So good morning, good afternoon for those of you on the live stream. It's my pleasure to be with you here today and to introduce BioNTech's second annual innovation series event. My name is Ryan Richardson, I'm the Chief Strategy Officer of BioNTech. It's a particular pleasure to be here in Boston. Just a couple of years ago, BioNTech set its U.S. headquarters here at Boston, really with humble beginnings, and we've grown our presence here now to almost 500 people, and I think it's fair to say we're really already embedded in the dynamic ecosystem here in the city, which is fantastic. We are gonna be making forward-looking statements today, which you shouldn't put undue reliance on, because future events could differ from, of course, our anticipated events or plans. For a full description of the risks, of course, please refer to our 20-F annual report and other documents filed with the SEC. So we have a packed agenda today, over the next couple of hours. We're gonna start where we left off last year with Uğur diving into the BioNTech approach to innovation before we introduce you to our new the CEO of our AI portfolio company, which we recently brought on board into the BioNTech group this year, to give you an overview of what we're doing on the AI front. We're then gonna go deep in our oncology strategy today, oncology strategy and programs. In addition, we're gonna talk about corporate strategy and the growth path ahead that we see for the company. We're gonna open up the floor at the end of the session for Q&A, and we're also gonna have a break in between, a 10-minute tight break in the middle of the day. It's gonna be sort of jam-packed. We're gonna try to get through a lot of material today, and I think we've hopefully have a lot of interesting content for you all. We're gonna try to end about 1:00 P.M. today. Our speakers, I think some of these folks you know well and need no introduction. We have both our cofounders here with us today, Uğur and Özlem. Very pleased to have them here. We also have, as I mentioned, Karim, who is the CEO of InstaDeep. And last but not least, we've also got two R&D leaders, two vice presidents that are integral to our program development, Ilhan and Michael, who are gonna join us for the second part of the day here up on stage to go through some detail of some of our programs. And without further ado, I'm gonna hand it over to Uğur Şahin. Yeah. Thank you, Ryan. Thank you everyone for coming, and it's a pleasure to welcome you here together with my colleagues, Ryan, Özlem, Karim, and several others. So many of you are following us now for many years, some featuring us, some supporting us as investors, and some working with us. And I would like just to start with something which makes our DNA, our accomplishment. The company is 15 years old this year, yeah, and we made history already by contributing, developing the COVID-19 vaccine. It's... It was the fastest development of any medicine in the history of medicine, less than 9 months. It was the strongest launch of any pharmaceutical product, with more than 4 billion doses shipped in the meanwhile, and this helped to save millions of lives and had an impact of trillions of dollars in global impact. We made history, and I'm saying that because we are not going to stop here. We want to accomplish more. We want to use our capabilities and our vision to make medicines, particularly to help patients with cancer. Even though during the pandemic, more cancer patients per year died because of cancer than because of COVID. So where we are today here, we are market leader in the COVID-19 vaccine space. We are shipping our vaccines still about 40% to low and middle-income countries, and live our healthcare and social responsibility. We have built an innovative pipeline, and this pipeline progressed in the last years. We have now 11 clinical trials in Phase 2 and Phase 3, and our team grow now with about 5,700 employees globally. Globally shows what we mean with that. We have now offices, labs, infrastructure, production units in 5 continents, including recently in Australia, Asia, Africa. And we see ourselves as a global pharmaceutical company. What is important is that the average size of our employees is 36. This is a young team, and this is in line with what we plan for the future to set up really a pharmaceutical company that is built for the future. Our key domain expertise is immunology, and we want to harness the full power of the immune system to fight human diseases. This is relevant because of two key aspects. The immune system is physiologically involved in many, many processes. It's a systemic body-wide organism, and that not only ensures self, but is involved in diseases like cancer and infectious disease. We are heavily invested in developing cancer therapies and infectious disease vaccines, but our interest goes beyond these diseases, addressing autoimmune cardiovascular diseases, neurodegenerative, and inflammatory diseases. These are things that will come later in the development of our pipeline. We expect to have first candidate products here in, in the years of 2026, 2027, starting in these indications. We are a research and development company, and we are focusing on five innovation pillars. This is the, on the one side, the understanding of the immune system. Yeah. The second aspect is about targets and mechanisms. Yeah. We built a multi-platform innovation engine, and I will show you what we mean with that. Yeah. We are now focusing and accelerating our competencies in digital technologies, AI, and machine learning, and we have manufacturing automation innovation. Yeah. We believe all of this, these key pillars are important to build our future company. The multi-technology innovation engine is driven by the idea that we are interested in mechanisms and targets in a technology-agnostic manner. We see that these technologies not only open up spaces, therapeutic spaces, but are also important because we can form synergy, we can do combination therapies, and we can prepare ourselves for the future of healthcare, which will be personalized. And importantly, these technologies are not isolated, but connected with each other. And everything what we have is going to be empowered by artificial intelligence and machine learning, which is some sort of a connecting piece here. You know that BioNTech is one of the mRNA leaders worldwide. We are continuously expanding our competence spectrum, our technology spectrum, and this includes mRNA formats. We are working with different type of mRNA formats, uridine mRNA, which is excellent for inducing T cells, pseudouridine mRNA for inducing antibodies, self-amplifying mRNA, and trans-amplifying mRNA, with a clear vision that we can reduce the amount of mRNA needed for future vaccines, 100-1,000-fold. There are additional new technologies in which we are engaged, circular RNA, chemically synthesized mRNA. We have to deliver these molecules, and we are using different type of delivery tools, nanoparticles based on lipoplex, lipid nanoparticles, but also polymer nanoparticles that come with specific features and should allow us to deliver mRNAs to different type of tissues, organs, and different routes of administration. The third level of differentiation, of course, is the drugs that are encoded with mRNA vaccines, our cancer vaccine program, infectious disease program, autoimmune vaccines. We are encoding antibodies and have IgG antibodies and bispecific antibodies, mRNA encoded into the clinic, in clinical testing. We are encoding signaling molecules, engineered molecules like optimized RiboCytokines, enzymes for genetic engineering, and transcription factors for reprogramming cells. We know that mRNA delivered in cells is able to rejuvenate and reprogram cells. And all this is driven by an additional technology layer, which is addressing multimodal optimization of the potency and performance. And this is work which is ongoing for decades and will also continue to evolve in the next years. It is not sufficient to develop vaccines and to develop mRNAs. We understood from the very beginning that this new technology also requires manufacturing competencies. And manufacturing competencies need to address, on the one side, scale. We have, in Marburg, a facility which has an annual manufacturing capacity of 1.6 million mRNA doses, which is one of the biggest worldwide now. The second aspect, the beauty of mRNA, is tailoring. We built digitized manufacturing of individualized mRNA vaccines with a turnaround time of vaccines in the range of 4-6 weeks. And the third challenge for new technology is global access. We want to democratize the access to mRNA technology, and we know that one limiting factor is building up GMP factories, which could require EUR hundreds of millions of cost and 3-5 years. We have developed a BioNTech containerized modular solution that could accomplish a copy-paste approach to, for mRNA production, with shippable containers and allow to set up manufacturing wherever it is needed. We are embracing the progress in AI. You are all seeing what is happening in the AI field, and we saw that coming for many years before. The trends that we see here is, on the one side, really the increase in computing power, which is dramatic, yeah, and which is going even beyond Moore's Law, yeah, with the recent developments. If you have an iPhone, you have the same storage capacity like a super Cray-2 computer a few decades ago. So this computing power gives us the opportunity to deal with a lot of data. But it's not only the computing power, but also algorithms, which are based on new insights, allowing AI to use AI for prediction of structure, but also for de novo protein design. What is coming next is the use of large language models to support general healthcare, to identify how a tumor will evolve, how a tumor can be classified, to, in real time, record the progress of the treatment of the patient. And we want not only to use these innovations, but we want to be part of the transformation. We want to drive this transformation. This will not only allow clinical progress, but allow us to develop new molecules. The goals for using AI are depicted here. These are some of our goals. AI, on the one side, for drug discovery, lead structure identification, optimization of mRNA, discovery of T-cell receptors, antibodies, our RiboCytokine programs, engineering vaccine scaffolds, identifying variants, yeah, and customizing new antibacterial proteins. This is modular technologies and modular know-how. But the bigger vision is to use AI to bring everything together for personalized medicine, starting from sequence analysis to manufacturing of vaccines. In the last years, we tried to identify collaboration partners. We screened a number of collaboration partners and identified about 3.5 years ago, InstaDeep, a UK-based company, started a collaboration with multiple projects and recognized that InstaDeep is a world-class AI company, which with capabilities beyond what we can deliver ourselves at BioNTech, even though we are very skilled in AI and machine technologies. And we also recognized that what we want to accomplish not only requires quality, yeah, but also scale, yeah. Scale in bringing AI into all our processes, and also speed. Can we do things faster? Can we accomplish to get the same in a larger scale, higher quality, faster? The overall idea is to combine our biology competences and the AI and machine learning competences at InstaDeep. The implementation, how we are doing that, is based on following concepts. On the one side, successful collaboration over the past years, focusing on high-priority projects. What is important is that even though both companies are tech companies, yeah, the culture in AI company is different, and we wanted to keep the integrity of the team untouched, so that InstaDeep is part of BioNTech, but acts as an autonomous engine. So we have here Karim with us, the CEO of InstaDeep, who will show some of the capabilities and to make clear why this is going to make a difference. Hi, Karim. Hi. Thank you, Uğur, for the introduction. Hi, everyone. It's a pleasure to be with you. Yes, there is a lot to talk about, so I'll just give you a sense of our capabilities and a few of the projects we're talking about. So if I get back to the one... Yes. See if this works. We can go to the next slides. Yeah. Yes. This one doesn't work. Yes. So... Yes. Okay, perfect. Thank you, Uğur. Excellent. So our capabilities. So what is important, and this is what Uğur mentioned, is the fact that we want to move faster. Yeah, there is a lot going on in AI, lots of opportunities to develop. I want to give you a sense of, like, the differentiating capabilities that we're bringing to the table. So the first one is the team. We have more than 300 experts, and importantly, those cover what I would call the sort of like the vertical of AI. So basically, researchers inventing new algorithms, but also machine learning engineers sort of using those algorithms in specific use cases. And finally, also, the MLOps, DevOps, the engineers and experts that can deploy those machine learning models at scale, and this is, this is important. So, as you know, there is global competition for talent in AI, and we are established over 10 offices in the world. So we have a presence here in Boston, also in San Francisco, in the U.S. We're headquartered in the UK, in London, with presence as well in Paris, in France and Berlin, in Germany. But importantly, and also we are hoovering talent all across the African continent. So we have a differentiated ability to attract talent, which is, which is very important, given, the competition you see today. And it's not just talent. We also have, supercomputing assets. To give you an idea, by end of 2024, we will be almost at exascale, level, which is, which is interesting. And to give you a sense, this is larger than the Cambridge-1, NVIDIA supercomputer in the UK. Having your own compute allows you to further optimize high- high-performance computing flows, which is, which is critical. And so with the talent we have, with the capabilities we have, we're capable of pushing, innovative research. To give you an idea, we have, like, 25 research papers published this year, which is a relatively high number in AI space at major conferences. So with all those capabilities, we focused on two angles, which are extremely important. One is large language models, and Uğur touched on that. And the second one is large-scale optimization. On both of these, our team are bringing original contributions, pushing the state-of-the-art. For example, language models, we have developed our own libraries, and same thing with optimization. We're bringing in novel methods at the cutting edge of what is happening, open sourcing some of these, and importantly, continuing to innovate. This allows us to really push the frontier of what's possible, and this comes together with, like, software capabilities, simulation capabilities. So in a sense, we have the critical size to do all this. So if we look at how does this impact the work we do at BioNTech, like Uğur mentioned, AI is the connecting piece between all these different, like, sort of, tasks and workflows. So if we look at, for example, what happens from target identification and personalized cancer vaccines to mRNA optimization to gene synthesis and functional validation. So if I look at some of the nodes on which, like, the BioNTech personalized immunotherapy platform operates, we are actually intervening at each one of those nodes, whether it is through simulation assets to basically simulate binding between macromolecules, developing large language models to have a better fundamental understanding of biological macromolecules or developing AI-designed vaccines. So I'll give you a few concrete examples to give you a sense of the exciting work we do. So if we look at gene synthesis, this is a critical component if we want to build like mRNAs, and so we need to first assemble the right DNA sequence. This needs to basically cut DNA in multiple small pieces, fragments, oligos, in order that these assemble in the right format. So, we have developed original methods to so to crack this problem. And at the moment, our, our software and the AI algorithms that power it allow us to... have allowed us to increase accuracy by 36% absolute percentage points. So we went from roughly, like success rate, around 53% in assembling those DNAs into now 89%. So this is significant. We're talking about reducing failure rates here by a factor of five, going from roughly 50% failure rate into something in the order of 10%, reducing failure rate by five. And as a consequence, the same gene synthesis platform is capable actually to delivering 68% more, like, ready-to-use DNA sequences. So this is a concrete example of the work we do, which is directly impacting the platform at BioNTech. And it doesn't end here. If we look, for example, at functional validation, which was one of the points described before, we have automated this task. It used to be done by manual experts, checking if you have, for example, like immune reaction or not. It is a time-consuming process, and as we are scaling the personalized immunotherapy platform at BioNTech, it is critical to add more automation, more scalability to the system. And the work we've done internally, building a visual AI system and embedding it in a software, allowed us actually to increase the accuracy. So we went from a human-level performance of roughly 90% into 98% performance for the AI system. And this is not surprising, because we know that visual AI systems can be more, like, more higher performance than humans. But nevertheless, it makes a sea change, because we not only improve accuracy, but we also actually accelerate and deliver efficiencies. So by building the software components, already experts spend 8x less time. They do not have to move, like, files with pictures and the like. But if you add the AI inference on top, conservatively, it is in the order of 40x improvement. And so an optimized workflow, and this is sort of a taste of the future, is really something where you have automation when AI has high confidence, and the specific cases where a manual expertise is required, then you can process these with manual experts. But we are talking about an order of magnitude improvement in the speed of the, sort of, like this task. And when it comes to manual experts, given that they have a lot less to do, essentially, like, the