Welcome to Moderna's third annual R&D meeting. Thank you for joining us in person in the room, and for those of you joining us by webcast, thanks for joining in. We're very pleased to host you today and to share some interesting new data about Moderna's platform that are important for patients. Before we start, let me remind you, we'll be making forward-looking statements, that making mRNA is a risky endeavor, and I invite you to check our risk factors on the SEC website or on the Moderna website. As many of you know, since we started the company, we thought that mRNA could be a very powerful new potential class of medicines. What excited us is the possibility of doing not only secreted proteins, but also intracellular and transmembrane.
When you think that the majority of a protein encoded in the human genome are coding for transmembrane and intracellular proteins, that could be a very large opportunity to help patients. We believe because mRNA is an information molecule that will have a true platform, the ability to learn from one product into the next product, which is very unique in our industry. We believe that if we invested in technology, in robotics, in IT, in manufacturing capabilities, we'll be able to go much faster into the clinic by speeding up the research process in the lab. Also going faster into the clinic because we don't have every time to reinvent how to make mRNA. We thought that over time, there'd be a very interesting CapEx advantage with mRNA, because mRNA manufacturing is very efficient.
We do not need cells to make the mRNA with very large reactors like the entire biotech industry. We make our mRNA in small reactors in a liquid phase reaction. Also very important to know about how we build the company, is the fact that it made no scientific sense to any of us that this could be a one-drug company. Because of the information nature of mRNA, it was very clear that either we're not going to be successful and we could not launch an mRNA drug, or there would be many. Because of this opportunity to help patients and to change the world, we spend a lot of time thinking about risk management. The obvious financing risk, execution risk, which are true to any company.
For Moderna, we spend a lot of time thinking, and still do to this day, thinking about technology risk around mRNA technology and biology risk. The way we built our company is to be risk across modalities. Think about modalities as different application of mRNA technology, where we use different route of administration and/or delivery systems. That's what is really unique about how we're building this company. To manage biology risk and to clear path technology risk and biology risk, we try when we can to start in a new application, in a new modality with a program that has low biology risk. Why is that? It is because if you run a clinical trial for a new technology that has never been tried by man and you have a failure, you cannot clear path.
Did that drug fail in the clinic because the technology didn't work, or because the hypothesis around the biology was incorrect? Because we care so deeply about learning, we're trying to keep those things apart as we start. If you think about it, when we started, we started with vaccines because we thought vaccine was less risky from a technology standpoint than what Tal is going to show you later, injecting IV, when we're going to do a rare disease for potentially a lifetime drug to a kid. We started with vaccines. When we started with vaccines, we didn't start with CMV, because that would have been a very high risk. We started with flu. Why? Because there's an FDA approvable endpoint for flu. It is used for seasonal flu every year.
By having this benchmark, we could know after the first phase I, are we home from a technology performance standpoint, or do we need to go back to the labs to improve our technology because we're not there yet? We're very happy because the great science our team did, that our first phase I that was run in Germany was successful. 100% of the subjects at 100 microgram dose showed more than 40 to 1 seroconversion. That was a great milestone for us that enabled us from there to go into things like CMV and personalized cancer vaccines. The other thing it enabled us to do is to invest in Norwood. We'll come back to it later with Juan. Norwood is a very important part of Moderna's story and the business model we have set up around this very unique technology.
Today, as some of you, I'm sure, have seen the press release, we're extremely pleased and proud to announce two very important new clinical milestones. The first one is, of course, around the CMV vaccine. We will talk a lot about CMV this morning, we are very pleased to announce positive interim phase I for the vaccine that we successfully immunized seronegative above the seropositive level. We boosted the seropositive subject in the study. The vaccines are generally well tolerated. We will share with you the details of why we think we can, in the near term, start a phase II, and are already preparing for phase III and we already have discussions with FDA.
I want to remind you that Moderna owns 100% of the rights of this vaccine, and we think it will be a very important product for the company. The second news that we shared this morning is the positive phase I data for our chikungunya antibody, which is the first time in the world that an mRNA has been injected by IV route to make a systemic protein. That is a big scientific milestone, and if you think about it, we made humans produce antibodies in their livers. When you think about that, this is just a scientific feat by itself. Usually, B cells make antibodies. I think my liver has never made antibody so far. That's really something very remarkable that speaks loudly about the possibilities over time of what we can do with this technology.
The important things about the antibody program as well, which was a big question in our mind, is how well will it translate from non-human primates? As I will show you, it translates very well. Before moving on, I would like to pause for a minute and to just acknowledge the scientific accomplishments of the Moderna team. CMV contains six mRNA in one vial, six mRNA molecule in each vial. This is borderline science fiction, the team made it happen. The chikungunya antibody is two mRNA in a vial that has to get into the same cells and make both protein correctly. They then have to fold up into an antibody, then to get secreting into the blood.
If you just count how many ifs I said in my sentence, a lot of things have to happen right in tiny little cells and in the body, so you can see the result we're going to show you today. This is just remarkable. It's really one thing that I think really differentiates Moderna, is the commitment we have to doing amazing science. I would like to take a moment to really acknowledge, not only the dozens and dozens of people that work in our labs, but the leadership of the science community at Moderna. Stephen, of course, Melissa, Kerry, Ern, JJ, Joe, Eric, Josh, Paulo, Andrea, and in Juan's organization in technical development, because it's one thing when Stephen and the gang invent those new technologies, then you have to be able to make it with high purity at clinical grade to go on a clinical experiment.
Trust me, this is not easy. Juan, Ari, Phil, Don, Hugh, Peter, also Nadim, Scott, and also Jim. Thank you so much for the remarkable. You guys keep amazing me, and I think we're not done yet. What have we learned in the last few years about Moderna's mRNA platform? Well, first we learned that we have invested a lot in science. We've had to publish a lot of scientific papers to share with the world of science. We build Norwood. Norwood is up and running, as Juan will give you a bit of an update. If you think about it, in the last three and a half years, the team has put 16 different molecules in clinical trials. 16. Remember, four years ago, in September 2015, we had not been in a clinic yet. I think we had one or two programming talks.
That was it. That was Moderna. That speaks loudly of the team, the platform, and the investments we have made when it was early and everybody thought this was too crazy of an idea. The team has dosed so far 1,300 humans around the world, healthy subjects and patients. This is not a body of two or three patients or subjects, in one study. It's really a very big body of data. We have also already many times over repeated injections. In the personalized cancer vaccine, in R40, the patients have gotten many, many doses of our mRNA. Not one dose, many, many doses over time. What I'm very excited about is that now if you think about it with CHIK map data this morning, for the first five modalities of Moderna, we have consistently shown that first, our mRNA is well-tolerated in humans across the board.
We have shown that the protein we encoded is active, which was not obvious. The only case where it was not happened is the Zika vaccine, the first Zika vaccine, and I will tell you a little bit about why. The other thing that is really profound is the translation from preclinical species, as you will see today again with CHIK antibody. We are very, very pleased of the progress of the company. As now it's five out of five that we've tried in the clinic that have positive results. Yes, there's still a lot of work ahead of us, but we are very, very pleased with the progress we have made as a company and what it could mean for patients over the mid to long term. Last December, we IPO-ed the company, and this was the state of the pipeline of the company.
One product in phase II, a few in phase I, and a lot in open IND or GLP tox setting. If you look at just what happened in the last nine months, we now have two products in phase II, two programs that are preparing for phase II, CMV and OX40 for ovarian. There's a long stack of products in phase I. The IND for MMA is open, we are working hard to recruit our first patients. We have GSD1A as our latest product that we moved from the labs into development. This is just in the last nine months. A lot of progress, and as we've shared many times on our quarterly calls, the company number 1 priority is to move those products as fast as we can to the market. That is really the priority number 1 of the team.
Where are we as of September 2019? Four products in phase II or preparing for phase II, 12 phase I, and now we have a very strong body of data, 10 positive phase I studies, six vaccines. Personalized cancer vaccine, as we shared at ASCO. VEGF, that AZ published in one of the Nature journal, and chikungunya antibody. We have four vaccines for large unmet medical need for which there is no vaccine on the market. Think about the impact we can have to have so many people around the world as we get those products moving forward. CMV, HMPV/PIV, RSV, and Zika. For a small biotech company that had no product in the clinic four years ago, we now have five immuno-oncology programs in the clinic. With one in phase II, and one in phase II very soon.
We are partnering, as you know, with Merck and AZ, who are some of the leaders in the IO field. We have five important rare disease that the chikungunya antibody technology, human data we're going to share with you today, is a very important de-risking. We talked about it, 1,300 humans have been dosed. The team is strong between Norwood and Cambridge, more than 800 members of the team. Norwood is a key asset for Moderna and a strong balance sheet of $1.44 billion as of the end of June. With this framing, I'm going to invite Tal in a minute. Let me share with you what we propose for the agenda this morning. As you can imagine, we're going to spend a bunch of time on CMV.
We have several guests that Tal will introduce in a minute, to talk about CMV, talk about the virus, the medical need, and we're of course, going to share with you the human data we announced this morning. I'll do something that I've not done yet in Moderna history. I'll come back on stage to talk about commercial. I was thinking about launching this product, which is quite different from talking about mice only, which was where we were a few years ago with Stephen, when we did our first collaboration. We'll do a Q&A just on CMV, because we think there's a lot of things to talk about. We'll take a small break. We'll come back to talk about immuno-oncology, and then Tal for the main dish of the day, we'll share with you the chik antibody data.
We'll talk about MMA, and then I'll do a brief conclusion. We'll do a Q&A for the second part of the presentation. With this, Tal, it's all yours.
Thank you, Stéphane. Good morning, everybody. Truly an honor and pleasure to be here this morning to start the agenda. We will start this morning talking about vaccines. We'll do a deep dive on CMV, cytomegalovirus. I promise you, if there's nothing else you take from this morning, you will understand CMV. We're fortunate to be joined here by three of the world's leading experts on CMV, Dr. Permar, Dr. Schleiss, and Dr. Riley. I've asked them to come and give you a sense of context for the disease, the unmet need, and throughout, we'll weave in the actual data that we have. I'm going to start just by a brief overview on our prophylactic infectious disease pipeline. This pipeline really, we think of it as divided in two big buckets. One is the opportunity to go after real unmet needs as they exist today.
Things like cytomegalovirus, where people are getting infected every day, and there's no vaccine out there on the market. There are clear significant unmet needs, and you can see in our portfolio, different vaccines going after them. There's going to be a disco there. The other bucket is those diseases which are pandemic threats. Of obvious concern for global preparedness, we think our platform can actually do good in that domain. That's obviously dependent on public-private partnerships, and we've been fortunate to date to have the strong support of both DARPA and BARDA on these endeavors. Why would you use an mRNA technology for vaccines? There's a number of answers, but let me give you sort of the top ones that as I joined the company almost five years ago, were sort of self-evident in potential.
The first is that our platform is essentially an mRNA and a lipid nanoparticle. On first principles, you've got a foreign nucleic acid with an lipid. Kind of looks like a virus. It's not a far technological leap to say, well, if we injected IM and we get some protein made, we could teach the immune system to then amplify that signal just like a vaccine does. In fact, because we're making the protein from within the cell, we're mimicking the way a natural infection would make protein. That's important because it not only activates the immune system, it ensures that you activate both the antibody arm and the T cell arm. You'll see examples from our platform that actually prove the point now in the clinic of the ability of this platform to both generate antibodies and T cells.
We can do a combination product. We can put more than one mRNA in a vaccine. We started with HMPV and PIV3, two viruses that cause very similar respiratory illness. We said, well, it makes sense that you would put the vaccine against both together because you're trying to immunize two causes of the same disease. Then we took it a step further with CMV. As Stéphane said, we've got six mRNA here in one vial, and I'll get to that in a minute. We've got, because of the, I think the foundation nature of synthetic biology and how we make mRNA, and frankly, I think because of our lean forward culture from the get-go, we have a faster discovery time than anything I've ever seen in pharmaceutical industry.
In fact, I joke with Stephen that he needs one-fifth the number of people that anybody else needs because every one of his people is five times as effective just because it takes them five times shorter to do discovery. That has translated for us the ability to actually go from an idea on a whiteboard to filing an IND in 12 to 18 months, routinely. Trust me, I can't hire the people fast enough in development to catch up with the opportunities we have there. That has been truly remarkable. Finally, the fact that you've got a single process and a similar way of doing all these diverse applications means that from a manufacturing standpoint, you have an agility and efficiency that is truly remarkable. If any of you have visited Norwood, or you haven't, on behalf of Juan, I invite you to.
I'll tell you, I'm a clinician. I'm a drug developer. You don't see people like me get excited about manufacturing almost ever. Come to Norwood. It is something different. If you understand what that means in terms of the ability then to launch and commercialize products and have the agility in the manufacturing space, not have that capital hurdle of needing to build a new plant every time you launch a new drug, I think that's a tremendous efficiency that's still ahead of us. With that, how have we done in the vaccine? This gives you an overview of where we are today. We started, as Stéphane said, with those that are relatively simple. Two cases of influenza vaccines where we understand the antigen, we know how high we need to hit it in terms of immunogenicity, and we've done that.
We've now published that paper, both H10 and H7. We took on additional higher unmet needs. We went after RSV with Merck, and that's worked. We may even have a better one. In 2016, in collaboration with BARDA and the big public health emergency that we all felt at the time, we went after Zika, and we did that with really not even putting the 12 months into discovery. We basically took the last sequence that CDC had described and put it into vaccine and ran with it in the clinic. Where we came up short was the level of immunogenicity at the dose that we tested wasn't quite strong enough.
At the same time as we were doing that, we put in the right 12 to 18 months preclinically and figured out, no, there's actually a much better way of doing that preclinically is at least 20 times as potent. Not only we think so, BARDA agrees with us. We've retained that collaboration. We've pushed that back up forward, in fact, within a relatively short time, we're back in the clinic in a phase I with an improved version. Anybody who's ever done drug discovery knows that it's very rare to be able to rescue a drug a year, 18 months later, because you just changed the sequence and you figured out what was wrong with it. I can't wait to see the results of that phase I. We went after more complicated antigens.
Both chikungunya and Zika are examples of viral-like proteins, so you need to make a much more complex protein that's got to be secreted, and that's the antigen. We've proven we can do this with chikungunya as a vaccine. We've disclosed those data. Further development there, again, depends on some sort of a public partnership. Finally, the two greatest unmet needs for vaccines that are wholly owned products for us are respiratory viruses, HMPV and PIV3. That worked in phase I. We've disclosed the data earlier this year, and CMV, which we're here to talk about this morning. These are more complicated applications, and you see the vaccine continues to perform as expected.
In fact, if you step back and ask yourself, well, what have you learned about the safety profile of the platform, we've now dosed well over 1,000 subjects across all these clinical trials in phase I vaccines, and the adverse event profile that we've seen is exactly what you would have anticipated from any active vaccine, whether it's a live attenuated, recombinant, subunit, adjuvanted, et cetera. You see the anticipated local reactogenicity, you see some systemic flu-like symptoms, and that's it. There's nothing magic about this being a platform. As soon as you get into the cell, you're making protein. The rest depends on the pharmacology of those proteins. Those three fundamental questions, you'll hear me revert to them again. Can we do it safely? Do we make proteins or protein active?
For a vaccine modality, starting with flu, we've now shown time and again that indeed that is the case. Without further ado, let's talk about CMV this morning. I'd like to invite Dr. Permar, one of the world's experts on this, to set the stage for us in terms of what we're looking for here. Thanks.
Thank you. I'm excited to come and speak today about one of the diseases that I think is the most important of our time to solve, and that's to prevent CMV, and especially the congenital transition of CMV. I'm an infectious disease pediatrician, and when I was in clinical training is when I really came upon this. Am I on?
Yeah, you can come in.
Okay. Am I on now?
No.
It's okay now?
No. You guys still have me on?
I think I Yeah?
We got it.
That's weird. Unless you want to get it.
No, you're good.
What I was saying is that I'm really excited to be here today because this is a disease that is near and dear to my heart. As a pediatric infectious disease physician, I am constantly bombarded with problems in our children that we see in the hospital with this virus. It is something that when I was doing my clinical training, I kept seeing child after child that was affected by congenital CMV, and I had been an HIV researcher, and I thought, who's working on this vaccine? I was disappointed by the level of activity with it at that time. I think things have really changed, and I thank Moderna for taking on this really important disease. I think it's achievable, that we can develop a vaccine.
I'm going to go over for you some of the basics about CMV, and I can also take questions as well. Starting from the viewpoint of a virologist, CMV is, in terms of virus sizes, the largest virus in the world. In comparison to a small RNA virus like influenza or HIV that only have around 10 genes, CMV has 250 genes. It's almost the size of a cell itself. It's a DNA virus as opposed to an RNA virus, which is most of the other ones that Tal talked about. Its cousins are other members of the herpes virus family. That includes herpes simplex virus, EBV that causes mono infections, and chickenpox virus. These are viruses that have been evolving with human populations for millions of years.
As they've evolved with the human host, they've evolved to evade the immune response and to be able to spread very easily from person to person. In fact, well over half of you in the room have this virus. You wouldn't know it because, generally, when you're infected, it's an asymptomatic infection. It remains latent. Once you get this virus, like all other herpesviruses, you have it forever. There is no way to cure this virus. The only way that we can cure it is to prevent it. It remains latent in host cells and really is not a problem for most individuals, except in two main cases. One being if you become immunosuppressed for any reason, such as having an organ transplant. HIV patients, before we had good HIV therapy, had major problems with CMV disease.
If you are a fetus, if you become infected with the virus as a fetus, you are apt to have lifelong brain damage from that infection. The congenital CMV disease burden is one of the major causes of long-term disability in children. It is extremely common. It happens in 1 out of every 150 live births. That's just under 1% of all babies born, and this is globally, not just in the U.S. It is the most common form of infectious causes of birth defects. When infected, about 0.7% of babies are infected, 20% of them will go on to have lifelong disabilities because of the infection, the most common being hearing loss. There can be other major effects, including neurodevelopmental delay, motor delays. There are children who can't walk, can't talk, have seizures, and just general learning disabilities.
Again, this is a virus that is contributing to the ongoing issues of neurodevelopmental delays in children, problems with learning, and it happens more in populations of poverty than it does in populations that are higher socioeconomic status. To me, it's really continuing some of the disparities that we're seeing in our economy and our society. A remarkable stat to me is that it is the cause of 25% of all of infant hearing loss. This is not a rare genetic disease that is only contributing a small proportion to infant hearing loss. This is a quarter of all infant hearing loss. With one vaccine, you could make that many babies hear again. The annual U.S. burden is estimated at $4 billion.
That would include things like cochlear implants, all of the social care that children who have neurodevelopmental delays that have been dependent on long-term care, as well as learning disabilities that we work through in schools. Some of the costs are in that $4 billion and some are hard to measure. If you look over here at the causes of pediatric long-term disabilities, CMV really rates at the top. This is the list for the U.S. It comes above fetal alcohol syndrome. It comes above things that you've heard more about, Down syndrome, spina bifida. Then come the things that we've done well at preventing, like preventing pediatric HIV. When moms take their antiretrovirals, we can prevent that. Haemophilus influenzae, which is one of the major causes of bacterial meningitis that did lead to a lot of hearing loss before we had a vaccine.
Congenital rubella syndrome, which we have a very successful vaccine that has eliminated that congenital infection in every country where that vaccine is used. Hopefully, I've convinced you that we need a vaccine for CMV. We need a vaccine that can provide protective immune responses prior to pregnancy in order to eliminate this infection. This would be sort of targeted at the adolescent timeframe. This has been a top priority for over 20 years, named by the National Academy of Medicine, yet we are still without even a lot of products that have been to late-phase trials. It really can build on the success of the rubella vaccine as an example. This is a graph of what happened to rubella infections at the time that the rubella vaccine was introduced.
