All right, great. Good afternoon. Good morning, everybody.
It is morning, yes.
Welcome to day two of the Leerink Partners Global Healthcare Conference. My name's Tom Smith. I'm one of the senior biotech analysts here at Leerink, and it's my pleasure to welcome our next company to the stage, Inventiva, represented today by CEO Andrew Obenshain and CMO Michael Cooreman. Gentlemen, thanks for joining us.
Thank you for having us, Tom.
Huge year ahead. 2026, coming up to phase 3 pivotal data for your lead program lanifibranor in MASH. Maybe Andrew, you could just kick us off with some introductory comments, a little bit of background on the company, and then we'll get into some of the Q&A.
Great. Thank you, Tom, and glad to see a sell-out crowd here again. Inventiva, we're a French-based biotech company out of Dijon, home of the mustard as well as the company. Inventiva's been around for about 14 years by Negma Fournier, whole pharmaceutical company. Really, the basis of the company was rational drug design for PPAR agonists. To take everything that had been learned from all the previous launches of PPARs and development of PPARs, take the best of those, leave the worst, and develop a drug which is now being used for, to, or being studied in MASH. We have a pan PPAR agonist called lanifibranor. We are in phase 3 studies.
That is our single asset as the company and the sole focus of the company. We finished enrollment in beginning of April 2024. We are expecting our data readout in the second half of this year. For those of you who don't know MASH, it is a market that has been received quite a bit of attention recently. I think there's been a decade or so where there was a question about whether this was a real market. I think that we can safely say that Madrigal has put that question to rest with a $1 billion-plus run rate of sales. We've also seen quite a bit of pharmaceutical interest too in terms of the acquisitions of the FGF21s.
We do believe that we have the best-in-class oral for F2 and F3 MASH and are looking forward to bringing that forward to patients.
Awesome. Maybe we could start sort of high level with the mechanisms, kind of compare, contrast. You alluded to Madrigal and Rezdiffra in their THR-β, but, like, your approach with this pan PPAR relative to THR-β, FASN inhibition, like some of the other oral mechanisms that are out there.
Michael, maybe I can hand to you.
Well, let me turn to Andrew.
Yeah.
I think Andrew gave a really nice lead in, right? People were looking to sort of get the biology of what these metabolic reprogrammers, pan-PPARs, could do, but get it through a different mechanism. The challenge on the table wasn't that PPARs weren't good. They could be phenomenal. Fenofibrate, for example, bezafibrate, even pioglitazone, an incredibly good drug still used 4-6 million times a year in the U.S. The challenge was that TZD scaffold, for example, with pioglitazone, carried with it, on the one hand, good outcomes. There were really a couple good outcome studies, but the safety baggage was really problematic, and the similar theme were sort of across many PPARs. What lanifibranor is, it's a novel scaffold, an oral-based therapeutic that gets all isoforms in two critical points, Tom, symmetrically and relatively low potency.
The combination of those two for a variety of sort of really nerdy but interesting reasons is that you get a really good pan-PPAR agonist that's got a very nice profile in that the efficacy elements hit all three of the underlying drivers of MASH, and at the same time, you have absent or blunted similarities, safety concerns compared to the other PPARs. An example of absent would be that we had shown previously in our toxicology program and talked about it many years ago, we don't seem to have any of the muscle or kidney concerns that plagued earlier alphas and deltas. On the other hand, we do have some of the gamma element, but it appears to be blunted compared to what you might have seen with pioglitazone or rosiglitazone.
Novel drug, novel scaffold, engaging the receptors in a sort of novel way, creating, we hope, all the benefits of that metabolic reprogramming with a sort of softer, gentler safety profile.
Yeah. That makes sense. You have, I think, really compelling phase 2b data. This is, we saw placebo-adjusted NASH resolution rate 26%, fibrosis improvement 18%, very early six-month time point. Maybe just talk about the population you enrolled in that phase 2b relative to what you've enrolled in the phase 3 NATiV study.
You go ahead and-
I mean, we should just leave this, the preamble standing there.
Yeah.
