Fibrosis vs. steatohepatitis in NASH: key prognostic and regulatory implications

Hi, my name is Vlad Ratziu and I work in Sorbonne University, Paris, France, and I would like to talk today about fibrosis and steatohepatitis in NASH. What are the key prognostic and regulatory implications? 

Non-alcoholic statohepatitis (NASH) 

So as you know, non-alcoholic steatohepatitis (NASH) consists of the abnormal accumulation of fat in the liver, which is called steatosis, but which goes together with lobular inflammation and hepatocyte cell injury, which is basically represented by the ballooned hepatocytes. The association of steatosis with hepatocyte ballooning and lobular inflammation makes for what we call steatohepatitis. Now, of course steatohepatitis can evolve towards fibrosis and cirrhosis and hence cause a significant health burden of this disease. 

Now, the current paradigm is that within this pathological framework, we can distinguish two different situations. Situation number one is when patients develop steatohepatitis. So when steatosis goes together with information and liver cell injury, and what is important about steatohepatitis is that it can drive the progression of fibrosis and fibrosis can progress through different stages with increasing amounts of fibrosis and architectural distortion and all this can result in the occurrence of cirrhosis. Alternatively steatosis rarely if ever progresses to cirrhosis and there’s almost no fibrogenesis involved in a patient that has near steatosis or only plain fat without information or liver cell injury. 

We do know that in some circumstances, some patients can switch from a steatosis form to a steatohepatitis form and then go on to develop fibrogenesis. This seems to be determined by the severity, the persistence or the worsening of the metabolic comorbidities or of the underlying insulin resistance, although not all determinants of this progression are well understood at this time. 

So therefore, there are two questions that can be raised –  one in clinical practice and the other when you do clinical trials, within the larger regulatory framework. The first question in clinical practice, of course, is to understand in front of a patient that has NAFLD. So the abnormal accumulation of fat in the liver, what are the important things you want to know about the state of the liver? Is it how much fat there is? Is it whether the patient has steatohepatitis as we pointed out earlier? Is it how much fibrosis is present, or is it simply whether the patient is running or not? Because that changes the way we will monitor the disease in the future. 

Registration pathway, phase 3 trials 

Now from a regulatory standpoint, the questions are a little bit different. As you know phase 3 trials have been ongoing for several years and the way those are built according to an agreement with the regulatory bodies is that these phase 3 trials are first comprised of an initial phase of treatment which usually lasts between 12 and 18 months, and which will be evaluated by an interim analysis. The interim analysis is intended to understand whether there is an efficacy of the drug on 2 histological endpoints, which are in fact, surrogates, and these two surrogates are first of all, steatohepatitis and second – fibrosis. Actually they’re both the same level from a regulatory standpoint. 

So, the way they are defined is that a drug is effective if it can induce disappearance of steatohepatitis or what we call resolution of NASH and / or depending on the on what type of regulatory bodies we’re referring to and / or the reduction in fibrosis stage, which is called fibrosis reverse. If these surrogates are met, then the drug can get in principle, conditional approval. But once this is obtained, the study needs to continue in order to demonstrate that patients, who achieve those histological surrogates would also go on to develop clinical benefit. So there will have less progression to cirrhosis and less offense of paddock decompensation throughout the computations. And if that is the case and that takes, of course, a little longer and high number of patients. But if that is the case, then the drug against, definite approval. 

So you can see here that the understanding of what steatohepatitis means and the prognostic implication of carries, and the understanding of fibrosis as a distinct pathological features are key to this whole regulatory framework and they’re also key to how we are going to explore and monitor a patient in clinical practice. 

Fibrosis stage and liver-related outcomes

So, let’s first, look at the data we have that had let us to understand the very important prognostic impact of fibroids. Many studies have been published, most of them are retrospective, actually, all of them. And they look into the correlation between the stage of fibrosis and clinical events, and they all show pretty much the same thing. The more fibrosis you have, the higher the risk of developing liver-related complications and liver-related death. But the main question is, when does the risk start to significantly increase? Is it at any stage of fibrosis? So stage 1 versus stage 0 or is it only a very advanced stages of fibrosis, for instance, stage 3 or even stage 4. 