Rubella vaccine was not developed to prevent the infection, which is a mild, self-limited infection in children. It was really developed to eliminate the congenital infection, which caused heart defects, hearing loss, vision loss. The red line is the number of rubella cases in the U.S. after implementation of this very successful vaccine. The blue line is the congenital rubella cases that went down in a related fashion. I am lucky to get to work with one of the developers of the measles vaccine. The measles vaccine was soon followed on by the rubella vaccine, being related viruses, Dr. Sam Katz.
He said something to me that I always remember, and it's kind of my mantra for what we can achieve with a CMV vaccine, is that when the rubella vaccine was implemented and so successful that the need for the schools of the deaf and blind decreased so much that those schools had to close because there weren't enough children to fill them anymore. That's what we can achieve with a CMV vaccine. The CMV immunology, unfortunately, it's not quite as simple as rubella, otherwise we would have that vaccine already. Unlike rubella, the CMV immunity is not completely protective. It's not protective against acquisition of a new virus, a new strain of CMV. It's also not completely protective against the congenital transmission. That makes vaccine development complicated.
What has really been complicated to the field and really was first identified about 20 years ago, that mothers with prior immunity to CMV could still pass the virus on to the baby. The majority of adults have, and women of childbearing age, have CMV. That means actually the majority of the transmission is happening in that seropositive populations throughout the world. This is just a comparison of the cases of CMV. If you had 1,000 pregnant women who were seronegative, meaning they have never been infected with the virus, or 1,000 pregnant women who had been infected with the virus prior to pregnancy, so were seropositive. The rate of acquisition in a woman who's seronegative that comes into pregnancy is somewhere between 1% and 3%.
A lot of those women are women who have older children who are toddlers, and especially in daycare, because that is where a lot of the CMV transmission happens with the saliva and urine shedding of the virus. That virus is very good at getting around a daycare room. That 1%-3% of new infections would lead to 10-30 primary infections in this 1,000 women. That then leads to a very high rate of transmission. 30%-40% of those newly infected women will pass the virus on to the baby. This leads to somewhere between three and 12 women out of 1,000 passing CMV on to the baby. A quarter of which will go on to have defects and many of those will be long-term.
If we look on the other side, what's really interesting is that that number at the bottom is almost the same. That's confusing because if you have some immunity, shouldn't it be protective, at least partially? This is breaking down the numbers that when you have CMV prior to pregnancy, you carry the virus in your body because you don't get rid of it. There's some risk that you'll reactivate the virus, maybe some virus is replicating in the blood and then is transferred across the placenta to the fetus. We don't know what that rate is. We aren't able to measure that. We don't know the contribution of reactivation to the congenital transmission that happens in people that are previously infected.
However, what has been measured is how many times does a woman who has immunity to CMV prior to pregnancy become reinfected during pregnancy. Again, natural immunity is not protective against reinfection. Because the populations who have CMV cohort together where CMV is very geographically distributed, racially distributed differentially with the Caucasian population have a lot lower seropositivity rate than Latino or African American populations. There's actually a lot of transmission that happens in a seropositive person, probably because the other people that they live their lives with. That leads to somewhere between 200 and 300 maternal infections out of 1,000 women reinfections.
We know that when we measure CMV in the baby afterwards, which we can do with a simple test of virus in the saliva, that's actually only a small proportion of those women who are reinfected during pregnancy are passing the virus on to the baby, somewhere in the order of 3%-5%. What's notable is that these numbers are tenfold different. Mark Schleiss and I wrote an article that laid this out because it is complicated to understand that even though we see the same number of infections in 1,000 seronegative versus seropositive women, that the numbers still reflect that there is partial protection from that natural immunity. What all this is saying is that while natural immunity may provide some partial protection against CMV congenital transmission, it's not completely protective.
Therefore, a vaccine has to be different than the immunity that's afforded by the virus infection itself. The most successful CMV vaccine tested to date is a glycoprotein B subunit vaccine. That's a vaccine platform that came about after live attenuated vaccines and killed vaccines. Examples of subunit vaccines would be the HPV vaccine, would be the new shingles vaccine. Those are subunit vaccines. That approach was tried with one of the proteins that's included in the Moderna vaccine, which is the glycoprotein B. That's the main receptor that the virus uses to enter a host cell. That seems like an appropriate target. The subunit vaccine made by Sanofi was added with an adjuvant, which is something that makes the immune response higher, brings in the immune cells.
They added an adjuvant that's a fairly potent adjuvant. They gave three doses to women who were recently postpartum from their delivery. The reason why they chose that population is because, like I said, women who have a toddler are much more likely to become infected. That was a way to increase the potential risk of acquisition. There were about 400 women who went into the study, and it was split by placebo or vaccine. The results were showing some protection by the vaccine. It was about 50%, just reached significance. This was a big win, but not quite high enough to go on to the next phase of clinical translation.
At the same time, the same vaccine was given to a separate population, an adolescent population, because this may be the true target of the vaccine, where you want to catch women before they go into pregnancy. The same vaccine schedule and the same vaccine was given to about 400 adolescent women, half of which got the vaccine and half got placebo. Remarkably, to me, is the same results were achieved. It didn't quite reach significance of that p-value less than 0.05, those of you who know statistics, but it still reached that right around 50% protection. Actually, the third trial that I'm not even going to show you data on is this vaccine was given to transplant patients, and the transplant patients went on to have about 50% protection against reactivation of their virus. Pretty consistent results from this one protein.
Again, just one of the two proteins that are targeted by the Moderna vaccine. In thinking about why other vaccine platforms may have failed. The virus, again, we talked about has co-evolved with the human immune system for so long that it has the ability to evade the immune response with several different mechanisms. There is this frequent exposure to high levels of virus when children are shedding the virus. In particular, they shed high amounts of virus in their saliva and in their urine. Previous attempts to the subunit vaccine were live attenuated vaccines, because that would follow from the work that was done with the measles vaccine, rubella vaccine. That type of approach was not protective against new acquisition and of course has concerns that that type of approach would have the ability to traverse the placenta as well.
The subunit vaccines that were only partially protected. There are some things that we know about immune correlates of protection, but this work is still going on. This is the type of benchmark that you really need to know whether your phase I or your phase II trial is looking on target to have a surrogate endpoint of protection. That doesn't yet exist for CMV, but there is continued work going into how we can establish that. Some of the things that have been established is neutralizing antibodies, which you'll hear about from the data from the vaccine. This has been the gold standard for virus vaccines since the beginning of vaccines. That's how a measles vaccine was developed, how a chickenpox vaccine was developed, polio vaccine, et cetera.
Neutralizing antibodies just seem to be very important, and we know it's the way that you basically prevent the virus from infecting the next cell. Neutralizing antibodies have been associated with protection against congenital transmission, as well as how well your antibody binds to its target. Some more refined data says that prevention of the virus infection of a certain type of cell, which is the epithelial cell, and the pentamer is required for entry into that type of cell. Also the glycoprotein B vaccines, in addition to the subunit trials, the glycoprotein B vaccines have been associated with protection, just in natural immunity. In studies that we did last year looking at breastfeeding babies who are exposed to the virus that's present in breast milk, because again, this virus is very good at getting out into mucosal fluids, that then is the way that it spreads.
Breast milk of CMV-positive women often has CMV in it. We studied in babies that were receiving CMV-positive breast milk from their moms, what were the antibody responses in those babies that prevented the acquisition? The glycoprotein B vaccine antibodies came up as potentially protective. Induction of T-cell responses we do think of as a potential important component, and that's because we know that people with T-cell immune deficiencies, like transplant patients, like AIDS patients, are the ones that go on to have problems with CMV disease. Just a little bit of data that we've generated from looking back at those subunit vaccines that I talked about that were partially protective.
It's actually the perfect setting in which to see what predicted if half of the vaccinees were protected, what predicted who was going to become infected versus remain uninfected. Between the two trials, the adolescent and postpartum trials, we had enough vaccinees who became infected versus didn't become infected to run a whole bunch of immune assays to figure out which one actually was different between the vaccinees who became infected versus didn't become infected. This is the type of work towards an immune correlate that's really needed to guide vaccine design. We had a finding of an unexpected antibody response that was predictive of protection in those glycoprotein B vaccine trials.
Here's a depiction of it, that the glycoprotein B protein of the virus expressed on the surface of a cell was higher magnitude in the women who were protected, that received the vaccine, versus the women who received the vaccine but were not protected. This was not the same result if you just looked at the binding to the vaccine itself. If we showed you the data from the binding to the vaccine itself, it would look the same between those two groups. This was a differentiating antibody response. In thinking about what that really means, that's really just a surrogate for what an infected cell would look like, the glycoprotein D on an infected cell.
We did that assay to see if we could repeat the same results, and did, in fact, see the similar results where the uninfected vaccinees were protected when they had higher magnitude of that binding antibody response to the cell-associated form of gB. This is the type of work that will help to guide endpoints for vaccines in the future. Why would the mRNA vaccine platform be very good for CMV? I think that for one, there have been several studies now that have shown the high magnitude responses that are elicited with mRNA, as well as durable antibody responses. That's very important because we're going to be immunizing women, likely in adolescenthood, around the time they're getting their HPV vaccine, because pregnancy can happen in a wide range of ages.
The vaccine response needs to be durable, and mRNA has proven to be a durable platform. What I just showed you, that it seems important that the antibody responses are able to recognize the cell-associated form of the glycoprotein rather than the soluble form of the glycoprotein. mRNA would be very good at that because mRNA is basically relying on the human cells to express the glycoprotein. That also fits with what we've seen from our basic research. Another piece that I didn't have time to show you is we've seen there may have been some distracting epitopes on that soluble GB protein that was partially protective, where a lot of the binding antibody response went to the cytosolic portion of the glycoprotein, which is the portion of the protein inside the cell.
That's not going to be useful in preventing infecting the next cell or preventing infection of the placenta, et cetera. This type of approach, the cytosolic portion of the protein will be inside the cell like it is in a natural infection. These are the reasons why I think the mRNA vaccine is a very promising approach for CMV. In summary, I hope I've convinced you that the CMV vaccine is highly needed. I think the next two speakers will drive that home also. Natural immunity is only partially protective, but there are lessons we can learn from natural immunity, and we should keep learning. We should keep learning from all our vaccine trials, what it's telling us about what's protective. Novel platforms are needed for this. We've run through the gamut of standard platforms in CMV, and they haven't been effective enough.
Novel platforms like RNA are needed. I think it's really important that mRNA vaccine will express the glycoproteins on the surface of a cell, instead of in a soluble form. That will induce the type of immune responses that we've been able to show now seem to be protective. Thank you very much.
Thank you.
Yeah.
Keep it on for Gene, I think. Thank you, Sallie. With that, let's talk about our vaccine and introduce the data. The actual vaccine that we have, as Dr. Permar said, will encode for these proteins from within the cell, and so will express them on the cell surface. We're really talking about encoding the two receptors, the two hooks that the virus needs to enter cells. One is GB, which we saw on its own can already give us 50% protection. The other one, and I think this has been a learning of more recent years, is the pentamer complex. This is a complicated protein that the virus requires to attach to epithelial cells. These are the cells that line our mucosal surfaces. These are the cells, the first port of entry, if you will, for any viral infection.
Teaching the immune system to recognize that receptor, that hook, we believe is likely a critical component. In fact, it's one of the components that has been missing from the history of attenuated vaccines. The ability to package all six mRNAs here in one lipid nanoparticle is really what allows us to effectively introduce a vaccine that will then code simultaneously for these two antigens. Let's talk about the trial design, and I have to give a shout-out here to Dr. Lori Panther from my team. Everything I'm telling you here, I'm just the privileged speaker to actually represent the work of a large team. Dr. Panther was a Harvard investigator for many years, and when I met her, I actually convinced her, rather than running clinical trials out there in the real world, come and help me do some of this work.
I'm really grateful for that. The trial that Lori and my team designed was essentially a phase I, a typical phase I for a vaccine. As you've heard, we need to think differently about the seropositives and seronegatives. They have different levels of preexisting immunity, and what the vaccine will accomplish is different in both of these. We've got cohorts that are either seropositives or seronegatives. What we've done here is your traditional dose escalation. The dose range that we studied here was quite wide, 30, 90, and 180 micrograms. Not surprising if you step back and think of the dose ranges that I've showed you data with for our platform today. We've studied anywhere from 10 to 300 micrograms writ large in our vaccine portfolio.
At about 100 microgram is where typically an mRNA vaccine reaches its peak efficiency for most of the vaccines that we've described. This trial launched about 18 months ago, and today we're here to describe the first interim results. The trial is actually doing zero, two, and six months of immunization. What I'm going to show you is the first set of data, the majority of which is after the second dose, so at month 3, a month after the second dose. Okay. These are the data. As a way to think about this, these are the seronegative subjects. These are the people who had no pre-existing immunity, and you can see the initial baselines are pretty close to zero, sort of below the limit of detection. Where seropositives are in these numbers is around the 5,000. What does that mean?
It means you can dilute the serum of a seropositive person about 5,000 times, you'll still retain neutralizing activity against the virus. That's the benchmark, because we know seropositives already have some immunity to the virus. Can we take the seronegatives who've got none and get them to that level? That was the goal of this vaccine. That's how you benchmark. There is no, yet, correlate of protection. What you try to do is bring those who've never seen the virus to have an immune response that's similar or above to those that have already seen the virus. We know, based on the work that's been shared in others, that it's the neutralizing activity against both epithelial cells and fibroblasts is the likely correlate that's meaningful to be able to protect these people. How do we do?
If you look at the seronegatives, we've actually got them to the level and above those of the seropositives. In fact, three to fivefold higher than the seronegatives after two doses. You can see here the prime and boost, that 30,000 is about fivefold or higher versus the 5,000 I've described. That's how we did against this complex antigen, the protein. How did we do against GB? We got them there as well. We got them to the level of the seropositives. That's the component that we already know in and of itself has 50% of protection. This is really the primary goal here, to show that we can take the people who are seronegative and boost their immune response to the level at or above that of a seropositive subject. How did we do on the seropositives?
Well, this was really reassuring, because if you think about it, seropositives, they've already got immunity. They've seen the virus. They're relatively protected, and yet for these, we can further boost their immune response. Now, if you think of other vaccines, if somebody's already been immunized or infected, we typically think of a two to fourfold increase of that basic immunization as significant. Yet this vaccine, as measured by neutralizing titers against epithelial cells, does at a sixfold process in where these subjects are already. That's against epithelial cells. We got to the two to four range clearly against the GB. Even in subjects who've already been infected and have some preexisting immunity, this vaccine that encodes for these two epitopes can further boost their immune response and the ability then in their blood to have neutralizing activity against CMV.
How do we do in terms of safety profile? No big surprises, I think is the bottom line. We collect safety here the way you do in all typical phase I vaccines. There's an FDA guidance of how you send up questionnaires. You typically look for injection-related reactions, pain at the site. You look for systemic flu-like symptoms. They're coded, they're graded according to a known scale. What you see here in general is consistent with what we've seen to date in the rest of our vaccine portfolio. Now I'm showing you the results after the second vaccination. Typically, after the first vaccination, they were milder, and so I'm showing you data that already encompasses that. The other thing that's worth noting is it seems that the frequency and severity of the adverse events is a tad higher in the seropositives than it is in the seronegatives.
That's not a surprise if you remember the numbers I showed you in the previous graph or chart. Right. If somebody's already got an immune response that's primed and ready to see this virus, in fact, has been cohabitating with this virus for decades, and suddenly they get a boost of titers above that, I've just woken up the immune system. It's not a surprise that that waking up is manifest not just as a lab assay on the titers, but actually in some transient flu-like symptoms. We saw nothing that was unexpected. There were no vaccine-related serious adverse events. Typical to vaccine-related adverse events, these come on on the evening, and then by a day or two later, they're gone. That's the safety profile that we've seen for this vaccine. Here's another interesting data point.
When we started this trial, we actually started a group of sentinels with an initial process. We then took a four-month hit and had a better manufacturing process. Along the way, what it meant was we were actually accumulating some data in the sentinels of the very first batch that we did, and that's the dotted line here. The data I've just showed you is the solid lines on this graph, and you can see the two processes are by and large the same. For the dotted lines, these are four subjects per group. We've actually now treated vaccinated them again. We've given them the six months boost, and we follow them out to a year.
So while these are small Ns, what you can see here is that not only do we get to the levels I described at three months after the second vaccination, we can actually further boost them later and maintain those levels at least out to one year at or above the levels of an immune response of somebody who's seropositive. We've gotten now data for persistence of the immune response. If you step back, there's not a big surprise here. First of all, this is what you'd expect from a vaccine that fundamentally mimics a natural infection and how it primes the immune response. We've actually already described data for persistence for our influenza vaccine, for our chikungunya vaccine, so these are not new data.
For the application that we're talking about for this unmet need, the ability to protect a woman for a long time is actually a critical feature of being successful for the end game here. It's very reassuring to see that we can maintain those levels of protective antibody levels of seropositive out to at least a year with this vaccine. The last piece I'd say is that it's hard to tell, but this graph is actually a log scale. If you look at the dose response there, that's actually a significant difference even out to six months and a year, the difference between the blue and yellow dot is still at least a twofold difference. We've demonstrated a nice dose response curve. We've demonstrated persistence. We've demonstrated that we can actually achieve levels of immunogenicity that are at or higher where the seropositives are.
We've demonstrated that seropositives, we can even further boost them by about tenfold where they live. We're boosting them against the specific receptors that the cell needs to then attach to other cells. If you think about the CMV genome that has all these other 250 genes, some of them wreak havoc with your immune system and actually dampen the immune response. We're actually going after the specific required antigen and are further boosting that. What's next? This was really exciting and reassuring to us to see as a result of phase I. Of course, the question is, okay, how do you get to the end game here? I think that the immediate next step for us is to go and confirm the dose, and confirm our pivotal manufacturing process in a phase II setting.
You'll see us in the near term launch a phase II study that's really honing in on the dose range, that's testing that pivotal process that Juan will talk about in a minute. We're already well on our way. We've submitted this protocol for review to the agency. We've picked the sites, and I look forward to this trial starting shortly. This is also designed to give us a relatively rapid readout. We'll make decisions based on the immunogenicity we will see at the 3-month time point, again, a month after the second dose. What's the end game here? Well, if you want to get to the ability to improve this unmet need to protect babies from infection from women who get infected when they're pregnant, maybe even have a benefit in the seropositive, the question is how do you design that pivotal trial?
I think this is another area where we're fortunate that there has been some movement in recent years. Historically, if you read the literature, there's been a lot of discussion on, well, if you want to say that you're preventing congenital infection, you have to go prove it. To go prove it, you actually have to go immunize tens of thousands of women and wait for them to get pregnant and collect enough cases, et cetera. Where I think we're in a fortunate place is after having talked about this for a number of years, we recently asked FDA for pretty specific guidance in a Type C meeting, what do you think about the approvable endpoint? The advice back was that you should consider the endpoint of preventing infection, just preventing primary infection in women of childbearing age.
If you can show that, and you can show that with a safe and tolerable profile, that could be the basis for licensure. That's really a game changer for us, because now if you step back and think about it, okay, yes, the end game is ultimately showing we prevent congenital infection, we'll probably go do that, but we can do that in the post-licensure setting with real-world evidence. We can actually in a phase III, just show prevention of infection. That means that the phase III can probably be done with about 8,000 subjects or maybe even less if you do the math, and we're still in the planning stages of here. It's a phase III trial that now looks feasible for a company like us.
We're super excited that we actually have a way forward all the way to licensure here and a basis for that. We've actually started the prep work to understand what the feasibility for that trial would look like. As part of that prep work, let me invite my colleague, Juan, to talk about where we are in terms of manufacturing.
