Well, welcome. My name is Stephen Harrison, and I'd like to extend this special welcome for joining us on today's webinar entitled Microbiome in NASH and Cirrhosis, sponsored by AASLD. It is my distinct honor and privilege as the NASH Special Interest Group Chair to moderate this session on microbiome in NASH and cirrhosis, and we have many exciting developments in this field, and here to tell us about them today are three world-renowned clinician scientists in the field. First off, we have Suzanne Dakota, Director of the Microbiome Research Center at Cedars-Sinai, Rohit Lumba, Professor of Medicine and Director of the NAFLD Research Center at the University of California, San Diego, and last but not least, Jaz Bajaj, who is Professor of Medicine at VCU and recently began as Editor-in-Chief of the American Journal of Gastroenterology. So without further ado, I'd like to turn the stage over to Suzanne to kick off our webinar. Thank you so much for listening. Thank you so much for inviting me to join this webinar, and today, I want to give just a brief primer on the microbiome. I'm going to start by sharing some basic fundamental principles that govern microbiome colonization in the GI tract, just to set the stage for how in the world we get all of these trillions of organisms in our bodies, because we weren't really born with them. So a little bit about the things that affect our microbiomes and how it evolves during the course of our life, and then I'm going to discuss some basics around methods that we use in microbiome research, and if you are interested in conducting microbiome research or just want to interpret the literature better, this will be a useful sort of primer to help guide you. I have no disclosures. So there are in the body, we have microbes all over our body, really, but there's true defined microbial niches that we know are the most dense sources of microbes in our body. So our nose and our mouth, our skin surface, usually on the skin, it tends to be in what we might consider the more moist areas like our elbows, behind our knees, behind our ears, areas like that. The GI tract, which has the densest community of microbes in the body, and then the urogenital tract. But there is a growing body of literature showing that bacteria have also been found in atypical sites, and some of these sites are the bladder, the placenta, and adipose tissue, which is something my lab is particularly interested in. The placental microbiome is highly controversial area of research, and the general sort of view is that the placenta truly is sterile, and there might be transient microbial DNA that passes through. So we, you know, for all intents and purposes, should consider the womb sterile, and essentially at birth, your first microbes are introduced during the birthing process, but not before that. But the question always is, what is real? You know, what is in those niches? What is true microbial colonization versus just, you know, hitchhikers or transients? And then what's contamination? We get contaminated by the external environment all the time on a day-to-day basis. So if we focus on the GI tract, which is all of our sort of area of interest, this is probably one of the more interesting areas, because there are 100 trillion bacteria, roughly, in our GI tract, and they colonize from the mouth to the anus. But the density is not equal throughout the GI tract. Actually, the microbes become more and more densely populated the more distally you go in the GI tract, with the colon harboring the vast majority of your GI microbes. However, in conditions where you might have stasis in the GI tract, so chronic constipation, or in certain cases like diabetes, you tend to get a little bit of, you know, slow motility in the GI tract. When there is stasis, microbes love to multiply whenever there's lack of movement. So just emphasizing that regular, you know, daily bowel movements are actually critical for keeping your gut microbial populations in check. And when we think about small bowel overgrowth, which is a growing problem in the U.S. and around the world, that's oftentimes sort of related to constipation and sort of that backup of microbes. So one of the primary things that determines each, all of our microbiome compositions in our body, and it's really like a, truly like a fingerprint. None of us have an identical microbiome composition in our guts, that we might all have the same general type of species, but the composition and proportion to each other is different from one person to the next. But your diet is a major determinant of what microbes colonize your body and where in your GI tract they colonize. So microbes really, I mean, it's a symbiotic relationship in the healthy homeostatic state, between the host and their microbial ecosystem. So microbes colonize where their function is best suited. You're in your small bowel, that's primarily your source for nutrient absorption, and your colon is primarily the site for fermentation. And so your high fermentative organisms will reside in your colon, because that's where that process is carried out. So really in the healthy state, form follows function. But in disease, that all goes out of whack, and it really becomes a case of survival of the fittest. And those organisms that can survive the best in a chronic inflammatory state, for example, are the ones that will survive. And often those are not the most beneficial ones for you. So how modifiable is your gut microbial community? Well, the reality is you probably can do it through diet. And what we don't totally know is how stable those are. If you drastically change your diet from, let's say, a Western high-fat, high-sugar diet, and you go completely, you move to Asia, and you start eating a lot of fermented foods, and your composition of your diet changes a lot. Can you completely overturn your microbiota? I mean, can you fix that in terms of maintaining that change? We don't entirely know that. So in the early part of life, this is really the most dynamic stage of your gut microbial community. It's highly changing, because the individual's daily exposure is everything is new. So what we do know is some of the big things that can affect your initial microbial community is, were you birthed by C-section or vaginal? Were you fed by breast or bottle? Because each of those introduce different types of bacteria. Now, one is not necessarily worse than the other. There's a lot of research going on on that, but it is different, and it will result in different microbial communities, and what's the long-term effect of that? Environmental exposures. What diet, you know, where in the world do you live, and what diet are you weaned on to when you start to move on to table foods? Are you given a lot of antibiotics when you're a child? Obviously, that's a major perturbation. And what sort of viral community are you exposed to? We know that viruses and microbes interact in quite an intimate way in the GI tract. So just a quick example of a cool study highlighting the importance of the indigenous diet that you are weaned on to, and how different that can, how much of an impact that can make on your gut microbiome. This is a cool study that came out of Jeff Gordon's lab at WashU, where he took populations from