Helen S. Te, MD, discusses her topic in Non-Alcohol Fatty Liver Disease.
[MUSIC PLAYING] HELEN S. TE: Hello, my name is Dr. Helen Te and I'm an Associate Professor of Medicine and Medical Director of the adult liver transplantation here at the University of Chicago Medicine. Today I will speak on non-alcoholic fatty liver disease, the current epidemic. Non-alcoholic fatty liver disease, or NAFLD for short, has evolved to become the most common liver disease in the United States. It consists mainly of the deposition of fat in the liver in the absence of significant alcohol intake. The pathological spectrum of NAFLD can range from isolated steatosis, or fat in the liver, and in some cases proceed onto have concomitant inflammation with hepatocyte ballooning degeneration, which represents injury to the hepatocyte with or without fibrosis. More advanced cases can have fibrosis that increase with time and later on some cases can evolve on to cirrhosis. The NAFLD prevalent in our country has been estimated to be at 19% or affecting 29 million adults. This was based on an NHANES III study that ranged from 1988 to 1994, where about 12,000 adults were included. NAFLD, in this study, was defined as the presence of hepatic steatosis on ultrasound, with no significant alcohol intake. When it comes to gender, there was a higher prevalence of NAFLD in men at 20% versus women at 15.8%. When it comes to races, however, there was also a higher prevalence of NAFLD in Mexican-Americans at 24% as compared to non-Hispanic whites at 17.8% and non-Hispanic blacks at 13.5%. In this study, NAFLD was clearly associated independently with the presence of diabetes or simply insulin resistance among those people without diabetes, dyslipidemia, and obesity. In population studies, however, different prevalences of NAFLD have been demonstrated. One such study was conducted in a tertiary referral center in San Antonio, Texas where the prevalence of NAFLD was as high as 46%. These patients were diagnosed by ultrasound. And those with NAFLD were offered an opportunity to have a liver biopsy to be able to detect the prevalence of NASH. Those who had the liver biopsy were found to have presence of NASH at 12.2% of the overall population, or about a third of all of those who were diagnosed with NAFLD. In another population-based study, consisting of an urban population called the Dallas Heart Study, the prevalence of NAFLD then was a little bit lower at 30% but still higher than the overall and NHANES data. In both population-based studies, there was still a higher prevalence of NAFLD in Hispanics as compared to Caucasians and blacks. The risk factors for NAFLD is clearly associated with the metabolic syndrome. And we all know what the metabolic syndrome stands for. But let me just review the most recent criteria as set by the International Diabetes Federation Consensus in 2006-- metabolic syndrome is characterized by central obesity with two of any of the following risk factors-- raised triglycerides higher than 150 milligrams per DL, reduced high-density lipoprotein at less than 40 in males and less than 50 in females, and raised blood pressure or hypertension with a systolic blood pressure of at least 130 or diastolic blood pressure of at least of 85 or a hyperglycemic patient with a fasting plasma glucose of at least 100 milligrams per DL or those who have been diagnosed with diabetes. It's clear that NAFLD incidence, or prevalence, is also following the prevalence of obesity in the United States. I am going to show you how this prevalence has evolved over the last 25 years through data that has been presented by the CDC. This map shows initially that most of the states are actually in the blue zone, which confers a prevalence of about 15 percent or less. And you can see that the darker blue is now appearing on the map as the years go by in the mid 1990s, representing a higher prevalence of NAFLD up to almost 20%. Now we're seeing the beige and the start of red. And in fact this red is now going to turn into maroon color, which signifies at least 30% prevalence of NAFLD in the country. Similarly, the prevalence of diabetes has also been increasing. And again in the last 25 years there has been a steeper rise in the curve. But why is NAFLD important? Well the NAFLD actually confers a predictable risk in terms of a patient's outcome. NAFLD or NASH is associated with a higher mortality from all causes including cardiovascular mortality, malignancy, as well as liver-related mortality. There is also higher comorbidity risk, in terms of cardiovascular events such as myocardial infarction, cerebral vascular events such as stroke, and type II diabetes, as well as kidney failure. The higher malignancy risk has been seen in hepatocellular carcinoma, or HCC, with patients with NASH that have cirrhosis having a cumulative HCC risk from about 2.4% over seven years to as high as 12.8% over three years. There is also a correlation with the incidence of colon cancer, where patients with NASH have been found to have an increased risk with colorectal adenomas as well as colorectal cancers. This is a forest plot up a comparison of the various studies that have shown that the overall mortality of patients with NAFLD as compared with the general population. And as you can see patients with NAFLD clearly have a higher odds ratio of mortality as compared to those who are considered to be the control population. So therefore we should be aggressive about diagnosing NAFLD and staging the disease to be able to intervene