SGS SA (SGSN) Earnings Call Transcript & Summary

Hello, and welcome to this SGS Life Sciences webinar on COVID-19. We'll be going over some of the regulatory considerations surrounding human challenge studies with the SARS-CoV-2 virus. And myself, the Scientific Director; Bruno Speder, Head of Regs; and Robin Rogiers, EPI and one of our clinicians, will be going over some of the interesting developments in this and some of the efforts being taken to tackle a pandemic of a virus currently circulating in our communities. Okay. In our agenda today, we'll be covering -- myself, giving an introduction of the virus and some of its characteristics. I'll also be talking about human challenge modeling and what that might look like. Then we'll have some comments from Bruno Speder on developing a challenge agent and what a human challenge unit looks like and how it operates, before we finish off with some of the clinical aspects and a round up. An overview. This novel coronavirus, we've now called it SARS-CoV-2, comes from WHO. It's known originally as 2019-nCoV, probably emerged from China over the close of 2019. We've done some mutation analysis looking back at the spike protein. Those regions are quite mutagenic, and we can see the mutation rate of about 1 single nucleotide mutation SNP per month. It may well have arisen in some region around the province and Wuhan, maybe towards the close of 2019, certainly appeared to be circulating in animals at that time. It may have transferred into humans. Before that, and there's certainly some evidence that maybe even as early as September or October in 2019, the virus had jumped into humans, whether it was the Betacoronavirus we know now, or whether it was something similar, is still under debate. But it certainly looks like the coronaviridae made their way into humans in a new form from a zoonosis around that time and producing these set of symptoms we now know as COVID-19, which is a largely upper respiratory tract disease but can also descend and cause bilateral pneumonias. We can see its spread to 210 countries or regions. I do know that there are already 195 countries recognized currently by the WHO. So 210 obviously includes some regionalities. We've seen a lot more than 2 million -- 2 to 3 million cases now as of the end of April, something like 200,000 deaths. And its spread has certainly outpaced other major infectious outbreaks in the past. And if you look at Ebola, the number of deaths were relatively small. I think they're somewhere in the region of 16,000, something like that. But certainly, the spread of that was not anywhere near as effective as this has been, probably due to the nature of its respiratory tract spread and droplet infections. It was declared an epidemic on March 11 by the WHO, and then as a -- pandemic, sorry. Declared as a pandemic on March 11, and so far has been an epidemic. We're used to seeing seasonal influenza epidemic, so seeing a pandemic was a shock, and we haven't seen that since 2009 with the H1N1 strain we saw. And I know, looking around the world, clinical infrastructures are not really prepared for such sudden increases in morbidity and mortality. And owing to insufficient ITU beds, specialist staff, ventilators high-dependency units, CT scan machinery, et cetera, it makes it very difficult. And including PPE, of course, makes it very difficult to cope with some exponential growth of cases coming into care pathways. It looks like the first vaccine will take 12 to 18 months to develop. We may have some early mRNA breakthroughs. The history of mRNA has not been built with glittering prospects because of the amount of product that's produced, but we are hopeful that something new may come through there. More likely, some of the subunit or attenuated vaccines or something similar will come through later on with the higher rates of efficacy. But certainly, to get this off the ground, even moving with all due speed and help from the regulatory agencies, it would be like 12 to 24 months, most likely. SARS-CoV-2 will probably end up as a globally circulating viral strain with a seasonality which is similar to that observed with other upper respiratory tract infections like RSV and influenza, with predominance in those colder tropical -- colder temperate regions, rather. And you'll see waves moving up and down the world according to the seasonality. Although antivirals and vaccines are being developed by governments through PPI collaborations, I think repurposing of existing drugs and vaccine platforms will hopefully form the first breakthrough in the first wave. So moving on to our next slide. SARS-CoV-2 virus itself, very similar to the original SARS 1 virus, share the same attachment process by the spike protein to ACE2. You'll see with others, they use a slightly different attachment process. Probably came from bats as a primary reservoir. Bats huddle together in caves in a nice moist environment, cool environment. Looks very much like these cool temperate environments where these tend -- viruses tend to circulate. So they all make very good hosts and they carry a lot of different viruses, maybe translate into humans through pangolins, Manis javanica. We've seen civet cats with SARS. We've seen camels with MERS. So we know there's often an intermediate host, a mammalian host. And those will be illegally imported, although it may have been translated directly into humans a number of times from interaction with bats. COVID-19 disease resulting from this looks like flu. It's a dry pyogenic type of virus, which gives you a fever. It gives you, after 5 days, a fever followed by dry cough. That may go on for some time. There are some other weird symptoms which we're noticing, like loss of sense of smell. 