Maybe talk. Okay, try them on. Good morning, everyone. My name is Juan Andres. I have responsibility for technical development, manufacturing, and quality for Moderna. I've been in the industry for a little bit over 30 years. I've worked for a dedicated company for 18 years in a number of different countries and locations around the world, and 12 years for Novartis based in Basel, in Switzerland. My last two jobs there are very consistent with what I'm doing in Moderna. I led technical development for the chromosomal synthesis there, also had quality responsibility for all the divisions across the company. The last job was leading manufacturing operations with 72 manufacturing sites all around the world, 35,000 people. For the last two years, I've been in Moderna. Why I'm here? The reason is very simple. I fell in love with the company.
I fell in love with the technology. I fell in love with the idea that we were attempting to do. We're trying to go and do something that has not been done before, and the opportunity to be realized, that opportunity is what brought me here. The reason I'm here today is to talk about the CMV and how we are planning to take the CMV vaccine to phase II, to phase III, and then to commercial. Okay? In order to do that, let me start with Norwood. Norwood manufacturing site is our own manufacturing site. It is located in Norwood, Massachusetts. For those of you that are less familiar, it is around 30-45 minutes away from downtown Boston, Massachusetts. It is very close to the location where we are in Cambridge. The facility was built and operationalized in 22 months.
Believe me, I've been involved with making manufacturing sites for a number of different years. That is a record time. That is a very small period of time to make a facility that can make larger molecules. We have three engines there. We don't only produce the clinical trials, like CMV and the number of products that we have in the pipeline for the clinic, but we also produce for research. We produce for Stephen's team, and we make around 500 batches of mRNA every month. Okay? The facility was finished in July last year. We produced our first GMP clinical batch in August. Since then, we've produced around 70 batches to date. That is more in one year than we had produced in previous years altogether.
That gives you an idea on how much we are accelerating and the opportunity that the Norwood site provides. Here are a few pictures of the site. We wanted to design for quality, for speed, and with tremendous flexibility. We wanted to do from scratch. It is completely digitized. It is paperless. It is designed in a way that fosters a lot of collaboration across different people. Some of you have had the opportunity to go there. For those of you that haven't, as Tal said, you're invited to come, just let me know. We finished the facility a year ago, and only a few months after we were surprised by ISPE, the International Society for Pharmaceutical Engineering. I am sure many of you are familiar with them. Basically we were awarded the Facility of the Future Award just a few months after.
We are one of the finalists for the Facility of the Year Award that is going to be awarded in a few weeks in Las Vegas. We are waiting for that. We are very proud of the site. The site is really working hard, including the personalized cancer vaccine area that it is personalized medicine even inside the site. Very proud of the team and what they have accomplished. How did we arrive into Norwood? Because we didn't have it at the beginning. Of course, in the earliest stages you don't need clinical, you don't need GMP. You start operationalizing all the areas as you move your pipeline. One of the things that we realized is we started to rely on CMOs in third parties, but they were all over the world.
They were analytical in the Pacific Coast, aseptic filling in Massachusetts or in Europe. We were plasmid in another place, mRNA in another place. We realized that integrating a number of different CMOs with the type of ambition that we had was going to be very tricky. This is the idea that led us into building the site. In this moment in time, we are putting all our growing pipelines in there. We believe that Norwood is a competitive advantage not only for CMV, but for the rest of our platforms. The reason I'm stating that is we are making all the different products and as a platform, all the products are made in a very similar way. Think about it. We have made 70 clinical batches, seven zero clinical batches in the same facility.
What we learn about one product in terms of process improvement, we can immediately apply to the next product. When we scale up, so for instance, when we go from a batch of 10 grams mRNA, because at the beginning you need small quantities and you need to go and scale to 50 grams, we don't need to be doing everything in terms of scale-up. We do it for one product and then we replicate that across the different products. That gives us a tremendous amount of flexibility, but also a tremendous amount of speed. It has been highlighted before as well that, yes, we produce large molecules, but we are not traditional biotech in that sense. I have built all over the world facilities for biotech with 10, 12,000 liter bioreactors. They are big monsters.
Only the diameters are as big as this thing, they go two or three stories high. You have several of them. That is immensely capital intensive, but at the same time it takes a lot of time to be able to go and operationalize a facility like this. We are a cell-free manufacturing operation. We have a bioreactor in the site and we call it the largest scale bioreactor. Actually, it's this size. Okay. That gives you an idea on how fast we can scale up. To put it into perspective with the CMV vaccine, we are anticipating with existing novel vaccine pricing that are already in the market or taking that into consideration, that we can produce the CMV vaccine with gross margins above 90%. Okay. That is something that will be very important for us. Are we ready?
Are we ready for phase II? Are we ready for subsequent phases? Yes, we are. The vials that you see on the screen are not taken from an image bank of lyophilized vials. These vials are real CMV vials. They are ready to go. We have produced them, and we just got all the quality control results associated with this specific batch. This is ready. Okay. This lyophilized image is going to be the one that we take to phase II, and it is the one that we intend to take to phase III and subsequently to commercialization. Norwood Site can support making that mRNA, making those LNPs. As we get ready, probably we are going to lean on partners to produce aseptic filling and finishing as we move forward. How much capacity would we have? How much capacity do you need?
The question is, we can produce $10-plus million if we take into consideration that we need to make many other products. If we dedicated the site only to a 100 microgram dose of vaccine, like CMV could be, we could be producing $100 million doses just in the site. Okay. Tal, Stéphane, we are ready to go. Tell me when and how much, and we will go for it. Okay. This is it. Let me introduce the next presenter. Dr. Mark Schleiss is coming to do the next presentation. Thank you very much. Thank you.
This is great, this image. I'm excited. We're ready to go. There it is. That is fantastic. I want to talk a little bit about cytomegalovirus vaccines and put it a little in a context for the audience. One of the take-home points I want to make is that this subunit approach that selects key CMV genes is fundamentally different than some of the vaccines we use in the clinic every day. MMR vaccine, varicella vaccine, chickenpox vaccine, these are whole virus vaccines that at some level attenuate a natural pathogen to a point where it induces immune responses that doesn't cause the disease. Now, there's a whole category of CMV vaccines that does that too, and that was alluded to a little bit earlier this morning. I'll kind of review those differences as we move forward.
I think that's a key take-home point, because one of the key things is safety. A vaccine that goes out into the market, into clinical care, has to be widely perceived as safe, particularly if you're talking about adolescent populations. We live in a culture and in a society where there's so much anti-vaccine pushback and so much skepticism and vaccine hesitancy that I think that's a key message that we have to have moving forward. We've heard about congenital CMV infection and how important it is as a cause of disability. I won't belabor these points. This is the classic photo from decades ago now of a baby with microcephaly, small head, and the CAT scan of the baby's brain showing the calcium deposits. That's one of the classic and tragic findings of congenital CMV when it involves the brain.
This is the kind of thing we're trying to prevent. One of the really exciting things about this project, in my opinion, is that from the beginning, it's been targeted at preventing congenital CMV. Many pharmaceutical ventures into the cytomegalovirus workspace have started out by saying, "Let's see if we can do something about CMV in transplant recipients." Important if you're a solid organ or bone marrow transplant recipient and get CMV disease, that's very devastating. The requirements of a vaccine and the endpoint of a vaccine program, I think, are probably very different in healthy women than they are in transplant recipients. I love the fact that this is a program devoted to congenital CMV. Sally Permar already talked about how the chief vaccine-preventable causes of brain injury, neurologic damage, deafness, have been solved in the past by vaccination.
I remember when my daughter was an infant, the Haemophilus influenzae type b vaccines had literally just been licensed days before she was born, and how anxious I was as a clinician seeing invasive Haemophilus disease every day, to find a pediatric practice in Seattle where I lived at the time that actually had the vaccine. I was calling all over town, "Do you have it yet? Do you have it yet?" That vaccine really changed pediatric care profoundly, as did the rubella vaccine, as we've already heard. I'm hopeful that a CMV vaccine will have the same impact. The total number of babies that are injured in the United States and Europe from cytomegalovirus every year is in the many thousands, and it's probably much higher than this actually in the developing world. We just don't have good data.
As we learn more about CMV in the developing world, where the prevalence of disease is actually going to be higher than the United States and Europe, and we'll come back to that point in a couple of minutes, I think globally, this can have a huge impact on child health. Does immunity protect the fetus? We've heard about this already this morning, this idea that if you get CMV and you recover from it and you go on with your life, well, you're not really quite completely protected. You reactivate the virus during the course of your lifetime. You get reinfected with new strains of the virus. That's really different conceptually than the measles, which if you recover from it, you don't get it again. You're immune. This is one of the underlying complexities of this crazy virus.
It has all of these immune modulation genes, these immune evasion genes that allow people to get reinfected. If you haven't sensed it already, I think one of the take-home messages is that eventually to protect babies against congenital CMV, we will need to give a vaccine to all women, whether they've had CMV in the past or not. That's why the data that we saw from Tal about boosting antibody titers in people who are already immune is really important, much more so than it would be for other traditional infectious disease pathogens. There is evidence, as Sally alluded to, that immunity does provide some protection. It provides a big reduction in transmission, and it probably provides a decreased severity of disease if transmission does occur. We've already heard about the economic benefits of CMV vaccination and how important those are.
I'll just sort of lay out four points that I think are important in CMV vaccines. Why don't we have one yet? Why has it been so difficult? The first one is we don't really know what the correlative protection is for the fetus, the developing baby, and the maternal circulation, the placenta, that whole maternal placental-fetal unit. What protects that from CMV if a woman gets an infection or is exposed or reactivates the CMV that she's had for decades? Is it antibody? Is it T cells? We know that both are important. You've seen various versions of this slide already. The viral particle is very complex and has a lot of glycoproteins, as we've heard, studding the surface. These are the important signals or hooks, as we heard earlier. I love that metaphor, hooking the virus to the cell surface.
CMV sets out to infect cells of very different types in the human body, respiratory epithelial cells, brain cells, the cochlea. These cells that we call epithelial, endothelial, fibroblasts, require different hooks for the virus to hook into them. The virus also has all of these proteins that are sort of under the surface in this region that we call the tegument of the virus particle, and these are targets of T cell immunity. T cell immune responses to CMV have been the cornerstone of many failed vaccine platforms that have come and gone in the last decade. I think we still don't really know the overall importance of those T cell responses in protection of the fetus. Clearly, this platform focuses on the outer proteins, the ones that are the so-called glycoproteins that involve the initial steps of infection.
These are some of the proteins of emphasis, and we've heard a lot already about the glycoprotein B. Sally nicely reviewed the efficacy studies from Pass and Bernstein, looking at adjuvanted glycoprotein B. It was pretty good. I'm from Minnesota, so we like to talk about how things there are generally pretty good. We need something better than pretty good for a vaccine that will protect against congenital CMV infection. That's where all of these other platforms and expression approaches have been used. I won't belabor all of these points. I'll just mention that a lot of these are in clinical trials. A vector known as MVA vector, this has been an interesting body of work performed by Professor Donald Diamond at the City of Hope Oncology Center in Duarte, California. Some virus-like particles, work done by Professor Bodo Plachter in Germany.
Polyepitope vaccines, soluble pentamer vaccine that's actually being developed by GSK. We just heard about the Moderna messenger RNA platform. All of these vaccines, I think it's very exciting to see them move forward in the clinic. The pentamer, that group of five proteins, hence a pentamer that's involved in binding of virus particles to epithelial cells, I think is emerging as a key milestone, as a cornerstone of vaccine development. I'll segue into just some comments about these live virus vaccines. Again, these are different than the messenger RNA vaccine platform or the purified protein platform because these vaccines are all generated against the backbone of an entire virus genome. Many decades ago, highly passaged strains of CMV went into human volunteers. These had a really minimal efficacy. Unfortunately, not much reactogenicity.
It turned out that pentamer that we heard about earlier wasn't really expressed naturally or normally in those attenuated vaccines. That led to a number of trials to try to optimize these vaccines to express that pentamer, but still retain safety. That pentamer, again, in this cartoon, shows the complexity of how this protein, this pentameric complex or PC, right? That doesn't stand for political correctness, it stands for pentameric complex, the five proteins that bind together to allow binding of the virus particle to these epithelial cells. This is the key addition that the Moderna platform adds that hasn't been in the adjuvanted gB vaccines that we heard about earlier. gB, as shown over here, is important in binding to fibroblasts. Given the complexity of cytomegalovirus biology, you really need to block both of these binding steps to prevent infection.
That's really, in a nutshell, what this vaccine is trying to do. Now, the Merck vaccine, V160, is different conceptually because it expresses the pentamer. That's a big selling point of these trials of the Merck product, but it does it against the backbone of the entire viral genome. Every gene in the virus is present. The genome is modified in a way that allows the virus to not replicate. Sometimes in the field, we call these DiSC vaccines. That acronym stands for Disabled Single Cycle, DiSC vaccine. The hypothesis is that you'll generate broad immune responses to all of the gene products that are in that virus particle that I alluded to earlier, including those T-cell targets in the tegument of the virus. In fact, we don't know that those are necessary for protection against congenital CMV.
This cartoon shows how the Merck vaccine is made. It's done with sort of a poison pill that's built into the virus, that can only be overcome using this synthetic ligand called Shield-1. It doesn't even exist in nature. That's why the vaccine is considered to be safe enough for use in clinical practice. In the laboratory where the vaccine's manufactured, you can generate these disabled particles, but you can only see these particles replicate for one round after inoculation into a subject. They're disabled. The idea then is that they'll induce antibodies to the pentamer, along with a lot of other gene products.
I loved in the last presentation how this idea of a cell-free workplace, a cell-free GMP facility, is really very elegant, because the Merck vaccine still very much depends upon having cells, cellular production lines where you can make these virus particles. Well, how does this vaccine work? It's gone into a lot of people. It's published now. This paper came out earlier this summer from The Journal of Infectious Diseases from Stuart Adler et al. What Merck did was they did, again, as we heard about already this morning, they did a dose escalation study where different amounts or different quantities of the vaccine were administered to volunteers. One of the arms of the study also included an adjuvant, an immune stimulation molecule. We use adjuvants all the time in clinical practice in your primary care clinician's office for many vaccines.
The best responses to this V160 vaccine were seen with the adjuvanted form of the vaccine. Using natural immunity as a benchmark, as you see in the dotted line here, the vaccines at highest doses with adjuvant were as good as natural immunity. Well, that's great. If the goal is to try to protect a woman who's never had CMV, and you can at least confer for her the equivalent of her neighbor who's had CMV in the past. We know that that doesn't completely protect against infection, but it does reduce the risk, reduce the severity. I think what we really need for a CMV vaccine program is something that's better than natural immunity. Think about that for a moment. You need a vaccine that's actually better than Mother Nature. That's a big challenge.
If we're going to prevent reinfections and transmission to the fetus, that's the goal. This is where the preliminary data that we've heard about from Tal is, again, very exciting because here's your benchmark of natural immunity, and we can see that the mRNA platform is actually better. Different laboratories, different manufacturers, different experimental conditions. Wouldn't it be exciting to see these run side by side, head to head in the same lab? At least based on the published data and the data we've heard today, it suggests that the mRNA platform is superior to the V160 platform in inducing neutralizing antibody responses against that pentamer at that epithelial cell surface, which seems to be emerging as sort of the key parameter for CMV vaccine. What's the population that we need to immunize? How do we deal with this problem of transmission from reinfection?
This idea that even if you've had CMV decades ago, that it never really goes away. It's a latent herpes virus that reactivates. You can get reinfected with new strains. Of course, we face this with flu every year, don't we? A flu shot every year, because new strains of influenza virus form that our immune systems don't have experience with. This slide makes the point that the prevalence of congenital CMV in a population is actually directly, not inversely, but directly proportional to the seroprevalence of CMV antibodies in women who live in that population. This, again, is what some authorities, some authors have called the enigma of CMV immunity, the paradox of CMV immunity. If you Google those phrases, you can find a lot of interesting papers sort of pointing this problem out. This is one of the major issues with CMV vaccination. Why do reinfections occur?
This is a very highly cited paper by Boppana and colleagues from the New England Journal some years ago now, in which they looked at the acquisition of new peptide epitopes and used this to define the presence of a new infection in a woman who already had immunity. These new reinfecting strains can be transmitted to the fetus. They don't tend to cause as much severe disease as a primary infection in pregnancy, but 10%-15% of those babies may still go on to have hearing loss. An asymptomatic baby at birth that looks healthy and normal, if that baby has CMV, there's still a substantial risk. We don't recognize those babies as clinicians because how can you make an asymptomatic baby better? This becomes a very powerful driving force for newborn CMV screening.
That's a whole other conversation in its own right, an interesting one. New York State passed a bill this year that now has opened the door to screening some newborn infants for congenital CMV. That might be a very exciting part of a post-vaccine licensure surveillance program as these vaccines move forward in clinical practice. We heard earlier about the power calculations that would be required if the endpoint is disease in the newborn infant or transmission to the fetus. Moderna has nicely laid out the rationale for, in their conversations with FDA, how an endpoint of preventing infection in the woman may be sufficient for moving a product forward to the licensure phase, and I think that would be very exciting and would be justifiable scientifically. This just summarizes some of the other issues with CMV vaccination as they relate to who you would immunize.
We've had some discussions earlier today about how the adolescent would probably be the key target, and that makes a certain amount of sense, giving to young women before they enter their childbearing age. Waning immunity from vaccination could be a problem, though, if you immunize a 12-year-old, and that immunity needs to last until she's into her 20s or 30s or even 40s during her childbearing years. How many boosters would be required? Would it be better to have a strategy of immunizing women within the context of obstetrical practices? There's even been some discussion about how immunizing all toddlers in group daycare might be the best target population. Let's get this virus at its source, because toddlers in group daycare centers are often the vehicle that brings CMV home to their mothers, who, if becoming pregnant, are then at risk.
I think this will be an interesting discussion moving forward who should be immunized. These are, again, just some of the other populations. Universal vaccination of all children, seronegative women. We probably need to immunize seropositives as well, although the phase III studies will, I think, most likely be done in seronegative. These are important issues moving forward. I'll just finish then by mentioning a little bit of the last challenge as I see it with congenital CMV vaccine, which is nobody knows about congenital CMV. I've had hundreds of babies over the years with congenital CMV that I've seen in my clinical practice or in the context of clinical research studies.
Almost without exception, I've never maybe I can count on one hand the number of times a woman or a family has told me that they ever even heard of this before it happened to them. That's a problem. Many of us in the audience with gray hair remember the public service announcements back in the 1960s, I remember the 1960s, for rubella, right? If you're pregnant, you should be careful. If you're around a child with German measles, you should get immunized. We don't have that same sort of media awareness, media savvy, and knowledge. That's a problem. One of the things I love about democracy is that when it works well, it comes from the people, and women have voted with their feet. They've talked to their representatives.
A mom who I worked with in Portland, Oregon, who had a child with CMV, actually met her representative at Starbucks in Sellwood, if anyone knows where that is. This is not an advertisement for Starbucks, but that's how they initiated the conversation. It went from that to a bill in the Oregon legislature to a debate on the floor to success. They have legislation now that tells the state health department that, "Hey, you've got to fund programs to tell women about CMV and inform them." Of course, when the medical school found out about this, they were a little bit mad because they weren't included in the discussion. A lot of states have done this now. We have a similar bill in Minnesota, that we're trying to get through.
I mentioned the legislation in New York. These bills have a number of unique flavors that vary from state to state. Some of them are done, the best ones, I think, are the ones that are done and passed with the collaboration and cooperation with the state health departments. It's hard to get a fiscal note for a lot of these, but the best bills are the ones that are funded to provide resources to increase education and awareness. Some of these bills are linked to a mandate that says if a baby in the newborn nursery fails their newborn hearing screen, which about 1% of all babies do, they just don't pass the newborn hearing screen, that baby should be tested for CMV. The first bill in the country actually had this mandate in Utah, and that's led to a lot of interesting data.
