All right. Good afternoon, everybody. Thanks for joining us at another fireside chat here at Guggenheim's Healthcare Innovation Conference. It is 2026, and really thrilled to be here with the management team of Inventiva. To my immediate left is Andrew Obenshain, the CEO, and to his left is Jason Campagna, Chief Medical Officer. This is definitely an exciting time, no question about it. We are going to be talking about the upcoming phase III NATiV3 readout for lanifibranor in MASH, and really the commercial opportunities. Andrew, maybe just to kick us off, phase III readout is this quarter. Maybe just provide us introductory thoughts before we jump into Q&A.
Great. Well, Seamus, thanks for having us here today, and hi to everyone in the room. Yes, so indeed, we have our phase III readout this quarter, and so I think you can all do math. I figure there is nine weeks left in the quarter, and it is going to be one of those weeks. This has been a long time coming. So, Inventiva has been around for almost 12 years now. We are a Dijon, France-based company, for those of who do not know us.
We have lanifibranor, which is a pan-PPAR agonist for MASH, and we are very excited to read out this top-line data, hopefully to be able to repeat our phase II-B data, which showed an 18% effect size on fibrosis, regression of one stage or more, and a 24% effect size on the primary endpoint, which is a little unique, and it is, I think, going to set the standard for the future in that primary endpoint for those patients who had to resolve your MASH, not just keep MASH stable, and improve fibrosis by one stage or more. So, when physicians were looking at the slide, they needed to see that fibrosis went back a stage, that there was no ballooning, and that the inflammation was scored as zero or one, which is very clinically relevant to a patient and certainly to a physician.
I think this readout will do a couple things. Hopefully, we will be able to hit an 18% benchmark on that primary endpoint, and also, I think it will change it for the field. I think this is the endpoint that the field will look at from now on.
Okay. I guess one, the dynamics between TZDs and PPARs on efficacy and safety. Jason, maybe you can just kind of go through the phase II data in the context of the primary composite and the individual components, what we've learned from that, and how it differentiates from PPARs and how Lani differentiates from PPARs.
Sure. Let's start with the second part, Seamus. There are several PPAR agonists that are approved in various diseases. We have TriCor, PPAR-alpha that's approved for cardiovascular risk reduction, bezafibrate in Europe, cardiovascular risk. We've got now seladelpar and elafibranor, both approved for liver diseases. What they have in common is they're either mono-agonist or dual-agonist of different pathways. None of those drugs, sort of distinguishing feature number one, none of those drugs have ever shown the breadth of the histologic improvement that we showed in NATIVE 2b. So, we had, at all metrics, whatever you look, the underlying inflammation called MASH, we improved or resolved, and then the fibrosis, we improved.
However you mix and match those data points, whether it's fibrosis improvement of one stage or more and no worsening of your MASH or resolution of your MASH, either way, the histology was quite strong across all of those. When you look at the individual sort of groupings of biomarkers, they all moved in the exact same direction as the histology. So, what you think about, we talk a lot, Seamus, we've talked about it before, the so-called extrahepatic and intrahepatic drivers of MASH, but keep it super simple. What are the three drivers that we think matter in the disease? Atherogenic dyslipidemia, glucose and adipose tissue health, and liver health. Each of the bundles of markers from NATIVE all moved in a dose-dependent manner for the most part. Certainly, they all tracked histology.
No other PPAR agonist, whether mono or dual, have ever shown all of those things in the same trial.
Got it. Great. The differentiation from thiazolidinediones, they do hit PPAR-gamma, as does lanifibranor, but in a bit of a different way. Could you just help us understand that?
Yeah, sure. Look, in short, PPAR-gammas are really good drugs, the TZDs, and they are commonly used in both the United States and Europe, but they carry a tolerability profile which many clinicians can find challenging in certain patients. The whole point of lanifibranor was to try to capture that biology and dial back some of that tolerability concern. It all comes down to the structure of lanifibranor and the way ultimately that drug lands in the pocket. It's a partial agonist—
Yeah
—in the end. So, where TZDs, call it pioglitazone or rosiglitazone, are a full agonist at PPAR-gamma, we're a partial, and it's that partial that we think accounts for the clinical profile that we've seen so far in phase II, which is it looks to be a more gentle or muted PPAR-gamma signature. Not absent, not at all, but certainly muted when compared to pioglitazone.
