Hello, my name is Mara Granella, and it is my privilege to welcome you to the 2021 Leon Schiff State of the Art Lecture. This lecture was established in honor of a giant in the field of hepatology, Dr. Leon Schiff, who was a founding member of the ASLD and the first president of ASLD. He received the association's Distinguished Service Award in 1981. To be invited to give this lecture is a tremendous honor. This year, we have the privilege of receiving it from Dr. Bernd Schnabel, Professor of Medicine and Director of the San Diego Digestive Diseases Research Center. Dr. Schnabel's research focus has been to elucidate the complex and multidirectional interactions between the gut microbiota and the liver. He has made several important observations in this regard and published them in journals of the highest impact. Bernd, we are very much looking forward to the insight we will gain from your lecture, entitled, Tiny Yet Powerful Microbes and Their Potential for Treatment of Liver Diseases. Hello and welcome. I would like to thank the ASLD and our ASLD president, Dr. Chang, to present the Leon Schiff state-of-the-art lecture, which is entitled, Tiny Yet Powerful Microbes and Their Potential for Treatment of Liver Diseases. I am Bernd Schnabel. I'm a Professor of Medicine and Director of the San Diego Digestive Diseases Research Center at the University of California in San Diego. These are my disclosures. I will predominantly focus today on fatty liver diseases, particularly about alcohol-associated liver disease and non-alcoholic fatty liver disease. As you know, most of the patients with chronic alcohol use disorder, they will eventually develop hepatic steatosis. However, the minority of these patients will progress to more advanced liver diseases, which includes steatohepatitis, fibrosis, cirrhosis, hepatocellular carcinoma, and alcoholic hepatitis. The same is true for non-alcoholic fatty liver disease. Here we have about 30 percent of the US population have a hepatic steatosis. However, the minority will ever progress to more advanced liver disease, which include non-alcoholic steatohepatitis, fibrosis, and cirrhosis. At this point, I just would like to remind you about our ability to get histology images from the liver, and this was enabled by Dr. Leon Schiff. He was one of the first physicians to use percutaneous liver biopsy for diagnosis, which is a standard technique today. Coming back now, why the minority of the patients with fatty liver disease ever progressed to more advanced liver disease, and what are the risk factors? The risk factors for alcohol-associated liver disease to progress to more advanced liver disease stages include comorbidities in particular concomitant liver diseases, such as hepatitis C infection. The drinking pattern plays a very important role, dietary factors, but also genetic factors. The risk factors for progression of non-alcoholic fatty liver diseases are very similar. They include comorbidities like type 2 diabetes, insulin resistance, dyslipidemia, and obesity. They include genetic factors, dietary factors, and sedentary lifestyle. Now, I would like to introduce another risk factor, which I believe is important for both diseases, and this is the intestinal microbiota. There is no doubt that many of these risk factors that I have described up here will affect the composition of the intestinal microbiota. To introduce the gut microbiota to you, these are the major components of the human microbiota and they include bacteria, fungi, and viruses. If we look at the absolute number of these microbes in our body, viruses, and in particularly bacteriophages, are outnumbering the bacteria and the fungi in our body. If we look at the total number of microbes and compare them to our human cells, we have about an equal number. However, if we look at the genes, the microbial genes are outnumbering the human genes that we have in our body. Why is the gut microbiota so important for the progression of fatty liver diseases? One of the reasons is that we have this delicate relationship between the gut and the liver. As you know, most of the venous blood is drained via the portal vein to the liver, so most of the dietary components like carbohydrates, proteins, and fat, they reach the liver as a first organ in our body. Many of them are metabolized in the parasites. Why the portal vein, secondary bile acids, other dietary metabolites such as trimethylamine or microbial-associated molecular patterns such as LPS or beta-glucan from fungi can also reach the liver via the portal vein. Now, this is not a unidirectional communication between the gut and the liver. The liver talks back directly to the gut via the biliary tract. Hepatocytes secrete bile acids and they maintain a constant bile flow. In the bile, you have primary bile acids, you have a large amount of immunoglobulin A, and you have also antimicrobial