productivity is massively increased. They can cover so much more than previously. So these are examples of what work we do on every specific piece of the workflows at BioNTech. But importantly, it's not just that. We are aiming to innovate fundamentally in large language models, and to Uğur's vision, you know, there is a lot coming in AI, and I think all of you have heard about generative AI foundation models. Here at BioNTech, we aim to take the leadership on some critical components of that. Here, I'm presenting the work done by the team on the Nucleotide Transformer. So this is currently the state-of-the-art model, state-of-the-art LLM for DNA. So what we have done is using our expertise in large language models to train at large scale, we're talking about 850 species, tokens in the order of 1 trillion. To give you an idea, GPT-3 was trained on 300 billion tokens. We're here at 1 trillion tokens in terms of like genomic sequences. What is interesting is that we see that those models can learn, and that they are capable of being competitive with methods that were specifically dedicated for a task. For example, if I look at Splice AI, which was developed by Illumina, this is a specialized software to detect splicing sites, which is very important in DNA and for RNA generation. What's interesting is that our model, which was trained in a purely generic way across multiple species, multiple human genomes, is actually capable of simply quickly fine-tune on this task to be competitive with a state like a state-of-the-art system like Splice AI. If you look at the bottom right, we are matching in terms of area under curve and top K, the performance of this specialized software. It's the same thing with DeepSTARR, which is here about predicting the activity of enhancer. So building those fundamental sort of like language model is gonna be critical for the future, because one way to look at it is that you are training for a general understanding of genomic sequences, and then you can use this understanding to specific downstream tasks that you care about. And so if you think about splicing, this is important because we know that some like tumor like cancerous mutations and others, there is sometimes a different type of splicing that happens. Same thing in terms of like predicting the deleteriousness of specific mutations. This is extremely important to help identify, for example, like passenger from driver mutations. And so at BioNTech, we are building those large-scale tools that allow us to cater to multiple, like, questions and provide data-driven answers and continue to scale basically data and insights in the future. And so our goal is to continue to be at the state of the art in terms of these systems, but we do believe this is a very exciting moment because we now have the tools to understand macromolecules in biology in a way that was never the case before. When you think about genomics, you know, like the exon part is a few percentage points compared to the whole sequence. Until now, it was extremely hard to understand the rest, and the new tools from AI, large language models, self-supervised learning, allow us to make tangible progress. So this is, in a nutshell, a few examples of what I wanted to share with you. And perhaps the takeaway is that we're building both fundamental innovation at the cutting edge of the industry and with critical size and accelerating, but we're also focused on delivering precise, efficiency improvement to BioNTech's pipeline, like we've seen on some of the examples. So now I'll pass back to Uğur to deep dive into the biology. Thanks a lot. ... Thank you, Karim. So we will now go to the next chapter, our oncology strategy. Let me start with stating a basic underappreciated challenge in oncology. We all know that oncology is or cancer is one of the biggest challenges of mankind, and the root cause of cancer treatment failure is depicted here in two aspects. One challenge is that every patient has a different cancer. And the second challenge is that even within a tumor of a patient, every cancer cell is different. The reason for that is that cancer is a disease by sequential acquisition of mutations. These mutations happen randomly, so that means the random accumulation of mutations results in a situation that every patient has a different mutation set. So if you compare two lung cancers, the overlap of the mutations is less than 5%, and this makes things complicated. Regardless what we do, many treatments are successful. We see most often in highly successful treatments that 99% of the tumors shrink, but the 1% of remaining tumors go up, and then the tumor is resistant. This is the fundamental problem in oncology, and we want to address that. We want to address that by, by delivering solutions and trying to understand the biology and connect that to solutions. What we want, what we want to accomplish is not to focus on a specific patient population with a specific stage of disease. We really want to address the continuum of cancer patients in the disease, starting with the early, early journey after surgery, but also following and addressing patients with advanced, resistant metastatic disease. We want to bring our therapies to as many patients as possible, yeah, and we want to use the full power of our platforms. The scientific and product strategy for that is that we are building a portfolio with compound classes that have synergistic mechanism of action, and we can categorize that as immune modulators. Many of you call them IO, yeah, but I want to specify that as immune modulators without specific targeting. We have the targeted therapies like monoclonal antibodies, T-cell receptors, CAR T-cell therapies, and a new modality here is antibody-drug conjugates. The third category, and this is something specific for us, are personalized mRNA vaccines. This category is completely different from the other categories because this is the only category where we can use a vaccine to target multiple antigens in parallel, personalized for the patient. And this gives us the opportunity to have a multi-specific attack. We believe that we can use the power of our technology, of what we deliver, only if we really combine it with other modalities. Therefore, BioNTech was built from the very beginning as a technology-agnostic company. We are not interested in a certain technology. We are interested in helping patients, focusing on our customer, which are patients and physicians, and delivering the best what we can do. We are going to do that by selecting molecules that are, that are either completely new or have the potential to be best in class. And in the second step, we want to go into combination therapies. This is a Venn diagram showing how this, this, different classes of molecules work. Immune modulators, the classical example is anti-PD-1. Yeah. We are working on immunomodulators that are just going beyond the classical targets, PD-1, PD-L1 blockade, CTLA-4, agonistic molecules, new checkpoint molecules. The second, second circle are targeted therapies, ADCs, CAR T-cell therapies, T-cell receptor therapies, but also we are open to small molecules if they synergize with our, with our key competencies. And then we have the mRNA vaccines, which come with the opportunity to target, in a polyspecific manner, multiple characteristic features, antigens and tumor cells. None of these approaches, except for vaccines, have the ability to really ultimately cure cancer. We see synergy of immune modulators. In the past, immune modulators with chemotherapy. Now, the future will be immune modulators, for example, with ADCs or CAR T-cell treatments or T-cell receptor treatments. These are the synergy space, and the space for curative approaches will be centered around combination therapies. Our immune modulatory toolbox is composed of several antibodies to which Özlem will go into detail. But this slide should show you that we are targeting multiple pathways. PD-1 blockade, CTLA-4 blockade, CD40 agonism, 4-1BB agonism, CD27 agonism, by using multi-domain antibodies. Combinations between targeting molecules and immune stimulatory molecules, but also molecules that combine validated targets, like, for example, VEGF blocking or PD-L1 blockade. A new wave of innovation, which is coming now in the last years, are antibody-drug conjugates. The concept is known or it was described more than 100 years ago. The first molecules came up around 30 years ago, but these molecules were not optimized. They were activated outside of the tumor. They were not highly effective. They had side effects. With the further development of the linker technology and the toxin technology, now, new generation of molecules arrive in the clinical practice, which are improving progression-free survival and improving overall survival. These are two examples recently, recently getting standing ovations on at ASCO and ESMO. Why I'm showing that? This is not because these molecules are always the solution for everything, but these molecules just show the starting of a new era, and this new era will come, will establish itself in the next 10-15 years. We will see replacement of chemotherapy in all indications. The standard of care will simply change, and we want to be part of this transformation because personalized cancer vaccines will, later on in advanced tumors, will benefit from this type of treatments. We want not only to benefit from innovation, but we want to drive this innovation because the solutions which are now already in the market can be further improved. We have seen in the OS curves that patients have an overall longer survival, but there is a room for improvement. There's not only a room for improvement of the efficacy, but there is room for improvement for the safety and tolerability profile. We are developing, for example, a HER2 ADC, where we believe that this HER2 ADC is not only distinguished by a higher efficacy, but also by a better safety profile, which is important for breast cancer patients. How are we doing that technically? We are screening for a distinguished ADC linker technology, checking for stability, improving the safety profile, and allowing high efficacy. We are looking for novel mechanism of actions for tumor-specific activation, improved novel payloads, which can be used for combination therapies. We are using our own core expertise in targets to develop new ADCs. We have an antibody portfolio, and we are weaponizing these antibodies now with these new ADC classes. For BioNTech, this means we are on the one side, partnering and acquiring new ADCs in the clinical stage, and on the other side, we are developing a preclinical pipeline of ADCs, which will come into clinical testing beginning 2025 and onwards. These are our clinical stage programs, and our colleagues will give you some data. Show you some data sets from early clinical development. The first of our ADCs have now entered Phase III clinical testing, and we expect data in 2026. Coming back to our synergy Venn diagram, this type of treatments, immunomodulators and targeted therapies, could allow us to further increase the response rates in patients. But ultimately, we believe that when we want to go to cures, regardless in the early setting, in the adjuvant setting, or in the late stage, we will need to use mRNA vaccines for polyspecific targeting. The way how we are addressing that is with two vaccine technology approaches. One is the fully individualized approach, which we call iNeST or individualized vaccines or mutanome vaccines. This is based on analysis of identification of mutations and tailoring of the vaccine according to the genomic profile of the patient. We have pioneered this approach, starting already in 2012. We brought this approach now into early-stage cancer, adjuvant stage cancer, and as then we'll go further into the details, what kinds of evidence we have generated, and why we believe that is the perfect domain for evaluating cancer vaccines. Then we have our FixVac approach, which follows the idea of polyspecific immune stimulation, antigen-specific immune stimulation, by providing multiple antigens expressed in the same tumor. This is not a fully personalized approach. This is more a tumor type-specific approach. So we have now multiple clinical trials ongoing, and as you can see, we follow our vision to go into multiple indications: non-small cell lung cancer, melanoma, head and neck cancer, breast cancer, ovarian tumors, pancreatic cancer. This, the first round of clinical trials are based on the use of single compounds, which have a single compound activity, and for which we believe that even a single compound activity will be sufficient to make a tangible difference for the patients. So, these are our ADCs, multispecific immune modulators, mRNA cancer vaccines, and cell therapies. The next question, of course, is how we can combine them? This will be the next wave of clinical trials. So that means 2023, 2024, we are starting a number of clinical trials based on monotherapies. From 2025 on, you will see multiple combination therapies, including also combination therapies with personalized cancer vaccines. We want to accomplish that in this type of development, this Phase 3 clinical studies give us approved products, starting with approvals and market authorizations from 2026 on, reaching in the range of 10 different programs coming to the market until 2030. In parallel, we are continuing to develop our innovations in other fields, infectious diseases, but also the fields of cardiovascular and neurodegenerative diseases, and this will become relevant for our long-term vision from 2030 onwards. Our vision remains to change the way how cancer is treated. This is a slide that some of you might have seen during our going public, yeah? It didn't change. It's the same strategy. We see the future in a personalized medicine way. We have the inter-individual variability on the right side. Every patient is different. We have the opportunity to get clinical samples to analyze that, and that we can do that faster than ever. We are building an armamentarium of molecules covering mRNA therapeutics, engineered cell therapies, antibodies, ADCs, small molecule inhibitors. And many of these molecules will be off-the-shelf drugs, because cancer therapy in future will also, of course, continue to have off-the-shelf drugs. But the personalization level will help us to move from prolonging survival, prolonging PFS, to improving, improving cures. And on the left side, you see why we are interested in AI, because the connecting factor for everything to make this vision really, really real is based on AI. And this slide did not change five years ago, but five years ago, it looked like a utopia. Yeah. Now it's much more tangible, and in a few years, this will become the standard of care. Thank you. Okay, so I'm going to shift a little bit, and zoom out for a few moments, and walk you through our growth strategy at the company level. I want to start by saying that, we, we feel that we, we have all the ingredients at BioNTech to put together a truly historic growth trajectory over the next 10-15 years. We think we have the technology, the innovation, the talent, the vision, and the financial resources, to make that happen. When we think about the growth pillars in the next stage of the company's, development, we see three key pillars at a very high level. The first, of course, is our COVID-19 vaccine. The second is our immuno-oncology pipeline, which we're going to really focus on today. The third is our infectious disease pipeline. While still early, still, we think, going to become an important pillar of growth. I want to talk about these each separately a little bit because they have different business models at this point in the company's development, and different aspects that are going to contribute to that overall growth story. So starting with COVID-19, our strategy here is quite simple, actually, and that is to continue to drive leadership in the fight against COVID-19 by leveraging our innovation power and Pfizer's global infrastructure. I'll talk more about that in a couple of slides. On the immuno-oncology front, here, as you've heard Uğur and allude to, and we're going to again expand on this, we are really building a diversified portfolio of products. Our vision, our strategy here is to build a fully integrated oncology company. That means discovering, developing, and commercializing therapies, on our, on our own and with partners. Then lastly, infectious disease. Here, our vision is to advance a pipeline while early today, advance a pipeline and broaden it, and we think, again, that could become an important driver of future growth in the next couple of years. Starting with COVID-19. I think our fundamental starting point here is that we believe that COVID-19, there will continue to be a need for COVID-19 vaccines, most likely on an annual basis for the foreseeable future. We think this is going to be a long-term business, and that's driven by the continued evolution of the virus. It's driven by remaining risks, in particular to at-risk populations, vulnerable populations, the elderly, that we think are going to persist, and we see that still in the hospitalization numbers and in ongoing deaths, unfortunately, attributable to COVID-19. We think it's also supported by accumulation of evidence that suggests that follow-on booster vaccination provides a benefit against the long-term sequelae of COVID-19, including long COVID. And we've seen now this pattern over two years. We're in the second year of this annual boosting and we have our XBB.1.5 vaccine, of course, that's being distributed now around the world. But we now have a precedent, and we think that that trend is gonna continue, like I said, on an annual basis going forward. Now, one of the defining features of our COVID vaccine business, which I think really is different from others, is the economic structure of this business. I think the first starting point I'd say is that, first of all, we have to recognize this is a very global business. It has been global, of course, during the pandemic. It continues to be global today. We've distributed COVID vaccine to more than 100 countries. We've just rolled out the XBB vaccine to dozens of countries, over 40 countries and regions, and we expect it to remain a global business. But at the same time, when we look at BioNTech's footprint to maintain and even grow that business in the future, we actually have a very limited