My name is Shadab Siddiqui from Virginia Commonwealth University and I'll be moderating this webinar. We actually have a very exciting webinar where we're going to delve into the liver heart connection looking at cardiac outcome physiology and potential endpoints in clinical trials and we have a very exciting set of panelists who will be doing that. I'd like to take this opportunity to introduce the panelists. First we'll have Dr. Michelle Long from Boston University. She is the director of NAFLD research center as well as the director of clinical research in section of gastroenterology, hepatology. She will be talking about the physiological changes in cardiovascular system in patients with NASH. Next we'll have Dr. Kathleen Corey from Mass General who is a program director for NAFLD clinic and has a long history of publications in fatty liver disease who will be talking about the clinical impact of cardiac impact clinical parameters in patients with fatty liver disease and then we'll conclude by Dr. Adam's talk who's an associate professor at the University of Western Australia and has published extensively in fatty liver disease and he will be talking about potential cardiac endpoints in clinical trials. So without further ado I will hand it over to Dr. Long to begin the webinar. Great thank you so much. It's my pleasure to be here to talk to you about one of my favorite subjects. So I will focus on the physiological changes in the cardiovascular disease system in patients with a non-alcoholic steatohepatitis. Here are my disclosures. So the interaction between fatty liver disease and cardiovascular disease is very complex and as you'll see in this presentation, my presentation, I'm sure the others will address this, that there is a lot of overlap between fatty liver disease, metabolic syndrome, diabetes, and cardiovascular disease and sometimes and often by the design of the studies we're unable to tell which is causing which. Many of these studies that we see are cross-sectional but the overwhelming thought is that we think that there is this interplay kind of goes back and forth and that there are inflammatory mediators that may make fatty liver disease worse and may also in turn worsen cardiovascular disease and their fatty liver is intimately connected with metabolic syndrome, obesity, and cardiovascular disease. What are some of these mechanisms? Well when you have an expanded and inflamed visceral fat that will lead to increased inflammatory cytokines and worsening your insulin resistance and increased uptake of fatty acids by the liver which will cause fatty liver disease and which in turn may also increase inflammation and worsen insulin resistance and this vicious cycle ultimately lead to a state of chronic inflammation and you may have hypercoagulation and hypoherbinolysis, atherogenic dyslipidemia, and dysglycemia and insulin resistance and together those mechanisms may lead to cardiovascular disease, atherosclerosis. So now I'll talk about some of the evidence for so the increased inflammation that we see in patients with NAFLD-C. So this is a project that we did in the Framingham Heart Study. So Framingham Heart Study is a multi-generational longitudinal cohort study of initially enrolling adults without known cardiovascular disease and then it tracked over many years the development of cardiovascular disease and in this study inflammatory markers have been measured over time and we also have CAT scans which have defined fatty liver disease. So in this project we looked at a cohort of Framingham participants that were completely unselected for liver disease but who had undergone CT scans. So we were able to characterize about you know 20% of the 3,000 or so participants who had increased liver fat and then we also looked at various markers of inflammation and we found a number of markers that were associated with CT-defined liver fat including IL-6 and highly sensitive CRP, ICAM-1 which is an intercellular adhesion molecule and p-selectin both of which are associated with atherosclerosis and urine isoprosteins which is the marker associated with oxidative stress and we found that these particular markers were all associated with CT-defined liver fat and even if we accounted for other cardiovascular risk factors including visceral fat or a general measure of adiposity body mass index we we noted that they are remained in association and then what we did was we created an inflammatory index. So we looked at the number of elevated inflammatory markers an individual had over them over the median value and determined what the prevalence of CT-defined liver fat was and and we noticed a very nice stepwise increase in the prevalence of liver fat for the number kind of burden of inflammation that an individual had depending on how many inflammatory markers were elevated. What about in in those with more advanced liver disease? So this is a study that looked at NASH and different fibrosis stages so you know as we know highly sensitive CRP is a strong predictor of future cardiovascular disease events and we others have shown that CRP is elevated in persons with NASH compared to those with simple steatosis so in this study they had a cohort of biopsy confirmed patients with NASH at different fibrotic stages and they noted again that as the fibrosis stage is increasing from zero up to four the higher there were higher values of CRP associated with more advanced NASH related fibrosis which may be you know maybe a mechanism behind some of the increased cardiovascular events that we see with various fibrosis stages. What about endothelial dysfunction? So this is a measure of subclinical cardiovascular disease and how do we measure it? Well one way is to measure something called a flow mediated dilation and basically you induce a stress over an artery typically the brachial artery and that stress is the inflation of a blood pressure cuff for a set amount of time and what you expect is that a normal blood vessel is going to respond kind of appropriately to that stress by dilating and when you have a endothelial dysfunction you'll see that the normal response will not be there