Okay. Can you talk about the safety and tolerability that we saw in phase II, and where the sort of questions of edema and weight gain may in fact differentiate from the patient experience?
Well, why don't we tag team that one?
Yeah.
I'll go to the data, and then I'll let Andrew pick up the experience. The data in phase II NATIVE, you can think about, I think, Seamus, very easily. If you look at that adverse event table, what's not there are side effects associated with PPAR-alpha or delta. That's another important element of lanifibranor's design. The goal was to sort of dial back a lot of that sort of stuff that sank other programs in the past. We don't see much in the way of alpha or delta, so no muscle signal, myopathy, muscle pain, creatine kinase elevation, no kidney signal. All very good. What we did see was a gamma signature of weight gain. It's all come back to fluid retention, shown itself as weight gain and peripheral edema, and to some degree, dilutional anemia.
Those numbers, when you compare to the full spectrum drug like pioglitazone, edema is maybe about half, where we were in like 7%, 8%-ish roughly. Pioglitazone in a similar clinical trial might look something more like mid-20s, for example. They are very different from one another, but on the other hand, they're also similar. That PPAR-gamma signature is important, but back to the market and how docs view that.
I'll pick it up there. If you talk to physicians, there's two types of physicians that we talk that segment very nicely in market research. There's ones that are familiar with pioglitazone. There's about 4 million or 5 million scripts of pioglitazone still. You show them this profile and say this is an attenuated effect of pioglitazone in terms of fluid retention, but it has the potential to resolve MASH, reduce fibrosis, and then it's a no-brainer for them. They know exactly how to manage that profile. From a commercial standpoint, they'd be very, very comfortable using it. Then there's physicians that haven't used pioglitazone, who ask more questions about that fluid retention.
Once they understand that this is a profile of a drug, it's an attenuated profile of a drug that's already on the market that their colleagues are familiar with, it's a very straightforward conversation. So commercially, we view this as a very acceptable profile for the market.
Great. The differentiation threshold. I don't want to sort of over-index to 18% specifically or 20%, but is there a minimum threshold for fibrosis differentiation where you basically look at the data and you say, "We absolutely have a drug here"?
Yes, and I think what we've got it very clearly is that we think that 18% on the primary endpoint is an absolute win. And when we say that, I'm talking about a certain patient population that we think is appropriate for the drug at that 18%. And that patient population is basically all the F3s and all the diabetics. And so that is about 75% of the market.
Now, let's just play the hypotheticals here, right? What if we're worse than that? What if we have a fibrosis effect that looks a little bit more like the existing therapies on the market? We still have a couple differentiations. We work quickly. I mean, we worked in six months in the phase II trial compared to a year or 18 months for the existing products. And we have these metabolic effects, HbA1c control, et cetera. So even if the efficacy was approached maybe what we see in the market today, that profile, we think, would still make us very competitive in the 1/3 of patients that are both F3 and diabetic. Now let's go to the other side. Let's say we're above 18%. We get up to, let's say, 20, in the low to mid-20s. Well, then I think we're appropriate for all patients at 100%.
You can see there's kind of a sliding scale based on the efficacy readout, anywhere between 1/3 of the patients to all of the patients.
Yeah, great. In terms of the power of the study and how the phase III is run and executed, Jason, can you just help us understand, or maybe you can do a compare and contrast of what REZDIFFRA has accomplished and how you are managing for a robust result in this phase III?
I think the core of that, Seamus, it's really about the power and the fidelity of the study, right?
Yeah.
Is that what you're getting at? Yeah. Look, 18-month clinical trials can be challenging to run. Patients can drop out of those trials. Of course, these are long trial, a lot of clinic visits. What do you want to do? From a powering perspective, what you try to do is be honest with yourself based on the data that you have and not the data that you fantasize about.
