Inventiva S.A. (IVA) Earnings Call Transcript & Summary

Good day, ladies and gentlemen, and thank you for standing by. Welcome to the Inventiva's KOL webcast, providing an update on Phase IIb NATIVE trial evaluating lanifibranor in NASH. [Operator Instructions]. I must advise that this conference is being recorded today, Monday, 4 of May, 2020. I would now like to hand the conference over to your first speaker today, Mr. Frédéric Cren. Thank you. Please go ahead, sir.

Thank you, Brian, and thank you, everybody, for attending. It's really pleased to be able to organize this webcast. We actually had planned to do this meeting live at the EASL, but due to COVID, we had to postpone it, but happy to be -- happy to have it organized it today. It's also an important day because today, we are actually resuming activity at Inventiva. So our labs are now open and people are back on the bench. It's also an important day because we are approaching the end of this trial, and we are approaching the hard look. And while the data remains blinded, we are really reaching the publication of the headline results that are due in June. We're very happy today to be sharing the floor, and I thank them again for their flexibility with Sven Francque, who is the PI of the trial. And also with Professor Pierre Bedossa who is the central hepatologist and will go through the -- how slides are read in this trial. And then, of course, also with Pierre, the CSO and Co-Founder with me of the company. The agenda would be very simple. Pierre will briefly go through why we think really there is a role for lani in NASH. Pierre Bedossa will go through the SAF scoring and why we believe it is a more appropriate scoring compared to the NAS one and to how we read biopsy NASH. And then, of course, Sven would spend time on the Native Phase IIb, the trial. And then, of course, we are looking forward to your [indiscernible] to your Q&A. So let's now me turn the floor to Pierre and go through lanifibranor in NASH.

Thank you, Frédéric. So hello, everybody. So well, it is very exciting times at Inventiva, as you would guess. And we are glad to see so many people attending the call. So before we deep dive into the NATIVE trial design and baseline characteristics, I would like to say a few introductory words on lanifibranor. So I will move now on Page 6. So as you know, lanifibranor is a pan-PPAR agonist with an unprecedented chemical structure. It's not a fibrate, it's not thiazolidinedione. The molecule is capable of activating the 3 PPAR isoforms concomitantly with a moderate potency. For example, for PPAR alpha, lanifibranor is equipotent to fenofibrate. And for PPAR gamma, equipotent to pioglitazone. And we believe that this unique profile is particularly well suited for the treatment of NASH. The next slide, please. So we know that NASH is a multifactorial and multistate disease with obesity, insulin resistance and dyslipidemia being associated features, which favors NASH development. And the development of NASH involves all the different cells present within the liver as well as the adipose tissue, intestines, skeletal muscle, where the PPAR play major regulatory roles. For example, PPAR alpha activation improves lipid metabolism by controlling lipid flux and regulating fatty acid transport as well as beta oxidation in the hepatocytes. PPAR alpha also reduces inflammation and improves endothelial function. For example, in models of cirrhosis, PPAR activation has been shown to decrease portal pressure. PPAR delta activation inhibits inflammatory macrophages phenotypes, and favors the alternatively activated protective cells back in the liver. PPAR delta is also involved in the glucose and lipoprotein metabolism and reduces insulin resistance in the skeletal muscle. And finally, PPAR gamma activation, as you know, improves insulin sensitivity within the adipose tissue, thereby reducing free fatty acid flux into the liver and steatosis. And PPAR gamma also prevents hepatic stellate cell activation, which is a key event in fibrogenesis. And in the context of cirrhosis, PPAR gamma reduces portal pressure, splanchnic inflammation, angiogenesis as well as portosystemic shunts. So altogether, 3 PPAR isotypes impact multiple pathways and mechanisms that are involved in NASH and fibrosis progression. And consistent with these, there are clinical evidence supporting the therapeutic potential of PPARs in NASH, as you know. Let's move to the next slide. So first of all, of course, you know that PPAR gamma activation by pioglitazone leads to a significant reduction of steatosis, ballooning and inflammation in NASH patients, actually doubling the number of patients that improved on ballooning or inflammation over placebo. Maximal efficacy is achieved within 6 months and sustained over time. And on fibrosis, there is a meta-analysis that indicates that PPAR gamma activation with pioglitazone can improve advanced fibrosis. Next slide. And in the pan-PPAR context, we think that PPAR gamma activity can be reinforced by PPAR alpha and delta activity, for which evidence of efficacy was reported, mostly on ballooning and inflammation with the dual alpha delta agonist, elafibranor. Next slide. So lanifibranor was investigated in multiple models of insulin resistance, NASH and cirrhosis. In diet-induced insulin resistance model, we observed a consistent improvement of insulin sensitivity with an inhibition of body weight gain, which contrasts with what has been reported in these kind of models with single PPAR gamma agonist for which you have the same level of insulin sensitivity improvement but with body weight gain. In more inflammatory models such as MCD and CDAA high fat diet models, we observed the rapid and sustained suppression of steatosis, ballooning and inflammation with SAF scores indicating NASH reversal. And in more fibrotic models such as CCL4 or TAA, we observed reversion of fibrosis and in the TAA model, an improvement of hepatic vascular resistance, leading to a decrease in portal pressure and a decreased number of [ animals with site ]. So in summary, lanifibranor improved all histological features of NASH, including liver fibrosis, combining and exceeding specific effects of single PPAR agonist in studies where lanifibranor was compared head-to-head to single PPAR alpha, gamma or delta agonist. And lanifibranor has also the potential to improve portal hypertension, which is now recognized as an early event in the progression of NASH. So Slide 7 -- sorry Slide 11, next slide. On the safety side, there are no concerns in safety pharmacology, low cardiotoxicity in long-term tox studies in rodents or primates and no carcinogenicity issue. Good safety profile that triggered the PPAR class related clinical hold to be lifted by FDA for lanifibranor. Next slide. So based on these, the potential of a pan-PPAR agonist such as lanifibranor in NASH is quite compelling, providing potential therapeutic benefit on multiple features of the disease, metabolic, steatohepatic, fibrotic and the NATIVE trial is designed to explore this. Next slide. So is there room for a candidate treatment of NASH that can potentially address NASH resolution and fibrosis in a space where several products are already in Phase III. And yes, we believe so, of course. As can be seen here, recent results achieved by different therapeutic approaches on fibrosis improvement without worsening of NASH. Only a couple of mechanism of action led to a doubling of the number of responders relatively to placebo. Next slide. So if we look now at the recent results regarding NASH resolution without worsening of fibrosis, you can see that the percentage of responders ranged from 12% to 24%, which can be considered suboptimal. So there is obviously room for improvement. Next slide. So we also believe that a pan-PPAR mechanism of action has the potential to provide a differentiated therapeutic profile compared to the Phase III products or other Phase II advanced products, with, firstly, a better control of insulin resistance, thanks to the PPAR gamma component. A good control of steatohepatitis, thanks to the combination of alpha, delta and gamma and potential anti-fibrotic properties due to the PPAR gamma component. Next slide. So in that context, we are working hard on the probation of the Phase III with the current assumption of one dose versus placebo, 1 year of treatment for Part 1 with an interim primary endpoint of NASH resolution with no worsening of fibrosis or at least one stage reduction of fibrosis with no worsening of NASH, and key secondary endpoints related to the anti diabetic potential of lanifibranor. We are currently foreseeing inclusion criteria that would allow to randomize patients with NAS above 5, with at least one point for inflammation and one point for ballooning or NAS score above 4 with at least 2 points of either inflammation or ballooning and fibrosis scores of F2, F3. So of course, this is work in progress. And changes may occur following the NATIVE trial results as well as clinical or regulatory informations that may arise from the competition. So next slide. Next steps: of course, NATIVE Phase IIb headline results publication next month, finalization of Phase III synopsis and protocol, which is ongoing, end of Phase IIb meeting with FDA planned Q4 this year, finalization of our Phase IIa study in NAFLD patients with type 2 diabetes conducted by Professor Cusi from University of Florida, and then the launch of pivotal Phase III study in NASH, pending outcome of NATIVE results and regulatory discussions. So with this, I would like to thank you for your attention and leave the floor to Professor Pierre Bedossa who will introduce you to the main histological methodologies, leading to the scoring of NASH and their relationship with one another. Thank you.