molecules transported from the liver to the gut. We have a third communication route between the gut and the liver, and this is made possible by the systemic circulation. Many of the inflammatory mediators, but also metabolites such VLDL, acetaldehyde, and TMAO, which is trimethylamine and oxide, are transported via the systemic circulation to the gut and they can now affect the intestinal microbiota. In the following slides, I would like to introduce you to changes that occur in the composition of the gut microbiota in patients with fatty liver diseases. I will start with gut viromes. How does a gut virome change in patients with fatty liver disease? We have recently published this. What we did is we extracted viral particles from the stool, from patients with fatty liver disease, and then we did shotgun sequencing on these viral particles. We had a patient cohort of about 80 patients with alcoholic hepatitis. We had a control group of alcohol use disorder patients with mostly early stages of liver disease, and then we have a non-alcoholic control group. One of the most significant changes that we found are an increase in the mammalian viruses in patients with alcoholic hepatitis as compared with the alcohol use disorder patients and as compared with non-alcoholic controls. In this graph, one column represents one patient, and you can very nicely see that the gut virome is very distinct in patients with alcoholic hepatitis as compared with patients with alcohol use disorder and as compared with non-alcoholic controls. Interestingly, the viral diversity now was highest in the patients with alcoholic hepatitis and lower in the alcohol use disorder patients and in the non-alcoholic controls. Now coming back to the mammalian viruses and asking what are the main differences. What we found is that in about 50 percent of the alcoholic hepatitis patients and alcohol use disorder patients, we detected parvoviridia in the stool, and this was much less in the non-alcoholic controls. In addition, we also found that in about 15 percent of the patients with alcoholic hepatitis, we had herpesviridia present, and this were completely absent in patients with alcohol use disorder and in non-alcoholic controls. When we looked at the MELD score from patients with detectable herpesviridia, you can see that this MELD score was higher as compared to the alcoholic hepatitis patients where we did not find any herpesviridia detectable in the stool. This might be a very important change in the gut virome and further studies are obviously needed to evaluate whether these are mechanistically related to progression of alcohol-associated liver disease. I would like to contrast this now to the gut virome in patients with non-alcoholic fatty liver disease. We did the same technique. We isolated viral-like particles from the stool from a patient cohort with non-alcoholic fatty liver disease. We had about 100 patients and we did shotgun sequencing. In this case, we did not find a significant increase in the mammalian viruses, either in early-stage liver disease with NAFLD activity score 0-4 or NAFLD activity score 5-8 or liver cirrhosis. However, what we also found is that we can discriminate patients with late-stage of liver disease, so NAS 5-8 and liver cirrhosis from early-stage liver disease, and this was predominantly defined by changes in the bacteriophages. When we added the viral diversity to clinical parameters on the left side, AST and H, on the right side, AST, H, and platelets, this improved the non-invasive prediction of histological disease severity in patients with non-alcoholic fatty liver disease using a random forest modeling. On the left side, we have the outcome was NAS 5-8 or liver cirrhosis. On the right side, we have F2-F4 liver fibrosis. Switching over now to the gut microbiome. How does a gut fungi microbiome change in patients with fatty liver disease? I will start again with alcohol-associated liver disease. The one very important part from the fungi gut microbiomes is that we observed an increase in the absolute amount of fungi present in the stool of patients with alcoholic hepatitis. We did very simple culturing of fungi on YPT plates from stool from these patient cords. You can see here that the colony forming units were significantly increased in patients with alcoholic hepatitis. When we did single colony PCR from these fungal colonies, we were also able to show that Canada albicans was one of the yeasts and fungi that were increased in patients with alcoholic hepatitis. Suggesting that we have fungi overgrowth in these patients with alcoholic hepatitis. We obviously also used non-invasive tests and sequencing methods to determine the composition of the gut microbiome. In this case, we used ITS sequencing. On the left column, you can see that we have a very nice diversity and about 25 percent of the fungi in these non-alcoholic