Sanofi (SAN) Earnings Call Transcript & Summary

So welcome again. Good morning or good afternoon, wherever you're coming from. It's Pete here from the Citi Pharma team. I'm delighted to be introducing Frank Nestle, who is the CSO at Sanofi and Head of Research. We also have with us the IR team. For those who have been on the conference, it's now day 3, 3 or 4. I'm sure you're hearing me say this, but I will, as always, I'm here to moderate, not monopolize. [Operator Instructions] I will make sure there's questions that I'll put it to Frank. This is going to be pure Q&A rather than prepared slides or commentary because there's a lot to go through. So maybe Frank, I'm kicking off -- sorry to interrupt you there.

So maybe just to kick off, before we get into discussing recent news flow or specific assets, maybe you could just set the scene because we're coming up to a 2-year anniversary of the CMD, where the management team laid out the sort of new vision for Sanofi and the new direction of travel. And I can say from a financial perspective, a lot of those targets are being exceeded. From an investor's perspective or market perspective, maybe too hang up on Sanofi's still perception on the pipeline. And with respect to vaccines, mRNA is being seen as a threat rather than a growth driver. But let's leave the vaccines part to one side. But you in your role as CSO since last year, Head of Research, perhaps you could start just by allowing us to understand what changes, if any, you have made to the organization to try and address those concerns, and then we can get into more of that specific asset.

Thanks, Peter. Pleasure to be here and obviously also discussing the emerging signs at some of the assets and everything else, which comes in the mind of you and your investor. So I mean, the bottom line is if we want to frame this, there's never been a better time to be an innovator or clinician scientists, leading teams of talented scientists like I can do or have the privilege of doing it at Sanofi. And actually, the ambition is to reimagine every single step of the value chain, how to define a target and make a molecule, moving it to indications and to patients with unmet need. Everything is up for grabs, and we saw that with actually a modality being born last year and 9 months, messenger RNA. We'll come back to it. Data science, we are changing how we think about time lines, return on investment, actually. And so it's definitely exciting times in life science in the biopharma industry. From my personal perspective, I started at Sanofi in 2016. I was actually attracted by Dupixent, a molecule, a sleeping molecule at that time, where I saw actually the opportunity of being a once in a lifetime drug, and that's sort of playing out as we speak. But also I saw a company which was mainly in the primary care space, cardiovascular, diabetes and actually was transforming itself into a specialty care company in immunology, oncology, rare neuro disease and hematology. And actually, with Paul Hudson and John Reed, we really are in this kind of hockey stick moment where we are transforming our pipeline, and I'm coming back to that. But in terms of my own research compass, I think one is really what I would call molecular disease maps, really understanding patient biology, at what I say is an atomic level. The atoms of the body are actually the cells. The cell is the defining unit of everything happening in terms of homeostasis and pathology, and we can now dissect these cells at a single cell level and actually get a 2,000 gene per cell resolution in patients. And we can do this not only in synovial fluids of arthritis patients, cerebrospinal fluid of multiple sclerosis patients, but we can do it basically across human pathology. And I think this will be very exciting. The other one is technologies to really unlock some of the targets we are seeing, which are typically and often can be undruggable. And I think we've added 6 new platforms in that space. And really with John Reed and Paul Hudson, we really went after opportunities where you often have assets, and we're going to talk about those, but there are platform technologies behind it. For example, in biologics, Ablynx, Kymab are just leaders in that space, and we can discuss it. Synthorx, synthetic biology to really craft completely new cytokine molecules. Kiadis as a next-generation cell therapy leader. And then something I really want to discuss with you, which is Tidal, essentially in vivo CAR-T cell therapy, in vivo immune cell programming. So really creating a node which is durable, sustainable for Sanofi. The final point is really around data science. I mean we're writing now our own algorithms, our AI algorithms to actually understand our single-cell genomic data. Our chemists are essentially cutting their time lines to get to a lead candidate by massively -- because they just need 500 molecules to synthesize instead of 5,000. And also [indiscernible] is essentially giving us the tertiary structure now of a lot of the targets in combination with [indiscernible]. So this is really a fantastic opportunity and in our research strategy 2.0, we really halved our costs of the molecule to bring into the clinic. We've shifted to biologics, first-in-class, best-in-class 80%, and we had a [indiscernible] refocus. Now how has that impacted our pipeline? We've started 10 Phase III trials over the last 18 months, 4 major assets: Dupixent, tolebrutinib, amcenestrant, itepekimab. We're building a comprehensive immunology pipeline. And if you have one aspiration, and I'm happy to come back to that, is to be really the immunology leader on the back of Dupixent, and we hopefully talk about some of the assets. Oncology, I would call as reemerging. Sarclisa, Libtayo give us sort of a credible play on the stage. But amcenestrant was the talk of ASCO. So that's how -- where we want to be. So 2 for 5 are IL-2. So that's basically where the aspiration is. And then also in genomic medicine, we are building on our Genzyme legacy and there's much more to come. So really exciting with catalysts such as Dupixent EoE and prurigo nodularis coming up, amcenestrant. And if you want to know more about vaccines, we have a vaccine event December 1.