Thank you, Pierre, and hello, everyone. So my name is Pierre Bedossa, and I am a professor of pathology, specialized in liver pathology and working in the field of NASH for more than 15 years. So the first slide that you will see next, please, is a typical pattern of NASH that we are used to see in the NATIVE trial. Can I have this slide, please? Yes. And you can see on the left part of the screen the central zone tree with a lot of ballooning, inflammation and fibrosis. And on the right, at B and C, the ballooning and the inflammation. And this type of lesion is very often associated with some fibrosis. And due to the choice we made -- can I have the next slide, the choice we made for the scoring systems, this is mainly the type of advanced lesion that we have seen in the NATIVE trial. So can I have the next, please. Okay. And indeed, we did not choose this trial, the NAS score, for several reasons. First of all, so the NAS score includes steatosis and steatosis is for sure not a marker of activity and not a driver of fibrosis. The second reason we don't use it is that ballooning is underweighted. While ballooning is a hallmark of the disease, only 2 points out of 8. Furthermore, there are some problems with reproducibility of inflammation and ballooning. And so far, NAS has never been demonstrated as having some prognostic value and the significant inter-observer -- can I have the next slide, please. Significant inter-observer variability in scoring is a real challenge in the clinical trial. And this slide shows the reproducibility, the inter-observer variability of the -- component of the NAS score. And as you can see, also steatosis or NASH diagnosis is quite acceptable in the several studies, which have been published on that, the reproducibility of the kappa value of inflammation or ballooning is close or less than 0.5, which is modest. So this is one of the major limitations of the NAS score. Can I have the next, please? So we turn to another scoring system that we developed now in -- 5 to 6 years now, and which is the SAF score. The SAF score for steatosis activity and fibrosis is another way to [ semi score ] in the disease. Steatosis and fibrosis is similar for NASH CRN, and it is not concerning in my talk and in the follow-up, but activity is the major issue. Here in the SAF score, activity is made only ballooning and lobular inflammation. And the activity is from 0 to 4 with ballooning being graded from 0 to 2, and lobular inflammation equally from 0 to 2. Can I have the next slide, please? So if you compare the NAS to the SAF, you will see that also NAS go from 0 to 8, the weight of ballooning, which again is a major feature and the hallmark of the disease, is only modest compared to the activity score of SAF where the ballooning can be at least half of the activity score. Can I have the next one? The other issue that I have mentioned before was a problem of observer variability, which, of course, is a major issue when we are talking about histology. And with the definition of ballooning that was used in the previous NAS score, this makes reproducibility of ballooning very limited. So I don't want to go in detail on how we score ballooning. But as you can see on this slide, we meticulously define the scoring of 0, 1 and 2, comparing, for example, to the NAS score where ballooning from 0 was no, 1 were few balloon cells and 2 was many ballooned cell. Here, what you see here is that for the scoring of ballooning in the SAF, we define clearly based on the shape, on the number, and on the size of the cells, what is 0, 1 and 2. And can I have the next slide, please? Then the consequence of this definition, which we made clear as possible, was better reproducibility between the various observer. And you can see here in the middle of the slide that now the reproducibility of ballooning and lobular inflammation is a kappa now, which is 0.8 for ballooning, which is as good as fibrosis and 0.7 for lobular inflammation. So this is highly reproducible and a very sensitive scoring system. Can I have the next slide, please? Now if you compare the diagnostic of NASH according to the scoring, either through the NAS and through the SAF, you will see in the next slides that for the NAS score, which is usually what is recommended in the clinical trials, there is a sort of gray zone because this is a study, which has been done on more than 800 biopsy of patients morbidly obese, where we have compared the diagnostic of NASH as percentage with the score either for the NAS score or the score of activity. You see that for a score of 4, we have -- you have something like 30% to 40% of the patients which are not NASH. By contrast, if you use the SAF activity score and a score of 2, 95% of these patients are in red, which you see, they are NASH. And if you put the bar higher as we have done with the trial, then you can see that with activity score of 3, then you get a score 100% of patients having NASH. Can I have the next slide, please? The second part of my talk will focus about the biopsies and the methodology, what we use and how we may try to make any progress to reduce the limitation of the biopsy. So the workflow of biopsy, can I have the next, please. It's -- as you know, it's a teamwork because it involves the one who do the biopsy, he's also a hepatologist and the radiologist. Often the local pathologist that will prepare the slide. And the central lab will blind the slide and stain the slide, and central pathologist. And to get a biopsy of good quality, you have to have all this part highly involved and having a good practice. Next one, please? So we try to inform and make a lot of information to the hepatologists and radiologists in order for them to provide adequate sample because if the sample at the baseline is not adequate, you cannot do anything. So to -- we will go over this quickly, inadequate sample, inadequate sections and the problem of histological criteria. Can I have the next one, please? So to avoid as much as possible inadequate sample, we provide information to the different center. But of course, as all these clinical trials are made with centers around the world, it is a very difficult part -- point to get good material, so we provide recommendation as the criteria for adequacy being optimally 20-millimeter length and not badly fragmented. We also suggest that the one who does the biopsy is someone who is trained to liver biopsy procedure and repeat that currently along the study. Of course, the major issue was to ask to the hepatologist and also the radiologist that sometimes do the biopsy to use a needle of at least a 16 gauge. Otherwise, you get a very thin core biopsy and very limited material. And we recommend to use cutting rather than aspiration needle in order to avoid the fragmentation of the biopsy because when you need aspiration biopsy, there is a lot of fibrosis. Then the biopsy will be fragment it. Can I have the next slide, please? And finally, if there is not enough material, we ask to the -- when you do the biopsy, to make eventually repeat passages and have a close discussion with the local pathologist. So the role of the central pathology laboratory is also essential because they will -- they are in charge to provide adequate bio -- sections to the central pathologist. So it needs -- it is as important as the one who does a biopsy and it needs, of course, to -- need to have the work done by trained histotechnologist. And this need also to process the biopsy in a very careful and standardized way using reliable staining method for H&E and Masson Trichrome. And Masson Trichrome is something sometimes to be difficult to have it reproducibly. And it needs also for the central lab to have the -- an expert in liver pathology that good quality control of the section before sending to the central pathologist. Can I have the next slide, please? Of course, this should be done into limited turnaround time in order to let the center having enough time to enroll the patient. The role, of course -- this is just an example of what you can get if you have on the left side of the screen, biopsy, which is performed by an experienced histotechnologist that will cut the biopsy with the needle. And you can see how the serial section are very similar. But if it is done with someone who's not experienced with a very small biopsy, then you can see that you can get at least one or maybe 2 sections of good quality, and then you lose the material. So next slide, please, this is very important for the study and for the pathologist to have an experienced histotechnology lab that we get for this trial. And finally, the robustness of pathology evaluation is important. And this is why in all the clinical trials, multicenter clinical trials, there is a role of -- there is a central pathologist that will review all of the slides. It needs to be highly experienced in clinical trial and in NASH pathology. And personally, I am involved in most of the Phase III clinical trials of NASH, and I review central pathologist several Phase IIb. So the experience is very important in order to have low inter-observer variability, meaning that you will interpret the same way of biopsy 1 day and 6 months after of the biopsy before treatment and the end-of-treatment biopsy. So the robustness of this evaluation is highly dependent of the experience of the central pathologist. So the same diagnostic criteria will be used all -- and have been used all along the studies. And for each of the biopsy, we had a scoring sheet, which was used all over the trial with several items, including the scorings that I have mentioned and other one. And this will also limited the -- limited the turnaround time. And I will conclude that by saying that we are very fortunate because last biopsy, end-of-treatment biopsy arrived in end of February, just before the outbreak of COVID around the world. So we -- this was -- we were not impacted by that. So thank you.

