Hello, I'm pleased to present 2020 NAFLD debrief for the ASLD meeting. I want to thank the ASLD for inviting me to give this NAFLD debrief. This is my biography. These are my disclosures and I want to thank many presenters who graciously made their presentations available to me for inclusion in this presentation. As has been the case past several years, there's an explosion of really good signs presented at this meeting. So it's going to be difficult to cover all of it. So the outline of my presentation would be treatment trials, which will be bulk of my presentation, followed by selected basic translational abstracts, natural history abstracts, and finally we'll end with some abstracts on non-invasive assessments. Many important treatment trials have been presented at this meeting. There are six with histological endpoints, we'll go into some detail. There are four proof of concept, early phase studies that also we'll touch on. Let's just first start with Professor Sven Franky's presentation on pan-PPAR agonist Lena Fibrenor in patients with non-sorotic NASH. Lena Fibrenor is a pan-PPAR agonist with moderate unbalanced activity against PPAR alpha, delta, and gamma. As you know, there's been interest with PPAR agonism as treatment target for NAFLD NASH since mid-90s. This was a 24-week trial where patients were randomized to placebo, 800 milligrams of Lena Fibrenor or 1,200 milligrams of Lena Fibrenor with a biopsy at baseline and biopsy at the end of treatment. And this was randomization was 1 to 1 to 1. And patients were stratified by the presence of type 2 diabetes. Main inclusion criteria was biopsy proven NASH by central reader based on SAF scoring system. There were 247 patients in the full analysis set, whereas 194 patients in per protocol. Secondary endpoint was decreased in inflammation and ballooning at week 24 as compared to baseline based on SAF activity scoring system with no worsening on fibrosis. And there were a number of traditional secondary endpoints, both histologic as well as liver biochemistry based as well. And there were exploratory endpoints. 81 patients in placebo, 83 patients in 800 milligrams of Lena Fibrenor, 83 in 1,200 milligram per day dose. Excellent completion, although this just was a 24-week trial, more than 90% of patients completing 24-week trial. Primary endpoint was met with Lena Fibrenor 1,200 milligram dose. Similar improvement in SAF activity score by two points or higher was achieved in nearly 50% of 1,200 milligram dose compared to 27% in placebo group. Similar efficacy was shown in part protocol population in diabetics as well as high risk NASH group, which was NASH patients with stage 2 or stage 3 fibrosis. When you look at traditional histological endpoints such as resolution of NASH with no worsening of fibrosis or improvement in fibrosis with no worsening of NASH or improvement in NASH and improvement in fibrosis were consistently significantly higher in Lena Fibrenor 1,200 milligram dose compared to placebo. And these results were consistent across per protocol population as well as diabetic and non-diabetic population as well. There were favorable effects on lipid profiles, both HDL cholesterol and triglyceride and glycemic control as one would expect with PPAR gamma agonism. Results concluded this was safe and well tolerated, but I draw your attention to higher treatment emergent SAEs in 1,200 milligram dose as well as there is the signal with increased weight, 2.7 kg with 1,200 milligram dose. Peripheral edema also increases in transaminases, perhaps related to PPAR alpha agonism. Online Lena Fibrenor in a 24-week trial showed exciting histological efficacy, but we just have to see how these safety liabilities will play out in phase three and long-term outcome studies. Next is an exciting study presented by Professor Newsome. This work is now, I think, believe published as a full-length paper in New England Journal of Medicine last week. This is subcutaneous semaglutide once daily versus placebo in NASH. GLP1 receptor agonists have multifactorial effects beyond glycemic control and make great sense as therapeutic agents in patients with NASH. This trial had three treatment groups, semaglutide 0.1 milligram, semaglutide 0.2 milligrams and semaglutide 0.4 milligrams. These are up titrated. These were placebo controlled and the population is typical with biopsy-confirmed NASH with fibrosis and NAS had to be four or higher. Primary endpoint is resolution of steator hepatitis with no worsening in fibrosis in high-risk NASH defined as NASH with stage two or stage three fibrosis. And the key secondary endpoint was improvement in fibrosis in the same population with no worsening in steator hepatitis. Patient population is shown here. There were a total of 320 patients were randomized. And interestingly though, only 83% completed the trial, but note this