This is a big challenge, I think, to the CMV vaccine field. I think any program of CMV vaccines, any phase III studies that move forward really ought to be done in conjunction with resources earmarked for knowledge and awareness to help put this on the landscape. The CDC has done a good job in recent years of increasing awareness, and the webpage there has a lot of valuable information, and it's been a much more highly emphasized part of the CDC mission in recent years. This is Kelly Fenton, who's a State Senator in Minnesota, introduced legislation, and actually one of my patients right here. I won't play the video. You can look at it if you'd like on YouTube. I'll conclude and just say there's a major public health need for congenital CMV vaccine. There are a variety of vaccines in the clinic.
I hope I've made the point about the importance of the pentameric complex and antibody responses to that and gB, presumably working in synergy to block binding of the virus to the cell and prevent infection. We've also talked about how live attenuated vaccines probably have some safety concerns that subunit vaccines based on individual cloned genes don't have, and these four areas of knowledge and awareness that I think are needed as the vaccine field moves forward. Thank you all for your attention. I look forward to the question and answer period.
Good morning. My name's Laura Riley, I'm a high-risk obstetrician here in New York now, I was at Mass General for 23 years. Not only am I a high-risk obstetrician, I also did a fellowship in infectious disease. I've spent pretty much my entire clinical career, unfortunately, taking care of moms with this disease, then helping to follow them through pregnancy. I think you've heard a ton already. Probably everything I was going to say, you've already heard from two or three speakers. I'm going to tell a story then sort of explain why I think, from an OB standpoint, why I think the vaccine is absolutely critical, and why I think we haven't made very much progress over my entire career for these children. The story I'm going to tell, I can tell because it's already been published.
It was a case report in The New England Journal, and it was a patient of mine, a woman who was 42 years old, 41, 42 years old. She had had years and years of infertility, and she had tried everything known to mankind to get pregnant. She and her husband were unable to get pregnant, so they decided to go to Ethiopia, and they were going to adopt a baby. They sort of went through that whole process. She goes off to Ethiopia. She had to spend, I think it was two or three weeks in the adoption agency, just getting used to the baby and doing everything that you need to do to get all prepared. She did all of that. This is a totally healthy woman. She did all of that.
She stayed for two or three weeks. When she came back to the U.S., she had a low-grade fever. She didn't feel herself. She went off to her PCP. Essentially, she had the mono-like illness. They screened her for CMV. It took a little while to get there. On top of it, this is someone with years of infertility. She was pregnant. She wasn't using birth control because she didn't think she could get pregnant. It was just like the perfect storm. Horrible. I actually met her at 14 weeks gestation. She was probably nine or 10 weeks when she got this mono-like illness. It turned out, long story short, it turned out to be CMV. For this woman, it turned out to be probably primary CMV. It was a difficult diagnosis to make.
The CMV immunity test, which is what we all rely on to tell us sort of the general timing, led me astray, actually. It didn't give me the information that I really needed to do good counseling of this patient. Long story short, we ended up doing, because it looked like she had primary CMV, looked like she probably had it sometime in the first trimester, and I'll show you some data why we are concerned about the timing. She decided to do an amniocentesis. Now she's home with her two Ethiopian children that she has adopted. She's newly pregnant. She's incredibly excited. We have this discussion, and I say, "Look, you need to have an amniocentesis." We do an ultrasound. The baby looks pristine. Everything looks totally perfect, and I'll explain to you why that was important.
We do an amniocentesis. Within 12 hours of doing the amnio, the lab called me to say, "This lady has CMV everywhere. This is the fattest we've ever seen this grow." I'm thinking, "Oh my God, PCR positive everything." I have this long conversation with the patient. I tell her what we know. Given the timing and given all the possibilities, this woman, this couple, is devastated. To this day, I have goosebumps over this conversation. Here's a couple that's wanted nothing more than a baby for 10 or 12 years. They finally get there. I tell them this devastating news, and they decide, "You know what? We've got two kids. We can't take any chances." We have to do a pregnancy termination.
Now we're doing a pregnancy termination at 18 weeks or, no, actually, it was beyond that because we waited a while to do her amnio. She was maybe 21 weeks gestation. This is not a good situation for any couple to ever have to be in. It is uncomfortable for the person doing the counseling and not being sure about what you're saying, and you're steering people in a direction, maybe not. You're putting them in difficult situations. This is where I see it all, as one of those people who's stuck in that small consult room trying to help people make these difficult decisions. I'm going to show just a little bit of data. This is the OB's perspective. You've heard this before. I think these are the epidemiologic factors that we're concerned about.
I think one of the speakers said early on, I think it's really important to understand that this is a disease, or CMV, I should say, seropositivity, is much more common in patients with lower socioeconomic status. It's seen much more commonly in Black versus white patients or patients with Mexican ethnicity. It's seen, obviously, in those with higher parity as well as residents in developing countries. If you think about it, the people who are at highest risk for becoming seropositive with CMV are usually the ones with the least resources to actually take care of these children that are now devastated. The other thing that's really obviously very tricky, and you've heard a lot about this, is the fact that there's multiple routes of transmission.
When we talk about how do you counsel pregnant women, well, there's a million things you're telling them that they can't touch, they can't eat, they can't do this, can't do that. It really is not practical, which I think is the other thing that's really important here. What are the annual seroconversion rates? Let me just say the background piece of information, just to think about the U.S. alone, there are 4 million births in this country a year. We're talking very large numbers. When you look worldwide, we're talking about a lot of children who have the potential to be devastated. If you look at just pregnant women in general, this is just one study in the literature, but it's the one that most of us use in our thinking about this.
Looking at all pregnant women, the seroconversion rate per pregnancy is about 2.3%. I think what is more important or equally important and becomes a counseling issue is those that are daycare providers, those that are stuck sort of with the kids slobbering all over everything, and it lives on the surface of every toy, and then they give it to each other child, including their daycare provider. That's really important. As you can see, the parents of children who are either shedding or not shedding, actually, is also an important vector. In terms of maternal disease, these are the things I think about as an obstetrician and become important when we're trying to counsel patients about what may or may not happen. You've heard primary CMV, it may be symptomatic, but for the vast majority of people, it's asymptomatic.
This patient sort of came in with a placard saying, "I have CMV." She had the mono-like illness, the vast majority that I've seen in my career didn't have anything going on. We figured out the CMV from their children rather than from them. Reactivation, again, may have a latent CMV, may have absolutely no symptoms. It's hard for us as OBs to know that this is something we should be worried about. Obviously, reinfection with a different strain, maybe you have symptoms, maybe you don't, most don't. Here's where we're left in obstetrics. The current recommendations, because of all of the things that you've heard this morning, the current recommendations are really only to test pregnant women who have that mono-like illness, because you need to understand that.
The second biggest group that gets tested during pregnancy are going to be those women who are carrying fetuses that have anomalies or findings that could be consistent with CMV. It can be quite the hunt. This first ultrasound on the right, this baby has calcifications, which is something that we're always concerned about. The second one is hypoechoic bowel, so the bowel is bright like bone. The problem with that one is that you end up doing a lot of CMV screening for kids that have a whole bunch of other things. If a baby swallows blood, it has echogenic bowel. If a baby has Down syndrome, it could have echogenic bowel. It usually isn't CMV, but we have to look for it. The last one is again, oh, I'm sorry. The first one had large ventricles, so ventriculomegaly .
The last one is calcifications in the brain. Any of those things that we see will prompt us to look for CMV and other viral infections. In terms of transmission rates, this is the data that we use to sort of counsel patients once we figure out that they've had primary CMV. The transmission rate, I should say, also depends in some ways on the gestational age at which we're doing the counseling and guessing on the timing of the infection. You can see that there is some transmission even if they get the infection just prior to pregnancy.
The reason this becomes an important issue is because what we see not infrequently these days is that people doing IVF actually will test for CMV. When they do that, every now and then, they get someone who's going to be CMV positive. Then we're trying to figure out when was it. Was it far enough prior to your getting your blood drawn that we can still do your transfer? You can just imagine you've taken a bajillion drugs, spent $20,000, ready to have your IVF transfer. They come see me. I say, "Ugh, it's probably not a good idea. You should wait three months." That doesn't go over well either. The other issue, obviously, is these other trimesters. I think the main point about this in terms of transmission rate is that transmission happens all through pregnancy. There's no safe time.
That's very disconcerting for patients as well. The timing of the infection is important for a whole bunch of reasons, but certainly important when you're talking about the level of damage potentially to children, right? When everything's being formed, when all those organs are being formed in the first half of the second trimester as well, that's when we see the most damage. The brain is developing all through pregnancy, right till the end, and kid's out in first year of life. In which case, anything that can affect that is going to affect the brain, hence the neurodevelopmental delays that we see. The other reason that the timing is important is because in most states in the U.S., you can only terminate pregnancy up until 24 weeks gestation, and in some places you can't terminate at all.
That's its own conversation, but it's an important one because that's really all we have to offer patients when they find out that they have that diagnosis, and they decide they don't want to continue the pregnancy. We sort of have this timeframe that we're working in as well. This has come up a couple times. I think it's important, obviously, to recognize that women will have reactivated disease as well as infection with a new strain. That's important. I think what's controversial is the severity of disease. It's very much less likely to happen. Is it less severe? I think it's kind of all over the board. At any rate, thankfully, it is far less likely to lead to what you see on the right. This baby has microcephaly.
In terms of reactivated and reinfection, I think the reason to even focus on it is because we know that that's probably the genesis of most of these kids with congenital infection. The true impact, I think this is a statement that is very true. I think we don't really understand 100%. Part of the reason we don't understand 100% is because we are strapped with just making the diagnosis during pregnancy. The only things that we have to make the diagnosis during pregnancy are the CMV IgG, CMV IgM, and then CMV avidity helping us try to figure out the timing. The problem is that those tests, like if you were to test most people, you're going to get IgG positive, IgM positive, and you don't know when the infection happened, and then you're stuck.
You're trying to guess when it happened and try and make some decision on telling this poor couple, "You should do X versus X." Very difficult. The current prevention strategies, I've gone over this, there is no routine screening. The reason there's no routine screening is because quite frankly, right now, there's not much we can do about it. We're very concerned that the screening could lead us in the wrong direction. In both ways, we'll miss people who are infected, who could have a devastated baby, and also there may be people who we say, "Well, we think that something bad is going to happen," and then they end up terminating a totally normal pregnancy. It's just not a good situation. In terms of what should we be doing? Well, right now the push is on education, as you just heard from our last speaker.
Trying to educate women about personal hygiene and kissing children less than six years old. I mean, really? Do you really think that works? I say all these things, but I can tell you, people just look at me like, "Okay, Dr. Riley," and they move on to the next thing. The breastfeeding situation, we know that it is in breast milk, and I think this does come up occasionally, and people will say, "Is it safe to breastfeed?" In general, people feel as though the benefits of breastfeeding outweigh the risks associated with there being CMV in the breast milk. If there's a recent infection, wait six months. I write that down there because I see a lot of patients. The thing in New York is to get pregnant at 50, I have found.
When we're doing all these consultations for the IVF crowd, and they get this new CMV infection, or it's the first time anyone's tested for it, that's really the issue. I have to tell them, "Yeah, save your money and wait until you're six months out before they do your transfer." This is the public awareness that our last speaker so nicely went through, so I'm not going to belabor it. I think that the public does know about it. I think that unfortunately, it's the public that's already got the baby that's been affected. They're the people that know about it, and everybody else is like, "I never heard of that. What are you talking about?" It's all quite scary. What are the current treatment strategies? There aren't any, basically.
We got very excited when a paper came out in the New England Journal all about CMV-IVIG. There was so much excitement, we went from that to, we're going to give it to everybody who we think has it. It doesn't work, right? Another paper that was a really well-done randomized trial came out, which suggested that it just doesn't work. We don't have anything. The only thing that is going to help us prevent congenital CMV is going to be your vaccine. That's it. That's all we got. Antiviral medications seems like a great idea, except that most of those, when used in the mother, are actually toxic to the fetus. We're sort of left with what you can do once the baby's out, not what we can do during pregnancy. This OB's perspective.
Here's the thing, these are the things I worry about, and these are the things that are going to keep me up at night when you launch this. Is it really safe? I think that you guys have convinced me that you will do all that needs to be done to tell me that it's safe. Is it efficacious? It's clear that it's going to work. Is it going to be durable for all the reasons that we talked about already? When are we really going to give this? When people say childbearing, well, don't wait till 15, forget it, because there's the 13-year-olds out there. When you say, is it going to work if you give it at 13? When I see all my IVF ladies and they're 48, I hope so. We're going to have to think about that.
Obviously it's a much needed intervention. Look at the numbers, right? Much can be done. With that, I will close.
Well, I would like to thank our three speakers for those very nice lectures. Thank Tal and Juan and their teams for all work done over the years. As I said, I'm very happy that for the first time since 2011, I'm talking commercial. As Tal described to you, our plan for CMV is the following. We want to go into a phase III, looking at prevention of CMV infection in healthy women. To get there, we're going to run and start in the near term, phase II that we have tapered will be very rapid. It's 152 healthy subjects, and we'll use a three-month interim data point to trigger the phase III start. How do we think about this vaccine? We obviously think that CMV has a blockbuster commercial opportunity. As you heard today, extremely large medical needs, no approved vaccine on the market.
If you think about the product life cycle, which we think is very important, and if you look Prevnar with Pfizer and HPV with Merck, it's a very important part of really maximizing the impact we can have with our vaccine. We want to start with women of childbearing age, because it's a very motivated population. That's where the medical need is, obviously, and want to start there. What we see with HPV is we want to go down to adolescents to broaden the protection in the population. As was alluded this morning, we have a possibility because I think most people don't appreciate that the reservoir of CMV are humans. Flu is birds and pigs. Zika is mosquitoes. The humans are the reservoir.
We have an opportunity to have a massive impact on public health if we're able to have a vaccine that's durable and that is given to infants. If you think about that opportunity, we think the addressable market is extremely large. If you just take a few different assumptions that you could have on what you could do after starting from the base of healthy women to go into adolescent and to go into children, you could talk very large numbers. Just to give you an order of magnitude, the MMR vaccines have sold 200 million vials in 2018. Just last year. 200 million vials around the world. This is a very large opportunity for us. We think from a commercial standpoint, the bigger part is going to be around OBGYN and then pediatricians.
We think there's a great opportunity both for the phase III enrollment and for the commercial uptake to use social media and to be able to really educate about the disease. What we spent quite some time doing already over the summer is to think about what activities are we going to start to do from 2020 to really accelerate the uptake of phase III, because of course, the faster you can enroll those subjects in your phase III, the sooner you can file your BLA to the FDA. It's critical for time to launch. All the activity that you do one or two year pre-launch are very important for the ramp of the product, obviously. We're going to start partnering with all the associations, both regional and nationals and around the world that care deeply about CMV.
We think there's great work to be done to partner across the globe. We want to really start educating, starting literally from today, that now the data is public. The team is going to publish the data, and we're going to start to go to several OBGYN conference, medical conference around the world, so that the team is going to start spreading the data to educate about mRNA vaccines or products. The data you've seen today with much more detail. We're about to, on the education front, do two things, both work inside the offices of the clinicians with pamphlets and information so people understand the disease. As you heard today, we do not talk about this virus a lot because there's nothing that can be done.
Worrying a pregnant woman or woman who wants to be pregnant is not a very good idea when you can do nothing about it. We think that it is a great opportunity to educate about the virus. We have seen the CMV Foundation already having pamphlets trying to use Zika, because Zika, given what happened in Latin America, has been really described a lot in the media. Because of the similarities between Zika and CMV, there's a great opportunity to use the understanding of Zika to educate around CMV. We want to do a lot around that. Of course, driving the awareness of the product. If you think about pricing, and I'll come back to it in my conclusion later, just to give you an order of magnitude.
Gardasil, the HPV vaccine in the U.S., has an average selling price of around $450 per treatment. For three doses, $450. As we've seen with vaccines are very long life cycles. We believe this is a great opportunity for us to have an impact. If you think about what Juan said, with the fact that we believe given our cost structure at this type of Gardasil price point, this product could be a 90 gross margin product. There only is around a 50% of it margin opportunity. We think it's a really exciting opportunity. We care deeply about the fact that we own the global rights to this vaccine, we are in the driver's seat.
We believe that the best way to have a great launch and a great impact on patients and public health is to really start working from now to educate about the virus, to educate about the data that we had from the phase I, and soon sharing the phase II data to be able to really uptake for the phase III and have a quick launch and then a good, big commercial update. That's the current plan. We will give you more details as we continue to progress that plan. We're going to be spending a lot of time with the team in the coming months and quarters to really sweat all the details to make sure we give justice to this vaccine and that we impact as many women and as many families as we can for the long term.
With this, what I would like to do is to get on stage all the speakers for the CMV section of the agenda, so we can take a section of Q&A before we go into a little break. Juan, Mark, Laura. Sally, if you don't mind joining us, please. Thank you. Sally, you want this?
Yes, sir.
For handling questions. If you have questions, there are a couple of mics going around, please just raise your hand. We're going to start here just for 15.
Hi, guys. Alec Stranahan from Bank of America. Just a couple of questions from me. My first question is on the dose selection in the phase II. In the phase I, you dosed up to 180 micrograms, in the phase II, it looks like the top dose was 150, and no real indication of anything higher, like 300 dose level. Does this mean you're satisfied with the neutralizing antibody induction observed in the phase I? Sort of what's the rationale for the dose selection in phase II?
Yeah. Let me take that. I think we are very happy with the immunogenicity that we've seen so far. There's not much, if you look at the data, to distinguish the 90 from the 180. There does seem to be a little bit more, expanded dose. The safety profile is overall similar. When you go into a phase III, you want to be sure that you got the right dose. What we've done is hone in a little bit on the dose range. The 300 is still ongoing. We will have the opportunity to see data from that dose just before we launch. If we need a course correct, we'll be able to. We've enrolled everybody there. We've seen safety through the second dose. We don't see any surprises there. If I look at this level of immunogenicity, it feels like we're there.
We just need to confirm that we've got the right dose with some bracketing around it to go into a pivotal trial.
My second question is on durability that you'd want to be seeing in the phase II and probably in the phase III. Through your interactions with the FDA, what sort of length of follow-up do you think you'll need? Obviously, it sounds like viremia rate might be the registrational endpoint of that study. Just wondering if that follow-up would need to be a year or something before you submit, or if that would be sort of a post-commercial launch follow-up study.
I don't have an answer for that yet. What we've had in our initial discussion with the agency is to understand the endpoints. I think we'll go through the usual process of an interface, do a discussion and aligning with them on exactly the pivotal trial design and duration and endpoints before we launch it. I can't comment on that. I expect it, if you look at the history, I expect that within a year, we expect to collect the rates that would be required to actually demonstrate that significance. Your point is valid. In terms of the durability, by the time we get there, we will have even more data on the entirety of the study of the phase I. Over time, we'll have more data emerging on the phase II. We'll get wiser over time in terms of those numbers.
Great. Thank you.
Thanks, yeah. Matthew?
Great. Thanks. Two questions from me, I guess. One, on the phase I, on the sentinel patients, you have here the epithelial cells. We have an idea against the pentamer, but not GB. Can you give us an idea of durability against GB? Second question is.
Can you be clear what was discussed at the Type C meeting? Was it just primary endpoint and then your ability to think about what the safety will look like off of that, or what other factors? I guess you mentioned durability already.
Yeah
just be clear about what other factors.
Let me answer both. First, in terms of durability, we see similar data on GB, just like you can get to the levels of seropositive in GB and seronegative, you can maintain those out to 12 months. We think the relevant point here is the pinpoint, that's why we've shown it. These are early days. Once we have the full durability for the entire study, we'll show you both against MGB. The Type C meeting, there was a whole bunch of questions, as there usually is. Some of them have to do with our manufacturing. Some of them have to do with the study design. I think the salient one for us was an understanding on the endpoint, in terms of how you eventually get there. That was, I think, the one that was the most important for us, and that's why we're sharing it.
