Hello, let me welcome you to the ACLD Japan Society of Hepatology Joint Symposium on Lean Nalphaldehyde. I'm Zubair Nassi, I'm the President of Innova Medicine as well as a Professor and Chairman of the Department of Medicine at the Innova Fairfax Medical Campus. I'd like to also take this opportunity to welcome you on behalf of my co-chair Professor Tetsuo Takehara, who is the Professor of Medicine in the Department of Gastroenterology and Hepatology at Osaka University Graduate School of Medicine in Japan. Now before we get started, let me just spend a couple of minutes reviewing some of the issues on non-alcoholic fatty liver disease. The global prevalence of Nalphaldehyde in general is about 25%, with the highest prevalence being from the Middle East and the lowest prevalence in Africa. The prevalence is of course higher in type 2 diabetics. Over half of diabetics have Nalphaldehyde and about 7%, 7.5% of children have Nalphaldehyde. The prevalence of the progressive form of fatty liver disease, which is called non-alcoholic fatty liver disease, and general population is estimated to be about 1.5% to 6.5%. Again among type 2 diabetics, this prevalence is about 37%. When you look at patients with Nalphaldehyde, the most common cause of death is cardiovascular death. In fact, a small percentage of patients with Nalphaldehyde will develop NASH, as we just noted, and maybe about 15% of patients with NASH can develop cirrhosis and be at risk for complication of cirrhosis, including hepatocellular carcinoma. It's important to remember that risk factors for non-alcoholic fatty liver disease are components of metabolic syndrome. In this context, there's obesity, type 2 diabetes, hypertension, and dyslipidemia. However, there is a group of patients that will not be considered obese and they have non-alcoholic fatty liver disease. The prevalence of lean Nalphaldehyde can range somewhere between 5% to 30% in different parts of the world, and this is, of course, this is dependent on how lean was defined. In most of these cases, lean has been defined by body mass index. In the United States, the prevalence of lean Nalphaldehyde was somewhere around 7%. This is from NHANES data we published a few years ago. Hopefully in the next hour or so, we'll see much more recent data about the prevalence of lean Nalphaldehyde in other parts of the world. What's important is to recognize that among Nalphaldehyde, body composition drives mortality. This is a study that we recently published from NHANES 3, and NHANES 3 was basically the data was collected between 1988 to 1994. We had over 20 years of follow-up, in fact, 22 years of follow-up with mortality data available. And the study patients were divided to Nalphaldehyde patients, as well as those who were healthy or healthy controls in terms of their liver disease. And then these groups were further subdivided based on body composition to lean BMI with normal waist circumference or lean BMI with obese waist circumference. And as you can see in the top panel, among Nalphaldehyde patients who are now considered to be obese by waist circumference, the prevalence of hypertension, hyperlipidemia, diabetes, cellular reactivity, and metabolic abnormality is substantially higher when you have waist circumference that define these patients as obese. What's actually even more important is that when you look at all-cause mortality, that among lean Nalphaldehyde patients, presence of visceral obesity was an independent predictor of mortality. And this will actually tell us that when you actually study lean Nalphaldehyde, it's much more important to pay attention to visceral adiposity rather than total body fat that's usually captured by body mass index. To address this issue of lean Nalphaldehyde, Japan Society of Hepatology and ASLD put together a joint symposium to emphasize that Nalphaldehyde is really not very rare among lean individuals. On the other hand, given that these individuals don't have traditional risk factors, metabolic risk factors, or they're not considered to be obese, that this entity remains under-recognized. So the specific goals of this program is to acquire new knowledge about the prevalence of lean NASH in general population, to understand the impact of metabolic abnormalities for hepatic fibrosis in lean patients with NASH, and to understand association of genetic background environmental factors and development of liver fibrosis in lean patients with NASH. Therefore, in the next hour, we have four superb speakers to review some of these issues for us. So next will be Dr. Kogisio, who's going to review the characteristic of NASH among lean patients in Japan. That will be followed by Dr. Allen's talk on NASH among lean patients in the United States. And then Professor Kawaguchi from Japan is going to review a data mining-based metabolic risk abnormality for lean patients with NASH from Japan. And finally, Dr. Vincent Wong is going to review the gene environment interactions in NASH patients who are considered to be lean. At the end, my co-chair, Professor Takehara, is going to review some of the data that is reviewed in this program and will make some closing remarks. Now, without any further ado, let me start the program. And again, thank you for participation and for your attention. Thank you. Thank you very much for giving me the opportunity to present our work. Today, I will talk about the characteristics of non-alcoholic steatohepatitis among lean patients in Japan. Non-alcoholic fatty liver disease, naphrodene, is the most prevalent cause of chronic liver disease, which affects 29.7% of Japanese individuals with chronic liver disease. Naphrodene is a lifestyle-related disease. This report shows the naphrodene prevalence among lifestyle-related disease in Japan. Naphrodene affects one-quarter of the adult population. The naphrodene prevalence has increased due to complications from other diseases. One-half of patients with hypertension and metabolic syndrome exist naphrodene, whereas more than 70% of patients with type 2 diabetes exist naphrodene. Among patients with obesity, more than 19% exist naphrodene. The global prevalence of non-obese naphrodene in the naphrodene population is shown here. Overall, about 40% of the global naphrodene population was classified as non-obese naphrodene, and about 19% was lean naphrodene. Lean naphrodene is becoming an important problem, especially in Asia. However, the definition of lean naphrodene is varied. The definition of lean naphrodene includes body mass index, BMI less than 25 among Caucasian