It's population. Broadly similar, if not identical in most aspects. Three critical differences are it's obviously a longer and a larger trial, 18 months as opposed to 6 months in the main cohort. We have about 1,000 patients randomized compared to 247 in NATiV3. Big population differences. NATiV had F1-F3, so fibrotic NASH, whereas, as Andrew said, NATiV3 is only F2, F3. Second is that it's a snapshot of NATiV3 of contemporary medicine. There are more diabetics out there with NASH, and those diabetics tend to present with more advanced disease. We have more diabetics in NATiV3 compared to NATiV, about 55% compared to about 45%. We also have therefore more F3s.
We have about 67% compared to about 40% in NATiV 2.
Got it. In NATiV3, just talk about how you're handling background concomitant meds. I think in NATiV, no GLPs.
Yeah, correct. Well, GLP-1s were allowed, liraglutide, but what everybody means by GLP-1 is semaglutide. First of all, I think it's just important to say every trial has concomitant medications, whatever you're running. In the case of NASH, concomitant medications are standard, right? You're on diabetic drugs, you're on blood pressure drugs, whatever you might be. In NATiV3, we allowed, as a nod to the modern world, GLP-1s. That includes not only the old ones, liraglutide, but of course semaglutide. It's allowed as a diabetic therapeutic. We'll have reasonable exposures of patients about between 10%-15% of patients in the trial that had been on a stable dose of a diabetic level GLP-1. And then we'll have some less than 10% of patients that start a GLP-1 for diabetes during the trial.
Maybe up to 400 patients across both cohorts. That's about 1,400 overall in the phase 3 program that will have been exposed to some degree to a GLP-1. Not all of those are semaglutide.
Interesting. Patients could start GLP therapy in the course of the study?
Yes.
Okay.
Just importantly, not the MASH dose.
Right.
Diabetes or obesity dose, but not the MASH dose.
Right. Okay. Just remind me, did we stratify by either GLP or SGLT2?
We actually stratified by diabetes status and fibrosis stage, which will allow us to untangle. Then again, neither here nor there for this venue, but statistical analysis plans sort of routinely manage the start of intercurrent medications, whether it's for blood pressure control or diabetes control. All of those things are managed within the analysis plan, and it's relatively straightforward.
Great. How, I guess how does the difference in patient population that you've enrolled in NATiV3 influence? I guess it's the patient population, it's also the longer treatment duration influence your expectations for what we would see on fibrosis improvement at 72 weeks.
Yeah. Do you want to start?
Yeah. I think the expectation is that there's been a deepening of effect over time, so that should be the main effect on fibrosis. We'll be hopeful to see a deepening of that effect between the 6 months and the 18 months.
Okay. Yeah. I guess like how are you framing expectations on kind of what a good-
Got it.
readout looks like?
Yep.
Yeah. Across both fibrosis improvement where investors are particularly focused. You know, both all histology angles.
I'm gonna use that as a lead into the market as well. We think that if we duplicate the Phase 2b, an 18% effect size for fibrosis, we have a winner of a drug. The way we look at the market is about 375,000 patients, F2, F3, under treatment care. We divide that population into four roughly equal buckets. Picture two axes. On one axis, you have F2 and F3, about 50% each. The other axis, you have non-diabetic and diabetic, and about 40% of patients are non-diabetic, 60% are diabetic. The four, if you picture four boxes, the two diabetes boxes are a little bit bigger. Let's say we have an 18% effect size.
We believe that the very natural place for lanifibranor is in that F3 patient population, and especially in that diabetic patient population. We believe that we win there versus Rezdiffra, which has about a 12% effect size after 12 months. The place where Rezdiffra probably keeps a natural, you know, if you picture two patients walking into a doctor's office, one with F2 non-diabetic, one with F3 diabetic, the day after lanifibranor launches, where is that doctor gonna try lanifibranor first? F3, which has the best fibrosis data, diabetes, because it's diabetic effect. Two years later, those same two patients walk in, that doctor is much more likely to use lanifibranor. Would certainly be using lanifibranor in that F3 non-diabetic patient population. If they had a good experience, maybe even in that F2 non-diabetic.
Now, if you double the fibrosis difference, say it's not 18%, let's say it's 24%, then certainly you would think that doctor's gonna use it in that F2 non-diabetic patient population, because then the efficacy is really clear there.
That makes sense. Okay. Maybe, so we'll talk about commercial. I wanna bring it back to clinical just for a second.