So what most of these studies have shown is that very early fibrosis does not increase the risk of liver-related mortality. So that is true for stage 1 versus stage 0. This is the upper left corner in this slide but as soon as the patient reaches stage 2 and of course, stage 3 and stage 4, then the risk of liver-related events increases significantly over that operation without fibrosis. So it seems like the threshold is significant fibrosis which in histological classifications, is turned stage 2. And this is why most studies with NASH only include patients with stage 2 or higher of fibrosis. 

You can also see on the right part of the slide that while the risk increases from stage 1 to Stage 2 to stage 3 and onwards, the increase in risk is much higher once you reach the cirrhotic stage than before which tells us something about the relationship between the amount of fibrosis on one hand and the architectural distortion which is what the stages are based on the other hand. 

These data have led many investigators to consider that the only lesion that has prognostic implications in NASH, is fibrosis. And that steatohepatitis does not matter. And you can see here some of the studies that have been published with long-term follow-up, although they were all retrospective, that tend to compare in a multivariate analysis, the weight of steatohepatitis versus the weight of fibrosis and they all determine that independently fibrosis alone is the only factor that determines a worse prognosis and that steatohepatitis carries no risk. 

Some of us have always been against this sort of interpretation and one of the things that you can notice from some of these studies, for instance, is the following. First of all, if you look at patients without fibrosis, F0 patient’s actually 35 percent of them have steatohepatitis but cirrhotics, the one that has the worst prognosis in this study here. 94% of them have steatohepatitis. So you can see right there that steatohepatitis and fibrosis go hand in hand. So the more advanced stages of fibrosis, the higher the probability that you end up with steatohepatitis. So the two are closely linked together. Another demonstration is that there is a constant increasing prevalence in steatohepatitis between F0 and F3. So the higher the fibrosis stage, the higher the prevalence of steatohepatitis, again telling you that the two lesions go hand in hand. Also, 17% of NASH patients have advanced fibrosis; only 2% of those without NASH have advanced fibrosis. So, this again, tells you that fibrosis and steatohepatitis occur together and NASH is associated with increased overall mortality, and only patients with NASH develop severe liver disease, and they do develop severe liver disease earlier than patients without NASH. So, basically there seems to be a very strong correlation and a very strong collinearity between steatohepatitis and fibrosis.

NASH increases liver-related mortality and correlation of ballooning with clinical outcomes

Now, if we look at studies that specifically, look at the impact of steatohepatitis, on liver-related mortality, some of the same studies that have shown the major impact of fibrosis have also shown in univariate analysis – that steatohepatitis increases the risk of liver-related mortality, as you can see, on this right here – in one of the earlier studies and on this other slide here in a study with a different set of patients, read by different pathologies. 

So in both cases, the presence of steatohepatitis, increased liver-related mortality, and the same has been shown when you look specifically at the cardinal – histological feature of steatohepatitis actually is the defining feature of NASH, which is hepatocyte ballooning. You can see here that patients that have a high ballooning rate also have higher rates of progression to end-stage liver disease compared to those that do not have ballooning. So, that tells you again that the presence of steatohepatitis Is not innocent. 

Now, one of the latest studies that have shown that trying to separate between patients that have steatohepatitis with or without fibrosis has also demonstrated that the mere presence of steatohepatitis, independent of fibrosis is associated with an increased liver-related mortality. And you can see from this very large database of almost 9,000 patients that have a liver biopsy and very long follow-up from Sweden. You can see here that patients who have (that’s the green line) steatohepatitis without fibrosis. Actually, the green line or right here, actually have a worse prognosis than those that have simple steatosis, which is the color of this green line here. So there is a difference here that appears to be significant. Of course, it takes years to see this difference here but that tells you that people with steatohepatitis without fibrosis can have a worse prognosis and the presence of fibrosis is not the one that carries necessarily all of the prognostic impact in this condition. 

