Sanofi (SAN) Earnings Call Transcript & Summary

Good afternoon, everyone, and welcome to our next fireside chat at the Credit Suisse London Healthcare Conference. My name is Matthew Weston from the European pharma research team. I'm joined by my colleague, Jo Walton. And it's a great pleasure today to welcome Dietmar Berger, the Global Head of Development at Sanofi for the next fireside chat. Dietmar is also joined by the Investor Relations team at Sanofi. Dietmar, thank you for taking the time.

Thank you, Matt.

[Operator Instructions] But we do have some questions of our own and we'll start with those. Jo, over to you.

I'm going to start with a hard one here then, really a criticism that Sanofi hasn't historically had a great track record in R&D productivity. And arguably, some of your successes come from focusing on modes of action that were derisked by others: Libtayo, Sarclisa, for example, are clearly second derivatives on an existing theme. Can I ask how you think that the mindset in R&D has moved on? And can you give us any examples of where you really moved to becoming more innovative, perhaps taking greater development risks? So something behind what we hear Paul Hudson say repeatedly, if you can give us some nuts and bolts to support that.

Yes. Thank you, Jo. I'm with Sanofi since close to 3 years now. And of course, I'm biased but I'm firmly convinced that Sanofi is undergoing and is actually making good progress in the transformation journey. The portfolio, you mentioned a few drugs, which are obviously drugs that have been around for some time. The portfolio has changed dramatically over the last 3 years, where we had a complete refocusing of the company away from, I would say, diabetes and cardiovascular to areas of much larger unmet medical need, like immunology, oncology, neurology, rare disease and vaccines, obviously. We've seen really a change in the portfolio. You've seen both the advancement of internal medicines, but you've also seen, obviously, additional bolt-on acquisitions. But there is a flow of molecules that is coming from our own research labs. When you think about the more recent molecules that we brought into the clinic, we brought 10 new molecular entities into the clinic last year with -- based on several different technologies, whether it's Nanobodies, whether it's antibody-drug conjugates, whether it's T-cell engagers. Again, those are based on the deep understanding of the biology, but then also really of work that's going on in our own research labs. And besides this increase in research productivity, we also see that change of mindset that you spoke about and that I think is really important with regards to risk taking, with regards to how do we move our molecules forward, with regards to really a focus on speed and delivery. And just to give you a few external validation points here, there's an industry-wide comparison. It's called the KMR/CMR data that looks at delivery of the portfolio, delivery of just the studies that we do. We're in the top quartile now for delivery of Phase I, Phase II or Phase III studies. So there's been a lot of change with regards to how quickly can we take decisions, how quickly can we actually deliver on our programs. As an example, I could mention here tolebrutinib, for example, where we went from having the readouts of our Phase II data to going into Phase III to including the first patient into Phase III was just 4.5 months. And you have to see that within those 4.5 months, you have to open the sites, you have to manufacture the drug. You have to ship it. You have to also take a decision about the program. And we've very quickly taken decisions to move tolebrutinib in into 4 Phase III studies, which then have been started in a very, I think, accelerated fashion. Another example of risk taking, and I assume we're going to talk about that in more detail, is, Dupixent, where we come from a deep understanding of the biology of the type 2, focus of the type 2 inflammatory focus. And we're going from classic type 2 diseases like atopic dermatitis, like asthma, like chronic rhinosinusitis, now in waves of other diseases where we are actually reacting to data sets, yes, but we're also exploring the biology further. We're adding to the knowledge in the field, and we hope to add more benefit for patients and, obviously, also more value than for Sanofi with activity in those additional diseases that we're looking at. And you see some of those reading out, right? You see the eosinophilic esophagitis. You see the prurigo nodularis and others. And then all the way, 2 molecules that are complete Sanofi developments like, for example, amcenestrant, which is coming from our labs outside of Paris, which is, we think, a potential best-in-class and first-in-class estrogen receptor degrader, where we're looking forward to have readouts very soon of our first pivotal study. So I think it's, as I said, I acknowledge I'm biased, but both when it comes to the portfolio composition, but also to really the delivery and the risk taking, we've made huge progress.