7 to 8 days, this can progress to a shortness of breath. And about 20% of patients who are symptomatic, they require hospital treatment at this point. And especially in the elderly, those with comorbidities like heart, lung disease, it can become a pneumonia, at which point, chest pains, shortness of breath appear. And these people do require supportive therapies such as oxygen. Following the infection and recovery, immunity for people who are symptomatic appear to be quite good, high titers, although we're not sure how effective those antibodies are in protecting you from further infections. Those who seem to have very few symptoms may not produce protective response. So these responses tend to be 6 to 12 months in mucosal assaults, mucosal challenges with upper respiratory tract pathogens particularly, so it may be that we will see reinfection at some point in the future. Okay. Quickly, let's look at the Coronaviridae. The family Coronaviridae, especially the subfamily Betacoronavirus, which we're interested now, with the genera Sarbecovirus. It's a typical classification for organisms: family, subfamily, genera, species. And species, in this case, what we're looking at is SARS coronavirus 2, SARS-CoV-2, but also includes the SARS-CoV, which came earlier around about 2003. Coronavirus is a zoonotic common cold and GI tract viruses with common characteristics, also human common cold viruses, GI tract, especially in avians. And we'll look later and see how those attachment processes reflect that. Single-strand RNA genome, 26- to about 32-kilobyte positive-sense single-strand RNA. They've got an enveloped structure. And we know that enveloped structure can be pleomorphic, can be spherical, a number of shapes, with -- characterized by these club-shaped projections, which is like a halo of glycoproteins on its surface. Hence, the corona or halo, which are the corona or surface of the sun. And the genome is one of the largest of the RNA viruses. We'll have a quick -- very quick look at that a bit later on. But is also subject to frequent recombination, 33% in some areas, which can create new strains, alterations and virulence. And again, that depends on the reading frame you're looking at. There are 7 strains of human coronaviruses: 229E, NL63, OC43. 229E and OC43, probably the most common we see. We've also seen MERS-CoV, SARS-CoV, and this new novel coronavirus, or SARS-CoV-2. These are one of several coronaviruses. They were recognized by the WHO in 2016 as a likely cause of future epidemics, potential pandemics in 2016, following the interventions around the Ebola epidemic. And I think maybe that was a shot across our bows at that point. We had some presages of things to come. And maybe we can learn from that in the future when we see such mutations arise, and maybe we can move on and potentially move more quickly towards viral and drug development. We'll move on to Slide 6 now. Okay, let's look at SARS-CoV-2 virus itself. Look at some of the structural elements and their functions of this B-lineage betacoronaviridae. There are A, B, C lineages on this. So if we look at the elements that actually make this virus up, you'll see the major elements probably are spike protein. You've probably heard that discussed quite a lot. This is the major ACE2 host receptor. It has 2 units, the S1 and S2 subunits. The S1 subunit is probably involved in the host range and tropism of the receptor binding site. And so it may have a larger or smaller host range, depending on mutations on that and adaptations. S2 subunit, involved with the viral and host cell fusion through these heptad repeats, 1 and 2 heptad repeats. It's just a 7-repeat basis. Unusually, you'll see in the A-lineage betacoronaviridae, you will see this HA-glycoprotein hemagglutinin fetal esterase, which you may have seen before in influenzas. It's probably been acquired from influenza C at some time in the past through some sort of transference. This is quite interesting because OC43, one of the other coronaviridae, uses Neu5Ac as an attachment, also hemagglutinin, a sialic acid attachment, very similar. MERS-CoV itself uses CD26. SARS 1 and SARS 2 use ACE2. So you can see from the different receptors, through the spike protein, you can see they're actually attaching to different receptors on the host, and that may alter the way that the disease develops, the pathogens, the virulence, and also some of the symptomologies, such as diarrhea you see in SARS-CoV-2 because of the ACE2 receptors you see along the GI tract. We have a membrane, M protein, this M glycoprotein, which enables a transmembrane transport and budding, envelope formation and stabilization, and also stabilizes the internal nucleocapsid protein. We have an envelope glycoprotein, the E protein, which is a virulence factor, has roles in viral morphogenesis and also is important during assembly and egress of the virus from -- during budding, where it takes part of the plasma membrane from the host cell. We have a nucleocapsid N protein, formation and maintenance of the RNA -- RNP complex, rather. We'll come back to the ribonucleoprotein complex. We'll come back to that a bit later on replication and regulatory replication and transcription of the internal RNA. It also inhibits the host protein translation. It helps hijack the ribosomes, and it also inhibits some of the cell proliferation which goes on. So it actually acts as -- again, helpful during control of immunoresponses. And finally, you have a lot of nonstructural proteins from the ORF1a, 1b section, the open reading frame, things like C-3-like (sic) [ 3C-like ] protein, papain, the helicases and NTPases. Again, very important for the actual replication of the viral RNA itself. So let's move on to the next slide, Slide 7. Okay. We'll now quickly discuss SARS-CoV-2 replication, how it hijacks a cell. There's 2 ways it hijacks it. First, it enters either by endosomal entry and cathepsin L activation of S protein facilitates that. Or preferentially uses plasma membrane fusion, and the serine protease initiates an ACE2 fusion. So