Thank you for the invitation. I will talk about molecular signature and epidemiology in NASH ACC in the U.S. Here's my disclosure. I will talk about four things today. NAFUD ACC in the U.S., trend and projection, challenging NAFUD ACC screening, risk stratification, and biomarker-guided chemoprevention. Starting from the epidemiology, this is a recent estimate of prevalence of NAFUD-associated ACC in the U.S., which is 15 to 20%. But this doesn't stay high over the past. From these four studies, from a variety of sources, CL Registry, UNOS, another transplant registry, the prevalence used to be single-digit a decade ago, which sharply has risen to 15 to 20%. And this prevalence will keep increasing. So this is the projection over the next decade up until 2030, plotting the incidence for decompensation and liberated death, and at the bottom, ACC in the U.S. So that the slope of the curve for ACC may not look steep compared to decompensation and deaths, but if you look at the number, both prevalence and incidence will be double over the next decade. And also, NAFUD is a newly proposed entity, a redefinition, which may take over NAFUD. And if you look at the global prevalence, which is reported in the recent CCH paper, the prevalence is simply doubled to 50%. So this is a huge population and keep increasing. So here is a challenge in NAFUD-ACC screening with this background. So here's a natural history of NAFUD patients with a number in the U.S. So as shown in the left, 60 to 80 million Americans are supposed to be affected with NAFUD, among which 15 to 20 million will develop into NASH. And then as is shown in the middle, fibrosis gradually progress, and as in the right side, eventually reach to the stage of cirrhosis. And once the disease fibrosis reaches the stage of cirrhosis, ACC, which is in the middle, start to emerge. And also, clinical studies have shown that ACC can arise even before reaching to the stage of cirrhosis, which is very unique to NAFUD. And this is the ACC deriving from the huge population. And if you look at the incidence rate, which is 1 to 2% from the cirrhotic patients. And from the patient with advanced fibrosis, but not yet cirrhosis, which is very low, 0.01% or even lower. So if you want to detect this ACC early stage, practice guidelines recommend regular ACC screening to this population. But if you think about the actual scenario, which is shown in this cartoon, that is identifying the one patient at the right bottom, shown as a red liver, out of this rest of the 99, or even more, NAFUD fibrotic patients without incidence of ACC. So the guidelines are kind of recommending one size fits all, ultrasound plus minus AFP twice a year, which is a huge commitment. So the question is if this is really cost effective, and if you can really detect early stage ACC as intended. So this is a meta-analysis done by my colleague Amit Sengal, summarizing the utilization rate of recommended ACC screening, which is surprisingly low. At the top panel, at the community clinics, which is less than 10%. At the medical center, at the middle, which is still less than 30%. Overall, less than 25%. So if you're saying this is our capability to offer guideline-recommended ACC screening, and if you think about you randomly allocate this 24% onto any of these random patients here, you will easily miss this one patient who developed real ACC. So here, the concept of risk stratification kicks in to resolve this problem. So you may think about that as a practical scenario, step-wise enrichment of high-risk population. Again, we have a huge population of NAFUD, or even larger, if not MAFUD. And if you look at the patient with F3 or 4 fibrosis, which is already a significant reduction, but still many patients. And you may think about further applying the clinical variable-based risk estimation to further narrow down the population. And if you have a molecular biomarker, you can identify even the smallest subset with the ultra-super high-risk population. That may allow you to laser focus on this population to make your screening more efficient, and possibly think about the intervention, medical intervention. So, as for the clinical score, there are several out there. And in this recent study, comprehensively and systematically assessed the several scores, including NFS, P4, Barscore, APRE, HEFS, for the detection of existing amount of fibrosis, as shown in the left panel. And also, the association with the future progression of lethal complication, as shown in the right side. So, it seems promising. Some of them, indeed, allow us to enrich the high-risk population to some extent. What about molecular biomarker? So, there are several promising developments ongoing. And let me talk about one example we are working on, which is called the prognostic liver secretome signature, or PLSEC. We call it as a liquid liver biopsy because this is a serum protein panel, which is intended, developed to surrogate the same information you can see in the liver tissue. And this serum-based surrogate of a liver molecular signature can classify the patients with a different ACC incidence at hazard ratio of 3 over the time frame of 15 years in our recently published study. If you look at the subgroup of the patients in this study, you see this biomarker is associated with ACC risk in a similar extent. And if you look at the etiology at the bottom, very encouragingly, this biomarker shows a stronger association with ACC risk in an aphrodisiac and a clitogenic liver disease patients. And interestingly, we see less association with patients with active ACV infection, which is the patient population kind of quickly fading away. And even more interesting, patients with ACV are still in the middle between an aphrodisiac and an active ACV infection, and is significant. So, we think this PLSEC as an etiology agnostic risk indicator. And recently, we have been developing another additional etiology-specific risk biomarker for NAFLD, which is called the PLSEC-NAFLD by Naoto Fujiwara in our team, which is shown here on the left top. On the right side is a summary of the molecular pathways associated with this PLSEC-NAFLD. Very interestingly, many of the pathways that involve innate and adaptive anti-tumor immunity. And in an independent validation, this biomarker can classify the patients according to the ACC risk. And very encouragingly, there is no ACC incidence in the low-risk patients over the timeframe of 15 years. And also, because these molecular pathways captured by this signature is very different from the signature associated with the etiology agnostic PLSEC, we ask, what if we combine both those two biomarkers? One is PLSEC as an etiology agnostic risk indicator as a backbone, and that plus PLSEC-NAFLD as a kind of plugin, which we find the risk prediction according to the liver disease etiology. And we simply ask, who are both high risk for this biomarker, or either high risk or both low risk? And if you classify the patient in this way, you can see even better separation of the risk, again, over the timeframe of 15 years. So with this promising clinical and the molecular tool for risk stratification, risk prediction, you may think about the strategy to refine our way of ACC screening. So here's our proposed risk-stratified ACC