individuals, whereas the proposed BMI is less than 23 or 22 among Asian individuals. Hegostream reported that 19% of patients in their study were diagnosed with lean non-alcoholic steatohepatitis NASH. Their overall mortality rate was 38%, while the liver-related mortality rate was 8%. These rates were higher in overweight patients or those with obesity. Proposed risk factors for lean NASH include alcohol intake, weight change, dietary change, and genetic background. A genome-wide association study identified a single nucleotide polymorphism, SNP, in PNPL3, which comprises a disease-sensitive gene in patients with naphrodene. In the general Japanese population, 21% of individuals had the GG risk allele in PNPL3. This risk allele was observed in 37% of patients with naphrodene and in 59% of patients with NASH hepacellular carcinoma, HCC. In contrast, it was only observed in 12% of individuals in general population in United States or Europe. Furthermore, it was observed in 16% of patients in naphrodene and in 28% of patients with NASH HCC. All of these rates were lower than in the Japanese population. Therefore, PNPL3 might influence the development of naphrodene in Japanese individuals. We previously investigated the frequencies of SNPs associated with liver-related disease risk and compared these between patients with liver cirrhosis and individuals without cirrhosis. We extracted genomic DNA and evaluated SNPs associated with glucose and fat metabolism. ADRB3 is a known risk factor from obesity. This gene was not significantly different between patients with liver cirrhosis and individuals without cirrhosis. KCNQ1 has been associated with diabetes mellitus. The frequency of the KCNQ1 PT genotype tended to increase in patients with liver cirrhosis. Furthermore, the frequencies of PNPL3 GG and GC genotypes were significantly increased in patients with liver cirrhosis. The aim of this study was to determine the characteristics of re-NAPHLD in the Japanese population, as well as the patient profile, comorbid lifestyle-related diseases, and genetic backgrounds. Here, I describe the method of this study. In total, 762 patients with biopsy-confirmed and clinically diagnosed NAPHLD were enrolled from among patients who attended our hospital for treatment. We first investigated the frequency of re-NAPHLD assessed by Bioelectrical Impedance Analysis, BIA, and computed tomography CT among patients stratified according to body composition using BMI categories of less than 20Y, 25 to 30, and greater than 30. We then stratified BMI into additional categories and examined the complications of lifestyle-related diseases, biochemical results, and genetic profiles. The survival rates and incidence of hepatocellular carcinoma, cardiovascular events, and diabetes were estimated by Kaplan-Meier analysis according to BMI categories. Finally, gut microbiome was evaluated in NAPHLD and Healthy Control. We compared the gut microbiome between patients with NAPHLD and Healthy Control and patients with mild fibrosis and advanced fibrosis in NAPHLD. The diversity of gut microbiome was evaluated by Shannon-Weaver Index. Overall, 35% of patients had BMI less than 25, 39% had BMI 25 to 30, and 26% had BMI greater than 30. Re-NAPHLD in patients with BMI less than 25 were observed more frequently than in previous studies. In our study of 762 patients with NAPHLD, approximately 25% of men showed re-NAPHLD. In women, nearly 40% of patients with NAPHLD showed re-NAPHLD. Furthermore, the prevalence of re-NAPHLD increased with age in both sexes. Regarding the fibrosis progression, approximately 30% of young patients with NAPHLD showed advanced fibrosis, while 60% of elder patients with NAPHLD showed advanced fibrosis. The findings were consistent between sexes. Among older patients, about 40% were diagnosed with NAPHLD by ultrasound. Complications of metaric syndrome did not significantly differ between patients with NAPHLD and those without NAPHLD. The associations of metaric complications with NAPHLD were not robust in very old patients. Sarcopenia is presumed to be an independent risk factor for the onset of NAPHLD. Aging reflects a gradual decline in the production of sex hormones, growth hormone, and insulin-like growth factor 1. The loss of regenerative capacity includes deprecated senescence, defective ophthalmophasy, or reduced hepatic-free radical scavenging. This loss involves enhanced oxidative stress, mitochondrial dysfunction, DNA damage, and terminal shortening at the end of chromosomes. Body composition assessment by BIA showed that the skeletal muscle and fat amount were significantly lower in patients with BMI less than 25. Both visceral and subcutaneous fat area were reduced, estimating by CT scan. BMI less than 23 was observed in 13% of patients with NAPHLD. Among patients with BMI less than 23, 60% were women. The age distribution skewed older in patients with lean NAPHLD. Complications of diabetes and hypertension were not significantly different according to BMI. However, dyslipidemia was observed significantly less frequently in patients with low BMI. Fibrosis status was not significantly different according to BMI. Mental disorders were more frequent in patients with obesity. This table shows biochemical results according to BMI. Serum levels of albumin, ALT, as well as bladder count were significantly lower in patients with lean NAPHLD. Immunoreactive insulin level, homeostatic model assessment of insulin resistance score, triglycerides level, and total cholesterol level were also lower in patients with lean NAPHLD. These findings indicated that the influence of metabolic commodities might be small. Analysis of genetic backgrounds revealed that ADRB3 and KCNQ1 polymorphisms were not significantly different according to BMI. Although PNPR3 was associated with disease development, it was not significantly different according to BMI. Survival analysis showed that the patients with lean NAPHLD tended to have a slightly lower survival rate, but the difference was not statistically significant. Newborn set ADC was observed in 13 cases, but the incidence did not differ according to BMI. In total, 26 patients had extrahepatic malignancy, including 12 patients with gastrointestinal cancer, and 8 patients with gynecological cancer. The incidence did not differ according to BMI. Cardiovascular events were observed in 20 patients, including seven patients with myocardial infarction and 15 patients with cerebral infarction. These events were not more frequent in patients with Lin-Nafu-D. Newer said