Yeah
Just talk about how you think about confidence in hitting on the histology endpoints. How are we evaluating the biopsy slides at 72 weeks relative to the phase 2b? You're using actually a composite endpoint as your primary endpoint. Just talk about how you expect that to influence and reflect in placebo response.
We have a lot of confidence in the NATiV 2 dataset. The histology was very strong, both doses. The 1200 milligram, obviously all three endpoints were hit, and the 800 milligram dose, two were statistically significant. Then layered out from that, the liver-related biomarkers, the systemic biomarkers, like blood pressure reduction, triglyceride movements, and then the extreme, as Andrew was just hitting on, the diabetic elements, right? Hemoglobin A1c, insulin, fasting glucose levels, all aligned very strongly. We have tremendous confidence. Now back to NATiV3. Longer trial, larger patient population. Question one. Does reading methodology matter? It does at the margins, but not on effect. The drug will work or it won't work, because if it worked in phase 2, it should work in phase 3. That's been independent of reading methodology.
I think companies like Intercept, Gilead, Madrigal, they paved that road for us today because FDA in that era was very concerned about reasonably likelihood. It's a surrogate endpoint. Intercept and Madrigal had one thing in common, Gilead failed, is that they both have about 11, 12% effect size. In that world, you're very worried that the lower bound, 7, 8% on those confidence intervals, may not reasonably predict. FDA made a real big deal about reading methodology and walking through, and the end result is simple, two plus one. That's what we all do today. Some version of two readers plus a tiebreaker consensus. Us, Novo, Madrigal was approved on that basis. Roche, when they move forward with their 89bio, all the same.
The lesson is that if you're positive in phase 2, you'll be positive in phase 3, sort of independent of reading methodology, and that's been shown. Beyond that, we think that the population is so similar otherwise and enriched in the right places for the more extreme patients with F3 and diabetes, that it's the natural sort of sweet spot for lanifibranor. In sum, I think we have really good confidence based on the totality of the data in phase 2, and that the reading methodology finally has sort of been worked on and codified by FDA, and we're now taking a path sort of blazed by those in front of us.
Yep, that makes sense. One other efficacy question we get from investors is around your non-invasive. You mentioned the biomarker data, which looked really good. Maybe talk a little bit about what you saw on VCTE at six months, and particularly I guess VCTE, PDFF, did we see a dose response? I guess, like how does the non-invasive and particularly like the imaging and elastography data like inform your-
Yeah
Your confidence on histology?
This is always tricky topics because I think we all want sort of a fast answer like, "Oh yeah, the histology improvement, therefore look at a non-invasive." The reality is that the correlation of these non-invasives and histology is complicated. Even in the wild, if patient goes to a doctor's office, simple question, do they have advanced fibrosis or not? It took a decade to work out what exactly are the TE cuts and what exactly are the FIB-4 cuts? It's not as clean as we'd like it to be. That being said, there's some important elements. At only 6 months, caveat number 1, so really short time for the liver to sort of biologically do its thing, and it's pretty short time for transcriptional activators like PPAR to do their work.
That being said, there's some instructive data that we got from NATiV 2. One, the Pro-C3 levels declined pretty dramatically. Pro-C3 is a marker of new collagen deposition. It's like the collagen clip tail. So what you see when Pro-C3 levels go down is that you're laying down less new collagen. Good, but that's not the entire scar. There's existing scar and there's resorption of scar. That's where ELF comes in. We didn't see much in the way of movement with ELF in NATiV 2. If you ask me, not surprising. Doesn't matter what I say though. I think it's consistent with the biology. It's unlikely to have really extensive extracellular matrix remodeling after only six months. It's just not something you see. The TE is a mixed bag. I think the TE did decline, but it's almost certainly not due to the fibrosis change.
If you look at our LEGEND study, we have really good improvement on MRI-PDFF. We do defat the liver roughly 40%-50% movement in fat. As you know from NATiV 2, really good anti-inflammatory MASH resolution. That comes with edema. The first thing that happens when you put a FibroScan probe on, it's looking for elasticity of the whole organ. Less fat, a little bit less fluid could be offsetting any fibrosis improvement you get. The TE wouldn't necessarily be expected to move because there's a lot of competing biology under there. Over 18 months, the dominant signal in TE will be fibrosis. We expect by 18 months that ELF and Pro-C3 and TE will look a lot like that Pro-C3, and they'll be not only directional, but magnitude of the effect will match the histology.
