Welcome, everyone, to our third annual Innovation Series event. My name is Ryan Richardson, Chief Strategy Officer at BioNTech, and I'm delighted to welcome you here back to New York, NASDAQ headquarters, which is particularly exciting for us, given that it's now our five-year anniversary from our NASDAQ IPO and public debut as a public company. So I'm going to kick off with a few introductory remarks, but we're going to spend most of the time today to focus on oncology and do a deeper dive on our late-stage programs. Now, the usual disclaimer here: we will be making forward-looking statements today, so please do take note of the disclaimer that's been filed with the SEC. I'm really delighted here to bring our management board and senior leadership team here to do that deep dive today.
I'm joined here by professors and co-founders Uğur Şahin and Özlem Türeci, CEO and Chief Medical Officer, respectively. I'm also very pleased here to be joined by our newest management board member, Chief Commercial Officer Annemarie Hanekamp. And in addition, we've got two vice presidents to join us today from our clinical development leadership team, Ilhan Çelik and Michael Wenger. So, as I said, today is going to focus more on oncology. We're going to have a pause in the middle of the session, and we're going to have time at the end for Q&A. And I think what we're going to focus on today is the next stage of the company's development. I'm going to provide just a few introductory remarks before we do the deep dive into our oncology portfolio. And I want to start by looking back since the company's founding in 2008.
What we see are sort of four distinct stages that led us to where we are today. It started with the founding of the company in 2008, the detailed work to develop our platforms and programs over a period of many years, iterative platform development that started once the company was founded and continued without a website, sort of iterative platform development, and already in those early stages, the company's vision was based on an idea of building a multi-platform company that could develop novel technologies to address cancer at its root. Hence the name BioNTech, New Biopharmaceutical Technologies, was really at the root of the founding vision of the company.
At the end of that first stage of development, the company went out and sought external capital in 2017, did a Series A financing, and also did a series of partnerships from 2015 to 2017 to validate certain platforms and take the company to the next level. The second stage of the company's development was our IPO and all of the preparation that comes with that in 2019. And of course, the IPO was instrumental in, first of all, giving us access to the capital market, giving us access to capital that could fuel our oncology portfolio and pipeline ambitions. But then, as we would later find out, four months later, it gave us the necessary access to the market to be able to fuel our COVID-19 vaccine program, which we initiated in January of 2020.
That leads me to the third stage of the company's development, which was the development of the COVID-19 vaccine, which we could say really was the primary focus of the company from 2020 to about 2022. Of course, we all know now that the COVID-19 vaccine program transformed BioNTech in a number of ways and highlighted on a global stage the impact that our technologies could make. I think since 2023, we've been accelerating into the next stage, the fourth stage of the company's development. Here, this is a stage that takes us back to our roots as a company, where we're focusing again, once again, on developing novel medicines to address cancer. As I said before, we're going to focus today on these later-stage oncology programs.
I think what's fundamental to understand here, and the point I want to make, is that this next stage that we're entering, we think, really builds on the previous stages of the company's development, takes us back to square one. Those many years, 15-plus years of developing our platforms and our programs, we think is going to fuel our ability to be successful in the next stage of the company's development. Where are we today? As I mentioned, BioNTech is a mission-driven company focused on developing medicines to address global health needs. We have one product on the market, a COVID-19 vaccine that's been distributed across over 80 countries, almost 5 billion doses, where we maintain a leading market position. We have a strong cash balance, almost EUR 17.8 billion as of Q3 this year.
We have a multi-platform oncology portfolio that today has 20 programs in the clinic and 13 ongoing Phase II and III trials, and a number of strong partnerships that we leverage along the way. We've also developed our infectious disease vaccine pipeline behind COVID-19. We have seven clinical stage programs, including two additional programs beyond COVID-19 that we've partnered with Pfizer. We've also been developing industry-leading capabilities at the intersection of biotechnology and AI through our wholly-owned subsidiary and BioNTech group company, InstaDeep. We've expanded our global footprint and expanded our capabilities across the value chain. This includes building in-house capabilities in individualized mRNA production, bulk mRNA production, modular RNA production, as well as extending into cell and gene therapy manufacturing. With the more recent acquisition that we announced yesterday, we, upon closing, will expand into biologic manufacturing as well.
So I think before we go into the oncology portion of the day, I do want to take a step back and talk, just share a few thoughts about our COVID-19 vaccine experience and the franchise that we have today, because we think it does provide us with a very unique anchor point and foundation to drive the business forward. Of course, we all know that the COVID-19 vaccine experience, still to this day, was the fastest development of a drug in the history of the industry. And we also know that the launch of this product, our first product, was the most successful pharmaceutical launch in the industry's history. But I think there's two other points that I want to make on why this is so relevant for the next stage of the company's development.
The first, of course, is that this experience highlighted the ability of our technologies to have a massive impact on humankind. And the second, of course, is that our ability to focus and execute in COVID-19 reflects the fundamental capability that we think is going to be essential as well in the next stage of the company's development. So on COVID-19, we believe that there's going to continue to be a need for seasonally adapted vaccines in the fight against COVID. And we believe this due to the continued evolution of the virus, but also the substantial evidence that supports the long-term health benefits and protective benefits of repeated vaccination. The fact is that the risk remains high, in particular for individuals of older ages and also the immunocompromised.
We do think that variant-adapted vaccines are going to continue to be an important line of protection for health systems around the world and for individuals to continue to protect against COVID-19. When we look at the epidemiology of COVID-19 over the past two years, we see a couple of interesting trends. The first, and this is based on data from the CDC, from a collection of surveillance sites across the United States. The first point is that we see markedly higher rates of COVID-19-related hospitalizations. We have seen this for much of the last two years versus seasonal influenza. I would note that during this period of time, we've seen multiple seasons where we've seen intra-seasonal dual or more than one COVID vaccine introduced into the market that was required. In fact, two of the last three years we've seen this.
Finally, if we look at the timing of COVID-related hospitalizations, what we see is actually that they haven't followed perfectly the example set by flu. You can see from this graph that what we've seen over the past two years is that COVID-19 infection rates and hospitalizations related to those have actually started to rise dramatically before flu. We saw that markedly was the case this year, with COVID infection rates rising in May and remaining at an elevated rate through the summer. We think that this highlights the potential for further benefit if this continues from an earlier release of COVID vaccines to help protect against those infections. Indeed, we've already seen a shift to earlier approval of COVID-19 vaccines.
You can see here, just anecdotally, a couple of the major markets: United States, Europe, U.K., Japan, where we have seen a shift to an earlier approval from 2023 to 2024. I would note that the shift has been somewhat muted, 20-60 days. But nevertheless, we see that consistent across the major markets. And we do believe that there's, again, if the epidemiology continues, that this would suggest that there's room for further alignment between the epidemiology of the disease and the availability of vaccines. I think our view of the COVID vaccine business is that this is going to continue to be an asset for us in helping us in the next stage of the company's development. And that's mainly related to three key business features, the first of which is our continued global market leadership.
We continue to have over 50% market share, and it continues to be a very global business. We benefit also from a very lean fixed cost base, and that then brings with it the potential for continued cash flow generation, and if you look at the first three quarters of 2024, you see that on display here. Now, this data is not full-year data. Again, it's just three quarters. It doesn't capture the full season of COVID-19. But what you see is that even in the first three quarters, our revenues have substantially exceeded our underlying cost of goods, R&D, and sales and marketing expense related to the COVID-19 franchise, and Annemarie is going to touch on that in her remarks as well.
But we think that this sets us up to continue to derive benefit from this underlying business as we continue to try to address global public needs in the fight against COVID. And you can see here how we're leveraging that as part of our core strategy. So continued cash generation from our COVID-19 franchise, coupled with a very strong balance sheet. And we're going to make strategic investments into R&D. We're going to talk about that today throughout the course of the day. And we think that those investments can deliver a progressive path and multiple levers for value creation over the coming years, starting in 2025 with rich data sets. We expect very rich data flow from our late-stage pipeline in 2025, and following that up with our first launches in oncology starting from 2026, on our way towards our goal of being a multi-product company in 2030.
So today's focus is going to be on our late-stage oncology portfolio. And you see here five programs that we think have transformational potential and also pan-tumor potential. The first is our BNT327 VEGF–PD-L1 bispecific antibody, which we just gained full control of with the acquisition that we announced yesterday. The second is our autogene cevumeran, our iNeST, our personalized RNA cancer vaccine, which is in multiple Phase II randomized trials in adjuvant stage tumors. FixVac, our off-the-shelf RNA vaccine, which is also in multiple Phase II trials. BNT211, our Claudin-6 CAR-T cell program. And BNT323, our HER2-ADC, which is positioned to be potentially the first oncology product that we bring to market. And Annemarie is going to also reference that. So this is going to be our focus for the rest of the day today.
With that, I'd like to hand it over to Uğur Şahin.
Yeah, thank you, Ryan, for the intro. Thanks, everyone, for joining us today. I'm happy to take the part of the presentation which is more conceptual to give you a background of what we are doing, why we are doing it, and how this is connected to our product development. So this is more or less the slide that describes our vision of personalized healthcare in the upcoming future. This vision is based on the understanding that every patient has a different type of tumor and that we live in a time where we see progress in many, many fields. Oncology is really moving ahead with new classes of compounds. We are seeing improvements of PFS. We are seeing improvement of OS.
But still, the cure rates, and with cure, we really mean that the patient is not dying because of its cancer. The cure rates remain suboptimal. They can be optimized. So increasing the OS for 10 months is fantastic, but it does not mean that the patient is cured. And this is something that we want to accomplish. And we want to accomplish that by a deeply scientifically rooted approach, understanding that cancer is not only an individual disease, that every patient has a different type of cancer. But even within a patient, cancer is a polyclonal disease. So that means often we develop treatments where the patient is responding first. But due to the polyclonality of the disease, after six months, nine months, cells which are resistant go up, and the patient is not anymore responsive. That's the root cause of cancer treatment failure.
The way how we want to address that in future is really understanding the polyclonality and individuality of the disease by analyzing at the top the nature of the disease. This can be done today now within 24 hours by sequencing the patient's tumor, identifying the mutations, identifying the RNA sequences. Our second approach is that we have defined therapeutic modalities. Modality is defined here, not anymore antibodies or RNA treatments, but the way how we influence cancer. The key system that can fight cancer in our body is the immune system. The reason for that is the immune system is as complex as cancer.
What we are doing with this type of treatments is we are not only directly killing tumor cells, but we are engaging the immune system to fight cancer. This is immunotherapy. And this immunotherapy was for many, many years some sort of a speculative type of treatment. Now, since 2010, 2013, everyone knows that immunotherapy works. And this is one aspect, strengthened by the immune system by immune modulators. The second is targeted therapies, so directly hitting the cancer and killing tumor cells. And this is also something which is broadly applied in the industry. What is somehow specific to us are the mRNA cancer vaccines. And the idea of the mRNA cancer vaccine is really a tailored, fully specific treatment to remove the remaining tumor cells. So these are the three pillars which we want to use.
This results in some type of a product portfolio that has some common features. The common features are off-the-shelf drugs. Every patient, even though every patient has a different type of tumor, there are principles, biological principles that we can address. The biological principle, for example, of checkpoint blockade, so this is not individualized. This can be personalized based on a biomarker, but this is off-the-shelf, but then we have also personalized aspects. That means really addressing the specific mutational spectrum of the cancer, so by the combination of both, we should be able to have a personalized treatment. To do that, we not only need to have these products, but we need to have capabilities. This capability is on the top, deep genomics and immunology understanding to analyze the patient's data, individualized treatment platforms, and this is what Ryan said.
We invested now more than many, many years to develop these platforms, allowing us not only to understand what is going on in the tumor, but being able with our multi-drug platforms to make treatments that address these pathways. The third thing is AI. AI is coming, and AI will enable us to do two things: to analyze complex data to make sense out of that. And the second is AI will help us to do things faster and take the computational power to deal with the individual patient. Last but not least, it's about manufacturing. Because when we want to get new drugs to the patients, and if we want to do that in a personalized fashion, we have to do it in an automated and digitally supported fashion. So this is the vision. But let's really translate this vision into a concept, how we can address that.
The concept you recognize now, again, the three classes of modalities of treatment is immunomodulators on the top, targeted therapies on the left, and mRNA cancer vaccines at the right or at the left. It's on the left, depending on how you look. And with regard to the immunomodulators, we, of course, know checkpoint blockade, anti-PD-1 treatment is almost in all patients. And what is next? We believe next is really combining these modalities. And for immune modulators, we are seeking particularly for bispecific immunomodulators. I will tell you later why. On the targeted therapies, we have now recently coming in the and with we, I mean really the whole industry coming up with cell therapies with the idea to target solid tumors. But another modality which is now coming back after discovery more than 30 years ago are ADCs, antibody-drug conjugates.