Before -- again, before we go diving into all these wonderful technologies and assets, you have been busy on the M&A front, and you'd be very consistent with your words. So I refer to the strategy laid out in December. My first question is how have you approached -- I mean you just left these companies to sort of operate very independently. Or how have you brought them into the Sanofi stable? Because there's quite a number of deals that you've wrapped up in the last 2 years.

Yes, that's always a difficult question and you try to develop a playbook and then you have to rip it apart when you meet the teams. I'll give you 2 examples, which are, I think, very successful integration. One is Ablynx. Ablynx 2018, we had a 2017 deal with 8 assets and really, I champion this as a next-generation biologics platform and obviously, a couple of years in the clinic. The platform is derisked, and we have multiple nanobodies coming into the clinic. But what is most exciting to me that we have these 300 drug hunters in Ghent, and they're all there. And they basically gave us that sort of biotech spirit, and really challenged our own large company presumptions and are now part of our talent family. And just the fact that we could retain basically all of the key players they really have, and there's this great dialogue now is really a success story. The other one is actually -- we just had a team coming out of LabCentral in Cambridge, the team from Tidal. And the Tidal team essentially is quite essential bleeding-edge biotech. They've shown fantastic data that they can deliver RNA to immune cells in vivo and actually turn them into CAR-T cells, which at least in [indiscernible] models are equally potent or better than your usual CAR-T approach. They moved into our Cambridge place, and they've been very happy ever after and are connecting with the environment. So I think those are the good examples. I have to say struggles are clearly Zoom, Zoom environment. Kymab, I mean one of the pioneers in biologic with Allan Bradley. We used to run the Sanger, and I know him for many years. I haven't been able to visit in person, you know, because -- so there are challenges in the Zoom world. But overall, I think it's what kind of inspires also our Sanofi scientists, in addition to the assets we've gathered in the technology platforms, which create nodes.

Very clear. And before we go, I do want to go back to title, but my last sort of big picture question. And apologies, I realize you're not the CEO and CFO -- or CFO, so it's a little bit cheeky. But just on capital allocation because, again, the message was very clear back in 2019 and the actions that you've taken to date are very consistent with you're looking for technologies, platforms, assets that add to your early-stage pipeline and beef up the Sanofi pipeline and deal size below $5 billion. Now there has been recent market speculation that of your potential capital allocation that doesn't fit that filter. And I realize that commenting on specific market rumors is not what the CSO does. But just kind of reiterate that the capital allocation [indiscernible] clarify the capital has no shift or change in Sanofi's capital allocation policy as it relates to M&A related to the pipeline.

Yes. I mean I obviously cannot -- as you know, M&A comments are for the CEO and CFO, and I know they have other opportunities. But as you've seen from our past strategy, the bolt-on acquisition strategy has been quite successful in terms of really getting the best of both worlds, typically a great asset and also a platform which creates a sustainable and innovation engine. I think that's what we need, that sustainability also for our pipeline. We don't want to be one trick wonders, where we have one success and then a lot of potential failures. It's about that sustainability of these multiple platforms coming through. And if you just talk about the latest, Kadmon, where essentially it's really an interesting place, an FDA-approved, ROCK2 inhibitor, which basically actually is in the -- mechanistically in the immunology and fibrosis space, where we're very interested in R&D. It's a first-in-class drug. It's exactly where we would be in R&D, but it actually shows that our gen med business is literally ready for business in terms of getting from a tail perspective into a really a growth perspective. So I'm very excited about those types of moves where basically you see that innovation and first-in-class perspectives are really touching every part of our business.

