Good afternoon. Good morning. My name is Ryan Richardson. I'm the Chief Strategy Officer at BioNTech, and I'm pleased to welcome you to our first Innovation Series event. I'm joined here today by my colleagues, Ugur Sahin, our Co-founder and Chief Executive Officer, Özlem Türeci, Co-founder, Chief Medical Officer, and Sierk Pötting, our Chief Operating Officer.
Just a couple of housekeeping items before we get started. The slides that we'll be presenting today can be found in the investor section of our website. I'd like to remind you that we will be making forward-looking statements, that you should not place undue reliance on such statements, as actual outcomes could differ from those currently anticipated. For a full description of the risks, both financial and operational, please refer to our Form 20-F, our annual report, which were filed with the SEC.
We have a full agenda today over the next couple of hours. Very excited to dig deeper into our Pipeline, our innovation engine, and the prospects for a number of our Pipeline programs. You can see here on the slide the agenda. Ugur's gonna kick things off with a welcome, and he's going to do a deeper dive into our innovation at BioNTech across a number of dimensions before going into our infectious disease Pipeline and our activities there.
We're then gonna have a short Q&A session, followed by a coffee break, and then Özlem will take over and provide an introduction to our oncology Pipeline. She'll go through our mRNA cancer vaccine Pipeline as well, touch on our protein therapeutics programs, and then proceed with the discussion around our cell therapy Pipeline and programs to address solid tumors. Finally, we'll close with a discussion of our RiboCytokine programs before some closing remarks from Ugur, and we'll have a second brief Q&A session at the end. With that, I'd like to turn things over to Ugur Sahin.
Thank you, Ryan. I would also like to welcome you. Today we call the meeting Innovation Day. It is about hardcore innovation. We would like to bring you into the machine room of BioNTech to better understand our technologies. Many of these technologies are known to you, some not. What is most important is that you get an understanding how we connect with technologies to develop our products and to follow our vision. First of all, start with our core focus.
We are immunologists for more than 30 years. We have a deep understanding of the immune system, and we believe that this deep understanding could translate into development of medicines for a number of different diseases. Why is the immune system so important? Just key facts. It consists of hundreds of billions of cells, a diversity of cell types, which have different functions. The function of the immune system impacts every tissue. It's about cell migration. It's about removal of deceased cells. It's about healing. It's about cell and cell communication.
The immune system is involved in about 80% of high medical need diseases. This is cancer. We know that cancer immunotherapy is now a cornerstone in the treatment of cancers. It's about infectious diseases. It's very clear about vaccines. It's about autoimmune diseases. Here, the challenge is to understand how can we address too much immunity. It's about cardiovascular diseases. It's about neurodegenerative diseases. These type of diseases are driven also by immune dysfunction. It's about regeneration and inflammation. We can use, we can trigger the immune system to balance cell removal or to induce cell regeneration.
What is important is that the immune system is able to establish long-term memory. This translates to the idea that once long-term memory successfully is established, immune interfering treatments might have a curative character. That's one of the most exciting aspects. The second thing is that we used this core idea of exploiting the full power of the immune system by using new technologies. BioNTech was formed by the idea of developing using new technologies that are flexible and that are versatile and that can be directly tailored according to the needs of different populations.
These are, of course, our mRNA vaccines and therapeutics. These are cell and gene therapies. But we also believe that new technologies apply to protein therapeutics and small molecules immuno modulators if you combine them with new modes of action. Our first product, Comirnaty, BNT162b2, is the first ever approved mRNA therapy. It was the fastest pharma product, which was developed within 10 months and launched. Within one year, this product was administered to more than 1 billion people. In total, 3.4 billion doses. 2 billion doses of this went into low and middle income countries.
By developing this product, BioNTech emerged as a global pharmaceutical company. We have now our products in more than 170 countries. We contributed to global health, and we contributed- also our vaccine has had a great economic impact. How was this possible? This was not just 10 or 12 months excellent performance. It was built on more than 25 years of research. This is just a timeline of some important milestones that we reached. We started our research for the technologies that we use in BioNTech already mid-1990s. We realized at that time that cancer is a heterogeneous disease, and that the idea of immunotherapy has to translate into targeting multiple tumor antigens.
We realized that we need new technologies. We came up with the idea of using mRNA. We worked more than 10 years on mRNA. We developed our first set of patents by improving the mRNA technology. We founded 2008 BioNTech, yeah, together with our investors, sharing the vision not only to have just a biotech company in Germany. Our vision was to initiate a company which could become the new Genentech of the 21st century using these new technologies. Accordingly, the seed financing was about EUR 180 million, yeah. We didn't talk about this at that time because, of course, this goal was too futuristic at that time, yeah.
We were committed from the very beginning with our shareholders to go this pathway. We had a series of successes- translating our technologies into clinical applications. The first individualized vaccine worldwide, the first nanoparticle vaccine worldwide. Nasdaq listing in 2019. Just a few months later, we started our Project Lightspeed, which resulted in the first approved mRNA vaccine. As you can see, parallel to the development of the vaccine, we were progressing. We are progressing in using our technology into different disease indications, f or example, in autoimmunity, in inflammation.
We expand our products in the oncology space, to ensure that the transformative aspect of developing a first product can be really used, to bring BioNTech to the next level. Today, BioNTech can be characterized- i t's always difficult to come up with a fact sheet that describes BioNTech, yeah- b ecause you have to deprioritize many other aspects. What we believe is what is important is we are a discovery powerhouse. It's about discovering new targets and discovering new type of molecules.
We have a diversified Pipeline of clinical stage products. We have world-class partners. We love collaboration, yeah. We love collaboration with leading pharmaceutical companies, but we also love collaboration with world-class institutes, yeah. We are now a global organization on three continents, yeah. We have more than 3,000 employees from more than 60 nationalities. Our common language here is English. We are an international organization. We are present in Europe, United States and in Asia. We have a diversified GMP manufacturing infrastructure, and Sierk will later on talk about that, yeah.
You will understand this is not just producing doses, it's really about diversification, yeah. We have a, with you, a strong shareholder basis, yeah. A good position with a balanced cash equivalence of more than $18 billion, which could give us now the opportunity to go to the next step of transformation and realize our vision. Our vision is by 2030, yeah, to become a multi-product global biotechnology leader. We are interested to expand from the disease focus that we have at the moment, which is infectious disease and cancer, and become a disease-agnostic company. We started as a technology-agnostic company, having diversified technologies. Now the next step is to become a disease-agnostic company.
We see that we can accomplish this long-term vision by two intermediate steps. At the moment, we have one marketed vaccine, and we are the market leader in COVID-19 vaccines. Our goal is to maintain and deepen our leadership. We believe that SARS-CoV-2 will stay with us at least for a decade. We believe that we have to adapt our technologies because the virus is continuing to change and continues to remain a challenge. At the first step, we want to accelerate our current stage clinical programs, which are mainly in the oncology field, towards marketing approval.
That means the key goal for the next years is really to engage into phase II registrational studies. There's at the moment five randomized trials in phase II, and the goal is next year to have multiple registrational studies intended in different oncology indications. What we also want to do is we want to expand our Pipeline from oncology and infectious disease into other disease areas like inflammatory diseases, autoimmune diseases. These are preclinical stage programs, and the next two years, you will see that several of these programs in different areas, like, for example, cardiovascular disease, will become clinical.
The midterm goal is that we have multiple product launches in the next three-five years, and of course, this can only happen if we have about five-10 IND submissions per year. This should result in 2030 in a company with multiple approved products in different indications. This is the vision and the goal, yeah. What I would like to do now is to show you, to bring you into our innovation strategy. It is based on five pillars. It's based on the deep understanding of the immune system. It's based on target discovery and characterization, the use of a multi-platform innovation engine, our digital AI machine learning competencies, and it's about manufacturing and automation technologies.
The presentation is intended to show you the elements that we use in these pillars, and why they are relevant and how they are connected. Start with the general idea. I already introduced the power of the immune system, yeah. It is involved in body-wide control of physiological and pathological mechanisms. Immune cells have access to any tissue, yeah. Having access to any tissue means they are able to interfere with any biological function, even in the CNS, yeah. That means there is no barrier for immune cells to enter body cells. That's one powerful element, yeah.
The other powerful element is the use of gene therapies like messenger RNA. mRNA is involved in essentially in all biological processes. That means we have now a pharmaceutical technology, mRNA technology, which is involved in all type of biological processes and connected to a powerful approach to interfere with disease processes. This gives an extreme power to enable to interfere with any type of physiological and pathological mechanisms. To give you just an understanding what kind- t his is a very simplified slide showing, for example, how antigen-specific immune responses are initiated with T cells, with Tregs, with antigen-presenting cells, with tumor cells as targets of the immune system as healthy cells, pathogens, antibodies, and so on.
This is extremely simplified. What is important is that there are general categories of interference. The first category is, of course, about targets. Targets are addressed by vaccines in infectious disease and in cancer. We explored the full spectrum of targets. We used different type of immune effector cells, including CAR- T cells, T cells, non-engineered cell therapies. We used immunomodulation to strengthen this cell-cell interactions, and we used also mRNA-encoded effector molecules like antibodies, but also mRNA-encoded cytokines. These are the key elements of interfering with the immune system and with immunology processes.
As you can imagine, the same system can be used also to interfere with pathological autoimmune processes. Let's go to target discovery. One of our major backbones for cancer immunotherapy is target discovery and target characterization. Why is this needed? Why don't we just use a few targets and go for universal targeted immunotherapy? That was the idea in the 1990s, and this idea is not doable because every cancer is different and every patient has a different type of cancer. This is the inter-individual diversity, and this inter-individual diversity is the biggest challenge for curing cancer, even in early stage.
The key question is: how can we address this inter-individual diversity in cancer heterogeneity? The only way to do that is to acknowledge that inter-individual diversity exists and that it provides a great opportunity to explore the individual targets for cancer immunotherapy. We started in 2012, or we showed in 2012 that this is doable on the preclinical level, identifying tumor antigens and using them for encoding vaccines, and showed that the preclinical proof of concept. Our concept was driven by the fact that mRNA can be used to deliver genetic information.
On the right side, you see the different type of human tumors, and you see that some tumors have a higher mutational load and other tumors have a lower mutational load. Even within tumors with high mutational load, for example, melanoma, you can see that a number of patients, a fraction of patients, have low mutational load. Yeah. You can't just go with one type of tumor antigens, for example, mutations, you have to consider also shared antigens. Therefore, our approach in targeting cancer is based on these two pillars, shared tumor antigens and mutations.
The concept that we have developed and which we are currently using in several clinical trials is individualized cancer immunotherapy. In the practical sense, it works in this way. We get the sample, blood sample from a cancer patient and the tumor sample from the physician. We use the sample to extract DNA and mRNA and identify the mutations by comparing DNA from healthy tissues and from tumor tissues. We use the mRNA to identify whether the mutations are expressed. In the totality, these mutations are the cancer mutanome. This is an individual pattern.
If you take cancer mutanomes from two different patients, the overlap is less than 5%. This is really an individual aspect. We develop neoantigen prediction tools to enable that we can identify and prioritize those mutations which are highly immunogenic and which can induce T cell responses that can enter tumors and recognize and kill tumor cells. This is done on person-by-person basis. Of course, these antigens, the neoantigens have to be produced as a neoantigen mRNA vaccine. We include multiple neoantigens into one mRNA vaccine. At the moment, up to 20 neoantigens per individual cancer vaccine.
Why is this at all needed? Checkpoint blockade is also able to address neoantigens. The reason why we believe that these individualized vaccines are extremely important is that checkpoint blockade addresses mainly pre-established immune responses. There is a category of neoantigens that are not spontaneously seen by the immune system. These neoantigens are called ignored neoantigens because they have certain features that don't allow them to be spontaneously recognized. The interesting aspect is, even though they are not dominant targets of spontaneous immune responses, they represent the vast majority of neoantigens.
This is the target population of neoantigens that we are using with our neoantigen vaccines. On one side, we further strengthen the immune response against neoantigens that are already recognized, for example, synergistic with checkpoint blockade. On the other side, we target ignored neoantigens, which are about 15%-25% of all mutations. This category is about tenfold larger than the other categories. What does this mean? If we would focus on the most frequent neoantigens. The most frequent neoantigen in this figure is the line at the very left side. In this case, it's p53.
It's highly mutated in many types of tumors, and therefore, it was selected by our algorithms most frequently. If you focus only on the most frequent line, then you would lose all the other types of antigens. If you focus on the first 10 most frequent, you would lose everything beyond 11. The individualized immunotherapy, it's the long tail targeting by the immune system. At the moment, we have more than 1,400 patients screened for our individualized neoantigen therapies in more than 25 indications. We have analyzed more than 1,700 samples, and most importantly, we have selected more than 12,000 neoantigens. This is not possible with any other technology.