or will be diminished and we know that this this measure is associated with different conditions including smoking, high cholesterol, hypertension, and metabolic syndrome. What happens when you measure flow mediated dilation in patients with NASH? So this is a study out of a single center in Italy that had 52 patients with NAFLD and 28 with control and they also had a subset with NASH and at baseline there was no difference in the brachial artery diameter between the groups but if you looked at the lowest tertile of the flow mediated dilation that was associated with a six-fold increased risk of having NASH and this was even after we counted for age, sex, body mass index, and insulin resistance. So some of the main kind of cardiovascular disease risk factors that are associated with NASH. So this suggests that there is endothelial dysfunction in patients with NASH. What about atherosclerosis? So you can measure the carotid artery intimal media thickness and this is these figures here show the intimate media thickness in millimeters in healthy controls versus those with simple fat and those with NASH. We see that the intima media thickness is higher in those with NASH and in particular if you look at panel B here by various stages of NASH fibrosis we see a stepwise increase in the intima medial thickness suggesting a higher burden of atherosclerosis. NASH is also associated with cardiovascular remodeling. So this is a study that had 171 patients mean age 40 they were pre-bariatric surgery so their BMI was over 40 close to 45 and they had 44 patients with NASH, 69 with steatosis, and sort of interestingly 58 with neither. And mostly they were non-advanced fibrosis only seven with F3 and there were no patients with cirrhosis. And what they did was basically these patients were coming in going to have bariatric surgery they had an echocardiogram as part of the study in the days leading up to the surgery and then during their procedure they underwent a liver biopsy and and that was how we got the pathological results that I just described. And they found that those that had NASH had an increase in their cardiac output and also their cardiac index which is your cardiac output index for your your weight and they also had an increase in their relative ball thickness and it should be a decrease in the left ventricular end diastolic diameter which suggests a concentric remodeling of the heart. And this is actually a response that is often seen in response to hypertension, insulin resistance, hyperglycemia, and systemic inflammation. So not surprisingly we see this in our patients with NASH because often they have these comorbid conditions as well. In this study of relatively young age participants there was no overt LB dysfunction. This is another study a similar setting that was in slightly older population but by no means old mean age of 47 and this was 65 patients also pre-bariatric surgery 14 who had NASH, 33 with steatosis. They underwent biopsies routinely during the bariatric procedure and in this study they had 74 percent had undergone a trans-thoracic echocardiogram as part of their pre-operative assessment and some participants 21 also had some form of dyspnea as an indication for the trans-thoracic echocardiogram. This was a retrospective study and they also found signs of LV concentric remodeling in this study but they also showed that there were some subclinical LV dysfunction. So they had increase in left atrial size and volume and left ventricular mass index and a decrease in the myocardial relaxation and decrease in mitral inflow velocity and these subclinical measures may ultimately lead to diastolic you know overt diastolic dysfunction which you know is also something that we worry about in our patients with NASH. So it may be that the increase in inflammation or the increase in the cardiovascular disease risk factors such as insulin resistance, hypertension leading to this chronic inflammation and LV remodeling may be a you know the mechanism behind an increase in risk of diastolic dysfunction in patients with NASH. We also see altered electrical conduction and arrhythmia particular atrial fibrillation. So this is a very nice study published recently in liver international that looked at a large clinical repository from Northwestern University and they had over 9,000 patients with ICD coding consistent with NAFLD or NASH. They also had a subset of biopsy proven NASH patients and what they did was they looked at the prevalence of atrial fibrillation which was about 5.6 percent overall and 4.2 percent in those less than 65 years in those among those that had a ICD-9 code consistent with NAFLD or NASH and the results looking in their biopsy proven cohort of NASH were very similar about four percent with concurrent atrial fibrillation and they compared this to the general population of two percent of prevalence of atrial fibrillation this is under 65 years old and also to a tertiary care center non-NAFLD NASH population of 2.7 percent so if you have NASH your risk of also having atrial fibrillation was essentially double so it may be that the cardiac remodeling is a risk factor for heart failure diastolic dysfunction and it also may lead to the development of arrhythmia. NASH is also associated with osteogenic lipid ratio so in this study 72 biopsy confirmed patients with NAFLD 36 with NASH 36 with simple steatosis 11 that were normal and we looked at various lipid ratios triglycerides HDL in the first panel cholesterol over HDL middle panel and the LDL to HDL ratio in the final panel and NASH and fibrosis were associated with atherogenic lipid ratios independent of obesity insulin resistance and diabetes so in summary there are several physiologic changes in the cardiovascular system that occur in patients with NASH now most of these studies are relatively small numbers and are also generally of a cross-sectional design so we don't know and cannot comment about causality but we think that an increase in systemic inflammation atherogenic dyslipidemia endothelial dysfunction cardiac remodeling and alter electrical conduction may all contribute to progressive progressive