The 8,000 patients that you referenced, can you just talk about how you got to that number?
Yeah
in one of the positions for them, I think I saw numbers that were like 2,000 to 5,000 or something. I'm just wondering what sort of criteria you're looking at.
Let me be very clear that I did not discuss trial numbers or that level of granularity with FDA. As I said, that will happen later than in the phase II meeting. The numbers are a function of the incidence that you see and some preliminary extrapolation. We may be able to do it with less than 8,000. The notion here is if you find the right population where the annual incidence of cases would be 2% or so, 2%-3%, if you can get there, then you can back calculate the math yourself and figure out what the relative power is you would need to show that, and that's how we come to that range of numbers.
Of course, in a trial like this, and I've got people who are far more experienced than me here on stage, what you typically do is make sure as the study progresses, that you're actually collecting enough cases that you're appropriately powered, so that if the actual incidence in your trial population is lower, you'll end up extending or enlarging it. If you're there or higher, great, you got there. That's typically how you're thinking about the numbers.
Okay, thank you. Kevin?
Great. Thank you very much for the updates. Just to kind of flush that out a little bit more because that sounds like an efficacy number. Do you have a vision that you would require a larger safety database? Would it be a single phase III study, or do you think even in that kind of 8,000 range from the prior subjects does that you would have sufficient safety data? Thank you.
My assessment at this time is I think that number should suffice in terms of a safety database as well. I also think that that number may end up including seropositive, not just seronegative. To the point of the power calculation, you may be able to get there with fewer subjects on purely seronegative. You're going to need to show overall, the safety profile for seropositive as well, because we expect ultimately to get an indication that doesn't distinguish between the two. That's how I come up with the numbers.
Thank you.
Kim.
From Oppenheimer.
Sorry.
Just a couple of quick questions. One is on just how you were talking about the background rate. We've got a lot of data presenting in the U.S. here, so phase II will be worldwide. Is there a potential that as you're looking at ex-U.S., the background rate of CMV infection, that you could see some differences there that would also affect your powering, and how would you sort of approach that? I've just got a quick follow-up after that.
Yeah. Our work is only beginning in terms of feasibility. We are looking globally. We're looking primarily at the major places in Europe, and similar geographies, similar in terms of incidence rates, so that the phase III, when we launch it, is fairly homogeneous in terms of expectations.
Just a quick follow-up was that I believe that vaccines, the uptake is usually also driven by a lot of organizations like ACIP, et cetera. It's not just legislation, which I think helps a lot. What are some of the things you're thinking of doing going forward so that by the time, assuming the vaccine and the phase II trials are successful, and get approved, then it just becomes, I guess that's algorithmic driven, right, in terms of uptake. Thank you.
My colleagues here can comment about the role of ACIP and what we need to demonstrate. We've had initial discussions with the folks down at the CDC. Obviously, in order to get a recommendation, you need to show your overall benefit, risk, and value proposition, which I think in this disease is going to be quite clear. I don't know if any of you want to comment on sort of from an ACIP perspective, what that would look like.
Well, as a former member of the ACIP, I would say that's exactly what we're concerned about, right? Is it going to work? Obviously, the cost effectiveness is huge now. Although the absolute vote within the ACIP, we're not supposed to take into consideration cost effectiveness, but it's always presented in the work group. I think it does influence in many ways what we do. I think the deal with this vaccine is going to be congenital CMV is the most common congenital virus that we see in the U.S. and probably globally. I think that it's going to be an easy argument, to be quite frank, assuming it works and it's safe. You bring up another point that I think is really important.
I think the timing for the vaccine is critical and a good one in the sense that finally, we're at a place where I think everyone believes that vaccines during pregnancy should be done for prevention. It's appropriate to test
During pregnancy, there used to be this, you can't touch a pregnant woman because of the potential effects on the baby. I think with the advent of H1N1 and all these other vaccine-preventable diseases wreaking havoc in the population, people are coming around to the necessity.
Thank you. I'm going to take Tim and then two more questions, and we'll have to move on.
Yeah. First a comment, and then a question with maybe a follow-up. This is a real moonshot vaccine, because not only does it have six RNAs in it, but five of them encode a pentamer. If any one of those RNAs don't work, the pentamer isn't made, right? It's really a complex vaccine. Now, my question is about cell-mediated immunity. I know it's your right to focus on the antibody-mediated responses, because they can block virus infection, right? They're neutralizing. It's great. It's a great endpoint. The great thing about our vaccines is that cells take them up, and they make the proteins intracellularly, and then those proteins can be made into peptides, and they get presented on the cell surface through MHC. They induce cell-mediated immunity, which subunit vaccines can't do, or they have a very hard time doing at all.
I'm very curious if you're getting any cell-mediated immunity, because there are cells in the body that are producing the CMV virus. If those cells can be killed by cell-mediated immunity, the virus titer is going to go way down, and you also may not only help that person but prevent transmission to other people. The last speaker before Stéphane spoke about shedding of virus. Obviously, you folks can measure shedding, and I wonder if we can measure either the cell-mediated immunity or the shedding in any trials or future trials.
Let me try to answer it. We're looking at shedding, and I'll ask Dr. Riley to answer that question because it's of keen interest to me, especially in the seropositive. For full disclosure here, I'm actually not a vaccinologist. I'm a cancer doc, and I've spent most of my career being a complete T cell chauvinist and believing that's what's important. I think when I step back, look at SHINGRIX, which is a one subunit vaccine that's had 97% efficacy against preventing reactivation of persistent infection. I think that teaches us that T cells may be important, but the ability of subunits just to generate antibodies may actually have a profound and deep effect. I think in the context of this virus, it's unclear to me what the importance of the T cell would be.
That being said, by virtue of mRNA's ability to make proteins from within the cell, while I haven't shown you any data and we're working on measuring it, I'm pretty sure that we're able to elicit T cells because everywhere else that we've looked, we've seen the ability to generate T cells. I just don't know how important they are as a correlate of protection or how to get there in terms of measurement as I am in terms of my confidence in the antibody levels.
I'll just tell you from a preclinical non-human primate standpoint, one of the things that I've took on in this field is trying to develop a non-human primate model of congenital transmission where we could test vaccines. At first we didn't know if the viruses in primates that are similar to the human CMV virus actually cross the placenta. We took about a strategy of immune suppressing the moms and then infecting them with CMV to see if we could see it. When we depleted their CD4 T cells of these pregnant dams, we infected them with CMV. All of the fetuses got infected, and 80% of the fetuses were aborted. We saw much more severe infection of the fetus and 100% transmission in the setting of absent CD4 T cells.
What was the % transmission when you didn't deplete the CD4 cells?
Small numbers so far, but it was at about half. 100% versus less than half, kind of similar to the human situation. We don't know if those T cells are needed for high good antibody responses or those T cells are needed to kill infected cells. mRNA vaccines though traditionally, what data's out there, induce great CD4 T cell responses. I expect to see that.
CD8 T cell depletion would be very interesting.
Oh, we're going to do that in a new grant that we just received from the NIH.
Great. Question over here. Yeah, Valerie.
What plans do you have to introduce a viral challenge to confirm the fact that you're providing protection here? Maybe a second question on whether you anticipate that the seronegative patients would need a booster at some point to remain above the seropositive level?
I don't think a challenge study is feasible here, certainly not a human challenge. I think to the degree that animal models are available, we've already done them and shown that it works. The next step is really the human study. For how much higher above the levels you need, I think that's an unknown, and it's one of the things we're going to have to figure out over time as we measure both the durability of our vaccine and its efficacy in a real clinical trial. I think short of that, it's going to be really hard to answer that question. Bob, you had a question. Oh, I'm sorry. I just saw Bob raise his hand from before.
Two quick things. Very nice, awesome presentation. Two quick things. One, I'm sure this is true, but these proteins are invariant across strains.
Variable strains.
Is there strain to strain?
Are they invariant?
or geographic variability of the antigens you're giving?
They're not. There's several different genotypes that make up each of the glycoproteins. That's one thing that we've been studying in the lab, and we'll be interested to see how it plays out here. I think one thing that can be done in the clinical trials is sequencing the virus that is actually acquired by vaccinees to see if there is a shift in the vaccinees, what viruses they're acquiring versus placebo recipients to see if the vaccinees, if we see protection or even if you don't see protection, were they protected against the strain that the vaccine was made against, or were they protected against all strains?
Is there already evidence in your model that only certain strains are protected against?
Very little. It's still an unknown question, I would say. We are looking at it in the old vaccine trials.
Any sense of the fraction of the U.S. population, of the prevalence that would be covered by the existing sequences?
Which is it again? Which strain again?
The U.S. prevalence.
Prevalence
of CMV, what fraction of it?
I believe this is the one that's the most prevalent, but I'll get back to you on the exact.
Okay. One small question. The press release mentioned one Grade 4 elevated PCT.
Sure.
Which I just wonder if you could say anything more about that?
Yes. It was a subject that had a somewhat abnormal at baseline. It was temporarily associated with getting the first or second dose. We got the subject back for follow-up by the time we found them, and 10 days later they came back, it was back to normal, and there were no clinical sequelae of that whatsoever. Technically, it reached the level of Grade 4.
What do you make of that? Was that a lab error?
I don't-
Was that what you think is a real finding?
I don't think it is a significant finding. It's not something we've seen anywhere else. From my perspective, I don't think it's of clinical significance.
Great.
So.
Can I please suggest one thing, just process-wise? We're going to have a clinician at the end. I just want to be careful, we have a very heavy agenda, and we're already a bit behind schedule. Just maybe a 10-minute break. There's coffee just outside, fresh from outside. We will start within half an hour, okay? Thank you.
For the next section, we're going to be talking about immuno-oncology, and then we'll get into the rest of our programs. To sort of set the stage and give you a sense of what it is we're trying to achieve in this part of our pipeline, I'm going to start with the cancer vaccines modality. In a nutshell, what we have here is a personalized cancer vaccine that we've been talking about for a while. We're now in the clinic. We'll, in fact, have started the randomized phase II and the KRAS vaccine, which is part of a partnership with Merck. If you step back and ask yourself, why does mRNA make sense in immuno-oncology? Everything I told you before is true about the ability to stimulate the immune system.
It is also true that we can get, by injecting the mRNA into a tumor cell, we can actually change the local microenvironment. That's a unique feature that's sort of taking it a step further from vaccine and into localized therapeutics in a way. We can engineer the mRNA so that we avoid off-target effects. If any of this mRNA finds itself in organs we don't want to, it'll actually get shut off, and we've discussed that in the past. We think with these features of being able to do combination, being able to get into the tumor as well as induce a vaccination, it opens up the opportunity for many applications.
Always when you have a wide opportunity set, you have to ask yourself, "Okay, what is it that strategically I'm trying to do?" What made sense for us from the get-go here was really to build on the ability and our modern understanding of what the immune system can do for cancer patients and our ability to modulate that and try to use our technology in ways that would further boost that. Dr. Flaherty will come up after me. We'll talk a little bit more broadly about where the field is. For us, it's an opportune time to use mRNA technology to try and boost the inherent activity of checkpoint inhibitors.
If you think about ways in which you can do that, I've sort of already described one, which is a vaccine, and if we could vaccinate against cancer as a therapeutic vaccine, that would be great. The other obvious place to go is can you change the local microenvironment so that you're turning, to use an oversimplification, you're turning a cold tumor hot, and in the background of a checkpoint inhibitor, you're now unleashing the immune system in a more potent way. In terms of why this should work for a T-cell vaccine, just to give you the cartoon version, because mRNA translates protein from within the cell, we end up presenting the epitopes to T-cells in the way they're presented naturally. In a way, our immune system has been trained over eons to actually recognize peptides.
We're leveraging that potential from basic biology in order to generate T-cells, which actually tells you, if you think about our CMV vaccine, all those proteins that are getting made and shuffled as proteins to make antibodies, there's a part of that protein that you would expect would get shuffled into the MHC molecule can actually lead to T-cell production. For a personalized cancer vaccine, this is not just theoretical. It's actually a manufacturing tour de force, because what we do here is we start with a biopsy from a patient, and we fully sequence it, and from that point on, all the information we need from that patient is up on the cloud. It's all digital.
There's a bioinformatics process that happens on the cloud that actually picks the best epitopes, and then it comes back down to the manufacturing site to make a bespoke vaccine just for that one patient. We've started on our first application with a concatenative 20 neoepitopes, and we've recently increased that to 34. Our personalized cancer vaccines now pack 34 neoepitopes that are predicted to be the ones relevant for that patient's immune system to recognize cancer into every vaccine in a bespoke manner just for that patient. We've done now over 60 or 70 of these. We can consistently get this done in approximately 50 to 60 days. We're working on ways to even further shorten that. What we do here is once the vaccine is ready, we actually start repeat immunization.
Now every time, every 3 weeks, the patient will show up to get their KEYTRUDA, they actually get a vaccine. We've now given this repeatedly every 3 weeks up to 6 months. We've done this now at a 1 milligram dose. If you recall the doses of the infectious disease vaccines, we're talking about 90-180 micrograms. That was already a dose that was quite potent. In oncology, we've taken it a step higher because obviously you want to make sure that you get the most that you can. We've got already a buffer from the infectious disease vaccines where we know we're already potent as an immunogen. The side effect profile has been benign. We've seen no great 3s or 4s in the study that we've done so far. It's been fairly well tolerated. These are the data, just a very quick recap.
This is the first subject we treated in the phase I study at the 1 milligram dose for which we have apheresis data. That's a procedure where you actually collect a lot of lymphocytes, so you can actually do very refined measurements of the activity of the vaccine against every epitope. The salient point here is that there were 18 epitopes that were predicted to bind to the class 1. Those are part of the immune system. Of those 18, we actually proved recognition in 10 of them. A hit rate of about 60%. How good is that? Well, I have to tell you, based on preexisting literature and what the competition has done, I have set 60% as a benchmark for our success sort of going into this a few years ago. We're right there. How many epitopes are required?
Well, if you get just one, but it's the right one, that should be enough, and we know that from adoptive cell transfer. The question is, are you able to generate enough of the T cells and the right kind of them to actually mediate a response? For that, we're conducting a randomized trial. The only example that I'm aware of that has actually been proven to be a neoepitope from an oncogene that is shared between patients is KRAS. I think Keith can talk to this a lot more eloquently than me, but KRAS is the most commonly mutated oncogene in common epithelial cancers.
What we did here, based on initial preclinical data that suggested that you should be able to generate T cells against the KRAS mutation in the right context, we went ahead and made a vaccine that's a concatenative of the four most prevalent mutations. They cover together between 80% and 90% of all KRAS mutations seen in epithelial cells. That trial, when we showed this to Merck, they actually said, "We want to run the trial. Let's partner, but let us run the trial. We want to make sure that we actually recruit the right patients, those that have the right immune system." It's Merck's team that's today running the trial. They've already started dosing both as monotherapy vaccine and in combination with KEYTRUDA. The essence here is exactly the same as I've discussed. This is where the trial is.
The other part of our portfolio here has to do with intratumor immuno-oncology. Here it's a very different approach. In a vaccine, what I'm doing is I'm figuring out what I need to stimulate the immune system against specifically for every patient, and I'm injecting a vaccine into their arm in the way we've done for CMV, but I'm doing it repeatedly every three weeks to boost that specific immune response. With an intratumor approach, I'm waking up the immune system without knowing what I'm waking it up against. It doesn't have the specificity of a vaccine, but what it does is it actually uses the local tumor microenvironment. We inject a combination of cytokines to wake up the immune system and draw in the T-cells and change an immunosuppressive local environment into an immunopermissive one. That's the idea.
You go into a tumor lesion, whether it's superficial on the skin or a deep-seated lesion that you need to inject, Keith will talk a little bit to the complexities of that, but the goal here is that if you can do that, then you can sort of wake up a local microenvironment and draw in T cells. The right antigens are already there in the cancer. These are two very orthogonal approaches to build on the fundamental ability of a checkpoint inhibitor to unleash your T cells. Where are we here? The first one that was really the pharmacology proof of principle for us was OX40 ligand. It was just one membrane-bound protein. We've injected it now to patients repeatedly. We've gone up to 24 injections given over six months every two weeks at doses of up to eight milligrams.
By and large, it's been well-tolerated. We've shown that we can actually lead the protein production. This was all discussed last year at SITC. What you see here in the graph is the red field is actually staining for OX40 ligand within a biopsy of a patient that got mRNA. You can see on the left the field before they got the mRNA. There's no red there. Okay. We've shown we can lead the protein production. We've shown we can do the tracking. The question, does this work? I don't know yet. We've seen an interesting signal in a couple of patients with ovarian cancer where the injected lesions shrunk. It doesn't constitute a formal RECIST response. It was enough for us to go ahead and start expanding that trial into a phase II in ovarian cancer. That cohort should start.
On top of that OX40 ligand, we then said, "Okay, we understand the fundamental safety profile of the delivery technology, and we've proven that this delivery technology can lead to protein production in tumors." Let's go make more proteins that have a more profound effect on the immune system. We did this. We combined two active cytokines, interleukin-23, interleukin-36 gamma, together with OX40 ligand, with the same idea of now injecting tumors, ultimately in combination with a checkpoint inhibitor. For this, we've partnered with AstraZeneca, and we're using durvalumab, their PD-L1 inhibitor. This is a trial design. The idea here is to go both in monotherapy but then quickly get to combination dose with durvalumab, and once we do that, we will do dose expansion cohorts, testing the ability to actually mediate true resist responses in a variety of tumor types.
The red check marks there indicate where we are. We're currently dosing at the 2-milligram level of monotherapy, and we're well ahead, advanced in the combination with a checkpoint inhibitor. There's one last opportunity here in this intra-tumor space, which is interleukin-12. Interleukin-12 has been discussed or has been studied for a long time as a potentially potent anti-cancer cytokine. You cannot give this systemically because you run into toxicity before you ever have a chance to see efficacy. You could potentially give it locally, and in fact, there are data showing with other approaches that giving IL-12 locally could actually lead to tumor regression, not just where you've given it, but some abscopal effects, and that's been described. The idea here is can you do this using messenger RNA technology?
This was one of the molecules that got AstraZeneca most excited about the opportunity set in oncology when we partnered with them several years ago. This has now progressed. They've taken it into the clinic, and they're actively dosing patients. I'm not going to belabor the preclinical data here. Suffice to say that for all of these molecules, the preclinical data is what you'd expect. When we use these kind of agents in the relevant preclinical models, we actually can mediate a sustained and durable remission. Unfortunately, translating mice to humans is not as straightforward as it is with rare genetic diseases for oncology. You got to go run the right clinical experiments here, and that's what we're in the process of doing. Let me give the floor to my mentor and colleague, Keith Flaherty. Thanks, Keith.
Thanks. Thanks, Tal. I've been around these programs since a quick while before the INDs were filed. I don't know, I think that's right. I'm not going to tackle the issue of interpreting interim data, or we can discuss that in Q&A. Rather, my task is to kind of contextualize these programs in the landscape of immuno-oncology and try to address the issue of if successful where would they fit in or what would they contribute. I've titled the talk as I have because really, the magnificent event of the past decade has been the introduction of antibodies that disrupt PD-1, PD-L1 interactions. That's the big moment.
Really what that has allowed us to then do is to try to understand what is the basis or what are the determinants of response in patients who achieve at least partial, if not complete responses to PD-1, PD-L1. A bit more difficult to do that in combination, particularly with cytotoxic chemotherapy. VEGF-targeted therapy is the same. The group that I direct at Mass General tries to tackle this problem broadly. Because melanoma has always been my focus, I'm going to use melanoma data, and it's a good canary in the coal mine area, as I'm sure you're well aware. All of my comments are meant to really relate to ramifications of these programs broadly across cancer, and I'll come back to that point a couple of times.