diabetes mellitus was observed in 44 patients and it tended to be less frequent in patients with Lin-Nafu-D. The distribution of gut microbiota is a possible contributing factor in the development of Lin-Nafu-D. A previous report indicated that NASH fibrosis severity was associated with changes in gut microbiota. Unconjugated and conjugated biacid levels in stool were stratified by fibrosis severity and obesity status. Two biacid were elevated, especially in non-obese subjects with significant fibrosis. Results of bacteria and the metabolites were associated with liver fibrosis, especially in non-obese. In our study, we compared the gut microbiome between patients with Nafu-D and healthy controls. We found that the proportion of farmacutis was reduced and the proportion of bacteroides was enhanced in patients with Nafu-D. Evaluation of diversity using the Shannon-Weaver Index indicated a significant difference between patients with Nafu-D and healthy controls. When comparing patients with mild fibrosis and those with advanced fibrosis, the differences in the proportion of farmacutis and bacteroides were more prominent. Notably, the proportion of bacteroides was further increased in patients with advanced fibrosis. Evaluation of diversity using the Shannon-Weaver Index also indicated a significant difference between patients with mild fibrosis and those with advanced fibrosis. This analysis included a small number of patients. Thus, we plan to perform a subgroup analysis according to BMI, following accumulation of additional patients. Differences in metabolic adaptation between patients with rene and non-rene Nafu-D were proposed. In rene Nafu-D, bile acid levels were increased and altered by gut microbiota. Rene Nafu-D might have better metabolic and liver history profiles, showing obesity resistance. In summary, rene Nafu-D was most frequently observed in women and older patients. Both bislel, subcutaneous fat, and skeletal macella amounts were lower in patients with rene Nafu-D. Moreover, the incidence of dysrepedemia and mental disorder was significantly lower in patients with rene Nafu-D. Genetic background did not significantly differ according to BMI. Although there were no significant differences in mortality rate and complications according to Nafu-D status, new-onset diabetes tended to be less frequent in patients with Nafu-D. This biosis might be associated with development of rene Nafu-D. In conclusion, liver function should be carefully evaluated in patients with rene Nafu-D, especially among older people in Asia. I thank my colleagues who contributed to this work, especially Dr. Maki Tobari, who has done analysis of rene Nafu-D, and gut microbiota was evaluated by Dr. Makiko Taniai. Supervisors are Dr. Etsuko Hashimoto and Professor Katsutoshi Tokushige, and I appreciate for their outstanding advice. Thank you very much for your attention. Good morning. I would like to start by thanking the program chairs for the invitation to present at this joint symposium. I am Alina Allen from Mayo Clinic, Rochester, Minnesota. My talk will provide an overview of the epidemiology and the natural history of rene Nafu-D in the United States. My disclosures are listed here. Here's an overview of the presentation. We will discuss how rene Nafu-D is defined in the United States, its prevalence, the characteristics of rene versus non-rene Nafu-D, and we will conclude with a comparison of outcomes between lean and obese Nafu-D. Given that lean Nafu-D is categorized based on BMI, the cutoffs used to define lean, overweight, and obese Nafu-D follow the World Health Organization definition for non-Asian race. In the most strict criteria, lean Nafu-D is defined as a BMI less than 25, overweight Nafu-D for those with BMI between 25 and 29.9, and obese Nafu-D in those with BMI above 30. Several studies have grouped lean and overweight categories into a non-obese group studied in reference to the obese group. It is important to note here that BMI is used due to a more widespread availability in conventional databases, but it is a poor marker of body fat distribution and metabolic risk. Those with lean Nafu-D can have and likely do have an abnormal amount of visceral adiposity despite a normal BMI by conventional criteria. So how common is lean Nafu-D in the United States? The number of studies assessing the prevalence of non-obese or lean Nafu-D in the United States is small. The Dallas Heart Study, which included over 2,000 individuals, estimated a prevalence of Nafu-D in 17% of non-obese, that is in those with BMI of less than 30. The multi-ethnic study of atherosclerosis included over 3,000 individuals and estimated a prevalence of 11% among non-obese. It is important to know the impact of race and ethnicity on the lean Nafu-D prevalence. In this study, African-Americans had 6% prevalence of lean Nafu-D, whereas Hispanics had a prevalence of up to 18%, and this mirrors what we see in the conventional overweight or obese Nafu-D. The main study assessing the prevalence of lean Nafu-D when BMI is less than 25 used the National Health and Nutrition Examination Survey, or NHANES-III database, which included over 11,000 people enrolled between 1988 and 1994. In this study conducted by Dr. Yunos' team, the prevalence of lean Nafu-D was 7%. As Nafu-D prevalence increased since the early 1990s, the contemporary prevalence of lean Nafu-D is perhaps higher, but this remains to be determined. Next, we will review the characteristics of lean Nafu-D in reference to the conventional obese phenotype. There are only a couple of studies that explore the clinical profile of lean versus obese Nafu-D in the United States. These studies are listed at the bottom of the slide. One is from the NHANES-III database that I mentioned earlier, and the other one is from our group using the Olmstead County, Minnesota population, and this is an abstract that will be presented at this liver meeting. The findings were consistent among the two databases The first notable characteristic is that those with lean Nafu-D have a lower prevalence of metabolic comorbidities. For example, diabetes is present in 20% of lean versus about 40% of those with obese Nafu-D. The prevalence estimates of hypertension and dyslipidemia in the two groups are also shown for reference. The second important characteristic is that lean Nafu-D is more likely to affect women. There are also distinguishing characteristics between those with lean Nafu-D and those lean without Nafu-D. This study by Dr. Yannossi explored a lean population of over 5,000 