Thank you, Pierre. We will now move to the NATIVE trial with Professor Sven Francque.

Well, thank you very much, and good afternoon. My name is indeed Sven Francque. I'm Head of the Department of Gastroenterology and Hepatology in Antwerp in Belgium, working for more than 20 years now in NASH research for the pathophysiology and the pharmacological treatment. Next slide, please. So the NATIVE trial is indeed a Phase II trial, but as the title of the slide says, it's a Phase III enabling study. This means that you can look at the bottom of the slide to the scheme of the study. This is a [ pad ] biopsy study. And we only -- we do not only have, of course, a liver biopsy at the start of the study or during the screening period for the inclusion, but it's also for efficacy assessment that we use liver biopsy, and that's what you need before you can go to Phase III, you need histological proof of efficacy. This disease is still largely relying on liver biopsy for accurate diagnosis and also for an accurate assessment of treatment response. Now this trial is, of course, a randomized, double-blind, placebo-controlled trial, 24 weeks of treatment, and I'm happy to be principal investigator of this trial, together with Professor Manal Abdelmalek from Duke. Now inclusion is, of course, based on several criteria, but the main inclusion criterion is, of course, the diagnosis of NASH. And as Professor Bedossa already explained to you, we used the SAF scoring system for the inclusion, focusing on the activity of the disease because it's not the fat accumulation in the liver that drives disease progression, it's the damage to the hepatocytes and the inflammatory processes that are driving this disease. And that's histologically hallmarked by lobular inflammation and ballooning. So that's also why the inclusion in this trial and also for assessment of treatment efficacy, we focused on activity of the disease, so on lobular inflammation and ballooning. That's the main reason also to go for the SAF scoring system, both for the definition of inclusion criteria as well as for the definition of the primary endpoint and hence, so the efficacy assessment. So inclusion criterion means biopsy-confirmed NASH patients with inflammation and ballooning sum of 3 or 4 according to the SAF scoring system. So the activity component of the SAF scoring system as Professor Bedossa explained. Of course, there also needs to be a steatosis score of at least 1 that's evident because otherwise, it's not NAFLD. And the fibrosis score, we did not require F2 or F3 fibrosis for inclusion. So we did not preset a number of patients that needed to have significant or advanced fibrosis. I will come back to that later on. But we excluded patients with F4 score for fibrosis. So we only included non cirrhotic patients. The 3-arm study comparing placebo and lanifibranor 800 milligrams once daily or lanifibranor 1,200 milligrams once daily for a treatment period of 24 weeks, followed, of course, by a follow-up period of 4 weeks for safety assessment. Next slide. So as I already alluded to in the previous slide, the SAF activity scoring system was not only used for inclusion, but also for the assessment of efficacy, at least for what is the primary endpoint of this study. So the primary endpoint was defined as a decrease from baseline to week 24, of course, of at least 2 points of that activity component of the SAF scoring system. So the sum of lobular inflammation and ballooning as defined in the SAF scoring system. So it's not a decrease of 2 points in the NAFLD activity score, the NAS from the NASH CRN, it's 2 points in the activity score component of the SAF score. It's important to highlight that. Professor Bedossa is the central reader of this trial. So you really have uniform assessment of the biopsies all along the trial. And then the statistical hypothesis is based on the expectation that we will have about 10% responders also in the placebo arm. And then we require an excess rate of 20% that we consider clinically relevant in the treatment arm and that ends up with 72 evaluable patients per arm. There will be, of course, comparison of the 2 treatment arms with placebo, so the 1,200 milligrams compared to placebo for the main analysis of the primary endpoint and 800 milligrams compared to placebo. Important analysis will also be done by subgroups, and one of the important points is the presence or absence of diabetes. I will also come back to that. But we will compare the efficacy in diabetic versus nondiabetic patients and, of course, evaluate also the dose effects, so compare to 1,200 milligrams versus the 800 milligrams. Next slide, please. So the primary endpoint, as said, is a reduction of 2 points in activity of the disease, again, activity is the sum of lobular inflammation and ballooning. So it does not take into account steatosis, which is an important point. But of course, we have several key secondary endpoints that reflect other endpoints that have been used in clinical trials, including the ones that are included in Phase III trials for regulatory purposes. So key secondary endpoints are NASH resolution with no worsening of fibrosis, which is an endpoint for approval in Phase III studies, improvement of fibrosis by at least 1 stage according to NASH CRN, but that's the same as in the SAF scoring system, without worsening of NASH. And then we will also look into NASH improvers, which is a classical endpoint that has been used in many Phase II studies. Namely a reduction in 2 points of the NAFLD activity score. And here, NAFLD activity score is what is proposed by the NASH CRN. So that's the sum of the steatosis, the lobular inflammation and the ballooning. Other secondary endpoints that are, of course, clinically relevant is the combination of resolution of NASH and improvement of fibrosis by at least 1 stage; or the resolution of NASH with no worsening of fibrosis and NASH improvers, so a NAS decrease of 2 points; improvement of fibrosis by at least 2 stages without worsening of NASH, that are other histological important secondary endpoints that will be analyzed in this trial. Given the fact that this is a disease that is closely linked to