cardiovascular disease and we need additional studies to determine if improving NASH will also improve instant cardiovascular disease it's my pleasure to pass the mic over to Dr. Corey all right thank you so much that was fantastic it's my pleasure to be here today and we're going to switch gears a little bit and talk about cardiovascular disease outcomes and non-alcoholic fatty liver disease as well as risk factors for cardiovascular disease among our patients with NAFLD so as you all know and i think michelle has definitely convinced us all you know NAFLD and cardiovascular disease are intricately linked we know that patients with NAFLD have a decreased 10-year survival compared to the general population and that cardiovascular disease related mortality accounts for up to 25 percent of deaths in those with NAFLD but it still remains unclear if NAFLD is an independent risk factor for cardiovascular disease or does the concurrent medical disease metabolic disease that we often see with NAFLD lead to cardiovascular disease as you can see on the right our patients with NAFLD have very high rates of cardiovascular risk factors including obesity type 2 diabetes and the metabolic syndrome and perhaps those are driving this relationship but there have been different reports and an important one a meta-analysis by Giamatti Targer from journal of hepatology in 2016 concluded that patients with NAFLD compared to those without have an increased risk of fatal and non-fatal cardiovascular disease events but there were limitations to several of the studies that were included in this meta-analysis baseline cardiovascular disease and coronary heart disease were not universally assessed there was incomplete adjudication in some of the studies for cardiovascular risk factors and I'm sorry incomplete adjustments and there was a lack of adjudication for cardiovascular disease outcomes and so I'm going to talk a little bit about several studies that our group at Mass General has done to try to get at an answer of whether NAFLD is truly an independent risk factor for cardiovascular disease and then look at some of the risk factors. So what we decided to do was to compare rates in a randomized controlled trial and a nested cohort of incident major adverse cardiovascular events and you'll hear this a lot in my talk MACE these major adverse cardiovascular disease events are quite important and they're generally what's used and you'll hear more about this often in clinical trials and we wanted to see if these rates are different by steatosis status controlling for cardiovascular risk factors baseline atherosclerotic burden and use adjudicated cardiovascular disease outcome. So what we did was conduct a nested case control trial within the PROMIS cohort and the PROMIS cohort was a study of over 10,000 individuals without baseline cardiovascular disease who had stable chest pain who are randomized to functional testing versus coronary artery CTA. Now fortunately often with coronary artery CTA you capture the liver and the spleen at least parts of them and so you can look at the liver spleen ratio to determine if someone has steatosis and 3,756 did have this imaging. And we found that 25.5% had steatosis as we would expect in the population, and about 75% did not. We then went on to compare their rates of cardiovascular disease and major adverse cardiovascular events, which are defined as death, myocardial infarction, or unstable angina. Now at baseline, we found that steatosis was associated with a higher prevalence of metabolic disease, as you would expect, obesity, metabolic syndrome, type 2 diabetes, and importantly, a higher cardiovascular risk factor burden by the atherosclerotic cardiovascular disease, or ASCVD risk score. Now the ASCVD risk score is something that we control for a lot in studies, and it's important to realize that this composite score allows us to control for a number of variables, including age, gender, race, total cholesterol, presence of hypertension, systolic blood pressure, tobacco use, and type 2 diabetes. And this is a score that has been shown to be valid in the general population to predict cardiovascular outcomes and help clinicians determine whether a patient should be on statins, what their ideal blood pressure should be, et cetera. What we found at baseline is that there were very small differences in the atherosclerotic burden by steatosis. So when those patients entered the study and underwent their CT scan, there was very little difference in the burden of atherosclerotic disease, whether you had steatosis or not. There was no difference in high-risk plaque, calcified or non-calcified plaques. Those are the plaques that we generally worry about for major adverse cardiovascular events. Those with steatosis did have a slightly higher mean, what's called Lehman score, which is a score of total obstructive and non-obstructive burden, and slightly higher Agassiz scores, which is a measure of coronary artery calcification. Then when we looked at MACE, we followed patients for 25.5 months, and they had an overall rate of MACE of 3.1% over that time. What we did find is baseline steatosis was associated with significantly higher rates of MACE, 4.4%, compared to only 2.6% in those without steatosis on imaging. And the hazard ratio was 1.72, which remained significant after adjusting for significant stenosis, ASCBD risk score, so that's all of those metabolic components, obesity, and the metabolic syndrome. And even after we adjusted for every CT measure of plaque and stenosis at baseline, those with steatosis still had higher rates of major adverse cardiovascular events. Interestingly, the addition of steatosis seems to increase your risk, whether you have non-obstructive or obstructive cardiovascular disease. So we can see here in the orange line that those patients who have both non-obstructive CAD and hepatic steatosis have significantly higher rates. It has a ratio of 5.42 compared to those with just non-obstructive CAD. And we see the same relationship in those with significant stenosis. So the addition of fat in the liver increases the risk of major adverse cardiovascular