Novavax, Inc. (NVAX) Earnings Call Transcript & Summary

Ladies and gentlemen, thank you for standing by, and welcome to the Novavax provides update conference call. [Operator Instructions] Please be advised that today's conference is being recorded. [Operator Instructions] I would now like to hand the conference over to your speaker today, Silvia Taylor. Please go ahead.

Thank you, and good afternoon, everyone. Thank you for joining today's call and webcast. Earlier today, we issued a press release announcing top line data results from our recent Phase III trial in the U.K. and Phase IIb trial in South Africa. That press release is available on our website at novavax.com as our slides we will be using during this afternoon's call. An audio archive of this call will also be available on our website later today. Before we begin, I need to remind you that we will be making forward-looking statements during this teleconference that could include financial, clinical or commercial projections. Statements relating to future financial or business performance, conditions or strategy and other financial and business-related matters, including expectations regarding revenue, operating expenses, cash usage and clinical development and anticipated milestones are forward-looking statements within the meaning of the Private Securities Litigation Reform Act. Novavax cautions that these forward-looking statements are subject to numerous assumptions, risks and uncertainties which change over time. Joining me today are Stan Erck, President and CEO; Dr. Gregory Glenn, President of Research and Development; Dr. Filip Dubovsky, Executive Vice President and Chief Medical Officer; and John Trizzino, Executive Vice President, Chief Commercial Officer and Chief Business Officer. I would now like to hand the call over to Stan. Stan, go ahead.

Silvia, thanks, and thanks to everybody who's listened in on the call. We are at a great point for the company right now. We've been working for -- this team has been working for a decade building a platform that can make important new respiratory vaccines, and we've been doing so with RSV and flu. And for the last year, we've been virtually solely dedicated to developing a vaccine against COVID-19, the pandemic flu that surrounds us. We have breaking news today, and we'd like to tell you about it. And that breaking news -- I can't see slides. Are you turning the slides?

You don't need to. We're turning them.

Okay. Good. Sorry, just catching up here. So we had -- we've unblinded 2 trials, 2 efficacy trials. The first one was in South Africa, was a Phase IIb efficacy trial of 45 volunteers and a 15,000-person Phase III pivotal efficacy trial, where we have an interim look at the data and have shown, I think, profound efficacy against COVID-19 in these trials and with both the U.K. variants and the South Africa escape variants that you've been reading a lot about over the last few weeks. We've shown strong efficacy in the Phase III U.K. trial and over -- even with over 50% of the cases attributable to the now predominant U.K. variant and the remainder attributable to COVID-19 virus. In addition, we've shown clinical efficacy demonstrated in the Phase IIb South Africa trial, even though over 90% of the sequence cases are attributable to the prevalent South Africa escape variant. And in the U.K. trial, I should point out that we've got 89.3% efficacy against the combination of the COVID-19 strain and the U.K. variant. And just a couple of hours ago, we received data that allows us to estimate what our efficacy was against each of those strains individually. I'll highlight it now. And then in the presentation, Dr. Glenn will additionally present some data. So on the COVID -- in the U.K. on the COVID-19 trial. And on the efficacy data against COVID-19 was 96% efficacious; and with the U.K. variant, it was 86% efficacious. And I think both those numbers are dramatic demonstrations of the ability of our vaccine to develop very potent immune response against a like strain and also against drifted strains. So with that, I'll move on to the next slide, which is the agenda. We'll cover the Phase III results in the United Kingdom. We'll cover the Phase IIb trial in South Africa. And we'll also give you a brief update on our enrollment because we have a Phase III trial in the U.S. and talk about our plans for what's next. With that, let me turn it over to Dr. Glenn.