All right. All right. So let's talk to the science and some of the assets. And let's start with -- you already mentioned nanobodies from Ablynx, the mRNA therapeutics with Tidal, the triagonist, but when we think about actually candidates entering the clinic, could you sketch out sort of any time lines that you're willing to give at this juncture when we might start to see candidates enter the tale?

So you're talking about nanobodies or Tidal or...

Well, all of them, I'm afraid. All is one -- is -- [indiscernible].

All right. Yes. So maybe we start with nanobodies. Obviously, 2018, we brought Ablynx into the Sanofi family. We had that quite successful merger. And what we did immediately kind of developed multispecifics from scratch, essentially moving them into the clinic. And I think we've shown at our Capital Markets Day in February that we have now multiple bispecifics, which are basically knocking at the door. And typically, it takes you 4 or 5 years to get a molecule into the clinic. And here, we have a whole range of nanobodies, which are actually moving into Phase I. And we talked specifically, for example, about our TSLP IL-13 nanobody and revealed that, but we also said that we would move nanobodies into rheumatology, into gerontology indications, into a gastroenterology indications. So more to come about that. But just to put the nanobodies into perspective, the industry has failed in I&I to really deliver good bispecifics. There's a lot of failures there. And with a nanobody platform, I think is an IL-17 A/F nanobody, which is really a very nice Phase II data, that the platform is derisked also. Cablivi is in -- went all the way through approval and is in the clinic. So this platform will give a lot, and it can do more than just bispecifics where like a string of feeds you have this kind of modular single domain nanobodies are connected with a linker. You can think about tissue conditional activation. You can think about tuning your half-life to basically make it very short-lived, but you can also give it an antibody-like half-life. So there's a lot you can do with it, and you will see more of that to come. Now in terms of Tidal, that's probably our -- I would say, it's like one of the biggest goals. And I didn't really get back to your messenger RNA question, and maybe I can address it. We've been in messenger RNA from 2018 when we actually had that deal with Translate Bio and basically occupied their vaccine activities. Why Translate Bio? Because they've been in there from the early days. They've been in there from the days when Moderna was born, and they know a few things about making RNA molecules, about making LNPs. And we are very excited about to see what's happening with our Phase I/II readout now coming up this year. But we're also in other areas of RNA. We've been -- we have had a collaboration with BioNTech on cytokine messenger RNA. We're using messenger RNA to basically deliver our -- basically a nuclease to stem cells in sickle cell patients. And the latest is we're using messenger RNA with Tidal to basically wrap them into a lipid nanoparticle, covalently link a nanobody to it and then specifically target them to immune cells such as CD8 T cells. We can transfuse about 40% of T cells in preclinical models. We see that the activity is [indiscernible]. So this gets us into what we call in vivo immune cell reprogramming. This is the big next frontier. I mean we can do a lot of things with messenger RNA, but can you really reprogram immune cells? And next steps might be in that we reprogram other cell types. So the Tidal technology I'm very excited about, but also, I speak as CSO. I'm responsible for the future in transformative science. It's certainly not your Phase III asset, which will create -- which will show up in your spreadsheet immediately.

So when can I start things on a spreadsheet for the nanobodies and these RNA [indiscernible]?

Yes. I told you. I mean, we will have multiple Phase Is kind of starting this year and then also in the next year. And for Tidal, we are obviously at this moment, working on the clinical candidates.

And just a clarification. On your nanobody, you said IL-17. Are you exploring A/F? Or is it in IL-17...

I just said that there is a -- so this is not a nanobody we have in our portfolio, but it's something which is out there and derisk the platform. I mean it's basically a previous collaboration Ablynx had.