was a 72-week study and the completion rate was not any worse in semaglutide compared to placebo. Even placebo had a sizable withdrawal because mind you, but this is a once a day injection therapy. Resolution of steator hepatitis with no worsening of liver fibrosis was significantly higher in all three study groups, active drug compared to placebo, but this was impressive. It's semaglutide 0.4 milligrams and high-risk intended treatment population with stage two or stage three fibrosis, nearly 60% had resolution of steator hepatitis with no worsening of fibrosis compared to only 17.2% in the placebo. However though, there were no statistically significant differences in fibrosis improvement and this could well be due to very high placebo response rate. As you see, 33% had improvement in fibrosis with placebo. This is some of the highest we have seen placebo response for fibrosis in NASH clinical trials. As one would expect with semaglutide, there was a dose-related weight loss with 12.5% in a 0.4 milligram dose as well as improvement in A1C and with an improvement almost by 1.2% and favorable changes in lipids. And the AE profile seems reasonable and nothing unexpected here. And in terms of other safety outcomes of interest, there was some clustering of gallbladder-related disorders and hypoglycemia in semaglutide-treated patients, but overall seemed like this was well-tolerated with impressive efficacy on necroinflammation, lack of significant histological benefit for fibrosis is a concern. We just have to see how this will play out in larger phase III trials and whether these benefits would translate into clinical outcome benefit. We just have to wait and see. Dr. Sanyal has presented Part C, Week 48 results from a FLITE-FXR trial with a propifexor. Propifexor is a non-bile acid agonist of FXR. This trial had three Parts A and B were proof-of-concept, 12-week studies, lower doses, whereas Part C is a 48-week study. A primary endpoint was MRI-PDFF at Week 12, which was met with doses propifexor 140 micrograms and 200 microgram doses. And here in this, though, Dr. Sanyal presented the 40-week, 48-week data from Part C. The study population was typical, and this was the high-risk group consisting of NASH and Stage 2 and Stage 3 fibrosis. Approximately 50 patients were randomized to placebo, 140 micrograms of TXR, 200 micrograms of TXR once daily. And as I said, we are showing the results of Week 48 data. Study discontinuations were significantly higher, only 75% in propifexor groups completed the trial, which is probably due to some of the AEs and other liabilities. Body weight was significantly improved with the propifexor in a dose-related fashion compared to placebo. MRI-PDFF changes were reproduced at Week 48 as were presented at Week 12, and both in terms of absolute reduction as well as proportion of patients who had a 30% reduction or higher. However, though, histologically, there was no significant benefit with propifexor, either with NASH resolution or improvement in fibrosis, which was disappointing. Overall safety, you know, there were higher dose reductions and discontinuations with propifexor of 200 microgram dose as well as the pruritus as well as lipid signal that one would expect from FXR. Bottom line, propifexor isn't going anywhere due to lack of histological benefits as well as some mechanism-based, some of these safety liabilities. This is an interesting work presented by Dr. Harrison with Aldofermin, which is NGM282, which is a FGF19 analog. And there are many reasons why FGF19 analog is expected to improve inflammation, steatosis, ballooning, and fibrosis. And in this particular trial, 25 patients received placebo and 53 patients received Aldofermin, one milligram subcutaneous every day for 24 weeks. Multiple MRIs were performed, liver biopsy at baseline and week 24, and key inclusion were typical of NASH trials, NAS more than four, biopsy confirmed NASH, but liver fibrosis stages two or three. Primary endpoint was liver fat, number of important histological and other secondary endpoints. Part of and unique aspect about this trial is this compound or this mechanism is expected to increase LDL cholesterol, therefore, if there was a 10 milligram increase in LDL cholesterol at week two, the study required a 20 milligrams of roswastatin proactively. There was impressive effect on liver fat, both in terms of absolute reduction as well as relative reduction, more than 65% of patients had either 5% or greater drop in absolute fat content or more than 30% relative reduction in liver fat, which was impressive. And numerically higher number of patients who received Aldofermin had either fibrosis improvement by a stage with no worsening of STRA hepatitis or resolution of NASH with no worsening of fibrosis. However, I believe these numbers were not statistically significantly different. In