screening. So here you have a NAFLD patient with AF3 or fibrosis upfront, and you do ACC risk assessment, which can be a combination of one molecular score and this classified patient into the D3 risk group. And you may offer altered intensity of screening according to the risk estimation upfront. More specifically, you may consider choosing the different screening modality. For example, for patients predicted as having high risk, you may offer a more costly, but high performance modality like an MRI or a new CFDNA biomarker. We quantitatively assessed if this stratified strategy is really beneficial or not by using Markov model-based simulation. So here, the scenario we tested is any combination of these possible screening imaging modality for each of these risk groups. So you can choose, for example, from the three of these for high risk patients, and the four of these for intermediate patients, and the two of these for low-risk patients. For example, one strategy may be MRI for high-risk people, ultrasound for intermediate people, and doing nothing for low-risk people. And we comprehensively assess the measure of cost-effectiveness in the simulation. So in the simulation, you need to assume the clinical context, especially the liver disease etiology, because annual ACC incidence will be different according to the scenario. For example, in patients with ACV sclerosis, anticipated annual ACC incidence is between 3 to 5 percent. Once these patients achieve the viral cure, this range should drop below 2 percent. And in our current population of interest, NAFUD, S3O4, which is between very close to 0 percent to 2 percent. And here is the result of the measure of cost-effectiveness as a measure of cost-effectiveness. As a measure of benefit of stratified screening, and the y-axis is showing the measure of cost-effectiveness, ICER, which is the amount of money you need to pay to gain one more life year by introducing this new strategy, comparing to the spinal care strategy, which is ultrasound twice a year. And a very, very accepted cutoff, which is called the willingness to pay cutoff, is $50,000, which means we are willing to pay up to $50,000 to gain one more life year for the patient. And any strategy underneath this cutoff is regarded as cost-effective. So, here at the top, you see the one strategy, which is ultrasound four times per year for high-risk people, ultrasound twice a year for intermediate-risk people, and doing nothing for low-risk people, which is cost-effective under this threshold when annual incidence rate is high, for example, AC resources patients. But in our patient population with interest, NAFLD S304, this is becoming quickly too expensive. In contrast, other strategies, like AMRI, which stands for the abbreviated MRI, which is simplified MRI, tailored for screening for high-risk and intermediate-risk people, and nothing for low-risk people. This is barely no additional money, just $2,000 to gain one more life year. So, theoretically, this analysis suggests this is a viable strategy. But you may think this is too complicated to implement in a real clinical practice. But I wanted to show this very encouraging survey-based study just published last year in CGH. So, this is showing the summary of percentage of physicians to the answer of the question, which of ultrasound, which is shown on the left, or CT or MRI, shown on the right, you would offer as an ACC screening tool for the patients in front of you, in case you can predict the future ACC risk in the three scenarios, as shown in the right region. One, 1% per year, two, 3% per year, or 5% per year. So, in case the predicted ACC risk in the future is 1% per year, most of the physicians will offer ultrasound, as shown in the most left bar. But as this predicted probability of ACC, risk of ACC, increase, the more physicians would offer CT or MRI. So, I think this is encouraging evidence that physicians are receptive to tailoring the ACC screening modality if they can know the ACC risk upfront reliably and quantitatively. So, let me move to the last part, which is a biomarker-guided NAFLD-ACC chemoprevention. So, with this stepwise approach, assuming we end up with a molecular biomarker-based risk population. So, whatever biomarker you choose, typically this biomarker is bound to specific molecular pathway dysregulation. And in parallel, you can think about the variety of molecular-targeted agents possibly associated with some of these dysregulated pathways in association with this biomarker. And you may think it would be the rational approach to give this drug to these patients as a measure of reducing the risk as a chemoprevention. So, we came up with…came to this idea with this very encouraging observation in a recent study. So, here, we measure the ACC risk level by this serum protein marker, PureSec, which is shown on the y-axis, in a patient series for the four-time point, time series assessment of the ACC risk in patients who achieved ACV cure by antiviral therapy. So, we know if you successfully clear the virus by antiviral therapy, especially DAA, most of the patients will be free from ACC, but a small subset of patients are still at risk of ACC. And we asked if there is any difference in the change of this ACC risk level measured by the biomarker over time between those two patient groups. So, we have a four-time point before antiviral therapy, at the end of antiviral therapy, four weeks after completion of antiviral therapy, and one year after completion of antiviral therapy. And in patients who are free of ACC years after completion of antiviral therapy, as shown in these orange boxes, that ACC risk level measured by this biomarker gradually decreased. In contrast, patients who develop ACC despite successful viral clearance, which is shown in these green boxes, shows that persistently the same level of ACC risk level measured by the biomarker. And also, we independently observed in another cohort that this modulation is observed or not is associated with 20-year ACC incidence. So, we think this suggests this biomarker, or any other similar ACC risk biomarker, can be used as a surrogate endpoint of any chemopreventive experimental intervention to estimate the future benefit to reduce ACC incidence. And we are thinking of a sterile agent to test under this scheme, and this is one example, lipophilic statin. So, this is a retrospective study published by our colleague Tracy Simon and Wei Chang of Mass General Hospital, showing that use of lipophilic statin use, not hydrophilic, is associated with lower ACC incidence in a national cohort. And we simply asked how this biomarker-based risk level looks like in lipophilic statin users versus non-users, and very interestingly, we saw a striking difference in the biomarker-based level. Based on these premise-promising observations, we are about to start a clinical trial where we test atrovastatin, which is one of the lipophilic statin given for one year compared to placebo for patients with histologically confirmed cirrhosis. And we measure the biomarker before and after the therapy and measure the change of this ACC risk level as a primary endpoint of this trial and also for the future association with ACC incidence. So, this trial, along with other trials we are planning or