirrhosis. Stellate cells after activation eventually need to undergo apoptosis, and hepatic cells should not undergo apoptosis. Of interest, there are a number of ways a cell can initiate apoptosis, through the mitochondria, endoplasmic reticulum, lysosomes or through nuclear receptors (affectionately called "Death Receptors" by the Hepatologists). Apoptosis of hepatic cells is one of the pathologic features of NAFLD, but also is one of the later links in the chain and will not be discussed here. For our patients, the more important questions are 1) Why does steatosis occur? 2) Can we prevent or reverse it? and 3) What are the second hits which cause a steatotic liver to go down the road to fibrosis?
We are all aware of the theory concerning GLUT-4 and it's possible link to the IR seen in our patients on Protease Inhibitors. Not surprisingly, the story about IR is more complicated than imagined.
When insulin engages the insulin receptor, there are at least 9 post-translational pathways activated.
In normal muscle, after insulin is engaged in the insulin receptor, Insulin Receptor Substrate-1 (IRS-1) activates translocation of GLUT-4, a glucose transporter protein, to the cell membrane, which then facilitates glucose transport into the myocyte. A general mechanism of up or down regulation of enzymes is via phosphorylation by either tyrosine or serine kinases: upregulation in the case of tyrosine, and down regulation in the case of serine.
Insulin resistance causes downregulation of IRS-1 through negative feedback on the tyrosine phosphorylation of IRS-1. Free fatty acids also have a negative effect on tyrosine phorphorylation of IRS-1. Thus, one can see the interrelationship between increased levels of trigylcerides and insulin resistance, and a possible link to HIV medications.
Of interest, thiazolidinediones (TZD's) increase glucose disposal via up-regulation of GLUT-4. Another link is that protease inhibitors increase trigylceride synthesis in the liver. One postulate for this effect is increased sterol regulatory binding element binding protein-1 (SREBP-1), which will be discussed in the next section.
Development of Steatosis
FA's come from one of 3 places: de novo synthesis in the liver, chylomicrons after a meal, or from adipose tissue via lipolysis. In the fasting state, circulating TG's are from VLDL destined for muscle and adipocytes. VLDL formation is the way the liver packages TG's for export out of the liver through the endoplasmic reticulum and Golgi apparatus. VLDL formation is complex and is subject to impairment at multiple sites. Formation of Apolipoprotein B-100 is the rate-determining step in VLDL formation and is blunted in IR states.11 Additionally, in obese NASH patients, the Apo-B secretion seems to totally dissociated from postprandial rise of TG.12
Other postulates include defective Microsomal Triglyceride Transfer Protein, which is necessary to insert the TG/VLDL package into the endoplasmic reticulum for eventual secretion into the circulation. Another theory postulates that the VLDL/TG package is derailed from its normal route of export out of the cell through phosphatidylinositol-3 (PI-3) kinase overactivity. PI-3 kinase is involved in insulin signaling. The end result of either of these postulates is the defective export of VLDL out of the cell, and thus lipid accumulation within the hepatocytes.
Sterol regulatory element binding protein --1 (SREBP-1) is a key regulatory transcription factor controlling fatty acid synthase, as well as other lipogenic enzymes. SREBP-1 is known to be markedly increased in animal models of obesity and diabetes and partially the cause of elevated TG levels in those disease states. Troglitazone therapy in animal models of NASH blocks the rise in SREBP-1 and development of hepatic steatosis. As previously mentioned, SREBP-1 is one of the postulates of the elevated TG's seen with PI therapy. Ritonavir blocks the degradation of SREBP-1, which results in increased TG synthesis. Insulin also causes rises in SREBP-1, and thus increased levels of TG synthesis.13 Again, one can easily see the interrelationship between elevated TG's, IR and HIV therapy.
As previously mentioned, mitochondria are felt to play a central role in the development of NAFLD because:
1) They are the primary site for the production of ROS
2) They metabolize free fatty acids (FFA) therefore dysfunction could cause accumulation of lipids in hepatocytes. Hepatic steatosis is inversely proportional to the basal lipid oxidation rate, i.e. mitochondrial fatty acid beta-oxidation
3) Drugs that cause NASH, such as valproic acid, do so through mitochondria damage.