We're certainly going to come on and discuss amcenestrant in more detail. I'm sure Matthew is getting laser pen-ready for that one. I'm just going to ask another broader R&D question about the use of digital technologies. So we're asking this question of all of our companies that we're speaking to in whatever function it is, and obviously, it's different on different functions. But it's clearly important with the use of Big Data, new tools, would you -- again, with your history and being at Sanofi, how do you see Sanofi in terms of an adopter there? Are you also-ran? Are you a leader? What more do you think that could do for you?

Yes. First of all, we firmly believe it's a very important space. And we have a lot of activity actually, especially more recently in the digital space. If I want to, for example, talk about development, right, the COVID pandemic has really led to a situation of much further acceptance and real acceleration of new tools, of new approaches, including digital approaches, remote clinical trials, decentralized trials, new endpoints really across the board, right? And we're obviously now also entering into more and more discussions about how can digital endpoints work with regulatory authorities. We're even speaking about digital therapeutics. And giving you a few examples, we've been able, during the COVID pandemic, to maintain activity on the vast majority of our clinical trials. We've been able to maintain more than 90% of the ongoing patients in our clinical studies. A lot of that was due to utilization of tools like remote monitoring, remote patient contact, shipment of drug to -- directly to patients, et cetera, and also close collaboration with regulatory authorities. Another example is that, for example, in neurology, specifically when it comes to movement, when it comes to gate, there's definitely a lot of activity. We have our own validation study at this point of a digital endpoint. It's called the PD endpoint, which we've discussed with regulators around the globe and especially also with FDA. And we're in a validation study of that digital endpoint, which then can be used in our Phase III trials. We've also made great strides when it comes to, for example, management of adverse events with digital tools, both the capture but then also the processing, which leads to real efficiencies when it comes to cost of adverse event management and registration and analysis. We've also made good progress when it comes to utilization of digital tools more broadly in clinical trials. How do you reach out to patients? How do you potentially also get to what we call an end-to-end digital data flow? Basically, getting the initial data from the electronic health record, then integrating those data directly into the databases and then all the way to the regulatory submission, utilizing the same data so you don't have to touch that data. You don't have to basically send the monitor out to the manual capture that we had before. And that end-to-end digital data flow, I think, is another really interesting concept. The one additional area that I want to highlight is really digital in research, where we've recently announced 2 collaborations, one with Owkin, one with Exscientia, focusing on how can we be much more efficient when it comes to identifying new targets, coming up with new molecules, integrating new tools like AlphaFold with regards to really the 3-dimensional protein structure, with regards to how can we target these structures. And we have really clear objectives with those interactions and with those digital approaches, how many molecules do we want to see to come out of them and how do we analyze data, for example, in different ways with Owkin to help our clinical development pathways?

I think with that, Matthew, do you want to kick off on some product questions?

Yes. Before we get to products, I'm just going to take one more general question, Dietmar, if I can, around proprietary technologies. And so you laid out that there had been such a significant transition in R&D and innovation at Sanofi. And part of that has been a number of acquisitions that have brought exciting platforms into the fold. So I'm thinking in particular of Ablynx but also the novel technology that came with Synthorx. We often get asked, particularly about Ablynx, as to when we're going to see meaningful evidence of pipeline candidates using their technology coming through the Sanofi pipeline. So I'd just be very interested. I guess it was acquired early '18. I presume you've now had a real opportunity to work through the technology, the portfolio that they have and bring clinical candidates to market. Are there -- anything you want to call out to us that are coming along the way?