conducting, is successful. We are expecting this strategy could transform our strategy to develop chemopreventive intervention in ACC. So, here's a key takeaway. NAFUD, NAFUD is sharply increasing as a dominant ACC etiology in the U.S. Cost-effective strategy of NAFUD ACC screening is an urgent and main need. Stepwise enrichment of high-risk population may be a viable option to enable cost-effective risk-stratified NAFUD ACC screening. And liquid liver biopsy, for example, may be a useful tool for risk-stratified ACC screening and chemoprevention. Thank you for your attention. Hello, everybody. My name is Junko Tanaka, professor at the Department of Epidemiology, Infectious Disease Control and Prevention, Hiroshima University of Japan. It is my honor to make presentation in this ASLD, Japan Society of Hepatology Joint Symposium. Thank you very much for inviting me as a speaker in this session. And I present about the epidemiology of NAFUD in Japan, a result from nationwide population-based study. So, this is the title. And this is my conflict in interest. Okay. So today's contents include four titles. Let's start. So, disease burden of NAFLD. So, Dr. Yamashita's report, the global prevalence of NAFLD is estimated to be 25%. Meta-analysis on NAFLD prevalence shows the Middle East as highest prevalence at 32%, followed by Asia, 27%. U.S. CDA, Center for Disease Analysis, estimated the change in NAFLD disease burden in United States until 2030, based on the modeling study by Markov model. They estimated NAFLD prevalence was increased from 25.8% to 28.4%, and the number of NAFLD cases were also increased by 21%, from 83 million in 2015 to 100.9 million in 2030, which equal to the 21% of total population in United States. Now, in Japan, in 10 years ago, Eguchi and his research group reported the prevalence of NAFLD among 8,000 residents who received ultrasonography at health check-up in Japan. NAFLD prevalence in male was 41%, and female was 70.7%, and highest in male in their 40s, and females in their 60s. In joint collaboration with CDA, we estimated the change in NAFLD disease burden in Japan from 2020 to 2040, using the same modeling study used for USA in the previous slide. NAFLD prevalence was increased in Japan from 2020 to 2040, but the number of NAFLD patients will decrease because of low birth rate and declining population in Japan. Consequently, due to the effect of aging society in Japan, the number of advanced fibrosis cases are estimated to be increased. Short summary, NAFLD disease burden differs by countries and regions. In Japan, NAFLD continue to be growing liver health trait. It was number of advanced liver fibrosis cases expected to be increased with age of population in Japan. We move next the new concept. NAFLD has been proposed. Should we change NAFLD definition into NAFLD, which is regardless of alcohol consumption? I will show you some of the results from our study. We studied the epidemiology of fatty liver disease in general population, healthy people, by alcohol consumption. So definition of NAFLD is shown in slide in advance. NAFLD is diagnosing case. The evidence of hepatic steatosis is confirmed either by imaging or histology in state of no significant alcohol consumption. However, the definition of significant alcohol consumption is uncertain. Although the criteria from AASLD guideline is shown in this slide, they are still under-resolved issues like under self-reporting and cases of abstinence with the history of heavy drinking. Metabolic and alcohol factors are also overlapped. Therefore, in recent years, as a new positive criterion, the concept of NAFLD was proposed as a single overarching term regardless of drinking history or presence or absence of other liver diseases. However, epidemiology prevents, for instance, of fatty liver by drinking status among general population but has not been fully clarified so far. So we conducted a large-scale retrospective analysis of 850,000 health checkups called data over 11 years duration. It is called alcohol consumption study. We exclude hepatitis B, hepatitis C positives, and we include 820,000 residents having alcohol consumption information. This figure shows the frequency distribution of alcohol drinking by age and sex. In the right side, the light blue color indicated the person who do not drink every day. That blue color indicates more than three times per week, not every day, but orange color indicates drinking every day. The percentage of males who drink every day was the highest in their 50s, about 50%, but for women in their 40s, just 20%. Next, we calculate the fatty liver prevalence among 75,000 who take ultrasonography during their health checkups. This slide shows that prevalence of fatty liver in free alcohol consumption level. In this drinking level, the criteria was different from male and female as shown in the slide. The prevalence of fatty liver based on ultrasonography in non-drinkers, moderate drinkers, and heavy drinkers was similar to 27, 28% respectively in total. It seems there is no difference in fatty liver prevalence by the level of alcohol consumption. We also conducted incident study in the same cohort, and 31,000 patients were included. The incidence rate of fatty liver based on ultrasonography tended to be slightly higher depending on the level of alcohol consumption. In heavy drinkers, it's higher and was higher in old age group, but there is no statistical significance. So far, I showed the result of fatty liver prevalence diagnosed by ultrasonography. However, in health checkups, less than 10% only were undergoing ultrasonography. It seems that the estimation of fatty liver prevalence may have biases toward the higher overestimation because people who diagnosed with fatty liver in the past are more likely to undergo ultrasonography again. As a keynote, we found that drinking status has no effect on the prevalence or incidence of fatty liver in Japan, but we noticed only less than 10% are undergoing ultrasonography. This is one of the limitations of fatty liver assessment using health checkup ultrasonography data. It remains challenging. So we conducted a study for prevalence of fatty liver and liver fibrosis in a randomly selected general population. From K-City and O-City in Hiroshima Prefecture, Japan, 6,000 people are randomly selected based on residence register. We provided biohepatitis screening and liver ultrasonography and fibroscan tests were conducted. Among them, 1,043 were participated for biohepatitis screening and randomly selected 488 people were received ultrasonography and fibroscan together. Prevalence of fatty liver diagnosed by ultrasonography was 24.6% in the middle of the slide in brown color. Prevalence of fatty liver, including mild steatosis diagnosed by fibroscan was 51.0%. And look at the right graph. Fatty liver diagnosed by fibroscan but not ultrasonography in 28.3% in green color in bar chart, so in total. So 52.9% were diagnosed with fatty liver by ultrasonography or fibroscan. This study suggested that the prevalence of fatty liver in general population of the people was 52.9%, including mild cases, could be double level compared to the previous report. It's very high. If so, how to detect advanced fibrosis cases among in general population become critical challenges. We