infrastructure that's needed to support that global business. You see here a global map, in all countries except for China, we commercialize our COVID vaccine alongside our partner. What's notable here is that actually only in two countries do we have sales and marketing capability that we have to deploy, and you see those highlighted here, Germany and Turkey. Everywhere else in the world, Pfizer is responsible for commercialization, and that means that we've been able to keep our sales and marketing capabilities very lean. You see here in Germany, we have approximately 55 FTEs in our sales force, and we've had costs roughly year to date, sales and marketing expense in total for the COVID vaccine of about EUR 45 million. That's a completely different ballpark than, I think, what you see from other peers that are operating at scale in the space. And just to give you a few more economic data points to illustrate the point, it extends beyond just sales and marketing. So the economic structure, again, of our relationship with Pfizer, we think, will make this a truly differentiating and profitable business for the foreseeable future. You see some of the data points here. So gross margins. Generally speaking, over the last 3 years, we've maintained gross margins above 80% on COVID-19 specifically. And again, sales and marketing expense, extremely lean, an average of about EUR 60 million a year over the course of the same period, 2021-2023. Even on the R&D side, where we are investing in innovation and will continue to invest, the overall percentage or the contribution of COVID-19 R&D spend as a percentage of our total, as a component of our overall R&D spend, has remained well below 50% generally on average, even though there has been some fluctuation between 25%-45%. Again, that's due to the fact that we share R&D expenses with Pfizer 50/50, just as we share gross profits. As we look to the next couple of years for the COVID franchise, I mean, there's a couple of features that I'd just like to highlight here to keep in mind. The first is that we've already largely reset our manufacturing base for COVID-19 to serve the future endemic market. That has meant downscaling or downsizing our network of CMOs and partners that we use, that we used to initially get to 3 billion+ doses with Pfizer. We've now downscaled that to a more fit-for-purpose capability. We're expecting over the next 2 years that a variety of shifts will continue to take place. That includes shifts to commercial model and commercial pricing for the COVID franchise. That'll be gradual, and it'll be incremental, and it'll be geography specific. And in addition, we expect continued shifts to single-dose vials and prefilled syringes, and again, this will be geographic specific. We're already serving the U.S. market with a combination of prefilled syringes and single-dose vials, and we expect other countries over the next couple of years to follow suit in that shift. Then finally, as we start to look at 2025 onwards, we do see potential for increased vaccine uptake if combination or next-gen vaccines are successful in the clinic, and we do think that could be a growth driver for the franchise. Even as we proceed through this transition period, we expect COVID-19, the product franchise, to remain highly cash generative, again, due to the economic features that I outlined. So turning to oncology. You know, when we started to emerge from the COVID pandemic about a year and a half ago or a year ago, we had, I think, an industry-leading early-stage oncology pipeline in terms of size, breadth, innovation level, but we lacked a late-stage pipeline. So we've taken steps, both organic and through external innovation, to try to rapidly bolster the late-stage pipeline and help bridge to the vision that Uğur described in terms of building the type of diversified company that we're aiming to build. You see here the start of that. We have a number of Phase III trials that have been initiated this year. We also have a couple of Phase II trials that have registrational potential if successful, and we're gonna... You can see that it's diverse across modality. You can see that by the color coding here. You can expect this trend to continue rapidly over the next 12-24 months as we build out the late-stage pipeline at BioNTech. One of the ways that we're doing that, and actually doing it more rapidly than we could have done on our own, is through partnerships. I think it's important to note that partnerships have been an important part of our model for a long time. It didn't start with the Pfizer collaboration. Actually, pre-Pfizer, we had 50/50 partnerships with companies like Genmab and Roche Genentech on specific modalities. We built on that model, obviously, with Pfizer, and a feature of that was that we tended to share cost, R&D costs, and also draw from partner capabilities in the development, mostly late-stage development, but also future commercialization of these drugs. At the same time, though, we preserved our rights to commercialize in the major markets like Europe and the United States. That was a feature that already goes back to the deals that we did in the 2015-2018 timeframe. What we've done since that time, this year, has been very active in building out and expanding our list of partners, and you can see on the right-hand side of the slide, that we've done that. We've kind of shifted models. We've done that typically with more innovative, younger biotech companies, some of which have just been founded in the last 3-5 years, many of which have already brought forward their lead assets into either on the door of Phase III trials or already into Phase III trials. We're really building this innovation ecosystem around BioNTech, and that activity is gonna continue. What we've done so far in 2023 is bring in-house seven clinical stage programs. While some of the terms differ by collaboration, generally speaking, we're sharing costs, drawing from partner capability here in development, with an aim of accelerating development towards the market of these assets. Generally speaking, for these new collaborations, we've also retained even greater commercial rights on the back end, typically here in the form of global commercial rights for BioNTech outside of Greater China. And, of course, we've supplemented this partnership strategy also with more of some of the more traditional funding mechanisms for the global health portfolio. We've done a deal recently with CEPI to partially fund some of our global health projects, and that builds on an earlier collaboration that we did with the Gates Foundation. InstaDeep, of course, was a unique acquisition that we did, highly strategic acquisition that we did, as we've outlined already today. So I think as we look across the portfolio, I think it's important to note now that as the portfolio really scales, that we're taking a very active portfolio management approach. So a couple of principles are guiding the R&D investments that we're making and looking to make in the near future. The first is that we're prioritizing late-stage programs, and we're gonna do a deep dive in the rest of the session today on these late-stage programs with the aim of bringing them to market. And as Uğur mentioned, our goal is to have 10+ programs in oncology approved by 2030, and of course, to get there, that starts with initiation of more pivotal trials. One example of this strategy is that we plan to have at least 6 different programs in 10+, potentially pivotal trials by the end of next year. Now, again, some of those have already started, so these are not all new trials, but this just highlights the scale and scope of our ambition in the near term to build out the late-stage pipeline. Second, we're gonna continue to access external innovation to complement growing internal innovation and investment as well, and we're gonna continue to do so in a capital-efficient manner. I think the best translation point here that I would highlight is these 7 clinical stage programs that we've brought in-house, 2 of which are now in Phase III, are about to go into Phase III trials. We've done that with only approximately EUR 500 million upfront capital commitment. So we've really made efforts here to be capital efficient and to try to bring in-house assets that we think could be game-changing or transformative to the company without having a, a massive upfront impact on our balance sheet. And I'll come to that a little bit more in terms of the, the importance of the balance sheet and how we, how we view that. And then finally, as we scale the, the portfolio, we are implementing ever-increasing rigor in our late-stage development and portfolio decision-making. I think it's fair to say we've always been extremely rigorous in early-stage portfolio decision-making, and that started with extremely rigorous preclinical work, which led to a very high percentage of our Phase I oncology assets throughout the history of the company, being successful in showing single agent activity well above industry standards, and we hope to continue that track record into the late-stage development decision arena. So it includes, in addition to the preclinical rigor, of course, demonstrating, where possible, single agent activity for early-stage oncology compounds, even where we anticipate bringing them into combination studies. So that's gonna continue to be a hallmark of how we do development. The overall aim of this, of course, is to continue to generate a very high return on R&D investment. We think our track record speaks to that already, but we aim to continue that as we scale the portfolio and pipeline. All right, so we talked about select oncology programs that have the potential to be among those Wave 1 oncology launches in the 2026-2028 timeframe. Here are a couple of those assets. Again, we're gonna go through these in much more detail, so I'm not gonna do that here. I just wanna point out a couple of features now about this late-stage portfolio. The first is that we've got already a diverse set of technologies, platforms, and mechanisms of action depicted here. We have cancer vaccines, we have antibodies, we have ADCs, we have bispecific antibodies, and we have a cell therapy here, all of which have shown single-agent activity and all of which are either in late-stage trials or about to go into late-stage trials. We also have a mix of partners here, again, reflecting the partnership model that I talked through. I think the important point there is that as we think about potential commercialization, what this already implies is that we're very likely to be commercializing not only alone, but also alongside specific partners. And I think that's... as that's been a hallmark and an asset for us in COVID-19, I think that's gonna continue into the oncology arena. Even though some assets we will bring to the market ourselves, we will continue to leverage partnerships. Here's just a little preview of our thinking about building the commercial front end on oncology. More details to come in a future meeting, but in a nutshell, our plan is to build a dedicated oncology sales force to commercialize this first wave of oncology assets. Our thinking is to focus on major markets, as I said, of U.S. and Europe, and selected other geographies. We will leverage partnerships, our partners, as I mentioned, and their capabilities as we grow into our own capabilities. Because we can have the luxury of not having legacy infrastructure and assets, we can really tailor this using the latest technology, digital enablement, to create a lean, highly potent sales force that really is strong in the areas of medical communication and MSL engagement. To fulfill our ambition to launch products starting in 2026, we aim to be commercial ready in 2025. So I think we've gotten to the break period, so I think we'll take a, we'll take a quick 10-minute break. I think we're just on schedule, actually, and we'll reconvene, pick things up with a deep dive into the programs in, at 20 after. So I would just continue at from that point where Uğur stopped, namely talking about our pipeline with a special focus on our more advanced and prioritized products, where we think we have the highest probability of success of delivering licensable compounds within the 2030 horizon. And I have help here with me. My VPs, Michael and Ilhan, will support this presentation here. So you have already heard from Uğur that we have complemented our multimodality pipeline with next-generation versions of two of two additional modalities which have transformational potential. One is IO agents, and we all know that IO, in particular, the first-generation IO compounds, PD-1, PD-L1 axis blockers, have already transformed the oncology space. They have revolutionized across indications, across treatment lines, and have become modalities. There are many who still regret that they have not moved earlier and with much greater boldness into the anti-PD-1, PD-L1 area. The new generation is being explored. These are the anti-TIGIT, anti-LAG-3, and other compounds, which have to be added onto this anti-PD-1, PD-L1 axis. Already the next generation is at the horizon, and we are, in particular, interested in this next generation. These are, these include agents which are converging multiple proven modes of action of immune modulation into one molecule, including bispecifics, which with one arm antagonize PD-1, PD-L1 axis pathways, which means that they probably don't need to be added onto the PD-1, existing PD-1, PD-L1 backbone. So this is, these are agents from which we believe that they will transform the oncology space and will become backbones with which every one of us has to combine or against which every one of us has to compare their products. And this is also the reason why we not only want to be in this space, but also shape this space, as Uğur has already pointed out. The second modality are ADCs, and you have already heard that the co-evolutionary maturation of toxin and linker technologies has poised this modality for transformation as well of oncology space. The targeted cytotoxicity of next-generation ADCs is much better. All of them have bystander killing effect, which means higher potency, which also means that lower target expression tumors can be addressed, which again increases the market size. And also this is an area where we want to enrich our pipeline and have done so. With regard to the ADC portfolio, we have licensed in four clinical-stage ADCs with broad, yet minimal overlapping indication opportunities. That means the targets are expressed, as you can see, in a complementing way. Our frontrunner is an HER2 ADC, which is in a Phase 3, at Phase 3 stage. We also have compounds in Phase 1, 2 stages, directed against Trop-2 and B7-H3. These 3 compounds are partnered with DualityBio, and we recently have also licensed an HER3 compound, partnered with MediLink. We not only select these compounds based on their targets and their coverage of different tumor indications, but also, with regard to their potential to differentiate from other compounds in this modality. And we think that for several of those, we can differentiate based on the safety profile, which again means opportunities to-... Move these compounds, these ADCs, in combination with IO or other compounds from our pipeline into neoadjuvant and frontline settings. Both modalities, IO and ADCs, are highly interesting for combination for combinations based on their mode of action. IO-IO tandems allow to synergize different immune modulatory functions. ADC-ADC tandems allow to increase the patient population by dual targeting, and also target heterogeneous tumors. ADC cancer vaccine combinations, for example, which are also very of high interest for us, are interesting because they not only synergistically work on the Kaplan-Meier curves of survival, ADCs having an early effect on survival, whereas cancer vaccines have a late effect on remaining cancer cells. But ADCs can also support immune modulatory functions, for example, for irinotecan, which is basically the systemic cousin of our ADCs, which are all topoisomerase inhibitor-based, is very well known to modulate immunity, to deactivate Tregs, for example, and that again is very synergistic with cancer vaccine modes of action. With this, I would go into concrete projects and hand over to Michael, who will start with our front runner, with HER2 ADC. Thank you, Ozlem. Happy to be here. I think Ryan called this section the deep dive. I would call it the scratching on the surface, 'cause we're doing quite, quite a few things here, and probably leave you with a few questions that I'm happy to answer in the break or when there's Q&A. So the first one is, as already alluded to, BNT323, also known as DB-1303, comes from DualityBio. DualityBio is an interesting company. It was built or founded in 2020. This was their first asset, and here we are, three years later, discussing an achievement from them. So, BNT323 is targeting HER2. As already alluded to, a Topo-1 payload. The DAR, the drug antibody ratio, is 8. And, in the clinical stage, you heard now a lot, we're in Phase 3. I'll show you the Phase 3 study in a moment. We haven't yet FPI, but we're expecting it very soon. And we have several, Phase 1, 2, plans and, things that we're building actively in terms of combinations. This is just a brief overview over, the ADC landscape in, Enhertu. You're all aware of these, ADC, well, classes, the oldest one being Mylotarg, which was, really one that, where basically the linker fell off, already in the vial. Then Enhertu, it's the most advanced one. In the middle, there's Kadcyla, which has a DAR of 3.5, and perhaps not the most ideal linker and toxin. Now, why do we go after a seemingly similar asset than Enhertu? Well, we think it's actually similar, but better, right? So there's several features which makes it better than what we think Trastuzumab deruxtecan can or is doing. One is already expressed here. You see the higher dose that we're able to see in patients, but a few others are shown here. And if you focus just on the top gray box here, the in vitro plasma stability, this is really one of the key features of this molecule. So basically, in both in the vial, which is also happening, that the toxin can fall off. But more importantly, in the circulation, this is about 20% or so more stable and in the circulation than the you know, potential comparator Enhertu. Why is this important? Well, most of the toxicity from ADCs comes from