Hello and welcome to the Non-Alcoholic Fatty Liver Disease SIG Programming. I'm Stephen Harrison. On behalf of myself and Dr. Yaron Rotman, I want to thank you for joining us for today's symposium that will focus on the clinical management of non-alcoholic fatty liver disease in NASH in 2020. We have an all-star cast today that will be giving you updates on how to identify patients with NASH who are at greatest risk of disease progression. We're still a long way to go, but we have narrowed the aperture a bit on how to identify these patients and I can't wait to hear what's in store for us in that lecture. We're also going to talk about current treatment options for patients with NAFLD and NASH and explore where we may be headed. There's going to be an interesting debate today and the debate is going to focus around targeting NASH resolution or fibrosis improvement. We know those are the two FDA-approvable surrogate endpoints. We're going to have one proponent for NASH resolution. We're going to have another proponent for fibrosis improvement. We're going to talk about what the future holds, what the future holds in reference to biomarkers, NASH diagnostics, and even therapy, and then we'll wrap it up at the end. And in this power-packed one-and-a-half-hour session, I hope you walk away knowing significantly more than when you came in. We're thrilled you're here today. Can't wait to get started. Thank you. Good morning. I'd like to thank the ASLD and the chairs, Dr. Yaron Rotman and Stephen Harrison, for this kind invitation. My task today is to answer the following question. How do I identify patients with NASH who are at greatest risk of disease progression? Those are my disclosures. So the objective of today's talk is to learn how to address the most common questions by your NASH patients in terms of staging and monitoring their disease using diagnostics, whether they're available today or in the near future. Whether you're a researcher or a clinician, when your patient comes to you, the very first question will be, how severe is my disease, which translates into staging workup. Once you stage the patient, the patient's going to ask you next, what do I do next? And this is when you start if therapeutic interventions are needed or not. And once you decide, you're going to follow up the patient. And along the journey, the patient is going to ask you, how am I doing? Which is disease monitoring. So let's start answering the question, how severe is my disease or staging? This is the natural history of non-alcoholic fatty liver disease. As you know, NAFL or steatosis can progress in some patients, especially those at risk, such as those with metabolic syndrome and diabetes to NASH. And NASH over time can progress to fibrosis and cirrhosis. We now know that NASH with stage 2 fibrosis and higher, F2, have more risk for liver-related outcomes, such as ascites, hepatic encephalopathy, and esophageal variceal bleed. And there are many studies that have showed that. This is an example of a study that at baseline, patients with F2 and higher had higher risk for overall mortality and liver disease, severity of liver disease. These studies were done with liver biopsy. However, liver biopsy has its own issues that we know of, such as liver histological features have low kappa, it's invasive, trials can take longer, there's sample size variation. And because the disease is so prevalent, it's impractical to do liver biopsy longitudinally. Therefore, we started examining and testing non-invasive testing. And here I summarize the most commonly used and the most promising, one in the near future. They are in two buckets, one's serogical markers, the other one is imaging markers. The serogical markers can be simple FIP4-NAFL fibrosis score or complex L4C3. The imaging page can be transient allostrography or VCTE, the MAR-based allostrography, multiparametric, and ARFI. And here I show you with errors what the ASLT guide has endorsed. The FIP4-NAFL fibrosis scores are simple, includes age, as well as PMI with NFS, and some lab's parameters. What they do is the following, they have two cutoffs. If you're patient in the low cutoff zone, that means your patient is fine and he just need to follow up with primary care physician. If you're patient in the high cutoff zone, that means your patient likely has advanced fibrosis and needs an intervention. However, there's that group of patients that they fall in the indeterminate zone, and I'll tell you in subsequent slides what we do with these patients. Pro-C3 is a promising serum biomarker for type 3 collagen and is independently related to fibrosis. Studies have shown correlation with fibrosis stages, but also has shown correlation between Pro-C3 levels and other histological features, such as lobular inflammation, steatosis, NASCOR, and baloney. I also show you here studies where Pro-C3 was combined with other parameters to increase its accuracy. Again, it's promising, and I'll show you later on its role of monitoring disease progression and regression. ELF is another very promising fibrosis biomarker, has been around for some time, has gained some more interest recently, and many studies have been done. And ELF includes hyaluronic acid, P3MP, and TMP1. In a recent meta-analysis, the ELF test showed high sensitivity, but limited specificity to exclude significant fibrosis, which is F2 and higher, and advanced fibrosis, which is F3 and F4. For instance, the threshold of 7.7 had sensitivity of 90% to exclude fibrosis. Nevertheless, if you wanted the specificity of 90%, the cutoff for significant fibrosis, which is F2 and higher, was 9.86, and for advanced fibrosis, which is F3 and F4, was 10.18. Again, we need more studies on ELF, and I will show you later on its role in disease monitoring. I showed you that in the nafl-fibrosis cordon and FIB4, also the Pro-C3 and ELF, they have two cutoffs. And many patients fall in the indeterminate zone. In this important study by Ensti and colleagues, he suggested what to do with these patients in the indeterminate zone, which is very much doing a second non-invasive testing, and even a third non-invasive testing, and you will pin down your patient and know exactly what's the severity of his disease. Moving on to imaging-based staging. The VCT, or FibroScan, is known by most of us. It has probes that quantify steatosis via the cap or liver stiffness, which reflect fibrosis. Easily done in the office. Another very accurate technique is the MR or listography. It gives you an accurate estimate of the stage of fibrosis from F0 to F4. For instance, its accuracy reached 0.93 for those with advanced fibrosis, showing you how accurate it is, however, it's not available everywhere. If you mention MRE, you want to mention also MRI-PDFF. It's not for staging, but it's similar to the VCTE. MRI-PDFF can quantify fat, and here it shows you how fat is quantified in different areas of the liver and colloquialized at the end of clinical trials, for instance, to assess fibrosis improvement or worsening. So if you couple this MRI-PDFF with a model listography, you'll have an idea about fat as well as liver stiffness. Now you stage your patient, and you told him or her how severe their disease is. They're going to ask you, what do I do next? This is when you start thinking about indications for therapies. Back to the same natural history I showed you earlier, NASH with F2 and higher are those at increased risk for morbidity and mortality, and thus NASH with F2 and F3 are the group that we focus on in phase 3 clinical trials in NASH. There have been multiple attempts to predict when pharmacological therapy can be started non-invasively, which is similar to the F2 and higher patients. So for detecting steatohepatitis, the multi-parametric scan can be very helpful. There have been multiple studies looking at how to assess the amount of fat and dose with F2 and higher using these various techniques. And recently, there has been a concept about creating prediction for NASH patients with NASCore, which is what we use in clinical trials histologically, of 4 and higher and F2 and higher. The FAST score has been published, as well as the NIS4, and I'm going to show you from this meeting the mass and the massive score. The liver multi-scan, it's a very helpful technique that provides quantitative biomarker of hepatic fibroinflammation. So it combines the iron-corrected T1 sequence with the MRI PDFF sequence, and this corrected T1 or CT1 is basically a reflection of inflammation of fibrosis. It even gives you a score called the LIF score, which represents inflammation of fibrosis, and you can monitor it over time. Moving to the concept of NASH patients with NASCore of 4 and higher and F2 and higher. Philip Newsome and his colleagues published a paper using transient osteography technology, and what they used is the CAF or steatosis, the liver stiffness, and ASD in a formula to predict those that in histology have NAS4 and higher and F2 and higher. And they named this the FAST score. It has excellent area under the curve in the derivative cohort as well as the validation cohort, but again, you see the same concept in the rule-in zone, rule-out zone, you have two cutoff values, and in between you have indeterminate zone. If your patient has a score of 0.67 and higher, this patient likely has NAS4 and higher and F2 and higher, so clinical trial indicated and in the future likely therapies. However, if the patient falls in the