appropriately when it is necessary. Diagnosis is made by liver biopsy as the gold standard. Liver biopsy is usually about 1.5 centimeter in length and 1.2 to two millimeters in diameter of a specimen that contains usually, at least, six to eight portal triads to be considered adequate. However this little piece of tissue represents only about one out of 50,000 parts of the liver. And therefore its' validity, as a true representation of the liver, has been challenged by many people. This comes then into the question of whether there are other modalities that could be more useful in terms of diagnosing NAFLD. We are now going to start looking into the non-invasive imaging as a potential tool for diagnosis consisting of ultrasounds, imaging studies such as CT scans, or even MRIs. This is a meta-analysis of the different imaging modalities that can be used to diagnose NAFLD. And you can see the accuracy of each modality as listed here according to the amount of fibrosis that is present on the liver biopsy, as the gold standard. Overall the ultrasound and CT can diagnose NAFLD but they have less accuracy than an MRI or the MR spectroscopy. In a small subgroup of people who have small amounts of steatosis under liver biopsy, however, the MR spectrography was much better than the CT, particularly in those with less than 25% steatosis. We also have other tools to diagnose-- to stage NAFLD actually when it comes to non-invasive means. And these are the elastography tools. One of them is transient elastography, another is a MRI elastrography, and the last one is acoustic radiation force impulse, or ARFI. Transient elastography, or what is known as FibroScan in the market, is a tool that measures the velocity of a low frequency 50 megahertz elastic shear wave that is propagated through the liver. And usually will reflect the stiffness of the liver. That more stiff the liver is, the higher the number will be. A normal liver would usually generate about 5.5 kilopascals, but when significant fibrosis is present at stage three or four, then this number could be higher at above eight. This test has a sensitivity of 91%, specificity of 75%. Although it's positive predictive value is not very good at 52%. However, it has an excellent negative predictive value of 97%, so this would be a very good tool to actually just negate the presence of significant fibrosis in a patient with NAFLD. The limitations of this study, however, is that it is operator-dependent. And body habits does affect the accuracy of the result. So obese individuals will definitely need a different probe than lean individuals. The presence of fat also affects the results so that the range for fibrosis in patients with NAFLD will be different from those who have other liver disease such as hepatitis C. And the presence of active macro inflammatory activity also will affect the results. Very early fibrosis stages will also not give us as much accuracy as later stages of fibrosis or cirrhosis. The second elastographic tool that I want to talk about is the two-dimensional and three-dimensional MR elastography, or MRE. This utilizes still a low frequency wave of 60 megahertz waves and is part of the whole MRI scanning process except that there's an additional driver that's placed on the patient's abdomen that delivers all these waves. The wave that is propagated in the liver is captured by a contrast face scan that will later, after some software algorithmic work, will deliver a color coded map, as what we see here. The presence of blue represents elasticity of the liver but as we go from being elastic to more stiff, we will start seeing green, yellow, and eventually red take over those areas that are more stiff. So a normal liver will again generate a number of two kilopascals But 3.63 kilopascals will indicate the presence of advanced fibrosis. It is noteworthy though that this numbers are not comparable to the numbers that we obtained for transient elastography so they should not be compared head-to-head. The last modality is the acoustic radiation force impulse, or ARFI. This test is incorporated into an ultrasound exam. And it allows specifically localizing that place of examination using the b-mode of the ultrasound. It delivers a longitudinal acoustic pulse. And it generates localized micron-scale tissue displacement and again the shear wave velocity is measured by the ultrasound probe. Comparing all three of them ARFI and transient elastography are more operator dependent and ARFI is also subject to the same limitations as transient elastography. MRE would be less operator dependent but it would require the use of an MR machine and an MR scan, which is more expensive than over two. Between ARFI and transient elastography they have similar predictive value but ARFI appears to have more reliable measurements and allows the localization of the area of interest. So what about non invasive scoring systems? We now have several equations and numbers that we have seen being developed to be able to predict the presence of fibrosis. The most simple all of these is the AST and ALT ratio. Another one is the AST platelet ratio index, or APRI. Again utilizing tasks that are clearly just commonly available in our day-to-day management of the patient. The BARD score utilizes the BMI the AST/ALT ratio as well as the presence of diabetes its absence. BARD score it's very clearly specific for use in NAFLD rather than any other liver diseases. The FIB-4 test utilizes age which is multiplied by AST and divided by the product of the platelets