Yeah. Okay, that makes sense. I wanna talk about, we've talked a lot about efficacy and kinda confidence on efficacy.
Mm-hmm.
Let's talk about safety tolerability, and I guess what are the expectations? Where we get the most questions from investors is around the on target effects, the PPAR-gamma effects, weight gain. Just help sort of frame expectations here as we think about a 72-week time point. I know we've also given I think some blinded analyses around what we're seeing in NATiV3 with respect to weight gain, but just help kinda frame like how should we be thinking about safety tolerability profile now that we're going out to 18 months?
You wanna start? Go ahead. Okay. I'll start. We're very comfortable with our safety profile of lanifibranor. I think that where I opened earlier about the structure function relationship of lani, simply put the prediction is the drug should act like a pan-PPAR agonist on efficacy, but not carry the same safety and tolerability baggage, or at least to the same degree as prior PPARs. What we have shown very clearly in our preclinical program, which we haven't talked much about, but historically the company had done so back in the day, and we'll talk more about it as the year goes on.
Preclinical toxicology, NATiV too, our scleroderma study, which nobody talks about, what was a phase two randomized year-long study that had a really nice look at weight gain, for example, over a course of a year, and then of course, NATiV too, which is the sort of metric on the wall. All of that shows the same answer. The drug is largely free from some of the PPAR baggage on mono and dual agonist, as I mentioned, like muscle, kidney, for example, but carries the sort of shape and shadow, what I like to call a blunted PPAR-gamma effect.
What we are seeing in NATiV3, when we talk about that blinded reveal, Tom, that was part of the PIPE financing back in 2024, I think it was, and the investors at that time wanted some better view that the data in NATiV3 looked or not like the old scleroderma data. Scleroderma showed that the weight had sort of plateaued roughly at around six months. The goal of that reveal was to give the investors in that era confidence on a pretty limited data set on a blinded basis of what weight gain looked like, and they're looking for same or different. Pioglitazone, we know the weight tops out about three years, keep going and going and going, but you've gotta get way out there before the weight begins to plateau.
What that blinded analysis did was, at least for those investors, give them certitude that they were gonna put in $400 million. They did. They felt pretty good about it. We press released those to sort of cleanse the world of that wall cross, and we have not done any blinded analysis since. I think the message stands from that era is that our view is that the gamma effects are muted. We will see some fluid retention, and we will see some weight gain, but it will not look to the same degree like pioglitazone.
Got it. That makes sense. Although we get a lot of questions around like weight gain specifically, I guess what I think is important is the discontinuation rates and kind of patient persistency.
Mm-hmm
out to 18 months. I guess one of the things I'm kind of struck by is within the semaglutide phase 3, actually pretty decent persistence rates, despite what we know in the real world is quite a different story. I'm curious whether we have visibility into discontinuation rates in NATiV3, I guess, how we're feeling about that, and then what we would think of as like a good persistence number relative to something like a sema.
Yeah. One of the things the investors in October 2024 had that same concern. One of the milestones they put in was that in order for us to receive the second tranche of funding, we had to have less than a 30% dropout rate, which is how we powered our trial. To be consistent with how we powered our trial, and I can say that we received that second tranche, and we were below that 30%. We haven't guided beyond that, but just to say that we, I think we were within expectations of what you would expect from a trial in terms of dropout of something that we pay very close attention to.
Understood. Yeah, I guess expectations relative to sema. I mean, it would seem to me that there's kind of like puts and takes to that.
Mm-hmm.
It's like oral convenience.
Yep
potentially less GI side effects, maybe offset by with sema, you know, those patients obviously experience some degree of weight loss.
Mm-hmm
Maybe they're motivated by that degree of weight loss.
Yep.
How do you think about kind of the puts and takes and your expectations, kind of framing it relative to what we saw from semaglutide at 18 months?
Yep. That I think we're not gonna comment specifically on numbers. Just to say, I think that that when we look at our dropout rates, you know, we can infer that either the physicians are doing a really good job or that the patients are feeling pretty good, right? Because you know, we're down below that 30%, and we were to have plenty of data to actually be able to analyze that trial.