Changes in the linker and toxin technology now allow really in patients with advanced solid tumors to reduce the tumor size. The mechanism here, the most important mechanism, is reducing the tumor burden, so it's a different type of ADC. And the third one, this is special for us, is we need to get rid of the remaining tumor cells. Often, our patients are almost cured. Almost cured means that there are hundreds of millions of tumor cells left which are radiologically not visible, but they come back. And then there is a relapse. And our idea is that personalized cancer vaccines, mRNA cancer vaccines could work there, so we are building on that. And of course, we are interested not only to develop these modalities per se, but we see synergy. We see in the synergy of combination of the individual modalities and particularly also in triple combinations.
So what we did is in the last five, six years, we created a number of candidates. And some of the candidates I showed you last year. And you can see that we are clustering these candidates into these three categories: immune modulators, targeted therapies, mRNA cancer vaccines. We did not have in-house ADC experience. Therefore, last year, we in-licensed a number of compounds, allowing us really to close this gap. But we are building now also in-house capabilities with a number of bispecifics. And actually, we were seeking for the next generation immunomodulator. Some of you who are following us now for years know the presentation. We were always talking. We were searching the next generation immunomodulators. We have found some. And then we have our mRNA cancer vaccines. And these mRNA cancer vaccines, we have two types of approaching polyclonal immune responses. This is our FixVac.
The idea of the FixVac is combining multiple antigens in a tumor-type specific manner. Then we have the individualized vaccines where we really individually tailor the vaccine according to the mutational profile of the patient. We will talk later about this. Giving you an example, giving you some examples. Daina, you had a question last year. With all these combinations, you asked how you are going to combine. Because if you have 10 candidates on each side, you would end up with 1,000 combinations that you can do. Even if you are extremely productive, you will not do that everything in a clinical way. We have some thinking about this. We are, of course, prioritizing. We are focusing. Our prioritization and focusing is based on considerations. For example, for the development of how broadly is a compound applicable?
Applicable in the sense of how broadly are we going to expect meaningful clinical activity for the compound? So we have identified one of our bispecific IO molecules. It's BNT327, PM8002. It's a bispecific which turned out to fulfill the criteria that we wanted to see for a super IO, for a next generation IO. And this was not visible by preclinical data. And the preclinical data said everything looks great. You have to really do the exploratory clinical trials to understand what is truly working in humans and whether the molecule is working, not only whether it's working as predicted, but whether it's able to make a difference. So the BNT327, PM8002, we believe has applicability for a broad range of cancers. And we will show you why we believe that. It's not proven. You will only prove it if it's after five years.
But we believe it is broadly applicable. We believe it is going to increase the response rate. We believe that it's going to increase PFS. And we believe that it's going to improve overall survival. The second technology we are extremely excited about, which defined more or less our vision from the very beginning, is our personalized or individualized vaccine technology. And we believe that this technology is particularly suited in the current situation for patients with adjuvant cancers. So that means there are only a limited number of tumor cells left in the body and where we can train the immune system of the patient to recognize these tumor cells and eliminate them. And the endpoints for such treatments would be disease-free survival. So that means patient comes from the surgery and is almost cured, but then receives the vaccine and then hopefully has a lower likelihood to relapse.
So that we can really translate vaccination to improve not only overall survival, but also to cure. As I will discuss later on, what does it mean as a requirement for this technology to enable that we achieve this objective? Because it's not trivial. And the third example that we are presenting here is also something that is coming from our innovation toolbox. It comes with its Claudin-6 CAR-T therapy, a target that we have identified, discovered more than 10 years ago where we have seen it's not expressed in normal tissues, but in multiple types of cancers. And where we developed a Claudin-6 CAR-T therapy, developed a vaccine to improve the activity of the Claudin-6 CAR-T therapy. And we know in which type of tumors this target is expressed. And we believe it is applicable to several tumor types.
Of course, we have to start somewhere with an indication to prove that this type of treatment works. These are three examples showing the type of thinking that we have to prioritize and focus on assets. Let's discuss a little bit about BNT327, PM8002, which was developed by our partner, discovered by our partner, Biotheus. This molecule is targeting two validated pathways or molecules in oncology. It's on the one side, VEGF. It's a molecule which is for many known as a molecule that increases the generation of blood vessels, vascularization. What is not broadly known is that this molecule is also an immune modulatory molecule creating, modulating innate and adaptive immunity. That's one target. The second target is the PD-L1 pathway. I don't need to explain that. Both targets have been validated.
There are antibodies around targeting VEGF and antibodies around targeting checkpoint blockade. So why is a bispecific antibody different? And I would give you really an example where you can logically it's not explaining how it works, but it shows you that a bispecific antibody is not just the sum of two antibodies. So this antibody is a bivalent bispecific antibody. Bivalent means that both targets are addressed by two binding arms. And it's just like you can compare it with arms and you can combine it with legs. And of course, you can say, I can do a lot with legs. And I can do a lot with arms. But only if you combine legs and arms in a meaningful manner, you can do things that are not able to be done with legs and arms alone. So we say this is a cooperative mode of action.
If you divide this, if you come up with the concept, so what is now specific for a bispecific antibody? Two things that are very clearly conceptually is that this antibody binds to VEGF. And VEGF is not only a soluble molecule. VEGF is secreted by tumor cells. And it is binding to the extracellular matrix. So it's more or less a tumor marker. And you can see that in the staining on the right top side. So one thing that this antibody can do is to bind to VEGF in the tumor microenvironment and then bring PD-L1 blocking activity to the tumor site in a tissue-selective manner. So that's one of the specific aspects that you can do. The second thing that the antibody can do, it binds to PD-L1. And we know that PD-L1 is expressed in many tumors, PD-L1-positive tumors. PD-L1 is also a tumor antigen.
So what this antibody can do is to bring anti-VEGF activity selectively into tumors. And this is specific for VEGF–PD-L1 bispecific antibody because PD-1 itself is not a tumor antigen. PD-L1 is a tumor antigen. Whether this translates to a biologically and medically significant advantage, we don't know yet. But it is different. And we have to think about and investigate how this difference translates into a differentiated clinical activity profile. So this is just showing you that targeting VEGF has really multiple biological effects. So VEGF treatment is not only a treatment to reduce vascularization of tumors, but it's an immunotherapy. And one of the limiting factors of using only VEGF was the side effect profile associated with that. And we see a more favorable side effect profile for the bispecific antibody. So we do not know, or we did not validate so far the reason for that.
What we assume is that the targeting of the VEGF function to certain tissues reduces the risk associated with a pan-body VEGF activity. The molecule has been evaluated now in the meantime in more than 700 patients. We will hear some data updates or we will refer to some data that we have shown previously. We have seen clinical activity across multiple indications. We have seen a consistent favorable safety profile across multiple indications. We have seen the safety profile as a standalone compound alone and with different types of chemotherapies. We will see some data updates this year and in various indications and in various combinations. Let's talk about why we are so excited about this molecule. One of the key reasons I'll put it in this way. We see a number of interesting clinical features.
The first clinical feature that we are seeing is that particularly in combination with chemotherapy, we have an increased objective response rate as compared to treatments with anti-PD-1 alone. And this is what we call an instant effect. So the objective response rate, the objective responses are happening within two months. And we are seeing objective response rates which are 10%, 20%, 30%, 40% higher than historical objective response rates. We are seeing a longer duration of clinical benefit, which translates into a clinically accepted endpoint, PFS. We do not yet have OS data. But the clinically meaningful PFS data that we are seeing are very close to OS with comparative benchmark treatments. We are still waiting to get more mature OS data, but we would be very, very surprised if this meaningful PFS does not translate to a better OS.
This is one thing, really a higher clinical activity. The second thing that we are observing is that we are seeing that patients, regardless of their PD-L1 status, are responding. You can see in different color codes the response rates and the depth of the response in patients with low immune infiltration, medium immune infiltration, and high immune infiltration. What is exciting is that we are seeing even in patients who are PD-L1 negative or do not have immune infiltration response rate, which are very similar to those with PD-L1 positive tumors. To put that into a more simplified pattern is we believe that this bispecific molecule, that the bispecific activity on the one side translates into a better clinical activity in PD-L1 positive tumors.
But we believe that this molecule has the potential to bring immune checkpoint blockade to tumors in which PD-1 blockade is not yet established. So these are two types of excitement that we have, which would require different types of clinical testing. And we have to see whether this is really the case. So what does it mean clinically? You see in the light green on the left side and in the middle tumor types in which PD-1 treatment is approved. On the right side in dark green, you see tumor types in which anti-PD-1 treatment is not approved. You see with the dark small circles indications in which anti-VEGF treatment is approved.
And you see that if you would consider either/or or if you would consider an or relationship, a mathematical or both, one or the other, then you would see that this is covering all types of tumors or could cover all types of tumors. Let's look more specifically into that. On the left side, you see all the types of tumors in which checkpoint blockade has been approved. These are around 1.5 million patient cases. It's, I believe, only for U.S. and Europe, right? Yeah, thank you. And on the right side, these are tumors for which checkpoint blockade is not approved. So our working hypothesis is that this type of treatment could provide benefit also to this type of tumors. And to do this, we can't use always the same trick. So that means we have really to understand what are specific aspects of the tumor biology.
If we use BNT327, PM8002, some sort of an umbrella treatment, different types of tumors would require different types of combinations. To give you an example, if we want to do an ADC combination therapy, some ADCs may work in lung cancer, some in breast cancer, some in colorectal cancer. That's the reason why we have in-licensed different molecules to have tissue-agnostic tumor-type specific differentiation. Some tumors come with different types of immunosuppressive mechanisms. Therefore, we believe that our broad portfolio of bispecific immune modulators could become really an asset for later on combining with BNT327, PM8002. That means by selecting and focusing to certain molecules does not mean that we are ignoring and dropping the other molecules. We see them really as a high potential for combination. We are talking so far about potential synergies.
Of course, we have to prove the synergy, whether the synergy is existing. We have started to do that. I will show you only one slide for synergy, for preclinically validated synergy between immune modulator and targeted therapy in a preclinical system. You know how the studies work. You investigate compound A alone. You investigate compound B alone. Then you investigate the combination of both compounds. In this case, it's the combination of BNT327 with our TROP2 ADC, which is potentially applicable to a wide range of tumors, particularly epithelial tumors. We generated data for synergy. This data and the safety profile that we have observed went into INDs. We have now first INDs which have been approved and where we will start clinical trials. This already goes now to the next wave of combination therapies.
Where we are with clinical trials, we have clinical trials that are China-based. These are at the bottom. And we have global clinical trials. We have so far more than 700 patients treated across more than 10 indications. There are 19 clinical trials ongoing. And three of the clinical trials are global registrational trials. And you can see that many of these trials are chemotherapy, but some of the trials start with immune modulators and with our ADCs. And this type of strategy would fit to a staggered approach that we want to do. The staggered approach, what we call established, combined, and broaden, is start with established combination therapies. So established combination therapies, for example, with chemotherapy, where chemotherapy and checkpoint blockade is already approved. We will see some examples here. Then combined. The first set of combinations that we want to do is with our ADCs.
And as you can see, staggered does not mean we wait until we have approval-ready data for the first one, but we are doing it in a staggered way. And the third one is broaden, broaden the number of combinations and broaden the number of indications. So this explains a little bit our excitement. We had one year collaboration with Biotheus, a fantastic team, highly dedicated, culturally fit to us, driven by similar ideas. And given the very encouraging data that we are seeing, we decided to go for an acquisition. The key factors of the acquisition are at the bottom. You have seen the press release yesterday.
I would like to share a little bit more about the rationale, why we believe that this is the right step and how the acquisition could translate to increasing the likelihood that we are successful and increasing the potential that we can get out of this. In principle, we have four categories of objectives or rationales. The first one is really about acceleration, clinical trials. We have excellent collaboration with Biotheus, but doing that as one company gives a lot of additional synergies. We can do the things faster. We can come faster to conclusions and execute. It's about speed. The second is Biotheus is adding clinical development capability, particularly in China, but not only in China, but also in Australia and even in the U.S.. Many of you know that clinical trials in China could be very successful, but they are not easy to do.