Earlier studies had shown evidence of mitochondria depletion in NASH, however that is no longer felt to be true. Although not depleted, electron micrographs of mitochondria in NASH patients show the mitochondria to be large, swollen, with paracrystalline inclusions and loss of cristae. Megamitochondria can also be seen in liver biopsies but are not necessary for the diagnosis of NASH.
Before we can go any further, a review of mitochondria function is in order. For a free fatty acid (FFA) to be transported into the mitochondria for Beta-oxidation, it must first be linked to a carnitine molecule. The FFA-carnitine molecule is transported across the outer mitochondrial membrane by Carnitine Palmitoyl Transporter-1 (CPT-1). In the mitochondria, CPT-2 dissociates the FFA from the carnitine. The FFA is then free to go through beta-oxidation.
Beta-oxidation of FFA's results in production of acyl-co-A and short-chain acyl-co-A Beta-oxidation is in turn linked to the transfer of electrons to NAD and FAD. The mitochondria respiratory chain (MRC) reoxidizes NADH and FADH2 to NAD and FAD.
There are a number of MRC complexes involved in the production of energy:
Complex 1: accepts electrons from NADH and transfers them to ubiquinone.
Complex II = Succinate dehydrogenase complex: also transfers electrons to ubiquinone.
Nuclear DNA and not mitochondria DNA encode for complex II.
Complex III = b-c complex: accepts electron from ubiquinone and transfers them to cytochrome c.
Complex IV = Cytochrome oxidase complex: accepts electron from b-c complex and transfers them to O2 to form water. Complex IV is coupled to pumping protons from the mitochondria matrix to the intermembrane space.
Complex V = ATP synthase: converts ADP to ATP when protons flow back from the intermembrane space into the matrix.
Previous research had not found alterations in the MRC from muscle or platelets, however a recently published study which looked at the levels of the various MRC complexes in the liver of NASH patients found significantly decreased levels of activity of all the MRC complexes ranging from 42% to 70% of controls. The decreases in activity of the MRC's were greatest in those patients with the greatest fibrosis. 14
In this same study, TNF-alpha was also significantly increased in the NASH patients, as was the prevalence of IR. TNF, mitochondrial dysfunction and AODM are all interrelated. TNF causes electrons to be retained along the MRC, which then causes more ROS to be produced, which causes more mitochondrial dysfunction. TNF levels are increased in patients with AODM or IR. Adipose tissue is the primary source of TNF in the absence of active infection. TNF levels are proportional to the BMI (the higher the BMI, the higher the levels of circulating TNF), and TNF induces IR. In NASH patients, the amount of TNF mRNA expressed is directly proportional to the degree of fibrosis.15 And thus starts a vicious cycle linking IR, TNF, mitochondrial dysfunction and NASH. Previous researchers had proposed a carnitine deficiency as a cause for mitochondria dysfunction for NAFLD. In this study however, carnitine levels were WNL as were CPT-1 and CPT-2 levels. Carnitine deficiency can definitely cause hepatic steatosis in other disease states such as valproic acid induced hepatotoxicity and in TPN feeding. Interestingly, there is some older HIV literature that demonstrated decreased levels of carnitine in patients treated with the earlier nucleoside analogs. Carnitine levels were low in patients either taking AZT16, or with AZT-related myopathy17 and, in patients on either AZT, ddI or D4T with nueuropathy18. Another piece of the puzzle is that children with lipoatrophy secondary to inherited defects in mitochondria function have hepatic steatosis. Therefore, my own postulate is that we will be seeing tremendous amount of hepatic steatosis in our HIV patients on nucleosides. These patients have a double risk in that that have mitochondria damage from nucleosides19 and have some amount of insulin resistance from protease inhibitors.
Oxidative stress can originate from mitochondria, endoplasmic reticulum CYP2E1 and CYP4A, peroxisomes (hydrogen peroxide) and/or inflammatory cells.