three distinctly different parts of the world, and just across the lifespan, across different ages, sequenced the microbiome and looked for differences. And each individual dot is a person, their microbiome composition, and on the left-hand side, OTUs are just different bacteria. So the higher the number of OTUs, the more different kinds of bacteria you have, and then on the y-axis is age. And what you can see, the first thing is that in the U.S., which are the blue dots, we have lower OTUs than the other two populations in Africa and South America. This means the American population has less microbial diversity in their GI tract. And this has been shown in other studies of the Western diet. But not only are the number, the different types of bacteria lower in the U.S., this second figure is showing how similar or different your microbes are, and it's showing that the bugs that are actually there are completely different than in these two other populations. And in this study, they largely attributed this to Western diet versus native diet. It's really diet-driven. So what perturbs the gut microbiome? And we know, I was just explaining that with Western diets, that is sort of a hallmark of Western cultures, and we're talking about now folks in the U.S. here. And we know that with Western diet, we have a rise in chronic inflammatory diseases. So what are all the factors of Western lifestyle that we know in the literature can affect the gut microbiome? Definitely antibiotic usage, diet I mentioned. Hygiene is very interesting. They found that with increasing hygienic practices, you can actually lose some members of your gut microbiome. Prescription drug use, that's a new and emerging area. How do drugs interact with the microbiome? We know that drugs can synergize with or sometimes inactivate certain medications. Circadian rhythms in the gut microbiome is a very interesting field. There have been a lot of cool research coming out regarding jet lag and shift work and how that affects your microbial composition. Your gut microbes have their own circadian clock, which is quite interesting. GI stress, obviously, we know that would obviously have an impact on the gut microbiota. And then lifestyle. People are starting to research exercise and how, you know, things that you may do in your day-to-day life might affect your gut microbiome. And the downstream effect of this is that an altered gut microbiota, which some people might call dysbiosis, I don't tend to use that term, but it can lead to what we might call a leaky gut or increased gut permeability. And that has been associated with food allergies, an increase in food allergies. Compromised immune defense, which is involved in the pathogenesis of inflammatory bowel diseases. And then chronic inflammation and metabolic dysfunction go hand in hand because there's a theory that with impaired gut barrier and chronic inflammation, you actually get seepage of bacterial antigen into the bloodstream, leading to inflammation in adipose tissue and leading to metabolic syndrome and obesity. And that's called metabolic endotoxemia. So what is a healthy microbiome? This is a study from many years ago that was initiated by the NIH where they took populations from St. Louis and Houston. And I'm just going to highlight some things they found, but I also want to highlight that we really don't know what a healthy microbiome is. And the reality is a healthy microbiome is a very individual concept and your healthy microbiome might be very different than my healthy microbiome. But what we do know is that there's 10,000 different kinds of species that can colonize the GI tract. You and I may not have 10,000 species in our gut at any given time. We probably have about a thousand, but there's 10,000 different kinds that can colonize the gut. They were able to identify almost 99% of all genera of bacteria in these individuals. And there are 360 times more bacterial genes than human genes in the body. So these are, these organisms are doing something. They have a functional ability and most of the time it's a symbiotic relationship, not always. And there's significant functional redundancy. And this is an important principle I want to highlight. So that means that we have many bacteria that can do the same thing in our body. So why in the world do we have that level of redundancy? Why would we need to have, you know, 50 different bacteria that can do the exact same process in our gut? Well, given all of the perturbations that I showed you, many of those organisms can be wiped out or killed. If you wipe out, if one organism can do one function and you wipe it out, you no longer have that function in your GI tract. So it's sort of an evolutionary principle of, you know, if you, you want to have still that function carried out in the GI tract. So you don't, you know, it doesn't lead to, you know, an altered metabolic state or an inflammatory state, as long as those functions can exist, then if one happens to be killed, then that's, that's okay transiently. So the top is showing that across different sites in the body, the stool, the mouth, the nose, your organisms that colonize those niches are very different, just looking at the colors. Okay. The different colors represent different organisms. But if you drop down and look at the functions encoded in those bacteria in each of those same sites, it doesn't really vary that much. It looks pretty stable. So you can have a very different composition, but you still have the same functions carried out. And that is extremely important for a healthy state. So briefly, since we're running short on time, I just want to talk to you about some important methods that we use in microbiome research. These are just an overview. There's many, many more than this, but these are sort of the ones that are being used most often these days. We have DNA-based approaches, which are 16S and ITF sequencing for bacteria or fungi. And that's really understanding the composition of who is there in your community, what bacteria are present or not present. And then RNA-based approaches will get at what genes are actually transcribed in a given context. So it gives you a sense of the activity of the gut microbiome. Protein-based approaches is still very new, metaproteomics. Many of these same tools are used for human research too. And this is understanding, you know, what proteins are the bacteria producing, which is even a more functional readout. And then metabolite-based approaches, what metabolites are being produced and released into the milieu. DNA-based approaches, so sequencing, profiling, and metabolomics are probably the two that are most often used right now because those methods are the most robust. But all collectively, all of this, these data have a lot of variability. So you really need large ends to achieve the statistical significance that you need for these studies. And collectively, when you hear big data or multi-omics, it's these approaches that people are using. You have large sample sizes, you're generating a ton of data, and you're doing large correlations and associations with this data. So if we go to just look at really what is the bread and butter of microbiome analyses, and it's the DNA-based approaches, the community profiling. And you hear a lot about 16S rRNA sequencing, and you hear about metagenomic sequencing. They're similar but different, okay? And you often hear that, oh, metagenomic sequencing is far superior to 16S, and everyone wants to move to metagenomic sequencing, but it actually, that's not always the best approach for your samples. There are many cases where 16S sequencing is much more superior to metagenomic sequencing. So you really have to understand what the purpose of your analysis is. So for 16S rRNA sequencing, you're taking little short reads of the bacterial 16S gene and amplifying that. Because you go through this amplification process, it works great for tissues where you have a lot of human DNA and maybe very little bacterial DNA. So we do a lot of studies on fat where typically there isn't a lot of bacteria there. 