Here's some melanoma data, but I could show you any other population in which PD-1, PD-L1 antibodies have shown some monotherapy efficacy and make a very similar point, which is that we basically have two problems to try to overcome. We have a big problem of de novo resistance, even in a tumor like melanoma with a 40% response rate. MSI-high colon, cutaneous squamous cell carcinoma, Merkel cell carcinoma, these have the highest response rates, 50% response rates in some cases to monotherapy PD-1. You still see this rapid disease progression in a subpopulation of patients, even in those indications. A much bigger problem, of course, in cancers for which the response rates to PD-1, PD-L1 is 15%-20%, sufficient to drive overall survival benefits and FDA approvals but an even bigger drop-off in terms of early progression.
I would tell you clinically that it would appear that these patients are getting no therapy whatsoever. Some people call this hyperprogression. That was a term introduced by lung cancer doctors. Melanoma is a very aggressive disease, as aggressive. We never used that term hyperprogression. We just said this is the natural history of disease as though these patients were getting nothing. That really is a striking phenomenon with this therapy, very different than molecularly targeted therapy, oncogene-targeted therapy, where there's this very graded response to non-response profile across the population. Here it's kind of black and white. Patients either just blow right through therapy. That's about 45% of the patients in this overall pembrolizumab cohort, but those were a mixture of some fairly relapsed refractory patients. The treatment-naive patients are on the right-hand side. That's a better estimate. About 35%-40% of patients have de novo resistance.
They just blow right through therapy. You get an effect. There we have stable disease, which is not a great outcome. Certainly, PD-1 monotherapy just don't last very long. Partial responses, the next best thing. Those are absolutely durable, they're not permanent. Complete responses/near-complete responses, meaning you're left with some radiographic artifact. That's what we're aiming for, and that's what accounts for the tail of the curve. That is now supported by data like this. This is complete response data and the durability of that data now with subsequent years of follow-up, and there's trailing data and other tumor types that make this point. If you clear your tumor, you're in great shape for the long term. If you don't clear your tumor, you're going to progress, it's just a matter of time. That was not a statement that people made 4 years ago.
Starting two years ago, anyway, we began to get this very strong thrust of the data that partial responders, middle-of-the-road partial responders are going to progress if they've got viable tumor remaining. Deepening response will result in greater duration of response. That is my contention. Certainly, if we can help patients to achieve a complete/near complete response who are otherwise destined at a partial response, that's going to be beneficial. That's going to translate in an overall survival advantage. If you can overcome de novo resistance, that will be an enormous contribution. I'm going to harp on that point through the rest of the talk as I move more quickly. Okay, you've all seen data like this.
I'd like to go back to this 2014 paper because it was really the first observation in a melanoma pembrolizumab-treated cohort that described the phenomenon that no, it's not PD-L1 expression on tumor cells or infiltrating immune cells that really sorts response, non-response very well, which is really scoring de novo resistance versus the other group. Even infiltration of CD8 into tumor alone is not a great discriminant. If you overlay on top of that a marker of activation, in this case granzyme B, which is an obligate subset of the upper left-hand panel. You have to have T cells in there for them to be granzyme B positive. Those are the patients who respond.
If you have the right T cell repertoire, is the point that I'm driving at here, at baseline, even though the tumor is progressing, the patient's going to die if things don't turn around and patients don't have a treatment response. It's those tumors that have infiltration of an activated T cell repertoire of sufficient quantity, as this data suggests, are the patients who are one drug away from at least inducing a partial response, if not a complete response. That was the starting point for our entry into this area about four or five years ago now, I guess, where we took our serial biopsy approach, which we elaborated mostly for oncogene-targeted therapies in previous years and directed it to PD-1 antibody treatment. We began biopsying patients with superficial manifestations of metastatic disease. This is a cohort of melanoma patients receiving PD-1 antibody.
Began doing single-cell RNA sequencing to try to get more resolution on that data that I showed you before, which is a bulk phenotyping of immune populations in tumors. At the single-cell level, you get really amazing resolution and rather quickly are able to start to sort the T cell population that if it's predominant, associates with response, and if it is in the minority, non-response. This is a published paper from last year, so I won't go through it, but basically, we saw much better discrimination in terms of outcome prediction, ultimately shown here on the lower right-hand panels with these AUC curves.
If the majority of T cells in the tumor had the right state of activation, the right transcriptional state, that associated with response. It turns out that there's even a single transcription factor that we could use in that CD8 positive T cell population to denote the population that are in this proper state. They're not just granzyme or perforin positive. They have to be in the right transcriptional state, and it turns out that that's a state that has stem cell-like features. This is a pool of cells that we think can basically regenerate or repopulate itself and ultimately produce the numbers needed to clear a tumor. If you have exhausted T cells as the predominant set, then basically you're not going to get a benefit from therapy. Can we shape this T cell population?
That is ultimately the goal of these programs, which I'll shift into here in a moment. The alignment between T cell infiltration, if not activation, and tumor mutational burden has been shown by many groups, but going back to this PNAS paper, to be tightly correlated. Those tumors that have high tumor mutational burden have baseline high T cell infiltration. As you well know, tumor mutational burden is tightly correlated with FDA approval. As I've suggested, though, even in tumors that have overall survival benefit from PD-1 antibody, alone or in combination with other backbone treatments, 15%-20% response rate, that's a heck of a lot of de novo resistance, even in those indicated populations. You move out to the zone here of insufficient response rates to translate into approvals. You've got big swaths of cancer, right?
You have some very meaty indications, including much of breast cancer, all of prostate cancer, microsatellite stable colorectal cancer untouched by PD-1 antibody to a large degree. Again, I want to just highlight this point that this is not a net need. There's still a ton of unmet need in the PD-1 indicated tumors. I'm highlighting here in peach those indications where trying to present the entire solution in the form of cellular therapy has been effective, and I'll maybe circle back to that a little bit later. Now you know that there are three programs charging forward in parallel. I'm highlighting here the academic data that ultimately was the foundation of Neon's entry into this space, and BioNTech, as you're aware, also in the first-line cancer vaccine arena.
The approach that's been taken in common with the Moderna program as well, has been to start in settings where feasibility would be anticipated to be at the highest possible level, which is let's say high mutation burden tumors, where you're going to get a lot of predicted mutated neoantigens to start with. You have a pretty good menu, and you can then start putting your informatics/immunologic MHC presentation filters on that to say, okay, can we get an adequate number of peptides with which to immunize? As Sal said, predictably going into these studies, you expect that not all would generate either CD8 or CD4 responses. Start with a large denominator so you can start to hunt, and refine your understanding of predicted immunogenic peptides to actually responsive immunogenic peptides. That's the approach that was taken in this initial academic study.
As you can see here, there was a mixture of both CD4 and CD8 responses. It was actually one of the surprises, I would say of this initial study run by colleagues of mine at Dana-Farber, that there were so many CD4 responses. That was really not anticipated, certainly based on the mouse data that preceded this human investigation. This was adjuvant data. It was a striking degree to which patients not only were able to generate responses to immunized peptides, but also in an uncontrolled population, admittedly, the disease-free survival. A really striking finding presented in this paper, in fact, was that two patients who then subsequently relapsed after receiving vaccine alone got PD-1 antibody therapy and had complete responses.
I highlight that because the proof-of-concept trials that you've seen, not just from Moderna but from the others as well, have been trying to demonstrate feasibility, safety, and really work out the kinks in terms of the operational workflow in relapsed refractory patients, generally speaking. PD-1 antibody patients who are maybe warming up on PD-1 antibody therapy and then start receiving vaccination. We absolutely believe that the adjuvant setting is where we relieve the pressure in terms of the feasibility issues, because you've got a lot more breathing room in terms of having a patient's vaccine be generated then as opposed to a patient who's in the metastatic setting, and already maybe several weeks into it in a disease like melanoma, same for non-small cell lung cancer, head and neck cancer. These are aggressive tumors in the metastatic setting. The adjuvant setting gives us that breathing room.
I just want to emphasize the point that we think is probably also the optimal sequence to be thinking about or being able to interrogate optimal sequence, I should say, in the adjuvant setting where we have a little less of that freedom to operate in the metastatic setting. This is ultimately where we all see these vaccines ultimately taking hold. The BioNTech randomized phase II trial, as you're aware, is in the metastatic setting, but the adjuvant trial, I hope, believe is ultimately going to show the greatest potential impact of this approach. As Tal says, amongst the canonical driver oncogenes in cancer, there's one that shines out as having human proof-of-concept data, that it is immunogenic, that actually you can direct a T cell response to this epitope, and that's mutated KRAS proven by adoptive T cell therapy.
This is the most powerful use of this approach. I've always politely complimented my colleagues at the NCI Surgery Branch in saying, "This is a fantastic research tool, what you're working on here." Never going to be a scalable therapy for patients, certainly not globally, but I love what you can do with this approach to show what epitopes are capable of driving a tumor-clearing effect. As you know, they focused on cancer testis antigens in years past and lineage antigens, specifically in melanoma. This really powerful proof-of-concept results highlights the point that in fact, T cells actually can latch onto KRAS, and being a driver and founding mutation in a vast majority of the tumors in which KRAS mutations are found, certainly seems like the right substrate to direct a non-patient-specific, if you will, or shared antigen vaccine approach.
This is just a schematic that reminds you of a couple of points about where efforts have been directed in years past between lineage-specific antigens for vaccine, but also for cellular therapy. Of course, if you can afford to wipe out an entire lineage, like your B cell population, and you can get away with that for at least a time, then it's okay to direct a fully locked and loaded T cell population at a tumor that otherwise lacks the relevant T cell repertoire, has a hostile tumor microenvironment as well, but that's not very many tissues, unfortunately. You've got your tumor-associated antigens or tumor-specific antigens where most of the academic work in years past and some cell therapy work ongoing, adoptive T cell therapy, not CAR T-cell therapy, is directed now. As we talked about situating where does patient-specific vaccination fit in?
It's ultimately got a potentially much broader footprint in terms of the opportunity space, but it is largely non-overlapping with those domains where cellular therapy will continue to be directed. Let me switch to intra-tumoral therapy, just highlighting where there had been, I would say to a degree, I'm sorry, go back one there. Proof-of-concept findings, either with these pathogen-associated motifs, CpGs, multiple clinical trials corroborating each other that actually can move a needle in an intra-tumoral context or platform, not appropriate for systemic administration. Thing, I would say still not quite there yet in terms of having a clearly established proof of concept. The oncolytic viruses, as complex as they likely are mechanistically, they clearly establish proof of concept in terms of the intra-tumoral approach.
Admittedly, most of this work is in melanoma. As you know, clinical trials have now multiplied in other indications. A nuance of tumor-injected therapy that I want to be clear to highlight as you look down this cascade of waterfall plots with T-VEC in melanoma. In a disease like melanoma, we actually have a fair fraction of patients who have cutaneous, subcutaneous disease that can be injected. It's fairly straightforward to get a response in an injected lesion. In fact, for years we've known that interleukin-2, interferon, in my younger days, we used to inject just about everything, BCG. You could go through rounds of this stuff, you could get responses in injected lesions. The hard trick was what's shown here. T-VEC was really the first agent to show us non-injected lesion response in a melanoma patient that's usually nearby cutaneous, subcutaneous disease, sometimes regional lymph nodes.
Uninjected visceral, that's a big ask based on the precedent data. interleukin-2 and interferon could never do that or any of the other therapies that I mentioned. This is the first one that actually tipped over in terms of yielding that result, and we think that's why it improved overall survival. I realize that the FDA didn't put it in the label that it improves overall survival, but all of us involved in the development of this agent truly believe this data speaks to that point, particularly in lower disease burden patients. Patients without very burdensome metastatic disease, as the subset analysis here suggests. You really can alter the true natural history of the disease with just a single agent injected therapy approach. Modest as it may be based on this data, but it's real. That has just caused this wave of enthusiasm.
Yes, in melanoma where it's so feasible to do this, but there's a big swath of additional cancers as I'll come back to that we think this approach can be directed again. You're aware of this data also, which ultimately prompted the phase III clinical trials for which we're still awaiting results. The point being that on a PD-1 backbone, this same exact therapy now has a much more significant set of waterfall plots. Admittedly, many fewer patients shown here. Again, fairly straightforward in terms of getting injected lesion responses, but these non-injected, both non-visceral and visceral response rates are really what has prompted enthusiasm here. Now, uncontrolled data, of course, has steered us in the wrong direction before. Ultimately randomized data is what's needed. That's an obvious point.
More powerful proof of concept data that you actually are moving the needle both mechanistically and in terms of altering tumor immune interactions. That had never been demonstrated with epacadostat, the IDO small molecule inhibitor. With T-VEC, in fact, it has been demonstrated. Of course, Moderna with their programs in our discussions over the years, we're well aware of the need to be able to demonstrate a proof of mechanism, both proximate expression of mRNA, but also immunologic proof of mechanism, which of course, has been integrated in the early studies that we've collaborated on. Just to highlight here the point that unlike IDO, there are agents that have single-agent activity, have clear single-agent proof of mechanism, and an injectable platform, I think certainly have a basis for receiving our attention in the field.
I do just want to also quickly highlight the point that PD-1 antibody is the big gorilla because it has such widespread use in cancer. When administering either a vaccine or for that matter, a tumor injectable therapy, it remains possible that CTLA-4 and CTLA-4 related priming could actually be a more optimal combination partner. Just want to mention that because it's a trailing strategy, I think that one has to consider that at least spritzing in a bit of CTLA-4 antibody could be an appropriate strategy. This data, I think, was really a striking suggestion in that direction, corroborated by a much smaller cohort with CAVATAK, another oncolytic virus of a different sort, not the GM-CSF secreting HSV, but a coxsackievirus, where again, in combination with ipilimumab, admittedly a small cohort, much more striking results observed.
TLR9 or CpG specifically, as I've said before, have generated recurrent signals across three programs that they can overcome resistance in a subpopulation of patients. It just reinforces the point that injectable therapy really is the only category we have right now in next generation IO, where we've seen recurrent positive signals even in relapsed refractory patients. As you well know, in next generation checkpoints and myeloid cell targeted therapies and other metabolic modulators beyond IDO, we have yet to see those successes. This is, I think, part of the reason why the field's attention continues to be directed so firmly here, including repurposing agents for intratumoral application like CTLA-4, which does have a therapeutic index problem, as you're well aware, both monotherapy and in combination with PD-1. Academic work at Stanford initially sort of proved the concept that this really could engender systemic response.
In fact, a head-to-head comparison, modest-size trial in France is being conducted to see if intratumoral CTLA-4 actually could overcome or could supersede intravenous with a PD-1 backbone therapy as shown here. The idea of manipulating the microenvironment with multiple maneuvers, that's really what got me and colleagues subsequently excited about this platform in terms of intratumoral therapy. We know that even more simplistic manipulations actually can produce benefit to patients. If you could pack into a single therapeutic multiple manipulations to really shape the tumor microenvironment so that those patients who have an adequate T-cell repertoire that's elsewhere in the body, not in the tumor, and it's not activated in the tumor, potentially could have all of the alterations necessary to facilitate activation and entry of that T-cell population.
That's the way in which I see this intratumoral platform as being so complementary to the vaccination platform, which is really trying to boost and create the T cell repertoire. Modulating the microenvironment is a complementary and different approach, which really could have very widespread ramifications in terms of use in clinical practice. I will admit that both academics and the FDA were a little conservative about the idea of just assuming that it would be safe to put "a stick of dynamite" into a patient's tumor, if you're introducing multiple manipulations in particular, there were concerns, I would say go back maybe three, four years ago, that liver metastases, for example, or relatively peripheral thoracic tumors and other abdominal sites could be safely approached with intratumoral therapy.
We've got now phase I data that's quite mature with multiple of these platform approaches that suggest this is in fact not a problem. I know that many people register the complaint that, well, this is going to require interventional radiologists and community-based practices, and it's going to be stressful. If this has efficacy, people will do what they need to do, right? The point being that, as I've said, we know what the phenotype is in terms of an immunotherapy response that produces long-lasting, perhaps permanent benefit, which is clearing the tumor.
If that's what's achieved with combinations like this, a few intratumoral therapies that require a trip to interventional radiology over the course of a couple of months period of time on top of a systemic therapy backbone, if that's what's needed to produce a durable remission that lasts for years, trust me, practices will be rewired to do that, not just academic practices where we've already done this for the purposes of clinical trials, but I mean even in community-based and global practices as well. I will not read these points to you, but I'll leave them up here as we transition on to the rest of the agenda. There'll be Q&A coming a little later. I'll stick around for that and happy to address questions then. Thank you.
Okay, I'm going to continue seamlessly into our systemic therapeutics. This is where we've got a little bit of news today to share with you. To put it in context, as we think about that X and Y or the technical risk and the biology and medicine unmet need, CMV represents, I think, a unique cross on the unmet need horizon. This one represents us extending what we can do in terms of our fundamental modalities. Systemic secreted therapeutics was always envisioned to be a lead-in to the ability to do quite a large significant pipeline of medicines. The first ones that we have that are purely secreted are going to be this antibody program that I'll touch base in a minute. We have a collaboration with AstraZeneca around Relaxin, which is meant to provide a secreted hormone for the benefit of patients with congestive heart failure.
We've got a program looking at Fabry disease as well to prove that we can actually supersede what traditional enzyme replacement therapy can do. There's also going to be intracellular programs that we'll talk about after this. Let me cut to the chase and talk about this program. Admittedly, it took me about a year to learn how to pronounce chikungunya in one flow. The virus is really sort of a part of the show, but not the entire story. What we're doing here is using the platform to teach the body how to make its own medicine in a way. We're giving mRNA intravenously with a goal, let it go to the liver and other organs, and makes a protein that then gets secreted into the blood. We chose a protein that happens to be an antibody for three simple reasons.
One, it's actually a complicated protein, it's not a simple one. It's a heavy chain and a light chain that have to assemble together intracellularly and be secreted in the right forms of the blood. Second, because it's a protein in the blood, it's relatively straightforward to measure, and because it's an antiviral protein, it's got neutralizing activity against the virus, we can actually measure the activity and prove that the protein that we make with this technology is functional. Finally, this has always been of interest to DARPA, who have been funding us from the early years as a target of interest to them. We're using the ability to demonstrate that we can use mRNA technology to transfer passive immunity as a proof of principle for what this platform could potentially do for other systemic therapeutics.
This drug, mRNA-1944, is essentially encoding for the heavy chain or light chain of an antibody. CHKV-24 is the name of the antibody. This antibody was derived from a collaboration with Jim Crowe at Vanderbilt, and it was really taken from the blood of a patient that was convalescing from chikungunya and defined such that it had very potent neutralizing activity against the virus. The idea here is passive immunity, but it's information transfer. We've got the antibody sequence from one patient, and we know that antibody is potent. What we do here is we take the information and we encode that same antibody and have somebody else's body make that antibody in vivo for them. We don't actually have to make the recombinant antibody.
We can react within a few weeks in the case of a future pandemic and just take the information from one antibody and use it to induce passive immunity in a recipient. Okay? That's the idea. This was predicated on a lot of preclinical work, and I'm showing you some of it. We've published this. On the top are the mouse models, and what you can see here is that mRNA-1944 actually encodes for a functional antibody in mice because the mice who gets the drug survive 100%, you can see on the top right, and those who don't all die. mRNA can lead to the translation in a mouse of an antibody that's fully functional. We get very nice expression levels in the mouse.
You can see there the red dots at about 0.5 mg per kg were above 10 micrograms per mL, and you can see the dose response curve. We're getting pretty clear pharmacology here, and this is a mouse that it's got immunological defects, so actually, this mouse requires a heck of a lot more protein to get that survival graph than a typical human would. How do we do translating from mice to non-human primates? That's here on the bottom. This is taken from the GLP toxicology study for this program, 1944. What you see here is the dose response curve, the green, the blue, the red, 0.3, one, and three mg per kg given to non-human primates. You can see very nice dose-dependent pharmacology where we get to levels of 10 micrograms per mL or above already at somewhere between the blue and red here.