NASH and Fibrosis – two simple analogies to understand the disease paradigm 

So the way we can put all this together is to think about steatohepatitis as the engine of a train and of fibrosis as how far the train has gone. So if you have a powerful engine, the train will go far and if you want the process of this to be stopped, you will have to, of course, put the brakes on the train, that would be with an anti-fibrotic drug, but most importantly would be to cut the engine because otherwise, if the engine is up and running, you would need to have very, very powerful brakes to stop the train and it’s not clear that the brakes will hold in the long term. 

And this was very nicely shown actually with drugs that have absolutely no anti-fibrotic impact and yet cure fibrosis because they’re able to get rid of the noxious agent, the cause of the disease. For instance, empty virals in HPV or HCV. They have no antifibrotic properties yet because they remove the cause of the disease. Fibrosis will end up going away slowly, it takes time but eventually fibrosis will regress, and we hope that NASH could be with drugs that are currently developed being the same scenario, meaning a drug that would address metabolic dysfunction and steatohepatitis and lipotoxicity, even without specific powerful antifibrotic effects might have an antifibrotic impact because the liver disease will be severely slowed, maybe even eliminated. So this is important to understand how the two things go together and the fact that steatohepatitis is the driver of fibrogenesis.

Fibrosis progression in relation to disease activity 

Is there any data to support this hypothesis? Which is that if you treat steatohepatitis, you also end up with fibrosis. So yes, there is data, there is data from natural history studies and there’s some data from interventional studies. First natural history studies. This is a very large database of patients from the NASH CRN in the United States that had a liver biopsy, and then follow-up and then a second liver biopsy. The idea is to see whether the amount / level of the activity of steatohepatitis on the first biopsy predicts, what will happen in terms of fibrosis on the follow-up biopsy and indeed it does, because when you look at the baseline value of activity which is measured by the NASH score for sort of a composite of activity means sort of a composite of lobular information and episode or death, and when the baseline In score is low, meaning very slowly active disease. Very low activity of disease. Not much inflammation, not much ballooning. 

On the subsequent biopsy fibrosis will mostly regress. You see the change in fibrosis stage will be negative as you can see here from 0 to 4. Conversely in patients, who had at the initial biopsy, high levels of activity, very high level of inflammation, very high grade of hepatocyte ballooning. These patients will progress into terms of fibrosis. So they will end up with more fibrosis at follow-up, then on the first biopsy. So that’s for Baseline mass. Now, if you look at changes in the activity, score in between the biopsies, you get the same result. Basically if the change is negative, meaning activity goes down, meaning lobular inflammation goes down, hepatocyte ballooning goes down, then fibrosis will follow – it will also go down to see a negative change in front of the stage. On the opposite iff the activity goes up because laboratory information goes up, hepatocyte side goes up, then you’ll see progression in fibrosis as you can see on the graph in terms of increased fibrosis stage on follow-up items. And this has also been shown in terms of no longer core, but a diagnosis. So for instance, if you look at all patients, including the study on their first biopsy, if they simply had fatty liver then on the second biopsy, very few of them actually less than 20%. You can see red here on the left bar, will end up with significant fibrosis. On the opposite, if the patient had steatohepatitis on the first biopsy, you can see here that more than 60% will end up with significant fibrosis and progression from steatosis to steatohepatitis was a very strong predictor of progression of fibrosis. 