16S works really well for that analysis because there's a lot of host contamination in there. But for stool, for example, where it's mostly bacterial DNA, the metagenomic sequencing is far superior. So you really want to think about, okay, what is my sample and what, you know, how much host DNA is there and how much bacterial is in there relative to that. 16S rRNA sequencing is also generally cheaper and faster. So for pilot studies, it's a fantastic first approach. And often what people will do is sequence all of their samples by 16S and then take a subset and go a little bit deeper and use metagenomic sequencing for that. The problem with 16S is you really cannot resolve beyond the genus level. So it's very challenging to, say, identify a species with 16S rRNA sequencing. So it's somewhat frustrating because you really kind of get a broad view, but you don't get much detail. And there's a little bit of bias in your analysis, but that's true for any analysis that we use, even for human research. Now on the metagenomic sequencing side, you can get species level resolution. And depending on how deeply you sequence, you can sometimes even get strain level resolution. The really interesting thing about metagenomics that kind of sets it apart is you actually get an assessment of the functional potential of your microbiome. So you're not understanding what functions are actually carried out in the context, but it gives you a sense of the gene content and what functions are encoded in those bacteria that are perhaps increased in a given context or treatment group. So you can get a little bit more insight into what might be going on, how the microbes are involved in a phenotype. But 16S sequencing is strictly just a community profile, who is there. But the other thing to keep in mind is metagenomic sequencing, you sequence all the genes. So it's not just the bacterial genes, it's every single gene that's in your sample. So you're getting fungi, you'll get viruses, you'll get human. But again, it really depends on your tissue type. And this is just a little schematic about the process of 16S RNA sequencing, just highlighting that you're really sequencing a teeny tiny fraction. The 16S gene is a teeny tiny fraction of the bacterial total DNA. And then within that teeny tiny fraction, you're sequencing yet a smaller teeny tiny fraction. This is why you can't resolve really deep information from 16S sequencing, but it is also why you can amplify it from a low microbial biomass sample. So if you want to go and let's say you've done the big multi-omics studies, you've done 16S, you've done metagenomics, and you want to go really understand the mechanism behind why that organism is increasing. There's many tools for that. You can use quantitative PCR, which can truly be quantitative. If you know exactly what organism you're looking for, you can target just that organism in the sample. Growing your bacteria out of a sample will allow you to prospectively co-culture it with maybe immune cells or host cells. And then it allows you to also grow that organism and put it into germ-free mice, which are mice that are grown without bacteria. And it allows you to use the mouse essentially as a living incubator to test what is the effect of your organism on a host. So I'll just leave it there. And sorry, I went over. Thank you very much. Great. Thank you so much. Are you able to hear me? Great. I'll discuss the role of microbiome in NAFLD. And thank you to Suzanne for setting it up. These are my disclosures. Now, when we look at non-alcoholic fatty liver disease, it's really a complex metabolic disease with a gene-environment interaction. In the human body, we have about 23,000 genes or so. But in the human microbiome, we are looking at about 3.5 million genes. And we know that there are certain genetic factors that are associated with NAFLD, including PNPLA3, TM6SF2, MBOD7, GCKR, and several others. And as recently, RHSD17B13 has been shown to be protective. And then these genes may be modified by specific nutrients that are important in NAFLD, dietary excess, fat intake, carbohydrate intake, fructose intake, as well as alcohol use. All of these, as Suzanne just told us, would also modify our microbiome. So could there be a clear interaction between these genes which confer the risk for progressive NAFLD, as well as some that may be protective, are interacting with microbiome to really see which obese individual develops NASH and steatohepatitis, and who is likely to have progressive liver disease. And this is an intense area of investigation currently. Now, what are some of the links that liver disease in general, and NAFLD in particular, is inherently linked to microbiome? This is work done predominantly by investigators working on cirrhosis complications, such as Dr. Bajaj, who will be discussing these in further detail. But we know that patients with cirrhosis have increased bacteremia, and increased LPS levels in blood, and there is increased intestinal permeability. Patients with cirrhosis have bacterial overgrowth in the small intestine. Now, selective intestinal decontamination with antibiotics is beneficial for patients with decompensated cirrhosis. We know that when we have a patient with variceal bleed coming in, we give them ceftriaxone. When a patient with a suspected SBP comes in, we do paracentesis, which is diagnostic, and then we're looking for clinical evidence of bacterial translocation, leading to increased mortality in cirrhosis, and that's why we're using fluoroquinolones for secondary, as well as primary prevention of SBP. These underscore the importance of gut microbiome in hepatic decompensation. Rifaximin for the treatment of hepatic encephalopathy as an example. And as previously studied and published by Patrick Cometh, Raj Reddy, and Jas Bajaj, the infections and sepsis are leading cause of mortality in patients with liver failure. So this really points towards a major role of microbiome in cirrhosis and complication. Now we are interested, could it be possible that these perturbations are happening much earlier in the pathogenesis of disease and may be contributing to disease progression? How can the gut microbiota affect liver disease early on? Well, changing calorie yield of diet and modulate fat storage. So it's possible that depending on your gut microbiota, your yield from what you eat might be different. The gut microbiota may regulate gut permeability to release