Got it. Okay. Given that the investors are on board, it would be remiss of me not to talk about some of the late-stage assets. And obviously, can we have a little bit of a post mortem on rilza. The market obviously pretty much zeroed that asset now, platform now. So maybe give your thoughts as to why there's still life in this asset. And then regardless of whether the commercial potential or the commercial indication as it relates to AMEERA-3 is modest, I think symbolically or from a [ central ] perspective, that data takes on some importance from a central perspective. So just how you're thinking about the setup on amcenestrant, especially as you have 2 very vocal competitors that are pushing their selves forward as the potentially superior.

Yes, happy to discuss that, rilza first and amcenestrant second. So I think if you take a step back, what I&I really needs is an oral immunology leader drug, and an oral immunology leader drug needs to be effective and safe. And we've seen that with the JAK class. Safety is very important. I mean what we're hearing is on the JAK side that there's now a false black box label, which relates to MACE events, cardiovascular strokes, myocardial infraction. And that's a very difficult situation to be in because you're then relegated, for example, in RA, to second-line. In dermatology, neurology, you have very safety-conscious prescribers because we have -- think about dermatology. You have often children or adolescents with atopic dermatitis. We just had pivotal data on -- that we can now go down with Dupixent to 6 months old. So it's really important to be safe. So if you look at the BTKi class from that perspective, I think it has been very nicely derisked that, with certainly our compounds, rilzabrutinib, for example, we can be very safe in terms of delivering in target engagement of BTKi. Now the next question is about the efficacy to be a real oral immunology leader. And I mean, you know that. We've seen really nice clinical data on multiple sclerosis and rheumatoid arthritis, CSU, which is urticaria. So really multiple times we've seen that if you block BTKi, that is a nonredundant mechanism, which actually impacts disease biology. Now very specifically, what we've seen for rilza in pemphigus vulgaris, and I've treated this disease myself, so what we see is blistering. It's actually very difficult to do a trial in it because patients are always on steroids. And the amount of steroid you co-medicate are basically deciding where your endpoints are like CRs. So a little bit like systemic lupus. You know how in systemic lupus, it was all about the steroid tapering. With pemphigus, we have less experience. The first BTKi trial in pemphigus, and we had, for this Phase III trial, we had like a 10-milligram tapering schedule down to 10 milligram. And what we are seeing -- what we're seeing and we just saw the data, I just saw it myself and obviously communicated them to the Street, high-level, top-line data. We weren't superior to placebo. Placebo got steroids, too. So it's like we didn't hit that sweet spot where the drug could shine. But if you ask me, what about the other -- is there a read-through to, for example, our ITP program, immune thrombocytopenia, where we saw in Phase II in 50% of the patients, essentially in just a matter of a few days or weeks, we saw really an increase of thrombocytes level to above 50,000. And that was due to the fact that if you think about the mechanism with immune thrombocytopenia is macrophages chewing up antibody-labeled thrombocytes in the spleen. That mechanism is very different BTKi-dependent mechanism from what you have in pemphigus vulgaris. So we are fully committed to these ITP trials as we are to our atopic dermatitis, asthma CSU Phase II trials we're going to start, where again, the mechanism is different. In this case, we have the IgE receptor, which signals in a BTK-dependent way, a different mechanistic activity of BTKi. So think about these diseases as sort of different buckets where BTKi can essentially just act, and we're going to go one after another and rilza is going to demonstrate the data it requires to get through the bar. So that's on rilza. On amcenestrant, I think it's really important to understand that we're reaching a new sort of era of ER degradation, hopefully with benefit for women suffering from ER+ breast cancer. And there are a variety of molecules in the race, as you say. But amcenestrant certainly has a great profile. And again, I'll probably start with safety, where essentially on the safety side, which is actually going to be match deciding if you want to combine it with the likes of palbo, if you want to move it in the adjuvant space. This is where the real -- the real need is. And I think we have very nice safety data. We're avoiding some of the things which you are seeing with competitor molecules, which actually could be bradycardia, visual disturbances. So that safety element is very important. Now on the efficacy side, when I talk to KOLs, they were already excited about the monotherapy data we were showing. But what really then convinced them was really the palbo combination data where we had objective response rates of 34% and a clinical benefit rate of 74% of patients. So I think the plot is thickening on that efficacy side. And I wouldn't argue that AMEERA-3 is the only readout. We will have many readouts, as you know. We've started AMEERA-6. We have AMEERA-5, where actually, AMEERA-5 is [ a group ] of people who are not in the AMEERA field. AMEERA-5 is where we're essentially running our Phase III combination of palbo plus amcenestrant. We are seeing an incredible pickup in recruitment. Now this is actually investigators and patients [indiscernible]. You get this if you have a great safety profile and if you have a profile which is beneficial to patients. So all of that kind of makes us confident. As you know, AMEERA-3, I get a lot of questions for it, is a trial which is essentially event-driven, and we'll report it at [ ASCO ] sometime this year. And it's a very interesting space because it's second-line, third-line metastatic breast cancer, where there's not a lot happening. It's in the post-CDK4/6 space. So it's a very interesting space, and we'll see what the data will show us. But there's a plethora of data now, which make this molecule very competitive, and we're doubling down in, obviously, letting it shine and making a benefit to patients with breast cancer.