Thanks, and good evening, everyone, and I'm very pleased to be here tonight. I'm going to start off the presentation with some science. We are a company that is, I think, has some very good technology. We -- here, we are making a fully prefusion stabilized spike protein, as you can see here. And we adjuvant this with Matrix-M, our proprietary atom. This is important. In fact, I think this, to a degree, explains the results that we have here today. Our spike is a -- has good fidelity with the native structure, and it presents the epitopes that are critical for protection, and Matrix-M greatly enhances that, and it allows us to create a broad level of highly functional immunity to the COVID virus. So I'm going to now turn to the U.K. Phase III study, and I'll take you through the top line results for that. I will say this is really just the top line results. So we're providing you all the information that we have analyzed to date. So this was a randomized observer-blinded, placebo-controlled trial. It was a little over 15,000 adults greater than 18 years of age. Over 25% were over 65. And as you can see there, the dose of our antigen was 5 micrograms and 50 micrograms of Matrix. It's given as 2 injections at day 0 and day 21. And the primary end point was a PCR-positive, symptomatic, mild, moderate or severe COVID illness, and that was diagnosed 7 days after the second dose. So this next slide shows you what we were up against. So as Stan noted at the beginning, we embarked on this trial with the understanding that this virus might evolve, but it's been quite a dramatic, I would say, evolution that has occurred. And this here -- this illustrates in a time line the evolution of the virus and the prototype strain, which you can see in the far left, all the way to the right, where you can see now this -- the B1 strain, which is the U.K. variant. And you can see what's superimposed here in the green box is the time during which we were doing the surveillance. So you can see from the beginning to the middle to the end, there's a rather dramatic change. And approximately 50% of the reported strains in the U.K. were of this mutant variety, and that's important. So if you go to the next slide, what you can see here is that we've met our primary end point. This analysis of -- was in fact interim, which meant it was yes, no. Once we got a yes, which was that the lower bound of the comps interval was above 30, this became the final analysis. So you can see here, there were 6 subjects in the vaccine group, 56 cases in the placebo group. And the breakdown is 1 in 15 for mild, 5 in 40 for moderate, and we had 1 severe case in the placebo group, and for an overall vaccine efficacy point estimate of 89.3% with confidence intervals of 75.2% to 95.4%. So this -- and in this trial, our data was showing us that more than 50% of the cases, as you might expect based on epidemiology, were attributed to this escape variant. And we are going to do, as you probably know, when these analyses work, we will continue to accrue cases, and we'll be able to do a final analysis since we have succeeded statistically on the additional cases. And we'll report on that at that time. So we just got some new information recently, and I'm sorry for the appearance of the slide. But this is a summary of the -- by PCR of the breakdown between the variant and non-variant cases here. You can see we have sequence data on most of them. But the -- you can see there's 4 variant vaccine cases, 1 non-variant; and you can see to the right, 28 and 23. So what this tells us is that the vaccine is working as expected, very, very well against the matched COVID-19 strain; also very well, but of slight different in the efficacy against the variant strain. And this is really -- of course, really encouraging data to see that the vaccine is working so well even though there were significant changes in this variant. But also, I think this replicates the -- the 96% efficacy is, I think, meets the expectations we have for this vaccine based on immunogenicity. So if we go to the next slide. We don't have a lot of safety data here today. I just would say we have a favorable preliminary safety profile. And you can see, the numbers there, a generally well-tolerated vaccine, we know from our previous experience in Phase I and Phase II. And they seem to be, the serious -- severe, serious and medically attended adverse events are at low levels, and they're relatively balanced between the vaccine and placebo groups. Okay, with that, I think we'll step into the South Africa trial, this Phase IIb study. This trial was a randomized, observer (sic) [ observer-blinded ], placebo-controlled trial also evaluating the efficacy, immunogenicity and safety. And we're going to report on the efficacy today at the top line. It had 4,400 adults, 18 to 65 years of age. We did have a special population that we recruited here in the HIV-positive. So there were 245 subjects that were recruited. Again, same regimen, same dose as we saw in the U.K., 5 micrograms with 50 of Matrix, given twice at day 0 and 21, and this is a 1:1 randomization. The efficacy analysis prespecified was between 23 and 50 events, and this represents a final analysis of this data as well. So just again to highlight the very dramatic evolution of the virus during this trial. Again, you can see the same thing which is a time line on the x axis. And you can see the different strains that were recovered. By the way, this is from Nextstrain, which is a gene save site, and you can see the reference to it below. So it's quite interesting to look at that. But by the time we began our trial, began surveillance in this trial, you can see the dramatic evolution of the virus to the -- what we call the V2 strain, which is a triple-mutant strain. And during our surveillance period, it's almost entirely composed of this -- the disease, almost entirely composed of this. And we know from on the ground, this is a very severe outbreak of the second wave of the virus. So if you go to the next slide, you can see there the breakdown in terms of the protection. We had 15 cases in the vaccine, 29 in the placebo. Without the HIV-negative population, we had an efficacy of 60.1% with a lower bound of 19.9%. When we included the HIV subjects, we didn't see any protection. It was -- the efficacy was 49.4% with a lower bound of 6.1%. And by the way, this was the primary end point, and in this trial, the success criteria was to exceed the lower bound of comps in all 6 -- sorry, of 0. So what you can see below then also is that by our PCR data, it appears that over 90% of the cases were attributable to the V2 escape variant, and that was, of course, aligned with what we saw in the epidemiology. Okay. So another very interesting feature here in the -- what we saw is that prior COVID-19 infection with the original strain may not provide protection against the South Africa escape variant. And what to -- maybe not to our surprise, we found about 1/3 of the study participants had prior COVID-19 infections. Now we did this by serology, where we could detect spike antibodies. And interestingly, we saw amongst the placebo rates -- a very similar rate, overall rate of infection in the seropositives and the seronegatives regardless of the spike status. And so I think that's a very interesting feature. Obviously, and this, of course, is the first vaccine to show protection despite the fact that natural immunity may not be playing much of a role in terms of protection, and yet we're seeing this really, I think, quite robust protection against this escape variant. All right. So we are considering what the next phase might look like. As you know, we're very -- we have a very agile technology. The difference between last year and this year is we have tremendous infrastructure in place for moving from a gene sequence to having a vaccine, and we've considered this. So to address the evolving pandemic, we've looked for the optimal vaccine may need to contain multiple strains, something like flu. And so we already have variant strains under development, including the South African strain. We have lab-scale manufacturing that's underway, and we believe we'll be able to rapidly scale up the production of additional recombinant protein candidates, and we expect to start clinical testing on some of these candidates starting in Q2. Okay. And then we also -- as you know, we are engaged in a U.S. Phase III trial as well in U.S. and Mexico, and this is called the PREVENT-19 trial. This is a -- as we've noted, another randomized, observer-blinded, placebo-controlled trial. It evaluates the efficacy, immunogenicity and safety. This is a 2:1 randomization, again, the same dose of 5 micrograms of our vaccine with 50 micrograms of Matrix-M adjuvant, 2 injections day 0 and 21; and obviously, 20,000 in the vaccine group and 2,000 in placebo. Again, the primary end point has been harmonized with these other trials. We have a very similar end point where it's PCR-positive, symptomatic COVID illness and diagnosed 7 days after the second dose. Here, the interim analysis will occur with 72 events as this began as an event-driven trial, and we'll have a final analysis at 144. As you may know, if you've been following us, that this enrollment is going very well. We expect to complete the enrollment in the first half of February. And we are focused -- very focused on the U.S. at-risk populations, and you can see the progress there. We think this trial is going very well. We have amended the protocol to incorporate a blinded crossover. So at a point where the vaccine can be deployed, this would allow us to offer to placebo recipients the vaccine and make sure everyone in the trial is covered with our vaccine. So with that, I think I'll turn you back over to Stan.