bacterial products into the portal circulation, modulate choline metabolism, produce endogenous ethanol, regulate bile acid, and regulate lipid metabolism. Now these are some of the mechanistic pathways where gut microbiota may interact with liver disease or liver physiology. Here we'll look specifically in NAFLD how gut liver access microbiome and microbial metabolites may play a role. This is a study that was published several years ago by Lixin Zhu and colleagues where they looked at pediatric NAFLD and they compared normal pediatric population obese children, and then compared them to biopsy proven NASH. And one key feature they found was that the blood ethanol levels were higher in children with NASH. So where is this ethanol coming from? And it was deduced that this is coming from the gut microbiome. There are certain bacteria that can form in alcohol and this could be seen in the blood and this could be potentially one of the reasons or it could be an epiphenomena with presence of steatohepatitis. But clearly linking gut microbiome to NASH as a phenotype in pediatric NAFLD. There was this follow-up paper from Anna Medeal and Jerome Bercier looking at 16S and we looked at where 16S might be useful versus metagenomic sequencing. And what they looked at was in patients with biopsy proven NAFLD, they did 16S in two specimens from patients with biopsy proven NAFLD. In patients with NASH stage two fibrosome higher, they noted decrease of the abundance in bacterioides and rivirtella and increase in the abundance of ruminococcus. And here are some of the results. We also had a really nice study from Spain in a group of women who were moderately obese and non-diabetic who had hepatic steatosis. They were coming in for bariatric surgery and this is that baseline. And what was noted that they compared those with hepatic steatosis versus those without hepatic steatosis. So normal versus NAFLD is the comparison here. And they noted that increase in enterobacteriaceae and streptococci and decrease in acromantia. There was another thing that was noticed and we wrote an editorial about it. There were key metabolites that were noted between NAFLD versus non-NAFLD. There were aromatic amino acid and branching amino acid. And in particular, they found that phenylacetic acid induced hepatic steatosis. And they confirmed that in mice model as well as in vitro in human hepatocytes. So this is telling us that not only the microbiome might be different in NAFLD versus non-NAFLD among obese individual, but they may be certain metabolites that could be picked up in blood. Then we looked at in our twin and familial cohort, Cyril Kossy, the endocrinologist, when she was visiting us as visiting scholar at UCSD, she looked at participants in the twin and the family cohort as well as in patients with biopsy for NAFLD in advanced fibrosis versus early fibrosis. And what we found was that there was a shared gene effect between liver fibrosis and hepatic steatosis. What that means is that people say that, well, liver fat has nothing to do with liver fibrosis. That's not true. We're studying fatty liver disease, both in alcoholic liver disease as well as in fatty liver disease. Quantity of fat plays an important role. So here we are showing that genetically there's a shared gene effect. Certain genes maybe have common determinant for hepatic steatosis and maybe also regulating liver fibrosis. And one of the common links we found was that this particular metabolite, 3,4-hydroxyphenylactate, again, similar to what was seen in the Spanish study, we documented it here in patients with biopsy-proven advanced fibrosis. That this could explain the shared gene effect, which means if you had hepatic steatosis and if you also saw this metabolite, then you were more likely to have liver fibrosis. So this is, again, putting the pieces together, linking gut microbial metabolites to liver fibrosis in NAFLD. Then we looked at patients with biopsy-proven NAFLD and did metagenomic sequencing of stool specimens collected very carefully in extremely valvenotype patients at UCSD. And we compared patients with early fibrosis with advanced fibrosis. And our intent was to develop a diagnostic biomarker using metagenomic sequencing in stool. What we noticed was that Firmicutes were higher. Firmicutes were higher in fibrosis stage zero to two, and proteobacteria were higher in fibrosis stage three to four Then at the species level, we documented that E. coli was three times more abundant in stage three and stage four fibrosis. Now, I just want to bring your attention to this. This is also, these patients did not have ascites or did not have any features of hepatic decompensation among those who were cirrhotic. So it's really important to see that all these changes are already happening much before patients developing ascites. And therefore it's possible that E. coli may be playing an important role in progression of disease much before patients develop ascites. Then we developed a metagenomic signature using 37 species with random forest model using machine learning approaches. And this was able to identify advanced fibrosis versus early fibrosis with a diagnostic accuracy of 0.93. So this is pretty significant that you could use gut microbiome as a proof of concept to diagnose advanced fibrosis, which is stage three or stage four fibrosis. We are doing validation studies in this domain now. Now, what is the mechanistic link, linking microbiome and bacterial metabolite with liver fibrosis? Well, LPS is a bacterial antigen that is released into the portal circulation from gut bacterial products and is an indicator of increased intestinal permeability. This leads to activation of a TLR4. And with the activation of TLR4, there's release of chemokine and adhesion molecules and recruitment of Hooper cell. This leads to activation of the quiescent hepatic stellate cell and this activation of quiescent hepatic stellate cell leads to collagen deposition and liver fibrosis. So here you can see a clear link between gut microbiome, gut microbial metabolite, activating hepatic stellate cell, leading to fibrogenesis. And this is a common path in chronic liver disease, particularly NAFLD as well as in alcoholic liver disease. Then we looked at a cohort of patients who had familial cirrhosis, which means one family member had advanced fibrosis and we phenotyped their first degree relatives. And we were able to pick, this is just using 16S, we were able to pick NAFLD cirrhosis in the index cases. And then using the same sequence, we asked the question that can we now detect first degree relatives? If we don't use any imaging, we only use the same signature that we developed in the probands with NAFLD cirrhosis, can we pick and identify advanced fibrosis? And the answer was we were able to identify majority of them just by looking at their gut microbiome. And this was the diagnostic accuracy of 0.92 and in the validation cohort, 0.87. So if we had a signature just with 16S, we could easily detect advanced fibrosis and cirrhosis in the first degree relatives. Then we moved on to looking at patients with cirrhosis due to NAFLD at UCSD using metagenomics and metabolomics. Did a very detailed analysis, Taghiou who works with Ron Evans in his lab at Salk Institute and