estimated the liver fibrosis stage distribution by fibroscan in this randomly selected general population. As a result, cirrhosis was found in 1% and advanced fibrosis was observed in 2% in total. Looking at the etiology of cirrhosis in these five cases, both hepatitis B and NAFLD is two people. Only NAFLD in two and unknown in one. Male suffered more than fibrosis than female and cirrhosis as well. 1.6% in male but 0.4% in female. By age group, in the middle graph in the slide, 3.6% had cirrhosis, 1.8% had advanced fibrosis in their 60s. Although we use fibroscan to detect liver fibrosis in general population, liver biopsy is still the reference standard for assessment of liver fibrosis. As we know, liver biopsy has weak pins such as variation among pathologists, sampling error, invasive, and can lead to severe complications, relatively high cost, and so on. But for diagnosis for NASH and liver fibrosis progression, liver biopsy is still the reference standard. However, nobody knows how frequent liver biopsy performed for NASH diagnosis in clinical practice. We analyzed large-scale medical claim data of 7 million insured people. We extracted all medical claims with NASH from 2013 to 2018 in Japan. Newly diagnosed NASH was defined as those who did not have disease named NASH in their medical claim at least 12 months ago, one year. Implementation rate of liver biopsy and estrography in the past, three months before diagnosis of NASH, were calculated based on the medical practice record. As a result, only 6.9% had been received liver biopsy before diagnosis with NASH, and 90% had been diagnosed as NASH without liver biopsy nor estrography. Prevalence of fatty liver in randomly selected general population was 52.9% by USC and FibroScan. It was about twice as high as previously reported. By using FibroScan, liver cirrhosis was 1%, and advanced fibrosis 2% in randomly selected general population. Liver biopsy implementation rate for NASH diagnosis in clinical practice is as low as 6.9%. So can we use 5-Fib4 index we use to identify the biased fibrosis in general population? It's a question. We can easily calculate 5-Fib4 index using routine blood tests. If we can apply 5-Fib4 index for primary screening for advanced fibrosis in general population, it may be very useful. So this figure shows the population of age group in each 5-Fib4 index value in the same cohort I introduced in previous slide. As we expected from its formula, 5-Fib4 index value were mostly found in the elderly. It is shown that 5-Fib4 index in general population is highly age-dependent. It seems that 5-Fib4 index is not applicable for primary screening for liver fibrosis screening among general populations. As this is full chart for identifying advanced liver fibrosis in Japan, according to Japanese guideline, if fatty liver detected by ultrasonography or metabolic risk factors are found, liver fibrosis should be evaluated by fibrosis markers such as type 4, 7S, or scoring system like 5-Fib4 index or some others. I'd like to show the small challenging result of the comparison between 5-Fib4 index and type 7, 7S for to know the liver fibrosis stage among healthy population, not among diagnosed fatty liver cases or patients. So this graph shows 5-Fib4 index distribution by age group in NAFLD diagnosed by ultrasonography among health population in 70,000. You can see 5-Fib4 index distribution also depends on age. The percentage of those who 5-Fib4 index are more than 1.3 and are recommended for referral to hepatologist is 3.6% in 50 under 50 years old. And as for over 70s in the right side, 88.1% is the percentage of recommended to referral to hepatologist. If we use 5-Fib4 index for screening, most of them already become eligible for referral to hepatologist. So this is the ROC curve in prediction of liver stiffness by Fibroscan compared using the type 4, 7S and 5-Fib4 index. Type 4, 7S seems to have a slightly better accuracy in liver stiffness evaluation than 5-Fib4 index. So summary, surprisingly, the prevalence of fatty liver in the general population may reach 52.9% and 1% of general population were found in hepatosis by Fibroscan. We need primary screening test for liver fibrosis which is easy and applicable to a large population but 5-Fib4 index seems not suitable for liver disease dependency and it may overestimate indirectly among healthy people. It seems our study, type 4, 7S is one of the good fibrosis markers but further study should be still required. Thank you very much for your attention. Hello and thank you for inviting me for this talk on NASH-induced HCC treatments controversies and unknowns from the ASLD and Japan Society of Hepatology Joint Symposium. I'd like to thank the course directors for this invitation. I'm a transplant hepatologist and associate professor of medicine at the University of Chicago Medicine. I am the director of our multidisciplinary liver tumor team. I'm going to start with this graphic which I'm sure you've seen before. As my colleagues before me have likely shown, the worldwide prevalence of NAFLD and NASH are rising. We know that HCC is an important and growing concern. Unlike viral hepatitis, one of the biggest reasons is that 90% of HCCs developed due to metabolic syndrome are NAFLD and NASH without cirrhosis. Again, striking contrast to viral hepatitis. Secondly, subcutaneous fat and liver steatosis in obese patients hinder ultrasound examinations in patients who undergo surveillance. This may delay earlier stage cancer diagnosis. These are some of the things that we need to talk about as we think about what we should do with HCC and NASH. I'm going to address several controversies. First is going to be NASH-related HCC and the use of immunotherapy. This was a recent study that generated some controversy. This was published in Nature by Feister and colleagues. They looked at a preclinical mouse model that you can see on the right of mice with NASH and HCC. They were allocated to anti-PD-1 immunotherapy or control arms. Interestingly, as you can see, for the mice in the PD-1 arm, none actually had regression of tumor. In fact, most had or several had increase in tumor size. Additionally, you can see on this graphic that patients, I'm sorry, the anti-PD-1 treat in mice liver tumor tissue also showed an increase in CD1 and PD-L1 T cells. And this has also been shown in other preclinical models of obesity. And so the conclusion in this preclinical model was that naturally the HCC had decreased response to immunotherapy. Now, the same logic was used in a subsequent meta-analysis of the three randomized phase three trials, Checkmate 459, Iambray 150, and Keno 240, and looking at testing inhibitors of PD-L1 and PD-1, more than 1600 patients. And it showed that, as you can see, that immunotherapy actually did not improve survival in patients with non-viral HCC compared to viral HCC. And two additional cohorts that you can see on the left, you can see that, again, overall survival was dramatically decreased in the NAFL cohort compared to those of other ideologies. Now, in this separate study, investigators examined the impact of obesity on T cell response, and they demonstrated a significant impact of obesity on the PD-L1 axis. And so here, we're looking at tumor volume of B16-fed melanoma