Under normal conditions, 1-5% of the oxygen consumed by mitochondria reacts to form ROS. The ROS can cause damage to lipid membranes, to mitochondrial DNA, and to enzymes that contain Iron or Sulfur. Under times of increased production of ROS, the liver increases production of anti-oxidants such as glutathione. A study that looked at the dietary habits of obese patients with and without NASH showed that the patients with NASH had a significantly lower intake of anti-oxidants in their diets, as well as higher intake of saturated fats and cholesterol.12
More importantly, one can see that any impairment of the MRC will result in an increased production of ROS, which then can damage mitochondrial DNA. Mitochondria do not have DNA repair enzymes, as does nuclear DNA.
Another side effect of impaired MRC is that beta-oxidation of lipids is done by the back-up system in the ER: CYP4A and CYP2E1. Hydrogen peroxide is formed from peroxisomal beta-oxidation, and, since the NADPH-dependant lipid peroxidation enzymes have a relatively weaker affinity for oxygen, there is significant electron leakage in the reduction of O2 to H2O.
Experimental models: MCD diet
A number of important studies have been done in mice. A popular experimental model is one that uses mice fed a diet deficient in methionine and choline but high in fat (the MCD diet). These mice develop significant steatohepatitis. Choline and methionine are required for the production of phosphatidylcholine (a component in lecithin), which is in turn required for VLDL synthesis, s-adenosylmethionine and glutathione precursors.
Another group of mice studied are the obese leptin-deficient ob/ob mice. Leptin is a hormone secreted by adipocytes, has a negative effect on appetite, and limits triglyceride accumulation. Leptin-deficient ob/ob mice develop tremendous hepatic steatosis, but no fibrosis, therefore leptin is not crucial for the development of steatosis. Alternately, leptin levels are elevated in NASH patients and in the obese. There is a direct correlation between leptin levels and the amount of steatosis. There is a not a correlation between leptin and either inflammation or fibrosis. Therefore, leptin is involved in steatosis but not fibrosis.20
Cytochromes 4A and 2E1
One postulate for the second hit of the 2-hit model, as previously mentioned, is that the CYP4A and CYP2E1 cytochromes increase ROS. CYP2E1 is elevated in NASH patients 21 and in the obese. With weight loss, the level decreases.22 An important study, which generated a lot of interest at the AASLD meeting, is as follows: wild type mice were fed the MCD diet and developed hepatic steatosis as expected. The researchers increased CYP4A expression 18-fold by using a PPAR-a agonist. (Note: The TZD's are PPAR-g agonists in the periphery. In the liver, the equivalent receptor in the liver is PPAR-a.) Turns out the PPAR-a agonist prevented hepatic steatosis in the wild-type mice fed a MCD diet. This is not what one would expect. One would think that increasing PPAR-a activity would increase ROS and result in increased NASH. Instead, the PPAR-a inducer (Wy-14,643) prevented NASH in the MCD mouse despite marked increases in CYP4A, by decreasing the FFA pool in the liver, which then decreased lipid peroxidation.23 The conclusion is that PPAR-a protected the liver by increasing hepatic lipid turnover, decreasing hepatic triglycerides and decreasing lipid peroxidation. 24
An interesting note is that when CYP2E1 levels are increased, there is decreased tyrosine phosphorlylation of IRS-1 and increased serine phosphorylation of IRS-1: an insulin resistant state. The end result is increased gluconeogenesis and decreased glycogen formation. (Recall the earlier section which discussed glucose entry into myocytes.)
Treatment of NASH
Since IR is central to the development of NASH, improving IR has been postulated as a treatment.
There have been a handful of studies that have looked at TZD's or metformin treatment for NASH.