You can see, furthermore, that if you come a week later, you can actually give another dose. You continue to boost the pharmacology because the protein has a long half-life. There's an antibody, in fact, been engineered to have a long half-life. This, should it work, will transfer passive immunity that will last for a long time, as long as the antibody levels are above a certain threshold. Those were the pre-clinical data, and with a modicum of excitement and trepidation, we actually took this into the clinic. The question is, can you get similar levels of expression safely in humans? We designed this trial, and this trial, I'll make two points. First, as a new trial for any modality, we care really about three things. Do we understand the safety profile? Is it safe? Do we make protein?
Is the protein functional? We've done that in infectious disease vaccine time and again. I've showed you data in oncology. We've got data for VEGF, and this is now a systemic modality for the first time. The second thing to note as we went into this trial, this has been a journey of many years. Getting a lipid nanoparticle delivery technology to do this is not trivial. As we went through the tremendous effort that Stephen and the team had to put in, it was really understanding the fundamentals of what leads to the expression and what leads to the potential adverse events, and how do we circumvent them. Now, as I've shown you, the pre-clinical data, the toxicology profile was actually quite nice.
What we did here, after a lot of internal discussion, and I'm grateful for the leadership of Allison August here, the physician on my team that's actually designed and run this study. After a lot of internal discussion, we decided that we were going to go first in man into a healthy volunteer setting. We want to understand the full safety profile. We were going to do this with minimal pre-medication. We'll give antihistamines, as is often used, but we were not going to use steroid pre-medications, which are usually given with these types of medicine. I can come back to that. This is a trial design. We start at 0.1 mg per kg, the dose that I've shown you in pre-clinical species. We went on to 0.3, 0.6, and potentially even higher. We do this very measured and safely.
We start with three sentinels, one at a time, and if that's okay, we go on and expose another group of five subjects, three of whom are going to get the drug and two of whom will get placebo. That was with the design. As I've shared with you in the past, as we told you in our August quarterly call, we're actually at the point where we've already treated six of the eight subjects all the way to dose level 3. When we step back to look at the data, we realize we actually have a cogent body of data now that is worth sharing, and that's what I'm going to walk you through. How did we do in terms of protein expression? What you see here on the right are the pharmacology curves for the antibody. I'm measuring the CHIKV-24 antibody.
That's not the drug I've given. I've given mRNA-1944 that encodes for it. What you can see in black is the 0.1, in red is the 0.3, in blue is the 0.6. The reason that you're seeing these graphs go out and get truncated is because these are interim data. The 0.1, we have the longest follow-up, the 0.3, we have up to 12 weeks, you see data for the blue, the 0.6. Just by way of reference, we had predicted, based on those pre-clinical models that I've shown you, that one microgram per mL in the blood should be a protective level against getting chikungunya disease. We got there by the synthesis of two lines of logic. One is the pre-clinical data that I've shown you. The other is simply epidemiological surveys of people who've been infected with chikungunya.
The question is, how much antibody do they have in the blood? We know they're protected against further disease if they get infected again. We had predicted one microgram per mL was a level worth getting to. What these curves show you is that everybody who got the active drug within hours already has protective levels. At the 0.3 mg per kg, we maintain those protective levels projected out to at least four months. For the first time, we've actually taken the sequence of an antibody and used it to generate passive transferred immunity. At the 0.6 mg per kg, we see even higher exposure, and you can see on the top right there the actual exposures. There's one other point that I would make in terms of the exposure here. Look at the variability.
If there's anything here that surprised me, let me just say this, we were all super gratified to see these numbers, but the one thing that stood out to me the first time I saw this was actually how tight these error bars are. This is a phase I healthy volunteer subject. We've treated people with different body habitus. We've treated men and women. We've treated people as young as 23, as old as 50. When you look at our pre-clinical data, the error bars are always tight. Okay, pre-clinical species are always kept the same, and they all kind of look the same. People are very different. When you look at the exposure data, the variability between the lowest and the highest, look at 0.3. It's six to 10 micrograms per mL.
That's as tight as you get with the injection of a monoclonal antibody. The same tightness we see at the lower dose and the higher dose. What that tells me fundamentally is two important things. One is that the fundamental efficiency of our platform is maintained. Because we're measuring, at the end of the day, translation of antibodies. Whatever the mRNA had to do, find its way to the liver, whatever cell, get into the cells, make protein, the protein gets secreted. The end result of that is what's measured in this, and you see very tight variability. The other thing it tells me is that with such a tight nice dose response curve, our ability now to predict and project what it's going to take to make therapeutics in other diseases is actually quite robust.
The half-life here is about 60 days, so if you can imagine, every time I go and double the dose of the exposure, you can expect 2 months more. We're already at a point where we can see 4 months or longer of protection against this vaccine. It is of interest to continue to explore this pharmacology. We're making protein. We're making the protein at levels that look reasonable. We've got the half-life of the protein that we completely predicted, but is the protein functional? The question is, yes. Now on first principles, you would expect it to be functional because how can you make a non-functional protein?
The naysayers will always find reasons why, "Oh, well, if you make an antibody in a liver cell, maybe it ain't going to work because B cells make antibodies." Well, actually, biotech industry taught us that cells make antibodies. A whole variety of cells in the body can make antibodies. If you can get the right sequence into the cells, evolution takes over and now you're making the right protein. Here's the proof. This shows you the neutralization against the virus of the blood from these participants who got mRNA-1944. What you see is that, already at dose of 0.1, you're starting to see a proportion of those subjects get as high as a titer of 1 to 100. It's an arbitrary number just to give you a sense of magnitude.
Every one of the participants who got mRNA-1944 is making some neutralizing active antibody in their blood. At 0.3 and higher, 100% of the participants actually have titers of 100 or greater against this virus. We're making protein, and we're making protein that is functional. All right. The pharmacology goals have been achieved. Now the question is, okay, what do you understand about the safety profile? Let's talk about the safety. You see here the totality of the safety data that we've had, and let me take a minute to walk you through it. At 0.1 and 0.3 mg per kg, we see no clinical significant adverse events whatsoever. Now remember, 0.3 mg per kg was the dose at which I told you we had already surpassed our predicted pharmacology.
Now I'm telling you that at that dose, subjects who got a one-hour infusion, neither them nor the principal investigator could tell whether they were getting a placebo or the drug. As we push up the dose to 0.6 mg per kg, we're starting to see infusion-related reactions. Now, these were not a surprise. These were actually already predicted within the protocol. We anticipated them based on the greater literature. These were the typical infusion-related reactions. You see changes to heart rate, blood pressure, fever, and that's what we saw here. Let's talk about this. We saw it in three out of the four patients, subjects, I'm sorry, participants. One of them had none, the other one had some grade 1. There was a participant that had a couple of grade 2 GI symptoms. In this case, it was nausea and emesis.
We had a participant that actually tipped over to grade 3. They had a rapid heart rate. It wasn't irregular. It was just a regular sinus tachycardia. It did dip over for about 15 minutes on the evening after the infusion into grade 3 territory. That counts as a grade 3. That subject also had additional grade 2 infusion-related reactions, adverse events. Among them was fever, emesis, and on a routine EKG that we do as a routine part of this trial, we noted some inverted P waves. There were no associated cardiac symptoms, and these completely resolved. In fact, all of the adverse events we've seen on this trial have spontaneously resolved, and they resolved without any medical intervention. There was not a participant on this trial that required as much as a TYLENOL. There was no serious adverse events.
There were no discontinuations of infusions or discontinuations from the trial. The adverse events that we see are typical of infusion-related reactions. They come up relatively quickly within a few hours after completion of infusion. By the time the participant goes to bed or wakes up in the next morning, they're pretty much gone. We did see one white count elevation the next day that tipped over to technically a grade 3 level. That too started to come down the following day and then completely resolved. Notable as well, for the first time ever for a technology that we're giving at this dosage intravenously, there were no other adverse associated laboratory abnormalities. We saw no adverse reactions in liver function tests, kidney function tests, other hematological parameters. Where are we? This is an ongoing trial. We think the body of data was worthy of sharing.
Clearly at 0.6 mg per kg, we got some work to do. The obvious next step is to go and see whether steroid pre-medication, as we commonly use for many other medicines of this type, will ameliorate and decrease the rate of adverse events or infusion-related reactions. We can also consider splitting the doses. We're looking at all those opportunities. The salient point here is that when you're giving a protein that you build the pharmacology over time. If you have adverse events that come up quickly over the first day and then completely resolve, that obviously gives you a window then to repeat this. That's what we're going to be exploring in the next trial when we get into patient populations. One last point here. How do we do relative to where we predicted to be?
For those of you who have been following us for years, from about two weeks after I joined the company, I've become infamous for drawing this curve on the whiteboard and saying, "Well, here's where I hope to be and here's where I'm worried about." Right? When we started this trial, I actually asked the clinical pharmacology team to go and draw out, based on all the pre-clinical rodent and non-human primate data, where do we expect the human subjects to land if indeed we can translate this without any loss of potency. When we started the 0.1 mg per kg, based on the predictions in blue, you can see where we thought we would land. This is modeling. That's why you see this shaded area.
That's a 90% confidence interval of where we thought we would be, and the straight lines are the means. Where did the human participants land? Right there. Smack in the middle of where we predicted they were, without any loss of potency in translation. What about the 0.3? This is what our prediction was. The shaded areas appear a little bit wider. That's an artifact of the fact that we're looking at absolute level. Percentage-wise, we're talking about the same level of variability. The red line is the mean. The shaded area is a 90% confidence interval. Six subjects at 0.3 mg per kg. Where did they end up? Right there. Exactly in this area, even a tad higher. How do we continue to dose escalate 0.6? What did we predict? It's in green, shaded area 90% confidence interval. Where do these four subjects land? Right there.
For the first time in history, we've actually shown we can teach a human body to make its own medicine by just giving the information required to make that medicine. I have to tell you, for me personal, while a passive immunity drug is perhaps not as intuitively exciting as an anti-cancer drug or a rare disease drug, the ability to do this from a foundational scientific perspective was really an aha moment. All of us in our careers have these moments that we won't forget. For me, that moment will forever be the day that we sat in the team, in the executive team, we looked at each other, people looked at me and said, "Okay, what are you saying?" Basically, I looked at it and said, "This stuff works." What does it mean for our portfolio?
Well, first, just to recap, we see dose dependence increase in levels as we dose escalate. We see that the protein we make is fully functional as predicted. We see that we have a dose at which we have a very tolerable profile where participants don't know if they've got placebo or they've got active drug, but they're making therapeutic levels of protein. 6-10 micrograms per ml, for anybody who cares to pull up the label of an antibody, is within the therapeutic range of many monoclonal antibodies. In this application, we've reached the ability to confer, or we project conferring passive immunity for at least 16 weeks or 4 months already at that dose of 0.3 mg per kg. We fully predicted the amount of protein that we would make without any loss of potency between pre-clinical species and humans.
Just like 100 micrograms will immunize a monkey and will immunize a human, and a few micrograms will give you VEGF in a rabbit ear and will get you VEGF in a human skin, a 0.3 mg per kg dose will get you 10 micrograms a ml of secretion of an antibody, whether it's a mouse, a non-human primate, or a human. That, I think, puts us in a really good spot for where we're going next, which is this really strongly supports our ability now to get into rare diseases, to go and do protein replacement intracellularly in places where we cannot measure the protein. Greg, Dr. Enns will talk about methylmalonic acidemia. That is the first one where we have the trial open, and as he'll share, what we're trying to do there is actually solve for a protein intracellularly.
We're not going to need to measure that by doing biopsies because we already know that in the case of a secreted protein, we can achieve this pharmacology. Thank you.
Sure. Sit down. It's my pleasure to be here. Now for something completely different or along the same track and set up so nicely. I'm a biochemical geneticist, and that often requires a little bit of explanation to start. I am a pediatrician by training. I trained in medical genetics at UC San Francisco, then I further trained and specialized in biochemical genetics. I take care primarily of children, but I do see adults as well who have inherited inborn errors of metabolism. These individuals often look absolutely healthy and fine for at least a few days or a few hours if they're more severely affected.
Many of the conditions that I care for are detected nowadays by newborn screening. Methylmalonic acidemia is one of these as an example, I was delighted to hear that Moderna is going into this area because there's still a large unmet need for these disorders as a whole and methylmalonic acidemia as an example, for sure. I was asked to give a bit of an overview of MMA or methylmalonic acidemia, I'm happy to do that. Here's my brief disclosures, moving along. MMA, like many of these conditions, is an autosomal recessive disorder. Like any of the genetic conditions I care for, these are not binary diseases. You're not 100% switched on or switched off if you have an enzyme difficulty or problem. There's a range of enzymatic activities that are determined by the underlying genetics.
The more severe form of the condition is, I call mut0. Chuck Venditti at the NIH says mut, but I will stick with mut because the enzyme name is mutase. The mut0 is the most severe form of the disease. A more partial form is called the mut-, where you have some residual enzyme activity, and these patients are typically more mild in presentation. The enzyme we're talking about is that methylmalonyl-CoA mutase. In North America, the prevalence is about one in 50,000, so this is indeed a rare disease. When you lump all the rare diseases together that there is screening for in California, for example, the incidence is closer to about one in 1,000, one in 800, depending on how you count them. Like many inborn errors, you have to have a pathway.
Sorry to hit you with a biochemical pathway, this is just to show where the enzyme has its activity. It's a nuclear-encoded enzyme, but it's localized in the mitochondria. It's synthesized in the cytoplasm and imported. It's a homodimer and requires a form of vitamin B12 or an activated form of vitamin B12 to work. It's in a pathway that involves the degradation of a number of amino acids, which conveniently make up the mnemonic VOMIT. When I'm teaching medical students, it's very easy. It's valine, isoleucine, fatty acids, methionine, isoleucine, and threonine. It sounds so good. You think, "Oh, that's just VOMIT." These patients often will come to us with that problem, and that's one of the things that we see in a viral infection or viral illness.
These kids will get severely ill, and the catabolic illness itself is what stimulates the metabolic crisis. When they're coming into a hospital with a problem, it's not that they have MMA per se. It's that they have an infection that is causing them to be catabolic that is then really stressing that pathway. The pathway is involved in line with the propionic acid or propionyl-CoA carboxylase, and this is a metabolic pathway that eventually yields succinyl-CoA that can be used in the TCA cycle. Biochemical pathways are typically listed like this. This is how we learn them. There's such an integration with biochemistry and metabolism here that it's always a little bit more nuanced or a little bit more complex. This is not quite PKU. In PKU, there's an enzyme deficiency. Phenylalanine builds up. Phenylalanine is toxic.
You decrease phenylalanine, you take away the toxicity primarily. Here there are multiple inputs in a bioenergetic system. What do we see? First off, although we have newborn screening, we tend to pick up all these children nowadays because of the newborn screen. Oftentimes they're coming in for a presentation before the newborn screening results are even back. These are children who are coming to the neonatal intensive care units with very acidotic, the pH is typically less than 7. They have often hyperammonemia as well, and they're going into a coma. They present with lethargy, and they have a severe presentation that is dramatic and might need dialysis to correct. If they present later, they will have more mild disease. Still, having some form of neurologic involvement is very common. Here are the sort of acute and chronic presentations of two common organic acidemias.
Propionic acidemia is included in this slide here. It's a nonspecific neonatal sepsis-like picture that is common in the neonate with a classic mut0 disease. Altered levels of consciousness in older children, having ataxia, lethargy, even mistaken for drug overdoses is relatively common as well. GI symptomatology, pancreatitis can develop in individuals. You can have vomiting and other problems. Interestingly, for MMA in particular, there's a renal tubulopathy, and these patients will develop kidney disease typically later in life. Later for a pediatrician, this is in adolescent years often, but depending upon the severity of the presentation, you might see kidney disease develop in the first year or first few years of life as well. The thing that keeps me up at night or keeps me pacing around is the fact that these children are very prone to brain injury.
We think this is in part due to the reliance upon the important parts of our brain, the basal ganglia, the deep gray structures on oxidative metabolism. Even if we have a diagnosis of MMA and we're treating a child, we'll go into therapy in a little bit. We're doing the usual medications, dietary interventions, and things like that. These children can come in acutely and have acute decompensation resulting in permanent brain injury that can lead to a movement disorder, or a child who was walking, talking, going to school might not be able to do any of those things again after they've had a crisis. We treat, like many of these conditions, if you have a block in a pathway, we decrease the flux through that pathway. There's special formulas that are low in the amino acids that are used in the methylmalonic pathway.
Every patient, every family has an emergency letter, a sick day protocol to come in just to make sure they can contact us at all time. Carnitine is often given as a supplementation, because methylmalonic acid and other metabolites bind to carnitine and deplete the body of carnitine, which is an important part of our energy process. Some of the MMA forms, not the classic mut0, typically can respond to vitamin B12. We give a trial of an active form of vitamin B12 and hope for the best, most of the time, they do not, unfortunately. When they're coming in, I already mentioned dialysis, because the ammonia levels can be just as high as you would expect to see in a classic urea cycle defect.
We're talking about children coming in with ammonia levels of 2,000 or 2,500 micromoles per liter or something like that, with normal being around 30. In the neonate, closer to 80. Renal failure, as I say, happens later. What we've done, and one of the things that we've started to do, is treat with a liver or combined liver-kidney transplantation. Others have used kidney transplantation alone. I'll go into a little bit of the data. The nutshell here is that kidney transplantation alone does not restore propionate metabolism as well as liver transplantation, and the combined liver-kidney transplantation does a wonderful job of restoring body in vivo propionate metabolism. I call this gene therapy with a scalpel. We have excellent transplant surgeons who are at the ready at all times, and they've done a wonderful job with our patients.
As I said, we've done mostly liver or a combined liver-kidney transplantation. I say we don't do a kidney transplantation, though we would consider it, and we're considering it in a case of an adult who has a more mild form of disease where that kidney transplantation might be a little bit less invasive. It's something to still keep in the back of our minds. The thought of doing the combined liver-kidney transplantation goes back several years. This is, I think, the first article that I know of that reported this. It's Dr. Leonard's group in the U.K., and this is from a now, gosh, going back a couple of decades in our patients because we had at least a five-year follow-up here. These are the first two patients that we did a combined liver-kidney transplantation at Lucile Packard Children's Hospital.
What you can see is that the MMA levels dramatically fall. We can measure blood and urine MMA levels, and you're seeing this is in micromolar, up to over 2,000 micromolar in one patient, and it looks like over 10,000 in another. Realize normal level is 0.3. These are exceptionally high levels. We get these exceptionally high levels in children who also have that kidney failure. The typical mut0 MMA patient will run levels, if they do not have really significant kidney involvement, between 200 and maybe 700 micromoles per liter. This is a paper that we presented on a group of patients, I think we had 14 or 15 reported here, that underwent either liver-kidney or liver transplantation. The mean age of transplantation in our group here was close to nine years.
This is, again, going back to when we first started doing this work. Currently, when we're seeing children who need a transplantation with mut0, we're typically transplanting them between six and nine months of age. Try to get things really early, and it's only going to be liver because the kidney is typically working just fine at that age. Just to give you an idea of at least this cohort, patient survival was 100% in the children we transplanted. We had one liver have a hepatic artery thrombosis that required a second transplantation. You can see the dramatic drop in MMA, just similar to other cases that I've already shown you. The main outcome here is you don't have that threat of hyperammonemia, which is a severe condition that can permanently affect the brain as well.
None of our patients had metabolic acidosis following the transplant procedure. Renal function seemed to stabilize following transplant, but this is an ongoing question, especially because some of the transplant anti-rejection medicines will also affect renal function. We're following these patients. The neurological outcomes, they don't reverse. If you have permanent hits or if you have specific damage, you're not going to change that. After the transplantation, these children tend to at least maintain what they had before and then improve. Socially and functionally, they seem to be much more interactive. They do better if they survive the procedure and go on. The complications have been described in a number of publications, mortality being number 1. Even if you transplant, some individuals have gone on to have what we call this metabolic stroke or the basal ganglia burnout.