So this demonstrates in the natural history context without any intervention that there is a very strong link between In the level of activity and progression of fibrosis. Now, there are also data from interventional studies here. This is data from PIVENS which is an older study that compared vitamin E  and pioglitazone versus placebo in treating NASH in the Flynn study, compared orbiting colic acid versus placebo, and they both show the same thing. Basically, if you look on the left part of the slide percentage of biopsies with fibrosis Improvement. So if the NAFLD diagnosis is worsened, meaning you go from steatosis to steatohepatitis, or you could go from not very active steatohepatitis to active steatohepatitis. The proportion of people who improve fibroids is very low. If the NAFLD diagnosis improved meaning, you now have steatohepatitis or it was severely reduced in terms of activity, then the percentage of people with fibrosis Improvement on the subsequent biopsy increases significantly and that is true both for PIVENS and for Flynn. So independent of the drug that was used to treat steatohepatitis and actually, the histological features that were associated with fibroids Improvement in these studies were the ones that are listed here and you can see that many of them have to do with inflammation or with steatohepatitis, because one of the strongest was in fact disappearance of the steatohepatitis is what we call NASH resolution. 

Responders vs Non-responders for the primary endpoint in completion on 120 mg

There is another set of data coming from interventional studies and those are well-designed randomized control trials. This example here comes from the early phase 2B trials of elafibranor.. Overall, what the drug was able to achieve was a resolution of steatohepatitis but without fibrosis reversal and that was if you look at the average between groups, however, and this is why it is interesting in this demonstration. If you look not at the average between groups, but if you compare people who are able to have resolution of steatohepatitis, that is versus those that did not achieve that end point, meaning responders for NASH versus non responders. You can see that the patient who had a response for NASH resolution of steatohepatitis also had an improvement in fibrosis, which was significant even. 

Changes in NASH activity index and Fibrosis evolution 

So that was not seen at the level of the whole group. So what that is meant to demonstrate is that while the whole group is not able to improve fibrosis, the only ones that do improve fibrosis are the ones who have less steatohepatitis or less active steatohepatitis after the treatment period. If you plot together with the activity score defined as a sum- a simple arithmetical sum, between inflammation and parasite ballooning, and you plot that versus changes in fibrosis scores, you can see that, there’s almost a linear relationship between the two. And if you have a change in activity index, in the sense that you have deterioration, meaning more inflammation, more ballooning, then you will see much less fibrosis improvement and much more fibrosis worsened on the second box. 

On the opposite. If you have a negative change in activity index, meaning you end up with less inflammation ballooning. Then you’ll end up with more fibrosis Improvement as you can see on the left-hand side of the graph and this relationship is almost linear. Now, this relationship between fibrosis improvement and ballooning improvement or information improvement is not reproduced. If you look at the relationship with the abilities, they seem not to be related. So, people who improved that score did not improve fibrosis or did not worsen it either. So this relationship between activity and fibrosis specific to that and for instance cannot be demonstrated when look at steatosis. 

Progression from NAFLD to NASH 

So what can be concluded for that? Is that overall NASH is associated with more severe hepatic and systemic disease. In terms of hepatic disease, people who have steatohepatitis as compared to those who have steatoizalone have more advanced liver disease, higher ALD, more advanced fibrosis but they also have more profound internal resistance more severe. Cardiovascular lesions, and more severe metabolic. comorbidities in terms of diabetes arterial hypertension, sleep apnea. There are also some data that have shown with repeat liver biopsies and our center was among the first ones we’ve shown that. That when you biopsy again, patients will simply have steatosis on the first biopsy, you know. Then go several years later and come back and biopsy them again, you can see that some progressed and many of them actually here in 16/25, no longer have just steatosis, some of them actually developed steatohepatitis, and some of them develop fibrosis, even advanced fibrosis. But what is interesting is that almost all patients, except one that have developed, severe fibrosis were the same ones who develops steatohepatitis. So the occurrence of steatohepatitis during follow-up, goes together with progression of fibrosis, which is also. well, the study from the national trend has shown. 

So, this resulted in an understanding of the spectrum of the disease, the interpretation is that the presence of information and liver cell injury is a prerequisite to the development of fibrosis and the higher the activity, meaning, the amount of fibrosis, the amount of inflammation, liver cell injury, the higher the chances, to develop fibrosis during the follow-up.