indeterminate zone or rule-out zone, like yellow and green that I'm showing here, be a little bit cautious because if AST is normal, they might fall within this green and yellow, but the stiffness might be high. So always keep an eye on the stiffness. Same concept, the NIST score came to look at these patients with NAS and NAS4 and higher and F2 and higher. And this time it's a serum test published by Harrison et al. in the GOLDEN trial, and they used microRNA, fibrosis markers, and hemoglobin A1C. It performed very well with area under the curve of 0.8 in the discovery cohort, and it was also good in the validation cohorts. Yet, we still have the indeterminate zone, and this is a promising test expected to be used in the future. In this meeting and in the recent EZO meeting, we presented the MASSIF score, which is similar to the FAST score. It relies on steatosis assessment via PDFF, stiffness assessment via model historiography, and AST assessment. And in the derivative and validation cohort, the area under the curve was excellent, and there was a roll-in and roll-out zone. Lastly, also in the last EZO meeting and this ASLD meeting, MASS score was presented by multiple authors, and this MASS score relied on the metabolomic testing, which is serum-based metabolomics to detect patients with NAS, NAS4 and higher, and F2 and higher. Also included these metabolomic lipids, AST-ALT, be in mind this meeting platelet was added, and the discovery cohort as well as the validation cohort have achieved excellent area under the curve, another promising test. Now you determined how severe is the disease, you determined who can benefit from therapy, your patient is going to ask you next how I'm doing in the next months or years he's following you. So you're going to start thinking about how to monitor the disease progression and regression, and also over there in therapy, you want to monitor response to therapy. So let's take a look. Remind you that most phase 3 trials are done with histology, and the outcomes or the primary endpoints that are required by FDA for approval is NAS resolution without worsening of fibrosis or fibrosis approval by one stage without worsening of steatohepatitis. This topic is going to be debated later today. But again, biopsy can be invasive and impractical for clinical practice and longitudinal follow-up. So let's see non-invasive testing in terms of monitoring disease progression. Fib4-NAF5 fibrosis score have shown correlation with survivor probability. Jérôme Brossier has shown that transient allostrography in NASH patients also can protect over all survival, showing you that non-invasive testing plays a role in predicting survival. Shadab Siddiqui and Arun Sanyal from the NAS CRN have shown that Fib4-NAF5 fibrosis score predicts fibrosis progression with an excellency statistic. Nevertheless, Hegstrom and recent two publications have shown that Fib4 can predict those that they will have severe liver disease. Nevertheless, many of those that develop severe liver disease, almost 50% of them had consistently low or intermediate Fib4, again, 50% of them, calling for more studies in terms of Fib4 monitoring over time. ELF. Very good two studies, one from the SEMTISIMAP trial presented by Stephen Harrison and another one from the GALACTIN trial came from Nagachalasani's group, where if you look in the left, if your patient had baseline ELF value of 9.76 and higher, this patient were more likely to progress to cirrhosis. If you look at the right and you raise the ELF threshold to around 11.3, the patient had 11.3 and higher, they were more likely to have liver-related event, very promising in predicting disease progression. This is a study we did in multicenter NASH cohort, and we said, okay, MR listography can give you the exact stage from F0 to F4, and we know these cutoffs. We asked this question, what is the cutoff on MRE that correlates with decompensation in NASH patients, and we found that this cutoff is 6.48. The take-home lesson from this study is when your patient reach about 6.4 or 6.3, you want to start thinking that they might develop decompensation. Again, this needs to be confirmed on longitudinal study. Let's look at non-invasive testing and how they monitor response to therapies. ALT has shown that it correlates with histological response. So in the Flynn trial that used the butylic acid OCA, ALT dropped by 17 units over 6 months has correlated with histological response. Great news, we'll keep an eye on the ALT. Rohit Lumba has proposed this MRI-PDFF concept that more than 30% reduction correlates with histological response, and he has shown that in initial very small studies that after that we had many MRI-PDFF phase IIa studies, but we wanted more evidence. And that has been shown recently. On the left, the Flynn trial, MRI-PDFF correlated with histological response. On the right, there is a mitral room trial where MRI-PDFF also correlated with histological response. So if you are an MRI-PDFF responder, you are more likely to have histological response with excellent response ratio. Moving on, because of the REGENER8 trial, which is the opiatocolic phase III study that has achieved the primary endpoint of one-point improvement in fibrosis, we were able to test non-invasive testing. Here, remind you that the biopsy was done at 18 months for internal analysis. In this figure on the left, VCTE stiffness has dropped both on the OCA and the placebo at 18 months when the biopsy was done, in addition to the 24 months beyond the biopsy, and continued to drop more on OCA than the placebo. Similarly, ALT dropped at 18 months, continued to drop by 24 months more than the placebo before it moved in the same direction. Jérôme Brossier also presented in this ESOM meeting the same thing from the same study, but this time using fibrometer-fibrometer VCTE and the FAST score that we talked about, and it shows you here that all of them dropped on the OCA by 18 months compared to the placebo, showing that or suggesting that non-invasive testing can be used in monitoring response to therapy. Also, the liver multi-scan has shown some longitudinal data. On the left, sub-analysis in a small number of patients in the OCA trial. On the top, you see the patient on the OCA. On the bottom, you see the patient placebo. And you start from the left. The red represents those with high LIF score. And you see over time, this red turned green, which translates into lower LIF score in the OCA group compared to the placebo. They did not have that. Liver multi-scan did the same thing in the NGM2A2 study. There were two doses, one and three milligram. And compared to the placebo, the CT1 as well as the liver FAT contact improved better in the treatment group, showing the ability of liver multi-scan also to monitor disease response. Finally, the PROCY3N-ALF. On the left, you see the MGL3196 study. And in red, you see PROCY3N-ALF in the treatment group that they improved compared to the placebo that they are in blue. On the right, the NGM2A2 study that also had histology, you see both PROCY3N-ALF dropping over time, in a short time here, more in the treatment group than the placebo group. So today I have shown you how to answer common questions for your patients, which is the principle of their disease diagnosis staging monitoring. So for the staging, you have a menu of tests. For therapeutics, we talked about it and we have this new concept and algorithm of NASH with NASH more than four, NF2 and higher. And finally, the monitoring, monitor disease progression and response to therapies. And with these menus, we always have to use clinical judgment as medicine is an art. We have to put the best test together for our patient well-being. With that, I'd like to thank the chairs again for this invitation. Hello, my name is Manal Abdel-Malik. I'm professor of medicine at Duke University, and it's with great pleasure that I present on the current treatment options for our patients with NAFLD and NASH. I'd like to thank the program organizers and Dr. Harrison and Rotman for the opportunity to speak today. We'll be talking and reviewing in the next 20 minutes the current treatment options for NAFLD and NASH. Clearly, this is a very dynamic process that can progress and regress. However, lifestyle modification is a must for all patients, irrespective of where they are in their disease stage. And in the absence of no FDA-approved treatment for NAFLD and NASH, treatment of comorbidities contributing to NAFLD is really only the backbone that we currently have. So it's essential initially to identify all modifiable risk factors. Now, clearly, there are non-modifiable risk factors such as ethnicity and age, family history and gender. However, there are several modifiable risk factors and things that we can change as practitioners, such as obesity, comorbidities of metabolic syndrome, diabetes, dietary risk factors, quality of sleep, or even behavioral factors that contribute to obesity and social habits, both smoking and alcohol consumption that can contribute to fibrosis progression. The current treatment of NASH, in my opinion, is one that is patient-centered and individualized therapy. With the intent of first tackling the overweight and obese status, which contributes so heavily to the development of NASH by addressing weight and sedentary lifestyle, and then really drilling down and focusing on the co-factors, such as dietary modifiers for disease, alcohol, smoking, fructose, coffee, followed