Well, good morning. It's just about over. I'm going to end up with a NAFLD debrief, and I want to thank the organizers for giving me the opportunity to do this with you. So my name is Anna May Deal. I'm a professor of medicine at Duke and the chair of the NAFLD SIG. Those of you who know me know I'm a runner, and we are going to run through about 5 percent of the NAFLD presentations that were presented at this meeting, and so I urge you to keep in touch with the information that you have so that you can review all the other ones. As an upfront disclosure, I know many of you are interested in the randomized controlled trials. There were three that were presented here yesterday at the late breaker, but the details of that came in actually after I had prepared this talk, so I want you to please go back and look at that, but the bottom line is that an analog of FGF21, an analog of p-agglutazine, and a beta-chloric acid all showed efficacy as treatments for NAFLD. So I've divided the rest of the talk up into a series of little bites, and I took the liberty of summarizing every abstract in a very standard format, hopefully so that you'll easily be able to see why they did the study, how they did it, and what the take-home messages were. So the first section is on clinical research modifiable risk factors, and I guess you can see from the logos that the two things I focused on were social determinants of health and weight. So the first abstract is about social determinants of health and overall mortality in cirrhosis and cardiovascular disease. So these authors want to acknowledge that there are social determinants of health effects on incident outcomes in metabolic liver disease, but we're not exactly clear about that. And so they wanted to determine the impact of social determinants of health in a large integrated health system, in this case Michigan Medicine. They identified individuals with NAFLD, and they compared incident mortality, cirrhosis, cardiovascular disease, obesity-related cancers in NAFLD cases that were at the highest versus the lowest quartiles of affluence after correcting for all the variables that we know impact NAFLD outcomes. So the study enrolled over 40,000 people with NAFLD, and what they discovered was that affluence independently decreased the hazard ratio for incident mortality, for cirrhosis, and for cardiovascular disease. So they concluded that social determinants of health are significant independent risk factors for overall and liver-related mortality in NAFLD patients. The next study kind of did the same thing in a pediatric cohort. So again, these authors knew that successful implementation of lifestyle interventions can prevent and improve NAFLD, and the ability to do this may be impacted by social determinants of health. So their aim was to determine how neighborhood resources impact pediatric NAFLD prevalence. They developed a geospatial map of pediatric NAFLD distribution in neighborhoods around Sinai in New York City, and they identified about 200 children with NAFLD. They were pretty young, 10 years old. Most of them were male, Hispanic. They were on Medicaid, obese, and had a mean ALT of 91. About half the children had transient elastography performed, and 68 percent were found to have significant steatosis, and 11 percent were found to have advanced fibrosis. A smaller subset of the children had a liver biopsy, and 30 percent of those who were biopsied had F3-4 fibrosis. Sadly, the NAFLD cases clustered in neighborhoods with the highest community deprivation index. So the authors concluded that social determinants of health are likely to impact both risk for NAFLD and the success of lifestyle interventions to treat NAFLD in children. So in summary of this part, the social determinants of health are significant, independent risk factors for all cause and liver-related mortality in adults with NAFLD, and social determinants of health are likely to impact both risk for NAFLD and success of lifestyle interventions to treat NAFLD in children. The next section talks about body weight, and the first paper is talking about a new drug that is an uncoupling agent. And these authors acknowledge that obesity is a major risk factor for NASH, and many of the most effective weight loss drugs have significant toxicity. So they wanted to evaluate the safety and efficacy of a novel mitochondrial uncoupling agent called HU6 that enhances the ability of liver mitochondria to burn fat. So they took 80 NAFLD patients with high BMIs, high liver fat, they randomized them to placebo or different doses of this drug, and they re-evaluated them after a couple months. And what they found was that the uncoupling agent caused a dose-related decrease not just in liver fat, but in whole body fat and in body weight. There were no significant adverse events, and the main side effects were mild diarrhea and flushing. So this may be a novel approach to treating NAFLD. Agents that selectively uncouple liver mitochondria appear to safely reduce fat accumulation in liver and other depots. This causes weight loss, and so it may be a new treatment for NASH. The next study used the more conventional