Okay. I've got a few questions that have come in on this topic, and I want to also layer on. We had a breast cancer doctor panel yesterday, some questions around that. So in no particular order, I mean our doc, [ Ian ], the doctor we had was very constructive on amcenestrant, the data you published in ASCO. The way he set up his thoughts on AMEERA-3 was that the standard of care, you would expect to show a PFS of 3 to 4 months, with amcenestrant maybe giving a 1- to 2-month benefit on top. So with that -- with the trial being fully recruited back in midyear, I mean, is it right to think that this data is around the corner? Is there anything you could say on timing? That chimes in with an investor question. And another question around from an investor is the interaction with CDK4/6. One of your competitors are saying that you can't fully dose amcenestrant because of drug-drug interactions. Any comments there? And then rounding out with my question, just the risks around adjuvant, ensuring compliance. We've seen the PALACE study fail because [indiscernible] patients weren't the drug. How are you thinking about mitigating that risk with your adjuvant study?

Maybe I can start with the last one. This is exactly the point I was trying to make is the patient not taking the drug. This is exactly what the profile is amcenestrant will have. They take this drug. And we've seen that with a pickup in the AMEERA-5 study where you combine palbo with CDK4/6 with amcenestrant. So I think one of the -- and that's probably the best summary of what I tried to explain is and you sort of rephrased that. You have to get patients to take that drug. I mean, you obviously -- prescribers will be receptive to that. And I think that's where amcenestrant will shine. In terms of the DDI, I mean, we've seen the data. We can have 200-milligram QD with a standard dose palbo. We get great objective responses and clinical benefit rate, as I say, and with a favorable safety profile. So I think we have no -- we don't believe there's a dosage issue at all, to be honest. I think it's purely constructive. And then I don't know. You had another question, which I didn't address.

Yes, the setup on -- the way that the breast cancer doc was setting up his thoughts around AMEERA-3. Is that consistent with you that you'd expect that control arm in AMEERA-3 to...

Yes, it's really interesting. I mean, 3 to 4 months is probably a good call, but that space has not been widely traveled. We're talking comparison to standard of care with docs prescribing fulvestrant and other works. So it's really -- we'll see where that standard of care then pans out, and that will obviously depend on what the control arm would show. So that's why I'm saying that the trial is great because it compares to really true standard of care, but we'll have to see what then these PFS will be. I mean, 3 to 4 is what has been circulating. I've quoted this number before, but the data will tell.

Okay. Can we go back to one of your scenario close to your heart, immunology. You're obviously very enthusiastic about Dupixent and you've got a strategy beyond that with OX40 and BTK and IRAK. But I have a question here from an investor that just adds on to that, and I want to add on something as well. Let me find it. Hang on 2 secs. You have a topical BTK in development for AD. Do you think topical is a real market opportunity? I think the point of the question is you know that compliance with topical agents is terrible, and payers don't want to pay for it. All they're going to ask that the patient has failed to topical-acting steroids before they give access to the self-selecting for the worst patients. So I suppose just can you talk about your enthusiasm for topical-acting agents in AD over and above your order portfolio? And if I could just chuck in one more thing, Frank. We've seen the [indiscernible] something quite exciting in psoriasis. Modest efficacy, but great safety. There are other sponsors out there trying to repeat that trick with PDE4 inhibition in -- orally in AD. Is that something that you find interesting? Or are you very comfortable with the current portfolio that you have?