With an experienced team, we have a clear understanding that doing clinical trials in multiple indications, early exploratory trials could help us for decision-making and then deciding which of the studies will go global. The third one is manufacturing. Manufacturing is always a pain and time-critical, particularly if you have a highly active compound that you want to approve quickly. Biotheus has fully integrated manufacturing capabilities I will show you later on the slide. The last one, which is also an important factor, is that this is not only BNT327, PM8002, but the company has really capabilities, engineering antibody capabilities, and a pipeline which could also create value if addressed in a meaningful manner. This is the Biotheus manufacturing site in Nantong. It's really an impressive, big manufacturing space with the capability with 2,000-liter plants.
Our first goal is really that sufficient antibody is generated for clinical trials regardless of how fast we want to go. The second goal is that this manufacturing site is supporting the launch of this molecule. Of course, we are in the process of transferring manufacturing also to outside of China and have identified vendors who will help us with the launch of the molecule, for example, in the U.S.. One important aspect, and I talked about this, is bispecific antibodies. Our prediction is it's not a prediction. It's actually visible. Bispecific antibodies are coming in the industry. In the last three years, 10 bispecific molecules were approved in different indications. We believe that this is just the beginning of a trend where the cooperative binding of antibodies engineered in a well-designed manner could help to increase the activity.
And we can categorize the bispecific simply in two categories. Bispecific that target two tumor antigens or bispecific which target one tumor antigen with an immune modulatory function. But this number of potential pathways is extremely high on the left side. So if you read a textbook, it appears to be everything is known, but it is not because most of the things have been tested in preclinical animal models. And we see that there is a lot of space for further discoveries, particularly in clinical trials. So Biotheus has world-leading capabilities. And I don't use this term often. They have world-leading engineering capabilities. So the founders have more than 25 years of international experience having worked in the U.S., in Europe. And they translated their vision into excellent antibody engineering capabilities.
One capability is how many bispecific antibodies can you make and how fast can you bring that into the clinic. They are in both categories at the top of benchmarks. They have shown that already for a number of molecules that they brought into the clinical stage, which are in Phase I testing on the left side. We are now in the process in trying to understand how this portfolio of assets can be used to further improve and augment the activity of BNT327. I'm sure that we will come in the next two years with some data showing that these assets really add additional value. Thank you. I stop here. Annemarie.
Yes, perfect. Thank you, Uğur. Needless to say that we're really, really excited about the future that BioNTech holds and the unique positioning that we have with the different platforms.
But in terms of the innovation and the research, the aspect of bringing innovations to patients and addressing the unmet needs also requires a capability that we haven't built yet and that we're now embarking on in this really exciting time moving into a commercial stage oncology company. Now, I will focus a little bit on the nearer term as we are getting ready to launch into our oncology area with a couple of our assets. Comirnaty, as Ryan already explained together with Uğur, is one of the fastest, the fastest developed mRNA vaccine in our history. And together with our partner, Pfizer, we continue to be very proud helping patients and people around the world protect from COVID-19-related hospitalizations and associated deaths. That landslide, though, has helped us as a company to build a pipeline that we're now ready to bring these innovations to market.
Now, specifically, we see the launch in the oncology area in three waves. Initially, we're talking about BNT323. So that is our HER2-directed antibody-drug conjugate ADC for patients with specific types of endometrial and breast cancer. We're hoping, as of 2026, that's our goal, to bring that innovation as the first launch in the oncology space to market. Beyond that, needless again to say, the bispecific PD-L1 VEGF is our next area of excitement that, as Uğur announced, is going to be studied in different types of lung cancer as well as specific types of breast cancer, triple-negative breast cancer. I'll be focusing on these two waves because the third wave, of course, is very exciting as we're starting novel combinations as well as bringing our mRNA vaccines to the market, is something that we will get ready for.
But the first wave is something for our commercial organization to focus on next. A few words on our Comirnaty business. As we shared, our business model, the unique feature of having a lean OPEX flowing through our P&L, has resulted in a very optimized business model, which we expect that will continue in the foreseeable future. This year versus next year, we expect to maintain or even grow market share in most of our markets, but also we expect to create and see greater visibility of the global vaccine demand for the predictability of where we go. This capability did not only, or this launch did not only establish the capability of the mRNA platform, but also created the value for us to now jump off into our commercialization for oncology. We're commercializing together with our partner, Pfizer. BioNTech holds the commercialization rights in Germany and Turkey.
For the other parts, we are leveraging the commercial capabilities from our partner. Now, with that, moving into our oncology slate. BNT323 is a launch that we see as highly valuable as a strategic launch. First of all, it will allow us, in a company that is naive to commercialization in oncology, to start building relationships, to start engaging with patients, physicians, payers, advocacy groups, which is going to be critically important as you think through the overlap that some of these sites have with different customer types, specifically in the U.S. The top 100 centers, if you look between endometrial cancer and breast cancer, there's an overlap of 40%. More importantly, if you think about lung cancer, the overlap between endometrial cancer, breast cancer-treated sites in the U.S. is 50% or more.
Early engagement for BNT323 with our stakeholders across the ecosystem will not only help us to build that capability, but serve as a springboard for future launches. Second to that is our internal machinery. Building an infrastructure and so-called running water through the pipes as we have a unique opportunity to start from scratch with AI-infused and optimized model to bring BNT323 to market and build capabilities that we can leverage for future launches to address a bigger patient population is going to be unique to our built-in commercial infrastructure. Of course, we will also focus on more traditional tactics like patient support system and making sure that we have everything in place to build a sustainable and agile customer model.
On endometrial cancer specifically, so as my colleagues will talk about a little bit later, our specific data package we anticipate to be very robust for second-line endometrial cancer across different patient segments, HER2 expression levels, prior therapy. And with that, we see and we're very confident in our ability to address unmet patient needs in remaining specific patient segments that are not being addressed today. In breast cancer, HR plus HER2 low breast cancer specifically, the patient segment where we expect BNT323 to launch after that is a huge market where current therapies already provide a lot of benefit. We also believe that 323 will be a valuable additional option which will offer physicians and patients the choice to tailor-make their therapy for their specific needs when it comes down to specific profiles, side effect profiles for patients, and where physicians will help and facilitate these discussions.
So all in all, a strategic launch for us for 323, building capabilities, building networks, building relationships, but also being able to help address remaining unmet needs that are still not being addressed today by the available options. That network, that infrastructure will then help us leapfrog into the launches that we're anticipating to come as of the 2028 timeframe for a bispecific PD-1 VEGF antibody. In different types of lung cancer, triple-negative breast cancer, despite the advancements that's made by the checkpoint inhibitors, as Uğur just explained, there is so much more unmet need. Patients respond, but patients relapse. We have the opportunity with this asset to provide more response to more patients with durable, but also move into areas where, again, the PD-1, PD-1s have not been able to help and address unmet needs. Now, the launch of 323 again will enable us to go with speed.
We recognize and understand that in this market, which will become highly competitive, speed not only in the R&D engine of our company, but also in terms of commercialization will be critically important in helping and addressing these unmet needs, so that's a little bit of what we're trying to do. The how, as I said, is phased into different stages. A strategic launch for 323 specifically where, again, from the ground up, we have the opportunity to leverage our internal AI capabilities, not just for our R&D activities, but also for commercialization activities, build an AI-infused, data-infused model to understand where our customers are, where our patients are, and follow them through the ecosystem and making sure we are there where physicians and patients need us in a timely manner to execute and bring these therapies to patients.
From that, we have the opportunity to accelerate into our next phase with BNT327, our bispecific antibody and make sure that we can scale these capabilities and that we are fast to the market as, again, this will be a highly competitive market. Then, of course, we also have our individualized mRNA platform where we're very excited to see and have the opportunity to bring BNT122 to patient segment, patient needs where we're still in the curative setting, namely the adjuvant setting. Now, one of the indications that BNT122 is being studied in is adjuvant colorectal cancer, CRC. This is a huge patient population, but specifically where one of our first studies is aimed at is patients with circulating tumor DNA positivity, ctDNA positive, which are patients that have a high risk of recurrence.
Specifically, 65% of patients will experience recurrence to the metastatic stage within two years after their surgery. Now, you'll see from the flow that this is a smaller opportunity compared to the bigger adjuvant status, but there's a theme here, right? So where we're thinking about 323 and 327 go small to go big, this is another area where we're small to go big. And specifically in this area where we're bringing a therapy that is on demand and personalized to patients, we require a full end-to-end commercialization capacity. It's akin to the CAR-Ts or the radioligand therapies where people within the organization have a lot of experience with, and we understand you have to get this right. This is not about us bringing a new therapy for our customers to just fit in their daily routine.
This is for us to meet our customers where they are and making sure that our new therapies can fit almost seamlessly in the day-to-day operations. That means that from the start, that a physician decides that this is the right therapy for patients, we help them, enable them with the visibility because we're dealing with tissue and blood samples. We're dealing with up and downstream, and Özlem will talk a little bit more about that in the next chapters. We're dealing with processes that are complex and that we don't want our customers to just send off. They end up in the abyss, and then they'll see whether or not the treatment will be delivered in the right time and the right place.
No, this requires a capability, a commercialization capability all the way from manufacturing to the moment that patients are being treated with great reliability, with great visibility, and again, going small to go big allows us to build this in a sustainable way to learn as we're already partnering with our clinical teams to understand how we're learning in the clinic and how we need to translate that to bigger patient opportunities when commercialization is there and making sure that we do this the right way from the beginning, and again, meeting our customers where they are, so this was a short session, but just to reiterate, we continue to be very proud that together with our partner Pfizer, we will continue to protect people against COVID-19 hospitalizations and death.
We expect that to be a very sustainable business for the foreseeable future, infusing our commercialization era in the oncology space. BNT323 or HER2-ADC, a very strategic and critically important launch for us and for patients. We will be able to help address unmet needs for specific patient populations, and we see and are very confident that we have unique opportunities with our ORR data package to bring that to patients, but also to then build relationships with stakeholders across the ecosystem as we're pressure testing our internal machinery to then scale and accelerate for our exciting bispecific anticipated launches in different types of lung cancer and triple-negative breast cancer. We continue to be very excited, and I'm very proud and pleased with the progress that we've seen so far in building our commercialization capacities to bring innovations to patients. Thank you.
With that, I hand over to our colleague Ilhan, who will spend a little bit more time talking about our CDP, our program with our bispecific antibodies.
Thank you very much, Annemarie. It is now my turn and my pleasure to share with you our currently available data and our vision and plan for the future, the clinical development strategy for BNT327, our bispecific antibody, and I will not repeat the mode of action and introduction already done by Uğur, but I will take more time and emphasis on the data and what we learned from this data as an exploration and how we implement this in our development plan for the future, to speak with Annemarie's words, taking the small and going to big, so this slide was already shared, and you can see our aim is to establish, to build and establish this molecule in the treatment of certain cancers in combination with standard of care, namely chemotherapy, and also, this one was already presented, but it is important to repeat.
There are areas where checkpoint inhibition is established, and on top of that, also anti-VEGF therapy is established. But there are also areas where the checkpoint inhibition is not established and where there is an unmet medical need for new treatment options with these innovative molecules. So namely mentioned here, TNBC, PD-L1 below 10%, or EGFR mutant NSCLC. So the idea for the moment is let's have a look into TNBC. And on the left-hand side, you can see the incidence of TNBC and also the staging distribution. This is for breast cancer in total, but as an example how this is really divided. And you can see this is very similar between U.S. and EU regarding the staging. But if you have a look onto the right-hand side, you can see really the outcome varies based on the PD-L1 levels of this indication.
In first-line TNBC, you can see below 10 the outcome, the efficacy, and the five-year expectation on the right-hand side where this is already established, checkpoint inhibition, also the numbers, but still room for improvement and an unmet medical need. Our partner Biotheus has conducted in the past a study in combination with nab-paclitaxel in first-line metastatic breast cancer. On this slide, which, by the way, was recently presented at ESMO, you can see the inclusion criteria. 60 patients were treated in the combination with nab-paclitaxel. The key endpoints were, of course, overall response rate and safety. For completeness, you can see on the right lower part really our benchmark comparator data, which you're familiar with.
But if we start with safety, which is always the first step in these explorative studies, having safety as an endpoint, you can see that on the right-hand side, really the AEs, the side effect profile, is very much in line with what we know of this class of compounds. And the AEs known like neutrophil count decreases, white blood cell count decrease, and anemia are very much related to the chemotherapy used here in this setting. So what we can say overall, the safety profile looks very manageable, and it is mainly grade one and two. And it is different from the individual molecules, which are known with a certain safety profile, which are numerically higher. So this profile leads to a low discontinuation number. The data for efficacy are already also shown by Uğur for TNBC, but just a quick repeat of that.