way to lose weight. And these authors, again, said obesity is a major risk factor for NASH, but we know that less than 10 percent of NASH patients can actually achieve or maintain sufficient weight loss to improve NASH. So they aimed to evaluate a mobile health-delivered lifestyle intervention, NOOM, to induce and sustain weight loss in NAFLD patients. So they invited 51 NAFLD patients to enroll in the study for four months, and they compared weight loss to the NAFLD cohort that was treated with NOOM versus a matched control group that received standard of care. And compared to the people who got the standard of care, the NOOM patients lost significantly more weight, 5.3 kilograms versus 1 kilogram. Also more NOOM patients achieved a greater than 5 percent decrease in body weight. Forty-five percent of the NOOM patients did that versus only 15 percent of the standard of care patients. So the NOOM-treated patients increased their physical activity by 35 percent. At the beginning of the study, they were only walking 2,000 steps a day. But by the end of the study, they were almost close to 10,000 steps per day, 7,000 in the NOOM group, and there were no adverse events. So the authors concluded that compared to the standard of care approaches, mobile health-delivered lifestyle intervention programs can significantly enhance weight loss and improve physical activity in NAFLD patients. So this part, I think we can say agents that permit selective uncoupling of liver mitochondria appear to safely induce weight loss and reduce fat accumulation in the liver, and so might be a new treatment for NASH. Mobile health-delivered lifestyle intervention programs are better than standard of care for enhancing weight loss and improving physical activity in NAFLD patients. We've all been struck, I guess, by the association between diabetes and NAFLD, and there were some nice papers about this at this meeting, and I'm going to tell you about this one that looked at the prevalence of NAFLD in advanced fibrosis, cirrhosis, and cancer in a prospectively assessed group of people with diabetes mellitus. So again, we all know type 2 diabetes associates with worse liver fibrosis in patients with NAFLD, but how about if we go into a diabetes clinic and say, what is the prevalence of advanced liver fibrosis? So they examined the prevalence of advanced fibrosis, cirrhosis, in a prospectively recruited cohort of adults with type 2 diabetes. They enrolled 500 adults over the age of 50 with type 2 diabetes who were being followed in either primary care, that we should be, or endoclinics. They did blood tests to rule out other liver diseases, and then they did imaging, either MRE or transient e-lactography, and a subset of the people had a liver biopsy. And what they found is, using imaging-based criteria, that 65 percent of these people in the diabetes clinics had a NAFLD, and about 14 percent had advanced fibrosis, because 6 percent had cirrhosis. In the cohort that had a liver biopsy, the prevalence of NAFLD was very similar, but it looked like they picked up a bit more advanced fibrosis and cirrhosis. Patients that were obese and were using insulin had an increased odds for advanced fibrosis. And amazingly, three of the people that had cirrhosis of the 29 cirrhotic patients already had hepatobiliary cancer, two had HCC, and one had a gallbladder cancer. So they concluded at least 10 percent of older adults with type 2 diabetes have unsuspected advanced liver disease, including cirrhosis and or hepatobiliary cancers, and so systematic screening of older diabetic populations merits consideration. The next study looked at rates of fibrosis progression in diabetes. So the authors said that, again, we know type 2 diabetes is associated with a higher prevalence of advanced fibrosis and cirrhosis and HCC, but is the rate of fibrosis progression faster? So they compared fibrosis progression rates on consecutive liver biopsies in NAFLD patients with and without type 2 diabetes. This was done with the NASH CRN, the database two, three patients who had two liver biopsies at least two years apart, and the authors compared changes in fibrosis progression after controlling for other variables that impact NAFLD progression. There are almost 450 people in this study, about half with diabetes, about half without diabetes. The median time between biopsies was at least three years. The cumulative incidence of fibrosis progression was higher in the diabetes patients than the non-diabetics. It was 23% versus 19% at four years, and 93% versus 76% at 12 years. And the precedence of diabetes was an independent risk factor for fibrosis progression. So they concluded that diabetics with NAFLD have a significantly higher rate of liver fibrosis progression than non-diabetics. And this is an important thing to consider when designing both patient management and clinical trials. So this part of the talk, at least 10% of older adults with type 2 diabetes have unsuspected advanced liver disease, including cirrhosis and hepatobiliary cancer. Diabetics with NAFLD have a significantly higher rate of liver fibrosis progression than non-diabetics. Chronic