you're seeing that ACE2, the angiotensin-converting enzyme-2 plus S protein fusing. And that is preferential. Once you have viral entry, polyproteins are translated by the region's open reading frames, ORF1 and 2, and these are nonstructural proteins. And these form then helicases and RNA transcriptions. It breaks open the RNA and starts the transcription process through the RdRP complex. So this replicates the -- viral helicases, replicates the RNA, genomic RNA and the nucleocapsid, then starts production of S1, S2, E, M proteins, envelope, matrix proteins, et cetera, via the endoplasmic reticulum. The N protein, the nucleoprotein, is actually made in the cytoplasm. And these are then all assembled with the M protein, so you get the RNA, you get the nonstructural and the structural proteins assembled from genomic RNA products, and they fuse with the virion precursor on the endoplasmic reticulum, move to the Golgi apparatus and then released as a cell surface vesicles. So you've got 2 products going on there. Genomic RNA either gives you these nonstructural proteins, which are involved in RNA synthesis; or we get structural proteins, which are involved in viral assembly. Clearly, if you have these 2 methods of entry, the endosomal, and you also have the plasma membrane fusion, you have 2 different ways of sort of avoiding -- blocking antibody. And also hydroxychloroquine, which is being touted as one of the treatments for this, may work within these endosomes at high doses, but they may be toxic doses. But also, let's not forget that's not the preferential method of entry, therefore, there are methods of escape. Okay. Moving on to Slide 8. We'll talk about some targets for SARS-CoV-2 drugs and vaccines. Just very quickly, just going back over the genome, very simple thing here. We have open reading frames 1a, 1b, you've got an S, you've got an M, you've got an N, you've got ORF6, all up to ORF10. You tend to see a lot of mutational changes on ORF1a, 1b, 3 and 8. ORF10 may or may not exist. We haven't yet seen it produce anything particularly structural at the moment. Anyway, so all these areas will translate into something, which gives us a chance both to interrupt them or to actually bind to them, to stop them being effective. And we've seen, for the drug, so we've got papain-like proteases. We talked about those protein agents. We talked about those earlier. The 3C-like proteases. So these things are responsible for the helicases and the RNA-dependent -- the RdRP, RNA-dependent RNA polymerases. So we can interrupt those, and we have various ways of doing that. RdRP can be inhibited by virustatics, and we've seen favipiravir and penciclovir being accounted as particularly working in there. Entry inhibitors, griffithsin, combined to spike glycoprotein, hopefully preventing entry. Also, we can see that some of these pH-dependent cysteine proteases can be blocked by these lysosomotropic agents such ammonium chloride. So there are various ways we can actually block either entry or attachment, and we have a number of targets there, either nonstructural, structural, including the host-specific receptors which we can attack. In terms of drugs and vaccines, as we've seen before, the difference between the different receptors between the coronaviridae is quite profound as to whether they're using something like the CD molecules to attach. And we've seen CD26 in that. We've also seen Neu5Ac. But in this case, if you're looking at the SARS, we're looking very specifically at ACE2. And if we're looking at the ACE2, the angiotensin-converting enzyme-2, it depends on what particular type of intervention you're looking at there, whether you're looking at an RNA, an mRNA vaccine; whether we're looking to make a DNA vaccine; whether we need to make subunits; whether we make something attenuated; and whether we're going to make carriers. And each one of those, if you look at the graphic on the right, you'll see going from the more simple, which are mRNA, which you can easily design and assemble, from back translation from the actual structural side itself, those are fairly easy to make and they're fairly easy to express in a carrier when you come through to attenuated organisms, live attenuated, which will probably give you better T cell responses, which may offer more direct interventions on disease. And then those are more difficult and probably take a longer time to develop. Drugs. And again, you can see if you look at the logo or the graphic on the right-hand side, certain ones of those have been around for a long time. And if you look at nelfinavir, lopinavir, some of the PIs on the HIV which are being repurposed, they've been around for a while. Whether they'll have an effect on here in terms of protease inhibition is an unknown. I suspect the differences will be too large to see any large translatable effects. And some of the new ones, remdesivir, which has been around since original SARS-CoV and Ebola outbreaks, again, is being tried. It does have a target, we have more hope for that one coming through. The chloroquine, again, has been around for a long time. It's probably less likely to have an effect because of these 2 methods of entry. And if you look on the -- there, you'll also see how far these have gone through, whether approved, not approved, or with -- they're still in early phase. Most of each, you'll see they are in quite early phase. We're still doing some very early work to see how effective these will be in reducing replication and, hopefully, transmission. For vaccine development. Okay. If we look at vaccines development, and currently, there are some 160 vaccines development. I suspect, while I've been speaking, 2 more will come online. We can see ones that already gone into Phase I vaccines, which you would expect to be quite rapidly rolled out, like these non-replicating viral vectors, which will be carrying genomic material