NASH resolution is associated with Fibrosis regression – how does steatohepatitis drive disease progression in NASH? 

So, how does theater of that is drives disease progression, NAFLD. Well, as I explained NASH is associated with more profound insulin resistance, more severe etabolic disease, higher ALD and more advanced hepatic fibrosis. It significantly increases liver related mortality over patients with just steatosis. It drives fibrosis progression. Very importantly progression from steatosis to bridging fibrosis. So, steatosis without fibrosis. Progression to bridging fibrosis. Transitions through steatohepatitis, meaning, most of the people who will develop advanced fibrosis, have also developed steatohepatitis in the process and the concept of disease activity in NAFLD is that non drug-induced changes in NASH score correlated with fibrosis progression and drug-induced resolution of steatohepatitis and improvement in the correlation also correlates with fibroids regression as we have seen from the PIVENS trial, the golden trial in the Flynn trial. 

A placebo-controlled trial of subcutanuous semaglutide in non-alcoholic steatohepatitis 

Now, do we have data from the current drugs in development that support this hypothesis here, the situation gets a little bit more complex because we have examples from the studies performed so far that provide data that can be interpreted in different ways. So the first example is, for instance samaglotide – a GLP 1 receptor agonist, resulted in an 18 month study in resolution of steatohepatitis, you can see on the left side of the slide at a much higher rate than placebo, 59% versus 17%, despite this very strong and despite a long term of treatment 18 months, the drug was not able to significantly and statistically show a difference in terms of fibrosis regression, there was a numerical difference, but not a statistically. A significant difference. 

So that seems to go a little bit against the hypothesis that we have developed previously. Another example, that of a abetacolic acid in a very large phase 3 trial, which is called regenerate. This drug was able to achieve fibrosis reversal without formally being able to demonstrate resolution of that. So it’s the opposite scenario. However, when you look at the impact of the drug on the elementary histological reasons of steatohepatitis, you can see that actually, the drug improves activity of the disease, improves inflammation, improves ballooning. 

The reason why it failed to show a statistically significant difference in terms of NASH resolution, might be may be insufficient exposure in terms of time, or maybe insufficient power because the same data set have shown that if you expand the analysis to a larger population, then you can see a significant difference in terms of resolution of steatohepatitis, but still the drug was to achieve fibrosis reversal without necessarily inducing steatohepatitis resolutions. So, this again has to be taken into account. See, whether the hypothesis were formulated and carries enough evidence. 

Hispological improvement in the NATIVE trial 

Now, there are examples to the opposite charge short-term trials, only six months, that have shown an ambiguously, the diisappearance of both lesions, of course, these trials, the contrary of the ones that have been shown previously are much smaller sample size. So there’s always a risk of type 2 error. However if you take for instance lanifibranor in a six months trial, you can see that the drug was able to achieve both NASH resolution and fibrosis reduction and the magnitude of the difference versus placebo is higher in terms of NASH resolution than fibrosis. 

Also, an argument in favor of the fact that one transitions to another. 

Another example is, of course fgf21,affirming the results of this smaller study have been released recently. Again, a very short study – a small number of patients. But again, showing unambiguously that there is a very strong effect in terms of NASH solution, a strong effect in terms of fibrosis improvement. And the fact again that you have a higher magnitude of effect for NASH than for fibrosis tells you that probably sequence of events here, makes it that first NASH disappears and then fibrosis goes away as well. But again, this was a six-month study, so very short time for all these lesions to improve. 

Is the diagnosis of NASH relevant for clinical practice? 

So then the question becomes beyond this data is the diagnosis of steatohepatitis even relevant for clinical practice. We have now many non-invasive tools that can measure some of the aspects of the disease. For instance, that can measure the amount of fat that can indicate whether there’s the other that is or the amount of fibrosis or fat. We have ultrasound, which is a qualitaive diagnosis and we have for research purposes only the measurement by MRI of the fat fraction is called PDFF and that measures quantitatively the amount of steatosis.