kidney disease is also coming into play in the NAFLD field, and so there was a nice abstract about this. They actually looked at two groups of people with cirrhosis. One was NAFLD cirrhosis, and the other was non-NAFLD cirrhosis, and they looked at what happened with kidney disease. So as we just said, maybe chronic kidney disease associates with NAFLD. What are the risk factors? So they defined chronic kidney disease incidence and predictors in NAFLD cirrhosis. They compared cardiovascular parameters and renal function biomarkers in patients with cirrhosis due to NASH, group A, close to 900 people, versus cirrhosis due to other liver diseases, group B, about 1,000 people, and then they followed them. So the NASH cirrhosis group, group A, were older, they had more women, more diabetics, higher BDMI, lower MELD, and this is interesting, higher wedged hepatic venous pressure gradient and pulmonary vascular resistance, but similar SVR and a lower cardiac index. After controlling for all the confounders, they found that NAFLD was an independent predictor for chronic kidney disease. Other predictors were age, MELD prior AKI, lower SVR, and diabetes. The renal markers, biomarkers, were also higher in the NAFLD cirrhotic patients, both at enrollment and follow-up. So I think this study tells us that NAFLD is an independent predictor of chronic kidney disease in patients with cirrhosis. There's a lot of interest in noninvasive tests, and so I've selected a couple of pictures that talked about this. Here are two that talk about transient electrography that I think everyone will find very interesting. So the first one acknowledged that prognostic significance of increased liver stiffness on transient electrography is unclear. So their aim was to study the prognostic value of liver stiffness simply as measured by elastography for predicting both progression to cirrhosis, which they defined as liver stiffness greater than or equal to 14.9 kilopascal, or worse composite clinical outcome, and their clinical outcomes were death, decompensation, liver cancer, or MELD greater than 15. They followed the NASH CRN database two, three participants who had a baseline score less than 12.1, so less than cirrhosis, and probably equivalent to F3 fibrosis, and at least two transient electrography testings done between 2014 and 2022. And then they identified factors that distinguished the progressors from the non-progressors. In this cohort, 13 percent of patients progressed to cirrhosis after about five and a half years. The progressors were more often white. They had higher baseline BMIs, higher waist circumferences. But look at this, the baseline liver stiffness was also higher in the progressors, 8.8 versus 6.1. And, of course, they had lower baseline albumin and platelets. The incident combined clinical outcome endpoint was significantly higher in the progressors, 26 percent versus 10 percent with a habituated ratio of 1.84, and that was statistically significant. And then after controlling for all the confounders, liver stiffness, as measured by transient electrography, defines cirrhosis. If that was used to define cirrhosis, it could independently predict an increased risk for combined clinical outcomes. So the authors concluded that progression to cirrhosis, as defined by VCTE, independently increases risk for poor clinical outcomes in NAFLD. A very similar sort of gist of an abstract was done by another group. And again, they were saying that the utility of transient electrography for predicting clinical events in NAFLD is unknown, and they wanted to address this. And they were looking for, as their clinical outcomes, liver cancer, liver failure, ischemic heart disease, and death. They had close to 1,000 people with biopsy-proven NAFLD who were followed for at least three to about three and a half years, and they had outcomes. Twenty-four patients developed HCC, nine had liver failure, 14 had significant ischemic heart disease, and there were 16 deaths, four were liver-related. Correcting for all the confounders, they too found that liver stiffness measurements were independent predictors for the liver outcomes, HCC and liver failure, with quite respectable hazard ratios. Patients with liver stiffness measures greater than or equal to 10 kilopascals had already had lower platelets, higher BMI, age, diabetes, complications, AST, GGT, but most importantly to us, if your liver stiffness was simply greater than or equal to 10, you had an increased risk for hepatocellular carcinoma and liver failure. So, they conclude that VCTE is an independent predictor of liver cancer and liver failure, and that a cutoff of 10 kilopascal identifies the at-risk group. So, from the non-invasive summary part one, we can say progression to cirrhosis as defined by transient elastography independently increases risk for poor clinical outcomes in NAFLD, and that VCTE is an independent predictor of liver cancer and liver failure, and a cutoff of 10 kilopascal identifies the at-risk group. So, now let's move to serum biomarkers. So, we all know that we're trying to get a biomarker that can be rolled out into the real world, preferably