What we observe here is independent of the PD-L1 level. We are seeing really clinically very meaningful response rates, high response rates, which are translated also to durable responses and to PFS. With that, we can assume really this is adding, and one plus one is not equal two in this case. It is adding synergy and an increase in overall response rate, which hopefully we will have mature data on OS also to be presented soon. You can see in the upper right part, the plan is we will have new data available for the San Antonio meeting in the beginning of December. This observation is important. From a clinical perspective, being able to see that independent of PD-L1 expression, there is benefit. Even in the PD-L1 negatives, there's benefit. This opens really this treatment option for patients in need, especially in TNBC.
Is this unique? Is this something only we observe in one indication? Is this specific for TNBC? Therefore, we have investigated also other indications. And just to complete, sorry. So this observation led, of course, to some conclusions on our end. So since summer this year, we have started a dose optimization study, which is depicted on this slide. And you can see we are testing two doses of BNT327 in preparation for an upcoming Phase III trial, which is planned in 2025, so next year. And now, this is the right slide indicating. So it was the question, is this overall something which is typical for this bispecific molecule or very specific for TNBC? A similar study was conducted, a Phase II trial by our partners Biotheus and EGFR mutant NSCLC. And also here, you can see on the left-hand side the inclusion criteria.
BNT327 was combined with carboplatin for four cycles, followed by the maintenance BNT327 plus pemetrexed. The main or the key endpoint in that case was overall response rate. Of course, also safety and the secondary endpoints were investigated. Also here, you can see the benchmark numbers in the setting for the standard of care or the currently available treatment options. The safety profile, which you can see here on this slide, is like in TNBC, mainly driven by the side effects, which we know for the chemotherapy part. Again, white blood cell count decrease, anemia, neutrophil count decrease. All this is really mainly driven by the chemotherapy. In this case, also the side effects known for, for instance, the anti-VEGF treatment like hypertension, proteinuria are in a lower value and mainly grade one and two.
Also here, this safety profile looks manageable and is leading to a low level of treatment discontinuation. The results, interestingly, showing a similar pattern independent of the PD-L1 level, negative, low, and high. You can see here in this waterfall plot on the right-hand side that there is benefit. Around 40% in the PD-L1 negative, 61% in the low, and accordingly, over 90% in the high expressers. On the right-hand side, you can see this is not only a short increase of response. It is durable across the different sets of PD-L1 expression and populations. This was the data for this EGFR mutant NSCLC. Of course, if we are talking about NSCLC, the NSCLC without EGFR without mutations is of interest. The incidence is very high. We know this is the biggest indication in cancer.
And also the staging here is depicted, which indicates that it is between U.S. and EU very similar. If you look to the right-hand side, also here you can see there is a dependency on histology and also the PD-L1 level expression. So it is dependent on both factors. But also here we see clearly a room for improvement and a need for additional more efficacious innovative treatments to have more benefit for patients. Our partner Biotheus has conducted an early study in different cohorts, but I will focus today on cohort number one, which is the first-line NSCLC EGFR wild type population, 17 patients. We need to make a clear disclaimer. This is a small sample size, but this cohort was treated in monotherapy with BNT327. And the key endpoints were overall response rate and safety as usual.
Referring to the currently available treatment options for non-squamous and squamous, you can see really data depicted in the table in the lower right table, but this is something which we should really see as a benchmark. Looking into the safety profile, you can see also here the monotherapy treatment is leading to a safety profile which is manageable and which is really below what you would expect from treatments of this class of compounds like checkpoint inhibitors. Also here, this side effect is leading to a low discontinuation rate. Most of the side effects were grade one and two. No grade four or five AEs were observed. This is important. Manageable and tolerable safety profile is a cornerstone really to develop the structure further also in combination. This is a depiction for the efficacy on the left-hand side, the waterfall plot.
You can see here, we have overall 47% overall response rate, and the circles above the bars indicating the PD-L1 status. Without going into details and analytical explanations, this is too small to really state that this is independent, but it is roughly half-half between low and high PD-L1 expression, and the efficacy is very similar. You can see in the spider plot this is also a very durable response, and what we see here is a PFS of around 13.6 months. Again, low numbers, but very, very encouraging from the perspective to build on that, and since this data was available, we had really consideration and strategic focus on NSCLC, and this study is in preparation Phase II/III , which is also to be initiated by end of this year already. Lung cancer is an important indication, as we all know.
The next one, of course, of high interest is small cell lung cancer. Talking about small cell lung cancer, it is very clear that this is an indication with a very high unmet medical need since treatment options are limited, and especially in the extensive stage disease, which is indicated in this dark green part of the left-hand side bars. You can see that the distribution in this indication between the poor prognosis, extensive stage, and the limited stage is very similar between U.S. and EU. And you can roughly say that extensive stage is 60%-70% of the disease. Again, the right-hand side is indicating clearly that especially patients with an extensive stage disease have the very poor prognosis and requiring really attention and new treatment options. Currently available data depicted here. This is a study design of a study which our partner Biotheus has conducted.
You can see here 29 patients with extensive small cell lung cancer were treated in combination with pemetrexed, followed by BNT327 maintenance. The endpoints, again, safety and efficacy of our response rate and safety profile, of course, and the benchmark comparator data again in the lower right-hand corner, so the numbers are interesting. The most recently approved treatment with tarlatamab is ending in 14 months overall survival. The current standard of care chemo is with eight months below that clearly, so the safety data which we are observing here, again, the same theme, the same observation. It looks like really that the main side effects we are observing are directed to the chemotherapy part of the treatment. Overall, we observed grade 1-2 side effects and low level of grade 3 and higher, and the discontinuation rate is very low, it was 2.1% on the left-hand side.
That's important for us that we can say in different indications. Obviously, we are seeing the same observation. Manageable safety profile, low discontinuation because of the safety, which is giving us really the confidence that we can develop this drug in combination with chemotherapy. This is the data for this small cell lung cancer, a Phase II trial, second-line small cell lung cancer in this case. And you can see from the waterfall plot on the right-hand side, we have an overall response rate of 61% in the ITT. And since this population is mixed, some of these patients are treated with an IO approach and some others are naive to that.
This was divided and you can see the data that the IO naive patients benefiting even more around 73%, but also the IO treated patients benefiting in the range of over 40%, which is more what you usually expect from the benchmark data which I shared with you. The spider plot on the left-hand side indicates long durable responses in this combination. Also here, we will really build on this observation further development in this indication with a study, a Phase III trial, which is planned to be initiated end of this year. This is a dose optimization trial as a basis for this plan to start Phase III trial end of this year. This is ongoing since already June. We are testing different doses of BNT327 with different chemotherapy regimens to give us really the possibility to move into different regions of the world.
So dependent on where you are running the studies and where patients are treated, the backbone chemotherapy is different. So this data will help us really to pick not only the right dose, but also have data available for the different chemotherapy regimens we have in mind. I mentioned already the plan to initiate this trial by end of this year, which is in a few weeks from now, as we all know, in November. And sharing with you this data available, and we have published this data at several occasions. The second line small cell lung cancer data, for instance, were published already last year. The most recent data for TNBC, EGFR mutant, and NSCLC were published this year at ASCO and ESMO.
Stay tuned because, as Uğur has indicated in his introduction, some updates will be shared regarding these indications for TNBC very soon in San Antonio, but also for the small cell lung cancer and the other indications we are planning really updated data to be presented next year. My next two slides, and these are the last two slides, will be dedicated to the vision and the explanation already detailed by Uğur in his introduction that this molecule gives us really great opportunities, not only in combining with standard of care with chemotherapy, which is the fastest way in bringing innovative drugs to patients, but it has the capabilities really to be combined greatly with novel approaches like ADCs and other novel treatment options. I will focus mainly on the combination with ADCs.
As you heard already, we have different ADCs in-house and in our ammunition box in our fight against cancer. This is important to have this flexibility with different indications really to be able to go across tumor indications in the treatment of patients. Again, here this slide just high level shows you the four different ADCs we have with the respective target. The rationale why these targets were selected and these assets were really implemented into our pipeline is explained in the table below. As you can see, looking into the expression of these different targets and having really the plan to go across different tumor indications, the table gives you clearly one message. With these four assets, we are able to cover all the relevant indications which are important in cancer treatment.
So this is a great match of a molecule which has really unique universal capabilities to be combined with different approaches with chemotherapy and with ADCs. And forward looking, of course, the idea would be to start the combinations on top of our backbone development in combination with chemo as soon as possible. Now, is this just wishful thinking? No. We are already on the way to test exactly this combination. The combination with our TROP2 molecule is already ongoing, and we have ideas, and we are in preparation to go also with other two ADCs into the combination as soon as possible. I stop here, and I believe I can only hand over to the break. Thank you.
All right, we're gonna get restarted here. Thank you.
Okay, I think we can start now after the break. First of all, also a warm welcome from me. It's really a great pleasure to see many of those who have been supporting and accompanying us for many years now here in the audience, as well as new faces. So it's really great. What I would like to do is to kick us off after the break with reminding everyone that mRNA cancer vaccines are not created equal. Our vaccine, mRNA vaccine platform, is clearly differentiated on many levels, actually. It's the uridine chemistry with adjuvanticity effect. It's the non-coding backbone, which has been engineered. It's the routing signals to antigen presenting pathways.
It's the RNA lipoplex formulation, which allows for intravenous systemic administration and targeted expression of the vaccine antigen body wide in lymphoid compartments. I'm not sparing you the technical details, and you can read them up in our publications, which we have put out there over the last two decades because these technological details really matter and we think are the reason why our platform enables us to ensure strong T cell responses against the vaccine antigen, also the novel ones, and these are the key, basically, surrogates for the activity of the vaccine. We are using this vaccine platform, as you know, in two ways. On the one hand, we are using it for what we call iNeST, an individualized vaccine approach with autogene cevumeran, BNT122 being the compound we are developing here.
An individualized vaccine using cancer mutations, neoantigens as vaccine targets, a platform which requires on-demand tailored manufacturing for each and every patient, and the other way we are using this vaccine platform is for off-the-shelf vaccines, which are developed in an indication tumor type specific manner and where we use a different form of antigens, shared tumor associated antigens or viral antigens for virally promoted tumors, and all the data we have accumulated in our cancer vaccine trials over the last couple of years, and in the meantime, more than 700 patients have been treated across all the vaccine trials we have been and are still performing, show us that irrespective of what vaccine antigen type we use, neoantigens on the top against cancer mutations, but also tumor associated self antigens, which are known to be subject to immune tolerance.
So this is really tough to get immune responses against them, as well as viral tumor promoting antigens. Against all these, we get strong immune responses. And as I said, there is data out there, and there will be more publications. A Nature Medicine article we just got approved and further publications over the next couple of months, which will further confirm that the immune responses we induce are special. We observe, firstly, that a large fraction of patients we are immunizing, depending on cancer type, and mostly these are above 75%-80% of the patients we are immunizing, are developing de novo immune responses by the vaccine. We also observe high magnitude immune responses.
You can see them here on the slide, up to 10%, 15%, sometimes 20% of the circulating T cells of a patient recognize the tumor and only recognize these new vaccine antigens after the vaccination, not before. We also observe that we get memory type T cell responses, which ensure that these T cells are around over many years and can still be activated against the respective neoantigens or the tumor associated antigens. We also observe polyspecific immune responses, so against several antigens in the same patient, meaning that several Achilles heels at the same time are targeted by our vaccine. This is a selection of some of the clinical trials we are conducting with our iNeST platform, but also with our FixVac platform. On the left-hand side, you can see the main iNeST individualized vaccine trials we are conducting.
We have advanced trials in pre-indications, colorectal cancer, pancreatic cancer, and what just started in urothelial or bladder cancer. It has a reason why we have selected these tumor types. For one, these are all high medical need cancer types. Second, these are indications which are believed, in particular colorectal cancer and pancreatic cancer, to not be immunogenic tumors, not to be treatable by IO or immune therapies. We think with the immune responses we see at this magnitude that we can dare to go into these tumors. Another important aspect, how we select the patient populations in which we develop, in particular, our individualized vaccines, is that we believe that the best setting in which to position our iNeST platform is the adjuvant setting.