Yes. The Ablynx technology, the nanobody technology is a really interesting technology, right? It gives you this opportunity of literally like beads on a string to combine different targeting moieties and also to modulate the half-life and the pharmacokinetics of the molecule. Obviously, you need to get to that point where you say, what are the targets that you want to look at? How do you want to combine them? And what are the target diseases, right? We have made good progress, in my mind, also on that side, where, as I said last year, we brought 10 new molecular entities into the clinic; 4 of those were based on the nanobody platform from Ablynx. And just highlighting a few of those. There's, for example, an anti-IL-13 and TSLP Nanobody that combines targeting, of course, the type 2, but then also the TSLP moieties, which then are obviously targeting the asthma space, right? Or you have an anti-TNF-alpha/IL23 Nanobody that again goes into these inflammatory diseases. So we have these 4 that have entered the clinic. We're working on additional ones. We're definitely utilizing that platform more broadly. And there's also good synergy with other parts of our research portfolio. For example, you mentioned Synthorx, which gives us this possibility to use synthetic biology to use really very targeted modifications of biologic molecules to eventually lead to new characteristics of those molecules. We've applied that to a non-alpha IL-2, obviously with Synthorx, but there are other cytokines and other targets that you can think about. And there's even some cross-fertilization where you can use some of the Synthorx technology with the Ablynx technology and come to entirely new molecules. The recent acquisitions, if you think about those that we've done, whether it's Synthorx, whether it's Principia, whether it's Ablynx, whether it's now the most recent one Amunix, for example, with conditionally activated biologics, were usually centered about a front-runner drug, but then also about a platform that we really feel adds to our ability to specifically target the biology that we want to target. And we will see in the coming years, I think, a constant stream of molecules from an antibody perspective, from a Synthorin synthetic biology perspective and, for example, also from a conditionally activated biologic perspective, and it just will some time for that to come through -- will take some time for that to come through.

Perfect. Well, we look forward to it. So enough of generalities, I'm going to jump in, and I've got to start with Dupi. And there's been a great deal of attention in the upcoming Immunology Day, particularly given Paul's comment that Sanofi will revisit the peak sales target for Dupi. Now clearly, I'm not expecting you to share any details in advance. But what has been a real investor discussion is how you're going to frame that? Because the original more than EUR 10 billion number was framed ex COPD within this kind of upside on top if COPD were to work. And I think we're having to wait till 2024 until we get the COPD data. So my question is simply, should we expect a kind of similar reference when new numbers are discussed of Dupi ex COPD? Or could we compare -- could we end up comparing apples and pears when we talk new numbers in a few weeks' time?

Yes. I will obviously not lift the suspense there, right? And you're going to have to dial in for the Immunology Day. And I think it will be a really good day because we are not only talking about Dupi and the change in expectations, there, we're also talking about the broader portfolio. And I firmly believe that we are on track to be the leading immunology company based on the portfolio that we've put together. For Dupi, what I want to say is, obviously, these pictures and frameworks around how you look at forecasting, how you look at expectations, both from a patient perspective but also from a market perspective, it's a constantly evolving picture, right? And we've seen remarkable resilience and very good take-up with Dupi during the pandemic even, right, when a lot of things came to a halt. We think this is really based on the unique mechanism, the central role of targeting IL-4 and IL-13. And we also see that some of our plans are clearly panning out. So I've been speaking about the wave of studies that we're delivering. We obviously have positive pivotal data now in 6 different diseases. We didn't know that a year ago. We're waiting for the -- we were waiting for the study readouts. Now we have positive data in atopic dermatitis, asthma, chronic rhinosinusitis where we're already approved. Then you've got the eosinophilic esophagitis. You've got chronic spontaneous urticaria. You got prurigo nodularis. So you really have more and more patients that are eligible for Dupi. We also have an expansion when it comes to age groups. We're going to younger and younger group when it comes to, for example, to atopic dermatitis and asthma, which, of course, satisfies a real unmet medical need because children, for example, with atopic dermatitis, having something for them, something effective for the first time makes a real difference. And we also have, I think, a really successful expansion when it comes to regional expansion, for example, in China. So all of that is really at the basis of the new projections. And I think we will make it very clear what the projections are based on. So I don't think you will have to compare apples to oranges.

Perfect, okay. Well, that in itself is certainly going to be an interesting discussion, and we all look forward to it. It's interesting that you raised prurigo nodularis, where, as you've said, you've announced headline positive data. The one interesting thing for that one is if we think of Dupi relative to other modes of action, particularly I'm thinking of AD, itch has never been your key selling point. You have good itch data, but other people, I guess, can lead with itch. Whereas PN, as I think of it, is a disease very heavily driven by itch. So how confident are you in that category? It's a very competitive market. Lots of PN trials are reporting in the coming months. How confident are you that Dupi is going to be the right mode of action for that disease?