One study published recently used 4 mg BID of rosiglidazone for 48 weeks in 30 patients with biopsy proven NASH. LFT's normalized, IR improved and in the 22 patients who had evaluable post-treatment biopsies, 10 had reversal of their NASH. There was a price to pay: weight gain. 67% of the patients had an average weight gain of 7.3%, and the weight gain continued even after the medications were stopped! (This is an evil thing to do to an already obese patient: give them something to make them gain weight.) Also note the connection to the use of pioglidizone for antiretroviral-associated lipoatrophy. (Recall that in the studies using pioglidizone for antiretroviral-associated lipoatrophy, there was weight gain and total body increase in fat.) Six months after discontinuation of the rosiglidazone, the LFT's were back up to pre-treatment levels suggesting that treatment may need to be life-long. 25
Metformin is another possibility, and has the added benefit of being weight neutral, or even causing some weight loss. With metformin treatment in mice models, LFT's normalize, steatosis improves, TNF levels come down, but the level of IR does not change (surprising: one would assume that meformin would lower the levels of insulin). 26 27
Statins and fibrates have not been well studied for NAFLD and NASH. Fibrates exert their influence via the PPAR-a receptor, therefore one might assume they would be efficacious for NAFLD and NASH. Early research has not born this out. Likewise, studies using statins have been disappointing. There have been small studies using one or another statin, some of which show improvement in liver enzymes. In general, the studies did not do paired liver biopsies in order to document improvement. One study using atorvastatin did show improvement in histology. My own concern, shared with many hepatologists, is that statins cause mitochondrial damage, as evidenced by the infrequent but potentially life-threatening side-effect of myopathy.
Another problem with the use of statins or fibrates to treat NAFLD is that lowering the levels of circulating lipids might actually have the unwanted effect of increasing hepatic lipid deposition. Simply decreasing serum levels might lower risk of heart disease, but may actually increase intra-hepatic lipid deposition.
Anti-oxidants added to the diet have been studied and the results have not been universally positive. Vitamin E has shown promise on and off, with and without concomitant Vitamin C use or various TGD's. 28
A drug which has not received much attention is betaine, which decreased steatosis and reversed fibrosis in one small pilot study. Betaine can act as an anti-oxidant of sorts to increase the production of glutathione, and is a methyl donor for the production of lecithin. Lecithin is an essential FA required for lipid metabolism. Choline and methionine are both required for the production of lecithin, and betaine can act as a alternate donor of the methyl group for converting homocysteine to methionine.
Before we discuss another set of anti-oxidants, recall that one model of hepatic steatosis is the mouse fed a methionine and choline deficient (MCD) diet. This results in the depletion of S-Adenosyl-L-Methionine (SAM-e), decreased synthesis of phosphatidyl choline ( a component of lecithin) which is required for VLDL synthesis, and decreased synthesis of glutathione precursors. Thus among the agents studies and S-adenosyl-methionine, betaine, and N-acetylcysteine.
Lastly, diet, exercise, and life-style changes do work, but with some caveats. Significant weight loss (average 34 kg) improved steatosis but worsened inflammation in one study. Diet plus exercise in another study improved steatosis but there were non-significant changes in the inflammation and fibrosis scores. For this reason, a small study of 31 patients published in abstract form at the Liver Meeting generated significant interest. 21 (68%) of the 31 patients were able to maintain the weight loss and continue their exercise program. Repeat liver biopsy in 14 patients showed dramatic reversal of steatosis, and 7 of 14 the showed an improvement of fibrosis score. 29
In short, the metabolic syndrome causes accumulation of fat in the liver. Accumulation of intracellular fatty acids (FA's), oxidative stress and mitochondria dysfunction all contribute to hepatocellular injury. Stellate cells are activated which then causes connective tissue to be laid down. With ongoing insult, inflammation and fibrosis continue, ultimately leading to cirrhosis.
1) Liver Disease in the 21st Century: Clinic-Pathologic Correlates; B Bacon, Z Goodman, E Brundt. AASLD Postgraduate Course 2003, Oct 24-25, 2003
2) Neuschwander-Tetri, Caldwell. Nonalchoholic steatohepatitis: Summary of an AALSD Single Topic Conference;; Hepatology. May 2003; Vol 37: No 5
3) World Health Organization. Definition, Diagnosis and Classification of Diabetes Mellitus and its Complications: Report of a WHO Consultation. Geneva, Switzerland: Department of Noncommunicable Disease Surveillance, World Health Organization, 1999
4) National Cholesterol Education Program. Third Report of the National Cholesterol Education Program (NCEP) Expert Panel on Detection, Evaluation and Treatment of High Blood Cholesterol in Adults (Adult Treatment Panel III). Circulation. 2002; 106:3143-3421
5) Mofrad et al. Clinical and histological spectrum of NAFLD associated with normal ALT. Hepatology, June 2003 Vol 37, Number 6
6) Brundt et al. NASH: a proposal for grading and staging the histological lesions. Am J Gastroenterol 1999;94:2467-2474 7) Matteoni et al, NAFLD: A spectrum of clinical and pathological severity. Gastroenterology 1999; 116:1413-1419.