terms of patients achieving both fibrosis improvement as well as resolution of NASH at week 24, 22% of patients with aldofermin group had this endpoint, whereas 0% in the placebo. Mind you, this is a very difficult endpoint to achieve. Within 24 weeks, 22% achieving this endpoint, I must say, is impressive and relatively well tolerated and no safety signals were identified. I must admit, though, is that more than 96% of patients on the aldofermin group ended up on rosvastatin by week 24, saying that this therapy isn't necessarily aldofermin alone, but aldofermin plus rosvastatin. I believe this is moving forward into phase three trials. We just have to see how this would translate into registration trials and with clinical outcomes. Dr. Harrison also presented the data on FGF21 fusion protein. This is a furexifirmin. Earlier, Dr. Harrison published a paper with FGF21 showing really impressive histological effects with relatively short treatment duration. This is a 16-week trial. 20 patients were randomized to placebo or three different groups of EFX, 28 milligrams, 50 or 70 milligrams. The study primary endpoint was week 12 MRI-PDFF. Biopsy was performed at week 16 only if patients achieved more than 30% reduction in hepatic fat at week 12. This design led to very few people in the placebo group meeting this imaging endpoint, therefore getting a liver biopsy. What you see here is this FGF21 fusion protein administered once a week, sub-Q, had dramatic effect on hepatic fat by imaging. There seems to be impressive histological benefit, but placebo comparison is very difficult because of the study design. We just have to see how this will play out in a phase three and outcome trials. Dr. Luba presented the results of Fascinate 1, which was proof of concept first-in-class fatty acid synthase inhibitor TVB2640. Fatty acid synthase is very important for de novo lipogenesis, which as you know is hallmark of NASH. In this trial of 12 weeks, fatty acid synthase inhibitor had significant reduction in hepatic fat, also favorable changes in other non-invasive markers such as ALT, lipids, adiponectin, fibrosis markers, and this was safe and well-tolerated. I assume this will move into a phase 2B-like with liver histology endpoints. We just have to see how this imaging results will translate into histological benefits. Dr. Harrison had a poster on potent and selective PPAR delta-cellar-del-PAR, which you know the program had to be stopped because of false alarm for histological lesions leading to study termination. This was a 52-week trial with three different doses of cellar-del-PAR. Unfortunately, due to very high placebo response, there was no signal from imaging standpoint. Similarly, histological response, I believe, were not statistically significant. Bottom line, cellar-del-PAR is not being pursued for NASH, but I believe this is being pursued for PBC indication. A few other notable mentions. A study from Israel tested novel mechanism A3 adenosine receptor agonist. In this proof-of-concept trial at week 12, there was imaging benefit with a CF25 milligram dose. I believe this is moving to a histology-based trial. We will hear the results in two or three years or so. Elifibranol, this is Resolvit. This is PPAR alpha-delta. This was a disappointment phase 3 trial due to lack of histological benefit. This was a 1,000-patient trial. 717 patients in Elifibranol, more than 350 in placebo. Unfortunately, either fibrosis improvement and NASH resolution was not different between Elifibranol or placebo groups. The treatment, this NASH program has been discontinued. Finally, for treatment trials, Dr. Alakoury shared the data on semaglutide with added FXR agonist or ACC inhibitor. My take on this trial proof-of-concept is that adding FXR agonist or ACC inhibitor did not dramatically improve response to semaglutide, but there was the liability in one patient with increasing serum triglyceride by 500 milligrams or so. I believe semaglutide is not going to be pursued with these combinations from what I take with this trial. Moving on to other aspects of NAFLD presented at the liver meeting. As you know, there's been a lot of interest to redefine NAFLD as metabolic associated fatty liver disease. This interesting late-breaker poster from Sri Lanka was based on a community-based prospective study in suburban Sri Lanka initiated in 2017. There were approximately 3,000 patients in 2007 when they applied NAFLD, MAFLD definitions. Almost equal number of patients met both definitions. Interestingly, the fast-forward from 2007 to 2014, both groups, whether you define population as NAFLD or MAFLD, the frequency of incident, metabolic traits or incident, fatal or non-fatal cardiovascular events were identical. Authors concluded other than increase the captive population by very small proportion, i.e., 1.6%, redefining NAFLD