representatives of SARS-CoV-2. We're also seeing a peptide-based and an RNA and a DNA virus. Again, these are relatively quick, and platforms exist to roll these out very quickly. The efficacy of these may not be sufficient in terms of the amount of product to give us a fully protective effect. So if we look back in the exploratory confirmed and unconfirmed, you'll see quite a few viruses there, which are probably more what we'd expect to be seeing coming through in time, which are peptide-based, inactivated and live attenuated virus vaccines. And those will hopefully give you a better T cell response and, therefore, we should see effects on disease as well as on neutralizing and blocking antibody. So I think, at the moment, we have some promising candidates coming through. But I think behind that, we have a very large tranche of, potentially, quite effective viruses. This is a virus that does not include itself into the genomic material of the host. It reproduces in the cytosol. Therefore, we're seeing probably a lot of mucosal shedding. It's not associated with viremia. It's not associated with rate of persistence. So we may be able to see something which is eventually protective, but may have a short window of protection. Some COVID facts and figures. Now we'll move on to the meat of the matter, which is, obviously, human challenge. Infectivity rates, SARS-CoV expressed as R0, it's probably around about 2.8. SARS itself was about 2.5. Any number higher than 1 means it's spreading rapidly. You can see a range of illness, but look at the asymptomatics there. We are seeing a lot of asymptomatics. And maybe that was represented early in the year. We've not seen it. People are saying 14% to 18% of the U.K. population may have had infection. This is in late April from probably beginning of the year, maybe end of 2019. That is not enough for herd immunity. We may well see reactivation. 80% of infections are mild. We've all seen people who've had things with coughs, colds, fevers. We haven't done yet the antibody test there, but it would be useful to see how many of those have had it. Maybe it is greater than 18%, which would be good for herd immunity. But worryingly, up to 14% of those are symptomatic to severe, 4.7% become critical. And we may see anywhere between 1% and 2% of reported cases as fatal. New antibodies are suggesting that the case fatality rate is around about 0.18, 0.19 from U.S. studies. U.S. doesn't appear to have same case fatality rate, and we're not sure whether that's because there are new strains out there. And there appear to be some quite rapid mutations going on within the hosts to show different virulence in some of the very recent data coming out. So maybe it's only twice as bad as flu or maybe it's 10x as bad as flu, we just don't know yet. But we do know the risk of death increases the older you are and doubles roughly every 10 years after 60. The good thing about this compared to, say, influenza RSV, is very few cases are seen in children, and the majority of those are mild. So we don't know the death rates yet. We do know they vary from country to country. It's probably influenced by both median age; societal effects, such as the age of communities and how they're put together, in homes or they live with families; population's average age; policies of testing regimens; but maybe also, we're seeing some effects of viral mutation. This is a single-strand RNA virus after all, and we're expecting it to mutate. Moving on to -- okay. Let's talk about some of the risk factors for severe COVID disease. We know that people turning up with raised D-dimer, raised IL-6, serum ferritin. These people are more likely to progress. Inversely, those with lower lymphocyte counts do poorly, and people on corticosteroids and some of these comorbidities, even obesity there, diabetes, chronic kidney, heart, lung, I mean, usual things you'll see, especially COPD, are associated to progression. LDH is another one. But I think D-dimer is very interesting because maybe that's associated with some of this microvascular coagulation we're seeing at some of the lung problems, from descending infections, actually causing occlusion, not just death and blockage of the alveoli with this sloughed material and mucus, disruption of the basal layers, but maybe also seeing the microvascular coagulation also causing necrotic areas from poor oxygen perfusion. So I think we have a number of markers there which are quite clearly defined. And maybe we can use those when we move forward into our human challenge agents to screen for some of these to make sure we aren't seeing people with D-dimer issues. We aren't seeing people with problem with lymphocytes, any pancytopenias or cytopenias at all. And maybe we can also look at LDH and make sure LDH are within normal levels. And we do have -- we are working on synopsis on this to make sure that all these risk factors are taken into consideration. But I think we do have some idea from the populations we're seeing and from some of the soluble and cellular markers as to what the risk factors are. They're not always predictive, but I think they are very useful if you're going to streamline your patients into such a study. So moving on to the next slide. Just about relative humidity. I've talked before about that and how they live in these nice humid and cool environments. And you can see a coronavirus prevalence. You can see that warm temperate and cold temperate climates are perfect for this. You don't want too high temperatures, a lot of UV, it's bad for viruses. And you don't want high temperatures because it also dries them out. And also humidity. You want a high relative humidity, which you tend to see during the temperate seasons, a little bit colder seasons as the air can't hold as much water. So you have these nice, relatively dull -- I mean, the U.K. is nice for this. Poor weather and lots of