in primary care type settings, and that FIV4 is widely used to identify individuals with at-risk NASH, defined as NASH with F2-3 fibrosis, but the optimal cutoffs are unclear. So, these authors actually leveraged non-invasive screening measures for a large clinical trial that's ongoing to evaluate the accuracy of FIV4 cutoffs, those low cutoffs that we say exclude people with significant fibrosis. They used a cutoff of 1.3 or 1, and they then asked, what are the sensitivity and specificity of those FIV4 cutoffs? How many people were misclassified? And what they found was actually a little bit disappointing. They found that the FIV4 only had an ROC of .68, and that a FIV4 less than 1, even, or 1.3, incorrectly excluded 42 percent of people with F2 fibrosis and 17 percent of people with F3 fibrosis, and this was particularly inaccurate in young patients. And that's not so surprising because age is a part of the FIV4, so you can imagine if it's inaccurate in young patients because of age, it also may be inaccurate in old patients because of age. So, the conclusion is that FIV4 may underestimate fibrosis severity, particularly in younger patients. So, another study kind of did the same thing. They looked at, here, combining some of the non-invasive serum markers with some imaging. Again, and they were looking at a group of patients that have a particularly high risk for fibrosis that we just talked about. We know that diabetics have more fibrosis and they're progressing at a higher rate. How did these biomarkers perform in that population? And they were trying to identify the best non-invasive tests for identifying fibrotic NASH in this population with a very high risk for liver fibrosis. They had over 300 patients with type 2 diabetes and an increased ALT that was done at least two months apart. They underwent a liver biopsy, and then they were screened very close to when the liver biopsy was performed using these non-invasive tests. They found that the prevalence of fibrotic NASH in their population was 43 percent, that combining imaging with AST, or one of the non-invasive tests, outperformed the other tests for identifying fibrotic NASH in the diabetic patients, and the FAST test, which uses fiber skin or transient elastography plus AST, was cheaper and more available. So the take-home message is that adding imaging to seromarkers is better for identifying fibrosis in diabetics. This is a nice study trying to understand whether we can use adult things to predict things in children, and so they wanted to determine the best non-invasive tests for identifying fibrotic NASH in kids, and they looked at the existing models in over a thousand children in the NASH CRN cohort, and that they found the existing models that we use in adults performed poorly in kids. They developed a new model that has an ROC of .81, 77 percent sensitivity, 72 percent specificity. It includes age, sex, height, weight, and then multiple lab values, so it's a little bit clunky, but it works. And so they conclude that kid-specific NITs are better for identifying NASH patients with advanced fibrosis than standard NITs that we use to do this in adults. So the non-invasive summary testing part two, FIF4 may underestimate fibrosis severity, particularly in younger patients. Adding VCDE to AST is the most accurate way to identify fibrotic NASH in diabetics, and kid-specific NITs are better than staging fibrosis in kids than standard NITs that we use in adults. Now I'm going to conclude with my love, basic science. This is a really frightening poster, or presentation, by Dr. Sukhionian's group. Who knew your liver has a microbiome? Not your gut, your liver. So we have bugs in our liver. And the microbiome differences may explain differences in navel pathogenesis, and host genetic differences may influence the liver microbiome composition. So these authors wanted to explore the relationship between the liver microbiome and the liver epigenome. And so they took fresh liver nuclear extracts from over 100 navel patients, and they looked for the bacterial DNA, and then they correlated it with things that regulate epigenetics. So this would be protein acetylation and DNA methylation. And they found that the bacteria, different strains of bacteria correlated with differences in liver protein acetylation, and those differences in liver protein acetylation correlated with differences in severity of steatosis, liver damage, and blood glucose. And that global decreases in DNA methylation associated with increases in other bacteria. So the bottom line is, the microbes that we have are regulating the systems that control epigenetics in liver cells, and this provides both an opportunity, I guess, to manipulate the microbiome and see if we can change the epigenetics and control the pathogenesis and progression of NASH. This is a nice paper that talks about viscoelasticity. So we know that non-serotic NASH patients can get liver cancer. And we know that this is particularly likely to happen in NASH patients with diabetes. So why is diabetes a risk factor for liver cancer? So in diabetes, we measure advanced glycosylation endpoints, or