the metabolic syndrome and also given the mode action of this pan-PPAR drug, it is also important and included in the secondary endpoints to look into changes in the parameters of glucose metabolism, fasting insulin, HOMA-IR, glycated hemoglobin and so on. Classical liver tests will also be evaluated, transaminases and canalicular liver tests, total bilirubin. Again, very important in the context of this disease and both from a safety perspective, but also given the mode of action of the drug, an important factor to look at is the lipid profile, that will be analyzed in detail. We will look at inflammatory markers, like high-sensitive CRP, alpha2 microglobulin, haptoglobin, fibrinogen and so on. And we will analyze changes in fibrosis markers, so noninvasive tests like TIMP-1, TIMP-2, the FIB-4 score, the ELF score, Pro-C3. These are all included in the secondary endpoints and will even so be analyzed along with the primary endpoint. Next slide, please. It is important, of course, in every clinical study, but especially in the context of metabolic syndrome and long-standing disease that we look into the safety profile of the drug, and we need drugs with excellent safety profile because we think we will have to treat patients for a long time. Now for the NATIVE trial, there have been 4 formal meetings of the DSMB according to a prespecified scheme. The last one was in September of last year. As you can see at that time point, more than 90% of the patients had already been on treatment and 57% at that time point had already completed the total duration of treatment. Important is that at every meeting, the conclusion of the DSMB was that there was no safety signal and that the trial could go on without any modification from the protocol, which was reassuring. Next slide, please. The next slide will show you the disposition of the patients, which is quite classical for a NASH trial that we have a screen failure rate mainly based on liver histology of about 72%. So about 860 patients have been enrolled, but 621 screen failed. And in the end, 247 patients were randomized. From these 247 patients, there were 19, so 8% that withdrew prematurely. Compared to other trials, this is at the lower end of a premature discontinuation rate. So that's reassuring. From these 19 patients, 9 withdrew because of adverse events, 2 were lost to follow-up and 8 were for other reasons, including withdrawal of consent or a patient that moved or noncompliance. So this means that 228 patients or 92% of the patients that were randomized, completed the 24-week treatment period. Next slide, please. In terms of patient distribution, about 90% of patients came from Europe and the United States. So that are the 2 main recruitment regions, as you can see from the next slide when it appears, there it is. So -- and if we go then to the next slide, you see some of the important baseline characteristics. In terms of sex repartition, it's about 58% female and 42% males in this study, mean age is about a little bit less than 54 years, with 80% being less than 65 years of age and the vast majority of the patients is from Caucasian descent. Now what you can also see on this slide, and I already mentioned it previously, one of the sub-analysis that we will do is to compare patients with diabetes to those that do not have diabetes. Diabetes is an important risk factor in this disease as you know. The relationship is a little bit double. If you have diabetes then, we know that there is an increased risk of having more advanced fibrosis and on the other hand, the fact of having NASH also predisposes you to an increased risk of diabetes. So these 2 conditions are intimately linked. And as you have also heard in the presentation by Pierre Broqua, PPARs have an important role in the regulation of glucose metabolism. And so PPAR agonists have also been used in the treatment of diabetes. So it's, for sure, something very important to have a close look at in this disease and especially in every pharmacological treatment of patients with NASH. There has been stratification for diabetes in this trial, but there was no prespecified number of patients that needed to be diabetic or non diabetic. As you can see from this slide, we end up with about 145 patients that do not have diabetes and 102 patients that have diabetes. So that will allow for the analysis that was planned. And as you can also see from this table, the patients with diabetes or without diabetes are comparable in terms of sex repartition, age and other parameters, which is also important to make the comparisons that are intended in the analysis plan. Also on the next slide, we continue with the baseline characteristics, again, comparing also a group with diabetes and without diabetes. And as you can see, mean BMI is about 33 and comparable between the 2 categories, if you split it by presence or absence of diabetes. And also the repartition in the different classes of BMI. We have a proportion, small proportion, but interesting, of lean NAFLD, about 6%. All the other patients are overweight or obese, which is classical, of course, for this population, about 30% are overweight and 65% have Class I or II obesity. Next slide, please. Of course, this disease is intimately linked to the metabolic syndrome. So we have to have a close look at the different components of the metabolic syndrome, which is highlighted in the next slide. As you can see, so the number of components of the metabolic syndrome is more at the bottom of this table. And as you can see, the vast majority of the patients have 3 or more components of the metabolic syndrome, meaning that they have the metabolic syndrome according to the preset criteria, but also the other ones, except for 1% that do not have any component of the metabolic syndrome, all the other ones have at least one component of the metabolic syndrome. And about 70% has the metabolic syndrome. Next slide, please. The next slide gives you and also the subsequent slides will give you an idea about the histological baseline characteristics of these patients. And on the first slide on histology, we have the data on the NAS scoring system and the SAF scoring system. So again, the