So a setting where the disease burden is minimal, where you have a situation where tumor mass is low, where resistant mechanisms and immune suppressive mechanisms have not been established so far by the tumor disease, and a setting where the immune system is much healthier. Patients have not undergone multiple rounds of immune therapies, for example, which weaken the immune system. And this is where all three of our indications are positioned. This is a setting of high unmet medical need, in particular in the indications in which we are developing pancreatic cancer, for example. Patients who undergo full adjuvant treatment and appear tumor-free after their treatment will relapse 80%, 75%-80% of those within the first five years after adjuvant treatment. This is a tumor which is checkpoint inhibitor resistant, and therefore an ideal place for us to develop our vaccine. Colorectal cancer, a huge patient population.
Many large portions of these patients are being diagnosed already in the early stage. However, after adjuvant treatment, up to 35% of these patients will relapse and need then further treatment. So starting with pancreatic cancer, the reason why we decided to develop in pancreatic cancer were observations from a small study we have undertaken in this indication in the adjuvant setting, a study only of 16 patients. However, what we have seen in these 16 patients was that only half of these patients develop a vaccine-induced immune response. This does not really fit to our observations in other tumor types where we, as I said, frequently or almost ever or almost always see 75% and higher incidence of vaccine-induced tumor responses.
However, this brought us into the situation where we had two cohorts to compare: the cohort with vaccine-induced immune responses and the cohort which did not respond to the vaccine, and investigating these cohorts, we observed that those patients who developed vaccine immune responses did much better in terms of disease-free survival as compared to the patients who did not respond to the vaccine. In the meantime, and there are publications out, and there will be one in the next couple of months, we have investigated these patients, each of these patients, very deeply, and we can confirm that those patients with immune responses have long-lived tumor T cells against their vaccine antigens and highly active and cytotoxic T cells.
Based on these observations, we have set up a Phase II trial in the adjuvant setting for patients with pancreatic cancer where we use our vaccine in combination with the adjuvant treatment, which means that these patients undergo surgery with full resection of the localized tumor, and they after that undergo standard of care chemotherapy with modified FOLFIRINOX, and we built the vaccine together with atezo, which is a checkpoint inhibitor meant to further support vaccine-induced T cells into this adjuvant treatment. This trial is ongoing, and we are very excited to see whether what we have seen in the Phase I can be reproduced in this larger setting. Moving to colorectal cancer, this is data from our own epidemiology study where we wanted to characterize the population in which we want to develop our vaccine a bit better.
These are patients with stage two high risk, high risk defined based on standard clinical pathological features, and stage three patients which qualify for the adjuvant setting of treatment. That means surgical removal of the tumor followed again with an adjuvant chemotherapy. What you can see in this patient population is that when you look for ctDNA positivity, you get further stratification with regard to disease-free survival. Patients who are ctDNA positive have only a quarter of the 18-month disease-free survival as compared to the patients who are ctDNA negative, so an even higher risk. This is the patient population which we are treating in our ongoing Phase II trial. These patients undergo surgery. They get their standard of care adjuvant chemotherapy.
What would be done normally is that after the standard of care chemotherapy, there would be a wait and watch strategy for these patients. What we do instead, and this is the randomized arm, is that we treat patients with autogene cevumeran, our individualized vaccine. Also, this trial is ongoing. We have a small biomarker group in this colorectal cancer trial, and this allows us to have, so to say, earlier peeks into what is happening to our patients, in particular when it comes to immunogenicity. So also in colorectal cancer patients, are we able to induce immune responses, T cells? And this is the first analysis in this biomarker group. Only 12 patients we were able to analyze who were evaluable. And what you can see here is that we get a fast induction from zero to high magnitudes of T cells against the neoantigens, which are vaccine-induced.
These vaccine immune responses are persistent. So we have patients here who are on this trial for more than two years. We still find the immune responses circulating, and it's each and every patient who develops this immune response. It was also interesting for us to see that all of these 12 patients, a small sample size admittedly, were still tumor-free when we did this study. Third indication, which we have just started a couple of weeks ago, is bladder cancer. Also a high medical need indication, patients with muscle invasive urothelial cancer who get neoadjuvant treatment, who get cystectomy, so the localized tumor removed, and thereafter, when eligible, are treated with a checkpoint inhibition as an adjuvant treatment, are eligible for this trial. Adjuvant checkpoint inhibition gives clinical benefit, but still a large portion of patients relapse and have metastatic disease. In fact.
This is where we are positioning our vaccine. What we compare is checkpoint inhibitor treatment in the adjuvant setting, plus combining this checkpoint inhibitor nivolumab, which is the standard of care with our vaccine. With this, I would hand over to my colleague Michael Wenger to talk about the other piece.
Thank you very much. I'm happy to continue with the vaccine on the vaccine front. Now we've been talking about autogene cevumeran. I'll talk about the FixVac on the right-hand side. Of the three that are in active development, I will not talk about BNT111, with perhaps the small comment that this drug, BNT111, is kind of the oldest of the vaccines that we have in development, also the most advanced, and also the one where we have completed a randomized Phase II study.
You might have, or you might remember, we issued a press release on the positivity of such a trial where we randomized the vaccine versus cemiplimab versus the combination of the two. The reason we're not showing anything here is basically that we have not published anything, and we will do so over the course of next year when we have a little bit more follow-up. But we're very excited about this drug as well. So what I will talk about more is BNT113, our HPV16 vaccine, and about BNT116 in lung. So what you see here is the very early results for BNT113, which are derived from an investigator-sponsored trial run by Christian Ottensmeier from the University of Southampton, now in Liverpool.
What he did is using single-agent vaccine in patients both in an adjuvant stage after resection of the head and neck tumor and in a stage where patients have full-blown head and neck tumors, sometimes metastasized. On the left-hand side, you simply see that the vaccine actually does work. What you see here is the ELISpot data on the left of the bar charts for the adjuvant in the adjuvant situation, so that we see about 80% expansion of the response and expansion of the antigens of the T cells that are derived from the antigens. On the right-hand side, a little lower response for those patients who are in the metastatic setting.
Again, emphasizing the place for these vaccines, both the individualized and the FixVac, is more likely a setting where tumors are smaller, don't need to be absent, but the adjuvant situation is probably a good place for these. But we need to, of course, test them first in a setting for ethical reasons where there's no other options. On the right-hand side, you see another experiment done now from the ongoing study, which is called AHEAD-MERIT. You see just three patients, and I don't want to claim causality here, but you see a little bit of the expansion that may be correlated to also the response, right? You see three patients. The patient that's depicted in gray did not have any discernible response to the vaccine and also didn't respond in the clinic.
The other two patients seem to benefit either by having confirmed stable disease or by actually turning a stable disease into a PR. Again, these are data that I don't want to oversell, but it's notable because there's a pattern here, and you see that also for the other vaccines that I'll come to in a second. Now, when it comes to clinical data, based on the data from the very early data from Ottensmeier, we started a randomized Phase II study in head and neck cancer for those that are CPS more than one, where Pembro is one of the standards. Some people use Pembro plus chemo, but I guess a sizable number use just Pembro for these patients.
We had to establish the safety of the combination, so we had to do a safety run-in, and that enabled us also to then publish some information on the very first 15 patients that were part of this safety run-in, our then-ongoing randomized Phase II study. That's what you see here. On the left-hand side, you see the data on response and PFS. When it says unconfirmed, this was a snapshot that we had to use for the publication. In the meantime, all of these unconfirmed responses have been confirmed. Just take this for what it is. Unconfirmed does not mean none of them were confirmed later. It just means that at the time the snapshot was taken, they were unconfirmed. You see a response rate of 40%. Now, the standard for these patients in head and neck cancer is about half of that, right?
So around 20%, sometimes a little more, but definitely not much more. Again, small numbers, but it's remarkable that we're seeing some response with the addition of the vaccine. On the right-hand side, what you also see is when you look at these swim lane plots, is that some of these patients stay for a remarkably long time without much happening, right? So what I mean by this is they come in, they have stable disease, they don't progress, but they also don't regress. But then some patients you see here, for example, this patient number 114, after half a year of treatment, turns to a responder, right? And for those that know head and neck cancer, it's a fairly aggressive tumor that tends to be not spontaneously regressing that much. So again, one of the features of the vaccines, they probably need longer treatments.
In this setting, we do seem to have a case that we are able to show that patients benefit from this. This is the actual study we're doing. You see on the 15-patient run-in cohort that I just showed. On the bottom is the randomized part. This is ongoing, 280 patients. We actually did discuss this study with the FDA, and the FDA is quite interested in working with us with this study. We had a few comments from them. We implemented them. This may very well also have registrational potential. The study is ongoing, as said, has recruited around two-fifths or so, so around 120, 130 or so patients. We see a ramp-up of the recruitment. We're quite excited about this. You see on the bottom the benchmark, but I already alluded to this. Changing gears, again, to BNT116.
That's our youngest, newest FixVac for advanced lung cancer, non-small cell lung cancer. The number one question we got in the past is, what are actually the tumor antigens that are included in this vaccine? You see them listed here. It's in the meantime published. What you also see, similar to the experiment you saw earlier, which is probably the number two question, like, how do you actually know that this vaccine works? How sure are you that all patients with lung cancer would mount an immune response? That's the data we have here that basically shows of a collective of patients that we treated, we are pretty sure that around 90%, perhaps 95% of patients will mount an immune response against at least one of the TAAs, but for most, it would be more than one.
This is important because for an off-the-shelf vaccine, we don't just want to have it off the shelf. We also don't want the CDx for it, right? So we basically want to develop these vaccines so that they can be used truly for any lung cancer patient that comes into the clinic without the need for testing of the presence or absence of any of these TAAs. Now, for a trial, of course, we do test them, but we also don't use this for inclusion. It's not an inclusion criterion. We let them in without. On the right-hand side, you see the expansion of the expanded T cell responses, de novo and after in vitro stimulation, which also shows that there's something quite substantial going on.
Now, this platform study that we do for LuCa-MERIT is our entry into human and a platform study where we try to combine, or where we are combining, the vaccine with a variety of different agents. You see them listed here, maybe a bit confusing, many different arms, but essentially, we're combining this with checkpoints in patients that are checkpoint pretreated, right? We're also combining this with standard chemotherapy, like docetaxel. We're using this in frail patients, which is a large group that would not be qualifying for chemotherapy. We're using this after radiotherapy, and we have a neoadjuvant concept, and we're not talking much during this meeting about our CTLA-4, but there's also now a cohort that combines it with our CTLA-4, which is an IO-IO combination, if you wish, which is also kind of interesting, specifically after PD-L1 treatment.
Now, what you don't see here is forthcoming cohorts. An obvious combination is with BNT327, right? So this is something that will also come. Two more combinations are with our ADCs. And again, we focused this particular meeting not on the ADC, so I won't talk about this, but I definitely want to highlight this that this is planned, and this is also coming, and it would be a nice addition for these FixVacs for specific patients that do have active tumor ongoing and where immediate treatment is necessary. So what kind of data we have of the first three cohorts? There are some data. On the left-hand side, this was the very first cohort, the entry into human cohort, last-line patient, single-agent FixVac. You see a bit of response here, but probably nothing to write home about.
Importantly, though, we established the safety, and we also established the dose for this FixVac. In the middle, you see probably something already quite interesting that actually sparked the interest of many of our investigators in the lung cancer field, ranging from many sites, MD Anderson, etc., that want to work with us now, because if you know after checkpoint blockade in frontline, second-line outcomes aren't that great. Docetaxel is still used as standard of care, mostly because even though it's a bad treatment, nothing has beaten it by now. And we're adding here docetaxel to the vaccine. And again, I alluded to the pattern earlier. You do see responses, and they are sustained. They may not go down to zero in the sense of a CR, but this is where we can have also the next hit, right?
We can add something else to this, perhaps an ADC or so, to eliminate the tumor completely. So we're quite excited about these data, and we had got some very nice feedback from our investigators who want to work with us. On the right-hand side, you see the combination with sugemalimab. Note, these are patients who have received a PD-1, typically Pembro or Nivo in frontline already, right? They're either relapsed or refractory to a checkpoint inhibitor. And we do see here, again, a lot of stability, disease stabilization, and some responses kind of suggesting that perhaps the vaccine may play a role in reestablishing susceptibility to a PD-1. So for all of this, again, we're quite excited about the FixVac. We're expanding this program, and we hope to have a few more vaccine studies very soon.
I'm switching gears now, continuing with ADCs, and Özlem will also continue then with BNT211. But before I start with BNT323, which is obviously our HER2 ADC and the study or the studies where we also will have soon some regulatory action, hopefully, I want to emphasize that for BNT324, which is our B7-H3, again, not the main topic of this meeting, but I have to just say that we will have our coming out for BNT324 at ESMO Asia. So for those that are interested in this, please watch out for these data, because we do have a sizable number of patients for BNT324 or B7-H3 ADC, and with a focus on small cell lung cancer. And I encourage you to watch out for the abstract and the data presentation at ESMO Asia. Plus, for BNT326, we will also have some interesting data later next year.