Yeah.
The data that we had in NATiV3 were very clear, and we ran several analyses on those data, all of which are in the paper. What it showed is that we had a fairly large effect size of between 24% and 26% on that composite. What we did was we took that and we said, "What if it were wrong? What if we came in at some number much lower than that?" Not that we want that to happen, not that we think that it will happen, but what if we did? Let's size the trial for that kind of an effect size.
Yeah.
When we talk about that we're well-powered, we chose a much more conservative estimate of that treatment effect and the placebo response than what we saw in NATiV3, and then we over-enrolled the trial slightly. Where have other people landed? Intercept, back in the day when this was all brand new, was running in the sort of low 20s at 18 months for loss of patients. Madrigal, who came later, ran about 18%-ish roughly at only a year. There's a curve. When efruxifermin came in, their six-month data looked a little worse than what we saw in NATiV3, about 14%, 15%. By the time they get to week 96, they're at 35%.
You can see that as these trials go on, you can lose more patients, back to what I think where you want it to go. We're very comfortable with the powering of the study because I think that the teams back in that day at Inventiva, neither Andrew nor I were there, but did a really nice job of just being honest. You build a phase II trial for success, not on aspiration and hope.
Yeah.
I think they did a nice job of that.
Great. So appropriately powered and the fidelity as we kind of move to the final read-out you feel very confident in.
Underpowering, right?
Yeah.
I think the effect size is why you are going to clinical trial. That is the whole point of this. But from reading out statistically significant, I think we at Inventiva feel good about where we are.
Can you talk about the choice of moving forward with two doses, so 800 and 1,200, and what are the possible outcomes that drove the choice to bring 800 and 1,200 into the trial?
I will let Andrew pick up the relevance of all of that, but the simple point is that lanifibranor is a PPAR, a pan-PPAR agonist. It is a transcriptional modulator, as you know, Seamus. Those things take time to work.
Yeah.
We just put out some posters back at Paris MASH in September, finally getting to talk about some of our pharmacology.
Yeah.
You can see that that 800 mg dose, crosswalked from the animals, is at the low end of the efficacy range.
Yeah.
Six-month trial, preclinical pharmacology, clinical pharmacology says you are at the lower end of the band. We had every reason to believe that if you gave that drug at that dose more time to sort of mature, that it would ripen into more of a clinical effect that we saw. I think that was the rationale for including it, but the benefits of that in the market may be several.
If we suppose that there is a dose effect both on efficacy and tolerability in terms of the fluid retention, you have higher efficacy with the 1,200, then you have lower fluid retention with the 800, you can imagine that very easily segmenting the market. For maybe your F3 diabetic patients, you are using the 1,200. For your F2 non-diabetics, where you are not as concerned about the urgency to treat there—
Yeah
—maybe you use the 800. But we will see what the results say.
Got it. In terms of just, these are trial-specific, so whether or not it makes sense, but in terms of splitting the alpha for the dose selection and success of the trial, how does that lay out for each dose?
We have not guided on alpha allocation in the trial. I know that I am mindful of the note that you all wrote, and other people have written similar notes making assumptions. I think the point that we want to emphasize is that alpha is limited in the trial, and FDA gets a very strong vote on how that alpha is utilized. But you can think of it as one of two ways. Either you will allocate alpha across the doses and you can test them in parallel, or you will have one alpha and you sort of cycle through that alpha and test one dose followed by the other—
Okay
—and you can keep going until the alpha is off. But either way, you are getting the same alpha allocation to those groups. It is just the manner in which you test.
Right.
We have not given any more guidance on that, so I think, as Andrew said, sometime in the next eight weeks, we will have data and you will get alpha.
You will get it all. Okay, great. In terms of the profile of the patient tolerability of the product, you have had a 26-week trial. You have had this trial ongoing for quite some time. There is a lot of blinded data. Can you talk a little bit about just where the sort of tolerability dynamics shake out when you talk to KOLs who have had experience with lanifibranor?