NAS is the NASH CRN system, where for the NAS score, the NAFLD activity score, you combine the 3. And you also have heard about the differences in the assessment of ballooning and lobular inflammation between the scoring systems. At the bottom part of the table, you see the corresponding SAF scoring system where activity is also mentioned, but again, this is the sum of inflammation and ballooning. And we also added there fibrosis, which is the same in the 2 scoring systems. So that's why it's not in the upper panel because it's completely the same in the 2 scoring systems. And we have a mean fibrosis score of about 2.1. Again, this table also shows you that the population without diabetes and diabetes is comparable, which will be important for the analysis once we have the results and for sub-analysis of nondiabetic and diabetic patients. Next slide, please. Now as I already said in the beginning, our inclusion criterion was based on the activity of the disease because the activity of the disease is the driving force of this disease in terms of fibrosis progression but also extrahepatic consequences of the disease. And so a drug that had or a drug that has -- should be promising in this field is a drug that can have a significant impact on the hepatitis component, the steatohepatitis component of the disease because fibrosis will then follow. And of course, there are also direct, potentially direct antifibrotic effects of these drugs. So that's important. But if you want to really tackle this disease, you need something that tackles the metabolic and inflammatory drivers of the disease. That's again why the emphasis of the trial was on the activity of the disease. And we did not set a criterion for fibrosis, except the exclusion of patients with cirrhosis. Nevertheless, using inclusion based on activity according to the SAF activity scoring system, so lobular inflammation and ballooning of 3 or 4. As you can see from the slide, we end up with the vast majority of patients having F2 and F3 and only about 23% of patients having F0 or F1 majority or F1, we almost have by using this mode of inclusion, almost no patients with F0. And this, if you then compare to, for example, the Phase III study of obeticholic acid, at least the first part for the interim analysis, there, of course, they included patients with F2 and F3 and included an exploratory arm with F1. And if you take that whole population, you see that in terms of presence of F2 or F3, we have a comparable composition of the patient group. So the fact that you select patients based on activity of the disease and you include patients with really active NASH automatically enriches your study for F2, F3 patients, which is the target population in Phase III, as you know. And that makes a difference with other trials as shown here. Of course, comparison between trials should always be done with caution because populations are different. But the main conclusion here is that by selecting patients based on severe activity of the disease, you end up also with patients that have significant or advanced fibrosis in more than half of the patients nearly or almost 80% of the rest of the patient having important degrees of fibrosis. Next slide, please. Also in terms of the repartition of the NAFLD activity score because that's still an important parameter, of course, to look at. As Professor Bedossa illustrated with his study from, I think, 2014, by using a SAF activity of 3 or 4, you end up with, first of all, 100% of patients having NASH, having steatohepatitis. But you also end up with a large proportion of patients having high values of the NAFLD activity score, so the NAS. And patients have, as you can see from the slide, a lot of patients, more than almost 72% have a NAS of 6 or more, which is considered really highly active steatohepatitis. And you only have a minority of patients with lower NAS scores. Next slide, please. Again, if we compare with other trials in the field, and probably, again, it's interesting to look at the Phase III interim analysis cohort of obeticholic acid. You can see that in that cohort by their selection, which was a NAFLD activity score of 4 and then F2, F3 and then an exploratory cohort of F1, but with some other criteria, like a NAFLD activity score of 5, they ended up with 64% of patients having a NAFLD activity score of 6 or more, which is, as I said, 72% in the NATIVE trial. Again, illustrating that by this selection, we end up with comparable population as in this Phase III trial with a selection criterion based purely on the activity of the disease, so on lobular inflammation and ballooning. Next slide, please. So this brings me to my conclusion. So the NATIVE trial is a well-designed trial, where we include the most up-to-date NASH endpoints, classical endpoints of NASH resolution, impact on fibrosis, NASH improvers as defined by the NAFLD activity score. All these endpoints are included in the trial. The trial has run smoothly with only 8% of patients that discontinued prematurely. So 92% of the 247 patients randomized completed the study. There were 4 DSMBs during the trial period. And as I said, at every occasion, we were recommended by the DSMB to continue the trial without any change to the protocol because there was no safety signal. The baseline characteristics of the patients were perfectly in line with other NASH trials, registrational Phase III trials with more than 40% of the patients randomized having also type 2 diabetes, which is important that there is a -- this is a stratification factor and that will hence also allow the sub-analysis that I mentioned in one of the first slides. Importantly, also is that the screening strategy and the inclusion based on the activity component of the SAF scoring system allowed us to include patients with an elevated score of steatohepatitis activity and of fibrosis, with, as I explained, 77% of the patients having F2 or F3, so significant and advanced fibrosis. And 73% of the patients having a NAFLD -- classical NAFLD activity score of 6 or more, which is a highly active steatohepatitis. And as announced previously, we should have the results available next month. Thank you.