humidity and not too much UV. And you'll see a lot of transmission going on. And I think we have a nice graphic that shows the virus spread by latitude and season. It moves up and down, up and down according to the months, according to the season. And you'll see some obvious peaks and troughs in countries as that occurs. Epidemiology. So rapid antigen testing. What are we looking at in terms of testing? We've only got 2 ways of identifying this -- well, maybe 3 ways of identifying this. Rather look for the antibodies from the host themselves, which is the response, or we look at rapid antigen tests to directly detect it. I think if we're going to put people into human channel studies, we have to have some way of detecting those people who have an antibody. There are a lot of tests out there. These -- the antibody tests tend to be more sensitive and specific. If you look at those, the antigen tests show lower specificity and also quite poor sensitivity. So I would suspect we're going to be using these nucleic acid tests, qRT-PCR, that will be the gold standard for this. We'll be looking for people screening and potentially for actually having the virus in themselves both with antigen NAT and antibody tests. And for taking people into studies, we may be looking at people with either no evidence of antibody, limited amount of antibody, or we don't know really what the infectivity this is. Maybe we'll have some people in there who also got relatively high levels of antibody. We'll now discuss human challenge modeling and some of the principles of performing CHIM modeling within the context of COVID-19. Looking at human challenge studies per se. Human challenge charts or controlled human infection modeling, CHIM, we noticed a number of things that may appear to be in conflict with usual pronouncements for any medical or clinical intervention, which is, first, do no harm, primum non nocere. So do we conflict established principles here where we are actually causing disease in people? But do those rights persist when the right of the individual may be trumped by the right of the many? Maybe some suffering for the individual can be justified by the right of the many? But can we protect both individual and society in such a circumstance? So some people do get ill. Do the benefits of actually learning things, protecting more people, give some kind of equipoise in our rights there? But also, who can champion the subject's right, considering that SARS-Cov-2, especially the BSL-3 organism, that means it needs to be grown under a biological safety level of 3, much like tuberculosis or cultured HIV. It's considered highly infectious organism with potentially serious adverse events. So we've got a few things we need to consider on the following slides. Slide 19. We can see that going from January in 1796, which is a deliberate inoculation in 2vaccinia, going through to, allow me to picture the common cold research unit in 1958, post war, using some facilities left over from the complex. Took 150 years to go from what appeared to be a very simple concept of using a similar virus, homologous virus to protect, to actually doing challenges with homologous viruses to test, in this case, not just vaccines, but mostly drugs. Then we move on. Another 50 years, things are accelerating, we're doing proper human challenge studies, and we're doing those in very promising facilities such as in the east end, such as in the U.S. at the NIH. And then 5 years later, we're in 2020, we've moved on for some epidemics and pandemic strains mostly associated with the influenza into coronavirus. So we have an evolution of a challenged model, it appears to be accelerating. Can we keep up both with regulations and with our challenge agents? Pretty mainstream now, I think, challenge studies because of animal testing as practical and ethical issues. If you want to use a SARS-Cov-2, you've got ferrets or you've got these humanized transgenic mice from the Jackson Laboratory. But they can only make so many mice. So we actually have supply issues. Not just PPE we're running out of; we're running out of some of our preclinical modeling. Some of that has poor predictability. Sometimes you can't predict -- sorry, infect animals with the correct strains. You have to use similar strains, and we're looking at in RSVs, often the case. Or say with asthma and arthritis, we don't really have good animal models. Also, when we look at things like T resident cells, T effector memory cells, B cells, are these responses the same? Do we see the same tissues as well? Are these gold bulk tissues a ring of Waldeyer, et cetera? Are they similar in these animals? Do we get a similar progression of immuno exposure? I don't think we always do. Field studies are also quite prolonged, have slow recruitment, they can be complex. You get very diverse populations which may have morbidities and different mortalities associated with those. Also, some of these may be seasonal, and that will slow you down if you're trying to produce s vaccine for something which only appears for a few months a year. HCTs, these human challenge tests or controlled human infection models tends to be a bit more crisp because you can infect people directly. You can look after them all the way through. They're regulated control. And we also know we've characterized the challenge agents. So not being exposed to a relatively unknown agent in the environment with all sorts of other co-infections going on. So we tend to use it with various clients for discovery, for approvals; and also looking at novel emergent strains, such as SARS strains and SARS-Cov-2, in particular; and also looking forward into some of the parasitic models such as malaria. But obviously, wherever you have something which is acute and not chronic, with heavy sequelae, it gives you a lot of options to be able to apply it. Okay. So before I hand it over to Bruno