controls were Canada species shown in red. Canada essentially takes over in patients with alcohol use disorder and in patients with alcoholic hepatitis. This also was associated with a decrease in the fungal diversity, which was similar in patients with alcoholic use disorder and in patients with alcoholic hepatitis. Suggesting that the changes in the gut microbiome are secondary, predominantly to chronic alcohol abuse, and are probably not so much dependent on the liver disease stage of the patient. We also analyzed the gut fungome in patients with non-alcoholic fatty liver disease. We similarly did ITS sequencing. Interestingly, what we did not find is a significant difference in the beta diversity shown here on this PCOA plot between obese NAFL and obese NASH patients. However, when we looked at the non-obese NAFL patients and NASH patients, we found a significant difference. This was characterized by a significant increase in Canada albicans and mucor, especially NASH patients. In NAFL patients, we found that we have this Saccharomyces sotalis were present in the patients with NAFL. We also compared the fungal dysbiosis in patients with cirrhosis, secondary to NAFLD, or alcohol-associated liver disease. We were able to show that the beta diversity is significantly different in patients with cirrhosis, secondary to NAFLD versus alcohol use disorder. Compositional changes in the gut microbiome were predominantly caused by an increase in mucor and penicillium in the patient with NAFLD versus patients with alcohol-associated liver disease were predominantly characterized by a increase in Canada species and Blumeria species. Finally, I would like to switch over to the bacterial microbiota where we have obviously the most ongoing research. I would like to show you how does the bacterial microbiota change in patients with fatty liver disease. In this case, I just want to show you a summary slides. We start off with the healthy gut microbiome. In patients with alcohol use disorder, we have a dysbiosis that is characterized by an increase in many of these potentially bad bacteria, which we call pathobionts, which include Prevotella, Vilonella, and Streptococci. We have also an increase in the gram-negative proteobacteria, while we have a decrease in some of the good bacteria such as Akkermansia. Similarly, in cirrhotic patients, we have an increase in these pathobionts while we see a decrease in many of the short-chain fatty acid producing bacteria in particularly Lachnospiriceae, Ruminococciae, and Fecalibacterium prausnitzii. We also know in patients with non-alcoholic fatty liver disease, we can very nicely discriminate using a random forest model patients with early liver disease stage from NAFLD patients with late fibrosis. This random forest model not only contained bacterial microbiota, but there was also some clinical parameters and the bacterial diversity factored in. What are the compositional changes that occur in the bacterial microbiota in patients with NAFLD and advanced fibrosis? You can see that we have an increase in Escherichia coli, that we have an increase in Acetaminococcus species and Klebsiella pneumoniae, while we have a decrease in Eubacteria, a decrease in Fecalibacterium prausnitzii, and we also see a decrease in some of these Doria species. Now, I don't have time to talk too much how about the gut microbiome now contributes to liver disease. I have just shown you here one summary cartoon, which we have also recently published. On the top, you can see the gut with the gut lumen and the portal vein, we have drawn here like a funnel, so everything which comes from the gut will eventually reach the liver by the portal vein. What are factors from the gut microbiome that affect progression of liver disease? We have certain metabolites, I will show you which ones in a second. We have microbial associated molecular patterns like LPS and beta-glucan, we know that viable bacteria translocate via the gut barrier to the liver, and this might be more important in alcohol-associated liver disease. We know that short-chain fatty acids are decreased, likely in both diseases, alcohol-associated liver disease and NAFLD. We know that indoles, which are tryptophan metabolites, are decreased in particularly in alcohol-associated liver disease patients. We know that in both diseases, total bile acids and secondary bile acids are increased. Toxins play a role in progression of especially alcohol-associated liver disease. Finally, trimethylamine is increased in patients with metabolic syndrome and NAFLD and NASH. In the second part of my presentation, I would like to show you now how the gut microbiome can be used as target for therapy to treat liver diseases. And I divided this part into two sections, and initially I want to show you how we can use untargeted microbiome-centered therapies for liver disease. So what are untargeted