Again, I don't want to talk too much about these because it's not the topic here. Switching back to BNT323, you're all very familiar with the mode of action of an ADC. The point that I tried to make here is it's the most modern form of antibody drug conjugates. Like we call them third generations, there's some classifications around ADCs, but this is clearly something using the most modern toxin. It uses a cleavable linker, which is also the standard of care these days. Both are proprietary, and both may mean that we have an edge regarding stability, which may lead to also better safety. It may also mean that we have an edge over established drugs regarding efficacy, even though it's a bit too early to tell. Preclinical data showed specifically plasma stability was better than other drugs.
Again, how this will translate into the clinic we'll have to see. This drug was developed originally by our colleagues at DualityBio. They did a fairly standard Phase I/ 1B study with a dose escalation. We came out with a recommended Phase II dose of 8 mg per kg, which then went into expansion cohorts. You see them listed here, both in endometrial cancer, which back then when this program started was untapped potential for any HER2 ADC, and of course in breast cancer and several others. We decided for strategic reason mostly to focus on breast cancer and on endometrial cancer. For breast cancer, we started a randomized Phase III study already that is ongoing and also recruiting nicely. It's run by our colleagues from DualityBio, and we have also some other plans for another breast cancer study that we talk perhaps another time.
But for now, I want to just show you the little data we have published thus far in endometrial cancer, which you see here. These are kind of old data from last year, has been shown a few times, only, what is it, 13 patients or so. But I think you can appreciate that this drug simply works beautifully in endometrial cancer, specifically when it comes to response, but also duration of remission and PFS are top notch. You see on the bottom the data of the 40 patients that were published by our colleagues at AstraZeneca and Daiichi Sankyo. So we're more or less on par with that, perhaps a little better. But by now, we have concluded the Phase I study with around 110 patients.
In-house, we have a lot more data, and we're quite confident that these data are on par or perhaps better than what you see here. We have a much larger data set than they do. In terms of safety, this is a safe drug. We do see side effects similar to T-DXd, such as ILD, but not to an extent that is worrisome or is any different or worse, I should say. Maybe it's different, but it's not worse than what we see for T-DXd. These data will also be updated next year, early next year, and we'll encourage you to look at this as well, because these data from this study, again, have been pre-discussed with the FDA and will form the basis of a potential submission that we plan.
Because we also then, in case we get to a successful submission and potentially approval for this drug, we will need to do a confirmatory study, and this is a study that is already listed on clinicaltrials.gov. We'll soon enroll the first patient, and it's a fairly standard 2 to 1 randomized Phase III trial that we are very hopeful to enroll fairly fast, because we work with the two largest global study groups, the GOG and the ENGOT, for this trial and have established really good relationships with them. So we're quite confident we can enroll this. You see here the benchmarks. Again, you can probably extrapolate what we are looking in terms of completion of this study. As I said, the study is already through regulatory approval. We're waiting for the first patient very soon. With this, I'll hand back to Uğur. Thank you.
Yes, this brings us to the last asset we want to talk about today, our Claudin-6 CAR-T cell approach. There are many reasons why CAR-T cell therapies have not made it into the solid cancer space. One of those reasons is that there are not many highly or sufficiently selective target antigens around. And this is one of the things we address with our Claudin-6 CAR-T cells. We have discovered it 10 years ago, but actually 20 years. This is how old we are in the meantime. Claudin-6 as a highly selective solid cancer target. It's not expressed in normal tissues at all, which we don't see frequently in solid tissue antigens because it's an oncofetal antigen. However, it is activated in many tumor types. It's a pan-tumor marker. Some of them, not all of them, are shown here.
Testicular cancer shows the most frequent expression of Claudin-6, germ cell tumors. More than 90% of tumors express this target. So there is something also biological going on here, but also other tumor types, ovarian, endometrial, gastric, many rare tumors, also pediatric tumors. Claudin-6 positive. One tumor, you have heard that lung cancer is a space we are particularly interested in, and we are, of course, also interested in providing something for the later stage patients. CAR-T cells are of interest here, and we also see a large portion of lung cancer patients, in particular non-squamous cancer patients, expressing Claudin-6. So it's in principle, Claudin-6 is in principle the portal to a large pan-tumoral patient community.
One of the other reasons why CAR-T cells are not in the solid cancer space yet is that in solid cancers, the persistence of adaptively transferred CAR-T cells is short. They are not encountering their antigen, which is compartmentalized in the solid cancer lesions, and also this is addressed with this program, our Claudin-6 CAR-T cells or BNT211. We address this by providing the target antigen in form of a vaccine, Claudin-6 CAR vaccine, again, based on our RNA lipoplex platform, so what we do is that we adaptively transfer the CAR-T cells and then iteratively, repetitively give our vaccine when we see that the CAR-T cell levels drop, and in preclinical models, you can do this infinitely, many, many times.
And we in the meantime have also experienced that also in patients, you can do this multiple times and therefore ensure that and thereby ensure that CAR-T cells stay persistent. This is the design of our still ongoing Phase I/II trial, which is across different Claudin-6 positive indications. This is a dose finding or regimen finding trial where we test different doses of CAR-T cells with and without the vaccine. We call this approach CARVac, CAR-T cells amplifying vaccine platform. We have in the meantime treated more than 100 patients across different indications with different dose levels. And we have already reported and published that we see clinical activity in a number of those indications, small cohorts, but clear signals. I want to remind you, these are patients which have been multiply pretreated with different compounds.
Here the germ cell tumor cohort, testicular cancer patients, for example, we see a clinically meaningful objective response rate of 40% and disease control rate of 75% on the dose level, which we see as one of those for further development. Ovarian cancer patients, also here we see clinically meaningful activity signal. Other cohorts are not filled sufficiently with a sample size, which allows definition of a clear activity signal. One of the cohorts we are continuing to recruit is, for example, our non-small cell lung cancer cohort. We only have three patients here. However, in two of them, we already see partial responses of large tumor masses, as shown here, and this is also one of the indications, as I said, which is of interest for us and might also serve as one of the indications we might continue to develop this compound.
What we also have already reported that adding the vaccine indeed increases the persistence of CAR-T cells and thereby confirms our is a proof of our principle to combining both compounds. The indication in which we want to go into Phase II testing, the first indication in which we want to do this is germ cell tumors, which are resistant and refractory. This is our real-world evidence study in which we have shown that patients who have not responded or relapse after high-dose chemotherapy or conventional dose chemotherapy have only a median overall survival of 7.7 months. So a high medical need indication. Most of the patients with germ cell tumors respond to primary treatment. Only a small fraction of patients come into the situation that they need some sort of salvage treatment.
But for these patients, as you can see, the medical need is high and survival is short. So this is a small patient population. However, it is a patient population with a very high medical need. These are patients who have the highest number of lost life years outside of the pediatric population. The clinical trial for which we are preparing is depicted here. These are germ cell tumor patients who are relapsing after the first-line of salvage treatment, either high dose or conventional dose chemotherapy, which first go into a safety run in phase in which we want to confirm in this population and the specific patient population the regimen to which we have narrowed down in our Phase I/II dose finding, regimen finding exercise.
And one of these two arms shown here, different dose levels with and without vaccine combination, will be then selected for further use in the main part. And in this main part, we will treat 62 patients in a single arm with the selected regimen. This trial will start sometime early next year and will recruit in the U.S., but also in Europe. With this, I would hand over to Ryan. There he is.
Okay. I'm just going to provide a few closing remarks, and then we're going to bring the whole management team up to the stage and take your questions. And I'd like to start just by looking back from 2023 to 2024 and the progress that we've made in the company. With COVID-19 vaccine franchise, we've maintained a strong market share.
Indeed, we actually believe that in several major markets, we've increased market share in 2024 versus 2023 with our partner Pfizer. On the oncology portfolio front, we've started a significant number of trials, seven Phase II or III trials started, and we've dramatically increased our enrollment rates in oncology across the whole portfolio. This was a major focus as we came into 2024 to be able to increase our execution power to handle the large amount of trials that are coming. On the infectious disease side as well, we didn't talk about it today, but we've started a number of additional Phase I trials, again, beyond COVID-19, and have progressed through those to data, which we hope to share with you next year in some cases.
And on the corporate development side, we've announced two acquisitions, one of them InstaDeep, which has been closed, the other obviously yesterday with Biotheus. And we've in-licensed six molecules. So we've been very active on the corporate development side. And we've done all of this while maintaining a cash balance, actually growing our cash balance. If you look at Q3 2023 to Q3 2024, our cash balance has expanded. Now, of course, it doesn't include the latest acquisition yet that we've just announced, but I think it demonstrates the power of our model and our strategy. And if we look at our capital allocation over the past two years, 2023 and 2024, you see here how we've spent our funds. So R&D investment has increased from 2023 to 2024. Of course, that was deliberate with an increased focus on late-stage trials, in particular in oncology.
You've also seen M&A and licensing be a supplemental lever for us to deploy capital. We've been judicious in how much we've spent and what kind of assets we've pursued, but we think that's already created value in a very short time period. On the CapEx side, again, we've been judicious as well, investing for long-term growth. Now, in terms of the outlook for 2025, I'll just say a few points. We'll give more specific guidance next year, but we intend to strategically invest behind our late-stage programs, in particular some of those that you've heard about today, 327, RNA cancer vaccines, and others. We're going to continue to be very active in terms of portfolio management to create additional P&L headroom to enable that investment in strategic growth.
And of course, we expect to continue to benefit from our large strategic cash reserve in the form of our ability to generate interest income. So looking ahead over the next one and a half years, we do see a very rich set of data news flow that we expect from the pipeline. I'm not going to go through these points one by one because we've covered a number of them today, and we're happy to deep dive with you in the Q&A session or subsequently. But I think suffice it to say this 2025 is going to be one of the most busy years in terms of pipeline news flow that the company's had, and we're excited about that. I think just a couple of data points to note. We talked about potentially our first filing data set for the HER2 ADC.
We've also talked about our lead adjuvant study in colorectal cancer for our personalized RNA cancer vaccine, where we do expect data late 2025 or early 2026. That's going to be an important news flow point for our pipeline. And of course, the ongoing news flow around BNT327. Many data sets there in 2025, likely to be Phase I, two data follow-ups. And then as we go beyond 2025, we expect to start to get into the Phase II/III data updates across different tumor types. So with that, I'll close, and we're just going to get some chairs up on stage. Give us a minute here, and we will open up the floor for questions. Thank you. Okay. Okay. Daina.
Hello. Thank you, guys. Daina Graybosch here from Leerink Partners.
I want to start with a question on BNT327, which is targeting PD-L1, where a lot of the competitors are targeting PD-1. And I like sort of a two-part question, why you think it's differentiated from clinical data. And the second thing is you spoke a lot about similar activity in PD-L1 low or negative and high tumors, which is really different than what you reported with the PD-1 and PD-L1, where activity was quite restricted to PD-L1 high, which makes sense because that's where you're getting the PD-L1 actually anchoring to the tumor microenvironment. So how do you reconcile your hypothesis on the PD-L1 mechanism with the similar activity in PD-L1 negative and high so far?
Okay. Thank you, Daina. I tried to repeat the question. The first question was, how do we differentiate in PD-L1 positive tumors where classical checkpoint blockade is already active?
Is that your first question?
The first is why you differentiated versus the PD-1 bispecifics.
The PD-1 bispecifics. We do not know if we differentiate. It's too early. It's, of course, different clinical trials, and you can't cross-compare. But what we are seeing, what I have already mentioned, is the PD-L1-positive tumors are more or less tumor types which are positive for the tumor antigen PD-L1, which is targeted by a bispecific with PD-L1 binding properties, but not with a bispecific which has PD-1 binding properties. So this is the functional feature. Whether this functional feature translates into a higher efficacy, we don't know. But you can imagine and do some sort of studies preclinically. It would be more effective to do that clinically. What happens in patients who are PD-L1 positive? So far, we have only small data sets.
One data set is in the EGFR-positive second-line advanced NSCLC with an objective response rate of 100% in this patient population for the PD-L1 high population, so this is interesting, but it's a small population, and in the TNBC data set, we had also 100% response rate in the PD-L1 high population, also interesting, doesn't prove anything, but goes into a direction that the PD-L1 high population could particularly benefit from the treatment, and we have to see.