events. So here we found that baseline hepatic steatosis was associated with a 70% increased risk of major adverse cardiovascular events, and it was independent of traditional cardiovascular risk factors and the presence or extent of CAD. Well, what we were interested in looking at is really delving into the patients that we care for most, those patients who have NAFLD. Do they have the same risk factors and should they be treated in the same way? And so we wanted to see our traditional cardiovascular risk factors valid in those with NAFLD, and are there unique risk factors that we need to be aware of? And we looked at this in radiographic NAFLD as well as biopsy-proven NAFLD. Again, we went back to the PROMIS trial, focusing on those patients just with steatosis who had MACE or not, and compared their baseline risk factors. Interestingly, age and gender, as we would expect, male gender, were significant independent risk factors for baseline cardiovascular disease. So cardiovascular disease of greater than 50% stenosis at baseline. And we also found that traditional risk factors like diabetes, hyperlipidemia, and something we probably don't pay enough attention to is hepatologist smoking was a very important risk factor for cardiovascular disease, and may be a risk factor, as we all know, for fibrosis, liver-related death, and transplant. So it's something important that we talk about with our patients. We also found that the ASCVD risk score was associated with significant baseline CAD. Additionally, lipids, antiprobian P, and high-sensitivity troponin were also associated with significant CAD. But most importantly, we're interested in what are risk factors for those hard outcomes, those major adverse cardiovascular events in NAFLD. And what we found was interesting. It showed that the ASCVD risk score was associated with an increase after adjustment by about 3% in major adverse coronary events for each one-point increase in the ASCVD score, showing that we can use this score to predict future outcomes in our patients. But importantly, the strongest risk factor that we found was sedentary lifestyle at baseline. And certainly that's important because it's a modifiable risk factor that we often probably underestimate, underemphasize over diet for our patients with NAFLD. So here we found that CAD severity and burden in NAFLD is associated with traditional risk factors as we would expect, including several modifiable risk factors like dyslipidemia and smoking, antiprobian P and high-sensitivity troponin, as well as lipid levels are important. But for predicting MACE, the most important things were that baseline ASCVD risk score and sedentary lifestyle. Now there are limitations to this study. It lacks histology. And so we know that some of these patients will have NASH and some will have steatosis or fibrosis, but we're not able to tell who. And so an additional study looked at identifying predictors of incident cardiovascular disease in biopsy-proven NAFLD. So we looked at 285 patients and looked at incident cardiovascular disease over a median of 5.2 years. Incident cardiovascular events occurred in 9.1% of patients. And we found the strongest risk factors, again, were smoking. So something that we can modify in our patients and counsel them about. Low albumin, suggesting that poor overall health or poor liver function was associated with cardiovascular events. And advanced fibrosis was strongly associated even after controlling for all relevant covariates. So stage three or stage four fibrosis was strongly associated with incident cardiovascular disease. Again, interestingly, no other component of NASH histology was associated with incident cardiovascular disease. So steatosis, hepatocyte ballooning, lobular inflammation, or the presence of NASH were not associated with significant CVD. On histology, only advanced fibrosis predicted cardiovascular disease events. And you can see here that incident CVD was significantly higher at 9.6% in those that are F3, F4, advanced fibrosis compared to those who are F0 to 2. And when we removed advanced fibrosis from the model and replaced it with a NAFLD fibrosis score, that score could also predict the incidence of cardiovascular disease. And so a higher non-invasive score of NAFLD fibrosis score was also associated with incident CVD, suggesting that while we're looking for fibrosis in our patients, we also should be counseling them about the risk of cardiovascular disease. So in conclusion, adults with advanced fibrosis at baseline had significantly higher rates of incident cardiovascular disease compared to those with less fibrosis. And among those with biopsy-proven NASH, smoking and advanced fibrosis was the highest predictor of incident cardiovascular disease. So overall, I hope that I've convinced you that hepatic steatosis is associated with a significant increase of major adverse coronary events independent of traditional cardiovascular risk factors when compared to those without steatosis and independent of baseline atherosclerotic burden on imaging and with adjudicated cardiovascular disease outcomes. So as we discussed with NAFLD, baseline CVD severity and burden is associated with traditional risk factors, but it's important to remember that we can use the ASCPD risk score to help re-stratify our patients and that sedentary lifestyle is an important and modifiable predictor of MACE in NAFLD. And further, advanced fibrosis should also give us pause and make sure that our patients are getting appropriate re-stratification and treatment for their modifiable risk factors. Further studies are certainly needed to understand this mechanism between fibrosis and cardiovascular disease, and those are ongoing. And with that, we will move on to our next talk and we'll be taking questions at the end. And I am happy to introduce Dr. Leon Adams of the University of Western Australia, who's gonna talk to us about cardiac outcomes in NASH clinical trials. Thank you very much, Kathleen, and thank you for the invitation to present today. This is my disclosure slide, although not relevant. And so the first question I was asked to address was