Thank you, Sven. I think we now move to the Q&A, and the operator will explain how to do that.

[Operator Instructions] And your first question comes from the line of Ed Arce from H.C. Wainwright.

Great. Thank you for taking my questions. Appreciate it, and very much appreciate all the detail in this presentation. So 3 questions for me, if I may. First is on the SAF score that you went into some detail about. Just wondering would this be the first Phase IIb trial that used the SAF score as a primary endpoint? And the broader question, I guess, really is just making sure this was accepted a priority by both the FDA and the EMA, correct?

Maybe I can take this one. I think Novo is also using the SAF score, maybe in the secondary. You should also -- as Sven, showed that it's been -- we're using the NAS and measuring the NAS on the secondary. And this, yes, is our accepted regulatory endpoint. So maybe Sven or Pierre Bedossa can comment more on the FDA and regulatory approval of the SAF scoring.

It seems like the relationship between the activity of the SAF score and the NAS, as Sven has shown, clearly demonstrated choosing this scoring really correspond to advanced disease with advanced NAS. So I don't think that there's any problem about this choice.

Okay. Great. And then I was just wondering as well, one of the assumptions that you made going into the design of this study, as you mentioned, was a 10% placebo response rate. I was just wondering how you came to this number? In particular, wondering how you reached that as a sufficient number given the rather high historical variability we've seen in other trials to date?

Maybe I'll ask Pierre Broqua to answer this question about the SAF.

Yes. Yes, sure. So thank you for your question. So I think the reduction in ballooning and inflammation that we are following as the primary endpoint is something which you can put relatively close to NASH resolution, okay? You would agree with that, with probably our endpoint is slightly less stringent than NASH resolution. And if you look at the placebo response, for NASH resolution across the different trials that I have highlighted in the introductory slides, you probably remember that this placebo effect ranged from 5% to 12%, which means that this 10% is -- that we have fixed for our study, is relatively well in line with what has been -- what we came out recently from all those trials. So we are -- yes, we are quite comfortable with the placebo effect size we have been working on.