between control and diet-induced obese mice. And as you can see, in the PD-L1 arm, you can see a decrease in tumor volume. And so because of this, the investigators noted that the PD-1-mediated T cell dysfunction in obesity remarkably leaves these tumors more responsive to checkpoint blockade. And they looked at 250 human cancers and patients treated with PD-L1 checkpoint blockade and stratified by BMI. And as you can see, in those patients with BMI greater than 30, both progression-free survival and overall survival were better. So again, this is differing from the other studies showing that these tumors are more responsive to IO across multiple species and tumor models. Now, in this study by Ho and colleagues, this was a meta-analysis looking at overall response rate between viral and non-viral HCC. And they looked at the cancer genome atlas and the tumor microenvironment. And you can see here that on the six studies of about 567 patients, there was really no difference in overall response rate between the two groups. And they also did not see any relationship between viral ideology and features of a tumor immune microenvironment. And the tumor mutational burden was the same between virally infected and infected patients. And again, this shows that the non-viral group did just as well as the hep B or hep C groups. Now, this was a recent meta-analysis that was published in Gastroenterology of eight randomized clinical trials, about 3,700 patients. And they looked at both immune checkpoint trial inhibitor trials, as well as TKI anti-VEGF trials. And here again, you can see that the red is the viral arm and the green is a non-viral arm that suggested that in this meta-analysis, immunotherapy may be less effective in non-viral ideologies than in viral related HCC. So the first question is, do we have enough data to defer the use of immunotherapy in HCC, I'm sorry, in NASH related HCC? And the answer is no, we do need more data before we can make that recommendation. Now, our next controversy is looking at NASH related HCC and local regional therapy. Now this is the HCC NAFLD Italian study group, and they were looking at clinical patterns of HCC in NAFLD. This was a multi-center study and looked at 145 patients with HCC from NAFLD versus 611 patients with HCC due to hepatitis C. And as you can see, the size of the tumor size infiltrated tumors and tumors out of Milan were more commonly seen in patients with NAFLD versus hepatitis C, and also more commonly seen that these cases were detected outside of specific surveillance. So again, indicating that these tumors tend to present later and maybe that our surveillance strategies are not quite what they need to be. Now, looking at different local regional therapy modalities, the first one I'm going to talk about is SIRT or selective internal radio treatment. And in this retrospective single center study, Schotten colleagues looked at the impact of NAFLD related HCC with SIRT compared to viral related HCC. And they had 87 patients with NAFLD versus 67 with hepatitis B. And the only difference as highlighted on the left is that the size of the nodules were slightly higher in patients with NAFLD as I've shown you in this previous study. And again, you can see there was no difference in overall survival of NAFLD versus hep B using Y90, which was used in this specific case. In another retrospective study, Wong and colleagues looked at overall survival of, this is the NACIR database, and looked at the use of radiofrequency ablation for HCC and looked at NAFLD versus viral ideologies and alcoholic liver disease. And here you can again see that there was no difference in the use or outcomes of radiofrequency ablation and survival in patients with NAFLD HCC versus other ideologies. And now looking from radiofrequency ablation and SIRT to taste, this is a retrospective study by Young and colleagues. They looked at overall survival and time to progression in NASH HCC versus non-NASH HCC. And again, they noted that NASH cohort had larger lesions, but again, you can see there was no significant difference in either overall survival or time to progression. And also the number of complications did not differ. Now, interestingly, in contrast in this study by Wu and colleagues, which was also a retrospective study of about a hundred procedures performed though only in 57 patients, they looked at patients with a higher versus lower BMI and the higher BMI was cut off as BMI greater than 25. And what they noted is that in patients with higher BMI compared to those with the lower BMI, they had higher progressive disease and more residual disease. And so what this diagram is to show that post-taste in panel B, the lesion was actually bigger and there was an additional lesion. So the question is, is there data to recommend a specific regional therapy modality in NASH related HCC? And the answer is no. What we really need is improved surveillance tools and biomarkers to predict those at higher risk of progression to HCC in this population. And these are important areas for future research. Third controversy is NASH related HCC and surgical options. Is there a difference in outcomes when you compare these patients to others? I'm going to first start off with this retrospective cohort study, looked at patients who underwent ultrasound exam for cirrhosis related indication. And as you can see, nearly one in five ultrasounds were inadequate quality for HCC surveillance in patients with cirrhosis. And in a multivariable analysis, this is more prominent with patients with higher BMI and also with NASH cirrhosis. So again, we have to find a better surveillance strategy for these patients. Now looking at different surgical options, this is a multi-center Italian study, and they were looking at liver resection for HCC in patients with metabolic syndrome, and they compared it to HCV related HCC. And you can see on the far right, there was no difference in recurrence-free survival, or in the middle, overall survival between the two groups. And when you look at the union multivariate analysis over here, you can see there was also no difference in overall survival or recurrence-free survival in those factors. Now, when you look at parameters of predictive of severe morbidity and outcomes, what you can see is mouse core greater than eight cirrhosis rather than steatohepatitis and major hepatectomy for portenders of higher morbidity and liver failure rates. So again, to show that liver resection can be safe in patients with metabolic syndrome related HCC. Now, this European liver transplant registry analysis looked at the overall survival for NAFLD-HCC compared to others. And you can see on the bottom right, the NASH-HCC is slightly lower when compared to overall survival and compared to alcohol related, but not different from HCV related or cryptogenic cirrhosis. And here you can see there is no overall difference in NAFLD-HCC compared to non-NAFLD-HCC in these two groups. Now, this is a study looking at the long-term outcomes of patients undergoing simultaneous liver transplantation sleep gastrectomy by Julie Heimbach and colleagues published in Hepatology. And so what they did was they allocated patients, obese patients, as BMI greater than 35 or less than 35 and used a non-MESA weight loss