should cardiovascular endpoints be included in clinical trials of NASH patients? And really to unpack this, I thought we should first address where the cardiovascular outcomes, or what are the cardiovascular outcomes in NAFLD, NASH, and cirrhosis patients? And are these outcomes likely to be influenced by NASH drug treatment? And when researching for this talk, I came across some lessons from type 2 diabetes, which I think are pertinent, which I'll discuss. And the second main question was if we should be including cardiovascular endpoints, how would we do that? And therefore, what is the appropriate cardiovascular endpoint to choose? And what sample size would we need? So Kathleen and Michelle have nicely outlined that cardiovascular disease is a major problem in our patients. And this cohort study is one of many which has demonstrated the high prevalence of cardiovascular death in our patients. Now, this was a biopsy study of actually over 10,000 patients who were followed up in a national health system for a median of 14 years. And you can see the cardiovascular death in this cohort was accounted for 28% of all deaths. Now, because it was a biopsy proven cohort, they're able to look at the prevalence of cardiovascular death according to baseline histology. And you can see that in the orange bar, cardiovascular deaths is responsible for a majority of deaths, but this does decrease in proportion with the more severe your liver disease. Where in contrast, the proportion of deaths related to liver disease increases as the severity of your liver disease increases. And we can see that in particularly in cirrhosis patients, liver disease becomes the dominant cause of death. And this does have implications if we're going to examine cardiovascular endpoints in this population, in that liver disease will become a competing event and that you'll likely require larger patient numbers to determine effect on cardiovascular disease. Now, there are a number of potential impacts that drug might have on cardiovascular outcomes. And so if we have a successful NAS drug, it might have a neutral cardiovascular effect, but due to its successful liver effect, it may result in a reduction in overall death in that patient population. Hopefully a successful NAS drug will have a beneficial effect on cardiovascular disease, and this may increase the overall benefit to the patient. However, some drugs may in fact have a detrimental cardiovascular effect. And even if this is a fairly marginal effect on cardiovascular risk, because cardiovascular disease is a common cause of death in these patients, it may actually change the balance of benefit and be responsible for an increased risk of mortality overall. So this brings us, what is the actual effect of existing NAS drugs on cardiovascular endpoints? So this table tabulates drugs which are suggested or recommended in clinical guidelines, namely vitamin E and pioglitazone, along with candidate drugs, which are finished phase two and entering or currently in phase three trials. And you can see that vitamin E and armchol have no effect on lipid or glucose profile, an unknown effect on blood pressure. Well, there's resmitron, semaglutide and nanofibrinol all have various beneficial effects on these cardiovascular risk factors of cholesterol and glucose levels. The odd one out is the beta-cholic acid, which really has a deleterious effect on LDL cholesterol and HDL cholesterol. So if we go into a little bit more detail in each of these agents, pioglitazone has been around for over a decade now in the setting of type two diabetes. And so there's good data demonstrating its effect on cardiovascular outcomes. In this particular recent meta-analysis demonstrated that pioglitazone reduced major adverse cardiac events, reduced non-fatal myocardial infarction as well as non-fatal stroke. However, it did actually increase heart failure hospitalizations, demonstrating that it's not just atherosclerotic risk that we need to take into account when we're assessing cardiovascular endpoint, but we also need to look at other endpoints such as heart failure hospitalization. And the overall effect of pioglitazone on all course mentality is actually null. If we look at vitamin E, now remember this had no effect on either lipid or glucose parameters, but this has been trialed in clinical trials as a preventer for ischemic and hemorrhagic stroke. And this meta-analysis did show that vitamin E actually reduced the risk of ischemic stroke, but conversely increased the risk of hemorrhagic stroke having a null overall effect. And similar to pioglitazone, the HOPE trial, which was a large randomized controlled trial of vitamin E on cardiovascular outcomes among others, did demonstrate that there was an increased risk in heart failure. So I think this also draws caution to the interpretation of the effect of drugs on surrogate outcome and endpoints, sorry, such as lipid parameters and glucose, that there may actually be unseen or off-target effects that are unanticipated. Now, beta-cholic acid is known to have adverse effects on lipid profiles due to its mechanism involving FXR receptor. However, these changes can be ameliorated somewhat by the use of statin. And this is illustrated by the study from Liver International, where they compared placebo with varying doses of a beta-cholic acid. And then after four weeks, patients were prescribed varying doses of atorvastatin. And you can see that this strategy certainly reduced the LDL cholesterol and had a variable effect on HDL cholesterol. This strategy, however, does increase cost of treatment and also may potentially increase toxicity. So when thinking about the cardiovascular endpoints for clinical trials, I think it's pertinent to look at other agents in terms of their regulatory experience. And the case study that I want to examine in this slide is the experience of rosiglitazone. And so rosiglitazone was approved for use in type 2 diabetes due to its effect on HbA1c back in 1999. However, eight years later, a meta-analysis of observational data on clinical trials suggests that it was associated with an increased risk of myocardial