free drug, right? It does not come from the drug that is released, but most of the toxicity comes from the drug that is in the circulation after injection. On the lower right-hand side, you see another feature, and that's a short half-life. So once the drug is released, it gets takes less than 48 hours to do what it's supposed to do. It's a highly potent toxin, which is totally capable of killing cells during that period, but then it doesn't stay longer because it gets degraded faster than the other top one that you see here. So does this translate into efficacy as well? These are still non-clinical data, in vivo data, and you basically see in these two diagrams, in a HER2-positive and Enhertu-low model, that given this that when this drug is given— You see a pretty rapid deceleration of the tumor size in both models. And for us, this was mostly important for the HER2-low, as obviously this is evolving a new market for ADCs, and this just shows a beautiful, you know, in vivo data that it does the trick. Now, on the toxicity side of things, again, preclinical data point in the direction that we might have a really differentiated safety profile as well. There's two or three things that stand out. One, or perhaps the most important one, is what's commonly referred to as pneumonitis or Interstitial Lung Disease. Hard to quantify in the cyno monkeys, but we saw less than that of a potential comparator. Also what was not shown in the cynos, but then I'll come to this in a second, is also that we may see lower alopecia rates and perhaps some other features I'll come to in a second. So some of these are clinically really meaningful in this patient population of mostly women with deadly cancer. And we'll need to see if they pan out, but from now it looks pretty promising. We did a fairly standard 3 + 3 dose escalation study in a variety of tumor types, obviously all expressing HER2. The dose escalation part was around 90 patients, but by now, this study has reached around 290 or so patients in total. And you see on the right-hand side, the indication variety of HER2-positive and HER2-low patients with breast cancer, non-small cell lung cancer, endometrial and a few GI cancers were done. And these are the data that have been published on safety, and I'm sure most of you have looked at them. From our point of view, this looks quite promising. So on the one hand side, we were able to dose this quite a bit higher than the trastuzumab deruxtecan. We didn't find any dose-limiting toxicities, or we're landing at probably the 8-milligram dose for most of our studies now, and we didn't see anybody dying through a side effect. ILD, the preliminary hope was actually confirmed in this early study. We'll yet need to see how this pans out in the Phase 3, obviously, but for now, we're quite hopeful that frequency, but also severity of the ILD may be less than what others have shown. We do see the smattering of other things like neutropenia and things, but nothing of big concern. Efficacy-wise, again, Phase I data quite encouraging. These are mostly breast cancer patients. This data was presented by Katie Moore at ASCO. You see around 40%-50% response rate and quite some durability here with most of the patients actually getting into a disease control state. Which is nice considering these patients are pretreated, right? This is not frontline treatment, obviously, and they're in a metastatic state. Yeah, we do see a dose response correlation here. So the higher the dose, the better it gets, but probably not more than 8 milligrams necessary at this point. Durability is shown here again in these swim lane plots. On the upper side, these kind of more for beige or orange, whatever, lines are the latest ones with 8 milligrams, hence they don't extend so far. This is an ongoing trial. But you get a sense on these lower doses, how far out we now have data in this relapsed population. Now, where does this lead us to? We think in the HER2 low patients is one of the biggest opportunities for this, for this molecule. And this just gives you an overview of what we think the size of this segment is. Luckily, most patients with breast cancer get cured through surgery or through the initial therapy, but those that do not and have HER2 low expression since a year or so are known to benefit from, from the ADCs quite, quite substantially. While there is a label in this third-line space, we and others are now conducting or about to conduct Phase III trials in this segment that is chemo naive, but has been treated with endocrine therapy. This is a design of this said trial, a randomized Phase III trial that is about to start in the next few weeks. It's a fairly standard design, one to one. We use also stratification factors, which probably don't surprise you too much. The design of the treatment is the three-weekly BNT323, compared to standard of care. And you see here the historical comparisons, and we think with 500+ patients, we're well underway to have a positive trial in a few years from now. I said this study is up and running in a few weeks from now. We'll certainly let you know when that happens. So, another opportunity is endometrial cancer, which is one of the few areas where we're not thinking that we're following somebody else, but that we are actually in the lead. It's a quite common gynecologic cancer, and we feel the standard of care is really dismal in this disease, right? Like, basically, patients get chemotherapy, or since a few months, they get chemotherapy plus a checkpoint inhibitor in frontline. Most patients in the metastatic or advanced setting will relapse, and so those would be eligible for this kind of treatment in the HER2 expressing population, which is about half of all patients with endometrial cancer. We have some pretty nice data in endometrial cancer. This is published data. We have actually quite a bit more by now. The waterfall plot, I guess, speaks for itself. 7 or 8 milligram is also the same dose that you saw before. We get to a response rate of roughly 60% or so, with, again, the disease control rate, which is much higher. This would easily beat any kind of chemotherapy that is out there if confirmed. Now we are discussing this data with regulators around the world and have obviously some plans, but we're not quite ready to share. I'd like to stop here and hand over back to you. Sorry. So the next one is also an ADC we are quite excited about, targeting a Trop-2, trophoblast protein two, which is a target which is highly expressed in a wide range of indications, including indications with amplification of topoisomerase, which helps with the payload we are using. Also, this is a topoisomerase inhibitor based ADC. The antibody BNT325, also one of our programs, which is partnered with Duality, is a humanized anti-Trop IgG one with, again, a cleavable linker. This is to compare with a known and also approved Trodelvy, Trop-2 targeting ADCs, where we think that BNT 325 compares favorably, again, on the dose side, which we can use according to toxicology toxicity studies in animals. We have some preclinical data here, which also shows that, as already described for our HER2 ADC, that BNT 325 inhibits tumor growth and leads to tumor regression very nicely, not only in Trop-2 high mouse models, but also in Trop-2 low or negative mouse models. And in the latter one, outperforms, in this case, for example, DS 1062. We have a running Phase 1, 2 trial with our partners from DualityBio, which is being conducted in advanced unresectable cancers. The dose escalation part has already been finalized. We are in the dose expansion part, where we are testing this antibody in non-small cell lung cancers with and without actionable genomic alterations, ovarian cancers in hormone receptor positive, HER2 negative breast cancers, and TNBC without prior Trodelvy treatment, or after failure upon Trodelvy treatment. And this trial is still ongoing. From what we see in terms of safety data, we think that we are at a comfortable place. This is data from the dose escalation part, where we have observed that in the dose levels of 2 and 4 mg per kg, we have a very nice and manageable adverse event profile. We see dose-limiting toxicities at the 6 mg per kg dose level, so that the 5 mg per kg dose level has been determined as maximum tolerated dose. We have observed one ILD in these 44 patients who have been treated in the safety set, and no treatment-related adverse events led to death. In terms of efficacy data, this is a data set from 23 evaluable patients, where you can see that we get, in particular in those dose levels, which I already highlighted, very nice change in target lesion diameter. Objective response rate is in across all those 23 patients, 30% with a disease control rate of 87%. And if we focus on the subset of patients with pre-treated non-small cell lung cancer, 30 patients, we can see that the objective response rate is 46%, and disease control rate of over 90%. So to summarize the key takeaways for our ADCs, you have seen that our BNT 323 program is very advanced. We are initiating a Phase 3 trial in breast cancer, HER2-low. We have ongoing dose expansion cohorts in different cancers in which we see interesting efficacy data, and we'll pick from them for the next pivotal studies. And for the earlier stage ADCs, including the TROP-2 ADC, which I have shown, but also our HER3 ADC and our B7-H3 ADC. We have ongoing dose ranging studies and indication testing studies, which will also lead to data-driven decisions for trials with registrational intention. With this, I would move to our immune oncology targets, which our immune oncology agents, we have an entire pipeline of them built in the meantime. We will hear about our anti-CTLA-4 from Ilhan in a... No, from not Ilhan, from Michael, right? In a second, and Ilhan will present our front leaders from our Genmab corporation. With this, Michael, it's again your turn. Careful there. Yeah. CTLA-4, you heard the word registrational before, so this is actually the registrational study that is ongoing. But before I show you the trial design, I'll walk you a little bit through why we have partnered with OncoC4, which is a small U.S. biotech, which also has an interesting story, but maybe that's for another time. So basically, there's Ipi and Tremi, of course, out there. CTLA-4, arguably the second most important checkpoint that we've discovered by now, next to PD-1, PD-L1, of course, and but probably more important in blocking this checkpoint than other things like LAG-3 or TIM-3 or all the others. But why isn't it such a big success? Arguably, it is a success, but maybe not the biggest one. It's usually because of tox due to thyroiditis, due to GI tox, namely, and all the other immunotoxicities that happen. So we decided to partner with OncoC4 because this molecule has at least the promise to have a broader therapeutic index than the other two, which obviously would then enable it to be better combined with both other checkpoints, but also with the other modalities we have in our pipeline. Why is this likely to have a broader therapeutic index? It's depicted on this slide. It's basically what happens when the antibody binds to CTLA-4 in the cell surface, the complex of antibody and antigen gets internalized. What happens with non-pH-dependent CTLA-4s is that they travel to the lysosome, get degraded, and are gone. What happens here is that the molecule was designed to be pH sensitive. So once the pH is below 7 or 6.5 or so, which is the case in the cancer cell, or actually in any cell, the two parts, antigen, antibody, separate and get recycled. So both the antibody gets out of the cell, again intact, and the antigen also gets expressed again at the cell, which enables then the antibodies to again attach to the cell. And if they find the corresponding T cell to destroy the cancer cell, then that could lead to a higher degree of on-target efficacy versus other CTLA-4s. So that's kind of the story. It's pretty well documented in the preclinical experiments that you see here, here listed, and we see also early evidence that this might actually work. So again, a version of a 3+3, you've seen now a couple of times during this presentation, was also done here. What's different to the other 3+3s is that this was a quite extensive study of dose, and also an extensive study of indications. So for monotherapy, panc was looked at, various versions of non-small cell lung cancer, head and neck, triple negative breast cancer, ovarian, several others, actually. This study has by now enrolled over 450 patients, most at the targeted dose. Plus, there's a variety of combinations with Pembrolizumab that also have yielded some data already. What we see in terms of dose is that we can dose this molecule much higher than Ipi or Trem. You know that Ipi is approved at 3 milligrams, mostly used at 1 milligram. And here we're talking about 6 milligram as a standard for combination, and perhaps even higher when used as monotherapy. So that alone shows that we will be able to differentiate. Whether this works in all combinations, we'll have to see. So we're still actively studying the dose, also in the Phase 3 that you'll see in a moment, but for now, we're quite confident. In terms of efficacy data, this Phase 1 has yielded quite a variety of data points already. Some of the published data is shown here in ovarian on the left-hand side, in a variety of tumor types, with Pembro, and on the right-hand side, the melanoma data, again, in combination with Pembrolizumab. So quite encouraging. Perhaps not surprising data, in the sense that the drug is active, and a little bit too early to say that this is differentiated enough in all of these indications, but for now, we're pursuing all of them further. Coming to non-small cell lung cancer, which is the pivotal program in this CTLA-4 development plan. This data has been presented by Kai He at ASCO, and a few days ago, actually, at SITC in an oral presentation. Which was quite interesting to see in that we do see single agent activity here at dose. The dose here is 10 milligrams followed times two followed by 6 milligrams resulting in about a response rate about 30%, but a much higher disease control rate. With some of these patients, you see that on the right-hand side, actually improving over time in this, in these spider plots. This is an example of what happens. This is a patient with non-small cell lung cancer, with squamous cell carcinoma, who had several disease sites, metastatic sites. You see here two spots in the liver and a large lesion in the spleen. This patient was on drug actually for quite some time, which is what very, very unusual for CTLA-4, because most patients cannot tolerate this CTLA-4 over the course of a few months, and this patient is over a year now on treatment. You see that over time, the liver lesions disappear, and the spleen lesion get shrunk to a very good PR. So again, where does this lead to? Where are we going with this? Non-small cell lung cancer, our key indication, our first indication for Phase 3. As you all know, this field is changing rapidly, with IO platinum based chemotherapy firmly established as a standard of care, but also firmly challenged with several other triplet combinations right now. For second line, Docetaxel is still around here, but the Trop-2 are scratching on that, you know, avenue. And we think we can probably deliver with the Preserve 3 study that you see in a second, also an alternative, definitely better than chemotherapy. This is a study design. It's a two-stage design, 600 patients. The first stage is a dose confirmation part, which is a bit unusual in a Phase 3 trial, if you think about it. But we had discussions with regulators that demanded us, despite having quite a body of data on dose, but no randomized body of dose data on dose, to do this in a two-stage fashion. This study is now well underway, so we're in the recruitment projections. Study has begun around, I believe, June or so, and we're quite hopeful to deliver results, you know, when the time is right. So this may take two years or a little longer. With this, I'd like to close and hand over to Ilhan. Good morning. Pleasure for me to present the already mentioned, collaboration and partnership with Genmab products, which are our bispecific, antibodies, which you see on the screen. For the sake of being focused on the more mature programs, we will present to you today, mainly it's BNT311, which is our PD-L1 x 4-1BB bispecific, and, a little bit later, the BNT312, which is our CD40 x 4-1BB. I start with the PD-L1 x 4-1BB. A few words to the mode of action and the rationale. Conditional bispecific molecule for two validated targets, PD-L1 and 4-1BB. PD-L1, known as receptor ligand expressed on tumor cells, that inhibits proliferation of PD-1 positive cells and has a role in immune evasion. The 4-1BB is a co-stimulatory tumor necrosis factor expressed on T and NK cells. Activating the 4-1BB pathway enhances T-cell proliferation, T-cell effector functions, and prevents T-cell death. Insights and information about the preclinical data for this molecule. And as you can see on the left-hand panel, this is an assay, an in vitro assay, indicating that BNT-311 blocks the PD-1, PD-L1 axis in the absence of 4-1BB binding. So this is, in fact, in a nutshell, showing that the PD-L1 specific Fab fragment is working and functioning as a classical immune checkpoint inhibitor. On the in vivo side, the right-hand panel, as you can see, BNT-311 is exhibiting also anti-tumor activity in vivo. The upper panel is a typical tumor volume experiment in tumor-bearing mice. These are double knock-in mice for human PD-L1 and for 4-1BB, and you can see that GEN1042, so BNT311, is leading to complete remission in all animals in this experiment, 9 out of 9. Compared to the controls, you can see that there is high volume increase in the tumors over the whole period of experiment, within 2-3 weeks even. The lower part is a kind of surrogate for the PFS in humans. So this is an experiment showing the percentage of mice with tumors below or highest 500 cubic millimeters. This is a surrogate, as mentioned, really, for progression-free survival. And you can see mice treated with BNT311, they all survive, and the tumor are not growing. The control group is leading stepwise, really, to death of the mice over the period of time. So a few words for our first human data, dose escalation. So this is the scheme, Phase 1, in 61 patients treated with BNT-311, IV flat dose Q3W until PD or