adults from the NHANES-III database, of whom 581 had Nafu-D. Despite normal behavior, despite normal BMI, there is evidence of a higher metabolic risk profile in those with Nafu-D who had higher rates of diabetes, hypertension, and metabolic syndrome than those lean without Nafu-D. So if we were to summarize these data, the metabolic profile of lean Nafu-D is somewhere in between those lean without Nafu-D and those with obese with Nafu-D. The last part of the presentation will focus on outcomes in lean versus obese Nafu-D, which is critical in the natural history of these phenotypes. Probably the most important piece of this is the paucity of outcomes data of lean Nafu-D in the United States. Most of the published studies are cross-sectional, therefore, relevant data and therefore, relevant longitudinal outcomes in this population are largely unknown. These important outcomes are development of cirrhosis, risk of cardiovascular disease and malignancy, because these are the top two causes of death in Nafu-D in general, as well as mortality. This major gap in the field prompted our group to conduct a study on the natural history of lean Nafu-D in the population. This study will be presented tomorrow in parallel 12 session of Nafu-D epidemiology and outcomes. Briefly, this was a retrospective longitudinal study, including all adults diagnosed with Nafu-D in Olmstead County, Minnesota, between 1996 and 2016. Their entire medical history and outcomes following the diagnosis of Nafu-D was ascertained from a medical record linkage system. We aimed to compare the natural history of lean, overweight and obese Nafu-D in the population. We identified almost 5,000 people with Nafu-D who had a BMI recorded within a year of the Nafu-D diagnosis, we then grouped these in three categories based on the BMI groups mentioned at the beginning of this talk. Of these, approximately 400 patients were lean, whereas over 1,000 patients were overweight and over 3,000 patients were obese. Note that the lean Nafu-D represented approximately 9% of those with no Nafu-D. Over the 23 years of follow-up, we ascertained the onset of cirrhosis and decompensation, cardiovascular events, malignancies and death. I will present the main results and direct those who are interested in the details of this study to attend tomorrow's session. In this cohort, lean Nafu-D, shown in black in this figure, had a trend towards lower risk of cirrhosis development in reference to obese, shown in blue. The hazard ratio was 0.33, with a p-value that did not reach statistical significance possibly related to the low number of events in the lean cohort. There was no difference in the risk of subsequent cardiovascular events or diagnosis of cancer between lean and obese Nafu-D, or overweight for that matter. And this is shown in the two figures and in the results of the multivariable Cox regression analysis. Interestingly, there was a difference in mortality, which was higher in lean Nafu-D when compared to obese. And referenced to obese Nafu-D. The hazard ratio was 1.63, and this was obtained after adjustment for several variables with likelihood of confounding mortality, including development of metabolic comorbidities after the Nafu-D diagnosis or at baseline. This is obviously an intriguing result and needs to be explored in other cohorts. If validated, it would have significant implications in the diagnostic and management of lean Nafu-D. Again, I will direct you to the session for further details about this study and an opportunity to ask questions. The following key takeaways summarize the data we have on lean Nafu-D in the United States. It is estimated to affect 7% of adults. It is characterized by a lower metabolic risk profile. There is no evidence of worse liver-related outcomes in the studies from the United States, and it is potentially associated with higher mortality, but this needs to be validated. Overall, the population data are limited in the United States. More studies are needed to understand the natural history in contemporary cohorts, and the management remains unclear. I will conclude this presentation with a list of important questions that remain unanswered. In regards to screening, who is at risk? Because these patients do not fit the usual phenotype that we consider in clinical practice, such as abnormal BMI, high prevalence of diabetes, or other metabolic comorbidities. With respect to diagnosis, what is the impact of genetic risk in lean Nafu-D? Regarding the management, what is the role of dietary changes and exercise in this population? What parameters do we use for guidance? Because BMI, being normal, is not a reliable one. It is likely that markers of visceral adiposity should be used, but further research should explore which are the ones that could be applicable to large populations. In regards to therapeutics, we do not know what the role of novel agents is in lean Nafu-D, as the majority of the current studies exclude patients who do not have an abnormal BMI. So I will end with a call to action for more studies of this particular Nafu-D phenotype, which is not rare and is difficult to manage due to these gaps. Thank you. Thank you, Professor Inosui, and thank you, Professor Takehara, for your kind introduction. And I'd like to express my sincere thanks to Professor Bezzera, the president of ASLD, for giving me an opportunity to present our work in this session. Today's my talk is about the data-mining-based metabolic risk abnormalities for lean NAFU-D in Japan. This is my CV. This is my COI. Backgrounds. Our recent meta-analysis shows that the prevalence of non-nafeas Nafu-D is 51% in Europe, 43% in the United States, 38% in Asia. We also analyzed annual changes in the prevalence of non-nafeas Nafu-D in Japan. Red bar shows the prevalence of non-nafeas Nafu-D. The prevalence is 47% in 2009, and it goes up to 53% 2019. So the prevalence of non-nafeas Nafu-D is gradually increasing in Japan. So far, we have analyzed So far, what we know is the prevalence of non-nafeas Nafu-D is high in Eastern and Western countries. Patient with non-nafeas Nafu-D appears to be healthy. However, they are metabolically unhealthy. And what we do not know is the difference in hepatic fibrosis between obese and non-nafeas Nafu-D. We do not know which is more severe, obese or non-nafeas Nafu-D. In addition, we don't know about a metabolic risk of abnormalities for hepatic fibrosis and non-obese NAFUDI. What is responsible factor? Diabetes, central obesity, or hypertension? The main topic of this study is first, the