infarction. And so that year, the FDA had a black box warning on the drug, and this led to a plummeting in the sales of the drug or the use of the drug. And as a response, the FDA mandated that cardiovascular endpoint safety trials should be conducted for anti-diabetic drugs. Subsequent to this, the record trial was published in 2013, and this actually demonstrated that there was no increased cardiovascular events with rosiglitazone, and subsequently the black box warning was removed from rosiglitazone. However, really the drug never recovered, and pyoglitazone suffered to a degree as due to potential class effects. And so I raise this because if we do find that one of the NAS drugs does have signals for adverse cardiovascular endpoints, then I wonder whether the FDA will take a similar approach of mandating cardiovascular endpoint trials. This is all speculative, but I think it is prudent for us to keep in mind when we're designing these trials. One drug that did go through this mandatory cardiovascular endpoint safety trial was semaglutide. And now this is in patients with type two diabetes. And it did, these trials did demonstrate that there was non-inferiority for major adverse cardiac events as well as cardiovascular death. And so whilst this cannot be directly translated to patients with NAFL, it's certain that this population of type two diabetics will be enriched with NASH patients, which shows some promise. So the second part of my talk is, if we are going to include cardiovascular endpoints, what should we be looking at? And if we're thinking of surrogate endpoints, there really is a plethora of different surrogate endpoints we can use. We can use the traditional risk factors or biomarkers. There are anatomical measures of atherosclerotic burden, such as chronic calcium scores, or there's these functional tests mentioned by Michelle, and in particularly endothelial dysfunction. Notably, none of these surrogate measures are actually accepted currently by FDA for regulatory purposes. Now I wanted to just briefly go over endothelial dysfunction, which was nicely outlined by Michelle previously, but this is thought to be a early subclinical marker of future atherosclerotic burden. And the reason I was particularly interested in this when I saw this study was that performed in NASH patients where they gave 40 grams of dark chocolate to patients for two weeks, and then a control group got milk chocolate for two weeks, and they measured endothelial dysfunction before and after. And you see, rather gratifyingly, the dark chocolate had an improvement in endothelial dysfunction, which was significantly greater than those taking the milk chocolate. However, before we rush out and advise that all our patients should be on a diet of dark chocolate, it is pertinent to note that these are just surrogate endpoints, and the utility of surrogate endpoints has been illustrated in this interesting study where the authors looked at 151 positive clinical trials, which were using cardiovascular surrogate endpoints, and these were published in high-impact journals over a period of a decade. And then they followed these clinical trials to see how many of them actually followed up with a clinical endpoint trial, so using hard endpoints of hospitalization, death, et cetera. And they found only approximately a third of these trials had clinical endpoint trials, and of these clinical endpoint trials, only 24 actually were positive after an initial 151 positive surrogates. So just a word of warning that surrogates are surrogate only and suggestive of a direction or a signal, but need to be followed up. So if we're gonna use hard cardiovascular endpoints, such as the combined composite endpoint of major adverse cardiovascular endpoints, what kind of power would we need in the trial? And this is a back-of-the-envelope power calculation. If we use the serious adverse cardiac event rate in the placebo group from the REGENERATE trial, which was a trial of NASH patients over 18 months, their incidence was 2%. And if we're looking for an agent which reduces the risk of MACE by 25%, which was the same as semaglutide and type 2 diabetes, we would need over 6,000 patients enrolled for five years to give us 80% power to determine the effect. And so this is three times larger than the current phase three trials that are ongoing and certainly a major undertaking. And I would suggest something that's not going to happen in a hurry. There's also the problem of competing risk for liver-related death for cirrhosis patients, which I mentioned before. However, a possible strategy may be to include patients with prior cardiovascular disease to enrich your population with events. So what's the other way? Well, there is phase four post-marketing trials to monitor for cardiovascular endpoints. And these trials really look at the effectiveness of the drug and safety in the real world, as opposed to the efficacy of a drug in a tightly controlled clinical trial setting. Now there's challenges to performing these kinds of trials, particularly the standardization of data collection and reporting, ascertainment of the data, and also questions about who may own the data such as pharma or government. And lastly, who's going to fund these trials? Will it be pharma or will it be government? So in conclusion, cardiovascular disease is certainly the leading cause of death in the patient population that's currently enrolled in our phase three clinical trials. The cardiovascular effect of NAS treatments may well make or break candidate drugs. And it's important to realize that traditional atherosclerotic rejectors may well guide an impact on ischemic cardiovascular events, but not necessarily other cardiovascular endpoints such as hemorrhagic stroke or heart failure. And I suggest that cardiovascular outcomes in NAS trials may well likely be assessed in phase four due to the difficulty in getting sufficient power in phase three trials. Thank you for your attention. So thank you for an excellent talk. And I want to congratulate Dr. Adams, Dr. Long, Dr. Corey. Very well-given talks and very thoughtful. And with this part, now we're going