Understood. Okay. That's helpful. And then given the strong evidence that suggests a significant antifibrotic effect, in particular, as you showed, last fall at ASLD with lanifibranor. What are your expectations here? In particular, you have a number of fibrotic markers that you're going to evaluate as secondary endpoints. And if you do see a significant effect, could the reduction of fibrosis be added as a co-primary endpoint to the planned Phase III?

Yes. I think it's -- I don't know. I think it's a good point. What can we expect relatively to fibrosis? So actually, if we refer to the Belfort trial on pio, with this study that lasted 6 months. The population was slightly more milder than the NATIVE population. So -- but if you look at the individual data of this Belfort trial and you focus on the patients that had F2 and F3 at the beginning of the trial, you see that half of them actually had a one point reduction after 6 months. So it's a small number, but that goes in line with the meta-analysis that I was alluding to, showing also that pio was able to impact or to reverse fibrosis in patients that had a rather high level like F2 or F3. So of course, NATIVE trial, you've seen the data. We have a majority of F2, F3s. We can analyze the data also by clustering F2, F3s over the others. We expect to see -- well, I would like to see, of course, a trend, at least a trend of an anti-fibrotic effect of lanifibranor in this population. And according to the data, as you say, we can readjust, of course, the design of the Phase III and the endpoints, but let's first look at the data.

And your next question comes from the line of Patrick Dolezal from Dovetail (sic) [ LifeSci ] Capital.

This is Valentyna on for Patrick. Just 2 quick ones from us. First, could you please speak to any differences you'd expect to see play out in the data in patients with type 2 diabetes versus without? And if there are any other subpopulations within the broader trial that we should be paying particular attention to? And I guess eventually, would it be a possibility to potentially seek approval exclusively in diabetics, if results weren't positive in nondiabetic.

Pierre or Sven, do you want to take the question about the -- what can we expect on type 2 diabetes patient?

I can start, if you want. So -- well, based on this meta-analysis, relatively to [indiscernible] and pio on fibrosis in NASH patients, showing a reduction of fibrosis in advanced patients. This observation was independent from the diabetic status. So pio is -- I mean [indiscernible] mechanism of action of PPAR gamma is able to impact fibrosis in type 2 diabetes patients as well as in NASH patients without type 2 diabetes. So we will see, I think, that, of course, the [ time-to-time ] mechanism of action can be very helpful for treating insulin resistance. And we know insulin resistance participated to the progression of NASH. So I don't have expectations really here. Just to say that the data with pio in NASH patients show improvement in fibrosis also NASH resolution independently of diabetic status. And regarding the -- and maybe Sven, you can go ahead with giving your opinion, and then on the approval in type 2 diabetes. I think that you will anyhow need a Phase III to do that. And it's a particular design. So we are currently working on that.

Yes, I can probably just add that indeed, we know from pioglitazone, but also from other trials like molecules like elafibranor, you have effect on NASH, both on diabetic and nondiabetic patients. So of course, we will have to wait for the results. But I don't think that probably it will only have -- if it's beneficial, based at least on the pathophysiology of the disease as well as the mode of action of the drug. I suppose it should -- if it's beneficial, it should be in both categories. Of course, patients with diabetes, we know that they are more prone to more severe disease and probably also to more progression of the disease. So it's a particular population to have a close look at. That's why it's important to have them well represented in the trial from that perspective, but not the other way around.

And your next question comes from the line of Lucy Codrington from Jefferies.

Just a couple from me. I guess looking at the screening failure rate, how does that compare to other studies that you're aware of? And if it is particularly high, does that reflect the more stringent inclusion criteria perhaps versus other trials? And then just based on, I guess, the description of why you chose to use the SAF activity score is very compelling. Is there a move to -- for NASH trials to start adopting this across other trials? Or is this still something that's relatively new? And then finally, with regards to the Data Safety Monitoring Board, was there anything, any particular factors that they were looking for at each of the reviews? Or was it just a general safety review?

Thank you, Lucy. Maybe I'll turn to Sven for your question about the screening failure rate. And if there are other trials using the SAF score or maybe Professor Bedossa can answer that. And then Pierre will take over the DSMB and safety question.

According to the screen failure rate, in fact compared to other trials, the screen failure rate is in line. We know that we have a very high screen failure rate if we have a prescreening on noninvasive parameters and all trials try to offer some strategies to narrow that down, but it ends up not to be very easy. So again, the accurate diagnosis relies on liver biopsy. It's true, of course, that we have a little bit more stringent inclusion criteria because we take activity and a high level of activity. But I think the results, at least of the inclusion clearly show you that it's also the population that has a higher risk of more significant fibrosis. And what we have also learned from other trials like the elafibranor trial or the recent data on obeticholic acid is that you have the highest benefit in patients with more active disease. And the lower your disease activity is, the more placebo effect you have and the less benefit you have from the drug. So I think that is a logical choice to make, and it's one of the strength of this study. But to answer the question, if this is an unusual high screen failing rate, I think it's not. It's perfectly in line with other Phase II and Phase III trials facing the same problems.

And then relatively to the -- your question about the DSMB, Lucy. So I think it's both. So it's both general and specific safety questions. And of course, among the more specific ones, you have the cardiovascular safety, the renal safety, the liver safety. So I think everything has been looked at in depth. And the fact that these 4 DSMBs actually was positive on a meaningful proportion of the population is fixed to the, you could say, the profile of the molecule.