difference in hepatic fibrosis between obese and non-obese NAFUDI. Second, data mining analysis of metabolic changes for hepatic fibrosis and non-obese NAFUDI. In this study, we used the Health Checkup Examines database 2009 to 2019. The total number is over 9,000. From the database, we excluded subjects with absence of fatty liver, alcoholic intake greater than 20g per day, hepatitis B virus infection, hepatitis C virus infection, and lack of data for FIFO index. So we found approximately 6,000 subjects with NAFUDI in the database. Then we excluded follow-up visit. And finally, we analyzed individuals with NAFUDI at initial visit. Total number is approximately 1,500. Patient characteristics. Non-obese NAFUDI patients are significantly older. The prevalence of central obesity, hypertension, hypertriglycemia, and low HDL cholesterol were significantly lower in the non-obese NAFUDI group than the obese NAFUDI group. There is no significant difference between obese and the non-obese NAFUDI group in alcohol intake habits, exercise habits, steps in a day, and sleep disturbance. This slide shows the difference in the severity of fatty liver and hepatic fibrosis between obese and non-obese NAFUDI. Severity of fatty liver was evaluated by Fatty Liver Index. Red color shows High Fatty Liver Index. As we expected, high prevalence of high fatty liver was seen in the obese NAFUDI group compared to the non-obese NAFUDI. Then we evaluated hepatic fibrosis by using FIFO index. And here we got unexpected result. FIFO index was significantly higher in the non-obese NAFUDI group compared to the obese NAFUDI group. Since the age is a factor for FIFO index, we then used age-adjusted FIFO index. Even after the adjustment, high prevalence of high FIFO index was seen in the non-obese NAFUDI non-obese NAFUDI group compared to the obese NAFUDI group. So, in non-obese NAFUDI, the prevalence of severe fatty liver was lower. However, the prevalence of hepatic fibrosis was higher. Then we investigated an impact of high FIFO index on serum AFB level and risk of cardiovascular events in non-obese NAFUDI. There is a significant increase in serum AFB level in the high FIFO index group compared to the low FIFO index group. Risk of cardiovascular events was evaluated by Framingham risk score. The Framingham risk score was significantly higher in the high FIFO index group compared to the low FIFO index group. Risk of cardiovascular events was also evaluated by Swita score, which has been developed for Japanese patients. The Swita score was significantly higher in the high FIFO index group compared to the low FIFO index group. According to these results, high FIFO index was associated with higher AFB level and risk of cardiovascular events in patients with non-obese NAFUDI. Moving on to look at the effects of high FIFO index on patient-reported outcomes, we used CLD-q NAFUDI. CLD-q NAFUDI is a disease-specific instrument for the assessment of health-related quality of life in patients with NAFUDI. CLD-q NAFUDI has been sufficiently validated in two international phase 3 clinical trials. CLD-q NAFUDI consists of four domains correlated with SF-36 and two disease-specific domains. CLD-q NAFUDI has been developed by Professor Inoshi, and we've been working together and validated the usefulness of CLD-q NAFUDI in Japan. This slide showed an association of CLD-q NAFUDI with hepatic fibrosis in non-obese NAFUDI. The data is based on our clinical dataset, and we found that fatigue was the most impaired domain. More than 75% of non-obese NAFUDI feels fatigue, but there is no significant difference between high and low liver stiffness group. Significant difference was seen in the activity domain. High prevalence of impaired activity was seen in high liver stiffness group compared to the low liver stiffness group. So high prevalence of impaired activity was seen in non-obese NAFUDI with hepatic fibrosis. Our results showed that hepatic fibrosis in non-obese NAFUDI was associated with higher AFP level, cardiovascular risk, as well as impaired health-related quality of life. So such patients should be treated. To think about a therapeutic strategy, I think it is important to understand the changes in metabolism in non-obese NAFUDI. A recent meta-analysis shows that the prevalence of central obesity, diabetes, hypertension, and low HDL was higher in non-obese NAFUDI compared to the healthy subject. So the patients with non-obese NAFUDI are metabolically unhealthy. However, a limited information is available for the factors associated with hepatic fibrosis in non-obese NAFUDI. To find out independent factor associated with high FIFO index, we performed logistic regression analysis. In obese NAFUDI group, none of following factors were not identified as an independent factor for high FIFO index. The result may indicate that obese NAFUDI is heterogeneous, and these six factors may intricately associate with high FIFO index. While in non-obese NAFUDI, hypertension and alcohol intake were identified independent factor. The odd ratio is 1.6 for hypertension, 1.5 for alcohol intake. These are statistically significant. Accordingly, hypertension and mild alcohol intake may be independent factors for hepatic fibrosis in patients with non-obese NAFUDI. Independent factor associated with high FIFO index was also examined by data mining analysis. In the study, we employed two types of data mining analysis. The one is decision tree analysis. Decision tree analysis reveals a series of classification rules by identifying priorities. So the advantage of the analysis is identifying the most important factor. However, the disadvantage is the result depends on the database. Another data mining analysis is a random forest analysis. Random forest analysis consists of many decision tree based on a random sampling. Random forest analysis actually performs decision tree to identify the factors distinguishing between the case and control groups. This slide shows the result of decision tree analysis for high FIFO index in non-obese NAFUDI. In a data set, 17.5% shows high FIFO index. And the most important classifier for high FIFO index was hypertension. The prevalence of high FIFO index was 22.6%. The prevalence of high FIFO index was 22.6% inpatient with hypertension. While the prevalence was 15.5% inpatient with no hypertension. Inpatient with no hypertension, the second important classifier was alcoholic intake. The prevalence of high FIFO index was 19.1% inpatient with alcoholic intake. While the prevalence is 13.5% inpatient with no alcoholic intake. This