Yes, I'm just looking at the clinical trial data, right, where 60% of patients are seeing, as you said, right, are seeing that reduction in itch. We see that very consistently in the 2 clinical studies that we have, where we see the 60% of patients meeting the primary endpoint there of itch reduction in Dupi and compared to 18% in the placebo group. And we're really excited about the prurigo nodularis data. Yes, there are different studies coming. But if you look at it right now, there's no other approved systemic treatment option in the market for prurigo. I think we have clearly a first-in-class and at this time, best-in-class biologic. With Dupi, it's based in, again, the understanding of the biology. We feel we hit 2 targets with IL-4 and IL-13 that play a key role there. And obviously, when it comes to Dupi, I think the important piece is that we're treating the underlying immunologic mechanism with really good efficacy, looking at these data, but also really good tolerability. And that's why I feel Dupi is very well positioned also in the prurigo nodularis market, right, but obviously also in other markets.

Can I ask if you think that the speed of onset of that itch reduction is competitive? I mean, it was a 60% reduction, but it was at 24 weeks. That's an awful long time to wait unless you're confident that actually you get 75% of that action within the first couple of weeks. And that's where I think others are looking for perhaps a faster onset of action against what's the -- a very annoying and debilitating condition.

Again, I think it's really the overall profile that Dupi brings to these patients. At this time, it's a really unique profile that it brings, right? It's the first biologic that works in this disease. The onset of action, we're acting somewhat earlier in the cascade, right, when you think about the different cytokine cascades and immunological cascades in prurigo nodularis and itch. But it's a really meaningful and sustainable efficacy that comes with very good tolerability. So as always, it's going to be a balance, of course. But I think Dupi has a really unique profile that will work well in patients with prurigo nodularis also.

You've also had some recent data with the failure of Dupi in CSU refractory to Xolair. But we understand that you're still looking to file that in the Xolair-naive population. How does that work in practice? Can you look at a PN filing and a CSU filing at around the same time? In Europe, as we understand it, you have to sort of go with one, wait for that to complete and then go with the next one. It's more difficult. And is that also true in the U.S. or would you sacrifice one in order to focus on the other?

So we can have -- the very specific regulatory question, we can have separate filings and we're not restricted on that side. But just a minute on the CSU data, on the chronic spontaneous urticaria data. We have these 2 studies. We have the first study, which was basically after antihistamine failure. And there, we saw very good activity. That's what you call the earlier-line setting. And then we have these results in the difficult-to-treat subset of patients who have failed all available therapies, omalizumab, high-dose antihistamines, potentially also immunosuppressants. And we're obviously disappointed in these results. We had hoped we could make a difference for that patient population as well. There may also be a difference in the underlying biology and in the mechanisms of resistance, of course, that we need to understand further. But we do think that we have the ability to really deliver potentially a best-in-disease option for those patients that are earlier line, as you said, and those discussions are ongoing, right? We're discussing with regulatory authorities. Obviously, that's one study at this point. We have a second study that is ongoing. So we cannot speculate on the time lines for the CSU filing at this point, but we have the firm intention to really bring the benefit that Dupi can bring to that earlier-line patients to the market, absolutely.

Can I then move on to some of the Dupi follow-ons and then Matthew will take us into AMEERA and the SERDs? But from a Dupi follow-on point of view, you've done a fair amount of in-licensing. You've got the OX40, you've got rilzabrutinib, et cetera. How do you -- how do these model -- these molecules add to your type 2 focus? I think people are looking as to, is this just a replacement one-for-one? Does this really increase the breadth? And I know this is a problem for many years down the line, but when we were looking at something like Humira with AbbVie, people are looking at the constellation of how you replace something so big. So it's just how these other assets fit together with Dupi, please?