This is the power of targeting individual neoantigens. We believe by exploiting this individual neoantigen target, we have a higher likelihood to generate meaningful immune responses translating into tumor rejection and cure. These are our targets. The third level is our multi-platform innovation technology- y ou know these technologies. These are our mRNA vaccines, our cell and gene therapies, the targeted antibodies, small molecule immunomodulators, biologics that means immune immune effector molecules that are encoded by mRNA and delivered as lipid nanoparticles and translated in the body to the effector molecules.
We have protein-based next generation immune modulators like bispecific antibodies. All of these platforms are already in clinical stage. On the left side, you see technologies for targeting of tumors. On the right side, you see immune modulators enabling to augment this immune cells. The way how we exploit our technology, this is the example for our mRNA technology toolkit, is that we don't focus just on a single mRNA format. We have in total four different mRNA formats. Of course, you know the nucleotide modified mRNA.
Of course, you know the uridine-based mRNA that we use for cancer. We don't have yet clinical trials with our self-amplifying platform and with our trans-amplifying platform, which are still in development, but come with extremely powerful potency, and will have their place, particularly when it comes to producing vaccines with extremely low doses. Just the left side, this is the payload for delivery of nanoparticles. On the right side, we have the nanoparticles itself. We are working on different type of nanoparticles.
We are doing nanoparticle research since 2010 and have a broad portfolio of formulations based on lipoplexes, lipid nanoparticles, polyplexes. The goal is to enable the delivery technologies to deliver local, intra-tumoral, tissue-specific or systemic into all bodies. Just to explain to you a little bit about which technology, what are the characteristics of these technologies. We started with our uridine mRNA technology, and the uridine mRNA technology comes with an inherent adjuvant function. That means we get T cells which are powerful and which are long-lasting, okay? We use that for our individualized neoantigen vaccines and for our shared mRNA vaccine.
We have as a second technology, the nucleotide modified mRNA technology, which is extremely useful for generating antibody responses, okay? It's a non-immunogenic vector and becomes only immunogenic if it's combined with the right lipid nanoparticle. We are using this technology to deliver infectious disease vaccines or mRNA-encoded antibodies or cytokines. The latter two are developed intravenously. We have our amplifying technologies. This is a self-amplifying mRNA. That means delivering the mRNA, the mRNA starts to amplify and has a sustained expression.
Then we have a trans-amplifying mRNA, where we deliver multiple mRNAs. The target mRNAs are not able to self-amplify. The trans-amplifying mRNA, the replicase, is able to amplify the target mRNA, but not itself. This has a key advantage of safety, but it has also the advantage that the target mRNAs that are used for this technology are short and small, yeah. We can combine multiple mRNAs to deliver with one replicase, for example, five or six different target mRNAs. Just to give you an impression why this is important.
With self-amplifying mRNA, if you deliver nucleotide-modified or uridine mRNA into the body, what happens is, after a few hours, it starts to degrade, okay? With self-amplifying mRNA, we have about 10-12 days of amplification. That means we reach a higher sustainable expression. We can go lower with the dose, about 100-500 fold lower. With the lower dose translates, of course, if you can use a lower dose, one batch of mRNA can be used for producing more doses. Our amplifying technology, which we believe has the highest potential is the trans-amplifying technology.
This is shown here how this technology works. You deliver a replicase, and you deliver one or several transgenes. After delivery, the replicase is translated and start to amplify the transgene. The transgene is amplified. After certain time, the amplification process stops. This results in a safe approach, but it results also in production of extremely large amounts of protein with low amounts of mRNA. We could show in several animal models that we can reduce the dose of mRNA 400-fold as compared to mRNA, which translates into- i f you just calculate that for 1 billion doses, we need about 30 Kgs of mRNA.
Using nucleotide-modified mRNA, reduce that factor 100, you would just need 0.3 Kg, which is a single batch. You can in principle, with this technology, if we are able to show that it works clinically, with a single batch, deliver 1 billion doses. That's the promise. We are testing that for different type of infectious diseases. This is a list of infectious diseases there where we have been able to show that this technology works in a preclinical system. Don't see this list as a potential list of preclinical candidates in development. This is just showing on which type of infectious diseases we are working to check. Yeah.
There is a lot of research ongoing on different type of infectious diseases to understand how this technology can be exploited for relevant infectious diseases. This was about mRNA technologies. Of course, we are doing deep research on lipid formulations- l ipid formulations for targeting dendritic cells. Our cancer vaccines are based on an intravenous platform. Özlem will share later on details where we deliver only about 100 mcg, 50-100 mcg of mRNA, to induce extremely strong T cell responses.
Just to give you an impression why this dose is so low. If you go to publications that people use in mice, which is a species about 33,000-fold or 3,000-fold lower weight as compared to humans. The doses used in mice for this type of T cell responses is in the range of 40 mcg. We were able, by targeting into dendritic cells, to reduce the dose to improve the tolerability to get extremely potent vaccine. This is a technology. Targeting dendritic cells means targeting the place in the body where immune response is either initiated or modified. This is relevant for cancer, infectious disease, also immune diseases.
Therefore, this targeting platform is extremely relevant for all of our applications. We know of course some limitations of lipid nanoparticles. One of the limitations is, for example, that all the LNPs that are currently used are based on PEG formulation. PEG is. In principle, for intramuscular application, it's a suitable formulation. But particularly if you want to deliver intravenously, you will encounter inter-individual variation because of low levels of anti-PEG antibodies in about 30% of subjects that receive this formulation. It is not just a fancy question of dealing with other formulations. It's a key question.
How can we replace PEG by other formulations? One of the options is instead of using a sugar molecule, which is depicted by PEG, to use peptide-based formulations. For example, polysarcosine, which is nothing but a glycine chain. That's the same. To avoid that these particles stick to each other. We have now a polysarcosine formulation, which is as potent as the PEG formulation, and we expect to bring this type of formulations in the next 12 to 18 months into clinical testing. These are- w e will, of course, later on, also talk about cell and gene therapies. We just wanted to provide you some aspects from our mRNA technology.
AI and digital technologies will become essential part of product development. It's not only about data, getting access to data, and trying to understand and to make a clue and conclusions from this data. It's really about driving and we are using machine learning and AI since we started our research in 2012, yeah. For individualized neoantigen vaccines, w e have our different platforms, the individualized mRNA vaccines, the NEO-STIM approach, where we take T cells from cancer patients and define the mutation spectrum in the patient and stimulate the T cells in vitro to be reactive with neoantigens and adaptively transfer the T cells.
We are developing neoantigen-specific T- cell receptors, identifying, for example, KRAS-specific T- cell receptors, restricted to different types of HLA molecules and thereby enabling targeting clonal mutations. There are clear examples that targeting clonal mutations with adaptive T- cell therapy could be a powerful approach. Why does this require machine learning and AI? The reason for that is that every patient has different set of mutations. That's the first complexity layer. The second complexity layer is that every patient comes with his own HLA haplotype.
Just to give you an example, a KRAS mutation is a certain mutation, is not immunogenic in every individual. It really depends on the HLA allele that the patient has. That means sometimes it's useful to select KRAS as a target, but often it is not useful because the HLA mutation of the patient does not match even though the patient has the mutation. That means we have to understand which mutations are presented. This is a very complex process because the presentation of mutations depend on many factors- w hether the mutation is present, whether the mutation is clonal, whether it's expressed, and it has to compete with other presented molecules and then it has to reach the cell surface and then induce an immune response and the peptide affinity binding to the MHC molecule.
It is a lot of biological features, and these biological features can be brought to machine learning algorithms, but you need, of course, large datasets. That's what we did at BioNTech for many years now, created millions of datasets. For example, from MHC ligandome analysis and evaluated how this the analyzed ligands that we have identified correlate with expression in the tissue, with properties of MHC binding affinity and came up with learning algorithms. We have what is now known in the field- t his is a new field in cancer immunotherapy- i t is called neoantigen prediction.
It is- i t started essentially after we published our first paper because it became now clear that we need tools, machine learning tools that allow us to identify individual targets. What did the field learn so far? That the type of mutation is relevant, whether the mutation is clonal or not, whether which position is mutated, how much the mutation is expressed, whether the mutation has some similarity with normal tissues or with infectious disease pathogens, and of course, the binding affinity. That is what we do. Yeah. That means on the left side, you see the process. We get the DNA from the patient, define the sequence from the tumor and germline. We define the mutations, we define the expression.
We use the machine learning algorithms to predict the mutations. This is the typical approach, starting with more than 20,000 exome variations, and at the end of the day, come up with a prioritized list of about 20 targets. This is not just a concept. We are using that for each individual patient. Here is a table that is generated. A table example- i t's a small piece of the table that is generated what kind of information we get. You see this is a table generated for individual patient with the first 20 ranked mutations. You see the mutation, which is ranked first, comes from the gene SNF8.
The position of the mutation, the length of the amino acid proposed for as a vaccine candidate for this mutation, whether the transcript is expressed or not, the MHC binding score, the MHC class two binding score, CD four, CD8 T cell, the coverage in the tumor, the expression in normal tissues. Okay. Many more pieces of information. Yeah. Based on these pieces of information, we select a mutation. Of course, the machine learning algorithm, yeah, continues to learn which mutations have been selected, and we feed as a feedback loop, the information that a mutation was immunogenic.
That means this is a process, which is based on the power of machine learning. It is getting better and better and better and better. Yeah? Even though if we use the same technology by just getting more data, we get a better result. Not surprisingly, we are publishing every year papers showing that the prediction is improving over time. As so the blue curve shows that we have a higher sensitivity with the improved algorithms to detect the right mutations, and with lower number of targets. We use that, of course, to ensure that the quality of mutations selected for patients are improving over time.
The technologies that we are using for cancer are in a different form applicable for infectious disease. In cancer, the question is which mutations are relevant as cancer antigens for tracking SARS-CoV-2 variants. It is relevant to understand which SARS-CoV-2 variants could be dangerous. Every day, there are about 10,000 sequences uploaded to databases. In these databases you can see, of course, the prevalent SARS-CoV-2 variants, for example, SARS-CoV-2 Omicron BA.4 or BA.3. But you can also see that these variants, within the sequence of these variants, there are additional mutations.
We detect that, and we can see whether these mutations, these additional mutations provide a benefit to binding of this new variant to the receptor, to the ACE2 receptor, and whether the mutations increase the antigen escape, the recognition by neutralizing antibodies using the machine learning algorithms. By this, we can predict which of the new sequences could become new variants. Using this technology now for monitoring, we were able to identify all of the variants early, about 10-50 days earlier than the declaration by the WHO.
This is again based on analyzing data, doing predictions, and then feeding back the feedback from experimental results about ACE2 binding, about antigen escape. The technology is still in the early stage. It's not yet perfect. We believe that it will become more and more powerful in the next two years. What I would like to show you is how this looks like. We are getting now every week a report from the system. This is a list and the top-ranked, there are sequences which are- i n this case, it's a BA.1.1 variant, which has additional mutations that are reported to be new. This, in this week, this was the highest ranked escape mutation.
We flagged this variant, and we continue to monitor that. We will see next week how this variant is evolving over time. There's much more information about this, details about the mutations, about the binding affinity, about the escape positions. This is a project that we are doing with our colleagues from InstaDeep. We have uploaded the preprint paper in the meantime. We find the technology, and we believe that the first version will go online in about two months from now. Manufacturing- it s time for Sierk Pötting.
Yeah. Thank you, Ugur. I would like to talk to you about manufacturing. As we saw, we have a breadth of technologies in-house actually. The idea of BioNTech was from the start to have like a vertically integrated company, including manufacturing. In 2009, actually, we bought the back then called EUFETS, which is now the Innovative Manufacturing Services in Idar-Oberstein, one year after foundation, because we said we need GMP manufacturing in-house.
Why is that so? Because all the technologies that we are producing, mRNA, CAR- T, et cetera, we need in-house access to technology to invent the processes, actually the scale processes, to do tech transfer in-house, not wait for slots, have cost control and very quick innovation cycles, actually. In 2009, this started. What we have now is a variety of like GMP production sites. There is bulk mRNA- w e can produce many millions, even billions of bulk mRNA, for example, for COVID. We have individualized mRNA- t hat's the process that Ugur talked about, the iNeST process, where we can produce for individual patients after sequencing, finding the right tumor antigens, finding the target selection, and making one vaccine for one patient, actually.
We have cell therapy in-house, CAR- T therapy and TCR. And, a s the latest addition to the breadth of the manufacturing facilities are modular mRNA facilities called BioNTainers that you can basically put up anywhere in the world. This is a production network that spans the world. We have about somewhat more than 1,000 manufacturing employees actually right now deployed with Marburg, the biggest one, around 750 employees. Mainz has individualized mRNA, but also will have a commercial innovative facility that's coming up, so it's about 5 minutes from our current headquarters. The building is almost ready, so the equipment will move in next year.