unacceptable toxicity. So in this dose escalation part, the recommended Phase 2 dose was identified, so 100 milligram per Q3W. And then we went into the expansion Phase, which is a Phase 2 part here, so in different tumor indication, lung cancer, first-line in monotherapy, plus Pembro, in squamous, non-squamous, but also in PD-L1 inhibitory, pretreated cohorts, cervical, endometrial, and so on. So this is still ongoing. So the study is recruiting 13 expansion cohorts in total. We are collecting data here. Each cohort can recruit up to 40 patients. We are, of course, also collecting follow-up data, and with that, we have already some observation from the dose escalation part, first in human. You can see this is a safety description here. Any treatment-related AEs on the left-hand side is in the range of 70%, which is not unusual. Any grade three and higher are around 20%-30%. Also, very similar to other immune checkpoint inhibitor data in this area. Mainly, what we observe are liver enzyme increases, which are reversible. All grades around 20%, grade three or higher, around 8% and 3%. So in the dose escalation Phase, BNT-311 demonstrated a manageable safety profile and preliminary clinical activity in a healthy, treated population with advanced solid tumors. So the disease control rate in these patients was around 66% of patients at a median of around 3% follow-up time. We observed four early partial responses in triple-negative breast cancer and ovarian cancer, and two in CPI pretreated NSCLC patients. So this is an interesting signal, which led to some further evaluation of this population. Monotherapy-treated NSCLC patients, 25 patients indicated in this waterfall plot, and you can see these 25 patients were evaluated regarding their PD-L1 status. So the dark green is indicating the negative, the light green is indicating the PD-L1 positive population. And you can see that more patients with a PD-L1 positive status benefited from this treatment, from the monotherapy. Seven out of eleven, as indicated in the gray bar above the waterfall plot on the right-hand side. So this could be interesting and is considered to be maybe an element for further studies to be selective, maybe on the PD-L1 status. But you can also see that there are also PD-L1 negative patients benefiting, not in that extent, but there is some signal at least. So, for BNT-311, two studies are planned and ongoing. So these two studies are ongoing in NSCLC. This is in relapse refractory, second-line-plus, PD-L1-positive patient, prior treatment with PD-L1. So this is a three-arm study in a randomized fashion. We are testing monotherapy versus this known schedule of Q3W, versus another or different schedule, which is a Q6W schedule, so longer period of time in between. And the question is, of course, so do we see different results here, comparing these three arms? Recruitment is open, first patient dose in December 2021. We are collecting data, safety part completed, and the expansion part is currently ongoing. Too early to report any data at this point, but this will come up in the next year. Endometrial cancer, this is a new study which was initiated in August, September this year. The first patient to be dosed is projected for mid of November, so just around the corner, so to say. We will have the first patient in, and this is a trial comparing in two cohort, CPI-naïve and CPI-experienced. So the next steps for BNT311 are the engagement with health authorities on the design of a pivotal trial in post-IO non-small cell cancer patients, and to present data in one of the upcoming conferences in 2024. Moving now to the other bispecific antibody based on the collaboration with Genmab. So we are using the DuoBody platform from Genmab, but here in this case, this bispecific is targeting really two stimulatory immune stimulatory targets, so anti-CD40 and anti-4-1BB. Also here, again, a few words to the mode of action. DuoBody conditional dual agonist molecule for the preclinically validated targets CD40 and 4-1BB. So CD40 stimulatory receptor primarily expressed on antigen-presenting cells. Engagement of CD40 leads to activation and maturation of APCs. 4-1BB is the same as for the molecule before. There is some activation of 4-1BB pathway enhances the function of the immune system. And important to say, both molecules are engineered in a way that they have an inert Fc part, and this avoids really effects like ADCC and CDC. So what we are seeing here is purely the effect of these antibodies regarding targeting the 2, 2 targets they are made for. Also here, a little bit of preclinical data to share with you, and what you can see here, these are reporter assays and very nicely written, comprehensive paper is cited here. Please, for further details, go to this paper. And so in summary, what we can say, that these 2 targets are conditionally dependent of each other. So in the absence of 4-1BB, there is no exhibit of CD40 activation in the reporter assays. This is indicated on the left-hand side. On the right-hand side, it is in the absence of CD40, there is no activation of an exhibit of 4-1BB activation. And so these curves indicating really that this molecule is leading to the proposed mode of action, and in preclinical data, this is the foundation for the further evaluation in the clinic. Another property of this dual antibody is the dendritic cell maturation, and this experiment is measuring really the percentage of HLA-DR-positive, CD86-positive dendritic cells in the total population. And you can see, indicated by the color coding on the right-hand side, that the isotype, the control antibodies for CD40 and for 4-1BB in this turquoise and light green, not showing any effect. Even the dark green, a combination of these two, both control antibodies, is not changing. And the Fc inert analog for CD40, which is indicated here in the gray bars, is not showing a dose-dependent effect here in the maturation or increase of the maturation of dendritic cells, neither the analog for the 4-1BB, which is the light green. What we see clearly is a dose-dependent increase of dendritic cell maturation for the DuoBody BNT312. This is study design of the dose escalation part for BNT312. You can see the dose levels, the classical design, 3 + 3, to identify the recommended Phase 2 dose. 100 milligrams was recommended here. Primary endpoint, as usual, MTD recommended Phase 2 dose and other secondary, like safety, antitumor activity, but also exploratory endpoints were investigated. From here, we moved really further, and these are some data from this dose escalation part, monotherapy, single agent, 50 patients. You can see on the left-hand side in the spider plots, really some nice responses, partial responses, durable partial responses over time. The disease control rate is 50%. We have two patients with confirmed partial response, melanoma and neuroendocrine lung cancer. On the identified recommended Phase 2 dose, 100 mg Q3W, we have observed one dose-limiting toxicity, a Grade 4 transaminase increase. It's mainly attributable to the 4-1BB element, which is known in the literature and described, but this was resolved really after treatment with corticosteroids. No NTD reached here, no drug-related Grade 3 or higher thrombocytopenia or cytokine release syndrome was observed, no treatment-related death. This is now really the expansion part, which I want to share with you. From the dose escalation and identification of the recommended Phase 2 dose, we went into the dose expansion part, and here you can see two main studies are ongoing with sub-cohorts. One study is focusing on the combination with pembro, so BNT 312 plus pembro in different indications, like indicated in the on the right-hand side in the gray box, melanoma and NSCLC, first line in TPS 1-49 or higher, and in head and neck cancer. The lower part is really the triple combination. It is BNT 312 plus pembro plus chemo standard of care, also in different indications like head and neck, and NSCLC, squamous, non-squamous, and pancreatic. All of these trials are still recruiting or waiting for follow-up of patients, and I will share with you, for the sake of time, really, information about our head and neck cancer first line trial in combination with pembro and chemo. These are the observed safety profiles here first for this expansion cohort, and you can see that in general, on the left-hand side, for the IO/IO combination, so BNT 312 plus pembro, there are really toxicities described like transaminase increase, which I mentioned before, not unusual with this molecule. Rash, fatigue, pyrexia, nausea, most of them were mild to moderate, a few cases of Grade 3. And if we look on the right-hand side, we can really see here that the backbone toxicity is not majorly different from the IO/IO combination. So some additional signals here for pruritus, transaminase a little bit higher, but otherwise in the same ballpark on both ends. So encouraging safety observation and signals here, nothing really to be concerned about. And these are data from the head and neck cancer cohort, which we have collected, and we have analyzed with a data cutoff of last year, October. Here you can see in a limited number of patients, for patients here available for this analysis, partial responses and complete responses in all four patients. And on the right-hand side, you can see here really that these responses are also durable. So these patients were analyzed also for their PD-L1 status, so we can say that both low and high PD-L1 expression is benefiting, obviously here from the treatment, and all four patients were HPV negative. This study is further enrolling. We can enroll up to 40 patients here, so it is, at the moment, too early to report further data, but the readout is expected for next year, and data might be presented on one of the upcoming conferences next year. I stop here and hand over to Özlem again for the last bispecific. The next bispecific is exciting as the other ones which have been already presented. This one, PM8002, is partnered with Biotheus. And what we have here is a fusion between an anti-VEGF-A IgG with a silenced Fc, so an inert Fc. And this IgG is fused to anti-PD-L1 VHH to a VHH part. That means what we have here is the combination of two validated modes of action. And the way this bispecific works is that with the anti-PD-L1 part, it binds to and disrupts the PD-L1 PD-1 axis, inhibition by tumor cells in the tumor microenvironment, and thus works against the T-cell deactivation. This not only is PD-L1 antagonistic, but also fixes this antibody within the tumor microenvironment, where it then can scavenge VEGF and thereby reverse the tumor angiogenesis promoting effect of this molecule. And it's very easy to imagine that this dual mode of action also would pair favorably with ADCs, which now with improved angiogenesis can enter the tumor and the tumor microenvironment much better. So as I said, we have here we are leveraging the anti-VEGF effect, which is a very well-established modality and mode of action. Many of us know VEGF in particular because of its tumor angiogenesis promoting activity, via which it also promotes proliferation and survival of tumor cells. And that means this bispecific antibody acts and reverses this tumor promoting effect. However, VEGF plays also a role in the cancer immunity cycle, in particular by down-regulating T cell activation via inhibition of dendritic cell maturation. And this has an effect on T cell infiltration into tumors, and it also increases the inhibitory effect of myeloid-derived suppressor cells. So VEGF supports the cold, so to say, tumor microenvironment. And this again is also an effect which is antagonized by anti-VEGF. So, the activity, the mode of action of the anti-VEGF component, which via the anti-PD-1 VHH part, is fixed in the tumor microenvironment, is on various levels, including the immunological or immune modulating level. Anti-VEGF bevacizumab, for example, is a validated mechanism across various indications. In almost all of the indications which are shown here, anti-VEGF concepts are approved treatments in many of these in combination with chemotherapy, but also with IO compounds. And that means that we know exactly where to test our compound PM8002. In fact, this is also what our partner Biotheus has done in their Phase 1, 2 trial. These are the indications where PM8002 is tested in mono and in combo, in the meanwhile in more than 500 patients across these indications who have been dosed and where we have data. So this is the Phase 1-2 trial design. The monotherapy part, which is ongoing, PM8002 has been tested in the dose escalation part in doses which range between 1 mg per kg and 45 mg per kg, and also different dose regimens have been tested here. The dose expansion part is including a number of different tumor indications, again guided by the VEGF, the anti-VEGF approved indications, including but not restricted to melanoma, ovarian cancer, cervical cancer, hepatocellular cancer, and others. This is data from this monotherapy part, which has been just this year presented by Yi Guo at this year's ESMO. The data set comprises 254 patients across, as I said, all these tumor indications. Many of these patients have been pretreated with previous lines, and most of them were IO naive. And as you can see here, there is an objective response rate of 16% and a disease control rate of 74%. And on the left-hand side with durability, we see a median duration of response of 7.4 months and a median PFS of 5.6 months in this advanced patient population. The safety also presented at the ESMO is manageable, and in fact, PM8002 is well tolerated as a monotherapy. And this data set is further expanded across indications, across multiple indications. Most of these patients have been treated in China. The IND for studies in U.S. has been just accepted, and we will extend assessment of safety also in ex-China populations now.... This is the combination part of the trial, where we combine with paclitaxel, a second-line treatment, for example, small cell lung cancer. And this cohort is shown here. Patients with advanced small cell lung cancer who have progressed after platinum-based chemotherapy, with or without checkpoint inhibitors, have been treated in this Phase 2 trial. This is data from 48 patients who have been enrolled in the meantime. We will extend to 99 patients. Here is a preliminary analysis of this data in 36 patients who have been analyzed. As you can see here, we have an objective response rate of 60% in the IO-naïve population, 72%, and a disease control rate in the intent to treat population of 86%. So also here, efficacy data in this pretreated patient population, which is encouraging and which we will follow up. And also in combination with paclitaxel in this case, but also other chemotherapies, we have a quite tolerable safety profile. This is patient vignettes from the ongoing studies, where you can see on the left-hand side in the bottom a patient with monotherapy with PM8002 pretreated non-small cell lung cancer patients. A patient where we see a regression of a tumor, and the other cases, a TNBC patient treated in combination with nab-paclitaxel, and patients on the right-hand side with from the second-line small cell lung cancer study, which I just have presented, who also show re-regressions. So to come back to the safety profile for a minute, we have conducted a very rigorous cross-trial comparison and literature search for anti-VEGF compounds and anti-PD-L1 and anti-PD-1 compounds in order to assess how PM8002 relates in safety and adverse event profile terms in comparison to these two targets against which it is directed. And what we observed is that the safety profile appears favorable, or at least comparable with regard to adverse events or immune-related adverse events, which are known and related to the individual targets, so to PD-L1, PD-1 on the one side, and to VEGF targeting. So to summarize the key takeaways for the immuno-oncology modulatory agents we have just presented for our anti-CTLA-4 compound. As pointed out already by Michael, we expect additional data readouts in 2024, coming from the different indication cohorts of our ongoing trial. And based on that data, we will plan additional registrational trials in 2024 and beyond. And we are very much interested in testing this anti-CTLA-4 compound in combination with several of our pipeline assets, including our cancer vaccines. With regard to BNT311, we have engaged together with our partner, Genmab, in health authority discussions on the design of a pivotal trial in post-IO non-small cell lung cancer. We will present data from the ongoing studies next year. The clinical data and pivotal development plan for BNT312 is, so to say, under construction and will be also presented next year. The strategy, the overarching strategy here, for our immune modulators, as pointed out, is to leverage this next-generation immune modulators to unlock potential in novel patient populations and to provide backbones, foundational and improved backbones for novel combinations. So with this, I would move to the next chapter, to a solid tumor cell therapies, and would like to feature one of our programs here, BNT211. We all know that CAR T-cells are very successful in liquid tumors. They have not lived up to this success, however, in solid tumors. There are many reasons for this. Two of the more important reasons are, on the one hand, that there are not many suitable targets on solid tumor tissues, tissues which have the tumor or the cancer cell selectivity, which is required for this highly potent modality. Another reason is that, in solid tumors, we have compartmentalization of the antigen-positive population, whereas in solid tumors, where we have circulating cells, tumor antigen-expressing cells, in the vasculature, the circulating CAR T-cells get continuously survival signals. BNT211, our program, is addressing both these limitations or hurdles and challenges in solid tumors. On the target side, we have chosen claudin-6 as a target for our CAR T-cell. This is one of the rare targets on epithelial and non-epithelial solid cancers, which has the cancer cell selectivity which we need. It this is a carcinoembryonic antigen, which is physiologically only expressed in a very defined stage of embryonal organogenesis, a tight, very primitive tight junction molecule. And after this stage, it is tightly transcriptionally silenced. So that we don't observe expression in, over, across the entire body map