And I think the second question was around the comparison of the mechanism of action in terms of you asked, with the PD-L1 and VEGF, we saw more activity in the PD-L1 high expressors. How do we reconcile that with our assertion that we believe that the VEGF–PD-L1 is going to be different,
so you repeated the question, and I had hoped that you are answering it also. Okay. Okay.
In the PD-L1 negative population, you still have the option to bring in an anti-VEGF into the tumor compartment. And this is different from a perspective of a PD-L1 for 4-1BB, because for 4-1BB, PD-L1 is negative. For 4-1BB, it will be negative because for 4-1BB, it is correlated. So we are really bringing a new function into a cold tumor. So this is a differentiating factor.
I guess maybe if you allow me one more question, I wonder how you're thinking about the political risk with Biotheus and your manufacturing in China.
This is something that you should take.
Yeah. So obviously, we did the licensing deal with Biotheus already late last year, and it was our strategy from the beginning to establish ex-China supply for the 327 molecule. That hasn't changed.
I think what this acquisition does, one of the benefits that we alluded to was that it gives us control or will give us control even over China production, which we think is just an added de-risking element. But the core strategy remains to establish, as soon as we can, supply ex-China, and we think we can do that to support launch.
Oh. Okay. Hi. Tazeen Ahmad, Bank of America. Maybe Uğur, a question on Biotheus. So now that you have control over the full company, are there specific programs that you could highlight that you now will have full control of that investors may not have been paying attention to that we should be and that you're particularly excited about?
And then secondly, the question I have maybe is for Ryan as it relates to speed with which future versions of the COVID vaccine will be approved specifically in the U.S. What's your level of confidence that the efficiencies that were gained during COVID in terms of getting new versions approved will remain? Thanks.
Uğur, do you want to take the Biotheus pipeline question?
I can maybe take it. Tazeen has shown the list or a list, a selection of the bispecifics and antibodies which are in the pipeline, and we have not yet decided on what to develop further. All of them are interesting in particular also as combination partners for BNT327 at some point, and it's great to have all of this under one roof. However, we are going through the data for many of them.
There is already clinical data, which also is an upside, and we'll then see how each of them at which time fits into the overarching strategy, which is for these new members of our pipeline, very BNT327 anchored, part of the BNT327 program.
Then on the question of next-gen COVID vaccines, maybe I'll speak just to the regulatory pathway briefly, and then maybe Uğur also may want to add on the next-gen development. I think as we showed, the epidemiological data would support an earlier approval if that pattern continues into the future. I think we saw already to start this year a push, as I alluded to in my slides, that across countries, including the United States, to enable an earlier approval. I think that it ended up that Europe came first.
Europe was kind of the first mover in that sense with a 60-day improvement in the release of JN.1 versus 2023. I think the United States, what we saw was actually that the switch to KP.2 ended up delaying the rollout of vaccines. So while they were earlier in 2024 versus 2023, they weren't that much earlier. And that, again, seems to be directly linked to the timing that the KP.2 strain's rise had and the impact that that had on the decision-making by the regulators. I don't know. Uğur, do you want to speak to next-gen development?
Yes, I can speak as much as I'm allowed to say. So what we did from the very beginning is not only to assess the full spike protein, but modifications of the full spike protein with, for example, multiple additional proline sites.
We have published data on subepitope vaccines covering only the receptor binding domain. We have tested receptor binding domain membrane anchored, and we have, in the meantime, evaluated a number of additional domain and domain combinations. That's one level of changes. The second level of change is, of course, the mRNA technology is progressing. So we are learning about further improvement of the non-translating regions, about optimization of the secondary structure, additional codon optimization strategies, ways to manufacture and purify these vaccines, and we are seeing that this translates in the preclinical setting to a substantial improvement, and improvement is defined by getting higher antibody titers at lower doses or higher duration of the antibody titers.
So this will be a focus of our next-generation vaccine development, particularly also aiming to provide a better way of combining the COVID vaccine with influenza vaccines, keeping the reactivity profile of the combination low and immune responses high for both components.
And maybe one last follow-up to that question. Is the frequency with which you expect patients to have boosters going to change from, I think you had said, roughly annual? Could it be less frequent than that in your next gen?
I don't think so because we are seeing really the slide that Ryan has shown. The virus is really mutating, and even though these variants are still called Omicron variants, they have a substantial number of mutational changes. So we had change from XBB to JN.1 to KP.2 within a year.
So therefore, we believe that yearly updates, as it is done for flu vaccines, will be useful. I'm not sure whether these updates have to be in the winter season for the winter season or whether it could make sense to bring them earlier because the infectivity rate of the new variants is substantially higher than what we are seeing with the flu vaccines. So one can consider to start the vaccinations earlier in the year.
Thank you. Chris Shibutani from Goldman Sachs. Two questions on 327, if I may. I think we're all intrigued by this whole bispecific mechanism, and clearly, there are competitors who are active with Akeso. And then there's news this morning that Merck is partnering also with another bispecific.
Can you help us get a little bit smarter in terms of how we should be thinking about looking for ways that these molecules may differentiate? Your team, obviously, made your choice. And when we're looking at the limited clinical and preclinical data, what aspects do you think are very important for us to be looking at? And is that a factor in how you're choosing which tumor types and indications to go into, or is it more of a strategic kind of playbook that is guiding your thinking? My second question is, where are you on a potential subcutaneous formulation of 327, and where are you philosophically on testing in the adjuvant treatment setting? Thank you.
Do you want to start? Sure. Okay. Good. Okay. So differentiation. So we always see differentiation as a fact of two things.
It's about science, science and engineering, because engineering must follow the science. We believe that the antibody BNT327 PM8002 has been engineered in a very reasonable manner. I will not go into the details, but you should be aware that Biotheus had also the option to engineer the antibody in a different way with the PD-1, but they decided to go with the PD-L1 design. At the end of the day, the preclinical data are very similar for these antibodies. It's difficult to make a differentiation in the preclinical setting. The preclinical studies do not give you an answer for PFS. There is nothing in preclinical models that can give you progression-free survival.
So all this preclinical data suffers from the aspect that you have to find the dose for the tumors in a way that you can treat the mice because the tumors are just growing too fast. So therefore, particularly for the bispecifics, I believe the science will be done. Part of the science will be done in the clinical setting. So we are talking now about making scientific decisions based on biomarker studies, based on responses and response rate in Phase I studies, and for these exploratory studies are needed. And we are all aware that we might be wrong with our hypothesis. So being open and doing the studies in a scientific accurate fashion is important. The second is about execution. Fast execution is important.
There are obvious data where even without having OS, it really makes sense to engage into registration of trials because the PFS is so much differentiated that it would be very surprising if it does not translate into OS. The third thing is about combination. This was a topic that we addressed in a setting. We have to kill the tumor cells. Killing is an important aspect. This type of immune modulatory compounds don't work for themselves alone. We have to reduce the tumor burden. This is the first approach that we are going to do. The second approach, we would like also to remove the remaining tumor cells. Here is where vaccines are coming. We will not see that in 2025 coming, but in 2026, we would come up with what we call broaden the combination spectrum.
SubQ and adjuvant?
Yeah.
Yes, of course. These are things that are what I would say is the typical path to go. And we have adjuvant, our personalized vaccines. And if you consider to strengthen the personalized vaccines, you would go adjuvant.
Hi. This is Cheung from Oppenheimer. Thanks for hosting the event. So first question may be similar to the previous one. Just wondering from tumor biology perspective in the early stage cancer whether it makes sense where you can expect a superior profile with a bispecific versus checkpoint inhibitor PD-1, PD-L1. And thinking about the adjuvant setting, whether it makes sense to do combination there or you think that's a path to just do monotherapy with 327.
So just didn't hear that well either. Your first question was, was it relating to PD-L1 status?
Related from the tumor biology perspective, if you kind of can expect a similar superior profile with a VEGF–PD-L1 bispecific versus PD-L1, PD-1 checkpoint blockade there. But you can have a better profile there as well in the early stage in the either perioperative or adjuvant setting there.
Okay. I tried to answer what I have interpreted. So in the adjuvant setting, we have, of course, two types of patients: patients who are really free of tumors and they continue to perform well, and patients who have tumor nest that are hidden in the body and which are not visible. And because of this, the patient is called adjuvant. Otherwise, the patient would be called metastatic. So we are thinking about the biology of small tumors. The biology of small tumors are characterized by two things.
One thing is that these tumors need to grow up over time, and for this, they need additional blood vessels. So VEGF blockade could help to delay that. And the second is that small tumors have a less established tumor microenvironment. And this gives you, of course, better chance for immunotherapy because T cells, if the tumor microenvironment is not well established, is easier to deal. And for the adjuvant setting, therefore, we have two strategies. One thing is to consider to have a vaccine alone. So in the colorectal cancer study, we have randomized patients who received chemotherapy alone, and then the treatment arm is chemotherapy followed by vaccine alone. So this is already interesting that there is no PD-L1 blockade in this treatment arm. So the readout will give us the information whether T cells can do the job alone without being helped by checkpoint blockade.
In the pancreatic cancer trial, we have PD-L1 blockade. So this is the second type of dealing with that. And of course, as Michael stated, there is now a third option to have not only PD-L1 blockade, but the bispecific blockade. And this will come, and this will teach us what is essential and how this translates to a clinical benefit structure.
Thank you. If I can just ask a quick follow-up on the Biotheus, just first congrats on closing the deal. But I think Biotheus may also be interested in autoimmune and also inflammatory diseases. So wondering if that's something you may consider in the future as part of the BioNTech. Thank you.
Yeah. So yeah, you want to speak to it? No, no, please. Well, yeah, so you're right.
We only showed a subset of the assets that Biotheus has in their pipeline focusing on the clinical stage. And you're right. There's a long tail. As Uğur talked about, the productivity of their platform has been very high. So indeed, they do have autoimmune assets as well. And of course, those come with the acquisition. We're buying the whole company.
Great. Thanks for taking the question. Terrence Flynn, Morgan Stanley. I have a two-parter on 327 and then one on 323. So for 327, just wondering in the Phase II/III first-line lung study, if you're going to, I'm assuming you're going to use Keytruda chemo for the control arm, but just wondering if you can confirm that, number one. And then if you can tell us anything about powering assumptions for that study, like what kind of effect size you'd hope to see.
And then on the Biotheus deal, can you tell us what the royalty rate would have been that you would have owned on 327? So that's the 327 question. And then for 323, just wondering in endometrial what the bar is for accelerated approval. I know you had some preliminary conversations with FDA. What do you need to show for ORR and DOR? Thank you.
So maybe I start on the royalties because that's the easy part. And then we so double-digit royalties, and it was tiered double-digit. And we also had over $1 billion up to $1 billion plus in milestones from the first deal that we did, just to put that in perspective. You want to speak to lung?
Okay. So for BNT323, you saw data, right? So it has around 40% or so response rate.
What the FDA typically will ask for is this data plus the degree of durability, and we are very confident that we can show that. In the preliminary discussions we had with the FDA, one of the factors played, of course, the data from T-DXd on their 40 patients, and I think if we're getting into that range or perhaps even a bit better, which we are quite confident we can be, then this would be in line with their expectations for granting approval.
I'm sorry. The first part of the answer was the data set you were referring to?
The data that you saw here, like we—I didn't quote any DOR. So we haven't published our data, but in the paper of the PanTumor trial, there are some data for also the endometrial set, and we're quite confident we can show similar data or even better.
Regarding your question for 327 and the non-small cell lung cancer, right, trial, you asked regarding specific design assumptions or what is?
Yeah. first-line, first-line, non-small cell lung Phase II/III. Are you going to use Keytruda chemo as a control arm and then anything around powering?
Well, regarding the design of the study, it's a Phase II/III study because we are using the first part, of course, also for some dose optimization. The assumptions for the powering and the really detailed sample size, we would not like to comment at this stage. We are, of course, in constant contact also with the health regulators, and there are some discussions ongoing. Very soon, you will see anyhow an update on the ClinicalTrials.gov homepage with the specifics. But the comparator, according to histology, as known for squamous, non-squamous, I think that's the obvious one.
Hi.
This is Jana from TD Cowen. I have a couple of questions on BNT327 as well. First of all, what specifically gives you confidence that kind of these great PFS benefits that you're seeing are ultimately going to translate into OS? Obviously, we've seen previous kind of PD-1 and VEGF, not in one bispecific, but these combinations with Avastin have not really panned out in terms of OS. So why are you so confident that BNT327 will show OS benefit? And then following that, just some questions on the Phase III trial design for your TNBC study. You've noted in your presentation that outcomes in first-line TNBC differ based on PD-L1 status, but that BNT327 shows efficacy regardless of PD-L1. So for the Phase III, are you going to be looking at PD-L1 less than 10%, at greater than 10%, or are you going to look at both?