ages, and the age we use all the time in clinical practice is the hemoglobin A1c. So these authors said these ages accumulate in the livers of patients with type 2 diabetes in NASH. And so they wanted to see if that would change the properties of the matrix, and then maybe induce mechanosensitive signaling pathways, and the one they looked at was YAP, that promote liver cancer growth. And so they studied livers from NASH patients with or without type 2 diabetes, as well as from mouse models in NASH, where they could manipulate the amount of ages that accumulated in the liver. And they found that the livers from the NASH patients with type 2 diabetes do have more ages, and they have higher viscoelasticity. Now for everybody that just listened to all the transient elastography talk, and you want to equate it to stiffness, it's not stiffness, it's a different property. But the point is, these ages, if you have more of them, you have more viscoelasticity, and in preclinical models, if you increase the age, you increase the viscoelasticity, you upregulate these mechanosensitive signaling pathways that actually drive liver cancer, and you get more liver cancer growth. So maybe this is why more HCCs develop in NASH patients with type 2 diabetes, because they're accumulating these glycosylated proteins, changing the liver matrix, activating these signaling pathways that can promote liver cancer growth. So this part, liver of NAFL patients harbor bacteria that control the expression of genes that regulate NAFL progression. Type 2 diabetes increases the risk of liver cancer, at least in part, because it causes accumulation of glycosylated proteins that change the liver matrix properties to induce these pathways that promote cancer growth. Now we are almost done, two more. This one is, I'm going to tell you about two cell types in the liver that you heard from the first speaker, Dr. Carr, about how all these single cell studies are exploding now. So here's a couple that use this technology to advance our understanding of NASH. So these authors know, like all of us, that antifibrotic therapies are needed for NASH because severe liver fibrosis increases the risk for death from all causes and liver disease in NASH patients. Hepatic stellate cells are the major producers of fibrous matrix. And we all know that stellate cell populations are heterogeneous with regards to their fibrogenic and fibrinolytic activities, and now we have these tools that can figure out which of these subpopulations are the bad actors in NASH. So they wanted to characterize these populations to identify factors that are enriched in the NASH-associated stellate cells, and they used these fancy new techniques, single-nuclei RNA-seq and single-nuclei ATAC-seq, to profile stellate cells in human livers. And they took normal, NAFL, and NASH. And it was amazing. They identified genes that distinguish stellate cell populations in these different disease states. And they discovered disease-related differences in the stellate cell heterogeneity, and they found six genes that are particularly enriched in the activated stellate cell populations in NASH. And so this deeper characterization using these novel techniques are going to provide us with new targets for antifibrotic therapies, and hopefully will eventually lead to improvements in prevention of liver fibrosis in NASH. Last one. This is about macrophages. Okay. So liver macrophages drive inflammatory liver damage in NASH. Correcting this might treat NASH. So their aim was to determine what makes the macrophages pro-inflammatory in NASH. So again, they used these similar techniques. They profiled liver macrophages in humans and mice with NASH. They manipulated the macrophage metabolism in the mice and assessed the effects on NASH. And what they found is that macrophages in NASH livers had activated an enzyme, PKM2, that promotes glycolysis, and it makes the cell make lactate. And that lactate then reprograms the cells to be pro-inflammatory. But the beauty of it is they can reprogram them by manipulating that enzyme. So they can take a pro-inflammatory one and flip it to be an anti-inflammatory one. So they conclude that macrophages in the livers with NASH have metabolic abnormalities that make them pro-inflammatory. And treatments that reprogram their metabolism can make them become anti-inflammatory and improve NASH. So I'll stop there. After summarizing, we know that genes that are uniquely enriched in fibrogenic stellate cell populations in NASH have been identified. And they've revealed novel targets and potentially will lead to the development of new antifibrotic therapies. Correcting macrophage metabolic abnormalities in NASH livers can switch them from being pro-inflammatory to anti-inflammatory, and at least in mice, this improves NASH. So I'll stop here. And thank you. And I guess this concludes the session. Is that correct, Madam President? So congratulations to everybody for making it through the liver meeting. Have a safe trip home. Thanks for coming.

NAFLD

randomized controlled trials

social determinants of health

body weight

diabetes

non-invasive tests

liver microbiome