For steatohepatitis, we don’t have a direct marketer but alt markers, the serum alt levels can indicate the level of activity of the disease and then there’s of course the liver biopsy and in terms of imaging biomarkers may be corrected T1 and on MRI is marker that indicates the amount of steatohepatitis. And for fibrosis, we have many serum-based fibrosis markers. And we also have imaging-based fibrosis markers, whether fibroscan or the measure of stiffness by MRE, the question then becomes if fibrosis is the most important thing, can’t we just bypass the diagnosis of steatohepatitis and only when we manage a patient in clinical care, only be interested in the amount of fibrosis, or can we do that differently by introducing a new concept that has gained considerable traction? Which is the concept of at-risk NASH. What is at-risk NASH? It is NASH with active NASH with already established fibrosis. So meaning NASH with sufficient amount of activity and with significant fbrosis already established. Why is at-risk NASH important? Because that is when the risk of progressing to cirrhosis increases considerably and those are actually, and for that reason, these patients are actually the ones selected for therapeutic trials. 

Can we diagnose at-risk NASH? Yes, there are several composite methods that now exists that diagnose, this condition. And they’re all based on sort of a mixture of imaging and serum based markers. Three of them at least are – you can see here, the fat score, which is based on based on fabric and measurements of liver stiffness of the amount of fat and the measurement of AST, one of the two liver amino transferase. 

Another possibility is to measure the stiffness of organ, MR Based technology – MRE. And then to add the form which is a serum based model out of fibrosis, calculate a score and that’s the method score and then there is a purely serum based test, which is called minus 2 plus, which measures to indicators of fibroids and activity. And put together in the score, takes into account age and sex, and that gives you that needs two plus four, which tells you whether the patient has advanced fibrosis and with active steatohepatitis or not. 

Two conceptual approaches to screening

Therefore, there are two different strategies that we can use the patient care pathway or for clinical research in epidemiology and these two approaches are somehow different in what they try to achieve. The first one is to use these tools to exclude patients who have almost no disease with no referral to a tertiary care center, who can be sent back drawn practice and then being monitored loosely two or three years later. So, that is the first approach trying to exclude patients, that are actually at low risk of progressing disease. This approach takes advantage of the very high negative creative value of some of the tests for fibrosis that have been developed over the past years. 

The second approach is different, it is to try to identify those people, that would need more stronger and more intensive monitoring and also who need pharmacotherapy, because they’re at high risk of progression to cirrhosis. So this second strategy consists actually in identifying patients with at-risk NASH and that is can be achieved nowadays, the four different markers that have shown on the previous slide. 

Conclusions

So in conclusion, what we can say based on all these data is that fibrosis is definitely an independent predictor of liver related-mortality but fibrosis never occurs outside of the presence of an active liver disease. So it does not occur without steatohepatitis. NASH is the progressive form of disease. The one that confirms a higher risk of progression in terms of fibrogenesis and in terms of cirrhosis and therefore people who have NASH need to be monitored, depending on the amount of fibrosis they already have achieved or and conversely people who do not have NASH probably need less monitoring and less aggressive implementation of therapies at least with the liver in mind, NASH drives fibrosis progression – we have seen that the baseline disease activity, determines fibrosis progression. We have seen that progression from no disease to advanced fibrosis transitions, almost always through steatohepatitis. And we’ve also seen that changes in disease activity whether your natural history or own therapy result in same direction fibrosis changes. All this data are very important if you want to conceptualize how future therapies for NASH should work. Do we absolutely need drugs that are directly / exclusively anti-fibrotic or can you achieve fibrosis regression by using drugs that control the underlying comorbidities that drives steatohepatitis drugs that control The other that is activity itself. These drugs, even in the absence of direct antifobrotic activity still result in regards of fibrosis because they have improved steatohepatitis either by eliminating it or by severely reducing its activity. Thank you for watch this program, I think I am done.