we collaborated and developed a signature for NAFLD cirrhosis. And then we were able to validate that signature in a Chinese cohort, as well as in an Italian cohort. The Chinese cohort had predominantly hepatitis B and alcoholic cirrhosis and the Italian cohort had patients with alcohol and hepatitis C cirrhosis. So what we learned was that this was one of the first diseases, which is cirrhosis, where there was a universal signature across different ethnic and geographic groups. This is not common, why is that? Well, it's because cirrhosis is an extreme trait. And when you develop cirrhosis, you start seeing commonality irrespective of the baseline diet that there might be. Because diet itself may have an important role in shaping your gut microbiome, but when you develop cirrhosis, there are some common patterns that could be seen independent of which part of the world you're living in. So this is unique about cirrhosis. Then as a proof of concept, Stephen and I were involved in this trial where we were looking at FGF-19 and looking at treatment response in NAPFER-Li in terms of improvement in liver fat. There was no change in gut microbiome at baseline and week 12, but we noticed only one bacterial species, which is Vianella, that was increased 30-fold by FGF-19. So here we can show a dose-dependent response by FGF-19 activation, depending on the dose of the drug that we use. We had four different doses, 0.3 milligram, one milligram, three milligram, six milligram daily, where we were able to show that Vianella was increased. And so this could be potentially a marker or diagnostic marker of treatment response with FGF-19 as proof of concept. In conclusion, gut microbiome plays an important role in pathogenesis of hepatic fibrosis, progression of liver disease, and potentially complications of cirrhosis that Jas will discuss now. Gut microbial metabolites, such as aromatic amino acid, branching amino acids, and bile acids may be involved in the development of NAPFER-Li and progression of liver disease. Gut microbiome-derived signature might result in a non-invasive test for detection or advanced fibrosis or cirrhosis in NAPFER-Li. And I think this may happen in the next five years. The gut microbiome may also be used as a biomarker of treatment response in NASH and antifibrotic treatment trials. Thank you. Thank you to the ASLD, and thank you to the other speakers and Dr. Harrison for inviting me for this. I'm going to take over cirrhosis and beyond. These are my disclosures. And the question is, why is it relevant to study this microbial change, part of which has been answered by the excellent talk by Dr. Loomba? What are the levels of therapy in reducing severity of cirrhosis that can be affected by microbiota manipulation? Is microbial therapy going to be enough? And what are the barriers towards operationalizing these changes? Can microbiota and liver disease be gut biosensors? And the answer is yes, not only is all aspects of the gut-liver axis, as you've seen here, different in patients with cirrhosis compared to healthy controls, it is also associated with the different microbial composition and function in the saliva, skin, and the blood. And this has something to do with the inherent immune dysfunction that patients with cirrhosis get, which progresses as the cirrhosis progresses and potentially gets better after a liver transplant. So, and this is exemplified by an oversimplification of microbial composition in which bacteria that produce short-chain fatty acids, including one that Dr. Loomba mentioned, Vianella, were in the numerator, and Enterobacteraceae, which is gram-negative rods consisting of E. coli and Klebsiella are in the denominator. So a high score means, high ratio means things are better. So controls clearly had the highest cirrhosis-dysbiosis ratio, which went down all the way till infected inpatients who had spontaneous bacterial peritonitis. And this was stable over time, worsened with the development of first episode of encephalopathy, and was worse in people who were subsequently hospitalized. Moreover, this reversed and went back towards healthy controls when the patient received a liver transplant. So this is a proof of concept that the microbiota change with cirrhosis. On the left-hand side is the nice study that Dr. Loomba has already mentioned, but what I want to impress to you that this etiology of liver disease also has stamps beyond the cirrhosis development. On the middle part, after cirrhosis development, you have NASH cirrhosis versus non-NASH. And on the right-hand side is alcohol-related cirrhosis versus non-alcohol. I'm not going to go over the details, but it tells you that even after the development of cirrhosis, the etiology, the predominant etiology of cirrhosis has a likely effect on the microbial composition, which possibly predated the development of cirrhosis itself. So these can also be worsened in patients in quantitative metagenomics. Dr. Devkota had actually done a very nice job of telling us what the difference between 16S and metagenomics were. The prior studies were on 16S. This is metagenomics, which went deeper, but extends it into inpatients with healthy, compensated, decompensated. AD is acute decompensation, an acute and chronic liver failure, which is ACLF. And moreover, this determined, this was associated with prediction of alive, people who are alive or people who died at the metagenomic level as well. Microbial changes can also be linked to outcomes in outpatients. Encephalopathy, hepatic encephalopathy, which is associated with cognitive dysfunction, brain inflammation, and edema, as well as hyperammonemia, is related to multiple bacteria in the stool, multiple bacteria in the saliva, as well as bacteria in the colonic mucosa. And these are directly or indirectly linked to cognitive performance. And different sets of bacteria are associated with inflammation compared to hyperammonemia. So these bacteria that produce short-chain fatty acids are negatively linked with hyperammonemia, whereas E. coli are positively linked. Moreover, oral bacteria such as porphyromonidase are more linked towards brain inflammation. And indeed, in Alzheimer's dementia, you see a lot of the oral bugs associated with those progression as well. Microbiota can also be used to exclude significant dysfunction in cirrhosis. Moving on from what Dr. Lumba had said, using these as biomarkers, we can actually find out if someone has these bacteria, ruminococcus and Clostridium cluster-14B in stool and higher streptococcus in saliva, they're unlikely to be cognitively impaired. These can also predict who develops hospitalizations in outpatients. Both salivary and stool microbiota are able to tell us who gets these important clinically relevant outcomes. Moving forward in inpatients, we already talked about that metagenomic study that was a single center study. In another study on the top right, acute hepatic encephalopathy episodes had a microbial signature that was very distinct from people who are healthy, compensated cirrhosis and decompensated cirrhosis, which again predicted outcomes in these patients. In the bottom left is a multi-center study which shows that microbiota that are higher in those who died or received in hospice belongs to Enterococci and