Thanks, Greg. Nice job. And it's nice to have good data, isn't it? It makes it easier to discuss. So we're all excited. So you're going to ask me what's next. What's next is what's the pathway to licensing. And so we have a very aggressive, active regulatory program. We're talking with -- I think the plan is that the most likely regulatory agency for first licensure will be through the MHRA in the U.K., and we have been talking with them about what's really required in a pandemic under an EUA to get licensure, both on the manufacturing data and on the clinical and safety data. And so we're going into the details of that. But we're also talking to, I think, even today, we've had conversations with MHRA. We've had conversations with FDA, conversations with EMA, the European agency, Canada and Australia. And so we're all over this right now. So we've started what's a rolling submission in the U.K. right now, and we'll duplicate that in these other countries and regulatory agencies. So we're on our way. And we can't -- we have to finish the trial. The U.K. trial was an interim analysis, and we have to get to the final analysis, which will be a few weeks. And then following that, we'll probably get a few weeks beyond that before we can put together a package to file with the agency. So I don't have a date for you, but we're looking at 2 to 3 months. We're going to try to squeeze every day and week out of that schedule. And as we get more clarity on that, we'll keep the marketplace posted. Everybody's interested in getting this vaccine into the arms of people, and that's our sole goal right now. And so with that, let me just summarize what you've heard today. You've heard that we have the first clinical vaccine data on the U.K. and South Africa variant strains and both very robust. We've got preliminary results from 2 independent efficacy studies demonstrating statistically significant efficacy. Cross protection has been demonstrated with -- I think we're the only company to have demonstrated cross protection in the clinic at this point. And we have -- the variant vaccines are already under development against emerging COVID-19 strains, and we hope to have clinical testing expected in the second quarter of 2021. And so with that, I think we can turn it over to Q&A. And before I do that, I want to thank -- we've had a year-long effort here, and it's been 24/7 literally. There have been no weekends and no vacations. We thank the trial participants, which are crucial. We thank the clinical research staff; the Vaccine Taskforce in the U.K. has been just really instrumental in getting things set up so we could run this trial in the U.S. -- in the U.K.; Shabir Maddi and the Wits University who actually helped us run our pivotal Phase III trials down in South Africa for RSV, did the same here; other partners in South Africa; the Bill & Melinda Gates Foundation, who helped fund the South Africa trial; CEPI; and the U.S. government, who have all been involved with us; and in particular, Paul Heath in the U.K., who has been our principal investigator there. Everybody has done a great job, and we look forward to reporting on our progress toward licensure of this vaccine. So with that, we can turn it over to Q&A.

[Operator Instructions] Our first question comes from Eric Joseph with JPMorgan.

I hope you can hear me. Can you hear me okay?

Yes.