Okay. So that's interesting question. It's not only immunologists, but also previously practicing dermatologists. I mean it's clear in atopic dermatitis that often when a patient walks out of the door, the patient will -- the doctor will also prescribe a topical. That being said, I think the real changes we've seen is with biologics in that space like Dupixent. I think what you can do with topical, you can control lesions, and you can basically create a combination scenario, which is beneficial to patients. And I think it's great for us to have a topical BTK because we're in dermatology. We basically have a whole franchise of molecules and pipeline of molecules we're moving into dermatology, especially in atopic dermatitis. And if you look at immunodermatology numbers circulating in 2025, I mean, we're going to be right there at the top. So I think it makes sense. If you already have a great sort of pipeline of molecules in a franchise like we're building in dermatology, just a topical pure play can never compare to a biologic, which basically has completely different -- it's a different ball game in terms of efficacy and impact for patients' lives. Now coming back to the PDE4s. I think the topical PDE4s and they do a little bit, but if you look at oral PDE4s, there's obviously oral like apremilast. They're out there. I mean their Phase II data published by Simpson, et al, I think, in 2019. And the IGA 1 -- the IGA 0 to 1, which is -- this is a global assessment of 0 to 1, which almost, which corresponds to almost clear to clear, was only 11% of patients. 11%. I mean -- so I think that's for the PDE4. I mean I'm always data driven. Certainly, we can deliver them relatively safe, you flush a little bit of the [indiscernible]. But I don't see the efficacy at all in the AD space, in the oral space. And topical is, obviously, a different question. I mean, there are different comparators and you always get a lesion here or there or regressing in.

Very clear. Can we shift gears. BTK in multiple sclerosis, but also the neurodegenerative diseases, so beyond MS. So anything you are willing to say, again, a little bit like the SERDs. It's a competitive space and every competitor is saying their BTK in MS is the best and they're recruiting really well. So maybe you can chime in with your thoughts on the [indiscernible] program in MS. I'm very interested to hear about beyond MS, when might we see trials start for some of these other indications and what indications are particularly excited for that?

Yes. I think one of the most exciting moments post Capital Markets Day when Paul Hudson was announcing his new strategy and we were saying, Tony Gordon, it will be one of these assets. And then we saw the Phase II data, which basically we're confirming that we had biologic-like efficacy with a safe molecule, that was really a true moment of the new Sanofi, and we're basically going to have this Phase IIb study follow-up. We have an open-label extension, and we'll report it at ACTRIMS and other conferences. So that's going to happen. We also embarked with support from Paul on really a very ambitious Phase III program, 4 Phase IIIs, and we basically expect that sort of data will read out to 2024, 2023. So that's basically maybe as sort of the earliest readout. And in terms of where this could go, obviously, it could go in B-cell mediated neurological diseases. This includes the likes of myasthenia gravis, neuromyelitis optica. We have not really communicated this as a final fact. But you'll hear more about this year, but there are obvious candidates to move the molecule into. I mean, that's for sure.

So I'm going to be jumping -- jumping around a little on subject matter as we go earlier into the pipeline. So perhaps I personally, I'm quite liking the NK cell therapy, the Kiadis deal that you did. That looks pretty interesting, off-the-shelf cell therapy. So could you talk to your excitement post that technology, the programs and when we might see next data points?