When you look closer at the data, most of the time, those are patients who have been ignored or not treated aggressively following transplantation. In these disorders, these organic acidemias, it's a little bit different than the urea cycle defects. Once you've put in a new liver for a urea cycle patient, they don't have hyperammonemia anymore, and hyperammonemia is the main issue with those patients, so they're pretty okay. We treat our post-transplant organic acidemias, including MMA, the same as we do pre-transplant, and then we slowly start to liberalize protein, slowly start to decrease the carnitine levels, slowly start to let them have a little bit more normal diet and things like that.
What I think transplantation does is at least it increases the bandwidth of their ability to sustain an intercurrent illness, because if they do come in with a vomiting illness, they're much easier to treat. We never need dialysis. We never need anything else after the transplant procedure, so to speak. Other complications listed here, immunosuppression, of course, and the actual surgical complications as you would expect. Post-transplant outcomes are always sort of still slender a little bit there. Another publication showing accumulated data shows that after the transplant, their crises go down, and the body weight goes up because you can eat better and you have a more normal diet and more normal existence. There are several therapies in development. gene therapy, there's been some pre-clinical data in gene therapy publications. Of course, what we're talking about today, we'll get to as well.
AAV has been the primary focus of gene therapy in pre-clinical models. There have been some evidence of hepatic genotoxicity and immune responses, including neutralizing antibodies, which is, of course, a concern, as would be insertional mutagenesis. These are just some of the accumulation of data. Adenovirus, AAV, and lentiviruses have all been used in pre-clinical models, usually just mouse models, to show the efficacy. You can see that there is a reduction of methylmalonic acid as well as an improvement in weight gain using, I think, on the right, we have an AAV9 vector. Yeah. Sorry. On the right here, we have AAV8. On the left, we have AAV9. You can see that there is at least some pre-clinical data that shows a response.
What we're talking about today is something that it's exciting for me because although we do a lot of liver transplantations, we might be the leading place for transplantation for this specific indication for MMA, I don't like to do it. It's an involved procedure, and there are some of our patients, even with our outstanding transplant team, who opt not to undergo the procedure because of the risks. When I saw this article came out, this was before, I think I really was talking a lot. I presented it at a journal club. I was very enthusiastic about it. I don't think I ever told you that, Tom. The mRNA therapy, of course, has a potential to change the intracellular environment, produce a normal enzyme. I look at it a little bit like enzyme replacement therapy as it actually is.
We do a lot of ERT at our center, usually for lysosomal disorders, of course, and we used to take valuations well. We used the injectables treating inborn error of metabolism. This is quite exciting. There are a couple of preclinical papers that I put there, and I think I just have maybe one slide here that basically shows that the improvements in the metabolism, the production of protein intracellularly in the liver following the injection of the mRNA lipid nanoparticles in a mut- minus model on the left and a mut0 model on the right. You're seeing an improvement in the MMA levels and an improvement in the overall growth, very similar to what we experience and see in our patients. I look forward to the day we can start the clinical trial at our center.
In summary, this is a severe disease, especially the mut0 phenotype, and it's associated with high morbidity and mortality. We have used liver or combined liver-kidney transplantation which has improved the overall life of our patients, we think. It stabilized their metabolic crises, et cetera. I'm looking forward to working further. You see a crystal ball sort of waving on the side there. This is the conversation that I have with my families, and this is often the tipping point of why we even would pull the trigger on a transplantation. We might have a child look good for a period of three, four, five years, and then have a crisis and not be able to walk, talk, and have a permanent movement disorder. We can't predict that. Even with the best dietary therapy and aggressive emergency management, we still can't predict that.
That's often the tipping point of why we go ahead and do the transplantation in the first place. That's a little bit of the background. Thanks for your attention.
Thank you, doctor. Just a few slides to wrap up, and then we'll get the entire executive committee to wrap it there. We're very proud, of course, of the announcement of this morning, both for CMV and the CMV data, as well as chikungunya. I want us to step back and reflect on the fact that this was not easy. This required a lot of work and a lot of capital. I would like to take this opportunity to thank our investors and to thank DARPA, who had faith in our ability to hit it on the science back in 2013 when we were a team of 20 people and could barely dose a few mouse, to be able to see that this could be possible if the right team had the right resources to do the science correctly, and this is what the team has done.
I think it's really a useful combination of a great scientific focus with the right partners and the right capital. What is nice about Moderna, as you can see, is the company continues to mature. The clinical data that we presented today and we have over the last month just confirm our thesis that we believe mRNA could be a new class of drugs. We now have 4 programs in phase II or preparing for phase II. 10 positive phase I, which is a very large number. Third program is at phase I as we speak. If I look across the company, we believe we have a very strong vaccine modality. 6 positive phase I data. We have 3 vaccines which we believe have blockbuster potential. The CMV vaccine, we talked a lot about today. Let's not forget, we have also an RSV vaccine.
Same thing, very large unmet medical need, no vaccine on the market. The combination of a vaccine for hMPV and PIV3, two more respiratory viruses that are the third and the fourth after flu, RSV, hMPV, and PIV3. We're combining in one vaccine, two mRNA, one to provide protection against hMPV and one against PIV3. In IO, we have five programs that are in a clinic in phase II or in phase I, and we're partnering with the best company in the world. We want to see that data coming through, and we're very, very eager for those readouts.
With today's news about the efficacy and the safety profile of chikungunya antibody, we're very excited that not only we can show we can make a functional antibody, but of course, that the technology that we are going to be using for rare disease can be effective at a safe dose for those kids that are needing those medicines. Let me really spend two minutes on vaccines, because we believe at Moderna that vaccine is a great business for following reasons. We think that there are three big buckets of vaccines in the world. There are me-too vaccines. Those are not necessarily our key focus. There are innovative vaccines like Prevnar and HPV, then there are public health vaccines that are important for the world, but for which one should not anticipate to make a return. The reason for that, of course, is pricing.
You look at it, the me-too vaccines are very low prices. You see it when you go and get the seasonal flu shot at CVS. Innovative vaccines, we talked about it, Prevnar cost of treatment is around $750 in the U.S. HPV is around $450 for a cost of treatment. Yellow fever in some countries sell for literally $1. This has, of course, a big margin, very obviously. In the me-too vaccines, we have no intention to develop me-too vaccines with Moderna's technology. It would not be a good use of our resources or talent. Of course, if somebody wanted to come along and partner on a me-too vaccine with their capital, we'd be happy to do that. We really want to focus on innovative vaccines.
For public health vaccines, as you know, we have a thinner pipeline, we believe it's an important part of our mission and the responsibility of a company. We want to partner with foundations like we have done before. We want to partner with governments because we believe it is very important to use our technology to get those vaccines to protect millions. We will not do it with our shareholder capital. That would not be the responsible thing to do. We are really open for business like we have done with DARPA, with BARDA, and also with the Gates Foundation, with which we are partnering for projects that are in research. The big focus of Moderna is on innovative vaccines.
Let me share a few examples of innovative vaccines to share with those of you that are not as familiar with vaccines, what they could look like. Prevnar has a few interesting characteristics. This is a vaccine that Pfizer is commercializing. The turnover in 2018 was $6 billion. It's the number one product of Pfizer. There's no product that Pfizer has that has a bigger turnover on an annual basis. Look at the launch year. Prevnar seven, the first version of Prevnar, was launched in 2000. Look at the prediction for 2024, I can bet with you that Pfizer is going to be selling Prevnar for many, many, many years to come. Another example is Gardasil, the vaccine that Merck has to protect against HPV. Gardasil was launched in 2006, it's already quite a number of years ago.
This is the third product from Merck. The first product, of course, is KEYTRUDA, a wonderful drug. I think that by the time KEYTRUDA goes off patent, which will happen next decade, HPV will be bigger than what it is today and what it's going to be in 2024. Merck has announced they are building additional manufacturing capacity. As you know, HPV needs to be protected in children, both boys and girls, because it's very proven the long-term effect in cancer. It's a very important vaccine. We believe this vaccine has a very long life cycle, and it's going to generate a great business for Merck. Let me now close with SHINGRIX, because sometimes I hear vaccines are very slow uptake. It takes years to get to a couple of $100 million and to $1 billion is almost impossible. Look at SHINGRIX.
It was launched in 2017. First full year, $1 billion of sales in 2018. This is the fifth product of GSK to date. If you look at the GSK products that are on the top five, I'll bet with you that in a few years, SHINGRIX will be the number one product of GSK. I have no doubt. We believe that innovative vaccines are first very important for public health purposes, obviously, and they could be a great business for our shareholders' capital. If you think about our midterm strategy, the number one priority of a company is to really build on this strong vaccine franchise, and continue to innovate and do research on innovative vaccines. Stephen and Andre and the team are, as we speak, working on additional new innovative vaccines, but already in a three big blockbuster potential.
We believe the cap could be a couple hundred million dollar a year. This is what yellow fever sells every year. Remember, yellow fever is supply constrained. Every year, people are missing doses of yellow fever. It's even possible that it's higher. If there's a pandemic, you could see that it will come in full like big spike, but it's not sustainable. Executing on this really exciting immuno-oncology pipeline, where as Tal said, our strategy across our five programs is to see can we improve the response of checkpoint monotherapy. Checkpoints have been wonderful for patients. We are saving people's lives every day, but unfortunately, they do not work for everybody. There's still a massive medical need out there, and we are hoping that some of our products could help patients who are fighting cancer.
Now, of course, with chikungunya human data to go and to really start clinical trials, which we are very eager to do, not only in MMA, but PA is just behind it. We've just announced GSD1a earlier this year. We have a very nice portfolio of five rarities. For those of you who follow our announcement and our presentation, Stephen and team have some very interesting new science that they showed at Science Day in May. We anticipate in the future to also be able to expand into new therapeutic area. To close, there's still a lot of work. We're not done, and nothing's going to be done for a long time because we believe this technology is very powerful. We believe that by working together as a team, this company can have a big impact on patient care across many therapeutic area.
Think about what we've shown to date now. We've shown we can code an mRNA safely in a human to make a viral protein. It's called an antigen. We've done that six times. We've shown we can code an mRNA to code a human protein, VEGF or 4-1BB as demonstrated by. With today's news that we can code an antibody, we basically have a full toolbox of what you need to go help patients. We can do that in a secreted space. We can do that in a transmembrane space. We can do that in an intracellular space. I think we can barely imagine what the future holds for this technology and for what we can do for patients. A few years ago, when we were not in a clinic, I had no idea that one day we would have a KRAS vaccine.
I had no idea until the NCI paper came out, and Tal came back from NCI with a big smile on his face, saying we should really do a KRAS vaccine. There's good scientific ground now. We have an enormous tailwind. It's called the academic world. We are living a biology revolution around the world, where every day we learn more at an exponential pace about what proteins do in the human body. What a wonderful time to be building this technology where we have this team that has a serious commitment to science and to invest in science for improving and extending the technology. The ability to use Norwood to be able to make quickly high-quality GMP products, and the clinical team to put that into music, into the clinic, so we can try with clinical research to see is our hypothesis correct or not.
As we said before, and we are doing that with PCV, with the randomized head-to-head, we care about knowing are our drugs going to work or not. We've said it many times, we're not in a business of doing any conclusive study and another inconclusive study just to keep the product in a pipeline. We want to be very disciplined with our shareholders' capital. We want to put it behind the drugs that have a highest chance to get to approval. That's what we will keep on doing. We feel all very fortunate at Moderna of the mission that we have. We feel very lucky. We think it's a once in a lifetime opportunity to be able to go to work every day with amazing people to try to do something that has never done before to help others, to help patients.
That is what is giving us so much energy. With this, I would like to close. Thank you very much for attention, and I would like the executive committee to please join me so we can take questions. Thank you.
I believe there are two from there.
Yeah. For sorry. The last Yes, as well. Laurie and Megan, Stephen, what am I missing? Oh, it's coming. Good. Since we have some mics, Keith, we're going to start, not waste any time we have . Thank you, Ted.
Great, thanks. Two quick questions, if I may. Tal, I'm wondering, again, congratulations. I agree with you, that's a monumental achievement, really cool. I'm wondering how we should be thinking about immunogenicity for the proteins and antibodies that you are creating and are secreted. I'm not sure if you analyzed that, if not, maybe you can walk us through the thought process about what we should think about that, because it is being generated by the body, is there the potential for immunogenicity or anti-drug antibodies? Thank you.
I'll take a stab at it, and maybe follow it up with anything. It is something we'll naturally look for in our phase I and over in use. As relates to mRNA-1944, you see that very long predicted clearance. Certainly, there hasn't been a clearance reaction against that protein.
I think it's on.
Yeah.
I don't see it. I was like, "Damn it.
It is something we're always worried about, and we think about very specifically as we design the drugs. All of our therapeutic platform is actually a very different platform than our vaccine platform. We use different chemical components. We look at nanoparticles. We send them to different cell sites. Actually, we published also, we even put in microRNA sites in the mRNAs encoding our therapeutic proteins to prevent their expression in immune cells or antigen-presenting cells. All of that is belt, suspenders, and a couple of other things.
It's fair to say that as you look at our preclinical data, whether that's the GLP toxicology work that we've done in primates multiple times, or the publications even that some of the presenters put up there of our work, we've not yet seen an immune response to the encoded therapeutic proteins when we do all of those things. It's a mix of different process for the mRNA, different encoding elements, and very different lipid nanoparticles that we think is pretty essential for that. We will always be looking in the phase Is. Obviously, if we see anything, we'll let you know.
Yeah, I'm thinking about that specifically for chronic dosing. With this proof of concept in hand, if you can remind us the status of mRNA-3704, then also, is this the green light you were waiting for to really expand and continue to invest and evaluate other opportunities where mRNA could be used to treat orphan-type diseases like this? Thank you very much.
What we want to do at this stage, Ted, is to prosecute the portfolio we have. With 500 diseases, we think there's plenty. We want to make sure we focus on execution. We want also to be careful about our expenses. As we've done in the past, the strategy has always de-risk with first program, which can be antibody, try a few programs to confirm that you're on the right path, and then we can double down. Stephen and his team are working still on more rare diseases, but we do not anticipate right now to take more into the clinic because it's just totally correlated risk. We want to see the MMA readout, we want to see the PA readout. Trust me, they are not working idle. They have more stuff cooking in the lab. Now we have to be cautious.
Any more questions? Yeah.
On the cancer programs, I know a lot of it's partner-driven, but I'm just wondering if there's any data that we should expect maybe over the rest of the year or early next year.
On our partner program?
IO.
On the cancer, any data?
There's a lot to expect. The challenge, as Keith will readily tell us, is it's hard to know when it's going to come in because really what we're trying to do now with our programs in the IO space is see responses, right? At least for those phase I programs that are open label and looking for responses, their power to a certain size, we're treating certain cohorts, but of course you see the data as it matures. As we treat more patients, we have the opportunity to see more responses. I think where that's different is in the randomized phase II. That trial will take time, as you can imagine from the statistics. It's powered to actually enroll everybody and then follow them for a year of landmark analysis. Remember, these are patients in the adjuvant setting.
On one hand, as Keith alluded to, and I completely share his belief, that is the space you want to go improve the utility of something like a personalized cancer vaccine, especially now with the benefit of KEYTRUDA in that space. On the flip side, you actually don't see the benefit until you accumulate enough events over time and you look at it versus a control arm. In that aspect, that one we're going to have to be patient and let time do its thing. The other programs, and that's true of our partner programs with AstraZeneca as well on IL-12, it's open label. As responses come in, we have an opportunity to see them.
Thanks, Rod. Yeah, there's a question in the back.
Hi. Two quick questions. Rosenberg selected for immunoreactivity to KRAS G12C. I don't know if you're selecting for that, you're just targeting it. I wanted to hear your compare and contrast the NCI approach versus your approach. Second question is, there's a therapeutic index to the amount of LNP that you can give as defined by your results. A T cell question and an LNP question.
Let me try and take these in sequence. Rosenberg showed that that one KRAS in the context of that one HLA was actually a tumor rejection antigen if you grew the cells ex vivo and gave them back. Our vaccine is trying to expand on the mutation spectrum, the patients enrolling in this trial are going to be patients who have any one of the four mutations that is in the vaccine.
Then we will look post-hoc and see who did what in terms of seeing any responses or immunogenicity, because immunogenicity for some of the other epitopes concerned with some of the other HLAs is yet poorly understood. Where we are trying to mimic Steve's approach is on selecting the HLA, the people with the right alleles, at least as the starting point for a proof of concept. That phase I, the first dose confirmation for safety is in an all-comers population. Immunologically, who've got the four mutations. The next phase of the trial will be focusing in, at least to start with, on the HLA-A11 and C08, as in his case, and then testing anybody who's got any one of these four mutations, and then we see what we get. That's on the KRAS.
In terms of the LNPs, look, I think these are early days and preliminary findings. You're seeing a study that's ongoing. I think the fact that we're seeing infusion-related reactions at this rate of 0.6 means we got to do something different. For me, the obvious next step is to go to minor with steroids. To the degree that that ameliorates, as it does for many other injectables, I would expect that we could be able to give even 0.6 mg per kg at a safe and well-tolerated level. It depends on the pre-medication regimen. As an example, back when Keith and I were training, every fellow was taught that with RITUXAN, oh my God, if you've got infusion-related reactions, you'll give them steroids.
If you look at the label for Remicade, we give 100 milligrams of methylprednisolone every time you give it to a rheumatoid arthritis patient, every two weeks. It's not a showstopper. patisiran, that got approved last year, is a lipid nanoparticle with an siRNA. That's given at 0.3 mg per kg already with steroids. They never actually tested it without steroids. I think for us, it was important as the first step into this platform, before we characterize the safety and tolerability first without it, before we resort to that. I would point to the fact that the starting dose for MMA, for methylmalonic acidemia, is 0.2 mg per kg, and that's a dose that's predicted already to have the possibility of benefit. That's very nicely bracketed between the 0.1, 0.3 that you see here that had absolutely no significant adverse events.
I think it's early to say that we know what the top dose is, but we pretty clearly have cleared a dose at 0.3, where we achieve therapeutic levels of protein and the adverse event profile is so benign that you don't notice somebody's on drug as procedure.
Thank you. Any other question? Yes.
Thanks. Hartog, first of all, it's this amazing data on the antibody. You've created a memory for me at least, seeing that title slide with me, always smiling. Just a quick question on you can get, I guess, the primary structure of the antibody right from the mRNA, the coding structure. There's a secondary, tertiary structure to various antibodies. Can you just talk about when you're thinking about future antibodies to try to come up with these approaches, does that play into how you think about existence kind of screen there or not really? You think you can pretty much go into any antibody that's available there that the human body can do so?
I think the right person to answer that is actually Stephen, because he's much smarter than me in thinking about the tertiary structure. For me, they kind of look the same. You just take the sequence and then you get what you expect.
I think the question was, is the complexity of antibody structure impacting maybe our next choices on discovery programs? The short answer is, given that we've been able to make any protein we've chosen to make so far, it does cause us to look at what are proteins that people have struggled to make in other ways because obviously there can be a real advantage of this sort of approach. We're not ready to guide on specific programs under that, we'll continue to look at that. We are also in the preclinical space where you can increase your N. We're always looking to characterize the quality of those proteins and particularly not just tertiary structure, but also glycoforms, that glycosylation both for lysosomal and extracellular proteins.
We're actively looking at those things, and much of that data will probably come out in our preclinical publications where we do much of that work.
Thank you. Any other questions? You're good? Thank you so much. For those of you that can stay, there is food on the other side, and we will just go at different tables to be able to be available. Thank you.
Thank you, guys.
Thank you again.