to transition into the question and answer sessions. And there's a number of questions that have come up. And the first question I think is actually directed towards Dr. Long, but I think Dr. Corey can actually comment on this as well because we're sort of merging questions and Dr. Corey can actually comment on this as well because we're sort of merging the pathophysiology as well as the clinical context. And this is sort of, it's been alluded to in the talks as well, to what extent does patients' comorbidities have an impact on these cardiovascular risk factors and how well can we actually separate the liver from pre-existing pathophysiological mechanisms such as insulin resistance, which may then impact things like atherosclerosis and inflammation and so on and obesity. So I'll sort of let Dr. Long start and then Dr. Corey, if you can chime in after that, because I think it's important to sort of separate these out. Yeah, it can be very difficult to do that. We don't have a lot of studies that look at people that are pre-cardiovascular disease and follow them over very many years, which is what you would need. And also characterizing their liver fat and fibrosis along the way. So that study would be very time-consuming to do and would be very expensive as well. So we really don't have those types of studies. But I think the take-home point is that these patients, our patients with NASH are coming in, they probably have more cardiovascular disease than you think, and it needs to be addressed. And especially as a hepatologist, you may be seeing this patient more than others, or you may be the only one even thinking about cardiovascular disease in these patients. So it's important to calculate their ASCVD and also make very clear that if they need to be on a statin or it's recommended that they're on a statin, that their underlying liver disease should not preclude statin use. Yeah, I totally agree with Michelle. I think it's very hard to tease apart and say exactly what risk factors contributing to what, but we know from some studies that there does seem to be an increased risk in our NAFLD patients. And we really do want to make sure that they're, because they have NAFLD, and especially if they have advanced fibrosis, that they're not being undertreated for their cardiovascular disease risks. We have a study coming out that looked in patients who have lean NAFLD compared to obese NAFLD, and we did see equal rates of cardiovascular disease, suggesting that at least obesity is not the main risk factor. But a lot of those patients do have insulin resistance and dyslipidemia. So again, it's really hard to tease out and the take-home message is making sure that we really re-stratify these patients, because most likely cardiovascular disease is going to be one of their causes of mortality, even potentially less than liver disease. Yeah, and I think that's a really important point. I think sort of extending that to Dr. Adams, you know, when we sort of talk about this cardiovascular risk assessment and mortality, we're sort of talking about a combination of factors that can lead to that cardiovascular mortality, atherosclerosis risk, heart failure, arrhythmias, and so on. So as you were discussing, how do we incorporate these into regulatory practice and clinical trial design? You know, I think there is limited guidance on how do we actually manage cardiovascular risk, specifically in patients who have fatty liver disease as they go through these trials. Do you have any comments on how we can sort of optimize potentially? Because, you know, a number of us may be referring patients to it. So what to kind of expect, what the future may hold for these patients? Yeah, I guess in a regulatory framework, it's difficult to comment, but I am concerned, I guess, about the potential adverse effect on, you know, of the FXR, some of the FXR agonists on cholesterol. Even though the changes in the particle size of the, you know, the LDL cholesterol may not be as atherogenic as potentially we worry about, I still think that may raise a signal. And I think if one class of drug demonstrates that, then the experience with the rosiclitazone was that they mandated that all drugs do these large, expensive trials, which will necessarily prevent a lot of potentially successful medications being trialed. So I think this is something that, you know, that the people initiating and designing clinical trials really need to have, like you say, a really solid think about how the cardiovascular risk is going to be managed in a way that's done universally in that trial population and addressed well, because otherwise I think we may find ourselves without agents not being able to do trials. That's actually very true. And I think there was some questions that I'm going to try to combine regarding the atherogenic milieu and factors of glycoproteins, and this is for Dr. Long. How do you see the impact of these atherogenic factors as this relates to inflammatory cells, cover cells, and so on in the liver? What's the interaction that you can maybe shed some light on? I think it points to the mechanism here of why these things may be related, and the fact that we're seeing increased inflammation, we know that that's related to fibrosis in the liver, and we know that's related to increase in cardiovascular disease. So I think it sheds light on the pathway, and that potentially will help us think about how we are addressing this clinically, either by modifying risk factors or new agents. Be critical to kind of impact that inflammatory pathway. The next question is actually for Dr. Khoury, and I think it's something you published from the NASH-CRN trial, which is what happens to the lipid profile as we get fatty liver disease treated? Yeah, it's really interesting. I mean, when we looked at the NASH-CRN, we didn't see that most of our patients had those incredibly high LDL levels of 190, 180, and the averages are usually around 130, and sometimes that can be, although we know we'd like LDLs to be in the 60s or 70s, that can be a little bit falsely reassuring. And what we found is that while when you improve your NASH