Speder, Head of regulatory, who's going to talk on some of the regulatory aspects of human challenge, let's go back and see the relevance of what we've discussed to human challenge studies of the virulence and pathogenicity of SARS-Cov-2. Certainly, we are seeing a high virulence, a highly transmissible virus with a high pathogenicity. We have unknown AEs, unknown SAEs; some nuance, such as loss in taste and smell, maybe effects on the olfactory bulb. Do we use an attenuated strain, such as the 8aa furin deletion? Do we use our molecular strain? Which one do we do because we -- which one do we use because OC43 and 229E, pages different attachment pathways, and we can't use the other SARS, the original SARS virus, because it has a very high morbidity and mortality rate. Do we use a wild-type strain? And if we have a wild-type strain, what some of the ethical issues around that. And is it sufficient that we select populations that have low-risk factors and screen for markers of severity? Which age groups also do we target? Finally, I think we should think about what titer we're going to use. Do we start with very low titers? Do we start with titers that are almost impossible to titrate down to? Or do we start with more regular titers we've used in challenge studies such as 10^3, 10^4, 10^5? Will that be -- we see dose-related symptomologies on those and dose-related severities. What's the root of inoculation? Is it going to be pernasal? If it is pernasal, because I think oral may give a high-risk of adverse events going by influenza, how are we going to give that? Is that going to be given by drops? Or is it going to be given by an atomizer? Drops may be swallowed. You can see that in ferret studies. If you give drops, they swallow a lot of that. We have a2 receptors in the gut. Are we going to see diarrhea as an adverse event? And also shedding, how long will shedding be? So what will the length of stay be? These are questions I think we'll have to answer. And a lot of that will have to be taken to the regulators in [ SGAs ]. And hopefully, they can give us some guidance. So please over to Bruno.

Thank you, Adrian, for this -- for the introduction. In the next couple of slides, we will take a look at the MC and nonclinical development of human challenge agents. When we take a high-level look at the development pathway of a challenge agent, we actually see that it's relatively similar to that of a classic drug or vaccine. When you have a drug or vaccine, you will first have a GMP manufacturing, then a nonclinical testing according to GLP or good laboratory standard practices with an in vitro and in vivo development pathway. And then you will go to human clinical trials in your clinical development path. A human challenge agent is considered a medicinal product according to European Medicines Agency guidelines, and also the Food and Drug Administration considers the human challenge agent as a medicinal product. Therefore, you will have a development pathway that, high level, is broadly similar than that of a drug or a vaccine. You will have a GMP manufacturing, you will have a nonclinical testing phase, and then you will have a human clinical development path that will consist of a Phase I titration trial. One very important difference with a classic drug or vaccine is that for human challenge agents, we, as of today, will not have a formal licensure procedure. So therefore, there will be no marketing authorization application of a human challenge agent in Europe or an NDA or biological license application in the United States with the FDA. And in the next 4 slides, we'll go a little bit more into detail in how a GMP manufacturing pathway would look like and how the nonclinical testing and clinical testing would look like. If you look at the high-level overview of COVID-19 challenge agent development strategy, there are a few points that need to be taken into account. The first one is that there are actually no existing guidelines on the development of challenge agents. The main reason for this is that actually, the development strategy with challenge agents is very specific on the type of agent you use. If you would use a parasite, a bacteria or a virus, all of these agents have specificities and need to be developed in a different way. And so one of the main reasons there are no overarching guidelines on challenge agents is that it's mainly a case-by-case development approach. I've already discussed before, one of the major attention points in developing a SARS-CoV-2 challenge agent is about safety level of the organism. SARS-CoV-2 is considered a biosafety level 3 organism, which will create some logistic and tactical problems in the manufacturing of such agents. Manufacturing should be done in a biosafety level 3 facility, but also the human challenge study later on would need to be done in a biosafety level 3 challenge unit, and we will discuss that a little bit further in the presentation. So one of the major discussion points that would be raised with regulators, both with EMA, national regulators in Europe and the FDA, is if a attenuated challenge agent, if one would choose to go that route, it could be considered a biosafety level 2 organism. And that shows that actually the GMP and MC development strategy is very linked to the scientific strategy as the choice of challenge agents might influence the practical execution of such a development strategy. Human challenge agents are considered medicinal products, both in Europe and in United States. In the United States, the FDA requires an IND, investigational new drug application, to be filed for a challenge agent. The challenge agent would need to comply with good manufacturing practice -- manufacturing. It should be a nonclinical development program. And it should be first in human titration study before it can be used in commercial studies with therapeutics and vaccine. The manufacturing