Congrats on the data here. Yes, great presentation. Congrats on the [indiscernible] Just a couple of questions on the South Africa study. Of course, we've been hearing a lot around the invasiveness of the V2 strain. And really, the key question here is, is there any reason to think that the sort of diminished activity efficacy that you're seeing relative to the U.K. results would be specific to 2373 and not similar for the other spike-based vaccines. Have you at all been able to study this in the lab? Do you see any differences in neutralizing activity between the V2 and the original strain? And then I'll pause for follow-ups.

Yes. Thanks, Eric. How are you? So we're ecstatic. This really validates the platform and so it is interesting. There's a very consistent diminution of the vaccine efficacy as we went from the prototype to the drift in the U.K. to the variant in South Africa. We see this as very similar to flu. We -- it's an RNA virus. We expected it to drift, that the experiment got thrust on us. So we are the ones that have data in the context of this drift and change in the virus. And so we think that as we presented our data to many people, they're actually quite encouraged that we're seeing a 60% efficacy against such a big variant. So you're asking a good question. When you look out at the papers that's being published, there's really a big falloff in utilization with the Moderna and Pfizer vaccines. And so the question would be how much of that fallout will manifest itself as a follow-up to next season. We don't know. We have data that is in surveillance, which we think is very strong. So what we do know is there is some evidence that there's about a twofold decrease of our [ immunes ] to the strain in the U.K., and so that kind of lines up with a small decrement. So we'll see. I think there's going to be a lot of attention being paid to this topic. You're asking a good question. We do plan on doing [ immunes ]. We think our efficacy is really quite good. And the -- of course, we would like to collect more data on the more severe cases. It seems likely that the vaccine will protect very well severe as it is. We are also -- as you can see, we're taking into account there may be a need for a strain change, and we're going to take advice on that. Unlike last year, we have a very strong infrastructure for manufacturing. And as we noted, we've already begun -- in case that becomes the recommended course of action, we'll be ready with another spike protein vaccine, reflecting the spike from the V2 variant.

Just as a follow-up to the South Africa study. It's interesting to see here about 1/3 of subjects had seen prior COVID-19 infection. Is there any thought as to whether that experience might have had an impact on vaccine efficacy? And is there any meaningful implications when it comes the -- when it comes to thinking about potential booster regimens?

Yes. So you picked up on it. It seems that the round1, wave 1 illness experience is not providing immunity to the wave 2. And so that's -- and we saw that in our trial, where the difference between the attack rate and the seropositives and seronegatives were -- was indistinguishable. So that's a really important finding. So we don't -- we know approximately in the -- so the primary end point was seronegative population. And the seropositive population is an exploratory end point. We know that the vaccine efficacy is approximately the same and unaffected, but we don't have a precise point estimate today to that. So that's -- as I said, we're at the top line data. And so we don't think it's having any effect on the vaccine efficacy. As you know, what's really happening is these viruses are drifting away from the -- from what is in the prototype immunity. We can overcome that with an adjuvant to some degree. And we think that that's kind of the explanation for why you're seeing the protection we do see against the variants. Which we, again, consistent with what we saw in flu, where the adjuvant end particle allowed us to have an immune response that was really relatively unaffected by a pretty significant drift in the context of H3N2.

Our next question comes from Kelechi Chikere with Jefferies.

Congratulations on the data here. This is a really rich data set. I guess my first question is just related to, can you provide any color as to the vaccine's efficacy in the elderly. And I guess my second question is just can you opine as to why the efficacy may or may not be lower in HIV-positive patients. And does that portend to anything about the vaccine's efficacy in people with compromised immune system?

Yes. So that's a good question so -- too. So we don't -- this is just top line, so we don't have the breakdown. We will by age and gender and so on. We just don't have that today. With respect to HIV, I mean, we are having our own debate. We're not seeing any effect. It's a really small number. So it's 4 and 2. And so what does that mean? We're not sure. We think that one reason -- one of the hypothesis being put out there for this escape mutant is that there could be a pool of viruses that have infected HIV-positive subjects. And because they have relatively ineffectual immunity, it allows the escape mutant to mature, and that may be one reason why it was so prominent in South Africa. We kind of expect that the same escape mutant may arise in many places because the structure of the spike protein seems to move that way. So we don't know. We don't have enough data. We tried to collect -- our goal here really was to get some safety and immunogenicity in this population. And so we just don't have -- we have too few numbers of cases to say really much around whether or not it's going to protect or not. And that will be more data coming out soon. We will have the immunogenicity. We'll be able to look at neutralization. And I think all of that together will start to build a picture. But we're sure they need to be immunized. Not sure based on a few subjects we have in this trial, that we can say too much about how well the vaccine works there.

Our next question comes from Mayank Mamtani with B. Riley.