And maybe Pierre -- Professor Bedossa, can you address the question from Lucy about do you know of any other trials that are using the SAF score?

So far, I don't think that any other use SAF score as the primary endpoint, but I know that many, if not most of the clinical trials that are now running on NASH, use SAF activity trial as a secondary endpoint or exploratory endpoint. And just I'd like to mention a recent paper by [indiscernible], which was published in Alimentary Pharmacology and Therapeutics, which was an independent review by 15 U.S. and Australian pathologists. And one of the question was what was the most useful scoring for activity. And for the NAS score, the question was, "Is NAS an optimal index for measuring NAFLD disease severity?" So answer was it is uncertain. But the same question for the SAF, the answer was it's appropriate. So to me, it's something that should be regarded very closely.

And your next question comes from the line of Frédéric Gomez from Pharmium Securities.

First, can you just talk a little bit about the different statistical considerations and your confidence level in being able to manage through that? And also because it's not clear in the slide deck, whether or not there is a [indiscernible] analysis. I'm just wondering because if it is prespecified in the statistical analysis plan, that you're going to test first the high dose versus the placebo. And if you're able to show superiority, you will therefore test the low dose against the placebo. And again, if there is a superiority, maybe you will compare both doses to see if there is an effect between the 2? And my other question is also anything on surrogate biomarkers that you are going to release or something that you can show in June? And the last one is on the Phase III because you said that you were already working on the design and the protocol, waiting for the Phase II NATIVE data, but you were already working on protocol. Can you maybe give us more color on the link between the NASH resolution and the fibrosis improvement? And whether if you see NASH resolution, we should expect the fibrosis improvement will come or whether that is not necessarily the case here? And are you going to discuss this with the regulatory people with FDA during the end of Phase III meeting?

Thank you, Frédéric. Good question. So the first one concerns the SAF and the Phase III. Maybe Pierre, do you want to take over for this statistical analysis plan and then where were we stand on the protocol for Phase III?

Yes, sure. So your question was about the way we're going to analyze the internal endpoint. So the statistical analysis will use the [indiscernible] methodology, meaning that the smallest p-value will be compared to a 0.025 alpha. And if significant, the second p-value will be compared to a 0.05 alpha. That means that -- to your question about first 1,200 or first or not, it will be first smallest p-value, whether it is related to the 1,200 milligram or the 800 milligram. Then regarding the Phase III, I think the question was look between NASH resolution and fibrosis. So I think that this has been addressed or maybe Sven, you can complete it. But this has been addressed in the Golden GFT trial when you see that the patients that actually improved in terms of NASH resolution, the improvers, the responders were also the one that demonstrated a decrease in the fibrosis score. So I think that -- well, that's an example, but I think that there are sufficient evidence in the clinical literature, indicating that the activity part of the disease, ballooning and inflammation, is the driver of disease progression to fibrosis. So I think that we can have some, well, good scientific arguments to defend the link between NASH resolution and fibrosis. But Sven knows much more than me.

Well, I can just confirm that we have a lot of data also from natural history studies with their biopsies that, although there might be some progression in patients that really do not qualify for the diagnosis of NASH baseline, what we have clearly seen even in those studies, so outside the context of clinical trial, is that patients that do show fibrosis progression, they have some ballooning or some lobular inflammation already at baseline, and they virtually all have steatohepatitis on subsequent biopsy if they are fibrosis progressors. And we also know from other studies that the fibrosis process is highly dynamic and clearly, also associated with the steatohepatitis. And as Pierre has alluded to what a sub-analysis of the elafibranor Golden 505 trial has shown is that those patients that have a regression in their disease activity, meaning there also the sum of lobular inflammation and ballooning based then on the NASH CRN reading. But just taking those 2 components and summing them up, patients that improved in terms of their activity were also the ones that improved in terms of fibrosis. And if they worsened in terms of sum of lobular inflammation and ballooning, that were also those with associated increase in fibrosis. So clearly, we do have data both from natural history cohorts and from clinical data that steatohepatitis, at least the activity component of it, and fibrosis are intimately linked. So it's -- there's a very solid scientific evidence for this link. An important point, of course, is to take into account that the biopsy is a snapshot and that the disease activity is somewhat fluctuating, and that gets some discussion in the literature. But if you take all the data from our conducted studies, the link between the 2 of them is evident.

And Frédéric, regarding your markers, associated biomarkers we'll be measuring in the trial. We will have TIMP-1, TIMP-2, hyaluronic acid, P3NP, NFS, FIB-4 score, ELF score, Pro-C3 and other inflammatory markers as well.

Okay. But you're going to announce news on that on June or it will be published later?

Probably later.

[Operator Instructions] And there are no further questions that came through. Please continue, sir.

Yes. Thank you very much. I think we can call it a day. I would like, first of all, to thank Sven and Pierre for their support and for being with us today. And also all of you who attended, we are more than 100 people who attended the call. So that's very nice, probably it showed an interest in our results. And then I would like to give you rendezvous to in June, where we'll be press releasing and also having a webcast on our results, which, of course, we are looking forward with, I don't know if it is optimism, but at least with confidence given the design and how well the design -- how well the study was showed. So thank you very much for your attendance and see you in June then.

And that does conclude the conference for today. Thank you all for participating. You may all disconnect. And speakers, stand by.