slide shows the result of random forest analysis for high FIFO index. In this analysis, the importance of the factor is expressed by valuable importance, VI. Higher VI means higher importance. When we look at the data, we can notice the valuable importance of hypertension and alcoholic intake were more than three times higher than the other factors, such as central obesity, low HDL, high triglyceride, and diabetes. What I'd like to emphasize here is that three different types of analysis showed the same result, which is hypertension was the most important factor for hepatic fibrosis in patient with non-obese NAFUDI. It remains unclear why hypertension was associated with hepatic fibrosis in non-obese NAFUDI, but we found an interesting study reported by Wei et al. in Journal of Hepatology back in 2008. They found the elbow expression of running causes NAFUDI with fibrosis, but with no obesity. These are data for the body weight, and there is no significant difference in body weight at 9 weeks and 12 weeks between control and wearing transgenic mice. These are images for HD staining, and we can notice that accumulation of lipid droplets in hepatocyte, but the amount is not that much as like other NASH animal models. These images are serious weight staining. In wearing transgenic mice, hepatic fibrosis was seen in zone 3 at 9 weeks, and the progression of hepatic fibrosis was evident at 12 weeks. It also remains unclear why wearing causes hepatic fibrosis, but several studies show that wearing a hepatic cell leading to an upregulation of TGF-beta and MCP1 causing hepatic fibrosis. So, wearing may be a key factor for hepatic fibrosis in non-obese NAFUDI. The importance of wearing on hepatic fibrosis is also supported by following two papers. The first paper is reported by Noguchi, Yoshiji, and Fukui et al. They found that selective aldosterone blocker inhibits the progression of hepatic fibrosis in a rat model of NAFUDI. The picture shows a significant reduction of hepatic fibrosis in rats treated with aldosterone blocker. The another study is reported by Gore et al. They found the use of rat blockers had less advanced fibrosis in hypertensive patients with NAFUDI. When we look at the use of rat blocker, an adjusted odd ratio is 0.46 and adjusted odd ratio 0.37. Both are statistically insignificant. So, again, wearing may be a key factor for hepatic fibrosis in non-obese NAFUDI. In the following two slides, I'd like to talk about pathogenesis of obese and non-obese NAFUDI. Based on the result, there is no independent factor associated with hepatic fibrosis in obese NAFUDI. The result may indicate that the various metabolic disorders may intricately associate with hepatic fibrosis in obese NAFUDI. And here is the well-known pathogenesis of obese NAFUDI. Obesity and bacillary adiposity causes insulin resistance and subsequent hyperinsulinemia, which causes activation of a hepatic stell cell, resulting in progression of hepatic fibrosis. This slide shows a possible pathogenesis of non-obese NAFUDI. According to our result, hypertension and mild alcohol intake were independent factors associated with hepatic fibrosis in non-obese NAFUDI. And our result, along with previous study, suggests that wearing is a possible pathogenesis. Because wearing causes not only hypertension, but also activation of hepatic stell cell, leading to progression of hepatic fibrosis. This possible pathogenesis is also supported by a result of a meta-analysis. The study shows that in a non-obese NAFUDI, incidence of new onset of hypertension is 56.1 per 1000 person years. The incidence is more than 4 times higher than that of diabetes. In conclusion, the prevalence of non-obese NAFUDI is gradually increasing in Japan. Non-obese NAFUDI is not a mild disease. High FIFO index was seen in 18% of patients with non-obese NAFUDI. In patients with non-obese NAFUDI, high FIFO index was associated with high AFV level, high cardiovascular risk, and inactivity. Hypertension and mild alcohol intake were features of high FIFO index in non-obese NAFUDI. The features were different from obese NAFUDI, suggesting that these factors are selected target for the fibrosis in a patient with non-obese NAFUDI. This slide lists my collaborators. I'd like to say thank you for all the Japanese doctors. And I also express my sincere thanks to Professor Yunusi, Professor Lin, and all the members of Yunusi's lab. Thank you for your time. Hello, ladies and gentlemen. I'm Vincent Wong from the Chinese University of Hong Kong. I hope you are well. First, I'd like to thank the organizers for the honor to present at this meeting. Here are my disclosures. We have long known that NAFUD is caused by both genetic and environmental factors. NAFUD is more often seen among first-degree relatives of NAFUD patients. In classical gene studies, the heritability of both hepatic steatosis and fibrosis is estimated at around 50%. The first NAFUD-related gene polymorphism identified by a genome-wide association study is PNPLA3, patotin-like phospholipase domain-containing protein 3. After the publication of the landmark paper in Nature Genetics 12 years ago, various groups across the world have confirmed the association between PNPLA3 gene polymorphism and the presence and severity of NAFUD. The gene polymorphism leads to the production of an aberrant protein, which results in an accumulation of lipid droplets in hepatocytes from reduced ubiquitylation and proteosomal degradation. The reduction in hydrolase activity results in an accumulation of monounsaturated fatty acids and a reduction in VLDL secretion. Some but not all studies also showed that the PNPLA3 gene polymorphism affects the hepatic stellate cells and promotes fibrosis progression. One proposed mechanism is through reduced retinol esterase activity, resulting in retinal retention in hepatic stellate cells, as depicted by the figure on the right-hand side. With the right environment of liver injury and TGF-beta action, this leads to upregulation of profibrogenic molecules and cytokines. It is noteworthy that the knockdown of PNPLA3 gene in in vitro and animal studies actually protects against fatty liver. It is the aberrant protein that causes trouble. Therefore, in drug development, the aim is not to inhibit PMPLA3, but to tackle the mutant protein. This nicely written review by Dr. Yunosi and colleagues illustrates that PMPLA3 may at least partly explain the racial and ethnic differences in NAFLD epidemiology. In the United States, for instance, Hispanics more often carry the PMPLA3 polymorphisms