as well as ALT. And lastly the NAFLD fibrosis score or NFS is a logarithmic equation that incorporates the number of years in age, the BMI of the patient, the presence or absence of diabetes, the AST and ALT ratio, platelets as well as the albumin. The table at the bottom of the slide summarizes all the area under receiving operator curve values as well as the sensitivity, specificity, positive predictive value, and negative predictive value of all these tests. In NAFLD it seems that the FIB-4 test and the NAFLD fibrosis score may have a little bit of superiority, in terms of accuracy over the others. Let's move on now to direct serum markers which have been developed to again represent the presence of fibrosis in the liver. One of this is the FibroTest, which measures the haptoglobin, the a2 macroglobulin, apolipoprotein A-1, the total biliruban, GGT. And these are all corrected for age and gender. The area under the curve for this test is pretty good at 0.72 to 0.85 for advanced fibrosis in NAFLD. We also have the European liver fibrosis panel, or the ELF, which utilizes the age and serum markers of matrix turn over, which are hyaluronic acid, tissue inhibitor of metalloproteinase 1, and the amino-terminal peptide of procollagen III. Similar to FibroTest, it has a pretty good area under the curve of 0.87 to 0.90 for severe fibrosis. The most recent area of interest has been in cytokeratin-18 fragment. Cytokeratins are keratin containing proteins that form intermediate filaments and comprise the structure of cytoskeletons in epithelial cells. It is the major intermediate filament protein of the liver. And cytokeratin 16 fragments has been thought to be a representation of apoptosis in the liver and increases with fibrosis in NAFLD. Generating an area under the curve 0.73 for stage two or higher fibrosis. So once NAFLD has been diagnosed, how do we manage patients with NAFLD? Patients with NAFLD should be evaluated for the presence of metabolic syndrome, as this confers a significant amount of potential morbidity and mortality in other organ systems such as the heart or the kidneys or, in terms of diabetes, the endocrine system. None of the non-invasive markers have been validated for use in clinical practice at this time. But FIB-4 and NFS or NAFLD fibrosis score, have better diagnostic accuracy as compared with the others. Liver biopsy may be considered for those who are suspected to have advanced fibrosis or cirrhosis, or those who may have conflicting biochemical and imaging data. We should screen for heptocellular carcinoma in patients who have NASH and have cirrhosis at the same time. And vaccination for Hepatitis A and B should be performed if the patient is not immune. But our main intervention usually will focus on lifestyle modifications as well as pharmacologic therapy. So what would consist lifestyle modifications? Well my prescription would indicate weight loss via diet and exercise. Weight reduction has been shown to result in improved amino transferase levels, as well as the improvement in non-invasive invasive measures of steatosis. Other studies have shown improvement in steatosis and macro inflammation although there was no effect on fibrosis. The goal for weight loss should be about 0.5 to 1 kilogram per week, with a total goal of about eight to 10 percent of the total body weight. It is important to think that fructose, which is commonly found in soda, is associated with increased hepatic fat deposition, inflammation, and possibly even fibrosis. So soda is one factor that should be removed from the diet of the patient. Exercise has been shown to improve insulin resistance, liver enzymes, and imaging results of steatosis. Patients with NAFLD should limit their alcohol intake because, as we all know, alcohol also causes deposition of fat in the liver. One thing that patient can patients can drink though is coffee. Coffee consumption has been associated with less severe NASH as compared to those who don't consume coffee. So, in a study that evaluated for coffee consumption, 58% of patients with NASH, at earlier stages of stage 0 to 1, drank coffee versus fewer patients at 36% of those who had stage two to four fibrosis. So perhaps the initial prescription should include both weight loss and maybe coffee. When it comes to pharmacologic therapy, insulin sensitizers have certainly been the main focus of treatment. There are now a total of nine trials that looked at insulin sensitizers. Five that use glitazones three that used metformin, and one that used both of these medications. Let me just comment first on metformin because it's an easy result to report. Metformin failed to improve any pooled histologic, biochemical, and anthroprometric outcomes. So therefore metformin is not very useful in the treatment of NAFLD. glitazones on the other hand, have had some more promising results. Each of these forest plot represents a specific outcome that was evaluated in trials that were looking at glitazones efficacy. So glitazones resulted in significant histologic and biochemical benefit, particularly in patients without diabetes. And specifically, at the left upper corner there was a decrease in steatosis, right upper corner decrease it hepatocyte ballooning, left lower corner shows the improvement in inflammation, and the right lower corner shows improvement in fibrosis. But we have to keep in mind that glitazones also have known side effects. And this have been reported to consist of fractures in women, weight gain, and in some few cases, even cardiac failure. In