of adult healthy tissues. However, in a number of cancers, testicular cancers, ovarian cancers, uterine, lung, gastric cancers, and a number of rare cancer types, claudin-6 is apparently switched on. And several of these cancers can reach very high and homogeneous expression levels. So this is the target we have selected for our claudin for our CAR T-cell. And have constructed a second-generation CAR chimeric antigen receptor, which with a 4-1BB costimulation domain, for which we have shown that it has indeed selectivity for our target claudin-6. Which was not trivial, because the claudin family is a broad family with several members also expressed in normal tissues, and it's not trivial to get the selectivity just for this one family member, claudin-6, and thereby avoid targeting of other normal tissue-expressed claudins. We are combining this CAR T-cell with a vaccine. Why? Because we want to compensate for the lack of survival signals and activation and expansion markers in the periphery. And the concept is that whereas this lack of survival signals and activation in solid cancer CAR-Ts frequently leads to a yeah short-term peak of circulating CAR T-cells. We wanted to use the vaccine expressing claudin-6, the target of a CAR T, to constantly and repeatedly act expand and activate and stimulate the adoptively transferred CAR T-cells to maintain persistence of a CAR T-cells in the periphery. And also use this approach to increase CAR T-cells from sub-therapeutic levels to therapeutic levels. We have shown in our preclinical studies, and these are published, for example, in Science, that this concept really works very well in mice. So, our CAR-T cells are specific, and the CARVac concepts, meaning the CAR amplifying RNA vaccine, in animals, is capable of leading to persistence of CAR-T cells, and now we are testing this approach in a clinical trial. This clinical trial is in its dose finding and dose ranging Phase. We are including Claudin-6 positive cancers, defined as at least 50% of tumor cells with 2+, 3+ staining for Claudin-6 via immunohistochemistry. We include here tumors like ovarian cancers, germ cell tumors, testicular cancers. We have conducted the dose ranging first with the first manual process of manufacturing these CAR T-cells, which we have established. This data has been published just recently in Nature Medicine. I will just show one slide or two slides of this data in a minute. In the meantime, the manufacturing process is automated, and we are repeating the dose escalation part. Dose escalation relates to the CAR T-cells, where we are testing CAR T-cell numbers ranging from 1 × 10^6 to 2-5 × 10^8. But with the automated process, the CARVac, the vaccine, is kept at one dose and administered repeatedly after adoptive transfer of these CAR T-cells. And patients are obviously lympho-depleted with a conventional lympho-depletion and preconditioning regimen prior adoptive transfer. This is now the data which has been published from the manual process. And what we have already observed very early on in the first sort of cohorts of patients treated, and these are patients with testicular and ovarian cancer. Mostly you can see that at the color code, we get in these heavily pre-treated patients, objective responses, which are durable, both in the only CAR T-cell-treated cohorts, as well as in the cohorts where the CAR T-cells are combined with the vaccine, which is our RNA Lipoplex vaccine. This is one of the case reports from this first published data set. This is a patient with a mixed germ cell tumor, who had been heavily pre-treated with five lines of chemotherapy in total, had multiple relapses in their history. One of those shortly prior entering our clinical trial, with also multiple lung metastases. The tumor at entry into our clinical trial was rapidly progressing. Only between screening for our trial and the adoptive transfer of CAR T-cells, we saw an increase of almost 40% of the target sum. As you can see here on the CT scans, the patient very fast responded to infusion of these CAR T-cells with a regression of multiple lung cancer metastatic lesions. This is data from our automated process, which has been presented at the ESMO this year. This is a data set of a total of 44 patients. Still the dose escalation part. These are patients with epithelial ovarian cancer, germ cell tumors, but also with lung cancer, and for example, one patient with sinonasal carcinoma. All of these patients, heavily pretreated. We see a manageable safety profile, in particular in the lower doses. The adverse event profile depends, as expected, impacted on the CAR T-cell dose. Most of the treatment emergent AEs are laboratory findings, decreases in blood count, and elevations of liver function tests, or cytopenias. The treatment emergent SAEs are mostly infections. We see 4 DLTs in the higher dose ranges and are therefore adapting, continuing to test lower dose levels to determine the recommended Phase II dose. CRS, cytokine release syndromes are also mostly grade 1, grade 2. This is efficacy data. 38 of those 44 patients were evaluable for efficacy. And these are now all these, this waterfall plot shows all dose levels with and without CARVac, with and without the vaccine. And as you can see here, we across all dose levels and all indications, we have an objective response rate in this heavily pre-treated patients of almost 45%, and a disease control rate of 73%. In this patient population, you would expect much lower objective response rates. This is now the dose level 2 subset of patients. Again, plus and minus vaccine, not differentiated here. As you can see here, we have an objective rate or objective response rate across all the tumor indications of almost 60%, with a disease control rate of 95%. The follow-up is not so long. The follow-up time we have, because with this recent data, is not so long, but for the- we see that for the duration of 100 days, we can see persistence of CAR T cells and are continuing to follow up with patients. In the higher dose level of 1 × 10^8, we can also see an effect of using the vaccine. You can see here on the right-hand side, only CAR T cells without addition, repeated addition of a vaccine. On the left-hand side, we have at the same dose level patients with CAR-T cells at the same dose level and repeated vaccination. And as you can see here, we see in a couple of patients improvement of CAR-T cell persistence. So, key takeaways, we have a manageable adverse event profile and continue to determine the recommended Phase II dose for the CAR-T cells. We see encouraging signs of activity with 13 responses in 22 evaluable patients at dose level two. The pharmacokinetics points to an improved CAR-T persistence by adding our vaccine, the CARVac, and we are continuing to determine the Phase II dose. We see that in particular testicular cancer patients, and this is described in our publication, in our published manuscript, respond well to the CAR T cell treatment, heavily pre-treated patients. Based on that data, we received the PRIME designation for testicular cancer from the EMA. We also see that there is a high unmet medical need in patients with refractory-resistant germ cell tumors. There is no curative treatment option for these patients post-salvage cisplatin-based chemotherapy regimens. It's also a neglected patient population. Checkpoint inhibitors have failed, and this is the patient population in which we plan our first Phase II trial with registrational potential and are continuing to assess additional pivotal indications. With this, I would move to our mRNA cancer vaccines. Cancer vaccine mRNA-based cancer vaccine platforms are not created equally. Our cancer vaccine platform is based on Uridine mRNA, with our proprietary translational performance-optimized non-coding backbone. We are using our mRNA-Lipoplex formulation, which allows intravenous and thereby systemic application of mRNA, which again means targeting of lymphoid compartments and antigen-presenting cells body-wide, which we think is a success factor. This formulation is also optimal, and we have shown this also preclinically and also in our clinical trials is optimal for antigen presentation in lymphoid organs, where immune responses are generated physiologically. This formulation, in particular, the uridine mRNA, provides exactly that distinct innate immune stimulatory signature, which we want to see, so intrinsic adjuvanticity. We are using this platform, the mRNA Lipoplex platform for both of the flavors of our mRNA vaccines. Uğur already talked about this, our FixVac, as well as our individualized neoantigen-specific vaccines, iNeST, which we have partnered with Genentech, Roche. And in for both vaccine types or categories, we have ongoing clinical trials. We are particularly excited about our individualized and personalized cancer vaccines, which we have pioneered. We have, with our partner, Genentech, large Phase 1 data, Phase 1 trial, 200+ patients, where we have treated multiple tumor indications, with iNeST alone, plus in combination with atezolizumab. And this was a very important learning exercise because it has shown us that all the different solid cancer indications are feasible for neoantigen vaccine approaches, even if mutational load is low or biopsies are not easy to recover, for example, for cancers like lung cancer or pancreatic cancer. And this trial has also allowed us very deep analysis of immune responses which we get, because that is really a very important indicator for what we expect our in terms of performance of our vaccine. This data is being compiled right now and will be submitted soon for publication. We have our first-line melanoma trial, which is operationalized by our partner, Genentech. There we have been very unfortunate with slowing down of this trial during the pandemic, and we have not really fully recovered from that after the pandemic. So that recruitment of patients took longer than expected in this randomized trial, where we combine iNeST with Pembrolizumab and compare against Pembrolizumab in first-line melanoma, a space which is highly competitive and where many compounds, also interesting ones, have been tested. Based on these delays, we are still waiting for the readout of our event-based PFS. Then, we have managed to move into the space, which actually is the interesting one for us, namely the adjuvant space, where we want to position our individualized vaccine. We have ongoing Phase 2 randomized clinical trials with registrational intention in high risk colorectal cancer in the post-adjuvant setting. And we have just initiated, which we are operationalizing, and we have just initiated with our partner Genentech trial in the adjuvant setting of pancreatic carcinoma. The CRC trial is monotherapy with iNeST. The PDAC trial is a combination with atezolizumab and sandwiched into the standard of care chemotherapy. And we are very excited about in particular these adjuvant settings. Our 4-point strategy, in particular for gaining leadership in individualized mRNA cancer vaccines, has been for the last couple of years and continues to be to pursue these 4 strategic points or streams. We aim to establish commercial manufacturing capacities and also extend our clinical manufacturing capabilities for personalized vaccines, and we have come a long way there. We continue to decrease manufacturing time. We continue to improve neoantigen selection. We have already a very advanced algorithm, and which is further improved with every patient from whom we learn. We also continue to advance the pipeline, in particular in the adjuvant setting. You will see more clinical trials and more indications coming. So, I want to take a minute to look into our history, and this is not because I'm nostalgic, but because I want to share with you some new data. This is actually our very, very, the very first trial, individualized neoantigen-based clinical trial, which we started in 2013, and which we reported in 2017 in Nature. So this data is published. 13 patients with stage 3-4 melanoma, who were treated with actually the prototype version of our individualized vaccine, which at that time was non-formulated, so we did not have the intravenous formulation at that time. And the vaccine was injected directly into lymph nodes, individual lymph nodes of the patient. As I said, we have reported the data, and I will not go into details of what we have already published. The only thing I want to say is that this was actually the first trial in which a vaccine platform showed for the first time 100% immune conversion. So immune responses, T-cells induced by the vaccine in all patients who have been treated. This sounds trivial, but this was indeed the first time this was achieved. And you can see this exemplified here for one patient. And this is not only zero conversion, immune response conversion in 100% of patients, it's also that we see T-cells going, T-cell responses going from zero to hero. So this patient does not recognize these antigens, which are there, needs the vaccine to even build an immune response, and this immune response is high magnitude, as you can see here. What is new, and what I want to share with you, is that we have now follow-up data from these patients. Data where we see even now, 4.5 years after starting treatment of these patients, that the immune responses which were induced by the vaccine are still there, are persistent, even though the patients have been, have not been immunized for us for a while, for a while. And even though this is not even the most potent, upgraded version of our vaccine, and this is, yeah, pretty exciting. And another piece I want to share with you, we have reported in that publication back in 2017, the right-hand part of this, yeah, Western blots, so to say. And what you can see here is on the far right. Let me see. It's your left, actually, on the far left, you see patients' prior vaccination, and these recurrences. On the right-hand side, you can see that the number of recurrences across all patients is dramatically reduced by vaccination. And what you now can see, this is new data, the far right, is that many of these patients, even though they are not vaccinated anymore, are still living more than six years after starting treatment and are still lesion-free. And this is very encouraging for us. So this is melanoma, and you all know that, this is one of the preferred indications, similar to non-small cell lung cancer for vaccinologists because, the tumor burden, the tumor mutational burden is high. And these are immunogenic tumors, which also respond well, to checkpoint inhibitors, are regarded as the lower-hanging fruits, for exploring neoantigen vaccines. However, while neoantigens are individual, they are based on cancer mutations which are a hallmark of cancer, and therefore have a promise of being universal targets universally across a tumor indication. So the question really is: Can we go beyond melanoma and, lung cancer? Can we go into indications which are lower mutational burden, which are the typical, yeah, cold or immune-suppressive tumors? Examples are, for example, here, pancreas, breast, and colorectal cancers, which, as you can see here, and this is our own data, have a low mutational burden. We think yes, because we see very strong immune responses with our vaccine, as I have just demonstrated, and this is the space, as I already pointed out, we want to go. Pancreatic ductal adenocarcinoma is a high medical need cancer. The five-year survival rates after resection alone is, around 10, 10%, and, up to 75% of patients with pancreatic ductal carcinoma relapse, even though they appear tumor-free within 5 years after adjuvant treatment. Triple-negative breast cancer, also a dire patient population with up to 45% relapse rate within 4 years after adjuvant treatment, and a high-risk colorectal cancer is also a high medical need indication, and this is the space in which we are now. As I pointed out, we have a Phase 1 trial completed in advanced pancreatic ductal adenocarcinoma. We have started a randomized Phase 2 in triple-negative breast cancer. We have just analyzed our Phase 1 trial, and in colorectal cancer, randomized trial is ongoing, and I will just briefly share with you... Our data from our adjuvant, from our small 14-patient adjuvant, triple negative breast cancer patients, where the vaccine, our individualized vaccine, BNT122, was used post-adjuvant or post neoadjuvant standard of care in these patients. And what we have actually observed is, again, 100% immune conversion and T-cell responses against multiple antigens in each and every patient. High magnitude immune responses. You can see here one case, 10% of the circulating cells are against one of the vaccine antigens. These are expansion rates you only see with adoptive transfer, with CAR T-cell therapies, for example. And again, this is follow-up data, 600 days, yeah, after treating, vaccinating the patients. The dashed lines actually on the far right are the vaccinations, so the patients get short-term vaccination and stay without boost. We have still long-term immune responses. So, this is an adjuvant pancreatic ductal adenocarcinoma trial, the Phase one trial, where we have treated patients after resection with one dose of Opdivo to prepare for the vaccine. The individualized vaccine was given for a couple of priming doses, followed by the standard of care adjuvant chemotherapy, and then again followed by a vaccine booster dose. This is a small clinical trial, 16 patients, an IIT, which we conducted with colleagues from the MSKCC. What we have observed is actually in the pancreatic cancer population that only half of these patients have an immune conversion. This is not what we see in all other indications which are tested, which probably speaks to the fact that, pancreatic cancer is considered as, highly immune suppressive. However, those patients who have an T-cell response, a de novo T-cell response, which is not their prior vaccination, have, a high magnitude to, to, responses, which are of long duration. Here, we'll follow up, for, 2 years, and, which are also not compromised by the the standard of care chemotherapy they are combined with. The fact that, half of the patients just responded, with immune conversion gave us an internal control population. And as you can see here, when we look for recurrence-free survival, those patients who have an immune response, which is vaccine-induced, are still alive and have not shown recurrence. Whereas, those