and are most likely to have NAFLD and more severe form of NAFLD. In contrast, the polymorphism is rare among African-Americans, and we know the prevalence of NAFLD is also lower in this population. Interestingly, the polymorphism is particularly common among East Asians. This may explain the relatively high prevalence of NAFLD in this area despite a lower metabolic burden compared with North America. Studies generally show that the PMPLA3 gene polymorphism doesn't affect the metabolic profile including body mass index, glucose, insulin resistance and lipids. Thus, the effect of PMPLA3 on liver fat is primarily hepatic in origin and not secondary to changes in metabolic risk factors. Our topic today is Lean NAFLD. In 2017, a study from Nature Genetics examined the interaction between several NAFLD genes and metabolic factors in several community cohorts. When stratified by BMI categories, the PMPLA3 polymorphism had a greater influence on intrahepatic triglyceride content and ALT level in people with high BMI in the Dulles Heart Study. In fact, both liver fat and ALT level remained quite low in people with BMI below 25 regardless of the PMPLA3 genotype. This suggests that while PMPLA3 polymorphism increases the risk of fatty liver, its manifestations still require the right metabolic environment. On the right hand side, we can see a similar picture in the Copenhagen cohort. This time, the analysis focused on cirrhosis based on diagnosis codes in the electronic health record. Again, the effect of PMPLA3 gene polymorphism on cirrhosis was more apparent in subjects with high BMI. That said, the observation may also be explained by the small number of lean subjects with cirrhosis. As clearly explained in the previous talks, Asians tend to develop metabolic complications at a lower BMI because of the propensity for central fat deposition. Indeed, studies in the 1980s showed that the risk of diabetes, myocardial infarction and stroke started to increase as the BMI exceeded 23 in Asians. In Hong Kong, we conducted a population screening study using proton magnetic resonance spectroscopy and transient elastography to measure liver fat and fibrosis, respectively. We used the Asian BMI definitions of 23 for overweight and 25 for obesity. As expected, the intrahepatic triglyceride content was higher in people with high BMI. However, we found that PMPLA3 had an effect on liver fat in all BMI categories. If anything, the relative effect of PMPLA3 was even higher in lean patients in this cohort. Using an intrahepatic triglyceride content of 5% as the threshold to diagnose fatty liver, the prevalence of fatty liver was again the lowest in subjects with BMI below 23 and highest in those with BMI of 25 or higher. Nonetheless, the impact of the PMPLA3 gene polymorphism was the highest in lean subjects. Compared with lean subjects with the PMPLA3cc genotype, those carrying the at-risk GG genotype had a 4-fold increase in the prevalence of fatty liver, increasing from 6.5% to 27.3%. The difference in the prevalence of fatty liver between the CC genotype and the GG genotype groups dropped to 2.2-fold in subjects with BMI between 23 and 25, and 1.4-fold in those with BMI of 25 or higher. In the obese group, the difference in fatty liver prevalence among people with different PMPLA3 genotypes was no longer statistically significant. Looking at the data from another angle, if we focused on subjects with fatty liver, lean and non-obese patients were more likely to carry the PMPLA3G allele than obese patients. Taken together, the PMPLA3G allele was seen in around 80% of non-obese patients who had fatty liver. Our findings were confirmed by the multicenter entire study, which included patients with biopsy-proven NAFLD in Japan. The PMPLA3GG genotype was more common in NAFLD patients with BMI below 25. If we combine data from different studies, we can draw two conclusions. First, the effect of the PMPLA3 gene polymorphism itself only has a small effect on liver fat in lean and healthy individuals. Metabolic risk factors are required for the genetic effect to manifest. On the other hand, among lean NAFLD patients, the relative effect of genetic predisposition would be greater. If the PMPLA3 mutant requires the right metabolic environment to induce fatty liver, one interesting question is whether it would influence the effects of treatments for NAFLD and NASH. Our group conducted a randomized controlled trial testing a lifestyle modification modification program in 154 community NAFLD patients, half of whom had a BMI below 25 at baseline. Interestingly, we found that in the lifestyle modification group, patients carrying the PMPLA3G allele had a greater decrease in liver fat one year later. We also observed a dose-response relationship with the greatest decrease in patients with the GG genotype and patients with the CG genotype were in the middle. In contrast, on the far right of the diagram, we didn't see the same difference across genotypes in the control group that received usual care. Our data suggests that while the PMPLA3 polymorphism increases the risk of fatty liver, through its interaction with metabolic factors, it may also increase the response to lifestyle modification. Now that a number of agents are in phase 2 and 3 development for the treatment of NASH, the potential effects of genetics should be examined. Now let's switch gear and discuss another well-established single nucleotide polymorphism associated with NAFLD. The TM6SF2 gene was detected in an exome-wide association study, again based on the Dulles Heart Study. TM6SF2 stands for transmembrane 6 superfamily 2. Compared with PMPLA3, the minor allele of TM6SF2 is quite rare. In TM6SF2 homozygotes, the liver fat content is increased by two to three-fold. Interestingly, while the TM6SF2 gene polymorphism increases the risk of fatty liver, it reduces total cholesterol and protects against myocardial infarction. The opposite effects of TM6SF2 on liver fat and cardiovascular risk stem from its physiological action. TM6SF2 is expressed in the small intestine, liver, and kidney. It is found in the endoplasmic reticulum and Golgi complex of hepatocytes. The mutation results in downregulation of the TM6SF2 protein. In preclinical studies, TM6SF2 affects lipidation but not secretion of very low-density lipoprotein. Although the mechanism is somewhat different from that of the mutant PMPLA3 protein, the end result is similar with retention of lipids inside hepatocytes. Nevertheless, unlike PMPLA3 mutation, TM6SF2 is not an attractive treatment