by addressing each and every comorbidity, which contributes potentially to the acquisition of both NASH or fibrosis progression. So, the model of care for NAFLD and NASH is one of an integrated approach. It may require multidisciplinary care between a hepatologist for patients particularly with advanced disease, endocrinologists to manage metabolic syndrome, and nutritionists to help facilitate better education with regards to optimal dietary habits. Cardiovascular risk reduction is essential in all our patients in the management of dyslipidemia and diabetes as cardiovascular disease is the leading cause of morbidity and mortality. We need to screen and treat comorbid conditions and really evaluate the infrastructure we have with the intent of developing integrative weight loss centers or facilities that will allow for not only behavioral therapy, but pharmacotherapy or even bariatric or endoscopic approaches to the management of obesity, as well as individualized pharmacotherapy for the management of prediabetes or diabetes, as well as specialty care and hepatology that will address those patients with advanced liver disease for surveillance of epithelial cell or carcinoma viruses and, of course, vaccination for hepatitis A and B. The treatment of obesity is the foundation of care, not only for NAFLD, but particularly for those patients who have NASH. A 5% reduction in BMI translates into a 25% relative reduction in liver fat by MRI-PDFF. And, in fact, a systematic review of 23 studies demonstrated that weight reduction anywhere between 4% to 10% consistently improves liver fat and liver aminotransferases. And as you can see here, a recent meta-analysis published in JAMA in 2019 of several different studies consistently demonstrates that weight loss improves liver aminotransferases and, likewise, consistently demonstrated that weight loss across many different studies improves hepatic steatosis. But more importantly is the question, is what does weight loss actually do to the liver histology of NASH? And in a very eloquent study by Dr. Vilar Gomez, published in Gastroenterology in 2015, evaluating paired liver biopsies of 261 patients with biopsy-proven NASH who underwent a 52-week lifestyle intervention demonstrated that higher degrees of weight loss translates into not only improvement of simple benign steatosis, but resolution of necroinflammatory injury, NASH resolution in a majority of patients, and even a one-stage fibrosis regression in up to 45% of patients who were able to achieve a weight loss of greater than 10%. The unfortunate reality is very few of our patients can achieve these levels of weight loss, and under 10% can actually sustain it beyond one year. However, exercise is essential because there's evidence that it retards hepatocarcinogenesis, and if we utilize the Fafa mouse, which is a hyperphagic, hyperinsulinemic, hyperlipidemic mouse compared to the lean littermates, these mice, when subjected to a sedentary lifestyle, develop liver cancer. However, when introduced to a voluntary exercise, daily exercise intervention on the wheel, not only develop decreased cancer burden, but decreased number of cancer foci, retard weight gain, while improving serum insulin and insulin resistance, and in fact, there are epidemiologic studies that demonstrate that vigorous physical activity of more than five days a week decreases the risk of hepatocellular carcinoma by 44%. And what's even more exciting is that exercise and weight loss even decreases portal pressure. In a recent study of 50 patients with compensated cirrhosis, of which 92% were Child's A, and the hepatic vein pressure gradient requirement for the study was greater than six, but 72% had a hepatic vein pressure gradient of greater than 10 millimeters of mercury, who underwent a 16-week intensive lifestyle intervention. The average body weight reduction was 5.2%, and interestingly, 52% of patients with cirrhosis were able to achieve a weight loss of greater than 5%. The average change in hepatic venous pressure gradient was 1.7 millimeters of mercury, which translated into approximately 10.7% reduction, which was significant compared to baseline, but what was even more dramatic is for those patients who were able to achieve a 10% or more change in body weight had a 25% reduction in hepatic venous pressure gradient. We also know that diet matters in our patients. Excess calories, excess fats, the quantity of fructose consumption, which has been independently associated with liver injury and fibrosis, as well as modest alcohol intake and even coffee. The avoidance of fructose is essential for the management of patients with NAFLD and NASH. Not only does it increase appetite, but it alters the gut microbiota, is a potent inducer of de novo lipogenesis, induces insulin resistance via SREBP, and also because of its utilization of hepatic ATP will decrease hepatic ATP sores, induce hyperuricemia, and induce reactive oxygen species. So fructose has been associated with obesity, hyperlipidemia, low HDL cholesterol, even diabetes, and increased sweetened beverage intake has been noted in patients with NAFLD compared to matched controls, and sweetened beverage intake of greater than seven beverages per week has been associated with fibrosis in a dose-dependent manner in patients with NAFLD and NASH. So essential that we encourage our patients to discontinue all sweetened beverages. Coffee consumption has been demonstrated to decrease the risk of hepatic fibrosis and cirrhosis in a meta-analysis that was recently published in 2015. Consistent evidence that coffee is good. And how much coffee is good? Three cups a day has been inversely associated with not only the development of NAFLD compared to controls, but the development of NASH as well as hepatic fibrosis. So coffee consumption is protective against NASH compared to those with steatosis, and also favors milder hepatic fibrosis compared to more advanced hepatic fibrosis. So our patients should be drinking coffee. But what's even more interesting is that coffee has been associated with a decreased risk of hepatocellular carcinoma in a recent meta-analysis. And as you can see across both cohort studies and case-controlled studies, there is a reduction in the relative risk of hepatocellular carcinoma. And in over 2 million patients in cohort studies and over 1,800 cases versus 2,600 controls, two cups of coffee reduced the risk of hepatocellular carcinoma by 35%. So I encourage my patients, in addition to discontinuing fructose, to start drinking coffee. Red meat and carbohydrates have been associated with fatty liver disease compared to patients who have a normal liver. Not only do patients with fatty liver disease consume more meat protein, but clearly more carbohydrates. But interestingly, the meat protein in the form of red meat, as you can see, based on quartile ratios, that those patients who were in the fourth quartile had anywhere between a three and a half to nearly an eight-fold increased risk of NAFLD. And that patients based on meat protein intake, that those who were in the 14th quartile, about 46%, had evidence of NAFLD. So the Mediterranean-style diet, while there is not substantial randomized controlled studies evaluating diet, is consistent with what the data shows that complex carbohydrates, red meat, refined sugars, hydrogenated fats are all bad, yet the Mediterranean diet enriches for coffee or tea, glass of wine a day, omega-3 fatty acids, polyunsaturated fats, antioxidants, and even choline. So modest alcohol consumption in patients with NAFLD has been shown to actually be associated with improved hepatic steatosis and has multiple metabolic factors that may actually be favorable. Lowering of triglycerides, increase in HDL, and a lowering of systemic insulin resistance and proportion of patients who drink modestly have a lower risk of diabetes. Smoking also is essential to counsel our patients about because heavy smoking use increases the risk of fatty liver disease almost eight to nine-fold. And there's animal data to actually support this. So in the Zooker rats who were exposed to two cigarettes a day for four weeks, as you can see here, not only did the non-smoking controls compared to the smoking controls, the smokers got some necroinflammation, but the obese animals not only developed NASH, but developed NASH with fibrosis. And this correlated to increased necroinflammatory score and increased NASH activity score, as well as an increase in ALT with exposure to smoke. So both the European and the American Association of Study of Liver Disease has published guidances on lifestyle changes, suggesting that it is of course mandatory in patients with NASH, exclusion of fructose intake, a limitation of alcohol intake, avoidance of smoking, for the loss of three to 5% weight loss, improving steatosis, and seven to 10% weight loss being the target for lifestyle intervention and the target that improves not only liver enzymes, but liver histology. Dietary recommendations should be considered, should consider energy restriction and adjustment of macro composition in accordance with a Mediterranean diet, and both aerobic and resistance exercise reduces liver fat, but of course the challenging part is to tailor that exercise to a patient such that there's long-term maintenance and compliance. What about the anti-obesity approaches, both consideration of FDA approved weight loss drugs, as well as bariatric surgery in our