We are anticipating go live then in around two years. We have Idar-Oberstein that just expanded. We'll get into this. We have an East Coast facility for CAR-T, which is in Gaithersburg. You see the BioNTainers for right now in Africa. We are planning an Africa production network in Senegal, Rwanda, and South Africa, and we had a groundbreaking event in Rwanda last week actually to start the manufacturing, the building of the manufacturing plant, the BioNTainers sites in Africa.
Now a couple of scales. What actually did we achieve since foundation when we started to build our platforms actually, and we started to build our GMP facilities as well. In 2011, when we started our first iNeST, back then it was called IVAC trial, Individualized Vaccines Against Cancer. We started the first trial. The first clinical material was made in-house by our colleagues from the IMS, the Innovative Manufacturing Service in Idar-Oberstein, and we were able to basically produce around 10 patients per year. It was a highly manual process. We then built a semi-automated facility here in Mainz at our headquarters that could do 1,000 patients at full capacity per year.
The new plant facility that's five minutes away from here will have a commercial pilot launch facility capacity of 10,000. That's scaling over various orders of magnitude of capacity. The latest example of scaling we demonstrated in Marburg. The one-gram process, we were able to produce one gram per week, which is about 20,000 vaccine doses in March 2020 in Idar-Oberstein. We took the process, scaled up in late summer 2020 to a 350-gram process that then went into the Pfizer network. They scaled it up, and we implemented this in our Marburg facility.
This was the backbone, actually the backbone scale-up for the vaccine supply of 3 billion doses last year. We further scaled this up to 1.4 Kg process, actually. Basically, in future pandemics or now when it comes to booster production and so on, we have a 1.4 g- Kg process, 4x as big as the original one. We can have bulk mRNA, pure bulk mRNA, very quickly available in the future with this scale, actually. Now coming one page for each facility. Let's quickly discuss Marburg. Marburg we bought in fall 2020, late 2020, from Novartis for less than EUR 100 million. It used to be a vaccine factory turned into a biotech factory under Novartis.
What we inherited was basically eight production suites and around about 350 highly skilled employees. What we did is we transferred in record time, did a tech transfer of our now scaled process into Marburg, actually. Last year, Marburg actually contributed with almost 1.5 billion doses COVID mRNA to our network. That's the bulk mRNA. We also do the formulation in Marburg. That capacity is not quite as high, but overall, you have a really high volume mRNA hub in Marburg.
On the other hand, I mean, with these highly scaled operators, what we also did is we are transferring in all kinds of other mRNA technology for clinical production of oncology assets like FixVac, for example. Marburg is our BioNTainers development hub. Basically we have our prototype facility for BioNTainers sitting in Marburg. Currently, we are building and setting up the BioNTainers with the mRNA process formulation in bulk from Marburg into the BioNTainers, but there will be other stuff coming, for example, clinical production in BioNTainers, et cetera, et cetera. This is also our transfer hub where we send the BioNTainers into the world.
When it comes to the iNeST facility, this is the process that Ugur discussed briefly before on the with the artificial intelligence, where we pick like the 20 targets with the artificial intelligence tool and then make a vaccine. Until 2016, before we had the semi-automated facility that we have built here in Mainz, the process was highly manual, and we had like more than three months, sometimes even six months of production time, lots of hold times, et cetera. We slimmed this down to below 5 weeks actually in our current semi-automated facility. The goal is with the new commercial facility, but also with process improvements to get to the biological limit of two-three weeks, actually.
We are getting closer there and the ramp-up time, actually, it took us two years actually to get down to this. It's very encouraging because we know the process very well and we're encouraged to see the time coming down a little bit more. Now let me wait a second here. The last site to quickly call out as a site or sites is Idar-Oberstein and Gaithersburg. This is now clinical development, especially for CAR-T and TCR. We are implementing like automated processes as much as possible to reduce steps and time, have an operational model 24/7 that can really fuel our clinical trial studies in the CAR-T space, actually, where we have recently announced some data for the testicular cancer.
The BNT211, we're increasing, continuously increasing efficiency, reproducibility of the manufacturing process, being done in Idar-Oberstein and now with the Gaithersburg facility and the East Coast that we acquired from Kite last year. We're transferring this Idar-Oberstein process to Gaithersburg to fuel the U.S. studies as well. Now last but not least, it's not a site, but you could call it site, is our BioNTainers. The idea behind this is because it's now possible to have modular clean room solution with a synthetic mRNA, which is of a small scale, very complex, 50,000 production steps, many, many suppliers that have to be managed, many, many tests that have to be done- y ou can put this actually into a small, confined, modular unit.
The nice thing is it's a turnkey solution. You have to build a hall, and at the same time, while building this hall or setting up this hall with certain temperature conditions, but nothing, stellar, you can build the BioNTainer and ship it there. For example, on the right side, you see this is our Marburg prototype facility. Six weeks before this picture, there was grass. We built the hall while we built the BioNTainer. This is a picture from the Africa event when we revealed the BioNTainers, that this is the concept that we want to establish in Africa. It's a turnkey solution.
That also means we are talking to regulatory authorities to have, like, a prototype validated in Marburg and then go to the various local authorities, being in Africa, being in any other country, to say, "We have a pre-validated facility here. Let's talk about the last 10%-15% that needs to be adapted, but we don't have to start from scratch." It's because it's basically copy and paste from a facility that's running a commercial process in Marburg. What are the next steps on these BioNTainers? As I mentioned, we just broke ground in Rwanda as the first site in Africa for our three-country African network.
By end of the year, we are anticipating to have this hall up and running and the first BioNTainer shipped into this hall in Rwanda. We at the moment talking this regulatory concept through with the agencies. That means with the local authorities in Germany, with the EMA, with the African CDC, and the Rwandan local authorities to actually have a concept to get it up and running as quickly as possible.
Last but not least, we have been talking about the African network quite extensively, but of course there are other countries highly interested in having local manufacturing capacity, be it for pandemic preparedness or for clinical studies, joint ventures or collaborations with local universities, et cetera. There's more to come on the BioNTainers front. With that, this was a quick run-through of our current GMP network, actually. With that, I hand it over back to, I think, Ugur, for more on innovation.
Just two slides. Thank you, Sierk . Again, in Sierk's presentation, the key aspects that we discussed is the versatility, and the opportunity to diversify the technology for different uses. These five innovation pillars that we presented today are of course not separate pillars, but they interact with each other. Deep understanding of the immune system, targeting of different types of diseases, target discovery for as an initial start point of drug development, multi-platform innovation engine for developing the tools that can be used as medicines, AI and machine learning as a supportive tool, and manufacturing to ensure that these products can be produced for clinical stage, but also for the market.
The future vision for that is, of course, synthetic medicine. They are combined with each other. We see the technologies and the tools that we developed, and that they represent a toolbox. This toolbox can combine in any different way and can be complemented by technology collaborations and by acquisitions. We have in this toolbox the core concepts of our drug platforms, mRNA technology, cell and gene therapy, small molecules and protein therapeutics.
They should not be considered as separate tools, but they are connected with each other and with different types of technologies. For example, for using mRNA, we can use mRNA for producing antibodies, vaccines, different types of molecules. If we, for example, produce cytokines, we can combine that with cell and gene therapies. We can combine discovery, for example, of a new type of enzymatic molecules, lysins, to target bacteria and combine it with our mRNA technology to deliver this type of molecules. We will go later into details.
The important aspect is that gives us a extreme high flexibility. The second important aspect is that we can update these modules. We can continuously improve and ensure that this becomes a platform that increases in its potency to deliver different type of pharmaceuticals with better properties. The next is- I thought that we have a break now- i t's after the infectious disease. Okay, we have to continue.
People are not exhausted yet, so I think we can push out the break. The technologies and our capabilities you have heard about extensively from Ugur, and some of them were probably new for you. Obviously are therapeutic area agnostic, so that we also want to use them in infectious diseases. Three years ago, I would have said that these are new frontiers. They are not anymore after this baptism of fire of the COVID-19 pandemic and our contribution to it. We want to build on our COVID-19 vaccine leadership to address global challenges also in other infectious diseases.
We have a broad toolkit, not only of mRNA vaccine formats to do so, but we also have ribologicals, which is encoding non-vaccine protein recombinant proteins as mRNA, and we have so-called Ribolysins, you will also hear about. We have therefore not only a diverse Pipeline of next generation COVID-19 vaccines, but we also think that we can deliver breakthroughs against infectious diseases with high medical need and from difficult to target pathogens. We are already working on such non-COVID-19 vaccines and expect to launch first clinical trials and other infectious disease indications later this year.
As you know, the medical burden from infectious diseases is growing and is a global challenge. There is insufficient protection against a wide variety of known pathogens. 20% of deaths worldwide are caused by infectious diseases, with more than 10 million deaths in 2019, and we think that we can address this with mRNA vaccines, but also with RiboMabs, antibodies directed against infectious agents, but encoded by mRNA. That's not the only area. We also believe that future pandemics, pandemic threats have to be expected. There are hundreds of thousands of undiscovered viruses, which are thought to be transmissible from mammal avian hosts to humans and back.
A rapid pandemic preparedness capability, which we have built in the context of the current pandemic and are continuing to establish, will be of use. There is antimicrobial resistance. Man-made antimicrobial resistance, which is among the top 10 global public health threats, with more than 1 million deaths annually. Here we think that we can position a new technology Ribolysins, which we want to develop towards high-precision antibiotics. We think that the COVID-19 vaccine has not only validated our mRNA technology, but we have learned a lot on different levels, which will also pave the way for future mRNA products.
You know that we have developed the vaccine in 10 months. We have, in the meantime, supplied for 3.4 billion doses to be administered, and more than 1 billion people around the world have been vaccinated, which means a huge safety database for this platform. This is a foundation on which we can build also in other infectious disease areas. We also can and will benefit from the collaboration with our esteemed partner, Pfizer, a partnership which is not a transactional, but a relational one. We had already an ongoing program in influenza prior to the pandemic. Which we are continuing, and phase I data updates are expected later this year from that program.
We have extended our cooperation to shingles, and the first-in-human phase I trial will be initiated later this year. We are also well prepared for the next phase of the COVID-19 pandemic, which is the obvious immediate area on which we focus. The key drivers for this preparedness are threefold. It's pediatric use. It's variant-adapted vaccines on which we work, and it's government contracts signed for pandemic responses, which for us is just a formalization of a social responsibility which we feel to continue to have to contribute to the ongoing pandemic.
With regard to pediatric use, as you know, we worked ourselves step by step down through the pediatric age strata, and have just recently shown that our vaccine at a lower adapted dose of 3 mcg is also efficacious in the pediatric population between six months and under five years of age. We have proposed a vaccination scheme of three doses, which did not come as a surprise to us. As you know, when we first approved our vaccine, these were times where the circulating variant back then was addressable with two doses.
With Omicron coming up and in this timeframe, the pediatric trials have been conducted, the vaccine had to be upgraded to three doses in general terms, and we also see this now in the pediatric population. We have shown that for this age group the safety profile is comparable to placebo. The reactogenicity is mostly mild to moderate and short-lived, and we have a similar frequency of AESI in both the BNT162b2 and placebo group. We are working on next generation vaccine approaches which aim to provide durable variant protection, and this is a multipronged approach.
We are developing Omicron-adapted vaccines, and some of our data has been discussed yesterday at the VRBPAC, and you will hear more about that, today. We are in the context of Omicron adaptation, but also in general terms, working on mono and multivalent vaccines. Here, as a spoiler, there is no final truth. Probably what is better will depend on the circulating strain and on the type of immunity which has been established in the population by recurring cycles of vaccinations, boosters and diverse circulating and infecting variants. In addition, we have endeavors, efforts which currently are still preclinical, but will go into clinical stage, as well.
We are working on T cell enhancing vaccines. Our T cell immunity expertise makes us believe that it is very important to strengthen this part of the immune response. We fully believe that paying attention to T cell immunity in our initial vaccine work, and ensuring that our BNT162b2 is a strong T cell enhancer as well, is the reason why we still see the breadth of protection in that vaccine. We want to strengthen this. We are also working on pan-coronavirus vaccines. On the next couple of slides, I want to share with you one body of data from a larger effort which we have gone through to test our Omicron-adapted vaccines.
This is one sub-study which basically reflects what we have also observed in other studies. This all this data which has been provided to regulatory authorities and has been partly discussed at the VRBPAC yesterday. This here now is data from a study in which we have vaccinated triple BNT162b2 vaccinated individuals with a fourth dose, either of BNT162b2 30 mcg, or of an Omicron-adapted monovalent vaccine, which means basically that the Wuhan spike has been exchanged by the sequence of the Omicron BA.1 spike. At 30 and 60 mcg, this was tested. Then a bivalent vaccine combining the Omicron variant vaccine with BNT162b2, again at two doses.