Yes. We are sort of excited about it for a variety of reasons. I mean one is the cell therapy space is now you see more and more also large players engaging with that space. And you want to make your mark. And for us, getting into NK cells was because when we look at the CAR-T successes, I mean, they're mainly restricted to liquid cancers. They are individualized. They're expensive. They have safety signals. We just saw when they move into prostate cancer, you see [indiscernible]. You even see fatalities, CRS. So there's a lot going on there. And with NK cells, we know that they are very good in tumor immunosurveillance. They're sort of a next-generation approach to cell therapy because they're off-the-shelf products, really like something that comes with the company obviously likes [indiscernible]. They're well tolerated on the safety side. They are amenable to genomic engineering, so you can turn them into CAR-NK cells. And actually, we had, at ASH in 2020 and in relapsed/refractory AML, we had 66% of patients with a complete response of CRI, CR, CRI in that really kind of late line setting. So they're really clinical features of response. Again, liquid cancer at this stage, and so I think it's a very exciting time. We believe we have the best NK product. It's called K-NK. They are essentially expanding with IL-21 and 4-1BB ligand. There are obviously other NK cell types out there, but we believe that this is sort of a universal donor and best-in-breed NK cell approach. And they can go into a really interesting -- into really interesting indications. I would also say that we're making a pincer move on NK cells. We became excited about NK cells because of Sarclisa. NK cells are a major mechanism of the antibody that mediated killing of myeloma cells, and so we're investing in programs to make NK cell biospecifics. So essentially engagers, NK cell engagers. So like you have your T cell engager, think about an NK cell engager where you're not then facing the same, for example, CRS issues. So that's very exciting. Again, moving potentially into -- moving into a liquid cancer setting. And then we have our IL-2 program from Synthorx, where we actually -- we showed data this year where we can expand NK cells eightfold and T cells fivefold with our IL-2, Synthorin IL-2, which, just to remind everybody, it's a technology where you synthetic biology to really reengineer or reinvent the genetic alphabet, putting in a nonnatural amino acid in your IL-2. And by a click of chemistry, you click on a peg, and then very exquisitely, you can shield the alpha part of the IL-2 receptor which is where the Treg activation would happen and focus it on the beta and gamma part where the CD8 and NK cells are expanded. And we have shown clinical responses with this compound, both in pembro combinations but also in monotherapy. But the exciting thing for the NK cell story is that you can then actually combine it. Think about that. You can then bring in the IL-2 plus your NK cell [indiscernible] or the IL-2 [indiscernible] cell therapy. So we have a whole -- this is a whole throughput surrounding NK cells. Now we'll see where this ends, but if NK cells will be the next big play in cancer immune therapy, we'll be right there to exploit it.

Got it. I've got an investor question now. Jumping back towards atopic dermatitis. Just on the OX40 asset that you in-licensed or acquired through Kymab. Amgen has gone down the same route, OX40, but they're going after receptor. I think you're targeting the ligand. Is there anything to think that you're -- targeting one or the other might prove superior to the other?

Yes. I hope that once I stopped talking and making my argument that you're absolutely convinced that you want to go for OX40 ligand, not the receptor. So listen to me. So essentially, why did we go for OX40 ligand? It was when I came in 2016, we obviously -- we were digging deep into atopic dermatitis mechanisms and response, nonresponse and we found the OX40 ligand pathway is really one of the pathways we wanted to go after. So more to come in that space. But we also saw that Kymab had that great program. And in fact, they wanted to go from psoriasis at the beginning and obviously, then they switched to atopic dermatitis, I think, which was a good thing to do this based on the Phase IIa data with really Dupixent-like efficacy. And also, what I'm really excited about, durable responses. It's early days, but this could indicate disease modification, which you could expect from that mechanism. This might not only lead to extend its scheduling, every 2 or 3 months dosing, but also ultimately, you see some modification, which is a big goal. Now why did we go for the ligand and not the receptor? The receptor is widely expressed on a considerate level on a lot of T-cells. The ligand is only up-regulated on APCs like macrophages and dendritic cells during conditions of pathology. So if you block the ligand, you really specifically target your drug to that pathology moment. And the pathology moment is when OX40 ligand engages with an OX40 T-cell and drives it into a Th2 cell or Th17 cell, and T-, B-cell collaboration takes place. And that's where we specifically then come in and block that. There's a thing called target-mediated drug disposition. If you have a widely expressed antigen, that could be a risk and it might impact your dosing. So that's one of the things you could have if you don't go for the ligand. But then the other thing is if you go for the receptor, it's very difficult. You think you have an antagonist antibody, it's an antagonist. But in vivo, it gets cross-linked and all of a sudden it can cross-link a receptor and signal and become an agonist. So you want to avoid that. And then the final part is probably the most important part. OX40 is expressed on Tregs. And one thing you block the ligand, you actually increase the activity of Tregs. Now some of the other compounds, which are out there, and please check them, they're IgG1s, some of them are even [indiscernible] IgG1s, so they have the potential to deplete. So what actually could happen is if you deplete cells, you would also deplete Tregs. You don't want to have that. And for example, the fact that you see pyrexia is also something which could tell you that you might deplete cells. So the devil is in the detail about choosing a ligand and a receptor. Also in other receptor-ligand interactions, we typically go for the ligand and there's often a reason behind it. We have thought deeply about it. But we'll see. I mean, the data needs to kind of confirm it, but that's the reason.