clinical practice? Well, there are several FDA approved anti-diabetes, I'm sorry, anti-obesity drugs. One needs to consider, of course, the common side effects, which are anticholinergic in effect, can increase heart rate and blood pressure, and induce some GI symptoms of nausea or constipation. And in patients who already have obesity and diabetes, there's a consideration for drug-drug interactions. Overall, the placebo-subtracted weight loss for the FDA approved anti-obesity drugs varies between three to 10%, and unfortunately there's very limited data on anti-obesity drugs on the histologic outcomes of NASH. Two drugs have been studied in randomized control studies, Orlistat and loraglitide. Orlistat, in a parallel group, open-labeled 24-week randomized trial of 170 patients with obesity and fatty liver disease by MRI-PDFF, demonstrated that Orlistat actually downgrades the severity of fat by MRI after 24 weeks compared to placebo. And in the right upper end of the slide, you'll see a representative MR-PDFF decreasing quantitative liver fat from 29% down to almost 8%, with approximately a three to five percent weight loss. Now Orlistat has also been studied some time ago by Steve Harrison in comparison to diet and vitamin E in patients with biopsy-proven NASH, 50 overweight subjects with a BMI greater than 27, some of whom received vitamin E and diet with or without Orlistat for 36 weeks. And as you can see, the Orlistat group lost 8.3% weight and the diet and vitamin E group lost 6% weight. But what was interesting about this study is both Orlistat and diet and vitamin E were comparable in their ability to improve NASH and that this improvement in NASH correlated with weight loss. But those patients that achieved higher thresholds of weight loss had a more dramatic improvement in the necroinflammatory scores with an improvement in the overall NASH score. But as you could see, a slightly more improvement in steatosis ballooning and necroinflammation. Baraclotide has also been studied in a 48-week phase two study of 52 patients with biopsy-proven NASH with at 1.8 milligrams versus placebo, with the primary outcome being resolution of definite NASH with no worsening of fibrosis. Small pilot study, however, as you could see, 39% versus 9% on placebo with NASH resolution and fewer patients in the baraclotide treatment arm actually had fibrosis progression compared to placebo. This correlated with about a 5.5% weight reduction and an improvement in liver aminotransferases. And likewise, somaglutide has been evaluated in a 72-week phase two trial of 320 patients with NASH fibrosis stage one to three with placebo versus a dose escalation of somaglutide subcutaneously daily with the primary outcome being resolution of NASH and no worsening of fibrosis. And as you could see, there was an incremental improvement in all doses of somaglutide that were studied meeting the primary endpoint of improvement in NASH without worsening of fibrosis. The safety of GLP-1s in NASH is consistent with those of other trials in other disease areas and the most common adverse event was gastrointestinal. NASH also results in resolution with bariatric surgery and a five-year follow-up study from France demonstrated that patients who underwent bariatric surgery, 84% resolved NASH without worsening of fibrosis at five years. And there was an incremental improvement over a five-year period of time in fibrosis regression. However, not much if you already had a well-established cirrhosis. But fibrosis regression occurred in those patients who had mild baseline fibrosis, nearly 63% versus 45% fibrosis resolution at five years in those that had F3, F4. The ASLD and the ESL guidelines do speak to targeting insulin resistance and targeting oxidant stress. Clearly these guidelines need to be updated and it's anticipated that new practice guidelines will be coming soon. Although metformin did not demonstrate any improvement in NASH and is not recommended in the guidelines, I'm going to make a push and consideration for its use in clinical practice as it does improve insulin resistance and there is data in a meta analysis that it decreases the risk of hepatocellular carcinoma in diabetics and improves survival in patients with diabetes from hepatocellular carcinoma. The PIVNS trial was a 96-week study of 247 patients with non-diabetic non-cirrhotic NASH for whom vitamin E versus pioglitazone versus placebo was evaluated. And as you could see, both vitamin E and pioglitazone met the primary outcome of improvement in NASH without worsening of fibrosis. However, there are caveats to consider. The effects were not sustained. Vitamin E increases potentially mortality from hemorrhagic stroke and prostate cancer, and pioglitazone increases weight, fluid retention, osteoporosis, and potentially prostate cancer. Statins should be considered in patients with NASH. In an open-label study evaluating risuvastatin, 19 out of 20 patients improved liver enzymes and had resolution of NASH. And in a meta analysis of studies, statins are indicated for cardiovascular risk reduction. They improve LDL cholesterol and liver function. They're safe, certainly, to be used in patients with NASH. And consistent histologic data to support the use of statins, however, on NASH is lacking. However, statins are pleiotrophic. There are several reasons to consider why these class of compounds may have both an anti-inflammatory and anti-fibrotic effect on NAFLD. And there's certainly evidence that statins consistently in a meta analysis decreases the risk of hepatocellular carcinoma. And lastly, I'd like to point out that in a systematic review and meta analysis, there's evidence that statins decrease portal pressure and reduce variceal hemorrhage in cirrhotic patients. Lastly, angiotensin receptor blockers, although there's not human data, there is evidence in animals, the MCD fed, the methionine choline deficient mouse, that there is a reduction in hepatic steatosis with the introduction of an ARB, as well as a decrease in fibrosis in mice that received the ARB. And lastly, addressing sleep is very important in our patients as in meta analysis, not only was obstructive sleep apnea associated with histologic NAFLD, but also associated with hepatic fibrosis of any stage. So the current treatment of patients with NAFLD and NASH is weight loss, whether it be considered by lifestyle modification, anti-obesity drugs, or surgical approaches. Consider control of the metabolic syndrome with treatment of each and every feature of metabolic syndrome, followed by liver-directed pharmacotherapy for patients with NASH and at least hepatic stage 2 fibrosis, not knowing the long-term risk of use of both vitamin E or pioglitazone. So my take-home points is treating obesity is the foundation for the management of NAFLD and NASH, address all modifiable dietary and metabolic risk factors. Anti-obesity pharmacotherapy may be integral to the treatment of NAFLD and NASH, but requires further investigation. Bariatric surgery improves NASH and hepatic fibrosis, and standard of care management of comorbidities has the potential not only to improve NASH, but also hepatic fibrosis decreases the risk of liver cancer and portal hypertension. Thank you. Mr. Chairman, ladies and gentlemen, here are my disclosures. The basic principles of therapeutic development are first, biological plausibility. You need to treat the cause or the symptom. Linking the endpoint to the mechanism of action in the context of disease biology and establishing that the endpoint is likely to reflect the probability of eventual benefit. Now let's start with thinking about NASH and separating cause and effect. NASH is non-alcoholic, and the steatohepatitis we recognize is a composite of steatosis, inflammation, and ballooning, which are features of disease activity that are typically seen in zone three. Now the effect of having NASH is a fibrogenic response in the liver, typically starting with pericellular fibrosis, which progresses and leads to cirrhosis, which then leads to clinical outcomes. Now my worthy opponent is going to tell you, no matter how you get fibrosis improvement, you will prevent outcomes, and that is all that matters. But the case I would like to make for you is that it really matters how you achieve fibrosis improvement, and treating the root cause is essential. Imagine three different conditions, pneumococcal pneumonia, hepatitis C, metabolic overload, which progresses through activity, scarring the organs in which they exist and destroying their function and leading to death. Would you treat pneumococcal pneumonia with hypoxia? I would love to see a show of hands, but I'm not going to do that. Would you treat in 2020 hepatitis C by saying, forget the virus, we'll just give you an antifibrotic once you get cirrhosis? You would not do that, you treat the cause. And same in NASH, you have to treat the underlying metabolic overload that drives the steatohepatitis, which then eventually translates into fibrosis. So if you're going to treat the underlying root cause of the disease, and that targets the disease activity, then the first thing you expect to see is an improvement in activity, which should be followed by an improvement in fibrosis. So I'm not saying that fibrosis is irrelevant. It's an extremely important part of the story, but depending on what you are studying, its mechanism of action and the time course in which you are making that assessment, if you're