This data here is from subjects who are older than 55 years of age. The data I'm showing you are the regulatory relevant serology outcomes measured one month after this fourth vaccine doses. What you can see here is that the Omicron BA.1 GMRs are consistent with what regulators want to see, namely with simple superiority for all the Omicron-adapted vaccines, the monovalent and the bivalent ones. If we look closer, we can see that this endpoint and approval criterion shows even superiority for the monovalent Omicron-modified vaccine, which against BA.1, which does not come as a surprise to immunologists.
When it comes to Omicron BA.1 sero response rate, this exceeds non-inferiority for all Omicron-containing vaccines, so also fulfills the regulatory criterion. These here are the geometric mean titers in participants without evidence of infection, again, up to one month after the study vaccination. As you can see here, again, the GMTs are clearly boosted by this fourth Omicron by all- a ctually all Omicron-adapted vaccine variants. Most pronounced by the monovalent ones.
When it comes to the reactogenicity profile, both in this older adult population as well as in participants aged 18 to 55 years of age, which have been also tested in some of those studies I have referred to. What we see is the reactogenicity is overall similar to the prototype BNT162b2 vaccine. At the 60 mcg dose level, we see across all the studies in which this has been tested, mild to moderate injection site pain, fatigue and muscle pain, which are more common compared to the 30 mcg dose.
To summarize, this data which stands for all the data we have generated across the ongoing trials. We could show that reactogenicity profile of variant vaccines is overall similar to the prototype BNT162b2 vaccine, which does not come as a surprise because the reactogenicity is expected to be strictly platform dependent. We also could show that neutralizing responses for Omicron-containing vaccines are consistent with regulatory criteria.
[Simple]- superiority for GMR and non-inferiority for sero response, and this is true for both Omicron monovalent and bivalent vaccines, and we see superiority for GMR for Omicron monovalent vaccines. In principle, the clinical development program was successful. However, what we also have to do is to contextualize this to the reality of current pandemic status and circulating strains. Ugur will enlighten us about this now.
It's okay. Thank you, Özlem. I think we all know that Omicron BA.1, which emerged in November 2021, is not anymore the dominant strain, yeah? The development of this vaccine variant is of course a success. It shows that we can develop an adapted vaccine or modified vaccine against a new variant. This is not the first time we have generated similar data for the Beta variant, for the Alpha variant, for the Delta variant, for the Alpha Delta combination variant, and the data are always the same. The data show that we get a variant-specific improved immune response as compared to wild type vaccine.
The clinical data also show that the tolerability is comparable for these different type of vaccines. The tolerability is dependent on the vaccine platform and on the dose, yeah? The challenge that we have is if we continue to do that with a development timeframe of eight months for clinical trials, yeah. If Omicron or SARS-CoV-2 continues to evolve with this high mutational rate, yeah, then we will always be when we are done with the proof, and then we will face a situation that we are dealing actually with a new variant or a new sub-lineage. Therefore, we need to come up with a new way in addressing that.
Those who have joined yesterday's VRBPAC electronically have seen how this was discussed. Yeah. Yesterday's VRBPAC approved and support the introduction of a new Omicron-adapted vaccine that contains an Omicron lineage, but it did not come up with a proposal that it should be a BA.1 or BA.4, BA.5. There was a trend, and that the committee had an implicit trend for BA.4, BA.5, a supplement vaccine, but this will be part of the discussion. If this is the case, of course, we need to come up with a strategy how to show that. To avoid that we start retesting one day would be to use preclinical data.
We know from our studies that we can perfectly recapitulate what we see in humans with regard to immunogenicity in mice. We had already seen what is going to happen with the BA.1 vaccine, and even the four changes are very, very similar. The question is how can we respond to that? The most important is of course epidemiology. Yeah. At the moment, the epidemiology is in line with BA.4 BA.5 emerging in Europe and in the United States according to the report two days ago, BA.4, BA.5 is about 38% now in the US. In South Europe, certain regions have already 70%-80%. The trend is increasing, and it is likely that BA.4, BA.5 in the next months could become the dominant strain.
The question, is BA.1 and BA.4 comparable? They have a set of mutations which overlap. BA.4, BA.5 is antigenically distinct from BA.1. It has additional mutations, and it lacks some mutations, and the mutations are positioned in regions which are missing in BA.1. BA.1 induced immune responses that was already expected from studies in individuals with breakthrough infections of BA.1 neutralized BA.4 with a lower titer. This is what we have seen, and this is what all vaccine developers report. BA.4, BA. 5 are neutralized with significant titers, but with significantly lower than BA.1. It's about three-fold lower titer.
The question is it good to anyway go for a BA.1 adapted vaccine, which is expected with this neutralizing titers to translate into a clinical activity. But on the other side, we know that the neutralizing titers drop over time. Or should we consider a BA.4 vaccine? This is a discussion that we can't decide on our own. This must be decided together with EMA, with FDA, and we have to understand what kind of requirements are needed here. What is absolutely encouraging is- t hank you, Michael. Michael just uploaded the newest slide. What is encouraging is that we have now evidence in mice, and that a BA.4, BA.5 adapted vaccine might be able to induce significantly stronger neutralizing antibody responses against BA.4, BA.5.
On the left side you see this- t hese data are from a preclinical study in mice. Mice were pre-immunized with two doses of the BNT162b2 vaccine, and then about 80 days later got a booster either with wild type, the data I'm not showing here, or with BA.1 adapted vaccine or with BA.4, BA.5 or with a combination of the wild type and BA.4, BA.5. You can see that the mice recapitulates the findings that we have now reported. In the clinical trial, we get strong titers against the Wuhan strain, against BA.1 itself, reduced titers against the BA.2 strains and BA.2.12 strains, and significantly reduced titers against BA.4, BA.5.
When mice is immunized with BA.4, BA.5 spike protein, we see a different picture with strong neutralization of BA.4, BA.5. It's about more than 10-fold higher neutralization titers. The neutralization, what is nice to see is that also neutralization of BA.1 is accomplished when we take a BA.4, BA.5 spike protein vaccine. We generated also the data for the combination. The mice data would support a Omicron BA.4, BA.5 adapted vaccine. Of course, this is something that we that did not surprise us too much, because we are following the changes. We are analyzing the mutation pattern in the different variants, and we know that BA.4, BA.5 lacks mutations which are in the N-terminal domain of BA.1.
This lacking of the mutations allows us with BA.4, BA.5 to get antibody responses against the N-terminal domain as well as against the RBD domain. We are prepared to manufacture a BA.4, BA.5 modified vaccine, and it would be most likely in the same timelines that as a BA.1 vaccine. We expect that if October is the timeline of a potential authorization and deployment of a Omicron-adapted vaccine, we could come up with a BA.1 or BA.4, BA.5 vaccine, depending on the requirements of the regulators.
This findings are shown nicely here. Here also, both the Omicron BA.4, BA.5, as well as the bivalent vaccine shows a strong neutralization of the Omicron variants, including BA.1. In the next weeks, we will have extensive discussions with the regulators. We will provide the clinical data sets that we have generated so far. We will, of course, generate CMC data for BA.1 and BA.4, BA.5. Depending on the outcome, we will be able to provide a sub-lineage specific monovalent or bivalent vaccine.
What we also, of course, do is that we continue to build our manufacturing competencies to ensure end-to-end manufacturing delivery in different regions worldwide. We have signed a pandemic preparedness contract with the German Federal Ministry of Health in April 2022, which has the main goal to reserve and maintain manufacturing capabilities to produce at least 80 million mRNA-based vaccine doses per year. We combine all the technologies that we have presented to ensure that we rapidly recognize which type of variants are of particular concern and respond, for example, with fast DNA and mRNA manufacturing competencies.
This is our COVID approach. As you know, we have started already in 2019 tuberculosis and HIV vaccine development together with, in collaboration with the BMGF. One year ago, we started our malaria vaccine program. All of these diseases are relevant in low-income countries. We expect to enter into the clinical testing of the tuberculosis and malaria vaccine end of 2022, beginning 2023. We have candidates targeting multiple antigens in these pathogens. The concept is here to address different stages of these pathogens to ensure that we enable extracellular and intracellular recognition.
The whole idea is connected to the manufacturing concept that Sierk has presented a few minutes ago. The BioNTech concept is not only about pandemic preparedness, it's mainly intended to enable production of these vaccines, malaria vaccine, tuberculosis vaccines, when they are approved on the African continent for African people. We are collaborating closely with the partners. The whole idea that we are following is that we aim with the partnership to engage the local partners, to build local talent, to ensure technology transfer, and to ensure sustainability, not only of vaccine supply, but also of the competencies and support to build ecosystem for mRNA products in Africa.
This is our vaccine approach, but we are also engaged in developing a new type of antibiotics for infectious diseases, for bacterial infectious diseases. As you know, bacterial infectious diseases are highly relevant with emergence of highly resistant bacteria involving in many type of severe diseases, chronic infections and infections of, for example, the respiratory system. The way how we are approaching this is addressing a new class of antibacterials, which is a precision antibacterial class. These are endolysins.
These are enzymatic molecules derived from phages. So as you might know, bacterial phages are the type of organisms with the highest diversity on this planet. The infection of the phages in bacteria ends in the late stage with a lytic phase. During the lytic phase, so-called lysins are produced by the phages, which are bactericidal and dissolve the bacteria. These lysins have usually different type of domains. They have an enzymatically active domain, which is cleaving the peptidoglycans. They have a specificity domain, which is based on binding to the cell walls of the respective bacteria.
They are active in Gram-positive bacteria as well as in Gram-negative bacteria. This class is highly interesting, why? They are highly potent. Their activity is often below 1 mcg per ml. They are extremely precise. They do not harm the beneficial bacteria. This is the problem, the greatest challenge in using broad-spectrum antibiotics. The goal of using an antibiotic is removing the pathogenic bacteria, but the side effect is that it depletes also the useful bacteria, yeah, and thereby create highly resistant bacteria, but also destroy the beneficial bacteria which prevent the pathogenic bacteria from invasion. Yeah.
This lysins are also effective against biofilms. One of the key problems in infections and colonization is the biofilms. Bacteria are protected below a biofilm shield, and this lysins can enter the shield, enter biofilms and degrade the biofilm, and thereby ensuring that not only these bacteria are depleted, but also other bacteria which are co-colonizing are affected. There are a number of applications and due to their specificity, they are regarded as safe. They have no off-target effects. There are no peptidoglycans with the same structure in our human organism.
What do we want to accomplish with that? As already mentioned, this is a huge spectrum of antibacterial molecules. There are millions of sequences out. Since these sequences have two domains, the enzymatically active domain and the bacterial binding domain, which are connected by a linker, you can use domain shuffling. Yeah. You can take them separately. Of course, since this variation exists, you can use library approaches to identify those which have the highest activity and those which have the highest binding specificity.
By combining that also with AI approaches, you can engineer a, you know, synthetic lysins. The first approach that we are using for that is the development of a product candidate, an endolysin, against bacterial vaginosis. Bacterial vaginosis is, in a significant fraction of cases, driven by a bacteria species Gardnerella. We have developed a lysin using domain shuffling, which is extremely specific to the Gardnerella species. Co-incubation of this lysins with the Gardnerella species within 24 hours results in 10^5-fold reduction of the species, specific depletion of the different Gardnerella species.
On the other side, the beneficial bacteria and in the case of the vaginal colonization Lactobacillus, they are not affected. That means you can selectively deplete the pathogenic bacteria without affecting the beneficial bacteria. This is, of course, something that is not possible with broad-spectrum antibiotics like tinidazole or metronidazole, yeah, which are currently used, which come with increasing resistances. We believe that this approach, when combined with the mRNA technology, it could be extremely powerful.
The vision is here to build a new category of precision antibacterials where we define the most challenging bacteria classes, for example, Staphylococcus or Pseudomonas, yeah, and develop lysins that are specific and highly potent, encode them as mRNA. We call this approach Ribolysins, and deliver them in high enough concentration to ensure that we have a sufficient pharmacokinetic exposure to lyse the bacteria and ensure that even complex diseases like, for example, endocarditis or osteomyelitis can be addressed in a meaningful manner.
This is a platform that will go into clinical testing next year. In our infectious disease portfolio, we plan following milestones- I don't want to read that. We have program updates in 2022 coming for different infectious diseases. As already mentioned, tuberculosis and malaria will go into clinical testing. We plan to have four first-in-human clinical studies for our infectious disease vaccine program. I hope this was the last slide before starting the Q&A session.
All right. It's on, right?
Yeah.