target as it is highly likely that the improvement in liver fat would be offset by increased cardiovascular risk, which remains the leading cause of death in patients with NAFLD. Just now, we saw that the prevalence of PMPLA3 gene polymorphism mirrors the prevalence of NAFLD in different countries and ethnic groups. However, we don't see a consistent pattern in TM6SF2. The minor allele frequency appears to be similar among Caucasians and Hong Kong Chinese. In the original Nature Genetics paper, TM6SF2 was less common in African Americans and Hispanics than Caucasians. In terms of the gene environment interaction, we can see a similar pattern in the Dallas Heart Study. On the right is TM6SF2. Because of the low frequency of the gene polymorphism, the EK and KK genotypes were combined for analysis. The effect of TM6SF2 gene polymorphism on liver fat was only observed among individuals with BMI of 30 or higher. The same appears to be true for GCKL, which encodes the glucokinase regulatory protein on the left hand side. Because glucokinase is responsible for fructose metabolism, and we know that high fructose intake is one of the strongest dietary risk factors for NAFLD, I would be very interested to see more studies on the interaction between GCKL and diet. We don't have very good data regarding the prevalence of TM6SF2 gene polymorphism in lean NAFLD. One such study based on an Italian liver biopsy cohort suggests that TM6SF2 KK homozygotes are more common in lean NAFLD than NAFLD patients with BMI of 25 or higher. Nevertheless, because only 4 patients in this study were homozygotes, the findings are far from definitive. Other genes, like MBOT7 and HSD17B13, may also be important in the development or prevention of liver diseases and liver injury. We haven't covered these today because data on lean NAFLD are scarce. There are also some isolated studies reporting the association between some gene variants and NAFLD in lean patients, but most of these haven't been independently validated despite potential pathophysiological explanations. For instance, Leon Adams found an association between the cholesterol ester transfer protein and fatty liver in lean girls in Western Australia. CETP facilitates the transport of cholesterol esters and triglycerides between lipoproteins, and the mutation is associated with accelerated atherosclerosis. In another study of the Italian liver biopsy cohort, Salfopeta showed an association between interferon lambda 4 and histological severity of NAFLD in non-obese patients, and this is of course related to inflammation and tissue injury. In another study that focused on genetic changes in lean patients using whole exome sequencing, phosphatidylethanolamine and methyltransferase was associated with NAFLD. PEMT is an enzyme that converts phosphatidylethanolamine to phosphatidylcholine, which is critical in hepatic lipogenesis. There are also rare genetic disorders that can cause severe hepatic steatosis in lean individuals, like genes affecting hepatic triglyceride metabolism and de novo lipogenesis. That said, these are all rare conditions that tend to cause dramatic physiological changes. They are different from the more common genetic changes that we have discussed so far. Now, we have only focused on single nucleotide polymorphisms and mutations. Compared with other chronic diseases, we know very little about the effects of other genetic changes on NAFLD, not to mention in lean individuals. One notable area needing further research is epigenetics. We know one phenomenon, that is, in children born to malnourished mothers, they tend to develop obesity and fatty liver during adulthood. In the Young-Finn study, for example, low birth weight rate was independently associated with fatty liver later on in life. According to current understanding, both maternal high-fat diet and undernutrition can cause epigenetic as well as gut microbiota changes in the offspring. Apparently, if the intrauterine environment indicates malnutrition, the body responds to the cue and prepares the infant accordingly. Unfortunately, such changes become maladaptive when nutrition becomes abundant. The epigenetic and microbiota changes would modify gene expression, which in turn predisposes the child to obesity later in life. In summary, chairpersons, ladies and gentlemen, single-gene polymorphisms of common variants such as the PMPLA3 gene often require the right metabolic environment to induce fatty liver. However, the relative importance of these gene variants is higher in lean individuals with NAFLD. Except in rare genetic disorders, there have been no identified genetic changes that are unique in lean individuals with NAFLD. Finally, future studies using multi-omic approaches are needed to fully elucidate the gene-environment interaction in the pathogenesis of NAFLD and NASH. Thank you very much. It's a great pleasure to mention the closing remarks of this exciting symposium. I am Tetsuo Takehara at Osaka University, one of the moderators of this symposium. I am currently serving as the Director General of the Japan Society of Hepatology, JSH. Taking this privilege, I'd like to briefly introduce JSH, JSH. The Japan Society of Hepatology was established in 1965. JSH originates from the Japanese branch of IASU, which was founded in 1959. We have currently more than 12,000 active members, including hepatologists, transplant surgeons, radiologists, and researchers, as well as medical staff in the hepatology field. We have a monthly official journal, Hepatology of Research, publishing many excellent papers from all over the world. Unfortunately, 2020 has been the year of the coronavirus infection, but it's also the year when the collaboration between JSH and ASLD begins. This symposium is supposed to be held annually in the United States and Japan with the cooperation of the both societies. This year, it's scheduled to be held in November at JDDW, Japan Digestive Disease Week in Kobe, with the theme of HCC surveillance and treatment. And this week, it's being held here in the RIBA Meeting Digital Experience with the theme of And this week, it's being held here in the RIBA Meeting Digital Experience with the theme of Lynn Nash. We hope this symposium will contribute to the friendship between the two societies and the development of hepatology. Thank you very much.

ACLD Japan Society of Hepatology

Joint Symposium

NAFLD

lean individuals

genetic variations

lifestyle modifications

treatment strategies

gene-environment interactions

research

hepatology

international cooperation