treating the activity and by treating activity, you hope to reduce fibrosis, then the first thing you need to show is that your activity is going down. So this is the basic paradigm, which even is accepted by regulators, not only in the US, but in Europe as well, so that in the short term, we have to improve activity with or without fibrosis. Whereas in the long term, of course, we have to demonstrate reduction in progression to cirrhosis and clinical outcomes. So then the next point is, if you reduce activity, can you be sure that fibrosis will improve? So here are data from 454 pairs of biopsies from the NASH CRN. At the top, I show you a set of bars from people whose fatty liver disease or NASH diagnosis improved from the first to the last biopsy. The first biopsy is shown on the left, the last biopsy is shown on the right. And you can see over here, those who went from NASH to fatty liver, the proportion of people with fibrosis stage two dramatically increased. Similarly for movement from borderline NASH to fatty liver and from NASH to borderline. Conversely, if you went the other way, that you went from fatty liver to NASH, the proportion of people with low grade fibrosis decreased, the proportion of people with high grade fibrosis increased. And even when you look at the NAFLD activity score and plot the change in activity score versus change in fibrosis, what you see is that when activity decreases, fibrosis tends to decrease, activity increases, fibrosis tends to increase. And so this was in the setting of no clinical trial. This is in routine follow-up. Even in the setting of clinical trials, what we have shown, and in fact, my worthy opponent is a co-author on these, that the improved NAS or resolution of steatohepatitis is associated with a greater probability of fibrosis improvement. And you see the odds ratios over here, and they're both statistically significant. Lastly, if you treat NASH and you only notice fibrosis improvement, I would suggest you have not met the burden of proof needed. The ideal situation is when you improve fibrosis in the setting of a clinical trial, and you also reduce NASH activity, either by demonstrating resolution of steatohepatitis or with an activity score. And I disagree with the regulators that the activity scores are irrelevant. This has high biological plausibility because by decreasing activity, you're reducing fibrosis. On the other extreme, if you're treating NASH, not a direct antifibrotic, you're treating NASH and you notice fibrosis improves, but you have no effect on NASH. You have uncertainty over the impact of unrestrained upstream drivers. It may position your drug for combination therapy, but you have not met the burden of proof because your drug is supposed to act on the NASH. You didn't demonstrate that. And even with an antifibrotic, you have to consider what happens when all the upstream drivers of fibrosis are still in play. And of course, if you have some effect in the middle, it depends which effect predominates. Lastly, a thought about drug development reality. Fibrosis improvement takes time and is linked to mechanism of action. Now, there are trials which have shown fibrosis improves within a few months. Good for them. But I would say in NASH, as in diabetes and other features of metabolic syndrome, outcomes occur over a period of years. So this is a marathon. It is not a sprint. You really haven't impressed me that much by showing that you can get to a point of improvement in three months, then you have to show me that it is sustained. So if you look at bariatric surgery, one of the most effective ways of reducing the NASH activity and fibrosis, that takes years to actually demonstrate. And so it is very important to also consider the time that you have to be able to finish your trial and to be able to show that you have some plausibility of benefit. And that is linked to your mechanism of action. So to summarize, NASH causes a fibrogenic remodeling of the liver and leads to cirrhosis. Most therapeutics treat the root cause, i.e. NASH. Now, if you treat NASH, you should expect to see NASH, the cause of the fibrosis, to improve first, followed by fibrosis stabilization and regression. And it does. I've already shown you the data for that. Fibrosis regression may take a long time and depends on the mechanism of action. Furthermore, fibrosis improvement without biological plausibility increases the risk of false discovery. So for practical reasons, it is easier and relevant to show increased NASH resolution in the short term, especially for agents with metabolic targets. I will finish by pointing out that while everybody jumps on the fibrosis bandwagon and fibrosis is important, it is not just about the fibrosis. It's about the basic disease. And as Stephen Hawking said, the greatest enemy of knowledge is not ignorance. It is the illusion of knowledge. I would like to thank Dr. Harrison and Dr. Rothman for giving me this opportunity. The next 10 minutes, I would support the motion. We should target fibrosis improvement, not NASH resolution. These are my disclosures, and this is going to be the outline of my talk. And in the end, the public will decide which motion would you support. As we know, the liver fibrosis is the most important predictor of liver-related mortality in non-alcoholic fatty liver disease. Based upon this pooled analysis, we showed that it is patients with NASH who have stage two fibrosis or higher who have increased risk of liver-related mortality. And this is the reason why pharmacologic therapies for approval by regulatory authorities, you require to have NASH with stage two fibrosis or higher. We wrote an editorial, Dr. Naga Chalasani and I, a few years ago, based upon data from Paul Angulo's seminal paper, looking at both fibrosis and also NASH and linking it with outcomes. Based upon our entire assessment of the literature until then, we looked at what are prognostic factors on histology. And number one prognostic factor was fibrosis, followed by portal inflammation. And NASH was buried as the number three, third cause that was linked to a prognosis in patients with NAFLD. And ballooning was below that. This is just again, looking into pharmacologic therapies for NASH-related fibrosis. So you have to have NASH for stage two fibrosis or higher. Can have patients with advanced fibrosis or NAFLD-related cirrhosis or NAFLD cirrhosis with decompensation. So clearly we're not just looking for NASH, we're looking for significant fibrosis due to NASH. And then we include these patients into pharmacologic therapies because NASH is not causing mortality, especially early on in the disease. It's really once you start developing significant scarring in the liver that is reflected by stage two fibrosis or higher, that's when we think you need treatment. These are the end points for approval for new drugs for treatment of NASH. There's initially a subpart H conditional approval and the end points are one stage improvement in fibrosis without worsening of NASH or resolution of NASH without worsening of fibrosis. And then for full approval, we need progression to cirrhosis which is also progression of fibrosis to cirrhosis and clinical decompensation in patients who have cirrhosis in terms of presence of clinical ascites, spontaneous bacterial periodontitis, hepatic encephalopathy, genetic treatment or variceal bleeding or mild reaching 15 or higher. Now, what are the caveats regarding NASH resolution? How do we define NASH resolution? You need to have lobular inflammation go down to zero to one and ballooning go down to zero. But at baseline, you may have lobular inflammation ranging from one, two or three. And for ballooning, baseline may be one or two. So what you're looking at, the Kappas for lobular inflammation, especially inter-rater really low, it's like 0.45 and Kappa for ballooning inter-rater is about 0.55. So the concordance rate between these two histologic variables is very, very low. And for NASH resolution, you need to have not just one, but both of these variables combined. Therefore, the Kappa is even lower than that. So when you have an end point that even the expert pathologists can't really decide or agree upon, there is a real problem. Well, with fibrosis, it is less of a problem. So how do we define fibrosis improvement? It's much more simpler. One-stage improvement in NASH-CRN histologic scoring system and Kappa is about 0.84 based upon previous studies. There's a recent paper just published this year from the NASH-CRN where we see various Kappa statistics for different histologic features in NASH. And this is updated. So of course, steatosis, very high Kappa, but you can see for lobular inflammation is again at 0.46. Steatohepatitis diagnosis in NASH is 0.66. But what is the highest Kappa? It's for fibrosis, it's 0.75. So you can see even in the hierarchy of the end points, we're looking at a baseline assessment, then fibrosis assessment is much more credible by expert pathologists as compared to diagnosis of NASH. Now, what are various phase three end points? So if you drill down a little bit deeper, what you start realizing is there are more facets of NASH resolution that are seen as primary end point in phase three trial. So certain trials use NASH resolution without worsening of fibrosis, where ballooning of zero or lobular inflammation zero to one. Then for certain other trials, they've added additional criteria to have at least NAFLD improvement and NASH