therapies for liver disease. So what are untargeted therapies for liver disease? We can use bugs as drugs, so we can essentially replace a dysregulated microbiota by using fecal microbiota transplantation and restore intestinal homeostasis. We can add to this dysregulated microbiota some probiotics, which are life-beneficial bacteria, and then restore homeostasis. We can also use antibiotics or antifungus to eliminate some of these pathobionts, again, to restore gut homeostasis. These are largely untargeted therapies. And in the following, I would like to show you some of the clinical trials and evidence that at least some of these interventions work in patients with chronic liver disease. I will start with fecal microbiota transplantation in alcohol use disorder patients and in alcohol-associated hepatitis. One small clinical trial with 20 patients due to cirrhosis and alcohol use disorder. This clinical trial used a FMT enema from a rationally designed donor. This was a phase one safety trial. And at day 15, there was a significantly reduced alcohol craving, and there was also improved cognition and psychosocial quality of life. In terms of safety profile, there were less severe adverse events in the FMT group as compared with the placebo group. For the alcohol-associated hepatitis and alcoholic hepatitis, we have two clinical trials coming out from India. One I want to show you in more detail. This was a small clinical trial with NA8 patients, and these were patients who were steroid ineligible. They received fecal microbiota transplantation daily for eight days via a nasotodial tube. And you can see that in the FMT group, there was a significantly increased survival of about 88% versus in the historical control group, there was a survival of about 33%. Obviously, this trial needs to be repeated. This needs to be better controlled and prospectively evaluated, but at least there seems to be some benefit for FMT in patients with alcohol-associated hepatitis. What about fetal microbiota transplantation in patients with a metabolic syndrome and NA3D? I'm showing you here six trials, fairly small trials in patients with metabolic syndrome, obesity, hepatic steatosis, and NA3D. Except for the first trial, which was published in 2012 in Gastroenterology, which showed a significantly improved peripheral insulin sensitivity six weeks after FMT, all of the other trials were essentially negative, although there were no serious adverse events recorded. So these are somewhat disappointing results for using FMT in patients with NA3D. We have several clinical trials using FMT in cirrhosis and hepatic encephalopathy. The first trial is a retrospective case series using single FMT via colonoscopy. This trial only showed that there was a sustained clinical response in six out of 10 patients, and there were severe adverse events, in particularly one patient died within two months. However, in two very well-controlled clinical trials enrolling cirrhotic patients and recurrent hepatic encephalopathy, the first trial showed that significantly less patients receiving FMT develop hepatic encephalopathy, and the second trial even demonstrated there might be a potentially effective treatment effect in preventing long-term recurrence of hepatic encephalopathy. In terms of safety, both trials were safe. In fact, there were less severe adverse events in the FMT group compared with the standard of care therapy. So very encouraging results for FMT in patients with cirrhosis and hepatic encephalopathy. There are clinical challenges associated with FMT. One of these includes that we still do not understand how FMTs work. Are they mediated by bacterial changes? Is the efficacy due to changes in the bacteriophages, or is this mediated via metabolites that are transferred? In addition, chronic liver diseases are complex and multifactorial in nature, and this might be one of the causes why FMT was not very successful in NASH and NAFLD patients. So this might require, in fact, a longer and chronic treatment for these patient cohorts. The best route of FMT administration is currently also unknown. Can we use FMT capsules? Should we use a nasogastric or nasointestinal tube? Should we use an upper endoscopy? Should we use a colonoscopy? Or can we use FMT animals? It is also not known what the best content of donor stool is. How should we select the best donor? Should we add some beneficial metabolites, such as short-term fatty acids, to the donor stool? The long-term outcomes are unknown, and we know that we have some short-term severe adverse events, which was recently published when drug-resistant E. coli were transmitted by fecal microbiota transplantation, and the FDA now implemented a more rigorous screening of donor stool for these pathobionts. I want to switch over now to probiotics, another untargeted microbiome-centered therapy for liver diseases. I would