Yes. Most of these assets are somewhat earlier, so obviously, we -- this will need to be informed by data. But just conceptually, right, the type 2 inflammatory space is a space that has been characterized over recent years. We had a key role in that characterization. And we now understand that there are a lot of different diseases that you can address. There are also pockets in specific areas that you can address based on the biology and further characterization of patients. So we feel that overall market will grow substantially, so there's definitely room for more than one molecule in that market. The other really important point is as we have multiple shots on goal here, we really went to understand what is the disease course. And these different molecules, when you look, for example, at a disease like atopic dermatitis, can play different roles in that disease course. Let's say, in atopic dermatitis, you're generally going from a topical therapy, you're going to a safe oral and then you're going to a biologic. And then you can see different lines of biologics and different segments. So it makes sense to develop these molecules in parallel to address, for example, different stages. You know rilzabrutinib, which is an oral BTK inhibitor, could fit into that space of a safe oral, right? We also have a topical BTK inhibitor that can play a role in atopic dermatitis. And then we have different biologics, right? And there in the different biologics, really the question is -- and we need to see data, obviously, the question will be, are there different segments of patients? Are there different types of patients that you can characterize? For example, is there more of an OX40-driven group of atopic dermatitis patients and more of an IL-4 and IL-13 driven because there are differences in the biology in some of those patients. So there are various ways how this can play out. The important thing is we have molecules that have the clear potential of best-in-class and first-in-class molecules. For example, amlitelimab as an inhibitor of the OX40-Ligand has, we feel, clear advantages over, for example, addressing just OX40, the receptor. It has demonstrated strong Phase II data. We're following up obviously with a Phase IIb study, and we're also informing further at the Immunology Day about the plan for the molecule. It has a unique mechanism of action, and it really has clear potential. And the same is true for other molecules like itepekimab, like rilzabrutinib. So really looking at an overarching strategy to get to that #1 immunology leadership and also to maintaining and sustaining that portfolio.

I think Matthew is going to jump in...

It's me then on a -- Jo's determined that we get to amcenestrant, Dietmar. I don't know why, but no. First of all, a quick one, and I think we know this already because it came up, but we can confirm that we'll definitely get the results in 1Q '22. Is that right?

Yes.

Perfect. And I think we learned last week that the trial is now completed and in data lock. So we're looking forward to it. Before we see those data, one question. We've obviously all had the opportunity to see the Menarini and Radius data on elacestrant at San Antonio last year from the EMERALD study. And it's just whether or not there's anything that you would like to flag that means that EMERALD is meaningfully different from AMEERA-3 so that we shouldn't make direct comparisons or try and look at the control arm, for example. Is there anything that you would flag that just makes those 2 trials apples and pears when we look at comparisons?

I think the key point here is that for elacestrant, there was also the focus of -- on ESR1-mutated patients. That clearly positions the molecule then into a late-line setting, right? And because the ESR1 mutations tend to accumulate after aromatase inhibitor therapy, right? So we have not done that. We have focused our study on the overall population. We do have good activity with amcenestrant in the ESR1-mutated population as well. It's just a different design of the trial. Also, we have a combination of second-line and third-line patients. So you will see a slight difference in the control group when it comes to how many patients do you have in the control group that are on another [ surrogate ] fulvestrant, right? And how many are still on an aromatase inhibitor? So there are slight differences in the study. But eventually, both are studies in the second -- in the late-line setting, in the late-line setting of metastatic hormone-receptor positive breast cancer. The positive that I would read through from elacestrant is we can really make a difference in this setting with a novel SERD, and amcenestrant there, I believe, is really well positioned. The efficacy data that we've seen so far indicate strong efficacy. The tolerability has been stellar, so I think we're, again, having that possibility of a best-in-class and also first-in-class profile with the program that we've put together.

Understood. Well, we all look forward to the data very much. I guess the other element is that in earlier lines, which is always where the larger commercial opportunity is, particularly in breast cancer, a real battle seems to be forming of people trying to design the trials that get them the readout the quickest in as early a line as possible, basically. So essentially, people are playing catch-up. And I would just be very interested from your perspective as to what you think sets Sanofi's approach apart from your competitors. You obviously set your trials up before everybody else, so you led the way, but other people are now trying to beat your time lines with different designs.