Maybe a couple. On the trans-amplifying platform, you guys haven't presented your self-amplifying data yet, but it does seem like longer, more prolonged presentation of the antigen doesn't necessarily lead to better antibody titers. It does seem like, at least from the data we've seen from the Imperial College London, the T cell response with self-amplifying seems to be pretty prohibitive.
Pardon?
The T cell, like you basically get a very high T cell response, which leads to problematic side effects.
Yeah.
When you think about the trans-amplifying platform, you're using it for flu, you're not gonna use it necessarily for COVID. Is the secret sauce here the prime-boost, or is there something outside of that that makes you more excited about this platform than, let's say, the issues you've seen with self-amplifying?
There are several aspects. We have. I didn't talk about the challenges. The key challenge is to ensure that the trans-amplification or self-amplification starts, and that we go into this logarithmic amplification phase. This is counteracted by the human innate immune system. That means many things that we researched in the last years, and we have several publications about that, is about how we counteract the immediate innate immune response to ensure that we have a sufficient amplification.
This is true for the trans-amplifying technology as well as for the self-amplifying technology. The excitement about the trans-amplifying technology is that we can separate the mRNA encoding for the replicase. We can produce that and have that already available. If we come to a situation where we not understand what is the disease target, we can just produce the target antigens, yeah, quickly at low doses. That means we have the opportunity to respond quickly to any type of pathogen with the safety element of having replicase and the target antigens separate. Yeah.
The second advantage is if you deliver, you could also imagine that you have several self-amplifying mRNAs. But the self-amplifying mRNAs are about 11-12 kb. Yeah. Every time if you take a small antigen, you have to deliver 11-12 kb mRNA. But if you have a trans-amplifying approach, you have just this 9 kb replicase plus the small target mRNAs. These are the key advantage. Once this is successful, you can come up with a vaccine that addresses one disease or several diseases or several pathogens. Yeah. The co-delivery approach is an exciting opportunity.
Got it. Just a follow-up, sorry. This is Akash from Jefferies. I realize we're virtual here. Maybe the second question. We're having difficulty interpreting the BA.4, BA.5 data that you showed at the meeting yesterday. I say that because back in January when you showed your titers for your, you know, the original strain against Omicron, you did that data after three doses, right? The data that came at the VRBPAC meeting was after four doses, and it kind of reset the baseline titers about fourfold higher, right? So if we use the baseline titers that you presented at the VRBPAC y ou're gonna have 95% efficacy, and you might have durability out to a year.
If you use three doses, our calculations are suggesting maybe something more like eight months. I guess, were there any differences in the assays that you used back in January than what you used at the VRBPAC? If you're just, you know, if you were to just give your guess for most Americans who, you know, have only had three doses, they haven't had a fourth dose. They get a BA.4, BA.5 specific variant, how long of a durability do you think would last, let's say, above 50% for symptomatic VE? Do you have a sense of that right now, or is it too early to say?
It's difficult. First of all, the individuals that have been vaccinated who had already 3 doses, there was, of course, a time delay. As you might remember, and also in the VRBPAC meeting, we presented data that even after the third dose of the vaccine, the protection against BA.1 is diminishing relatively quickly within 12 to 18 weeks. Then you have to calculate that the BA.4 titers will be an additional three-fold lower, at least three-fold lower.
Right.
The titers that we showed before were BA.1 titers. Yeah. What we see is that we can, with our Omicron-modified vaccine, increase the titers from the baseline about 8x-10 x. This is after the third dose. After the second dose, we described an even higher increase, yeah. But this is understandable because after two doses, there were almost no BA.1 titers measurable in these individuals. Yeah. That means the benefit of an additional dose depends whether it's the second, third or fourth dose. Yeah. Yeah.
The next question is coming from Chris Shibutani, who's attending via Zoom. We hand over to Chris.
Thank you very much. Can you hear me?
Yes.
Yes.
Thanks. On the question of the Omicron-adapted monovalent versus the bivalent, the data that was presented yesterday at the VRBPAC between the different mRNA vaccines and the sponsors, there was a suggestion that perhaps your Omicron-adapted delivered some evidence of stronger immunogenicity than the bivalent.
This question of what would make the most sense across the recommendations seemed to be from the panel to go with a bivalent. Is it possible that your own strategy and the data is different from the recommendation there? What do you see as being the right approach to take as we think across different age groups as well? You seem to be pursuing the 55s and younger population versus an older population, and there was some debate about extrapolating across to younger population groups.
Yeah. This is a good question as you, since you have...
If you could address, if you could comment on those two issues, that'd be helpful. Thank you.
...yeah. Since you have participated or have seen the meeting yesterday, the VRBPAC members had the joke- I am in the monovalent fraction or I am in the bivalent fraction. Yeah. So it is indeed, I had also the impression that the bivalent fraction was a little bit stronger. The reason for suggesting bivalent is just feeling a little bit more comfortable to ensure a broader immunogenicity against the wild type variants. There was also the statement that these variants Alpha, Beta, Delta, Gamma disappeared. Actually none of these variants came back.
Yeah. You can argue in the one or in the other direction. We are open to any type of decision. We have shown superiority and have accomplished the basic requirements to go with monovalent or bivalent. We have the ability to go with a BA.1 adapted or BA.4 adapted vaccine, and therefore we will remain open to any suggestions from the FDA and EMA.
We hand the question over to Matt.
Thanks. Can you hear me?
Yes.
Yeah. Okay, sorry.
Yeah.
Matthew Harris from Morgan Stanley. I guess two from me. First one on neoantigen cancer vaccines. Can you just touch on a couple topics? One, obviously tumors are very heterogeneous, and we've obviously seen previously the peptide approach. I think obviously the number of neoantigens that you're picking is higher than the peptide approach. But at the same time, there's always been this question of, well, you can sample from one section of the tumor and maybe not get the right group of neoantigens versus others.
I guess the heart of the question here is what went wrong with the peptide approach, and what are you doing either in terms of picking your neoantigens or compensating for heterogeneity of the tumor that makes you think, you're gonna get a different result than we saw there? Then I have a follow-up on a different topic.
I can start if you want so. I would not sort of contrast peptide versus mRNA approach here. Both have their might have their strengths and weaknesses. In the peptide approach, you can, in principle, represent the same neoantigens as you can do with a mRNA approach. At the end of the day, the way you select your neoantigens and your computational pipeline will define a large portion of success or failure. In the peptide approach, it's a bit challenging that you might not be able to produce each and every peptide you want to produce, and in particular, the binders to frequent HLAs are hydrophobic. That might be a challenge.
Coming to the neoantigen computational pipeline, as you have heard, we are using AI to learn and improve. We learn about our neoantigens we have selected. We see that they are immunogenic, and we learn what features make them to a success so that we can implement this back into our Pipeline. In terms of heterogeneity, this has to be addressed by a multi-antigen approach, and we are using exactly that, and making sure that the neoantigens we or neoantigen candidates we select are not from one certain clone, but are sort of dispersed across different clones within the tumor.
No, I think this is a- t here are a lot of questions related to that. We wrote a recent review about 20 pages and could write 60 more pages about that. It's really a new deep science field. The field is making huge progress in short time. For example, one of the key learnings is that clonal mutations seem to be prime targets. The clonality of the mutation can be analyzed at the level of the primary tumor by identifying, for example, DNA from different regions of the primary tumor. We have a trend to take the primary tumor sample.
Taking also the metastatic tumor sample gives us the opportunity to add additional mutations which arise in the metastatic setting. I believe the more important aspect is to acknowledge that this type of vaccines are not suitable for advanced disease. We want to position that in the early stage of disease, ideally in the post-surgical setting, where we have a primary tumor which is removed. We can analyze the primary tumor, and then we address minimal residual disease, either defined by circulating tumor DNA, which are what we do in the setting of colorectal cancer, as Özlem share more information about that, or by the disease stage, knowing that certain disease stages are associated with 60%-70% of relapses. That's how we want to position that. We believe that the superpower of this technology is in this stage.
Okay. Perfect. Thanks. The follow-up is just on COVID-19. Have the regulators expressed a view around what sort of durability is necessary? Because I think that's, you know, we've touched on it here but they didn't really touch on it yesterday if there's a threshold and how they'd want that maybe even to be proved or demonstrated. I don't know if you have any thoughts on that.
Yeah, we are collecting the data. The clinical trials are designed to collect data from different time points after vaccination, and then we can- w e will see.
I think.
Yeah.
The next question comes from Tazeen.
Thanks, Ugur. Hi, I'm Tazeen Ahmad from Bank of America. Ugur , I just wanted to get your thoughts about the upcoming data for influenza and what would you consider to be good results there? Thinking forward, would it be your intention to have the flu vaccine just be a single shot by itself? For practical purposes, is it just going to be better to combine it with the COVID vaccine?
The flu vaccine is licensed to our partner, Pfizer. I can just give my personal perspective. The personal perspective is that the vaccine, an mRNA vaccine, should be as good, ideally better, than the current flu. We will see that only if we use that in a larger clinical trial setting. The general advantage of an mRNA vaccine as compared to the conventional flu shot is that we have here the chance to address CD8 T cells, which are prime targets for intracellular recognition and removal of infected cells.
Yesterday in the session, it was also made clear that control of COVID is not only antibody-driven but T cell driven. In the T cell fraction, we need tthe CD8 T- cells. mRNA vaccines could really give us the opportunity for the first time to ensure that we have strong CD8 T cell responses and get better control, particularly in elderly people. That's what I would love to see.
I think part of the reason I'm asking is because it does seem like a good portion of the population feels like they have vaccine exhaustion. When you talk about next generation COVID vaccine, for example, i s it possible at all for development of a vaccine that would prevent symptomatic disease? Because that does seem to be a topic of conversation in real life, where all of us are boosted. Many of us have still gotten COVID, albeit not serious symptoms, but still getting it. How realistic is it that a formulation can be made just based on the knowledge that you have that can actually prevent symptoms from occurring?
As Özlem said, we are working on different approaches. We are working on technologies to improve the antibody response. We are working on technologies to improve the T cell response. We are of course considering to develop vaccines which have a broader neutralization spectrum against multiple variants. This is ongoing research. What we have to acknowledge is that we are dealing with a new pathogen. The pathogen is different than, behaves different than many things that we have seen before. We have to learn.
\We have to be patient and we have to define reasonable goals. One reasonable goal would be to prevent severe disease. Another reasonable goal would be to reduce symptomatic disease. A third ideal goal would be to prevent infection. So these are the three goals from the very beginning. I expect if we go into a scenario where we have maybe yearly vaccinations that during the winter season, the pandemic would be more relevant. Ideally, we would have high antibody titers. Ideally at that time we would have prevention of infection and prevention of symptomatic disease. Later on, after six months, it is more about prevention of severe disease.
So we have to see how this works, but this would be reasonable goals. What is another challenge we have to speak about is that the outbreaks are not associated with season. We have seen that different strains could come really even in the summer. Yeah. We have to learn. Yeah. This reflects also the discussion yesterday on VRBPAC. Everyone said this is an extremely complex challenge, and I'm happy that we are not alone. It's the whole scientific community and a number of companies working on that. I'm confident that the results that this community is generating will become better and better, and we are going to get used to the new normal.
Okay. Maybe a last question.
Yes.
Related to the durability topic, do you think that based on where we are today, people might still need to be boosted more than once a year? Knowing that the goal ultimately would be once a year, but where are we now?
That would be speculation. Yeah.
Okay. Ingrid, do you wanna ask another question? We have Daina for the last question before we go to a break.
Okay, perfect. Thank you. Yeah. Let me keep a little bit, let me go back to the neoantigen that we were talking about. Do you believe that there is a better platform to target neoantigens? I think you have been focusing on vaccination, but you also spoke a little bit about potential cell therapy. How are you looking at those, and what do you think will be the best use?
I already said that the vaccines could be ideally suited for the early stage of cancers, actually when minimal metastatic residual disease. The larger the tumor is, the quicker and the poorer the prognosis. Therefore, we need intensive approaches like, for example, adaptive T cell therapies. Where we can deliver millions of T cells directly after diagnosis. We believe that T cell therapeutics will become a relevant pillar in treating advanced cancers. There is now increasing evidence in the field, and with T cell receptor therapeutics, and with adaptive T cell therapies with neoantigen-specific T cells, and that even advanced diseases in tumor types like pancreatic cancer could be possible. That could be the positioning of the cell therapies versus vaccine.
Hi. Daina Graybosch from SVB Securities. Two more on COVID. The first one I think came up in the committee meeting yesterday for the BA.4, BA.5 neutralization. The question was: How is the prototype doing as a third or fourth dose versus the BA.1 monovalent. I don't think you showed that data. You showed it in mice, but the mice wasn't as many doses. I think the implication was, you know, if you are boosting with the prototype, these conserved epitopes, may that actually be doing a better job for these new escape variants, which we can't predict, than switching to even a BA.1 Omicron booster.