program as well as combination of liver transplant and sleep gastrectomy if they failed the weight loss program and maintained the BMI greater than 35. And you can see here there was 45 patients that underwent liver transplant and 29 that underwent the combination of liver transplant and sleep gastrectomy. And you can see here on the right that the percentage of total body and weight loss was significantly less in the combined group versus the liver transplant alone group. And you can see on the graphic on the left that NAFLD, there was three patients with NAFLD and HCC. And what this study showed in the long term is three years post-transplant, those in the combined group actually had less hypertension or insulin resistance and other metabolic syndrome as well as hepatic steatosis. So something else to think about when we're thinking about how to reduce the risk of HCC in this population. Now this study is a little different and it's a cohort study looking at liver transplantation, I'm sorry, looking at laparoscopic sleep gastrectomy, pre-liver transplantation in morbidly obese patients. And then like the other study which had a BMI cutoff of 35, this has a BMI cutoff of 40. And you can see in this cohort of 32 patients, five patients at HCC all were within Milan. So once the patients underwent laparoscopic sleep gastrectomy, they waited on average about six months before they were eligible for transplantation. Now all five patients with HCC were able to undergo successful liver transplantation. Importantly, there was no significant perioperative deaths or liver related mortality. Okay. What you can also see that the percent of patients in the middle who had a BMI less than 40 over time after sleep gastrectomy, as well as the change in MELD after sleep gastrectomy over time. Now this study was aimed to determine the differences in severity of liver dysfunction, HCC diagnosis and long-term survival between patients undergoing curative therapy for NASH HCC versus hep C or alcohol related HCC. And what you can see here from the graphic on the right is there was really no difference in curative treatment and that included resection, transplantation or ablation between the two groups. And on the left, you can actually see that patients with NASH actually had improved overall survival. So this was one of the few studies that showed this, that these patients not only did well, but actually did better. So this just recently published review in a journal of hepatology looked at a different NAFL-driven HCC and different modality to see what each modality and their outcome in this patient population. And as you can see, when you look at both liver resection and liver transplantation was either better or similar to those of other non-NAFL ideologies of HCC. So do patients with NASH related HCC have better or equivalent outcomes with surgical options, including resection, bariatric surgery and liver transplant? The answer is yes. Now the last few minutes, I just want to talk about NASH related HCC prevention and the role of caffeine, statins and metformin. Now, this is a study here that was systematic review and meta-analysis by Kennedy and colleagues, and they looked at caffeine as a dose dependent effect on the reduction of HCC risk. And what you can see here is that there was a decrease in all these cohorts in case control studies with coffee. And you can see on the right, there was an actual 35% reduction in the risk of HCC with coffee. Now there were no randomized control trials and they may also be publication bias, but this is looking pretty favorable when we look at coffee. Now, the next thing I want to look at is metformin and the role in reducing HCC risk. The data is really in patients with type 2 diabetes. Now this is a Taiwanese study that looked at both MASH and un-MASH cohorts. And the MASH cohorts were about 20,000 users, metformin users were about 20, I'm sorry, 21,900 metformin ever users versus non-users. And they looked at a propensity score. And what they noted is that patients that use the metformin actually had a reduced risk of HCC in a dose response pattern. And this was similarly also shown not here is in a study of almost 200 patients from Indiana University looked at the association between long-term metformin and HCC risk and all cause mortality and found that it was also very similar. Now looking at HCC and statin or the risk of HCC risk when statins are used, this is a systematic literature study of literature search looking at that. And you can see the graph on the right looked at patients with type 2 diabetes, and you can see that it showed there was a decreased risk across all studies when looking at statins in patients with type 2 diabetes on HCC risk. And the graphic on the left looked at the same idea about patients with cirrhosis. So both those look like there was a decreased risk with the use of statins. However, when you look at the randomized control trials, what you see is that there's actually no statistically significant difference in patients that had statins versus no statins. So how does it stack up as far as prevention of HCC with coffee? It's likely yes. Metformin in patients with type 2 diabetes may be to possibly yes. And with statins, we just don't have any data yet, convincing data yet, I should say. So my key takeaways of this talk is that one, as we know, NAFLD and NASH-related HCC is on the rise with 20 to 50% occurring without the background of cirrhosis. We know that improved surveillance of parameters and biomarkers especially are needed to find and identify these patients at highest risk of developing HCC in this population. It's too early to predict the impact of immunotherapy on NASH-related HCC. And at this time, I would not hold immunotherapy in this patient population. There is no clear consensus on which local or regional therapy should be used for NASH-HCC. And surgical outcomes, including resection, bariatric surgery, and liver transplant have equivalent and then sometimes better outcomes in patients with NASH-related HCC. And although coffee has shown to decrease HCC risk, so for prevention, the role of metformin except maybe in patients with type 2 diabetes and statins still remains controversial and unknown. And that is it. Thank you very much. Hello, everyone. It's my great pleasure to speak to you. And before starting my presentation, I would like to thank the organizing committee for inviting me to this symposium. This is my biography. Here are my disclosures. Today's my talk has four parts. As an introduction, I'd like to show you how non-viral HCC is increasing in Japan, reflecting obese population and potentially non-alcoholic fatty liver disease. Then, I'm going to describe the characteristics of non-viral HCC based on a national-wide survey. Thirdly, I'd like to review studies regarding the incidence of HCC in NAPLD and NASH. Finally, I'd like to talk about how to conduct surveillance for non-viral HCC. This slide shows the proportion of background liver disease of HCC in Japan. For the last three decades, non-viral HCC, as indicated with red bar, continued to increase. In 2020, the proportion of non-viral HCC exceeded that of SCB. The increase in the proportion partially reflects the decreased number of SCB-related HCC. However, when we