Yes. So the -- when we're talking about earlier lines, there are 2 areas we need to distinguish. One is the first-line metastatic setting. We have our AMEERA-5 study ongoing in that setting. We're expecting pivotal results there basically earlier than anybody else. I think that's a study that has pretty much a standard design combining with the CDK4/6 inhibitors and different companies are following very similar designs then in that regard. It becomes more interesting when you look at the adjuvant setting. And in the adjuvant setting, we've seen, I think, some movement more recently with the activity of different CDK4/6 inhibitors, also with data on different hormonal therapies and combinations. And we believe that the study that we have put together, which is the AMEERA-6 study, which is basically in start-up phase, right, that the AMEERA-6 study is focusing on a high-risk patient population, on a patient population that has seen an aromatase inhibitor and in a patient population that will most likely give you events earlier. So the AMEERA-6 study, we believe, is designed to really give us results as quickly as possible. Those adjuvant studies tend to be long and obviously readout rather late, tend to have low event rates. The population we've selected will have a higher event rate, will hopefully give us data earlier, will hopefully give us a path into the adjuvant setting earlier than what other people have. It is important then to remark that, that leaves a few other areas in the adjuvant setting that we still need to address, and we have plans to also address those other components of adjuvant therapy. For example, when you think about long-term therapy, nowadays, hormonal treatment is given all the way up to 10 years after surgery, so addressing the long-term treatment, addressing also combination treatments and also addressing the broader low-risk adjuvant setting. But those are studies that we haven't spoken about a lot so far, but we definitely think about addressing those spaces as well.

Perfect.

If I can move us on to nirsevimab and RSV, GSK has obviously just stopped their maternal study. Does that help you with a clear run towards infants for your product? And can you give us -- remind us of the timing of submission? And my final question around this is how it's going to be used. From a vaccination point of view, you normally need an ACIP recommendation. How does that work for a monoclonal antibody?

Yes, so 3 questions there. I'll try to answer them quickly, right? We always have seen nirsevimab as a differentiated molecule and a differentiated approach to prevent RSV disease in children. And the study program that we have, the Phase IIb study, but also the Phase III, the MELODY and the MEDLEY results have really demonstrated the potential of nirsevimab to protect all infants, right, through their first RSV season with one single immunization. And I think that was always a real -- really differentiated approach and really differentiated offering. And this includes preterm infants and healthy late preterm and term infants as well as infants also with chronic lung disease and congenital heart disease and all those subgroups. So it's really a protection for all infants. It also does allow for a single treatment and really protects you through the RSV season. When you think about a vaccine candidate, which is a maternal immunization, it would require some time to allow for that maternal immunization to kick in and really generate robust antibodies. It also depends a lot on timing when do you give it because that maternal immunization wanes and it can only last about 90 days. So for some children, it may actually not protect them through the RSV season. And in addition, if you have preterm infants, you're also not protected because the maternal antibody production is not there yet, right? So that's why we always felt we have a differentiated offering with nirsevimab. And we've also received, as you know, priority designations from different health authorities. The submission has been accepted by the EMA under an accelerated assessment, and we do expect submission in the U.S. in the second half of this year, which means we're actually 1 year ahead of the plan and we anticipate to being able to introduce nirsevimab into the market in the beginning of 2023. With regards to the ACIP recommendation that's for vaccines and really for the standard immunization schedule, we anticipate that nirsevimab will be added to the routine immunization schedule for all infants. And the ACIP has established a working group for RSV, which then includes monoclonal antibodies, and we're actively engaged with the key stakeholders here to provide the clinical information on the nirsevimab data. And we think that there's a clear path forward.

And do you think they understand the differences then? I mean, they are happy that this is just a vaccine under another guise effectively? So they're happy to -- because funding and all the rest of it comes from the ACIP recommendation, doesn't it? So it's very important that you are captured in exactly the same way as a regular vaccine.