8) Sasaki et al. Mortality and Causes of death in type 2 diabetic patients. A long-term follow-up study in Osaka District, Japan. Diabetes Res Clin Prac 1989;7(Suppl):33-40
9) Adams, Keach, Lindor Angulo Time Course of fibrosis progression in patients with NAFLD. Hepatology. 2003;38(suppl 1):206A Abstract 104.
10) Paradis et al. High glucose and hyperinsulinemia stimulate connective tissue growth factor expression: A potential mechanism involved in progression to fibrosis in NASH. Hepatology, Oct 2001, Vol 34, Number 4.
11) Charlton et al, Apolipoprotein synthesis in nonalchoholic steatohepatitis, Hepatology, April 2002, Vol 35, Number 4
12) Musso et al. Dietary habits and their relations to insulin resistance and postprandial lipemia in nonalcoholic steatohepatitis. Hepatology, April 2003, Vol 37, Number 4.
13) Kakuma, et al Leptin, troglidazone, and the expression of sterol regulatory element binding proteins in liver and pancreatic cells. Proc Natl Acad Sci USA 2000; 97:8536-8541. Again, one can easily see the interrelationship between elevated TG's, IR and HIV therapy. 14) Perez-carreras, et al. Defective hepatic mitochondrial respiratory chain in patients with nonalcoholic steatohepatitis. Hepatology Oct 2003 Volume 38 Number 4
15) Crespo, et al. Gene expression of Tumor Necrosis factor alpha and TNF receptors, p55 and p75, in nonalchoholic steatohepatitis patients. Hepatology, Dec 2001, Vol 34, Number 6. 16) De Simone et al. L-carnitne deficiency in AIDS patients. AIDS 1992;6:203-205,
17) Dalakas et al, Zidovudine-induced mitochondrial myopathy is associated with muscle carnitine deficiency and lipid storage. Ann Nuero. 1994;35(4):482-487
18) Famularo et al. Acetyl-carnitne deficiency in AIDS patients with neurotoxicity on treatment with nucleoside analogs. AIDS 1997;11:185-190.
19) Cote et al. Changes in mitochondrial DNA as a marker of mitochondrial DNA as a marker of nucleoside toxicity in nHIV-infected patients. N Engl J Med 2002;346:811-820 20) Chitturi, et al. Serum leptin in NASH correlates with hepatic steatosis but not fibrosis: A manifestation of lipotoxicity. Hepatology, Aug, 2002 Vol 36 Number 2.
21) Chalasani, et al. Hepatic cytochrome P450 2E1 activity in nondiabetic patients with nonalcoholic steatohepatitis. Hepatology March 2003 Vol 37, number 3
22) Emery, et al. CYP2E1 activity before and after weight loss in morbidly obese subjects with nonalcoholic fatty liver disease. Hepatology, August 2003, Vol 38, no. 2.
23) IP et al, Central role of PPAR alpha-dependant lipid turnover in dietary steatohepatitis in mice. Hepatology 2003:38; 123-132
24) Emilia, et al. Central role of PPAR-a-dependent hepatic lipid turnover in dietary steatohepatitis in mice. Hepatology July 2003 Vol38 Number 1
25) Neuwchwander-Tetri, Brunt et al. Improved nonalchoholic steatohepatitis after 48 weeks of treatment with the PPAR-alpha ligand rosiglidazone. Hepatology, October 2003, Vol 38, Number 4.
26) Lin et al, Metformin reverses fatty liver disease in obese leptin-deficient mice. Nat Med 2000;6:998-1003
27) Maher, Antidiabetic treatment for NASH Hepatology, May 2001; Volume 33, Number 5
28) Harrison, et al,. Vitamin E and Vitamin C treatment improves fibrosis in patients with NASH. Am J Gastro. 2003;98:2485-2490
29) Hickman et al, Benefit of Sustained weight loss and exercise in overweight patients with liver disease -- indicators for success, Abstract #716