Thank you so much.
Okay. Okay.
Maybe I can start with the first question. What makes us confident that we will see an OS effect? That was the first question, correct? And we had that already. First of all, some of our data, as Uğur has pointed out, our partner, Biotheus, has data based on more than 700 patients, different cohorts in different indications, different combinations. And we can see at least PFS data, early OS data. Some of that will be presented next year. Do we already show something on San Antonio? Yeah. As I indicated, San Antonio will come with OS data. So stay tuned. So based on this still early data, as I said, this is not randomized or anything. These are signal-seeking cohorts. However, on a large number of patients and pan-tumor, we feel confident. This is one.
And the second is whenever we hear the OS question, it's clear to us that this is because of the bevacizumab experience, right? Because we all have made the experience that it might be underwhelming and PFS might not always be an indicator that we would also see the OS. Here, we need to keep in mind that we not only have the anti-VEGF-A arm, we also have the PD-L1 arm, right? And PD-1, PD-L1 compounds are the poster child for an OS effect, for the effect of elevating the tail. And this is an additional confidence bringer.
Absolutely. Maybe I can add. So it depends on indication where you can maybe speculate whether PFS translates into OS.
There are indications where this is maybe not the case, but in most of the indication, PFS, if it is meaningful and high enough on top of overall response and duration of response, the observation is usually that this translates to OS. Small cell lung cancer is an exception where it is proven that PFS is not in all cases translates into OS, and therefore the primary endpoints need to be used differently. Nevertheless, high overall response rate with long duration of response gives you some assurance that this cannot drop so much that you would not see a meaningful increase in overall survival.
So the totality of what was already mentioned across different indications, the safety profile, which indicates that it is not one plus one, so anti-PD-L1, anti-VEGF, so speculating that you would see the same outcome as you in individual molecules application, I think that's for us a confidence here based on this data in a large scale of indications and patients.
The second question was about the histology and the PD-L1 level in our study.
Yeah, just for the Phase III design for your breast cancer study, I know that you said there are different outcomes depending on PD-L1 status. So I was curious what the specific patient population was going to be.
Maybe I can comment on that. So not going into all details, which we cannot disclose at the moment. This study is planned and is preparation to start next year.
But we have mentioned which indications, which settings are really at the moment not used for checkpoint blockade. And I think it was on one of our slides that TNBC in less than 10% is an unmet medical need. So you can assume that, of course, we will try to cover this area. The above 10%, there are treatment options, but this is what we are currently considering overall, and we will make a final decision in which segment we will go.
Thank you so much.
Hi. This is Josh Zaharoff on for Cory Kasimov from Evercore ISI. We got a couple on 327, please. Will you be pursuing further development in the EGFR positive non-small cell lung cancer space? Why or why not?
And with 327 Phase II trials now running in the U.S., I know it's still early days, but how have adverse effect rates, generally speaking, trended versus your experience in China?
Okay. So did you get the first question? EGFR mutant lung or not? What's our thinking on that?
I think I can say that lung cancer and non-small cell lung cancer are indication, which is one of our focus indications. So we are looking into different segments of non-small cell lung cancer, and EGFR mutant ones are definitely one of those in which we will, which will be part of our clinical development plan. Also based on the data Ilhan has shown earlier.
Exactly. And we have different options also as we have presented with the opportunity not only with standard of care with chemo, but also with our molecules, the ADC molecules.
This gives us additional options here to combine. So we are considering really an evaluation also in this setting in EGFR mutant. Yeah, that question was? Can you repeat the second question?
Adverse rates. Right. So whether or not we have any read that we can share on adverse event rates from the global studies versus the China studies.
So we have presented to you the ongoing dose optimization studies, which are studies to evaluate different doses of BNT327. Recently started, the patients enrolled in the study are not showing any hint of safety signals which are concerning. To summarize it, it's too early to really compare against the studies already conducted. So we need to wait and collect more data. And at a certain point, we will be able to make a conclusion. But exactly for this reason, we are conducting the trials. The short answer is too early.
Hi. Etzer Darout, BMO Capital Markets. A couple of questions. First on 325, AstraZeneca is employing this computational biomarker approach, QCS, to help identify patients with TROP2 expression. Just wondered, as you're kind of thinking about longer-term studies, especially going into non-small cell lung cancer, we're going to get a readout from them later next year. Is this something you're thinking about as well in terms of enriching patients for TROP2 in your studies? And then on 327, sorry if you may have addressed this before, but in the first-line, triple-negative breast cancer data that you had up, you had sort of increasing response rates with PD-L1 expressions increasing, but sort of the reverse trend in PFS.
Is this sort of a small N effect, or is this sort of the actual trend that you observe in terms of response rate versus PFS in that first-line triple-negative breast cancer population? Thank you.
Okay. So I can try to answer the first question. As you are aware, we have our own powerhouse AI group. They're looking carefully at the options, but we also have a degree of skepticism regarding the AZ data, largely because it's too much of a black box.
I shouldn't comment on the strategy of other companies, but if it's not possible to disclose the methodology or how they derived with these results, I see a pretty difficult regulatory pathway for them to get an AI approved or an AI algorithm that is not really an algorithm because it's something that the machine does, but we don't understand what they actually do to approval. The data looked impressive, but the actual applicability is post hoc, right? So it's not something that we think we want to reproduce. We are though looking into this with our own AI capabilities to understand this further.
Can you please repeat your second question regarding the TNBC topic? So I got it partly, but you're referring to the PFS.
Sure. Yeah.
So the first-line TNBC data response rates sort of 100% at the PD-L1 greater than 10, but sort of the PFS was lower there relative to sort of PD-L1 that were essentially 1 to 10. So just wondered if that sort of is the trend or is this maybe just a factor of being small numbers where you would expect there to be sort of higher PFS and higher response rates in that PD-L1 positive population?
So the data we have shown in TNBC across the different PD-L1 levels. So high response rate, medically very meaningful, and the PFS in the ITT population is 13.6 months, which is compared to the class at which we are really presented at ASCO in a discussion. We are seeing really our numbers superior at the moment.
It is a small sample set, so we should not overstretch here the conclusion, but we have good reasons to be confident that the PFS is not a chance finding because really it speaks to each other, and as mentioned in San Antonio, we will present additional new information regarding OS. This will give you also a glimpse that if all this information is speaking to each other, then it is not by chance.
I get the sign that two other questions.
Hi, this is Na Sun here for Jess Fye from JP Morgan. Can you just highlight to us that beyond the San Antonio breast, what are the key updates coming for 327 in 2025 and 2026, and when should we be looking for them? Then secondly, I think you showed a slide with a bunch of milestones we should be looking for.
Which ones do you think are firmly coming in 2025, and which ones are more likely going to come beyond 2025? Thanks.
Yeah, thank you. Ilhan, do you want to speak to the San Antonio Breast conference data?
Oh, yeah. Yeah, so first on TNBC, San Antonio meeting around the corner. The other marquee indications you saw were small cell lung cancer, which will come most likely with an update in 2025, so next year. Not referring to any point in time, but so we believe some updates will come for second line and maybe as a kind of teaser also, first-line data might be presented in 2025, which we have so far not published, and this will be new information from studies conducted already and ready to go.
Our own Phase III study will start at the end of this year, but the dose optimization is expected to be completed next year. This data from the dose optimization trial for small cell lung cancer and for TNBC is also poised to be presented next year.
Yeah. Aside from the BNT327 data updates expected next year that we just mentioned, which are likely to be myriad, we're also expecting FixVac BNT111 data in melanoma next year. Again, this is the trial that we've already reported out positive result on primary endpoint ORR. We expect that data set to be published next year. We also mentioned that our BNT323, our HER2 ADC, we expect a data set there in endometrial cancer next year, which could provide a pathway for BLA submission.
Further to that, we expect CRC data for our iNeST program likely to come towards the end of the year. We've said consistently end of 2025 or early 2026, but I think so far so good in terms of study progression. Okay. I think we have time for, is it two more questions? Two more questions. Okay.
Hi, everyone. Gil Blum, Needham & Company. Thanks for taking our questions. Just a couple of quick ones on cancer vaccines over here. First, on the cancer vaccine program, is there any way to identify patients who are likely to respond to a cancer vaccine? Have you guys looked at that? And secondly, for BNT211, was there any relationship between duration of response in individual patients and the addition of the vaccine? Thank you.
The first question was, is there a way to identify patients who would respond to the vaccine?
Not really. If we now speak about the individualized vaccine, not really. Because what we observe in the vast majority of tumors and tumor indications is that a large portion of patients shows at least a tumor T-cell response. That means that the mode of action of the vaccine is also activated, which means that we don't really have, at least as long as we don't have clinical responder data in large sample sizes, the opportunity to stratify and ask a question, what is different in this versus that patient population. Once we have large sample sizes and have the clinical responders, we can start correlation analysis in order to understand that. An interesting observation, but this is trivia actually. We have made in the small pancreatic cancer trial because there we have the stratification patients who don't have an immune response.
We can ask a question, why don't they have an immune response? An interesting observation in this patient population was that in that non-immune responder population, which also relapsed earlier, the number of splenectomized patients, so patients who got their spleen removed in the context of pancreatic cancer surgery, we saw a lack of immune response, which supports our not only hypothesis, it supports the mode of action of our vaccine that it goes to these places to lymphoid compartments to induce a strong immune response. The second question was about 211. Was that a correlation of response and positivity for the target? Did I get that right?
CARVac. Does CARVac, I think the question was, does CARVac lead to, do we have any data that shows that CARVac leads to increased duration of response?
Yes, I have shown that data that CARVac leads.
It's small sample sizes because you have to look in the respective dose levels of the CAR-T cell, but we have, first of all, data that CARVac leads to prolonged persistence of CAR-T cells in circulation. And we have some trends which show that this also then translates into response and longer duration trend data.
Okay. Last question.
Yeah. Yeah. This is Manuel from Jefferies. Just one. So in the PD-L1/VEGF of NSCLC clinical trials, do you think the lab-mediated discontinuations versus the clinician-determined discontinuations could make a difference in PFS to OS translation? We see generally like the grade 3AEs higher than pembro in those studies, but the discontinuations are generally low. So how translatable are those treatment discontinuation protocols used in China-based studies to global trials? And also lastly, the median follow-up time in HARMONY-2 was around 8.5 months.
Do you think this was long enough to get a fair view of the temporal pattern of AEs with this approach?
Can you? I think the first, I got the first question I believe was, do we think that discontinuation rates are going to impact OS? Is that correct?
Yeah. The lab-value-mediated discontinuations versus the physician or clinician-determined discontinuations.
So discontinuations driven by elevated lab measurable levels versus non?
Yeah. It's a difficult question. I think what is very clear is that discontinuations are often associated also with immune suppression of patients. They are associated with a higher risk of death. But on the other side, sometimes they are also associated with a stronger response to the tumor. So I do not know if due to the treatment. I would say that in the I have to check publications.
It's a good question, but I don't know if someone has really proven that PD-1 discontinuation due to side effects results in a poor OS for all kinds of other drugs where you have to have a continuous effect by, for example, ADCs. This is associated with a poor OS. The longer you can treat the patients with an active compound, the better it is. In the case of immune responses where you create a memory T-cell response, this might be more difficult to answer. And the second type of question, can you stand up and?
Yeah. The second one asked about the duration, like the 8.1 median duration in the HARMONY-2 trial, was that enough to kind of see a temporal pattern of the AEs in general?
You're asking was eight months enough time to see the AE profile, to get a good sense of the AE profile?
I think this is now the development of the compound started three years ago. And the patients, there are patients who are receiving this compound now for 24 months. So we are not seeing this type of cumulative AE events, neither in the immune-associated adverse events nor in the bleeding aspects. It's really clean. And it is the PD-L1 molecules tend to have a lower rate of immune adverse events. And the one that we are seeing here is, as Ilhan said, really encouraging that this compound could be more amenable for IO/IO combinations because of the low basic AE events.
Thanks.
So what I have presented is really the low discontinuation rate. And in cross-indication, you can see that this is low. And the typical known side effects from VEGF and anti-VEGF treatment, the numbers are also lower.
So that's why my statement was, it's not comparable from the level from the individual treatments and the molecules. So this bispecific reveals a different pattern for safety and for efficacy. I think this is what we can say over the period of the development currently. And the class is providing similar information, I would say, of these drugs.
Thank you. Thank you.
Okay. Thank you very much.
Thanks, everyone.