approach. Before we can start manufacturing of a challenge agent, we, of course, need a strain to form your seedstock to start up. And then where would this strain come from? There are actually 2 possibilities. The first possibility is that you would use an existing isolate, an existing COVID-19 isolate. This whole [ backward ] strategy is that probably it's not very well documented where the isolate comes from and that you would probably not know the medical history of the patients from whom the isolate has been selected, which from a GMP, GCP point of view, is not the best approach. An alternative is that you start a community surveillance program. So you set up a small protocol in a GCP study, approved by an ethics committee or R&D, in which you isolate COVID-19 strains from the general population. You isolate the strain. You do a broad range of testings to establish which viruses and bacteria are in the samples. And based on the selection of samples, you choose the strain that you feel appropriate. And depending, as Adrian discussed in the first part of the presentation, would you go for a wild-type, would you go for a more attenuated strain, things that need to be discussed with regulators and based on development strategy. The manufacturing approach. As we discussed in the previous slide, there are no overarching guidelines for the manufacturing of challenge agents. But usually, vaccine manufacturing guidelines are used as the mate of the foundation of GMP manufacturing of a challenge agent. Then you would look at the combination of FDA and EMA guidelines to ensure that you have a robust manufacturing program that complies with regulatory requirements on both sides of the Atlantic, so both for the European Medicines Agency, a national regulator in all of Europe, and with the FDA. The challenge agent would be manufactured as a solution for intranasal use. The spending on medical and scientific considerations, but also manufacturing considerations, one could choose from droplets over intranasal sprays. This is also something that will need to be excessively discussed with regulators to see what would be -- would give the best -- the best results, but also to be the most acceptable for the safety of the volunteers participating in such studies. A second important step in the manufacturing strategy of our human challenge agent is adventitious agent testing program. Also here, you will base yourself on existing vaccine guidelines. But here, a more risk-based approach could be considered. Based on our experience of with our H3N2 influenza challenge agent and our ROC challenge agent, regulators are open to take a more risk-based approach in its adventitious agent testing program, specifically because you will only manufacture a limited amount of this challenge agent. So it's not a continuous production like we would have with the vaccine, but more of a single batch approach. So a more risk-based approach for an adventitious agent testing program can be considered. It will, of course, need to be considered and discussed case-by-case with regulators. But also here, in the case of the COVID-19 challenge agents, one could essentially base itself on the guidance issued by both FDA and EMA on some risk-based approaches for development of therapeutics and vaccines for COVID-19, and sort of the principles could also be applicable to human challenge agents. Once you manufacture your agent and once you do the complete testing program, of course, for the European Union, they will need to be reviewed by a qualified person. And of course, the manufacturing site where you manufacture this would have to have foreseen in its manufacturing license. And all the safety information will be discussed in the investigational medicinal product dossier for a safety application in the EU or in the CMC section of your IND for an IND submission to the FDA. As already been discussed and mentioned a few times in this presentation, it's really of critical importance that you discuss this risk-based approach upfront with regulators, EMA or national regulators in Europe, and the FDA during a pre-IND meeting. Once your agent is under manufacturing, you will, of course, also start considering the nonclinical development of the COVID-19 agents. And the nonclinical development will exist in 2 parts, the in vitro characterization and the in vivo characterization. The in vitro characterization will be a cell-based assay, in which you show the in vitro infectivity and characterization of the COVID-19 challenge strain to show that it's effective in vitro, but also that the COVID-19 strain can affect the COVID-19 strain. Once you've done your in vivo characterization, you will go to the in vivo characterization study. And this will be done in an animal model, and the study will be on -- according to good laboratory practice. And then you have to select the right animal model. Based on current literature and current state of knowledge, the ferrets will be the preferred animal model, but humanized mouse model is also possible. But the second model is the preferred model. There is also some evidence that the strain would be infective in cats. But for the moment, there is no GLP animal cat model available for the testing of COVID-19. So you would go for the ferret model. Such a study would be a positive and negatively controlled study. And the aim of both the in vitro and in vivo study is to show the efficacy and infectivity of such a strain before you would go to the first-in-human titration study. Based on the data of the animal studies, you would be able to establish the attack rate, but you would also be able to select the starting dose for your titration study, as you would have with any other medicinal products. So the clinical -- the nonclinical development will also form the base of the design of the first in-human titration study as to give you several important