one of the most recent glitazone study, that compared pioglitazone versus vitamin E versus placebo in the PIVENS trial, a total of 247 patients were studied. These patients were treated with 96 weeks of pioglitazone at 30 milligrams daily and another group was treated with vitamin E at 800 units daily and the third group was treated with placebo. The results are shown in these graphs. There was an improvement in the NASH composite score in 43% of patients treated with pioglitazone. Versus only 19% in those treated with placebo. Comparing the group of vitamin e and placebo, there was also a significantly higher improvement in vitamin E treated patients at 34% versus 19% in placebo. If we dissect this to the specifics, there was an improvement in the amino transfer raised in the pioglitazone and vitamin E treated group as seen in the left upper graph where vitamin E and pioglitazone are in the blue and green lines versus placebo, which is in the yellow line. Similarly, there was also an improvement in insulin resistance with pioglitazone as represented by the blue line in the left lower corner as compared with vitamin e and placebo, which did not have much effect on insulin resistance. However the right lower graph shows a higher weight gain in the pioglitazone group again represented by the blue line as compared to vitamin E and the placebo groups. It's again important to think about the side effects of these drugs and now this time vitamin E has also been associated with a higher all-cause mortality rate and prostate cancer in men in other studies. There were no side effects seen in this particular study. Other therapies have been tried for NAFLD and some of these has consisted of statins where studies were pretty small and there was no randomized controlled trial. But the statins have shown improvement in labs and histology. Orlistat has also been shown in two randomized controlled trials, although the results were quite conflicting. But overall there were improvements in the labs and steastosis as seen by ultrasound. But these were not anywhere outside of the weight loss affect that what's expected from the weight loss caused by the orlistat. Ursodeoxycholic acid, omega 3 fatty acids, both show the improved laboratory test. But there's no data or histology. Ezetimibe, or Zetia, was also evaluated. And there was limited data on histology. But there was an increase in the hemoglobin A1c, and the hepatic long-chain fatty acid levels. Sitagliptin was studied in one small trial, which showed improved labs but there's again no data on histology. And eicosapentaenoic acid was ineffective in a randomized controlled trial. Future therapies are being worked on and pentoxifylline certainly has also been evaluated in small studies where histologic benefits were seen. Pentoxifylline is a TNF-alpha inhibitor. The one that has gathered the most attention recently is obeticholic acid which is a farnesoid X receptor agonist. In fact it's phase 2b trial was terminated early, because of an obvious benefit from the drug. And they are now looking into a phase III trial. Cysteamine bitartrate, GFT505, cetuximab, and GR-MD-02 all are being studied currently in phase one and phase two trials as potential medications to be used against NAFLD. The last two cetuximab and GR-MD-02 are antifibrotic agents. Vitamin D deficiency has also been documented in NAFLD and this correlates with the severity of the disease. But there's no data to inform us if vitamin D supplementation is going to have any effect on the disease process of NAFLD. Last but not least bariatric surgery has certainly been studied as well as a treatment for NAFLD in those who are morbidly obese. And across the board bariatric surgery has been beneficial for patients with NAFLD. Steatosis, steatohepatitis and fibrosis appear to improve or even completely resolve in the majority of patients after bariatric surgery when they have achieved weight loss. Bariatric surgery resulted in 19% to 47% reduction in BMI. And there was also complete resolution of type 2 diabetes in those who had undergone a Roux-en-Y gastric bypass. If you look at the histologic effect of bariatric surgery, the plot here is a little bit scattered but overall there was an improvement in steatosis 91.6 %, improvement in steatohepatitis an 81.3% and an improvement in fibrosis and 65.5%. | So in conclusion NAFLD certainly has come a long way since it was first diagnosed although we are also dealing with a higher prevalence. NASH is increasing in prevalence in the United States and will continue to be a major cause of morbidity and mortality in the future. Detection of advanced fibrosis will soon be facilitated with non-invasive means, which consists of a combination of elastography and fibrosis score or direct serum marker. Obviating the need for liver biopsies, except perhaps during a time when these markers are contradicting the imaging modalities. Lifestyle changes improve NASH but these, as we know, are not easy to implement or sustain in many of our patients. pioglitazone has beneficial effects on NASH but it's side effects may potentially limit its utility. Vitamin E improves NASH with out serious side effects in NASH clinical trials. Although side effects from vitamin E have been reported in other trials. And several new agents are currently under study in the pursuit of better therapy for NASH. This concludes my presentation for today. And I thank you for spending the time with me. And I hope that was interesting and educational for all of you. Thank you.