patients who were not able to mount an immune response show recurrences in the time frame, which are expected. Based on this data and motivated by the immune responses and the clinical activity which we see, we have started a Phase 2 trial with our partner, Genentech. A randomized Phase 2 trial, where we compare a slightly modified regimen against standard of care, which is modified FOLFIRINOX. So to summarize the personalized cancer vaccine part, we aim to bring personalized cancer vaccines, in particular, into the adjuvant treatment setting, and we dare to also do this in tumors with low mutational burden and in cold tumor types, because this is where the high unmet medical need is. With this, I would hand over to you. Okay. Thank you, Özlem. I'm just gonna say a few closing remarks here, and then we're gonna open up the floor for questions, and we're gonna have mics that go around to hopefully indulge you guys. Okay, I'm gonna speak on the path to value creation that we see as we look to the years ahead and to 2030. And again, this is big picture. So going back to the framework that I talked about in the first section, the three basic growth pillars of COVID-19, immuno-oncology portfolio, and infectious disease portfolio. So what are we looking to do in the next stage of the company's development? So for COVID-19, of course, this means continuing in this transition with our commercial COVID-19 franchise into the next Phase, which we think could bring combination and next-generation vaccines from 2025 onwards. We just had an announcement yesterday that Pfizer and BioNTech will start a Phase 3 trial in the coming months, looking at some of the combination vaccines, COVID flu being one of them, and probably some others. On the immuno-oncology side, the next stage will involve executing multiple pivotal trials, and we think launching multiple products from 2026 onwards, as we've stated. And then finally, for infectious disease, we do plan to initiate our first late-stage studies as we broaden and expand our early-stage pipeline. So then looking at where we are today and where we want to be in 2030 across these different pillars, and I've added a fourth, which is our balance sheet, which we think is, as I said yesterday, an immense asset to the company, and is going to be very important in, in helping us, develop and grow and actually accelerate over the next couple of years. So today, we have a balance sheet of EUR 17 billion, approximately. This is data, of course, as of September thirtieth, as of the end of Q3, with EUR 2 billion in trade receivables as of that date. And of course, that's an interest income-generating asset on its own right. We then have a leading market-leading COVID-19 vaccine franchise, which, as I've described earlier, on a product basis, is highly cash generative because of our lean cost structure, and also global in nature in terms of the diversity of revenue that it brings. We then have an expanding late-stage oncology pipeline, which is also diversified by platform, mechanism of action, and indication. We have an expanding early-stage infectious disease pipeline, which we didn't profile today, but there will be more updates in the coming months and certainly in 2024 on that pipeline as well. We do have some interesting data coming in that we're going to seek to publish in the coming months. So that's where we are today. As we look to 2030, where do we want to be? Well, firstly, we want to still have a strong balance sheet in 2030. Secondly, we want to transition or convert our market-leading COVID vaccine into a multi-vaccine portfolio. And that could mean not only multiple vaccines to address the various forms of COVID-19 in various populations, but also extend through combinations into other adjacent indications with our partner, Pfizer. In addition, by 2030, we want to have multiple products approved in the oncology field, and also a leading late-stage pipeline to set us up for further growth through the 2030s. We think the innovation that we've touched on today, which is just a small part of what we're working on, will position us very well to really transform oncology treatment, as I said, for a decade and beyond. And then in infectious disease, by 2030, beyond COVID, we think that we can have our first approved products, and again, here, a late-stage pipeline behind that, giving us, in totality, strong balance sheet, but also a diversified cash flow-generating multi-product portfolio, which is where we think we can, we can be in 2030. So then to zoom in a little bit on the path to that and how we see that, what are the sort of key principles that we're, that we're building into our, into our decision making and our strategy? So again, 2023, this year, even with the low rates, relatively low rates of COVID vaccination in Western countries that we're seeing, we expect to be profitable in 2023 if we can hit our revenue guidance that we've given the street. In addition, we've grown our cash balance from the beginning of the year to the end, despite having also invested about EUR 1 billion in BD and M&A. As we look to 2024, we do want to increase our investments in R&D, in particular in oncology, for pivotal trials. We've talked about that, the goal of 10 pivotal trials ongoing, 10 or more, by the end of the year, next year. We do expect to maintain our lean cost base in terms of SG&A expense. We will continue our active BD and M&A strategy, as you, as you've seen on display this year. That's going to continue. We do see opportunities, and we have a pipeline of opportunities that we think could bring further synergistic assets to the company. And again, we want to maintain our strong balance sheet. And we think we can do that, given, again, the unique features that we have baked into the Pfizer collaboration and the COVID vaccine business. Then looking beyond 2024, as we look to 2025 and 2028, our goal here is really to, as soon as possible, get to a period of what we call sustainable strategic growth, which means to us multiple new product approvals, revenue growth from our first oncology launches, and potentially combination vaccines. Getting to a point of sustained profitability and cash flow positivity across the whole company as soon as we can, and of course, continuing to maintain a strong balance sheet. So that's the basic vision and roadmap that we're working against. So then to close, we see the path to value creation comprising four key components. The first is increasing investment in R&D with a focus on pivotal trials. The second is to continue investing in external innovation with a focus on synergistic assets to complement our organic innovation. The third is to build an oncology commercial front end, not only alone, but also in collaboration with our partners. And finally, we think the true path to value creation for the company is to commercialize multiple new products in infectious disease and oncology. And that's what we're focused on as we think about capital allocation and as we think about execution. And we think ultimately, that's the best way for us to create value for shareholders, but also for patients and society. And with that, I'd like to close and thank you all for your attention, and we'll open up the floor for Q&A. Can you hear me now? Okay. Tazeen Ahmad from Bank of America. I have four questions. I'll just ask them all together, for simplicity. For the BNT-311 combo that you talked about, you are going to be presenting data in 2024 at a medical meeting. Can you give us a sense of what you would consider to be good data at that time? And then I just want to clarify, would it just be non-small cell lung cancer, or would you also be able to include some endometrial data in that? Because I know that's just starting. And then the second pipeline question is on the 312 combo. What data exactly should we expect to see in 2024? And then I have a couple of COVID questions. Should I wait? Yeah, we can take that. So the first question was regarding 311? Yes. Definition of good data is, of course, not easy while the trials are ongoing, so we are expecting really to collect further data from the ongoing cohorts. As I have indicated, there are some of these cohorts still recruiting, so there will be cutoffs, and we will look into that. Of course, there are expectation regarding the number of patients to be enrolled until this point and what to see. As I mentioned, 40 patients maximum can be enrolled in this expansion cohorts, so this will guide us in certain directions. So the mentioned expansion cohort for the randomized study regarding the 2 doses is too early to make any conclusions out of that. So this is regarding 311. You wouldn't be in a position to make a go, no-go decision based on that data cut? It depends on which timelines we are talking about. More to the end of the year, we will have more mature data and more follow-up. Beginning of the year, this will be too early. But we have- Duration is an important- Exactly. Duration- metric for us, and therefore The follow-up of patients will give us this information, most likely more in the second half of the next year. Okay. Thank you. And then on 312. The BNT312 is also still recruiting. In the cohorts, we are analyzing the data in the different indication, mainly head and neck cancer, will give us most likely more to the mid of next year information about more mature data in head and neck cancer, first line. Okay. And then on COVID. So, Ryan, we used to talk a lot more about the China market. It's become quiet. Just wanted to know if there's any update on expectations of ever getting approval there to market and what the market could look like. And then the last question I have is just on timelines. So in the past, to get a vaccine approved, you've had expedited treatment from FDA because we were in a pandemic. As we move back to endemic, how should we think about timelines for Phase 3 for the flu COVID combo, if you have any data on that? Okay. Yeah, so on China, so of course we never give up, but I think it has to be said that at this point, the scope of the opportunity that we would see is much more muted. So you know, I think it's very clear that you know, there were no foreign vaccines approved in China during the pandemic. I think that was clearly a matter of policy. We don't have any expectation for any near-term change in policy, but of course, policies can change, and sometimes you don't know that they're coming. So, you know, we are continuing to build our presence, actually, in China. As you see from some of the deals that we've done, our strategy is actually to tap into some of the innovation we've seen coming out of China in the first instance. So COVID still, you know, I think is still an open question, but no immediate expectation of a change in policy. With regard to the timelines, of course, we are not anymore talking about pandemic timelines, but still also the timelines with regard to authorization of COVID flu vaccines are fast. The FDA has a high interest to bring in combination vaccines, respiratory combination vaccines. So with that, we don't expect this now light speed mode, but still fast authorizations based on Phase 3 data. Daina Graybosch from Leerink Partners. Let me ask one big question and one really weedy question. So the big question is: how and when do you think you'll transition to the more audacious combination development? I sat here and counted 10 combos I want to see, specific combos and indications, you know, tomorrow, and I'm sure you can do the same, and I'm sure you could do 30. So when, when do we get to that in, in Phase 3, and is there anything in AI that can help disentangle signal of combination trials, which we all know is really difficult? So that's the high level, and then the more detailed one is on the vaccine. So really interesting new data, and I just immediately reacting to these really durable T cell responses of intranodal injection and the neoadjuvant, as well as the lack of immune response in pancreatic, which I think was somewhat associated with splenectomy. I wonder how you're thinking about delivery overall. Are you thinking about doing just IV or a mix of IV and intranodal? And is there any way to get a vaccine into neoadjuvant instead of adjuvant when you still have the tumor in the lymph nodes and potential for epitope spread? Thank you. Yeah, uh- We need your combo list. Yeah. Yeah. ... Dana, these are great questions, and we will see engagement into combos starting already in 2024, with multiple exploratory combination arms in various indications. And the first set of combination therapies were based on two very simple concepts. On the one side, preclinical data showing synergy, yeah. And, in the slide with the Venn diagram, we showed the synergy fields, and these are indeed things that we have seen in preclinical settings, yeah. So these are obvious combinations to address. And the second aspect is what we have seen in single arm trials, yeah. So it makes sense, for example, to combine two compounds that have, in the same disease indication, shown single convincing clinical activity. For this, we don't yet need AI. We can do that just based on experience. And we are not only anticipating double combinations, but from 2025 on, also triple combinations, yeah, in some IO indications. And we expect to move from exploratory studies into registration, potential registration studies in 2025 for the first double combos. Okay, so this is. Can AI help to identify combos? I would not exclude that. I would see it more in the patient-centric way, yeah. AI could help to see which pathways are relevant in the tumor, and that was visualized in our last slide, yeah. So that we can read. We get from these tumors, this is really, really something extremely exciting. When we do this genome and RNA sequencing, we are not just getting the new antigens, we are getting the full spectrum of information in the tumors, which regions are amplified. We can recapitulate if the patient is estrogen receptor positive, if the patient has, in the metastasis, different other pathways activated. And we will come there, where we can even predict what will be the next escape mechanism of the tumor, yeah. So this is something exciting. The key question is: How can we connect that to a regulatory path, yeah? So this will require a few years' dialogue with regulators until we are there. To the second question, with the pancreatic cancer, do you want to take that? Yes, I can take that. Yes, we were also surprised to see this dichotomy in pancreatic cancer, because as I pointed out, we have various trials for Phase 1 trial with our partner Genentech, goes across indications, and we have not seen this for other indications. You have read our manuscript very carefully, and indeed, there are imbalances between these sort of post hoc two cohorts, the immune responders and the non-responders, including that in those who don't respond, we have a higher prevalence of splenectomized patients. The spleen is the largest lymphoid compartment, and we are the ones who preach that the lymphoid compartment targeting is essential for not only our, but all vaccines. That can have an impact. So we need to continue to monitor that. Splenectomy is part of the surgery technique, which is used for resection of pancreatic cancer. We are looking deeper into this data. The immunogenicity data is just capturing the high magnitude responses. So what we want to understand is: Is there at least some degree of response which we can boost further, for example, by a different dosing regimen of atezolizumab we combine with via different ways? So we will probably get a better understanding as data comes in. We are also expanding the spectrum of neoantigens. We have included in our pipeline, for example, now neoantigens coming from splice mutations, which adds additional mutations, and thereby increasing the likelihood that some of these mutations may add higher immunogenicity rates. Neoadjuvant versus? Neoadjuvant versus adjuvant. We want to go as early as possible with the treatments. The neoadjuvant is. So one challenge for the neoadjuvant is that the tumor is diagnosed, the patient is biopsied, and neoadjuvant treatment starts already. So that means the manufacturing time of 4-6 weeks comes into a situation where the biopsy might be right. These are small biopsies that we identify the mutations and deliver the vaccine in time. I believe we will come there, but at the moment, with regard to the feasibility, we believe that the adjuvant setting is the more appropriate one. ...Thank you very much. This is Sam Fazeli from Bloomberg Intelligence. Thanks, first, for putting this fantastic session together. It's created so many questions that I'm gonna have to be emailing the IR team endlessly. But I have three questions detailed, and one strategy. On clinicaltrials.gov, for the BNT323 Phase 3 trial, it's lists only one clinical trial site. And either that's an error, or if it's not an error, it's just a Texas site, and I just wanted to understand whether that meant something as regards to this trial. Is this proof of concept Phase 3? Which, that's fine. And second question is, the Trop-2 ADC, I know the N was quite small, but did you have any responses in squamous carcinoma patients, given what we've seen from AstraZeneca? And then, the 316 single agent activity, the CTLA-4, is it possible, are you confident that the activity isn't because of the over or the lack of washout from previous checkpoint inhibitor, so that you're actually seeing a combo effect? Which is nothing wrong with it, because you're still seeing an effect, but is this a true single agent in the non-small cell lung cancer cohort? So obviously, you have other data sets. And then the broader question is, you're heading for profitability this year. It sounds like you're trying, going to try and do everything you can to remain profitable. So the cash balance, I mean, there are a few biotech or pharma companies, biotech definitely, but pharma companies that are profitable, that actually maintain a positive cash balance. They all try and... Because it's too expensive and burns a hole in their pocket. Is the idea that you're gonna continue to use this diligently, I'm sure, on M&A and CapEx as required, or would there be a day that you go, "It doesn't matter whether we're profitable or not, we're just gonna pile another EUR 2 billion into our R&D because our trials are looking great"? Thank you. Absolutely. Yeah. So, so you start, then we can share. I'll start