Yeah, so I do not know if we can get into details about kind of conversations with investigators, et cetera. I think, but the question behind the question is, are you guys seeing any safety signals that you are worried about, right? I think one of the ways to talk about that is the fact that we do have a data safety monitoring board, right? They are going to pay attention to any kind of signal that the class has that would lead to something concerning, and let us just say the fluid retention leading to CHF, right? So, they better believe that that group is looking at what is the baseline rate of CHF in this population, and are they seeing any imbalance in the arms? If they are, you know—
Yeah
—to some point, they are going to flag it. We have had nine now, and they have said continue each time.
Got it. Okay. That is encouraging. Especially in the context of a longer trial, because my recollection of the TZDs when that was shown was actually pretty early in the treatment, almost an unmasking of underlying CHF. Is that correct?
You are talking the original pioglitazone?
Yeah.
Yeah, I think it is more right than not, Seamus. I think we could agree that on the one hand, pioglitazone showed cardiovascular outcomes, which were—
Yeah
—really positive. On the other hand, they had this excess of hospitalizations for CHF, and it really wasn't clear whether that was due to the drug alone, the epidemiology of the patients, but I think appropriately, the labeling language. I mean, it's a commonly used drug.
Yeah.
Right? When you're getting it in millions of people, you want to make sure the labeling language is giving providers what they need to make that call. I think back to where we were with lanifibranor, the entire sort of rationale for doing all of this back in 2010 or so was to try to work their way through those problems with just old-fashioned medicinal chemistry, a lot like the GLP-1s did. I mean, the original exenatide and the earliest GLP-1s, they did work, but boy, you got sick to your stomach and the half-life was terrible and—
Yeah
—some good medicinal chemistry, which Novo did to their great credit. I mean, here we are in the world that we are. I don't know, not comparing Lani to that world, that would be fantastic, but the same idea.
Yeah.
If we could do a little bit of medicinal chemistry, or could we dial out, tune out some of the things that we don't want and make sure that we keep some of the things that we do, that's the whole point.
Great exercise. What about the exploratory study that is baked into NATiV3? So, the opportunity to look at and understand better the F4 patient population.
You are right, you are correct. We do have that in the exposure cohort, between 75 and 100 patients. We will get that worked out when we unblind. It is highly unlikely that gets showing up anywhere around the top line datasets. It is just not the priority. So, I think I will turn it over to Andrew to talk about top line, but you should expect a lot of that to come out at scientific meetings.
Yeah.
There will be a fairly good amount of visibility into those cirrhotic patients right around the time we are beginning to talk about opening our trial. So that is our goal, is to have those line up with one another.
Got it. Perfect. I guess the one question, Andrew, is when we think about the sort of the F3 population and segmenting this whole population, remind us how big the patient population is that is F3, F2, and then where does the sort of diabetic patient population divide up within that?
Rough numbers. When we, by the time we hopefully launch this therapy, there will be around 400,000 patients with F2 and F3 disease that are under the care of a physician. About half of those patients are F2, about half F3, and it is about a 55%-60% diabetes. The largest segment, especially in the U.S., is that F3 diabetic segment. That is about 1/3 of the patients fall into there. That is not surprising because the rate of diabetes is a little bit higher in the U.S., and F3 and diabetes actually track together.
That is the largest patient population. But then there is about, as I said before, there is about, if you look at any patient that has F3 or any patient that has diabetes, that is about 75% of the patient population.
Okay. Got it. Pre-diabetes, is that still even a conversation any longer?
Well, when I say diabetes, non-diabetic, what I am really talking about is a continuum of metabolic risk factors.
Yeah.
I say that to be very simple. Physicians are actually going to sit down and look at a patient chart and say, well, for example, they have pre-diabetes, they have high cholesterol, they have some other metabolic markers. Well, maybe they're not obese, but they're overweight, right? I might start to lump that patient into a higher category of risk and look to be a little bit more aggressive with them. They're going to have a continuum of things that they look at, including the pre-diabetic or diabetic status.