several gamma proteobacteria, which are the gram-negative rods, on admission. So this tells you on admission, the microbial profile is already changed in patients who are destined for dying or hospice compared to those who are alive. And in this other study in the top right, the red are the people who died and the blue are the people who survived. And these are microbiota bacteria that were taken every day, contrasting to the studies on the left where the microbiota were taken only once. So again, it tells you that the microbiota on admission and throughout the hospitalization could have something to do with the consequences of that disease. And this is something that we should err on modification sooner rather than later, because these things once admitted may be a little harder to take care of. What about microbial origin metabolites? We talked about composition, but as Dr. Devkota mentioned, and also Dr. Loomba mentioned, what the microbiota are doing are also very important. On the right side is the study of acute and chronic liver failure with serum samples in people who had already either developed it, compared it to acute decompensation, compensated and healthy controls. They found a lot of energy metabolites that were different between those who had ACLF versus acute recompensation. And they concluded that there was some metabolites-related signature. On the left-hand side is our Naxal study, multicenter study on 600 patients, in which we actually want to use metabolites as a predictive model rather than just a diagnostic model. What we found is aromatic amino acids, secondary bile acids, and xenobiotic modulation from the microbiota were independently associated with the development of ACLF inpatient and 30-day mortality compared to those who did not have this. Moreover, phospholipids were protective in this situation and estrogen metabolites, which are not really affiliated completely with the microbiota, were also associated with the development of ACLF. So it tells you not only the microbiota itself can be a marker, but their product can also be a marker even in situations as extreme as patients hospitalized with cirrhosis. So what are the therapeutic strategies that we can do in these patients? They go across many levels. The levels of the liver, the level in the gut barrier, and multiple things such as diet, probiotics, prebiotics, non-absorbable antibiotics, FMT, and absorbents have been tried. And of course, we are very familiar with antibiotics and FXR agonists, one of which was mentioned by Dr. Lumba et cetera also. So this is important for us to study further. In the GI strategies that we can also do a non-selective beta blocker, especially the Predesky trial, which shows reduction in inflammation, and periodontal therapy in which you clean the patient's teeth and hopefully things will get better. But the current therapies requires more precision because probiotics, even though they have been studied extensively, their evidence base is not very good. Hepatic encephalopathy is the one part of cirrhosis where probiotics, prebiotics, antibiotics, and laxatives still have very good evidence. And these are used variably in combinations with other processes, other medications based on what is available and what your patient can afford. So what are the other therapies that we, levels of therapy that we use to reduce the severity of liver disease? We can control the etiology, we can affect inflammatory milieu, we can change gut microbial composition and hopefully function, and change bacterial interactions with others. Most etiologies are not directly microbial. In fact, if you ask the patients to stop drinking, it's the stopping drinking that actually impacts what happens to them, et cetera, et cetera. Whereas the oral cavity is very important for systemic and inflammatory control because there's an oral gut hepatic axis, as we already talked about, porphyromonadesis, which causes periodontitis is associated with brain inflammation in hepatic encephalopathy. Multiple series of studies across pre-cirrhotic and cirrhotic patients have found that the more periodontitis you have, the more systemic inflammation you have, the more brain inflammation and the more liver inflammation you're going to have. So this is an easy fix. You can actually treat those patients, treat their patient's mouth, and hopefully in addition to all these other things, we can help these patients feel better and have better outcomes. And this is a study that we had done in which we had actually randomized, took patients who did not have therapy from the mouth and patients who did have dental scaling. And we found not only did the MELSCOR and endotoxin and TNF get better, this was also associated with improvement salivary and stool microbiota, and also associated with cognition and healthcare-related, health-related quality of life in these patients. We can also manipulate the microbiota by changing PPI use, fecal microbiota transplant, diet, and as well as modifying bacteria. PPIs have been used for a very long time. On the left-hand side, a nice study by Dr. Schnabel in alcohol-related injury showed that enterococcus goes up in patients who are drinking alcohol and in the mice who were exposed to alcohol. And in our study, we found that in compensated and decompensated stages, if you give people PPIs, their oral microbiota increases in the stool, but even in decompensated patients, withdrawing this, we take that away. And why is this important? Because oral microbiota could have consequences, including causing SBP, including causing a lot of inflammation in these patients. And some, as we're going to talk about, produce a urease that can split the urea into ammonia, which we know is injurious in hepatic encephalopathy and disease progression. So hepatic encephalopathy can, in patients with lactulose and rifaximin, can also be potentially treated with fecal microbiota transplant. We've done these two trials, which are phase one trials, and larger trials are underway. And we found they were safe, not only safe, but there was potentially a role of, this could actually be associated with changes in microbial function, as well as microbial outcomes in these patients. However, larger studies are needed, and this is not bereft of risks. There's a lot of risks, so therefore it must be done under controlled circumstances and not at home. Diet can influence diversity and outcomes, even in decompensated cirrhosis. A lot of these products, the studies are from the Western literature. And in this study that we had done in patients with USA versus Turkey, there was no difference in diversity between controls, compensated and decompensated patients in Turkey, compared to USA, where it followed their usual pattern, as you can see on the left-hand side. But what is important is the Turkish diet contains fermented milk products eaten every day. So this was associated with lower risk of decompensation and lower risk of hospitalization as well. Moreover, the Mexican diet had worse diversity than the U.S. diet, and higher hospitalization because of Prevot-Lacey, whereas the Brazilian diet, again, had better diversity and lower hospitalizations due to yogurt intake. So fermented milk products are very important, and if it's part of