No, I do believe the data we have is very clear, and I also think they understand the underlying potential and also the biology, where this is a different approach, but it's a different approach for a reason because you want that protection for all infants for an extended period of time and we can provide that. We also think, and that's important for ACIP as well that this really helps to address disparities in health care and also ensures really the equity of preventive care. And that is a top priority for U.S. health care decision makers and also for ACIP. So I think we're actually in a positive situation there.

And would you expect a really broad indication using all infants? Or would you expect at least to start with that it would come in infants that would be perceived as higher risk, those with other underlying illnesses or disabilities?

No. We feel the profile of nirsevimab, the data that we have supports an all-infant approach and also an all-infant recommendation.

Okay. Matthew, do you want to take over on hemophilia?

Yes, I'm not going to go to hemophilia, Jo. I'm going to jump to oncology because I think there's a couple of other elements. Obviously, Libtayo, you recently withdrew the cervical cancer indication because you couldn't align with regulators on post-marketing requirements. I'd be very interested in understanding Dietmar, as to what you see any future trial readouts that could meaningfully grow the Libtayo franchise beyond where it is today and make it potentially more significant to Sanofi or whether or not actually the growth opportunities in the portfolio are just elsewhere other than Libtayo?

No, the Libtayo, maybe a quick second on the cervical cancer piece. The cervical cancer piece was obviously a clear decision based on the interaction with the FDA. And based on the fact that this was a second-line treatment setting, a rather small part of the market. And it was just not -- we were just not able to align on post-market commitments, which we felt doing another study doesn't make any sense. It's obviously bad news for patients because you have one option less, but I think there are also alternatives available, right? When you think broad about Libtayo, we always had a more focused strategy. We were always saying we want to have leadership in skin, and we want to provide a really meaningful option in lung cancer. And we've definitely demonstrated in non-melanoma skin cancer that we can maintain and consolidate our leadership position. There's also an adjuvant study that is ongoing. So we feel that that's an important area and we're there the first and only immunotherapy licensed in both advanced squamous cell carcinoma and basal cell carcinoma. Lung, again, we see actually good uptake there given the data that we have. This has confirmed our competitive strategy choices. And the key study to point out is really the first-line chemotherapy combination and the launching of that in the second half of the year, following the submissions that we had with FDA and also EMA in the fourth quarter of last year. There are other studies ongoing with Libtayo, but the strategy really is around these 2 key areas and then exploring also combinations with the ongoing portfolio that we have. And as you've seen our oncology portfolio, really switch more and more to immuno-oncology and with novel approaches in immuno-oncology, Libtayo can be an important combination partner there.

Okay. No, it makes sense. And if I can just pivot to CEACAM5 because I guess that's -- you mentioned novel combination partners or novel molecules coming through the portfolio. That's clearly one of the next ones in oncology other than amcenestrant. I think you've guided to a possible regulatory submission in 2023. So can you remind us when we're going to see some potentially pivotal data for CEACAM5?

Yes, this is the -- for CEACAM5, we are basically reading out 1 Phase III study and 4 Phase II studies in the near future, so we will have a meaningful data flow. The Phase III study, we call that CARMEN-LC03, that's an ongoing pivotal trial for tusa in non-small cell. We call it tusamitamab ravtansine, our CEACAM5 antibody-drug conjugate. In non-small cell lung cancer in the second-line setting, it's in CEACAM5, high-expressing nonsquamous non-small cell lung cancer patients that are pretreated with, obviously, immuno-oncology with checkpoint inhibitors and platinum-based chemotherapy. There's an interim readout that we could potentially also have in the second half of this year. But the pivotal data, the final data of the study is expected in 2023. And obviously, then we've also guided towards the regulatory submission. We feel tusa can play an important role in non-small cell lung cancer in those patients that express CEACAM5. You see a bit of a renaissance in ADCs, I feel, over recent years. Obviously, I was involved in ADCs earlier. We've seen early ADCs also in breast cancer in the HER2-positive setting. You see a continuation of ADCs in HER2-positive breast cancer. You also see ADCs now, for example, in lymphoma, and in other areas. So I feel there's a real potential to the technology. CEACAM5 is a unique target, and we feel we have a meaningful molecule there that can really make a difference for those patients that express CEACAM5.