Daina, just repeat your question.
Sorry. How is the Omicron BA.1 m onovalent booster doing on protection or immunogenicity for BA.4, BA.5 versus the wild-type prototype as a third or fourth booster?
Yeah. We don't have this for the wild type for the BA.4 lineage. The data, the only thing that we can do is, we can calculate and then the total amplification of antibody responses, and we find the fraction as compared to BA.4. This is a data point which is missing. You want to know whether we have with the Omicron adapted superiority against BA.4?
Yeah. Is BA.1 superior than prototype...
Yeah.
...in immunogenicity for BA.4, BA.5?
Yeah. Absolutely. This is a data set that is being created.
Do you have a hypothesis?
Yes, but I should not speak about hypothesis. We will have data.
Got it. On the proposal that you and Pfizer had to switch to preclinical so we can move faster, it seems sensible, but I think the question is do we know enough holistically, globally, about what the right version will be. We talked about monovalent versus bivalent, but I know you guys previously have talked about two doses of a new strain. I wonder what additional information you think we really need before we could shift to preclinical.
We need the totality of information. The preclinical does not come with lack of information, as it as compared to the clinical. If we really believe that the preclinical vaccination is a surrogate, accepted surrogate for the immunogenicity that we see in the clinical testing. So far, the data really looks very convincing to that. If we accept that, then we can go with the preclinical approach. The only thing that everyone struggles is the current information of the epidemiology the same, although we when we deliver the vaccine. If we decide today to go with the Omicron BA.4 strain, we know already that Omicron BA.1 is not any more relevant.
Do we believe it will come back? This never happened. Do we believe that Omicron BA.4 in October is still the dominant strain? We don't know. What is a reasonable approach is, since mutations are continuing on the prevalent strain, it would be reasonable to be as close as possible to the currently circulating strain. If something comes up that's completely new, then it's completely new, and it was per definition unpredictable. The science, the scientific approach that I believe is reasonable is to be as close as possible to the currently epidemiologically relevant strain.
One follow-up on that. The other proposal that we heard yesterday was, let's pick the strain that gives us the best breadth, b ecause we can't predict the future. I wonder, we heard competing hypotheses.
Yeah.
From WHO and FDA on which one would give you the best breadth from. From all your input into the deep work...
Yeah.
...do you guys have a hypothesis on one of the options which gives you the best breadth of antigenic response? Is that a valid hypothesis?
I heard this proposal. However, we have to be careful. The amino acid variation space is huge, yeah. So the problem with SARS-CoV-2 mutations are that the same amino acid can be mutated in two different amino acids, yeah. Amino acid change from variant one to variant two, the same amino acid could change to a variant three, which is not cross-reactive against variant one or variant two induced antibodies. That means the idea of the breadth is fine, but we are still in a very early phase. I would agree that we need for sure a breadth now against Omicron, yeah.
The BA.4 and BA.1 share a lot of mutations. BA.1 has in the N-terminal domain some more mutations, but BA.4 acquired additional antigen mutations, so there is not a big difference. You could even consider to have a BA.1 plus BA.4 vaccine. This is a hypothetical advantage. I would really go with a more down to earth hypothesis to stick to a sequence which is very close, closely mimicking. With BA.4 we would cover all BA.2 and BA.1, BA.3 mutations, and with BA.1 we lose many mutations which are relevant in BA.2 and BA.3.
Okay, thanks a lot. Now we are having a 10 minutes break. We return back at 4:25 P.M.
You have a display there, the important picture. Okay, so now moving to our oncology Pipeline. Sierk has talked about vertical integration earlier today, and this now is about horizontal integration. We, as you know, are interested in exploiting the various immunological mechanisms of the immune system to develop immune therapies on the basis of understanding them. Which means that we have different modalities and drug classes developed to be diverse and to have different modes of actions.
This is the reason why we have this diversification which goes across mRNA-encoded cancer vaccines, CAR, TCR and non-engineered T- cell therapies, next generation immune modulators, and mRNA-encoded effector molecules, which is antibodies and also cytokines. Based on those modalities, we have established a clinical Pipeline. All of our four modalities are at clinical stage in the meantime, and many of those programs, more than one dozen clinical programs have been initiated during the pandemic. Among these programs, we have phase II programs, five of them.
Our FixVac platform, our iNeST platform, partnered with Genentech Roche, and our next generation checkpoint immune modulator platform partnered with our esteemed partners from Genmab are at phase II stage. We are conducting clinical trials there, which are with registration intention. One of the reasons why we have a diversified Pipeline is that we want to use the success factor of the human immune system itself, namely to combine modes of action, immunological modes of action. Three examples for such combinations are shown here.
mRNA cancer vaccines, if they are efficient and potent, induce de novo T cells against cancer. Efficacious induction of an anti-neoantigen specific T cells means that these T cells would be PD-1 positive. The combination with anti-PD-1, PD-L1 agents is an obvious choice to make. mRNA cancer vaccines inducing antigen specific T cells, which then proliferate, would also open the need for mRNA encoded T cell homeostatic cytokines such as IL-2 and IL-7. This is another option. Third, mRNA cancer vaccines are capable not only to induce T cells which are endogenous, but also engineered T cells in vivo and can be combined with engineered cell therapies.
This all is not only scholastically inspired heuristic principles. We are indeed already developing these compounds clinically, and some of those combinations, mode of action combinations, are in fact already happening in our clinical trials, and we will expand on that, and I will show examples for that later on in my presentation. Starting with our mRNA cancer vaccines. You have already heard a lot about that from Ugur. Our mRNA cancer vaccines are based on using our proprietary mRNA lipoplex platform for IV administration, with which we ensure that the mRNA-encoded antigen, tumor antigen ends up at the best place thinkable immunologically, namely in dendritic cells, which are resident in lymphoid compartments body-wide.
We use our backbone optimized uridine mRNA. This is not nucleoside modified, but we want to make sure by preserving the uridine that we can maximally leverage the intrinsic adjuvant activity of mRNA. Our cancer vaccines, as Ugur has already pointed out, come in two shapes. We use shared antigens, which are tumor-associated antigens, and can be used off the shelf. We also work with neoantigens, which are very interesting targets, which need an individualized approach, which we have developed with our iNeST platform.
Ugur has also already alluded on this process, namely that we have in the meantime established to start from each and every patient's unique tumor. We have a computational pipeline which is continuously improved in terms of its algorithm with what we learn and reimplement into our AI and ML- machine learning systems. We select the best neoantigen candidates, design the mRNA on demand and tailor-made, and this is what we treat patients with.
We have reported during the pandemic actually preliminary data from a large phase one trial, which we are conducting with our partner Genentech, in which we have tested this platform, BNT122, in various tumor types as monotherapy and in combination with anti-PD-L1 agent atezolizumab. This trial was really meant to explore also the feasibility across various tumor types to implement this platform into clinical practice. What we could show, and this has been reported on AACR posters, is that BNT122 is safe, has a manageable safety profile.
It is feasible to detect neoantigen candidates, engineer vaccines, and bring them back to the patient in time in various tumor indications, including melanoma, pancreatic cancer, non-small cell lung cancer, breast cancer, et cetera. We also have shown in checkpoint inhibitor sensitive and insensitive tumor types that BNT122 induces CD8 T cells at high magnitude against the neoantigen candidates we have used in our vaccines, and that these are partly also de novo induced T cells, which have not been present before.
We have also selected cases in which we could show that these CD8 T cells go, in fact, into the tumor during vaccination and infiltrate the tumors of those patients. These were, as it always is in early clinical trials, this was first in human with this platform. These were advanced, heavily pretreated tumors and tumor types, many of them cold tumors. The place we would like to position our platform is obviously much earlier in the first line setting or even in the adjuvant setting with minimal residual disease, because here the tumor burden is obviously low and vaccine-induced T cells encounter a sort of cell-to-cell combat.
Size of the tumor matters. We also expect in those earlier stages that tumor resistance mechanisms are not yet fully established and that the immune system, the T cells, are still functional and therefore immune responses are more easy to induce. We have several ongoing trials in these earlier settings now. One of these trials is in first-line advanced melanoma, a phase II open-label randomized trial where we treat advanced melanoma patients with the standard of care, which is pembrolizumab, and compare against pembro in combination with BNT122.
We expect, and this is also based on preclinical and clinical data, that we will see a synergistic activity. The T cells induced by the vaccine will be PD-1 positive and will benefit from an anti-PD-1 agent. The clinical trial is continuing and is still enrolling, and the data is expected later this year. The second trial is at an even earlier stage in colorectal cancer. There is a high medical need in the adjuvant treatment of stage two high-risk and stage three colorectal cancer.
As you know, colorectal cancer is the second deadliest cancer worldwide and very prevalent. The standard of care of stage two high-risk cancer and stage three colorectal cancer is to first remove the primary tumor and add adjuvant chemotherapy and thereafter apply watchful waiting. It is expected that a fraction of these tumors will reoccur, will relapse. Patients who are at high risk for such relapses can be identified by ctDNA, meaning by liquid biopsy, which is a marker for minimal residual disease. The duration of disease-free survival of such ctDNA-positive patients is six months.
This is the indication in which we are testing in an ongoing clinical trial our vaccine, namely instead of watchful thinking, post-adjuvant treatment, these ctDNA positive patients are treated with a repetitive vaccination with our individualized vaccine, BNT122, and also this trial is ongoing. Then there is a third trial, a small first-in-human trial, which we have reported on the ASCO meeting just recently. This again is an adjuvant setting. In this setting, patients with pancreatic ductal adenocarcinoma have been treated by our esteemed colleagues at the MSKCC. What they have done is after standard of care surgery and removing the cancer, they have treated patients with one dose of atezolizumab to prepare for vaccination.
Thereafter, patients were vaccinated with our individualized vaccine for eight doses, and after that, received the standard of care adjuvant chemotherapy with modified FOLFIRINOX, and this was closed with one additional booster vaccination with our individualized vaccine. As you know, this indication PDAC is of high unmet need. Surgery offers the only chance of cure, but patients as those we have treated in our clinical trial frequently relapse and the five-year survival rates after resection alone are 10%. In this clinical trial, we have observed that we were able to induce immunogenicity against our individualized vaccine, which was featuring a maximum 20 neoantigen candidates of the patient.
We had a very high threshold for immunogenicity, namely patients had to be positive in two assays, one of them with low sensitivity, to make really sure that we would capture the high magnitude vaccine-induced immune responses. The first assay was a T cell clonal expansion assessment by TCR V beta sequencing. We could see that eight patients responded with a strong expansion of a vaccine-encoded neoantigen T cell responses. These T cell responses were of high magnitude, between 2.9% and 10% in median of the peripheral blood cells. The second assay was an ELISpot assay.
Again, here we could show that half of the patients were responders. The ELISpot assay allows to assess how many of the encoded antigens, neoantigens are detected by the patient. You can see that on the left-hand side that in these patients, at least one high magnitude immune responses were seen at least against at least one up to eight of the neoantigens. An interesting observation was that the responders had a higher median recurrence-free survival as compared to the non-immune responders. This data is obviously very encouraging and has motivated us to plan for a randomized phase II trial, which is being developed.
The second cancer vaccine platform, which we use, is our FixVac approach, which is based on using not cancer mutations, which are the basis for neoantigen candidates, but shared tumor-associated antigens, which are not mutated, but aberrantly expressed in the tumors we want to treat. We have several ongoing clinical trials here. What we do is that we select the antigens which, with which we want to immunize in an tumor indicational or tumor type optimized way, so that for each tumor type, we have a sort of perfect combination of multiple antigens.
Multiple antigens, because on the one hand, we want to address tumor heterogeneity, and on the other hand, want to ensure that cumulatively, with all these antigens, we cover as many as possible patients in that tumor indication. For melanoma, for example, we have four tumor-associated antigens with which more than 90% of melanoma patients are covered. We also have an ongoing clinical program for HPV16-positive head and neck cancer, where the antigens are the two oncoproteins of HPV16, which is the pathogenic virus inducing these cancers.
In prostate cancer, we have five different antigens, which are prostate-specific antigens, and another set of six antigens are used for non-small cell lung cancer, in which we also will start clinical trials this year. First, to come to advanced melanoma, which is one of indications we tackle with our tetravalent vaccine. Melanoma remains the deadliest skin cancer. Even though there has been many improvements in the treatment of melanoma, in particular, with checkpoint inhibitors, which have been approved broadly in this tumor, there is still a high medical need. This high medical need is in particular in the fraction of tumors which are checkpoint inhibitor resistant or refractory.