the diet, which means a sustained dose, rather than occasional probiotic use, is what could potentially be associated with all of these things. But what it tells us, as the prior two presentations said, diet is a very big determinant of microbiota composition, and potentially could be manipulated to improve outcomes in these patients. And this is the study from Brazil, where there was a lot more changes in the microbiota as well. So could we change certain bacteria to move this instead of putting this whole FMT in, which has multiple bacteria, but we don't exactly know what that is? In this nice initial experience with SynBiotic, with SynLogic, they had actually turned an E. coli nistle. Long story short, in healthy humans, as well as in mice, it worked beautifully in sucking up the ammonia and turning into arginine. However, and same again here, in humans with hepatic encephalopathy, it did not work. We do not know why, but what we want to figure out is we may need more than one bacteria, because as Dr. Defcota mentioned, the body has so many redundancies to take care of things that the other one bacteria may not. So therefore, the FMT or consortia may be the way to go at this point. What about other microbiota and other specific things that are present that can be changed in these patients? It is important to realize that we do not have microbiota, all microbiota, or all E. coli, or all enterococcus are not built the same. Some of them express virulence factors, some of them express antibiotic resistance genes, and some of them may actually be commensals that we actually need. It all is a balance between what your body's immunity is like, what your end stage, what your organ dysfunction is like, and how much of this bacteria is there. In this wonderful study, cytolysin positive E. enterococcus was negated by a phage in patients with alcohol-related hepatitis and then changed into this, and then in a mouse model, alcohol-related liver disease was reversed because of this. In patients with cirrhosis and as well as alcohol-related hepatitis, we know people with antibiotic resistance genes and virulence factors have a poor outcome, which in patients with cirrhosis, this burden is significantly higher in diabetes or dialysis patients, which tells you that the liver has such a major impact on the immune dysfunction and allowing these bacteria to grow. And this, it predicts hospitalizations and death independent of other complications. And moreover, when you give patients a fecal microbiota transplant in cirrhosis, it can actually reverse all of this. So antibiotic resistance genes is important, and this is another way of knowing which bugs are present rather than, which bugs are present with functional translation, rather than just the presence and absence of certain bacteria, because not all pathogens are, not all potential pathogens could be pathogenic, and not all commensals may be beneficial for you, given the context. We talked briefly about, we'll talk briefly about fungi and viruses, and fungi can work, worsen alcohol-related injury in murine models on the left-hand side, and gut fungi are related to bacterial diversity in patients with cirrhosis. And this actually can, this huge diversity can be changed after antibiotics to be replaced by candida, which is the number one cause of fungal infections in patients with cirrhosis. And so it is important for us to realize that indiscriminate antibiotic use is not a victimless crime. Not only does it worsen the antibiotic resistance genes, it also can improve, worsen the, increase the chances of patients getting fungal overgrowth and infections. Phages are very important, as we just talked about the phage changing of the E. coli, is of enterococcus that produces cytolysin, was important. In this study, what we found is the phages went down as well, and also as disease progressed, and with lactulose and rifaximin, it went down even further. On the top right, you see healthy controls and compensated cirrhosis. This is the correlation between the phages and the bacteria. You can see how dense it is in controls, goes down in compensated cirrhosis, and absolutely disappears when patients have decompensated cirrhosis. And this can be associated with hospitalizations, because this dense correlation between streptococcus phages, which are directed towards the production of urease, expressing microbiota, could be potentially one of the factors why these patients get hospitalized, because the high ammonia generation, and potentially also because of these, it could potentially be one way the rifaximin acts in these patients. So not just bacteria, it's also fungi and viruses. One thing which is very important, is for us not to forget that we are dealing with the liver disease, and the microbiota on their own may not be a factor, and this can, as Dr. Devkota pointed out, this can be potentially studied by germ-free animals. And this is what several studies have done, taken stool from healthy humans, non-drinking cirrhotic humans, non-drinking cirrhotic with HE, actively drinking non-cirrhotic human, and actively drinking cirrhotic human. But whatever combination you take, there is liver inflammation, and no cirrhosis. Only significant injury happens if you actually make the animal do something that would otherwise have caused that liver injury in the first place. So only if the mouse who was transplanted with the stool from an actively drinking non-cirrhotic human was fed alcohol, only then you had cirrhosis. So what it tells you, that it is complicit, it can amplify, but it may not necessarily cause this liver injury in the absence of the actual etiology of the liver disease. So long story short, if a patient is actually drinking and wants a quick fix, you do want to counsel them that we will fix, we will have potential to fix your microbiota, but please see if you can actually stop drinking again. We did a study on FMT in those patients, but that's right now, that's that the current story is, we want to make sure that the etiology is also fixed along with the microbiota. So what are the current means by which you can potentially improve outcomes? Regular dental cleaning and avoid periodontitis, withdraw unnecessary PPI use, carefully re-evaluate the need for antibiotics which could encourage fungal infections, and emphasis on fermented probiotic foods, including yogurts or fermented milk products. What are the current situations, the barriers, unmet needs, and future directions? Currently we know that this is associated progression of fibrosis, also mentioned beautifully by Dr. Loomba, associated with decompensation and gut-liver-brain axis change. We want to predict access to risk of decompensation and mortality. The barriers are, there are many dietary and lifestyle factors, there are differences in collection and storage of samples and access to timely, translatable, and cheap interpretation of results. Therefore, there are a lot of unmet needs with multiple etiologies, complications, and grading, and the future directions have to go beyond bacteria, we have to talk about microbial products, define microbial consortia, and focus on microbial metabolite and oral health. Thank you, and these are my acknowledgements. All right, well thank you guys, that was