Yep.
Tom, the other thing that I'd like to agree, no comment. I think retention rates in clinical trials are an area of intense focus for sponsors. We're paying attention. Second, we can actually point you to data that might help answer the question for those. In NATiV, too, the electronic data capture forms actually quantify. We know exactly how much everybody weighed at every point of the trial. So we have in our investor decks a very clear statement. About half of patients gained no weight at all. About 30% gained roughly 5%. So when we say no weight at all, 1%-2% background noise. About 30% gained 5% on average. If you look at the adverse event table for weight gain, it was a step down by a factor of three.
The reported incidence of weight gain as an adverse event was only about 10%, a little lower. The point is that docs are not looking at a patient who walks in their office. They might not even see the weight. The patient may not report it because they don't view it as a safety concern, which gets to your core question is, are they likely to be dropping from it? We like that experience in NATiV two 'cause it gives us a really good view of investigator behavior, and they're viewing weight gain as sort of part and parcel of what happens when patients are on investigational product.
Yep, that makes sense. I wanna ask about you enrolled an exploratory cohort of 410 patients who screen failed out, and that includes kind of I think the full spectrum F1 up to F4. Maybe talk about specifically like the F4 cohort and what we're hoping to learn from that cohort, and then, like, how are we thinking medium, long-term about prosecuting the F4 opportunity?
I'll start with the exploratory and you go to the F4.
Yeah, sure.
Great. It depends on when you look at the exploratory. It's gonna serve a lot of masters. It's a really important data set for us. One is that it's safety exposure data for a chronic treatment that we want approved around the world in F2, F3, so really important. 400 more patients with NASH, particularly advanced NASH, F3, F4, that have exposures for years, that's a positive thing. Second, we'll talk about in a minute, we're gonna leverage an outcome study that's gonna include almost certainly cirrhotics. Having safety data at both doses in that population, it can be really critical. I think third, it's a better view where you were going before talking about adherence to therapy in a more sort of natural real world environment 'cause they're not getting the same efficacy assessments.
These are patients that may not have the same schedule of assessments. It's more like they're on a drug in a clinical trial, but not under the same degree of burden for coming in for efficacy assessments. It gives you a real snapshot later of how patients are liking the therapy. Depending on when we look, there'll be different matters the data serves.
Yeah.
Right now, first and foremost is the F4 study.
Yeah, it will inform our outcome study. We are getting accelerated approval on the F2, F3, so we do have to do an outcome study. We're looking to design an outcome study where we find a patient population that is sick enough, so they have outcomes, but not so sick that we can't actually reverse the disease. We think we've identified a patient population there that is essentially a portal hypertension, a CACLD patient population. Patients with portal hypertension are the ones that portal hypertension or stiff liver that leads to blood flow backup is the way I think about it is what leads to outcomes. We will design our outcomes study around that patient population. That patient population is mostly F4, a little bit of F3.
That's gonna serve two purposes. Number 1, to confirm our F2, F3 approval, and 2, to expand the patient population to that F4 patient population. We have some pre-clinical data that suggests that we will be effective with that against portal hypertension.
Yep. Very interesting novel strategy. I think most are familiar with Intercept, Madrigal, basically taking your F2, F3 population through to outcomes. You guys are coming with a F4 specific outcome study. Maybe just walk through, I guess, level of regulatory alignment. I know this is an option that's laid out, I believe, in the guidelines, but I guess, like, level of regulatory alignment and the blocking and tackling, like you would have to have this off the ground at the time of submission, right, because it's a confirmatory study. Like just walk us through sort of the logistics around that.
Yeah, let me start off with the kind of just logistics, then we can pass it off to the regulatory side. Just from a logistical standpoint, you have to have the trial started by the time you file your NDA. What started mean is not well defined, right? Certainly have to have a protocol, certainly probably have to have your CRO. Whether you need to have a patient enrolled or not is up for debate. We certainly will get it going as fast as we can. We do anticipate that upon positive data, we would wanna raise some more money, fund that trial, have it going for the NDA filing, not only for regulatory reasons to comply with that, but because we think this is a really interesting patient population for lanifibranor.
Yep.