improvement of two points or higher and ballooning of zero and lobular inflammation zero to one. Why are these things being added? This shows that the end points are not credible. So people are trying to improve upon them by having some treatment effect delta favoring the drug over the placebo. What about fibrosis improvement without worsening of NASH? Well, it's standard. One stage or higher improvement in fibrosis in a NASH-TRN histologic scoring system. But that's consistent. Why is that? Well, because I think it is much more reliable, precise and accurate end point to assess compared to other histologic features that are required for NASH resolution. Can we show improvement in fibrosis? Has any trial shown that? Well, Flint trial for ovidicolic acid versus placebo initially showed that. That was the first phase two B trial to show one stage improvement in fibrosis over 72 weeks. And again, regenerate trial, the phase three program confirmed those data when they compared patients receiving ovidicolic acid 25 milligram versus 10 milligram versus placebo at 72 weeks. This trial is ongoing for long-term outcomes. And this is what we found, 23% of patients had one stage improvement with the 25 milligram dose relative to placebo showing 12%. So significant difference. What about individual features of NAFLD activity score? Of course, there's improvement of ballooning that was consistently higher in the 25 milligram versus placebo. Same thing with larval inflammation, but not with steatosis. Now you can imagine that larval inflammation by itself was significant, ballooning itself was significant, but the NASH resolution was not. Why is that? It's the Kappa issue here. If you look at state, more than one stage improvement or two stage or higher improvements, then you are seeing that there's a significant difference again with the 25 milligram relative to the placebo group. What about Elifibrinol versus placebo? Recent data that was in a press release showing that this was not significantly different for NASH resolution, which was the primary endpoint of this trial relative to placebo. So in my opinion, fibrosis assessment is much more reliable and precise than NASH resolution. So what are the trade-offs? Well, if you have a NASH resolution endpoint, that the treatment effect delta is less precise, the estimate is less precise. So you'll require a larger sample size. Fibrosis endpoint, treatment effect delta have more precise estimate because the measurement by the pathologist is more reliable with a higher Kappa. So relatively modest sample size could be new. Now the primary endpoint in a NASH trial should be decided based upon the mechanism of action of agent, not as to which endpoint is better just because I'm seeing it. Understanding these concepts that I just elucidated helps us make an informed decision on the chosen endpoint based upon the mechanism of action of our drug. In summary, fibrosis assessment by a pathologist is more reliable, precise and accurate and has higher Kappa statistic than diagnosis of NASH resolution. One-stage improvement in NASH CRN fibrosis state without worsening of NASH is standardized. NASH resolution definitions are ever changing and are less reliable. Ultimate choice of endpoint is dependent upon the mechanism of action of a drug rather than credibility of how that endpoint is assessed and the precision around those estimates. Future is bright for new drugs for treatment of NASH-related fibrosis. And I look forward to your opinion on this debate. Thank you. Hello, I'm Steven Harrison and thanks for joining us today. I was tasked with giving a 20-minute talk on what does the future hold relative to fatty liver disease beyond 2020? There's a lot of ways we could have gone with this. For the sake of time, I really wanted to focus on biomarker development and treatment. So here are my disclosures. So really the question we have in front of us is lots of fatty liver out there. How do we identify the at-risk NASH patient? That is data shows us convincingly now in natural history studies that NASH with fibrosis portends a worse prognosis, particularly at F2 and greater. So really that's the target of not only identifying these people in clinical practice, but also for drug development. So just very pragmatically, we can look at clinical data, data we can get every day in our practice and determine high risk features that put people at increased risk for NASH with fibrosis. And I highlight that here in this pyramid, many of you have seen before, just focusing on demographics like age, diabetes, obesity, and hypertension, turning to some very basic stuff like AST, lab values, AST greater than 40, putting that into NAFLD fibrosis score, FIB4, or looking at something as simple as an AST, ALT ratio, being highly predictive of a patient with more advanced disease. And then finally, FibroScan, we'll talk a little bit more about that, but targeting a number of around eight to eight and a half. And the more of these variables you put together, the higher the likelihood that the patient will be in that high risk group. Now, I wanna talk about liver biopsy just briefly, because we're transitioning through this phase now, where right now liver biopsy is still very, very important, but ideally we'd like to transition into non-invasive assessments. However, liver biopsy is still required to confirm the diagnosis and exclude alternative and secondary pathology, to appropriately stage the disease and to stratify risk progression. And it can also be used where there's diagnostic doubt, you're worried about coexistent liver disease, and maybe sometimes the patient wants that. Often we do it in clinical practice when our non-invasive testing strategies are indeterminate, they don't fall into a low risk or a high risk category. So how do we best select patients for liver biopsy? Now, for clinical trial purposes, it's still the endpoint that's required for accelerated approval, that surrogate endpoint. We'll talk more about that later on. So we need to do this biopsy to have them eligible for late stage clinical trials, not necessarily early phase two trials, but the later phase two Bs and the phase threes. And so the question comes up, if we're going to do liver biopsy, how can we improve liver biopsy interpretation? And I think that's a big part of moving beyond 2020, and that's optimization of histopathology processing, reading and interpretation. So it's not just about reading the biopsy, but it's about making sure that you get an adequate sample, that that sample is processed properly, stained properly, so that it can be read with the most optimal settings. And then we need a standardization for central reading processes. And we're in the middle of working through that now, taking more than one pathologist, looking at what their interpretation is, maybe adjudicating the differences. There's lots of different ways to do this, but that's something that is an active point of conversation, not only with the agency, but with sponsors and pharma companies as well. And then moving into, we'll talk a little bit more about that today also, automated reading through artificial intelligence methods. And I think we're gonna go through that period of artificial intelligence, where we can really come alongside the pathologist and provide additional helpful information to that pathologist to adjudicate the biopsy. And we'll hopefully base our non-invasive testing strategy off of that method as well. And then finally, into the non-invasive testing arena, really, we wanna develop non-invasive tests to identify that high-risk patient, to ultimately replace the biopsy, and really three different contexts of use there. We wanna stage the disease in the diagnostic paradigm. We wanna see how patients do on therapy. In other words, monitoring therapeutic response. And then finally, can we predict long-term outcomes? Now, just talking a little bit more about liver biopsy. Recently, we published data from the Eminence trial, a very large phase IIb paired liver biopsy study, looking at when we took those biopsy slides and had them looked at by multiple pathologists, what did we find? Well, unfortunately, we found some significant variability in the interpretation of those biopsy slides. And I think that's part of what's kind of spearheaded this closer look at how can we improve the liver biopsy interpretation and results that we're currently utilizing for clinical practice, as well as for clinical trial endpoints. And I want to look back a little bit. So this isn't unique to the imminence trial. Even way back in 2005, the NASH CRN pathologists got together and calculated their Kappa statistics on a small number, 32 cases, and showed that they were really good at fibrosis and at steatosis in harmonizing their response. Not so good for inflammation and ballooning. Fast forward about 14 years, same group, more cases, still about the same for steatosis and fibrosis. Unfortunately, not a lot of learning here has happened for the gestalt NASH diagnosis, ballooning, and lobular inflammation. Suggesting that even pathologists who sit around the same microscope for years and years and years and get to know each other, they still struggle with some of the finer points of understanding exactly what a balloon hepatocyte looks like. Is everybody in agreement on that? Same thing for lobular inflammation. So I do think it really behooves us to see if we can find collective ways to work