a wonderful presentation by each of you, I think you summarized it very, very well. We do have a few minutes for Q&A. Let's see, we do have two questions. The first one is, has anyone looked at whether ex vivo assays of whole blood, including immune cells, subjected to LPS stimulation are sensitized to produce higher levels of inflammatory cytokines in patients with NASH, cirrhosis, and decompensated cirrhosis? Not sure who's best to answer that, but I'll throw it out there. I'm not sure, you know, exactly of these experiments, but I think there is clear data in terms of, you know, at least one paper I remember in animal models by Akiseki, was in Nature Medicine, where they looked at gut-liver axis in terms of TLR4 activation and progression in NASH, as well as in alcoholic liver disease in animal models. There have been some studies, and actually a lot of studies by the groups in London, King's College on this. I only know about the cirrhosis and decompensated cirrhosis, and clearly there've been a lot more studies with this. In addition, there's been studies on the neutrophil function in these patients with and without alcohol use. So clearly this is something that is very important, the immune dysfunction locally as well as systemically. And two more questions here. What are your thoughts on pretreatment with antibiotics prior to FMT or MTT? Do you think FMT is superior to MTT for engraftment? These words are used very interchangeably, MTT, IMT, FMT, and, you know, obviously anything that you think about, you know, you take a large fermented milk product diet, that can also be an MTT. It's a microbial targeted therapy. Which microbes you're targeting is where the kind of folder. So it depends on the context, and please everyone else chime in. In our study, when we did it before cirrhosis, it actually decimated it and made things much worse. Because when we transplanted those tools, post-antibiotics with pre-FMT in mice, their brain inflammation, liver inflammation was much worse. So it did not set the stage. Unlike in C. difficile, where the context is very different, you really want as much of that acute disease process to form a nice foundation for the incoming stool and then stop giving them antibiotics in the future. Now in patients with end-stage liver disease, that's not possible because many of them are already on refaximin and SPP prophylaxis. So it all depends on context. Some studies in ulcerative colitis have said, yes, they have better outcomes with pre-antibiotic conditioning, but in some stages like end-stage liver disease, it may not be the case. So again, sorry. Depend. I agree with Jaz. I think on this one, just because I think the intent is slightly different in cirrhosis and its decompensation, you know, we are fixing the specific defects when they happen, say when we have an indication to use refaximin, when we have an indication to have SPP prophylaxis. But if we are talking about before that, it's probably the pattern and the diversity of the gut microbiome. And so having to reset that, probably we want different kinds of, I would say, healthy microbiome to set in as opposed to a randomly eliminating a big group of bacteria, which would further reduce the diversity and put the patient in danger. Yeah. I'm sorry. And, and the, the doc, Dr. Bernes just said that he wanted to capsulate it bacteria from like some of the products that are coming out now. So I don't, sorry. I, because C. difficile is such a universe away from liver disease or other chronic diseases that I'm not totally sure that those lessons can be immediately transferred to chronic diseases, which have low level, but constant issues going on rather than a sudden massive loss of antibiotic diversity. So those, everyone has to chart their own path. I'm sorry at this stage. You know, let, let me ask a question. Maybe this is directed at Rohit. So we, you showed the slide with Aldofermin and its effects on vianella, but just taking that a step further and maybe reversing that a bit and saying, is there, is there, do you see a day where we're modulating the microbiome to augment our pharmacotherapy for the treatment of NASH? And if so, how would you see that happening? Yeah, I think the answer would be in a specific bacterial metabolite. And so I think, so with vianella, what I think it's happening is vianella is extremely bile acid sensitive. And so whenever we perturb bile acid synthesis, and this is what FGF19 higher dose is doing, vianella takes over. But I think the, the true game changer would be if we can identify a group of species or bacteria in the stool that release a certain metabolite in a certain patient, and that's linked to, you know, progression of disease. And then we can identify that particular, you know, if that metabolite has a particular receptor in the liver and somehow we can block that and show that you can reverse their disease. It could also be, say, maybe in future, we develop antifungal prophylaxis as Jas just showed, or in alcoholic liver disease, where we have, say, there are type of enterococcus that are releasing a particular toxin. And then we can block that toxin by blocking the receptor for the uptake of that toxin in the hepatocyte. So I think those therapies are coming and especially in more advanced disease for prevention of decompensation. Got you. Got you. We're, we're a little bit over time. I do have one more question that I'd like to throw out to the group, and then we'll close the session. In terms of the local immune response in the liver, evidence suggests innate response is playing a role in NASH. Has anyone evaluated the role of adaptive immune response? And if so, what are the thoughts on how the microbiota affect this large compartment of the immune response? Yeah, I think there's plenty of evidence. The thing in immunology is everything is relevant. So if you start picking a particular area, I think there's a paper this month on the cover of hepatology, looking at B-cells and role of B-cells in steator hepatitis progression. Others have shown, you know, important role of CD8 positive T-cells, cytotoxic T-cells in progressional liver disease. We know that in NASH-related HCC, they're very important. They might even modulate the response to chemotherapy. So I do think that there is good data that both innate and adaptive immune responses are associated with NASH, NASH progression to cirrhosis, as well as to NASH-related HCC. Excellent. Well, you guys have done a fabulous job today. Just a special thank you to Suzanne Rohit and Jas for taking their time out of their busy schedules to not only synthesize their thoughts and put them down on their slides, but also to take the time to present this data in such a clear and concise manner. So I'll wrap up by saying a special thank you for attending this webinar that has been put on by the NASH Special Interest Group, and we look forward to seeing you guys again soon virtually at the AASLD Deliver meeting. So thank you for joining and have a wonderful rest of your day. Special thanks to you guys for participating in the lecture. So have a wonderful day, guys. Thank you, you too. Bye. Thank you. Bye.

webinar

microbiome

NAFLD

cirrhosis

gut microbiome

therapeutic approach

FMT

microorganisms

immune response

liver