Very clear. Listen, we're fast running out of time, but I do want to have quickly stop by oncology. And you've been very candid about Libtayo and Sarclisa and how you're thinking about the commercial opportunity and it's all about building on top. What about some of the early-stage assets? I mean SHP2, IL-2, all very interesting, we know that, but you're not the only one exploring this mechanism. So would you push back or agree that it's too early to really sort of make a big call? Doesn't look like the Sanofi assets have shown enough data yet for a big call to be made. And then any -- because it's late stage in Phase III on the CEACAM5, any updated thoughts there?

Yes. Maybe I'll start with the later stage and then go down. I mean obviously, Sarclisa and Libtayo launched. Amcenestrant, we talked about, the talk of ASCO and a part of the leading profile in the ER breast cancer space. CEACAM5 is very interesting. It's an ADC. We have very good CEACAM5 reagents, which basically goes for about 20% to 30% of non-small cell lung cancer. We have a Phase III going on where we're essentially in this post checkpoint space where we obviously want to make a mark, that's where there's a real high unmet need. But the real opportunity with CEACAM5 is to go first line because what it induces, it induces immunogenic cell death. It destroys the non-small cell lung cancer cell, and then you combine with a checkpoint, and you actually have this kind of very synergistic crosstalk between immunogenic cells death and a checkpoint inhibitor. So that's the ultimate goal of the CEACAM5 tusamitamab ravtansine approach. Then if you go a little bit earlier, I think the SAR'245 is the IL-2. We already talked about it, very exquisitely expanding NK cells and CD8 T-cells, just comparing to competitors in terms of what they have. IL-2 can induce vascular leak syndrome, can increase eosinophils. IL-5 check for that. That tells you that you basically are not as clean on the CD8 and NK cell side. We had 3 objective responses, including monotherapy response. So that's quite exciting. NK cell platform, we already talked about. And if we then go down the list, I think what's exciting is IL-2 and [ ILT2 ] blocker from Biond that gets us into really releasing the break on immunosuppressive macrophages and actually also NK cells and activating NK cells. So that's an interesting checkpoint inhibitor, moving away from T-cells to macrophages and NK cells. The SHP2, I think we saw the monotherapy data also from competitors. We have some objective responses. But SHP2 is essentially a kind of great [indiscernible]. The promise is really in this KRAS cancer space, where you combine, and this is what the trials we're currently running with our, obviously, partner, Revolution Medicines. And we'll see what the data say. And obviously, the combo with pembro is in -- and other or with other checkpoint inhibitors. So it's really a combination drug, and I don't think it's -- that necessarily the monotherapy is important. Important is that you really navigate the combination in a safe way. This is always what happens. So if you look at competitors, always look at their safety -- at the safety profiles. And then finally, Tidal platform, I already told you about. We have a trispecific, which we published in [ Science ] on where we have a CD3, CD28, CD38 [indiscernible] based trispecific. So there's quite a lot cooking in the pipeline. But we're sort of -- I wouldn't say we're a leader. I mean we're reemerging and the reemergence will be data driven. But we're setting us up for success, I would say.

Listen, I mean, we've barely scratched the surface upon [indiscernible]. But look, it's -- we are working within the constraints of 45 minutes, and we've already spilled over. But look, on behalf of the team, Frank and the IR team on line, thanks for your time and enjoy the weekend, and I look forward to the upcoming data.

Thank you, Peter, for having us, and it was a pleasure to have that great scientific discussion.

Thank you very much. Take care. Goodbye.

Bye.