together to improve our interpretation. And that gets me to artificial intelligence and machine learning. And I just want to highlight two areas where this research is going on. There are actually quite a few companies working in this arena. Here's two. This is Path AI. And some of the work they're doing, I just want to show you a theoretical construct. This is like a NASH comprehensive panel, what you would get from their analysis. So you get the NAFLD activity score, you get the CRN fibrosis score, but then if you look down below that steatosis, lobular inflammation, and hepatocyte ballooning score, you now begin to be able to fully quantify the degree of steatosis, the degree of inflammation, the degree of ballooning. And you see that highlighted in the different colors on the right. And we see in this slide that when we start to look at steatosis, ballooning, and lobular inflammation, you can see that they're color-coded here. But there's actually a nice correlation with pathology consensus grading as well for all of these parameters. And for fibrosis, we see the same thing. On the left hand, you see these box plots looking at F0 to F6, and the model fibrosis score on the y-axis. And on the right, you see pictures. There's a central pathologist interpretation showing F6. And then further on the right, the machine learning interpretation of fibrosis. You see this fully quantitative, very nice breakdown of the different degrees of collagen within that same liver biopsy slide. Now, moving onward to another artificial intelligence method, this is HistoIndex. And basically, they have a proprietary AI-powered algorithm using second harmonic generation. Basically, a laser that looks at the different components of the liver biopsy slide and can pull out collagen features, steatosis, inflammation, etc. You see that here on the right. Here's an example of that being used in clinical practice. So this is data from the Madrigal Phase 2 trial, where you see a nice correlation between the artificial intelligence score and the pathologist score. And way over on the right, when looking at those patients that had greater than or equal to one point reduction in fibrosis on liver biopsy, you were able to see some pretty significant differences, 47% of the F3 patients versus 0% for placebo. And then if you wrap up all the patients, it was 32% versus 12%, which allowed that to reach statistical significance. So really, it was a way to augment the pathology scoring, but also to help really show what was happening with this drug in the Phase 2b trial. And that might help power the next phase study or give you some guidance on, really, is your drug having an impact on fibrosis? So now let's just move through to non-invasive tests. So really, the proper use of non-invasive tests is the addition, the synergy of demographics with laboratory data, fiber scan, and imaging data. That allows us to enrich for NASH in advanced fibrosis, as well as to evaluate for therapeutic response. Now, historically, we've used NAFLD fibrosis score in FIB4. They've been around a long, long time. There's online calculators for them, and they give us two different cut points, a low probability and a high probability cut point that correlate to a high sensitivity and a high specificity. The problem with this is that there's a high indeterminate range in the middle, about a third of patients, depending on which study you look at and what population is evaluated. So ultimately, that's not ideal. The other thing is, while they're optimized for their negative predictive value, in clinical practice, really, they lack sufficient positive predictive value to be used alone to predict NASH in fibrosis. So there are others that are being utilized and in development. We see NIS4, Fibrospec, ADAPT, Pro-C3, FAST, which is FibroScan and AST, and then MR elastography are just a couple of the examples. This is not a complete list or an exhaustive list by any means, but there are tests being developed, both wet and imaging-based biomarkers. I want to highlight a couple of these. The first is NIS4. This has recently come to light, been published in Lancet Gastroenterology and Hepatology, and this is a four biomarker panel blood test that was designed to identify NAFLD patients that had NASH with an NAS of 4 or more and fibrosis F2 or greater, and you see the components A2M, YKL40, HbA1c, and then miR34a, and when you look at the AROC on the right, you see it performed very nicely relative to most of the other variables that we commonly look at to include Fib4, NAFLD, fibrosis, score, BARD, FibroScan, ELF as well. Now ADAPT Pro-C3 is another marker out there. It looks at age, diabetes, Pro-C3, which is the marker of fibrogenesis, platelet count, and what this study showed, published in 2018, was that it accurately identifies advanced fibrosis with a very nice AROC. In fact, Pro-C3 was independently associated with advanced fibrosis, which is encouraging since a lot of our clinical trials are using Pro-C3 as one of the biomarkers. Now I also want to highlight FAST. This is FibroScan plus AST. There's actually an app that you download on your phone called the MyFibroScan app. It asks you to impute the CAP score, the KPA, and the AST value, and it gives you, again, a range, a high cutoff and a low cutoff, that can be utilized to detect the NASH patient with an NAS of 4 or more with F2 or greater fibrosis. And this is data from the Lancet-GastroHEP paper showing, in blue, the FAST score relative to fib 4 in the NAFL fibrosis score for that patient population, that NASH with an NAS of 4 or more with F2 or greater fibrosis. And you see it performs very nicely, not only in the derivation cohort, but in other cohorts done around the world. And then there's MR elastography, and this is courtesy of Rohit Lumba, a good friend of mine who loaned me this slide. But you can see the AROC here showing a very high value of 0.92 for detecting severe advanced fibrosis to cirrhosis. You see it greater than 3.63 KPA. There's a very nice sensitivity and specificity, and this is something that we're really beginning to see more and more of. So, again, individual tests have varying degrees of sensitivity and specificity. Most have an indeterminate range in the middle. It turns out when you begin to combine multiple tests in sequence that we can narrow down that indeterminate band, and it gives us a much better sensitivity and specificity. So here's an example. This is published by Quint Nancy's group in Hepatology in 2019, and it was evaluation of about 3,000 samples from the Stellar 3 and 4 trials, which were trials looking at F3 and F4 fibrosis. Now, while the drug didn't work, we get some amazing data on some of these biomarkers. It turns out that single test, whether you use NAFL fibrosis score, FIB4L for FibroScan, led to about 50% of the results being indeterminate. But when you sequence them together, you're able to cut that by about half. Now, we talked about the context of use of diagnosis of NASH with advanced disease. What about those markers that can measure therapeutic response? And we're starting to see some of these pop up in ALT, MRI, corrected T1, ELF, and Pro-C3, and highlight here just some data that is very preliminary, but I think this is where we're headed, and this looks at a correlation with histologic response over 12 weeks of treatment, and you see that Pro-C3, ELF, and CT1 all were able to show nice changes that correlated with histology. So I think there's more to come here. This is just really the tip of the iceberg. Now, in clinical trials, we can look at various non-invasive endpoints, like ALT and proton density fat fraction, and we now have data that correlates the magnitude of effect change with actual histology. So for ALT, it's a 17 unit per liter drop, and for PDFF, it's a 30% relative drop. And so a lot of these trials are now including this data in their analysis, and I want to show you some data from resminarone, some trials being done there, some data being generated there, that really look at the association of PDFF response with NASH resolution and fibrosis reduction. And what you can see is, in the orange, this is at least a 30% relative reduction of liver fat. Now, when you look at the response rate on the y-axis, you can see that it ranges around 37% or so for NASH resolution and fibrosis reduction. If you look at both NASH resolution and fibrosis reduction, the percent response, if you achieve that 30% relative reduction, is very nice. It's very high. So, yeah, I want to just bring this to your attention, that greater than 30% in this so-called super responder group, the greater than 50% fat reduction, we're associated with about a 60% achievement of both endpoints, NASH resolution and greater than a one point fibrosis reduction. And turns out that's not just unique to that mechanism of action, it's a thyroid hormone receptor beta drug. But if we look at the FGF-19 analog aldefermin, we see something very similar. So, on the left hand, in the blue bar, you see those patients that had fibrosis improvement of at least one stage with no worsening of NASH at 24 weeks of treatment, and it was 30% for the aldefermin one milligram group versus zero for placebo. But if you take it one step further, and you say, what about those F3 patients that had a 30% relative fat reduction, you actually see about 46% of the aldefermin group had that achieved that mark of fibrosis improvement by at least one stage. So, pointing