Congrats team, just an incredible feat here. It really goes back to multiple decades of your work here. And great to see the entire ecosystem come together and collaborate with you on this outcome. So I have 3 buckets of questions, if I may. Can you clarify how much follow-up you may have had on average, on median, for the U.K. trial? And then the breakdown of moderate versus severe, if you're not able to provide that, when will you be able to have that data?

Okay. This is Fil Dubovsky. I can start on that. So the U.K. study is still ongoing. This was just an interim analysis. But pretty much at this stage, we have a 1-month safety follow-up on all the vaccinees in the U.K., so all 15,000, and we're also at a 2-month median time point as well. We're going to continue to gather safety information until the final analysis is triggered, and that will be triggered some time after we have 100 cases identified, probably in the next several weeks, as Stan said. As far as the severity of the cases, we had very few severe patients in both studies. We had one severe case in the U.K. and one in South Africa, and they were both in the placebo group.

And mild versus moderate?

Yes, we did -- if you look at the slide -- what is it, 9, we were able to break down the U.K. data by the mild, moderate and severe. And you can see there, we had 1 in 15 for mild, 5 in 40 for moderate and 1 severe. We don't have that data yet for South Africa.

And so maybe the next question I have on the South Africa strain. So is your understanding that higher the titers, kind of the better it is? And if you are able to comment anything on the -- I know CMI data takes a little while, but do you have any understanding of what the T cell response incrementally adds to this? And last part of it. The David Ho paper that came out a couple of days ago on the E484K mutation, what might be some of your plans to kind of interrogate that in terms of what neutralizing antibodies you're creating?

That's a really good question. And we have multiple collaborations going on, both at sort of the macro level where we're looking at neutralization using multiple assays but also at the molecular characterization level. So I expect in the future, we'll want to map this out. We've done some mapping in terms of context of the prototype virus. And we know we make both receptor-binding domain and turbo domain antibodies. We make monoclonal antibodies. The E484, that appears to disrupt one of the major epitopes that's neutralizing on the receptor binding domain. We know we make antibodies to the other major epitope. So it's going to be very interesting, and we'll map it out. I think one thing we have seen from our other technologies, our NanoFlu, especially is that we have epitope spreading, the adjuvant really matters. And kind of to your point, it's not just the quantity but the quality of the antibody. Both affinity and the epitope spreading is a feature of the adjuvanted, 2-dose vaccine. And that, to some degree, we think has overcome its very significant epitopic change and drift. And again, I just think that given that, this is really very good data in South Africa against such a big variation. So we will dissect that at the macro and molecular level.

Any comment on the T cell response?

We don't have that yet.

So our data comes from the earlier studies where the human data recapitulate what we saw in both rodents and large nonhuman primates. And that's -- we generated Th1 via CD4 T help cell responses. This is polyfunctional, probably effector memory cells.

And finally, has this data been shared with any member of the coronavirus task force yet? And what might, if any, updates we may get? We've been hearing a lot of supply agreement kind of updates on the x U.S. side. Are there any we could expect from the U.S. side of things?

Yes. As you can imagine, we have many collaborators and funders. And so we've been making rounds to make sure they're aware of the data. And as far as steps forward with that, I think, additional contracts, I mean, I don't think we have any more data than what we've already announced.

We have. Mayank, you know that we've announced several already. So including the U.S. as part of the original $1.6 billion award, New Zealand, Australia, U.K., Canada. And we would expect more to come, of course. And again, as I've said, we are seeing far more demand and supply especially now and probably through '22. So expect more updates to come from a purchase agreement perspective.

Our next question comes from Charles Duncan with Cantor Fitzgerald.

I wanted to offer my congratulations on the data not only for protective effect for COVID but also validating the platform more broadly. So nice work there. I had a question on Slide 11. I know that this is top line data. But in terms of the safety profile within the U.K. study, I guess I'm wondering if you could provide any additional color on the severe treatment-emergent AEs that you saw. I know that on a percentage basis, it's roughly the same between the 2 groups, but numerically, it's a little bit larger in the COVID -- or on the vaccine side. So any thoughts on that? Any color?

We just got very top line tables right now. So we don't really have any details about what's in there.

Okay. What's your impression in terms of the reactogenicity relative to what you've heard with some of the RNA-based vaccines? So how do you feel people are dealing with this? And did you see any difference with the older patient population, 65 and older?

Yes. So we think we have quite a favorable reactogenicity profile, and we published that on the Phase I data. And that's really recapitulated in the Phase II study we did. We presented data previously that shows that in older adults, it's less reactogenic. And that wasn't really a surprise. There's senescence in older adults, and that's really what leads to inflammation and some of the local reactions. We don't think we have a problem with react, though. We haven't had dropouts because of it. And additionally, we don't -- we haven't really seen a fever signal. When people get this vaccine, they get some pain and tenderness, and they get some headaches, fatigue and myalgia. And that's kind of it, all lasting less than 2 days.