highlight one abstract that will be presented during this ASLD meeting, and this shows an efficacy of six-month lactobacillus rhamnosus GG therapy in reducing heavy drinking in patients with moderate alcohol-associated hepatitis. These were 38 patients with alcohol use disorder and moderate alcoholic hepatitis. These were randomized to either receiving the probiotic, the LGG, or the placebo plus standard of care, and there was a very encouraging result shown down here that the probiotic-treated patients showed a significant reduction in drinking at six months. However, there was a considerable dropout rate, so nine out of 17 patients in the placebo-treated group and six out of 21 patients in the probiotic-treated group did not continue with medical management during six months. But overall, there seems to be some benefit in patients with moderate alcohol-associated hepatitis. Switching to probiotics and NAFLD, we have now several studies published, and I'm just presenting you two systematic reviews with meta-analysis. Both reviews conclude that probiotics appear to be beneficial for a patient with NAFLD. It looks like that ALT and BMI are improved. In addition, both studies agree that other liver function tests are improved with probiotics. Total cholesterol is reduced, and the insulin resistance is also reduced upon probiotic treatment. The problem with many of these probiotic trials is that they're using different probiotics, sometimes combined with other medications such as vitamins, and they're using a different treatment length. But this needs to be in the future, so there needs to be some homogenization of the clinical trial design. Some of these clinical trials need to be repeated and confirmed in independent studies. How about probiotics in patients with cirrhosis and hepatic encephalopathy? We have multiple studies published. I'm showing you here a contrarian database systematic review, and the conclusion of this review was also that probiotics improve the symptoms of hepatic encephalopathy in patients with decompensated cirrhosis. In the second part of this microbiome-centered therapies, I want to now focus and switch to more precision, so the newer treatment strategies that we have now available to use the microbiome as target for the treatment of liver diseases. So microbiome studies might identify new pathways and factors that contribute to the progression of fatty liver diseases, and this might be used to assign patients to subgroups and identify those associated with specific alterations to the gut microbiota. And the integration of this knowledge now might lead to increased and improved diagnostic accuracy, prognostic accuracy, therapeutic efficacy, and most importantly, to a reduction of side effects. So how can we now use this personalized treatment approach for our clinical trials? And I want to give you one example. This is about intestinal permeability and a disrupted gut barrier function in patients with NAFLD. Increased intestinal permeability is a very popular topic, which is discussed in many reviews and evaluated in many preclinical studies. However, if you look at this meta-analysis that was published by Dr. Luther, our ACT president, Dr. Chung, and Dr. Patel, you can see that only about half of these patients with NASH have, in fact, increased intestinal permeability. And the same is true for patients with alcohol-associated liver disease. You can see that only about half of the patients with alcohol-associated liver disease have increased intestinal permeability as compared with the controls, and the other half have a very similar intestinal permeability as compared with the non-alcoholic control patients. So suggesting that it's very important that before a clinical trial is started, that you stratify your patients according to the microbiome parameters that are targeted for therapy. So in the following, I would like to show you some of the newer treatments that we can now use as precision microbiome center therapy. One of these treatment approaches is to use engineered bacteria. And we are mostly engineering probiotic bacteria, so which already have some benefit for the host. And these engineered bacteria, they can either produce beneficial metabolites for the host, or they can metabolize or eliminate some of the toxins. And I will show you in the next slide that one of these targets could be ammonia. Two independent studies showed that engineered bacteria are able to reduce intestinal ammonia. One of these studies colonized mice with urease-negative bacteria. And when these colonized mice were then subjected to a chronic liver disease model, these mice were then protected from hyperammonia. In a second study, a company engineered bacteria. In this case, they used E. coli and Nestle as the isogenic probiotic bacteria. And they engineered these bacteria to metabolize ammonia into L-Arginine. When mice