with the simple ones. With the most simple one is, yes, it's a randomized Phase 3 trial, and yes, there will be more than one site. The number of sites is in the range of 120 or so. We haven't yet put them on ClinicalTrials.gov as our... The trial is actually operationalized by our partner, DualityBio. The update will happen coinciding more or less with the FPI, which we talked about will happen more or less imminently in the next few days or weeks. On the Trop-2 question, yes, but I—so you asked whether there were squamous cell carcinoma patients with regards to the ESMO data. Yes, there are, but we are at this point, have to talk about very small numbers, and so we're not ready to declare this drug works any differently than other Trop-2 agents with regards to squamous cell. We were all surprised by the data, but we looked post-hoc, and we think it may have to do with Trop-2 expression in the different categories, and we are now recruiting those slides and try to look at that for ourselves. So with all caution, there may be something to also squamous cell, but I wouldn't take that for granted right now. And as for 316, the- Washout. The washout. We do have patients. So the trial population was refractory to PD-1, so most of the patients had a relatively short interval. But there were some of them which had longer intervals, up to half a year, and yes, they also responded. But again, this would not lend itself to a statistical analysis that we could claim this is independent of prior PD-1. But we do see responses also in patients that were either naive to PD-1 or had longer washout periods. Yeah, and I think to your last question on cash balance and profitability. So I think what I would say is the first priority is that we see a big long-term opportunity in front of us, and we want to invest, we need to invest to realize that opportunity. So I think that's the first point, and we're in a very strong position to do that. You know, I think in terms of cash and cash balance, as I said, we see our current cash balance as an asset, especially given where we are as a company. And what I mean by that is that we're undergoing a transition on two fronts, right? We're undergoing a COVID transition as the COVID market restructures. We're also undergoing a transition to become a commercial stage oncology company over the next couple of years. I think that, you know, we view that cash balance as an important part to carry us into that next Phase on both fronts. So I don't think the typical—we don't view it as a sort of excess cash. In the next two years, we don't view that as excess cash. That's an important driver of our future and an enabler of us to realize the long-term ambition of the company. So it's not to say that we're gonna carry that kind of cash balance long-term. That's not what we're saying. But I think in the short term, that's a strategic asset to us, and we want to preserve it to the best extent we can, while using it for good investments. So that means continued, measured BD and M&A. I say measured meaning, you know, our core strategy there is really for bolt-on, bolt-on deals, whether they be in licensing or measured acquisitions, not a huge transformational acquisition. That's not a priority for us right now. Because as we said, we think from the base that we're starting on, we can we can transition, if we manage it effectively and we use our cash prudently. ... we can manage into that next stage of the company, over the next couple of years. Does that address your question? Okay. Great. Thanks so much for putting the whole event on. It's Brendan Smith from TD Cowen. Two, if I might, on the ADCs, and then two on iNeST, if that's okay. So first, for the ADCs that you're advancing, could you maybe expound a bit on what, if anything, differentiates your topoisomerase-1 moiety from the competition, just to try to understand the relative positioning there structurally? And then for PM8002, can you just clarify, is the second-line SCLC paclitaxel combo a registrational study, or do you think you'll need to actually do a head-to-head study there versus chemo? And then I have two for iNeST, so I'll do that. The second question we got, we didn't get the second question. Can you- Oh, sorry. For... Can you hear me? Yeah. So for PM8002, is the second-line SCLC paclitaxel combo study, is that registrational, or do you think you'll need to do a head-to-head study versus chemo? Okay. With regard to the ADCs, we have evaluated before licensing technology, the preclinical data and the PK data. We have now also PK data, which show superior stability of the linker and the antibody in the circulation. And for our BNT323 ADC, as you have seen from the presentation, we have selected a dose of 8 mg per kg, which is higher than the dose that are applied, for example, for Enhertu. Yeah. And we believe that this could be a distinguishing factor with regard to efficacy, but even more importantly, also with regard to tolerability. We see some side effects with a lower frequency that are observed for the ADCs in the same class. It is too early to make a big statement of that, but this could be an additional differentiation factor. And the most important note is we do not want to position these ADCs as stand-alone products. This is our path to enter the market, yeah, and then enable us to do combination trials, which could allow really to make a bigger impact in the same patient population. With regard to PM8002? Yeah. So I can comment to that. So you saw on the slide the ongoing activities in the different indications. None of these studies are registrational at this time point. So we are starting conversations on the further development path, and this indication is definitely in scope, and more to come. At this moment, we cannot comment more and in more detail, but the ongoing studies are the foundation for the further development into registrational trials. Great. Okay, and then, quickly on iNeST, can you just clarify, maybe, is there a particular reason why you aren't running, an adjuvant melanoma study with, BNT122, kind of just given the Moderna/Merck data there? And maybe what gives you more confidence in, in pancreatic and, CRC adjuvant settings for the Phase 2? And then maybe just more broadly on iNeST, what do you kind of see as the, the drivers that might explain why adjuvant might be more amenable to efficacy, than metastatic? And is there kind of... Do you expect any difference, in the data between those two settings in particular? So the first question was why have we chosen pancreatic cancer, for example, and not melanoma? Our assessment of which adjuvant indication to choose is not final yet. We have started with the first indication, CRC and pancreatic cancer. Based on some data, we have seen the pancreatic cancer data I have shown and also the CRC immune responses in our Phase 1, in our big Phase 1 trial, were motivating for us in order to choose CRC. There are additional indications which will follow, and we don't exclude that melanoma could, adjuvant melanoma could be among them. The second question was why adjuvant? From an immunologist's point of view, the metastatic, whereas the adjuvant setting of the same tumor type, are entirely different sort of beasts. In the adjuvant setting, you know, immune therapy, in particular, vaccination, that's a numbers game. You need a sufficient magnitude of T cells against your tumor, and if a tumor is a 1-kilogram football, that makes it difficult, right? And therefore, the adjuvant setting, or actually any minimal residual disease setting, where it is more about attacking micrometastasis or preventing recurrences. It's not only about tumor mass. Also, there is no established tumor microenvironment which could be suppressive or inhibiting, and the clonality of the tumors in early settings are such that you have a lower degree of different clones and heterogeneity. Hi, Bill Maughan, Canaccord Genuity. Thank you. So I have two fairly broad questions for you. So first of all, on your, personalized cancer vaccine, and you've spoken about your neoantigen identification, capabilities. How important is it to be differentiated in that versus just, put together a functional, neoantigen, identification algorithm? In other words, from our perspective, you know, we-- it's hard for us to diligence, different neoantigen identification, capabilities among companies because it's, it's somewhat of a black box unless you're a true expert. Second question is, when you're thinking about in-licensing late-stage assets, as you mentioned before, simply kind of adding in, late-stage near commercial assets where you didn't have it before, has value in and of itself. If they're differentiated, obviously that's better, but bringing something in that you can combine with what you're developing and bring near-term revenue stream has value in itself. So how are you thinking about simply adding high likelihood of success modalities versus something truly differentiated and maybe a bigger opportunity in itself? Thank you. I can start with the first question, which the second question. What? The first question? The algorithm. The algorithm. So to be very frank, I believe the algorithms are based on science. So it is. And because they are based on science, they will not be somehow mystic elements, but they will be just a clear understanding how prioritization will work. And I believe that neoantigen prioritization will become a commodity. Yeah. At the end of the day, we know the rules, we know what are the key elements for prioritization. This has been published. There are 20-plus publications showing what is really important. Yeah. And I would love that we end in the industry with a full transparency how neoantigens are identified. And of course, you can bet on different neoantigens. You can say, "I am interested particularly in fusion genes. I am interested in mutations that which have the anchor positions. I am interested in mutations which are clonal antigens." Yeah, you can tell that, but still make it transparent. Yeah. So how these algorithms are affecting the response, I would say it's like with any type of diagnostic, if you are hitting the right patient population with your diagnostic, you increase the likelihood of success. Yeah. It's not black and white. It's not just patient is going to respond or not respond, but it might be that instead of 25% of patients responding, 35% of patients are responding. Yeah. So this is an incremental science, yeah, incremental deep science, and we will see a lot of progress and publications in the field here. Yeah. And on your in-licensing, the in-licensing strategy question. So when we look to our pipeline organically, last year, for example, the disproportionate number or amount of our programs were going after novel biological targets of some kind. So we had already a high, very high level of novelty as a sort of base starting point. And it's true that what you've seen us do in the first set of deals that we've done this year is we have tended to go after more validated targets, but with technology that we can vet, where we have patient data that we can look at, where we think that these assets have a chance to be best in class. It is fair to say that in the first sort of set of deals that we've done, they've been more focused on, let's say, validated targets with less biological risk, but where, because we've assessed that, as Uğur talked about in his speech, that a number of these markets are about to really, the standard of care is really about to be reset. So here we think we have an opportunity to actually take part in that directly, and we think that's a very unique time-bound opportunity. But I think in totality, going forward, you can expect us to pursue both, a range of both. Novel targets are also, we're also looking at those as well as validated. Hi, can you hear me? Great. This is Stephen Sloan from Goldman Sachs, for Chris Shibutani. I'll have a broad question as well as a pipeline-specific one. Kind of following on the last question, for your goal to have 10 therapies from the oncology pipeline approved by 2030, how are you thinking about the split between assets that are advanced from your internal pipeline versus those that were brought in externally through in-licensing or acquisitions? And then on your Claudin-6 CAR-T program, just wondering if you can provide more color about how you're thinking about dosing going forward, both for the monotherapy and combination with the CARVac. As you mentioned during the presentation, that you're looking at lower doses, should we assume that's similar to the DL 2 level or below that, potentially? Thank you. So I can take the first one and sort of hypothesize that, you know, to get to the 10, we have multiple routes to get to the 10+. That's the good thing. But I think as a general, let's say, a general estimate, you know, it could be 50/50. It could be 40/60, 60/40, 50/50, but the point is that we have, I think, multiple, let's say, shots on goal from where we stand today with the organic pipeline, and now we have multiple, and we'll have even more from the external innovation pipeline. The second was the second question? ... 6 CAR dose. The CAR T-cell dose. It's too early to answer that question. We are in the dose testing Phase. We can see... We are actually in all dose levels, and are backfilling the lower doses. And we can see, which is expected, that the adverse event profile is dose-CAR T-cell dose dependent. We will probably land somewhere around 1 to 10^8, but as I said, it's too early. We need a careful assessment in a larger patient population, in particular, in combination with CARVac. Okay, one more question. Not working. There it is. Good. Hi, Michael Pye from Baillie Gifford. I have two very brief questions, please, one for Ryan and one for Özlem. For Ryan, you've highlighted, obviously, the business development you've done this year in M&A. Specifically on the late-stage assets that you've acquired, clearly these are differentiated, you know, particularly as you've highlighted around safety. And these organizations could have had a choice of organizations to work with. Presumably, they had other suitors going after them. Could you help us to understand why it is that they chose to work with BioNTech? And for Özlem, on the personalized cancer vaccines, I guess I'm thinking all else equal, and particularly with the capabilities that Karim's team bring, as you add more data, you're able to refine your algorithm and gain greater accuracy and hopefully greater efficacy of your vaccines. What avenues are available to you to significantly increase the patient population that are undergoing iNeST trials, so you can build that data asset? What are the limiting factors to dramatically expanding that? Thank you. Yeah. So thanks, Michael. I think. So I think you rightly point out, it's true that a number of those deals that we executed were, in fact, competitive. And actually, in several of the cases, we were going up against Big Pharma, who may have had a an in-line product that would have made them the natural suitor for those assets, and yet we were able to come in. And I think it's, you know, I think it's a couple of factors that starts with personal relationships and to being very proactive, and really being able to forge a strong joint vision early on with some of these company founders and management teams. Which goes a long way when you're trying to win them over to co-developing a product with you, which they're gonna live with that relationship as well for the future, right? So I think that's the first thing, and we come at that. We take that really seriously. You know, who we partner with from our side is a big decision. We look at it sort of as a marriage, and we wanna make sure that we join forces with like-minded people. And so that's. And we've been able, I think, so far to find multiple partners who share the kind of vision that we have. I think the other thing is that people, a lot of these companies recognize that we bring a lot to the table in terms of novel combinations that we could bring to bear, novel technology, know-how, expertise. You know, we haven't, until very recently, had a late-stage development organization. We've now built that, or we're building it. We certainly have one now, and it's expanding its own capabilities, you know, month by month. And so we did have to convince some people that, you know, that we're up for the challenge, and that we can go head-to-head with some of the big companies like AstraZeneca and others, in terms of late-stage development and ultimately commercialization. I think, you know, what you've seen here is that we've been able to win, win, let's say, hearts and minds in that respect, largely through what we can bring to the table in terms of innovation. How would you add anything? You said everything. Your question, Michael, was how we can increase enrollment into our iNeST trials. Actually, we are interested in increasing enrollment in all our trials. That's a continuous struggle to do that. We use all the measures other companies are also using: engagement with sites, better assessment of what a site can deliver, plus also maybe unconventional approaches. We, for example, have a partnership with UK, where we have a shared goal to mobilize basically the entire NHS network of clinical centers, with a very organized cooperation of referral centers to clinical trial centers, to mobilize each and every potential cancer patient would who would be of interest for the specific iNeST and other trials we are conducting in the UK. So these sorts of approaches. Thank you. Hey, guys. This is Nasim from J.P. Morgan, for Jessica Fye. I wanna focus on the iNeST program. You have three programs right now that's been highlighted: TNBC, CRC, and PDAC. When do you think we're gonna see the first sort of definitive proof of concept data from any one of these three programs? And then to follow up on that, when do you think the first iNeST program will be approved? First one. And then lastly, do you think accelerated approval is necessary? Thank you. Twenty-five. I don't know. You don't know? 2026. The very next trial coming to a readout will be the colorectal cancer trial in the adjuvant, of course. We are anticipating a readout end of 2025, beginning 2026 year. We will see, based on the data, how convincing this data are and whether they are opening up a path to registration. This concludes today's webcast. Thank you. Okay, thank you. Thank you. Thank you.