histology, you improve your triglycerides and HDL, but more importantly, you change your LDL profile. And so instead of having all small, dense, atherogenic LDL, you have more large, buoyant LDLs, which improve your atherogenic risk factor. And we also saw that reflected in an improvement in the Framingham risk score. So we do think that improving your NASH does change the type of LDL that you have, even if the absolute number in LDL doesn't change. Now, I wouldn't say that patients shouldn't be on a statin with their NASH, if their NASH resolves. If their LDL is still higher than gold, then they certainly should be. But we do know that there's a benefit in LDL type, and that probably could contribute to some improvement in cardiovascular disease, although that is a data-free zone right now. And this question is for Dr. Adams. And it sort of says, irrespective of pharma feasibility, do you think regulators will mandate cardiovascular trials for NASH drugs that clear phase three? I expect there'll be some requirement for phase four, but I think currently in the absence of any signals, I wouldn't think that that would mandate it for phase three. Having said that, I'm quite removed from any discussions. So this is all speculative, and someone else may well have more insight than me. No, so I think that's a great point. And I think from a NASH research perspective, a NASH regulatory development perspective, we're lagging far behind other disease conditions. And I think diabetes is a good condition, where we sort of started in the beginning just sort of making glycemic control better to get regulatory approval. But the bar has been raised substantially in diabetes, where not only do you have to improve glycemic control, but also address comorbidities such as cardiovascular risk. And given the strong association with cardiovascular risk in fatty liver disease, I anticipate that down the line, this is going to be something that's sort of in the forefront when we talk about drug approval and drug therapies for fatty liver disease. Sorry, go ahead. Having sort of just researched a little bit for this talk, I did get the feeling that the FDA was actually starting to draw back from that requirement for diabetes. And that was based on a fair bit of commentary, I think, that these trials are almost prohibitive in terms of getting drugs to market. And so I think the FDA may be being a little bit more cautious. And again, I think that's probably why they wouldn't mandate it for five straight trials for NASH. Yeah, no, I completely agree. And there's some actually, and this is sort of open to all panelists. And I think this is an important question is, A, in your clinical management of patients with fatty liver disease, do you sort of go off, use statin therapy right off the bat, irrespective of other risk factors or risk assessment tools, or do you actually do certain algorithms to start people on statin therapy? Yeah, I mean, for me, I use the ASCVD risk score. And I feel comfortable initiating statin therapy. And when I teach others, if they don't feel comfortable using statins themselves, being kind of the primary initiator of that therapy, then I recommend that they're very explicit in their documentation saying, look, this needs to be calculated. And as a hepatologist, go for it. I have no problem with them being on a statin, despite their mildly elevated liver biochemical tests. So either way, but I think it's just, it needs to be addressed in addition to the lifestyle risk factors, like Dr. Corey was saying with the sedentary behavior. I mean, these are things that we can really impact if we spend time with our patients. Yeah, I totally agree. I don't use NAFLD alone as the risk factor for statins. And I use exactly what Michelle does, the ASCVD risk score, which for those who have some EMRs, there are dot phrases that you can use that can calculate it for you. And if you do it online, it will walk you through the interventions that you should consider. And I will either start the statin or strongly encourage the primary care doctor too. And usually once they have the blessing of a hepatologist, I think they're more than willing to treat their patients. Do you have any comments on that, Dr. Adams? I had a question actually for Kathleen and Michelle. And so do you think the ASCVD risk factor operates differently in NAFLD patients compared to non-NAFLD patients? Should we be trying to develop a different predictive score? We're actually in the process of looking at that right now with some of our data from the PROMIS trial to see if adding the presence of NAFLD improves the ASCVD risk score or not. And then we're looking to see if the cutoffs look like they're about the same. I think what we can conclude right now is that it does predict cardiovascular disease risk and so can be used in our patients, but it's a great question of whether there should be something unique for those with NAFLD. I would suppose if we all agree that it's an independent risk factor, we probably should incorporate NAFLD, but that's certainly something we're looking into. And this equation, the pooled cohort equation, the ASCVD risk score, and remember the population that it was defined in, whether they knew it or not, a quarter, if not more, had fatty liver disease. So at least it was developed in a population that had or has fatty liver disease as well. So there's some kind of evidence that it perhaps would be a good tool in that sample as well, but I'm really excited to see Dr. Khoury's results. Yeah, that's a great point. That's a great point. I think so many of our studies had, as you saw in the PROMIS trial, 25% had NAFLD as we predict. So, but whether or not we add NAFLD to the score and it helps is still something we're hoping to sort out. So I think we're sort of at our time point here. So thank you all for an excellent talk and sharing all this valuable information with us. And thank you all for attending the webinar.

liver-heart connection

non-alcoholic fatty liver disease

NAFLD

cardiovascular outcomes

physiological changes

metabolic syndrome

diabetes

atherogenic dyslipidemia

endothelial dysfunction