were colonized with these bacteria and subjected to a chronic liver fibrosis model, you can see that with the engineered bacteria, this ZunB1020, there was a reduced systemic ammonia level as compared with the vehicle-treated mice or with the mice that were just treated with the E. coli and Nestle alone. However, in the press release, which is published on ZunLogic's website, an interim analysis of a randomized double-blind placebo-controlled Phase 1b2a study with ZunB1020 in 23 patients with cirrhosis and elevated blood ammonia did not show any evidence of blood ammonia lowering or changes in other exploratory endpoints relative to placebo. So this is obviously very disappointing. And we are waiting for this trial to be published to know more about why this was a failure. In the next section, I would like to introduce you to another concept, how we can use microbiome-centered therapy for treatment of chronic liver diseases. In this case, we can use drugs from bugs. So metabolites that are secreted or metabolized by bacteria. I will not focus on bile acids. As you know, primary bile acids are metabolized by gut bacteria into secondary bile acids. And some of these could be used for treatment of diseases. I will show you two clinical trials that used short-chain fatty acids in patients with the metabolic syndrome. Short-chain fatty acids are an end product of fermentation processes by gut microbiota. We had one larger clinical trial in with 60 overweight subjects. They were randomized and they either received inulin propionate. So in this case, propionate is a short-chain fatty acid versus inulin control for 24 weeks. And there was a significantly reduced weight gain, intra-abdominal adipose tissue distribution and hepatic lipid content. There were no increased adverse event while a smaller clinical trial which used butyrate for only four weeks did not show any benefits. So the first trial is obviously very encouraging. We need to see whether these results can be reproduced in an independent clinical trial. Finally, I would like to introduce you to a third concept how we can use microbiome-centered therapy for treatment of liver diseases. And this is currently still in a preclinical stage. We can use bacteriophages. Bacteriophages are small viruses that specifically recognize bacteria. They can inject their DNA into bacteria. They can replicate inside and then they can lyse the bacterium and kill it. We found cytolysine, which is produced by Enterococcus faecalis in about 30% of the patients with alcoholic hepatitis, but we could not find it in stool from non-alcoholic controls and in only one patient with alcohol use disorder. When we associated cytolysine positivity with disease outcome, in this case with mortality, you can see that most of the cytolysine negative patients survived while close to 90% of the cytolysine patients with alcoholic hepatitis who were cytolysine positive died within 180 days. In a multivariate Cox regression analysis, the cytolysine showed a hazard ratio of 14 to die within 90 days if the cytolysine is positive in the stool. So we used a bacteriophage approach to specifically eliminate the cytolysine positive Enterococcus faecalis in preclinical mouse model of ethanol-induced liver disease. You can see a bacteriophage with a head and a tail, they dock onto the Enterococcus faecalis which produces cytolysine. And then once the bacteriophage is inside the bacterium replicates, it then eventually eliminates and kills the bacterium. So when we use this in chromobiotic mice, so germ-free mice that have been colonized with stool from cytolysine positive alcoholic hepatitis patients, and when these mice were treated with ethanol and gavaged with bacteriophages against the cytolysine positive Enterococcus faecalis, you can see that we can improve liver disease shown here by ALT levels, so liver injury as compared with the mice that were gavaged with a control bacteriophage. We can also reduce hepatic triglycerides, and most importantly, we can also reduce the cytolysine that is present in the liver. So these are my key takeaways. I hopefully have been able to convince you that liver disease is associated with changes in the gut microbiota, which includes not only bacteria, but also fungi and viruses. Intestinal dysbiosis represents a very attractive target for therapy. Microbiome-centered therapies can be targeted or untargeted, but both require personalized approaches. Again, I would like to thank the ASLD and our ASLD president, Dr. Chung, to provide me the honor to present the Leon Schiff State of the Art Lecture. I would like to thank you for Zooming in today and for listening to my presentation. Bye-bye.

Dr. Bernd Schnabel

2021 Leon Schiff State of the Art Lecture

gut microbiota

liver diseases

alcohol-associated liver disease

non-alcoholic fatty liver disease

personalized microbiome-centered therapies