calculated the number of non-viral HCC based on national-wide survey, there was an approximate five-fold increase from 2,200 in 1992 to 13,000 in 2015. The most probable cause of increase of non-viral HCC is shown here. These figures show the proportion of overweight or obese population in Japan, according to gender and age groups. As you can see here, the overweight and the obese population has been increasing in the last three decades in all age groups in Japanese males. The fatty liver is also increasing in Japan. This slide shows the proportion of fatty liver in the health checkup center. As you can see, the proportion of males with fatty liver increased in the last two decades from 50% to 35%. The proportion also increased in females from 10% to 15%. The incidence of fat delivery is known to be influenced by genetic factors. Among various genetic polymorphisms that affects the sensitivity of fat delivery disease, PNPR3 is the most studied one. This figure shows a global distribution of disease susceptibility areas of PNPR3 shown in light blue. I'd like to draw your attention that approximately 40% of Japanese have the disease-sensitive area. So in summary, while SCV-related CC is decreasing, non-valid CC is increasing due to the rise in obesity, diabetes, and fatty liver disease in Japan. I'd like to introduce Inuyama Noboru's study, a much-anticipated study on clinical characteristics, treatment, and prognosis of non-valid CC involving 53 high-volume centers in Japan. We unload naïve CC with non-valid theology defined as negative for both HBS antigen and anti-SCV diagnosed at participating hospitals from 1991 to 2010. Specific liver disease like autoimmune hepatitis, PBC, and butyric carcinoma are also included. We collected various clinical parameters as shown here. I'd like to show you the result. We collected the data from more than 5,000 patients. This slide shows the distribution of alcohol consumption by gender. Whereas most female patients were non-drinker, there are variations in the amount of alcohol consumption in males. The proportion of male patients who drank less than or equal to 20 grams per day was 40%. Heavy drinker defined as more than 6 grams per day was also 40%. This slide shows the proportion of NAFLD, alcohol-related disease, and unclassified groups determined by attending physicians. The proportion of NAFLD was only 11%, probably because this is a retrospect study based on a chart review, and NAFLD was not well-recognized 20 to 30 years ago. We performed a follow-up survey collecting data from 2011 and 2015. The proportion of NAFLD increased to 50%. You may notice that 40% were categorized as unclassified. When we check the details of characteristics of this group, many of them were complicated with type 2 diabetes or hypertension, suggesting the influence of metabolic syndrome. This slide shows the proportion of comorbidities associated with metabolic syndrome in non-viral CC. Those with obesity, diabetes, and hypertension increased in the recent survey as indicated in dark blue. You may notice that the proportion of this behavior is relatively small compared to other factors. This is due to decreased cholesterol production in cirrhosis. The proportion of apparent fatty liver is rather small. In the recent survey, we checked the presence of fatty liver in the past. The results showed that the fatty liver disappeared in 7% at the time of diagnosis of CC. This is a schematic image of NAFLD progression. As shown here, the fat content tended to decrease according to the progression of fibrosis. The final status of liver cirrhosis with decreased or vanished fat is known as burned out NASH. This is quite important when you consider surveillance for NAFLD in NASH population because to find a fatty liver is not a good strategy. We performed a hierarchical clustering using various clinical parameters, including laboratory findings, age, gender, presence of metabolic factors. The non-viral CC was categorized into five clusters. Cluster one as moderate drinker with good liver function. Cluster two as moderate drinker with obesity and fatty liver. Cluster three as young heavy drinker complicated with liver cirrhosis. Cluster four as elderly obese female with liver cirrhosis. And cluster five as male non-drinker complicated with metabolic syndrome. So, in summary, non-viral CC are quite heterogeneous. However, they can be divided into several groups by risk factors like old age, moderate to heavy drink alcohol consumption, obesity, and fatty liver. This slide shows the NAFLD pyramid structure. As shown in the previous slide, the prevalence of NAFLD in the general population is 20 to 30%. Approximately 10 to 20% of them develop NASH. Among them, 10% will develop liver cirrhosis. Consequently, the annual incidence of NASH SCC is extremely low as at 0.006% in the general population. This slide shows the major cohort studies regarding SCC development in NAFLD and NASH. The blue table contains the population-based studies and loading NAFLD, and the red table contains the hospital-based studies and loading NASH with advanced fibrosis. Let me introduce the largest population-based study shown here. The authors enrolled more than 10,000 patients with biopsy diagnosed NAFLD. They selected age and gender matched control at a ratio of one to five. They collected events from nationwide registry. The red chart indicates the number of deaths per 1,000 person years. The presence of NAFLD increased the risk of all-cause mortality. When you look at the breakdown of the mortality, the largest impact was from malignancy from organs other than liver, followed by cardiovascular disease. Although the NAFLD increased the risk of cirrhosis-related deaths and liver cancer-related deaths by 18.2-fold and 11.1-fold, respectively, the relative impact on all-cause mortality was small. How about those with liver cirrhosis? I'd like to introduce the two landmark studies. These figures indicate cumulative SEC incidents by Kaplan-Meier measures. As shown here, NASH cirrhosis showed lower incidence of SEC than HCV-related cirrhosis. The annual risk was estimated to be 2.5 to 4 percent. I'd like to show you a more recent study with advanced fibrosis. The authors conducted an international cohort study to allow biopsy-diagnosed NAFLD with F3-4 fibrosis. The left panel indicates the cumulative incidence of SEC divided by fibrosis grade and childbirth class. As you can see, the incidence was clearly stratified by the fibrosis grade. The RNA incidence in F3, F4A5, and F4A6 was 0.3 percent, 1.8 percent, and 4.7 percent, respectively. I'd like to emphasize that risk of SEC can be further divided by the grade of fibrosis progression, even in the state of cirrhosis. The right panel indicates the risk stratified by fibrosis grade and fat content. And the inverse relation between the degree of fat content and the risk of SEC development was observed. So, in summary, only NASH cirrhosis can be a candidate for SEC surveillance. Since liver fat content and liver fibrosis are inversely related, finding a fatty liver is not a good strategy. This slide shows the proportion of risk factors in novel SEC from a nationwide study in Japan. They are old age, male sex, overweight and obesity, heavy alcohol intake, diabetes, hypertension, fatty