out that that drop in PDFF may actually be helpful in predicting the response, which could be a non-invasive surrogate to show us that patients are improving. More to come there. Now, there are two major consortiums I would be remiss if I didn't highlight. One is Litmus. This is a really a consortium of European academic centers and pharma that have come together to develop and robustly validate and advance toward regulatory qualification markers that diagnose risk stratify or monitor NAFLD or NASH progression for use in drug discovery. So, it's really a global effort here to validate these NAFLD NASH biomarkers. And I want to highlight the fact that it's not just the European group, Litmus, but there's also a U.S. group called Nimble. And together, you see here the Venn diagram, they're going to achieve all of these. The goal is to achieve all these, robustly validate these biomarkers, really enhance the qualification pathway to the FDA and EMA. And you can see that there's just lots of data being generated here, prospectively, that I think will be very helpful in helping us understand, really within the context of use of these biomarkers, how they can be utilized. So, I want to shift gears in the few minutes I have left and talk to you about therapeutics. So, I highlight here just a pathogenesis slide. And I want to highlight the parts of this where we really have activity happening. It's really in the metabolic space right now. So, if you have a drug that impacts metabolically to reduce liver fat, that's where we're seeing significant strides being made. It doesn't mean we're not having other drugs developed that are pure anti-inflammatory plays, that target gene expression, or gene silencing of overactive genes like HSD-17, beta-13, or stellate cells, for instance, as an antifibrotic. So, lots and lots of activity going on. Just want to highlight again the requirement for subpart H conditional approval. It's a liver biopsy. You can have NASH resolution without worsening of fibrosis or fibrosis improvement with no worsening of NASH. Full approval, you got to go all the way to progression to cirrhosis or a clinical endpoint that we have listed here. Now, I summarize all this phase 3 data for you right here. Sinocrivorac, elefibrinor, beta-cholic acid, solancertib, resmeterome, and aramcol. Just real quickly, where they're at. Sinocrivorac is, we're waiting to hear what their initial phase 3 trial results are. We should hear soon on their subpart H approvable endpoint. Elefibrinor, we heard earlier this year, did not make it. So, that drug is no longer being developed. OCA, we know they've reported positive results on fibrosis, but they are working through a complete response letter from the FDA. So, it's not approved yet. Solancertib didn't work in either the STELR 3 or STELR 4 trials. So, that's not moving forward as monotherapy. Resmeterome is marching forward also right now. They are still enrolling their phase 3, actually two different phase 3 trials. So, the thyroid hormone receptor beta class is still moving forward. And then, aramcol, which is moving forward as well in phase 3. So, right now on this list, we still have sinocrivorac. We still have a beta-cholic acid, although they're working through the the CRL. And then, we have resmeterome and aramcol. Now, just turning real quickly to the NASH resolution landscape for monotherapy. You see here where we are, right? So, if you look in placebo, that's the light gray. And then, you have the drug classes in the other bars. You can see response ranges of up to 67%. We have some really good news from the GLP1 semaglutide. We also have good news from FGF21 acaroifruxafirmin. And then, also Inventiva, which is a pan-PPAR agonist. You see almost 49% for the high dose and resolution of NASH. And then, others that behave relatively well when compared to placebo, aldefermin, the FGF19, madrigal, which is resmeterome, and the thyroid hormone receptor beta agonist. And even Cella delpar with their PPAR delta. Now, unfortunately, that was halted prematurely. Looks like they did get clearance to restart that based on the FDA giving them a green light to move forward. They're focusing more on PBC at the current time, but hopefully, we'll jump back in to NASH at some point. And what about fibrosis improvement? Again, you see the spectrum here of where we stand. Intercept way over on the right, you see the results there, 23 versus 12%. We see some interesting data. I would say what strikes me is that the placebo response rate is widely varied from 12 to 29%. The hope there, dialing back in to the variability we see on histopathology, bringing in that artificial intelligence, helping us to refine what we're really seeing, harmonizing our reads between pathologists, really making sure that there is a uniform way to do the biopsy, to get the right core, the diameter, to stain it appropriately, to cut it appropriately, all that we're learning about how to do this in a systematic way so that hopefully, we can have a placebo response rate that's more uniform. Of course, that can also be affected by the patient's lifestyle in the middle of the trial. So another thing we're trying to work on is really harmonizing across studies a uniform lifestyle management plan. So would be remiss to not talk briefly about combination trial therapy. There's lots of them underway. I highlight the ones that are ongoing right now. You see them here, the various agents. So there's a proof-of-concept trial looking at semaglutide, frosocustatin, and encelafexor that's underway now. There's also one with an FXR agonist and senacrevirac, and you see the rest. So lots and lots of data being generated on combination therapy, and I just want to say moving forward, I think the considerations for combination therapy really need to hone around safety. Got to make sure that safety is always put first when we do this, and then looking for synergistic mechanisms of action that affect complementary pathways. This is a multi modality disease, multi pathway disease, and so we need to target those different pathways. The higher up in the in the chain that we hit, I think the greater the probability that we hit multiple downstream pathways. To me, what interests me, the combinations that I think were intriguing are combining a thyroid hormone receptor beta agonist with either an FXR agonist and insulin sensitizer. I'll even throw a PPAR Delta in there as well. Looking at metabolic drugs plus anti-inflammatory, I think if it's the right anti-inflammatory that shows some data on its own as being effective, I think putting that with a metabolic drug would be attractive. And then finally, metabolic plus anti-fibrotic, I think. It doesn't mean that metabolic drugs can't have an impact on fibrosis. We're seeing that already, but I think if you had an oral anti-fibrotic agent that you could combine with the metabolic drug, you would just get there that much faster. So with that, I want to thank you for taking the time to listen today, and I hope you enjoyed this symposium. Take care. I'm Yaron Rotman from the National Institutes of Health. On behalf of my co-organizer, Dr. Stephen Harrison, and the members of the steering committee of the SIG, I want to thank the presenters for working, preparing, and recording the fantastic presentations that we heard this morning. Here are my disclosures. So it's 2020, and we have tons of information about NAFLD and NESH. Last time I checked at PubMed, there were more than 27,000 papers. Going from the seminal paper in 1980 by Ludwig to the most recently published one, we are publishing at an exploding rate. We're expected to have more than 4,000 papers in this year alone about the topic. And many of you in the audience are actually writing those papers and contributing to that knowledge. In fact, the rate of increase in publication is outpacing the increase in obesity in the United States, so we may be doing something right. However, we still do not have an approved medication to treat the disorder. How have we advanced? What have we learned today? So Dr. Namadzin Nooruddin spoke to us about who needs treatment and how to identify them, especially using non-invasive markers. Dr. Manal Abdel-Malek spoke about what are our options to treat those patients, either things that are available or things that we're studying now. Drs. Lumba and Sanyal discussed a very important, debated a very important question of what are we aiming for? What should be the endpoint of treatment? This is a question that's relevant for both clinical trials as well as for clinical practice. And finally, my co-organizer, Dr. Stephen Harrison, spoke about where we're heading next and hopefully going for a brighter future. Before we say goodbye, I would like to encourage anyone here who is not a member of the SIG to join us. Join the SIG, participate in its activities. For example, we had a fantastic webinar this summer about dietary management of NAFLD. It was led by Dr. Nooruddin and, as far as I know, broke the ASLD attendance records. It can still be watched online and please be on the lookout for future educational activities. With that, I'd like to thank the presenters and all the participants and I hope you all have a fantastic virtual liver meeting. Thank you.

clinical management

NAFLD

NASH

lifestyle modifications

medications

surgery

disease activity

fibrosis

NASH resolution

non-invasive tests

liver biopsy

therapeutics

treatment strategies