as MAFLD did not seem meaningful. So, more to come on this, but I thought this was a very interesting work from Sri Lanka. Many basic and translational studies were presented at the liver meeting, and due to time constraints, I'm not able to go in detail, but I draw your attention to these five abstracts were very interesting and were advanced field significantly. In terms of selected natural history and other clinical presentations, I draw your attention to this oral presentation by Dr. Pasanyal where he presented the results based on 1,000 child-saying natural cirrhotics who participated in Symtuzumab and Celeron-sortive clinical trials sponsored by Gilead. What he has shown is if patients with cirrhosis had fibrosis improve, regression of cirrhosis, it is significantly associated with lower risk of liver-related events. If you see here, almost 85 to 92% reduction in incidence of liver-related events. Furthermore, authors have shown that ELF and VCTE can be surrogates to identify who have improvement in fibrosis, i.e., regression of cirrhosis in patients with natural cirrhosis. I believe these are very important data as we conduct clinical trials in natural cirrhotic patients. These are, as you know, there's been a lot of interest in lean NAFLD. It is controversial whether the outcomes of lean NAFLD are different than obese with NAFLD. This is a population-based study from Olmsted County from Mayo Clinic investigators. Bottom line, lean NAFLD compared to only obese NAFLD was not different in terms of incidence of cancer, cardiovascular events, or development of cirrhosis. However, inexplicably, lean NAFLD had significantly higher mortality compared to obese NAFLD. We just have to see the peer-reviewed paper to see how authors can explain this observation. Finally, in terms of this natural history, I found this abstract from AMRA. This was an oral presentation to be quite interesting. Several cross-sectional studies have suggested this phenotype of subphenotype of patients with high visceral fat but low liver fat may be associated with heightened risk of coronary artery disease. In this particular study, AMRA investigators have conducted MRI, six-minute MRI protocol on more than 25,000 individuals. What they have shown is a couple things. One is visceral fat nicely stratifies patients at risk for incident cardiovascular events. Importantly, though, is among those with high visceral fat, if you have low liver fat, that seems to be the highest risk group for incident cardiovascular events. What does this mean? Authors conclude if you have a clinical trial that is selectively reducing hepatic fat with no impact on high visceral fat, that may not be good for cardiovascular events. We just have to see if this concept will play out in clinical trials. In the last few minutes, I want to review selected non-invasive assessment presentations at this liver meeting. This is a very interesting work from perspective diagnostics. As you may know, Perspectum is Oxford-based company that has a liver multi-scan, which gives couple imaging-based readouts. One is corrected T1. The other is MRI PDFF. What they've shown in the past is that corrected T1 values can separate patients with NASH, NAFL, and non-NAFLD patients, although there is some overlap. Recently, they also have shown among patients with NASH, corrected T1 can separate patients who have high-risk NASH from those with low-risk NASH. High-risk NASH is defined as NASH with at least stage 2 or higher fibrosis. More recently, what they have done, though, they have modeled two different models based on corrected T1. The first model is corrected T1 plus PDFF combined, had an AUROC of 0.75 for identifying patients with high-risk NASH with tight confidence intervals. Another model they developed is called CTAG. This is corrected T1 plus AST plus glucose. This had an AUROC of 0.82, once again, confidence intervals ranging from 0.79 to 0.85 for identifying high-risk NASH. There was a companion poster 1495 from Perspectum with investigators from Harvard doing a cost-effective analysis of different strategies for identifying patients with high-risk NASH, and they observed that liver multi-scan is most cost-effective compared to other strategies such as VCTE-based or MRE-based over a broad range of costs. That's interesting. In addition to that, I've also identified five other abstracts in the non-invasive testing, including the one that I presented on Saturday. These are very interesting due to time constraints. I'm not able to go in detail with them, but I encourage you to look these up as these certainly are to move the field forward. This is my last slide, and I want to thank you for your attention, and I hope this debrief to be of educational value to you and your colleagues.

NAFLD

ASLD meeting

NASH treatment trials

pan-PPAR agonist

semaglutide

FXR agonist