liver, and cirrhosis. Again, you may notice that the proportion of patients with apparent fatty liver was as small as 24 percent. This is probably due to Barnum-Nash, as was mentioned in the previous slide. This time, I'd like to focus on diabetes meritus as a potential candidate for SEC surveillance. This slide shows the risk of cancer-related deaths from various organ cancers in patients with diabetes compared to control population. Regarding the adjusted risk, liver showed the highest risk at 2.2 among other organ cancers. The Japanese Diabetes Society conducts a nationwide survey regarding cause of death in Japanese diabetes patients every 10 years. The latest results are shown on the left panel. Approximately 38 percent of Japanese diabetes patients died from malignancy. Among various organ cancers, the liver was ranked second after the lung, accounting for 6 percent of all cause of death. The mortality from malignancy in the general population is shown on the right panel. Now, the liver was ranked sixth in 2016. As shown here, diabetes patients' liver is a relatively important organ as a site of cancer. This slide shows pros and cons of surveillance. Where surveillance is beneficial, when it can reduce mortality by early diagnosis of a cancer, surveillance may have disadvantages as listed here. It is considered that from a viewpoint of cost effectiveness, an annual risk of at least 1.0 percent, ideally 1.5 to 2 percent, is necessary for starting surveillance. In response to the results, the Japan Society of Hepatology and the Japan Diabetes Society decided to conduct a collaborative project to elucidate the actual profile of hepatocellular carcinoma in diabetes clinics. This is a scheme of the study. It is estimated that enrolling 10,000 patients with observation period of 10 years is necessary to have 100 events considering the annual risk of SCC in diabetes patients. Therefore, we decided to use it as a cohort study and a case series. Project cohort was used to estimate the baseline SCC incidence, and the risk factor with SCC development in five years were analyzed, comparing SCC patients at diabetes clinics and those who did not develop SCC in the cohort study in five years. The detailed characteristics of two cohorts are shown here. The SCC cohort consists of 239 patients with type 2 diabetes who are diagnosed with non-valid SCC between 2010 and 2015, with more than five years of follow-up at diabetes clinics of 81 teaching hospitals in Japan before SCC diagnosis. The project cohort consists of 3,277 non-valid non-SCC T2DM patients were followed for an average of 4.7 years in the Mark Center cohort study. The figure shows the cumulative SCC development in the prosthetics cohort of patients. The annual risk incidence was calculated at 0.1%. This chart shows the characteristics of SCC patients at diagnosis. The mean age was 72.8 years. 80% were male. The median duration since the initial diagnosis of diabetes was 12 years. The mean value of VML was 25.5 kilograms per square meter, and fatty liver was diagnosed in approximately 5%. We also observed the proportion of fatty liver decreased within five years of the initial period. Hemoglobin A1c was 6.9% on average, and the value decreased in five years, reflecting decreased hemoglobin levels. Femur size was larger than 3 centimeter in 54%. Back spiral invasion and extrahepatic metastasis was observed in 50.1% and 8.4% respectively. This slide shows the adjusted hazard ratios of risk factors of SCC in diabetes patients. DDT and FIFO index were analyzed in quarters. Male sex, higher VML, lower hemoglobin A1c, lower albumin, higher DDT, higher FIFO index, and hypertension were statistically significant risk factors. HbA1c showed an inverse relationship because decreased hemoglobin levels were observed in HPC patients probably due to advanced cirrhosis. Among risk factors, FIFO index showed an outstanding performance. We calculated annual risk of SCC according to various cut-off values of FIFO index. We found that FIFO index of 3.5 or more, which is corresponding to the annual risk of 1%, may identify potential candidates of SCC surveillance in diabetes patients. This is the final slide. Non-valid SCC is increasing in Asian Pacific region, reflecting increasing obesity. Liver fibrosis determines the risk of SCC in NAFLD, whereas FAT content shows inverse risk. FIFO index is usually useful to identify at-risk population in type 2 diabetes patients. Thank you for your kind attention. Hello, everyone. My name is Tatsuya Kanto, a board member of the International Affairs Committee of JSH, or the Japan Society of Hepatology. First of all, I would like to express my appreciation to the speakers for the valuable contribution to the joint symposium between ASLE and JSH on the occasion of the Digital Liver Meeting 2021. This joint symposium between two societies has begun in 2019, one year before the COVID-19 pandemic. The main purpose of collaboration was to strengthen the relationship between us and to promote all members actively participating in the annual meetings and building social networking, not only by academic activities, but also comfortable friendship. However, in 2020, we were forced to refrain from going abroad and interacting each other in person. The last year, we had two virtual meetings at each side, which I believe drew much attention from clinicians and researchers from both sides. We appreciate for the members of active participation and fruitful discussion at that time. Today, we have two speakers each from JSH and ASLE, and they talked about epidemiology and screening strategy of NASHCC, which highlighted similarities and also dissimilarities in NASH patients between two countries. On Japan side, Professor Junko Tanaka talked about the high prevalence of NAFLD and its annual changes in Japanese general population. Diagnosing fibrosis stage is critical for the management of NAFLD patients, and Professor Tanaka showed the impact of FIP4 and Fibroscan for the assessment of liver fibrosis, by which she showed the surprisingly high prevalence of liver cirrhosis and advanced fibrosis in Japanese NAFLD patients, which was one to three percent. Next, Professor Tateishi talked about the increasing trend of non-viral SCC in Japan, which could be subdivided into several groups according to their lifestyle or metabolic conditions. He proposed a screening strategy of NASHCC from NAFLD patients. Of course, liver cirrhosis is a significant risk factor of SCC. He clearly showed that FIP4 is a good indicator of high-risk population of type 2 diabetes of NAFLD patients. I believe that the talks from two JSH members could provide useful information and help getting insights of the situation of NAFLD in Japan. Before ending my closing remarks, I would like to convey a special thanks to the SLD staff, who have been extremely cooperative in making the joint session happen. And finally, my deepest thanks are of course reserved for the attendees of members from both societies, and I wish we can get together in person at the next occasion of this joint session. Thank you very much.

non-alcoholic fatty liver disease

NAFLD

hepatocellular carcinoma

HCC

liver fibrosis

cirrhosis

surveillance

screening

biomarkers

FIP-4

Fibroscan

metabolic syndrome