And can you tell us, Dietmar, remind us, in that second-line lung patient population, is there an approximate percentage of patients that are CEACAM5 expressers or high expressers?

Yes, there's definitely around 40% of patients that express CEACAM5, right? So it's a meaningful part of that market. And we're also exploring other disease types where we have CEACAM5 expression, and we're also potentially thinking about different toxins that you would have to use for different types of diseases. There are other disease types where you see even higher expression of CEACAM5, for example, in gastrointestinal tumors, but we think you need another toxin in order to get there. So it's an evolving area, but we think there's real potential.

And clearly, we're about to get pivotal data. So we're not about to, but relatively shortly, going to get some pivotal data. Why do you think others haven't chosen CEACAM5 as a target? It's not often in a market as competitive as lung cancer that you don't have fast followers coming with you. We obviously saw it with Novartis with IL-1 and canakinumab, that so far hasn't worked out for them. TIGIT obviously got a huge amount of interest and has now become a giant battle as to who gets there first. Why is CEACAM5 -- is Sanofi-only?

Well, because it's not such an easy target. There have been different approaches, right? There's the whole CEA family. There's different CEACAMs. Some molecules have basically demonstrated adverse events, which are not tolerable. Others have not shown the right level of efficacy. Targeting CEACAM5 with a really well-behaved antibody and then having the right link at the right toxin technology is really crucial to get to the right clinical outcomes in this patient population. So overall, we think that there's a variety of factors that explains the competitive setting. Actually, this also gives us clear first-in-class and best-in-class potential with our molecule.

Understood. And just a quick one. I just had an e-mail question in from a client. I'll read it out to you. Can you push, Dietmar, of the 40% that express CEACAM5, what proportion are high expressers? And do you think you need high expression for efficacy?

Well, eventually, that's what the data is going to show us, right? We have a study that is broad enough that we can look at these different populations and eventually, the data will answer that question. There is more potential, obviously, beyond the high expressing population. And as you know, we are measuring these usually by immunohistochemistry and then you get these classifications of 3 plus 2 plus 1 plus. But we feel there's a large addressable patient population here.

Perfect. Jo, over to you to wrap.

Well, I'm just going to ask you one final question before I thank you very much. And that is which is your favorite child that we haven't actually been discussing today? We always look -- you must have something in that early-stage pipeline that you feel we should be doing a little bit of work on.

Well, it's hard to talk about favorite children, right? And that's already implied in the question. But the favorite ones are always the ones that readout next. And I want to highlight SAR'245, which is our non-alpha IL-2 simply because we've taken a decision to explore it broadly. We feel it's an important mechanism. Also, we feel we have a differentiated molecule based on the synthetic biology. It's obviously been coming from Synthorx. And we will have early readouts throughout the year. So we've seen initial, very encouraging biomarker data. We've seen increases in T effector cells. We've seen increases in NK cells. We've also seen initial clinical responses, right, in a smaller patient population overall. So we've seen monotherapy and combination responses, which will be important. And we're just looking forward to see what can we add with SAR'245, what's the data that's panning out from the multi-cohort study. We're doing different types of studies in thoracic tumors, in cutaneous tumors, in head and neck, in GI cancer and in lymphoma. So we're doing this broad evaluation. And it will just be exciting to see can we bring a new IL-2 principal with a differentiated profile to immuno-oncology?

Dietmar, sorry to jump in. If everything goes right with THOR-707, like it's the dream readout, what's the soonest you could file it?

There's different plans. It depends on the cohort where you have those dream readouts. But in oncology, obviously, you can work with expansion cohorts, you can work with accelerated approvals, et cetera. If we see dream readouts this year, that will take us to potentially reading out positive larger data sets, even in 2023 and then that would lead to an early filing in 2024. But that requires dream readouts, so take it with a grain of salt.

Well, with that, we look forward to hearing you within the month on the AMEERA study, given that you promised just the first quarter data. So I'd like to thank Eva, Natalie and Felix from the IR team and also Dietmar for being with us. Thank you very much, indeed, for your time.

Thank you.

Thank you, Dietmar.