It sounds like all to be -- all expected in line with what would -- what you'd hope. And then I guess the next question that I had to ask or that I wanted to ask is relative to the EUA and the pathway to licensure. I know -- I think that Stan mentioned 2 to 3 months or so in terms of filing. And it's dependent on cases in the U.K. But could that also enable a filing in the U.S.? Or is the strategy now to rely on the PREVENT-19 study for the U.S. and other geographies?

Yes. I think you have it exactly right. I mean this will be our first efficacy Phase III data. We think we have a relatively strong clinical package when we combine the U.K. data with the South African data as well as the Phase I and Phase II data. And certainly, we're in dialogue with the FDA to understand if that's going to be adequate to get us into an EUA with the understanding that we have a very large Phase III data -- study that's going on. And that data would be available at time of our BLA. Although, we're still awaiting final decisions from the agency if that's an acceptable approach.

Okay. And then with regard to the PREVENT-19 study, would you anticipate the case rate to be modulated, given the rollout of the vaccines here in the states? Although, I know it hasn't been all that robust yet. But what do you think about the expected case rate there?

I think there's a lot of disease everywhere in the U.S. When we first started, we were trying to estimate and model places that had high transmission. Well, guess what? Every place has high transmission. So I don't think we'll be lacking cases. And as far as I can tell, isn't clear to me that vaccination at the low levels we've had so far has really blunted the spread of the virus.

Okay. And last question, then I'll hop back in the queue. I appreciate you taking all my questions. I'm really intrigued with the booster and the -- or the bivalent approach. Can you provide us any additional color on timing there? I think Stan mentioned Q2 studies to start, but really a little bit more color on the product profile there for the booster?

Yes. So I think either Greg or Stan mentioned that we've been working on these variants now for a while. And the speed of our technology is quite favorable. So we are already at last scale, and we're moving toward GMP manufacturing, though we didn't find places we can do that. So I think that the guidance that Stan gave is Q2 start for clinical is about right. So if you remember back to our Phase I study, we had 2 dosage levels, 5 and 25 micrograms. And they were pretty comparable from a safety profile. So we think we have a technology where it's very easy to do bi or trivalent vaccines. And this is very consistent with what we do with flu, where we had a quadrivalent vaccine. And those were very high antigen doses, indeed. So I think the concept still isn't clear exactly how much of which we want to put in together. Right now, we're going to do a -- or planning a dual approach. One is do a co-formulation; and the other one is doing a sequential boost with one and a priming prior to that.

Our next question comes from Vernon Bernardino with H.C. Wainwright.

Congrats also for me on the fantastic results. I know it's been a lot of work and a lot of years. And not only is the program -- or speed that you're involved in, but the effort has also been at fast speed. And so congrats on this high-quality study. Greg, you had mentioned that -- or I think, Stan, you had mentioned that variant vaccines are already under development. What does that work entail, for example? And what do you expect would -- for example, do you need to do primary studies there, too? Or is it just a matter like when you do seasonal flu vaccines, it's just formulating it, conducting a safety study? And then if you do have success with that, what would actually be the strategic position of that vaccine versus the current one that's being used in PREVENT-19?

That's a lot of questions. And thanks. First of all, thank you for your congrats. I know you've been with us for a long time, and we appreciate that. I'll say a few things, then I'll pass the baton. There is a plug-and-play technology. We've got the gene. We made it, and put it into the back of our expression system. And as Filip mentioned, we're at the lab scale. It's a very quick pathway to get into GMP, we believe. So a lot of things now. In the last year, we've calculated that a lot of things could be done in parallel. This is similar here. Very quickly, we would get into nonhuman primates and mice and animals. It is a question of how much data we need. Because now this is representing a strain change. And I think Peter Marks with the FDA, gave some guidance on that and really looked at -- suggested that maybe a small immunogenicity trial might be sufficient for licensure. So we've had no discussions with them. We don't know for sure. And we're also talking to -- this is really new information, literally days, it's days old. But you can imagine, we're talking to many people about how one might solve the problem of antigenic drift. And that's what we have incorporated into our trials, the number of arms, as Filip mentioned, bivalence, monovalence and maybe some different permutations on priming and boosting so we would have the data to help make it a very informed decision about what the next steps are. So I don't know, Filip, if you want to add anything to that?

I guess there's just one point is that even those -- even though these escape mutants have drifted from the prototype enough to really invade the human immune system, they're very biochemically similar, right? They're 90%, 99% identical. So we expect our manufacturing and purification technology to be the same basically.

And can you compare as far as the efficacy that you're seeing with NanoFlu against drifted strains versus your prototypes used for PREVENT-19 and the South African variant vaccine that you might test?