From regulatory pathway, we're actually in really good shape. The FDA has been quite outstanding at making clear what the framework of those trials need to be. These are outcomes trials, and they're liver-related outcomes trials. They know exactly what they want. One can do, as you said, Tom, look at an F2, F3 population if you're willing to wait long enough, or you can look at an F4 population enriched for this sort of clinically significant portal hypertension, as Andrew said. You maybe have to wait a little less longer. That's the point. You might need less of them to get the same outcomes. Either way, the endpoints are exactly the same.
That being said, every sponsor, including Inventiva, will have to negotiate with FDA on exactly what that protocol looks like pertaining to our drug, and we won't have final visibility into that until we get data from NATiV3. Why? 'Cause we're gonna see the safety and the efficacy of both doses, so we can have final discussions. We continue, we're in constant dialogue with FDA about this trial, and I think the key point for now is that we're really thinking about this late phase 3, early phase 4 well within the regulatory framework of what counts as an outcomes trial, and we think it will let us run a really good clinical trial that makes sense both time and size.
Yeah. Interesting. Okay. I wanna ask, Madrigal also has an outcome study that's ongoing in that F4 patient population. They're guiding to data in 2027, which could possibly coincide, I suppose, with you getting your F4 study off the ground.
Mm.
How do we think about probability of success for that study, but also, like, how could positive data there would that influence, I guess, your plans in F4 at all? This is a massive, totally underserved patient population.
Yeah.
Yeah.
Would positive data change the way you think about your outcome study?
Well, we certainly can't comment on whether we think it's gonna be successful or not. That's their study. However, if it is successful, I think it's good for the whole field because then you've validated the fact that you have a surrogate outcome that's been validated by a hard clinical outcome, so all the surrogate outcomes that we're studying have extra validity at that point, so it's good for the whole field.
That being said, I mean, look, I can't risk adjust their study. I can only go on the same public comments you all know. They're good drug developers there. We know many of those people. That being said, you know, Intercept ran an F2, F3 study for 6 years and an F4 study for 4 years. I know what those outcomes data were. Many of them are in the public domain, and looking at those data, I am not clear on how we're gonna be able to see clinical outcomes in that F4 study. Could we see histology? Absolutely. And that may be really important, but that's not within the FDA framework of sort of counting for outcomes. Is it important clinically? Very well may be, but that's not a regulatory path forward in an F4 indication.
We might be in a world in which they show histology, which I agree then would be phenomenal if you can actually reverse histology. The question will become then that clinically relevant dataset, what can they do with it in terms of a regulatory pathway? REGENERATE at Intercept was the only study that I'm aware of that was ever given permission by FDA to be a true histology study that counted as an outcome. After that, in 2018, the FDA moved to outcomes only, no histology.
Very interesting. I know we only have a couple minutes left. I wanna talk about plans post-data, perhaps post-filing maybe, but just how you're thinking about commercial build-out. You touched on, like, how you view market dynamics.
Yeah.
I think Madrigal has shown us, like, an ability for a mid-size biotech company to build out a commercial launch, sales force, resource it well. Obviously, the Rezdiffra is doing quite well. Like, how are you guys thinking about commercializing lanifibranor?
Well, I think they've actually laid the path out quite nicely, right? They have proven that a mid-size pharma biotech can actually go to that GI and Hepatology call point with a relatively targeted sales force, and be successful. That provides the roadmap for what we need to do. We need to raise a little bit of money and post data to fund that, but then we can actually go it. We can do that ourselves. We can launch ourselves now. Will pharma have interest along the way? Maybe, but that's not our. You know, certainly, we don't have to be reliant on pharma to buy us out to be successful. We can follow the commercialization path.
Yeah. That makes sense. Just in the last 30 seconds, remind us on the exclusivity assumption, the IP estate that you have around Lani and your expected exclusivity.
We've exclusivity to 2041 based on polymorph patents. That composition of matter actually expires in December 2026, so not a composition of matter, but we do have good protection out to 2041.
Yep. All right. Awesome. Well, unfortunately, we're up against time, but thank you, Inventiva team for joining us and sharing the insights. Andrew Obenshain and Michael Cooreman, a huge potentially transformative year for you guys ahead, and looking forward to it.
Thank you.
Thank you, Tom, for having us.