This is the indication in which, in an ongoing clinical trial, the Lipo-MERIT trial, we have reported encouraging data. In this phase I/II trial, we have treated patients with stage 3C and stage 4 melanoma, which means that we had a mixed population of patients with non-evaluable disease, so small tumor mass, but also patients with evaluable disease, metastatic tumors with high tumor mass. In a Nature paper which we have published, we have reported data for the fraction of patients with metastatic melanoma.
What we have seen is that this subset of patients, which all are checkpoint inhibitor experienced, both in monotherapy and in combination with approved anti-PD-1 tumors, show a high rate of vaccine-induced T- cell responses. In more than 75% of these patients, these responses are strong. We also have seen durable objective response rates in this checkpoint inhibitor experience. Patients of response rate in combination with anti-PD-1 agents was 35%. We have in the meantime follow-up data from these patients who continued in this trial. As you can see here, this is follow-up data for three, four and longer years.
We can see that the duration of a vaccine-induced objective response rates is prolonged. This again is from that follow-up data package which shows tumor shrinkage which we have observed in patients receiving both monotherapy and combination with anti-PD-1 inhibitors. Again here we see a strong shrinkage rate. One of those patients has developed a late complete response. This is a prolonged effect, and it's worthwhile to continue to vaccinate these patients. We have also, in the meantime, and this is also on published data, and only published on a poster, investigated the immune responses in the non-evaluable cancer subset, so in those patients who have minimal residual disease.
What we can see here is that also in those patients, even though the tumor is not present anymore, we can induce and maintain high T cell responses. We have, encouraged by this data, an ongoing phase II randomized trial in patients with relapsed and refractory melanoma after anti-PD-1 treatment. These patients are being treated in this trial with our FixVac, melanoma FixVac, in combination with an anti-PD-1. This is a partnership with Regeneron, where we combine with their anti-PD-1 agent, cemiplimab, and have also calibrator arms in which we test the individual compounds, and this trial is ongoing.
We have FDA Fast Track designation and Orphan Drug designation for this vaccine. To summarize, we have several ongoing trials with our individualized platform, iNeST, and with our off-the-shelf vaccine, FixVac, with a number of milestones to report on this year and next year. Now to move to our cell therapies in solid tumors. We have three autologous cell therapy platforms and address novel targets with our cell therapy approaches.
One platform are chimeric antigen receptors, CAR- T cells. The second one is NEO-STIM, which is adoptive transfer of autologous cells, which have been stimulated ex vivo with their individual mutations and neoantigen candidates. The third is TCR engineered cell therapy, where we have a couple of preclinical programs with TCRs directed against a number of antigens, including KRAS and PRAME. Our forerunner program is the BNT 211, which is our Claudin-6 CAR- T cell.
Claudin-6 is a novel antigen which we have discovered many years ago. It is a carcinoembryonic antigen, so it is not expressed in normal tissues except for in organogenesis at the fetal, embryo-fetal stage, but is aberrantly switched on in a number of tumor indications, including ovarian, testicular, endometrial and lung cancer. We have engineered a CAR, a chimeric antigen receptor, which binds to this antigen with high selectivity, and have thus obtained a CAR- T cell, which is in an ongoing clinical trial. We not only test this novel CAR- T cell, but we combine it with our RNA-lipoplex vaccines, which encode for Claudin-6, so the very target of this CAR - T cells.
What we use here is our RNA-lipoplex platform. The concept is that adoptive T cells are administered and after they have engrafted, at some point, we can vaccinate with our intravenous systemic vaccine. Claudin-6 delivered by the vaccine is expressed in the resident dendritic cells of lymphatic tissue. This is the place where CAR- T cells also roam and they are thereby expanded and we want to ensure persistence of these CAR- T cells, which is one of limitations of using CAR- T cell technology in solid tumors. This trial is ongoing in a number of claudin-6 positive cancers.
We have in the meantime obtained EMA PRIME designation for this vaccine or for this approach actually, the CARVac approach in testicular cancer. We have data for 16 heavily pre-treated patients that have been assessed in this trial, and this goes, as you can see, across different cancer indications. These patients have been treated with monotherapy, meaning just the CARVac at different doses. We are still in the dose escalation phase or in combination, meaning the CAR- T cell plus the vaccine. These are patients who have been heavily pre-treated. CAR- T cell approaches are, so to say, a salvage approach in the different lines of cancer treatment.
We have observed that in the tested dose levels with and without combination with the vaccine, this approach is well tolerated with manageable toxicities. We have observed a cytokine release syndromes, which were all grade one or two, and we have not observed neurotoxicity. These are the swimmer plots of those patients whom we have treated in the different dose levels and regimens. As you can see, we have quite a number of partial responses, and also a complete response in different tumor types.
What was very interesting for us to see is that in particular testicular cancer, which is the cancer type with the highest and most homogeneous expression of our target of Claudin-6, we see quite encouraging rates of objective responses. In the higher dose level, we have observed in five treated patients, one complete response, three partial responses, and one prolonged stable disease. This is shown here again on the left-hand side, the best responses in testicular and ovarian cancer patients as well with considerable tumor shrinkage. On the right side, the spider plots, which show that we have some durability with this first prototype regimen in the dose escalation phase.
As I already pointed out, one patient with initial PR showed deepening of a response over time, resulting in a complete response. These are two examples of patients. The patient one is a patient who had several tumor masses, including in the lung. After two weeks of treatment, there were no new lesions detectable. The tumor marker was back to normal levels. This patient then went on to complete response. The second patient had a strong first initial response with multiple new lesions first disappearing. After this initial response at week 12, we observed new lesions which were growing.
We obtained an on-treatment biopsy, which showed still positivity of these lesions for Claudin-6. The target was maintained, and we redosed the patient again with the CAR- T cells, and this patient is now being followed up. Our second platform is NEO-STIM, which is an individualized neoantigen-targeted strategy that allows us to obtain tumor-specific T cells from the periphery, so from PBMCs liquidly, instead of isolating them from the tumor. The typical TIL approach, which has the limitation that you have to have the tumor which you can resect, it has to be large enough to get TILs out of this, and TILs need to be viable ex vivo.
What we do here is that we isolate T cells, obtain the neoantigen candidate profile of a patient's tumor. With these individualized neoantigens, we stimulate ex vivo, enrich for the neoantigen-specific T cells, and these are adoptively administered. This trial is ongoing in the dose evaluation phase. In PD-1 refractory metastatic melanoma and the status is recruiting. To summarize our activities in the third platform, TCR engineered T cells, we are not at the clinical stage yet. The basis for our effort here is that we have TCR discovery platform, which we have been using and extending for many years also by acquisitions and collaborations, for example, with Medigene.
We are creating a warehouse of highly selective TCRs, which at some point will be used for individualized and broad patient coverage treatments, treatment paradigms with these TCRs. Now to quickly come to our RiboCytokines. This again is a platform which leverages our mRNA toolbox. We use it to encode cytokines as mRNA. Here specifically, we are interested in T- cell homeostatic cytokines such as IL-2 and IL-7. IL-2 is a cytokine which has a high potential, we all know that, and there are efforts for decades now in making it a viable drug. The capabilities of our RiboCytokine platform allow to address a couple of drawbacks which are observed with recombinant protein cytokines.
Our RiboCytokine platform works such that we use backbone-optimized, and in this case, nucleoside-modified mRNA encoding the cytokine fused to human albumin. The albumin fusion, for example, allows an optimized PK, but also targeting to the tumor lesions and also lymph nodes. This mRNA is enveloped into a liver targeting LNP formulation for intravenous delivery. The cytokines are expressed in liver cells and are secreted. This allows us a prolonged serum half-life of the recombinant protein, also via the albumin, but also the depot effect of being expressed in the liver cells, high bioavailability, lower and less frequent dosing, and thereby lower toxicity.
One of the compounds in this Pipeline is BNT151, which is an IL-2, which encodes for an IL-2 variant albumin fusion. What we have done here is to mutate the IL-2 sequence such that the binding to IL-2 receptor alpha portion is attenuated. At the same time we have introduced mutations which increase the binding to IL-2 receptor beta portions. This IL-2 variant is designed to stimulate naive and effector T cells with low to no expression of alpha, namely CD25, without extensively triggering immunosuppressive regulatory T cells, and that this in fact biologically can be seen on the right-hand side.
On the bottom there, you also see that the combination with anti-PD-L1 agents increases the magnitude of antigen-specific CD8 positive T cells, which can be expanded by BNT151. Most importantly, we have also shown here that the CD8 positive T- cell to Treg ratio is substantially increased by BNT151, which was the objective of mutating this IL-2 sequence. In addition, we have two other cytokines, BNT152 and BNT153. BNT153 is wild type IL-2 encoded as a fusion with albumin as RNA and BNT152 is IL-7, also wild type, as a single molecule, again, a molecule which plays an important role in T- cell proliferation, survival and effector function.
The reason why we combine these two, and this is also supported by the data shown to you on the right-hand side, is that BNT152, so IL-7 stimulates recently activated antitumor T cells and regulatory T cells, so a situation which you would be interested in when you vaccinate. BNT153 IL-7 sensitizes T cells to IL-2, increases CD8 and CD4 T- cell expansion and survival, and importantly, also controls the fraction of immune-suppressive Tregs among the CD4 T cells that are stimulated by IL-2. That we control Tregs via a different path in this case. What we have observed preclinically is that combining these two with vaccines in difficult to treat mouse models leads to a strong antitumor effect.
In this case, this is gp70 in a mouse model, a vaccine which we have engineered as an RNA-lipoplex. We also see in our mouse studies that BNT152 and BNT153 are indeed complementary. Combining both of them induces the highest effect. This again is data for the combination of both with an RNA vaccine. Here, you can see that this works both with hot tumors, where T cells are already present in the tumor, while the vaccine alone induces a complete response rate of zero, combining with both cytokines results in complete responses in 10 of 11 mice. This model has been chosen to be very tough for vaccines.
This also works in cold tumors, as shown on the bottom. This is data for BNT151 also showing a strong synergistic activity of strengthening T cell vaccination. We have also combined, and this was already featured at the very beginning of this oncology part of my talk, with CAR- T cells, so with our CAR- T cell platform. Again, here we see that the magenta-colored data that we can expand CAR- T cells, which are at a subtherapeutic level, by combining both the vaccine plus, in this case, BNT151, and by this can improve CAR- T cell expansion in these mouse models, and their persistence.
This data also in a different mouse model supports this- a gain, in the magenta data package on the left, you can see that subtherapeutic amounts of CAR- T cells by repeated vaccination and BNT151 treatment can be maintained at high levels. On the right hand side, this then results in tumor shrinkage and prolonged survival of these mice. Phase I studies with both BNT151 and BNT153 and BNT152 combination are ongoing as single agents. As soon as we have the dose levels for phase II, we will start to combine with our compounds, but also with standards of care. With this, I would hand over to Ugur for closing remarks.
Thank you, Özlem. We presented you our technology toolbox today, and in the first part. In the second part, Özlem gave examples how we are using the technology package in the clinical setting. This is an ongoing process. We are with many of our products, we are in single compound clinical trials. The goal is to identify which of the products have a single compound clinical activity, and then combine them, and go towards combination therapy. I would end with the slide with which we started at the beginning, that this is our vision, based on our product Pipeline, based on the results that we are seeing. Of course, we would like to see that you continue to support us and remain partners with us. Thank you.
Thanks a lot. This is Akash at Jefferies. Maybe a few. For your Regeneron FixVac melanoma trial, you talked about, and you've had this data for a while, 33% response, but at the 100 mcg dose it goes to 50 mcg. Is that a pattern that you've seen across the board with FixVac? Like how real, quote unquote, "is that signal?" And how do we think about that when we frame your expectations, which is a win would be kind of 30% response rate, even though that seems pretty similar to the data that Bristol has generated with their LAG-3. You know, is the dose response real? And then B, is a 30% response rate clinically meaningful given the LAG-3 data already shows that with less toxicity than ipi/nivo?
I think number two, you've recently started talking about your prostate cancer FixVac antigens, and I know that was proprietary for you guys for a while. There are a few in there that I think your team personally discovered. Can you talk about expectations there? You know, what are those antigens that you've discovered, and why are you so excited about that program going forward? Then I think one thing that you guys didn't talk about today, you've talked about autoimmune and cardiology, and moving towards those areas. Those are areas where I think investors have a lot of interest in as well.
You have patents on silent mRNA, which seems to be important, especially if you're going after anti-inflammatory diseases. You recently did the Matinas collaboration on, you know, the lipid crystal formulation. What are the big barriers with using mRNA, which is, you know, naturally inducing some type of an immune response for things like autoimmune diseases and cardiology diseases? When could we start getting INDs into the clinic in those two clinical areas, for the future? Thank you.
Should I start with the cancer ones?
Yeah, please.
Your first question was about melanoma FixVac and how real the 30% is- d id I get that right?