Well, so just reminding you with the flu, that was like the program I just described to go to the next COVID strain, which is called accelerated approval, which allows you to capitalize on the concept that antibodies are likely driving protection. That's what we did with flu. So we did a Phase I, II and III trial where we showed safety and immunogenicity in the Phase III. It was non-inferior immunogenicity to a licensed product. And in that context, we have the same kind of experiment that we experience here where there was -- and during the trial conduct, there was significant change in the virus. And what we mentioned was something called [ AGI ], which is kind of an analog of neutralization. And we showed that we -- despite the fact there was drift, we actually extensively studied historic drift, the matched strain and the drift forward, the forward drift that we experienced and showed no decrement in the [ AGI ] titer. It's a little hard to compare things here because there's an awful lot of background immunity in the context of flu. But I think it's a good point. I think what we're seeing there, if we had a vaccine that worked this well in flu, we'd be very happy, even though there's this drift. And I think the other thing to keep in mind is here, we've got a case definition that's relatively mild, and so that could be another feature. So it's a little hard to compare, except to say, I think it validates the proof-of-concept that the Matrix-M adjuvant nanoparticle allow you to get this cross protection. And of course, the matched strain really validated the technology with a 96% level of protection. That really is -- I think those 2 features really validate the strength of the technology.

Our next question comes from Umer Raffat with Evercore.

We still can't hear you, if you're talking.

I'll go to our next question from Seamus Fernandez with Guggenheim.

Great. And congratulations on the data and I guess the fortuitous locations where you guys conducted your trials to get all this great data. A couple of just questions on the correlative protection. I just wanted to get a little bit of a better sense of when we might see some published data in terms of potential correlative protection. I know you guys just went into fairly extensive detail on how you could do a smaller immunogenicity trial for a second variant. But just wanted to get a little bit more color there. And then I have a couple of other questions.

Yes. So to get correlative protection, you first need to have protection, and we've achieved that in these studies. And we have a lot of partners in both the U.K. and South Africa to evaluate these under a number of different assays. In the U.K. as we're partnering with the U.K. task force, they've set up their own labs to evaluate correlative protection across all the platforms that they have. So they're testing in the U.K. And in South Africa, in collaboration with the Bill & Melinda Gates Foundation, we're doing a similar approach. Importantly, we've carried our own assays through all of our studies so we can compare between studies and get a consolidated view of the immune responses. Now to answer your question, and when we're going to start getting immune responses out. And from the South Africa study, it will be in the near future, next handful of weeks. We'll have some answers, and it may allow us to start to doing some early correlation to see if we have a signal there, yes or no. And in the first instance, they're going to be just IgG before the [ immunes ] are available. But if you remember back to our earlier data, we have a very, very tight correlation between IgG and immune responses. It's a [ percent ] of like 0.95. So it's basically the same thing, at least in the first 2 studies we did.

Yes. That matters because the assay for the spike IgG does require live virus. Does require BSL-3. And so -- and it's very robust that we can managed range. And so we -- if we had a preference, we'd love to see the IgG as a correlate for those practical reasons.

Perfect. And just in terms of the sort of initial supply versus how you're hoping to, let's say, exit 2021 from a supply perspective. Just hoping that you guys could maybe give us a directional sense of how that is likely to come on? And then the second follow-up to that is when you feel the bivalence could actually be coming into official production.

Well, great question. We've been spending the last year going from nothing to where we are now, which is where we have 8 large commercial-scale manufacturing facilities in 7 countries. Six of those are up and running and now making what we refer to as GMP material for stockpile. And we've got -- and the goal is to have production at the rate between all 8 of our places, included in our partnership with Serum Institute at a rate of approximately 2 billion doses per year, 150 million doses a month. And that's where we're heading. So we should have lots of product, it's -- yes, by now. And the answer to your second question is we'll be starting production in small scale of a bivalent vaccine within weeks, and it's the same process. And to the extent that we have requirements for it, we'll be able to make it.

Great. And then just one last question. As we've, I guess, monitor -- as epidemiologists monitor the evolution of different strains, as you look at other potential vaccines, do you think that there is a case to be made, again, relative to other vaccines that they don't have compelling at least pseudovirus data that perhaps -- and without correlative protection? Because, obviously, you guys have that data and perhaps your [ nabs ] can help educate the market on other vaccines. Do you think that there's a case to be made that given VE below sort of the low -- at least low end of the confidence, I know, for now, but you guys are at a 50% point estimate, 60% non-HIV, you clearly make the bar. But do you see that as a potential issue for some of the other vaccines in market?