Adding to that, Seamus, you asked earlier about the overall profile of lanifibranor. The HbA1c reductions with that drug are about maybe 0.5 percentage point. Back in the day, as you know, that would have been an approvable endpoint for diabetes.
Right.
Obviously, you need cardiovascular trials today, but in a world in which, take the saroglitazar data from AASLD, their median A1Cs on entry were like 6.2, 6.3, not exactly the same between them. The point is, you're getting pretty close to that pre-diabetic—
Right
—definition. But if you give patients F2, for example, that are on the bubble, pre-diabetes, early diabetes with A1Cs of like 6.1, you're lowering their A1C by 0.4, 0.5%. That's meaningful because you're converting them from diabetic to pre-diabetic or pre-diabetic not. This idea of sort of having the capability in one therapeutic to get at the histologic drivers of liver-related outcomes and work on those other elements that both worsen the MASH and predispose you from other elements, we think that that's going to be important out in the world. How important? We'll see when we get the data.
Right.
But we think based on what we are hearing now, that that is likely to be important to docs.
Okay. So let us talk about the. Imagine a world where the data are positive, and in the wake of success, what needs to happen from there? Time to filing, potentially, and then from a commercialization perspective.
Well, we are preparing already. So we have made a number of key hires, including our new COO, Chris Benecchi, to really start preparing that commercialization. So, assuming positive data, we will switch immediately to getting ready to file the NDA. We have said that we are going to do that in the first half of this year, and we will rapidly start to scale the team. We have added the brains so far, but then we will add the bodies as well afterwards to get ready for that launch, which we anticipate will happen in 2028.
Okay.
So already thinking about it.
Got it. In terms of just the go-to-market approach,
there are two distinct populations of physicians, I think, here, maybe certainly more, but the distinct ones being the gastroenterologists who treat, and then also, and the hepatologists that treat, and then the endocrinologists.
Yeah.
How do you kind of segment those patients? Who do you go to first? Or is it a little bit of a different approach?
Well, first of all, what is interesting is the profile of this therapy is one such that the two incumbents, one appeals to the GIs and HEPS, one appeals to the endocrinologist.
Yeah.
We actually have a profile that can cross over.
Yeah.
From a pure numbers perspective, the GI and the HEPS make up 80% of the prescribing volume right now. They're the most important, so we obviously have to focus there first. Then there are a number of endocrinologists as well that we could consider going after a high volume set of endocrinologists. We haven't made those decisions fully yet, but we certainly will be looking across those three populations, looking at what the prescribing behavior is in those three populations and deciding where we want to go. Again, I think we have a profile that appeals to both, so I do think we can go to both.
Got it. How do you think about the competitive landscape and the evolution there? Just the growth of the market itself today. You have two players now with labels with MASH. How does this evolve with Inventiva and Lani on market? What do you see as kind of the next actual competitors or potential competitors for Lani, even if we can, maybe there aren't any?
No. First of all, I think this is a market that's going to grow tremendously over the next decade, both from new diagnosis, treatment choice, and just a lot more players entering the market. So we are potentially the next one to be approved. You have the FGF21s coming down the pike. You also actually have some phase II data that came out at AASLD for at least two, if not three, different products as well. All of the new therapies have the same primary endpoint. It's actually we're shifting the treatment of this disease to really be thinking about resolution of MASH and resolution or regression of fibrosis at the same time. They're all going to be used together. This is going to be a combination therapy market.
If I had to predict the future, I would assume that for F2 and F3, there's going to be an injectable plus an oral as the standard of care. Pick your injectable with the GLP-1s that are currently available, or maybe these, in the future, GLP and glucagon combinations, and then there'll be an oral. Right now, we only see two orals, right? Ourselves and REZDIFFRA. But I think that'll be the paradigm of the future in a very rapidly growing market with lots of patients with whom we can benefit.
Great. Well, lots going on. Very exciting time, I'm sure. I'm sure you're gearing up to lock databases and have a great unblinding event. So, all the best in the weeks ahead. We're all waiting with bated breath, as I'm sure you are too.
Thanks.
Thank you very much.
Thanks so much for joining us.