The dose response signal that you've seen where it goes to 50% at the higher dose. Is that a pattern you've seen across the board with FixVac, where the higher dose you get, the more you improve your responses and kind of framing expectations for that trial?
Yeah. Yeah. What we see is that the dose range 25mcg -100 mcg is pretty much the same. This is what has been also previously reported for vaccines, that at some point there is a saturation. You don't get this steep dose response curve, which is reported from other types of modalities. What we see is that this is sort of the range in which we can move. We don't have so much data in other indications, so that we cannot answer the question whether in which indications and whether we can reproduce the objective response rates.
The respective trials are just ongoing- i n phase I trials which are more advanced, we have addressed adjuvant settings, so that in small trials, you cannot really. You can anyway not answer objective response rate, and you can also only in small trials, not really answer clinical activity. The important data will come from the ongoing trials. With regard to how meaningful could the 30% be, you have quoted other trials. However, we have to keep in mind that the population of really CPI refractory and resistant, so a clear and stringent definition of resistance, this is not only experience, and many trials are in CPI-experienced patients.
In that indication, which means patients have basically had everything and what remains is chemotherapy, 10% response rate. In that indication, we think that 30% is- o ur PIs tell us that 30% is quite meaningful. Your question about prostate cancer. We will have an internal readout of the first package of data later this year, and we'll learn more about the data and, in particular, also the responses we have induced. Because as you have said, these are new prostate-specific antigens which for them have not been really broadly tested before. A very important question for us is really, do we have immunogenicity and how high is this immunogenicity?
We are in very advanced metastatic castration-resistant prostate cancers. It might be difficult to see an objective response rate, and this was also not powered to see that. We might see some signals via PSA levels. We have to sort of wait for that data. Prostate cancer, we are very excited because these are new antigens and prostate cancer is really a tough indication for IO- has always been. Therefore, we think that we should invest our efforts also in this cancer. One of the next rounds of planning next steps will be to also think about combinations with other compounds from our Pipeline.
Could you repeat the other question?
Autoimmune.
Just on cardiology and autoimmune.
Yeah.
Just on cardiology and autoimmune, you have patents on silent mRNA that you didn't put in the slide decks, but- I know you're working on it. It does seem difficult. You have a platform that naturally does induce an immune response, g oing after anti-inflammatory types of diseases.
Yeah.
You know, is there anything early that you can comment about the signals you're seeing there? What are the kind of hurdles in cardiology and then also for autoimmune, for mRNA to become a successful kind of therapeutic modality?
Yeah. Yeah, absolutely. The preclinical research is ongoing for both type of indications. We have several indications in the cardiovascular disease area in mind. Yeah. We have preclinical data that we did not share so far, making good progress, and that could be one of the first indications in the field of new disease indications, which can go into the clinical testing. For the autoimmune diseases, we need to generate additional data to be 100% sure that we are not inducing harmful immune responses, but modulating immunity.
I think I'm next. Matthew Harris, Morgan Stanley. I guess two follow-ups for me. First one, you didn't talk in detail about the 4-1BB antibodies that you have in combination with Genmab. Maybe. I think we're due some sort of update maybe by the end of the year there. Maybe you could just talk about how you're thinking about those programs, what the data update might be, and also, you know, what subsets of patients you might consider those to be most active in.
Then the second one is just on the neoantigen vaccine. Sorry, I know I asked another one already, but on the neoantigen vaccine again, I guess a couple questions. On average, how many neoantigens are you able to get? Because I know the range is somewhere between five and 20, and maybe it's disease specific, so maybe if you could c omment on that. Second, just on the first-line melanoma study. I think Moderna- Merck is doing an adjuvant study there. Can you talk about your choice of first-line versus adjuvant and just how you think about obviously earlier is better and what that may mean? Thanks.
Okay. Do you want to talk about the 4-1BB or-
Yes, I can start with that one.
Yeah.
You can take the neoantigen.
Yeah.
I can take that one.
Yeah.
Actually, I just took out my 4-1BB slides because we were so out of our time schedule with oncology part. It's also among my favorites, those programs with Genmab. We are actually internally together with our partner Genmab asking exactly the questions you have just asked. For both programs, PD-L1 4-1BB and CD40 4-1BB, we will have additional indication-specific dose expansion cohorts, which will bring some data which will be analyzed, which will help us to understand in which indications do we want to develop both programs.
For the PD-1 4-1BB, as you might know, we have initiated a phase II study in checkpoint experienced non-small cell lung cancer in combination with anti-PD-1. There are additional tumor types from which we will get data soon. There might be decisions to go to further expand in other indications, and the same with CD40 4-1BB. If first-line versus adjuvant melanoma, a very difficult discussion and decision. Back then when we started our iNeST trial in first-line melanoma, the decision was already difficult. We then decided for first-line but we are now here and again discussing whether adjuvant melanoma would also be an option.
The average number of...
Do you mean average or median?
I'll take both if you wanna tell us.
The median is very close to 20. Yeah. We have just a relatively low number of cases where we can't fill the 20. The average is, I think, believable. Some tumors have 4,000, 5,000 mutations, others have 100, and some have just 10, 50. Yeah.
This is Ingrid from Kempen. I have a couple of questions on your CARVac program. First, I would like to know what you believe will be the ideal dosing schedule for the CARVac approach. You mentioned waiting for CAR- T expansion before you start it. Why not bring that even earlier to provide earlier antigen exposure? A follow-up is the data that you shared at the conference. We saw a little bit mixed information for different patients. There was not a clear dose CAR- T expansion after you gave the CARVac for all of them, or at least not to the same magnitude that you had at first dosing. How do you look at that, and what do you need to actually drive good persistence?
Ingrid, we have to learn, and we are in this learning phase. You are right- h aving the vaccine early on so that we don't even wait for the drop would be an interesting option. We are also starting to test that. You are also right that with the small number of patients, this is a living drug, yeah? With a small number of patients, we have heterogeneous data. We cannot say every time we give dose level two, we have clearly this kinetics, and after this and that day, it goes down to this and that level. It's not expected that we will be able to see that.
We have to collect data to sort of define common patterns. It might well be that it depends really on the tumor type which we are treating. It may depend on how much the patient is pre-treated. It may depend on immune cell health in general. This is something we need to investigate with a larger number of patients and understand and continue accordingly with the specific regimen of CAR- T cell transfers and vaccinations, which will then come out of those learnings.
Do you expect an update at the end of the year, right, for the program?
Yes. Yeah.
Okay, we have two more questions, one from Manos, one from Daina, before we close.
Hi. Manos Mastorakis from Deutsche Bank. This is about the iNeST pancreatic study. First of all, could you just help us interpret the median 2.9 post-vaccine T cell response, just to put it a little bit into context. Given that trial is kind of unusual trial design, could you just walk us through that rationale and how feasible this would be in the real setting potentially, to do the checkpoint inhibitor priming, vaccination, chemotherapy, and then vaccination again?
The 2.9%- it was actually 2.9%-10%, which is the fraction of neoantigen specific T cells within the entirety of circulating CD8 cells. You have to understand that this is for one of the neoantigen candidates, and each patient has received multiple neoantigen candidates. The responders, as you might recall, have had responses to one to eight antigens. In some of those patients, you will see more than 10% of the circulating T cells being induced by the vaccine. This is, yeah, quite substantial.
10%, this is what you would want to see with CAR- T cell approaches, for example, this fraction of circulating T cells being tumor-specific or antigen-specific. We are quite encouraged to see that. We have seen this also in earlier trial, so this is not surprising. The question was what is the rationale of this sophisticated approach, and the sequence of treatment. What is standard of care is surgery, and within a certain time, one follows with adjuvant chemotherapy, which is modified FOLFIRINOX.
What we have done here in this experimental trial, which was really about learning, was that we wanted to squeeze in the vaccine induction, so six cycles of vaccine prior to the chemotherapy to induce tumor-specific immune responses. Atezo was given prior to that to support the expansion of T cells and to block any PD-1, PD-L1 axis related suppression of T cells which we would induce with the vaccine. This is the reason for this sequence. One important finding actually was that this is easily feasible within the normal clinical practice flow.
Daina from SVB Securities. Two questions from me. One, now you guys have two different approaches for expanding CAR- T, the cytokine and the vaccine, or both, and I'm wondering how you're going to pick between those two approaches. The second question is on iNeST. You guys had an interesting publication of Genentech on the different innate response to your vaccines in mice versus humans, and I wonder if you could talk about that and whether there's any implications for the current programs or future programs.
I can s-
Take the first one.
Yeah, combining with cytokines or with vaccines as options for our CAR- T cells while we want to do both, at least explore both in clinical trials and then see where the data leads us. In mice, the data is pretty clear, but you know, we are working in mice and men for a long time and know the differences in the immunology and factor them in when we plan or extrapolate from mice to men. Therefore, we have to reproduce the data first in men, and this will happen once we are more advanced with those findings, et cetera.
Yeah. The study that you refer to is a mechanistic study. We made this type of observations several years ago that we understood that when we translate mRNA vaccines from the mouse to human that we have different sensitivity thresholds with regard to triggering of innate immune responses such as secretion of IL-6. We have already implemented the lower doses in the clinical setting which resulted in highly immunogenic and well-tolerated dose. The question is what drives this higher sensitivity?
With our colleagues from Genentech we found out that this is driven by an IL-1 pathway. Yeah. The part of the higher sensitivity of the human species is driven by the reduced or delayed IL-1 receptor antagonist pathway activity. What is the learning? The learning is if you understand a mechanism, you can start to play around with this mechanism, and this gives us even more room to adapt and tune our vaccines either into more innate activating vaccines, particularly addressing the type I interferon pathway, which is extremely important in the cancer setting, or further reducing the immunogenicity of mRNA LNPs to ensure that they are even better tolerated.
August Müller from Bernstein. I have a question on iNeST and basically your algorithm for neoantigen identification and how it has evolved basically in recent years. For instance, Neogen had mass spec engine, sort of. Is it currently flowing into the algorithm or what kind of, you know, updates there've been so far?
With the acquisition of our Cambridge site, we extended our ability to do mass spec analysis. Our predictions are relying on a huge database of MHC ligand data, also monoallelic data, which are much cleaner and provides better machine learning algorithms. This is one of the datasets driving the improvement of machine learning algorithms.
Maybe if you could comment a few words on different decay pathways for mRNA as you start targeting different...
Yeah.
...tissues, et cetera.
Yeah. You mean intracellular pathways? That's something that we are dealing with since 15 years. To understand how mRNA is degraded and how we can inhibit that, the 5-prime attack, the 3-prime attack, what is the inhibition of the translation, what kind of modifications in the 5-prime end are required, how long the poly-A tail should be, what kind of additional UTRs can be included. We have introduced the concept of the double 3-prime UTR. We have introduced the concept of in vivo discovery of UTR sequences. We are using library-based approaches, mRNA evolution approaches to identify mRNA sequences that allow us to have a higher stability in certain cell types. This is all ongoing research.
We have time for one last question.
Okay. Thank you, Akash from Jefferies. This is back to COVID. I apologize. It's the pandemic. The pan-coronavirus stuff, you're taking a candidate into the clinic in the back half of this year. It's, you know, it's interesting if you look at, you know- t hen it seems like when you're talking about pan-corona, you're also talking about T cell boosting, right? You know, SARS-CoV-1, 75 of the epitopes were on the spike. In the case of SARS-CoV-2, it seems to be a little more diversely s pread. It seems like you guys are thinking about maybe combining the N- protein, parts of the S- protein, and looking at the virus a bit more holistically to see what parts are really neutralizing and...
Yeah.
...both on a humoral and then also on a T cell perspective. Can you kind of sketch out if you were trying to make kind of a mutant antigen that only mRNA could make, that is a pan-coronavirus approach, what parts of the virus are you particularly interested in, from a targeting perspective?
I think we have to be careful though. The term pan-coronavirus vaccine I use, and we have also to use it to ensure that people understand what we are doing, yeah? There are certain epitopes in the RBD domain which were considered to be evolutionarily conserved. It turned out that Omicron BA.4 deleted exactly this sequence. We have to be careful. There are other domains in the C-terminal part of the spike protein, which appear to provide at least additional help for neutralization.
The question is, how can we ensure that the vaccine design and antigen choice triggers immune responses that are focused on these regions? This is part of ongoing preclinical research, yeah. I would be careful to expect that we come up with a universal pan-coronavirus. I could imagine that we provide solutions for more stable epitopes, yeah? That are more difficult to mutate, yeah, without impairing the function of the virus itself.
Thank you.
Yeah. Yeah.
Okay. With that, we'll conclude. Thank you very much.
Yeah.
Thank you, and have safe travel back, and thank you for joining us here.
Thank you.
Everybody who attended the webcast, thanks a lot.