Well, I mean, time will tell. It's not going to take very long for a couple of other vaccines to -- that we'll have some data to compare to with trials that have been run in South Africa. So I don't think we have to wait very -- I don't think we have to speculate a lot about that. We'll know the answer within weeks. And so let's see what we get. I hope all my colleagues in the industry succeed in their efforts to develop a good vaccine. So that's what we're hoping for.

Our next question comes from Geoffrey Porges with SVB Leerink.

Really appreciate it, and what a milestone. Well done. Just a quick follow-up on the manufacturing. Other than SSI, are you collaborating with any other vaccine manufacturers to accelerate your scale-up for any of the components of the manufacturing? And could you talk a little bit about what is rate-limiting to manufacturing right now amongst the key steps in the process? And then could you give us some information? Do you have any observations about the severity of the cases -- the breakthrough cases in previously exposed individuals compared to the seronegative individuals in the South Africa study?

Yes. This is Stan. I'll take the first, which is what challenges do we face in manufacturing. Well, we -- the first challenge that we faced 6 months ago was trying to start up manufacture of a biologic at commercial scale in 8 different facilities at the same time. Never been done before. And I think we're doing a good job. In addition to that, we had to make adjuvant, that we had to set up production in 3 different countries: in Denmark, in Sweden and the U.S. And so we've hired a team. We've gone from 100 to 150 people this time last year to well over 700 people right now globally, and most of that is manufacturing -- quality of manufacturing, and they've just done an outstanding job. And so we had -- as you can imagine, it took some time to get from 50 liter scale or 5 liter scale to 50 to 2,000 to 4,000, now we're at 6,000 liter scale. And I think this is January. Around 6 or 8 weeks ago, we identified a process and walked it, which you have to do and shared that with all the manufacturing plants on a daily basis. And now we've got a locked process of that kind of scale. So that part is decided. Now it's just a matter of optimizing, getting higher yields. But the hard part right now is coordinated on a global scale all the supplies that go into production. You've got bags, you've got filters, you've got media. And that's what our team -- now with a huge and very successful team is in our global supply chain. But we've identified a couple of different raw material components that are in short supply globally, and we're trying to get people to increase capacity as quickly as possible. Risks are that some of them don't get there as fast as we need, and we have some delays. But we're trying to mitigate those. That's what we spend a lot of time doing. So I'm confident that the team will get there. So -- and with that, Greg, you had a clinical question. I forgot what it was now.

I believe the question was about severity of breakthroughs in the seropositive versus seronegative baseline. And that's a -- it's a critical question. And we don't have that information yet. It's obviously going to be important to understand that to see if some of the baseline titer does modulate severity of the disease. Our population in that study is very young, and there were very few severe cases. So this may not be the most obvious place to answer that question.

Our next question comes from Jonathan Lim with UBS.

This is Jonathan Lim on for [ Naveen ]. A few questions, if we may. The first one would be, what are your thoughts on whether durability of response may differ against the wild-type per se versus the South African strain. And then our second question would be what learnings there may be from this data to the Brazilian P1 variant seen in Manaus in Amazonas so in Brazil. Then finally, we also want to ask whether or not you've collected routine nasal swabs to kind of interrogate the potential impact on transmission. When may we see that data?

Maybe I'll start at the end there. We have some nasal swab data, but probably from this study isn't going to be the best place to capture that. In the U.S.-based study, we are pursuing quantitative PCR of all the positive nasal examples, so we can get a view on it. I'm not sure if that's going to be informative in the end, but we're still going to give it a shot to see if we can find a signal there. As far as the Brazilian strain, I mean, when I look at what's happening in Brazil versus South Africa, I see a lot of similarities. And in a sense, it all makes sense, right? Some of the mutations make the virus more fit and some of them are meant to escape the developing immune response in people. And I think those are some good similarities between South Africa and Brazil. It will be an interesting experiment to see if those responses can cross influence.

All right. And what about the first question with regards to the potential durability of response against wild-type versus South Africa strain?

Yes. That's not known. And I think it depends if you are a believer that antibody is the one and only component. We showed some data where we generate high levels of IgG, which is quite durable, out to 6 months, so if -- which is superior or equal to that of convalescent serum, the 6 months the IgG. So it's possible that it's more the quality of immune response versus the quantity, which may be important in correlative protection.

Very good. Congratulations on the data.

And I'm not showing any further questions at this time. I would now like to turn the call back over to Silvia Taylor for any further remarks.

Okay. Well, I won't be Silvia Taylor. But I want to thank everybody. This is probably the biggest event of our corporate life right now in scientific and medical, and we're very proud of all -- everything our team has done. And I want to say that to all of my colleagues at Novavax, you've done a fantastic job. It's been great working with you. I'm sorry it's still virtual these days, but we'll have a big party one day. Thanks, everybody, and we'll see you.

Thank you. Ladies and gentlemen, this concludes today's conference call. Thank you for participating. You may now disconnect.