elements. Another important element, and we will discuss it somewhat later in the presentation, is also the shedding rate of the COVID-19 challenge strain. There are also a number of points where, compared to the traditional development of a drug, the nonclinical development of a human challenge agent can be more condensed. For example, pharmacokinetics and product metabolism studies are usually not required for human challenge agents. Also, no formal toxicology studies are expected for this type of agent. There could, however, be one exception that is if the virus is overly virulent or pathogenic. In that case, regulators might require this type of studies. And this is something that will highly depend on the choice of your challenge agents, as has been illustrated by Adrian in the first part of this presentation. And in case you would choose an overly virulent or pathogenic strain, this type of studies might be required by regulators, and we'd strongly advise you to discuss that in a scientific advise meeting. Reprotox studies are usually not required for this type of products at the condition that the volunteers participating in this type of studies take sufficient contraceptive measure and usually double at the conception as described in the CTFG-3 anticonception guidelines. All the results of the preclinical studies will need to be described in the investigation brochure and in the nonclinical overview that will be part of the IND and CTA application. Once you've described all the results, CMC results in the IMPD and clinical and nonclinical results in the IP, you will, of course, need to have some regulatory maintenance of the dossier. Also taking into account that as there is no licensure procedure for challenge agents, so there will be no formal marketing authorization application or NDA BLA application, there is, of course, the regulatory requirement to maintain the documents. And so the IMPD will need to be kept up-to-date with ongoing stability results. So your challenge agent will undergo a GMP stability results program. There, you can choose if you would follow a classic good 3 months, 6 months, 12 months, first 6 months, et cetera approach, or if you would just test the challenge agent before [ each target ] in the challenge study and reports the results back in the IMPD. That's something that you will need to -- will depend on your stability strategy. The investigator brochure, of course, will be subject to a yearly update and will be updated after each study with the challenge agent. And of course, as for each medicinal product, you will keep development safety update report and do the adequate safety reporting that were required. Now that we've very briefly discussed the manufacturing and nonclinical development of a human challenge agent for SARS-CoV-2, we would like to spend a little time in explaining what the regulatory requirements for a human challenge unit executing these studies would be. So as discussed, there are some regulatory requirements to which a human challenge with performing this -- SARS-CoV-2 challenge studies should adhere to. Broadly speaking, these biosafety level requirements depend on the biosafety level of the organism being used. As discussed, SARS-Cov-2 is currently considered a biosafety level 3 organism. This can be a discussion if an attenuated challenge agent could be considered biosafety level 2 or not. There are some examples of challenge agents that in the framework of a human challenge study have been downgraded. Example is plasmodium that have been used in Malaria challenge studies. It's considered a biosafety level 3 organism. But in the framework of human challenge studies with the highest containment measure, it's considered a biosafety level 2. So this is a precedent that can be used to discuss this with regulators. But as I said, regulatory requirements will highly depend on this. The biosafety requirements are duly discussed with the biosafety agencies of the country in which the human challenge study is based. In Europe, these biosafety agencies can be part of the human drug regulatory agency, but can also be a separate agency. So assuming one would consider a downgraded biosafety level 2, then you would need to have a biosafety level 2 permit. You will have to have biosafety requirements that are compliant to biosafety level 2: biosafety level 2 compliant beds; airlock/HEPA filtered negative-pressure systems; a dedicated lab; of course, all the standard operating procedures that will have to be established in consultation with the Biosafety Committee. But of course, this is something that will need to be discussed with regulators and are highly dependent of the classification of the COVID challenge agents. Two other important regulatory aspects to take into consideration for these human challenge studies are the prescreening of volunteers and the shedding of virus in the community. The volunteers entering such a human challenge study should be prescreened with an accredited and validated IgG and IgM serology system. And this where we need to require an update of the prescreening protocols secured with the ethics committee. The shedding of virus in the community is something that must be avoided at all times, in particular, in a [indiscernible], where it's even more important than with some of the more mainstream H3N2 in all of the challenge models that are currently being used. And the length of stay in the units, which currently is 11 days for H3N2 and other disease studies, will need to be confirmed based on the basis of the ferret study where shedding of the virus will also be measured. Based on this data, a proposal can be discussed with regulators and will depend on the view of the Biosafety Committee and the regulators also. So it's possible that for SARS-CoV-2 studies, a somewhat longer confinement, for example, 14 days might be needed. And once there is a sufficient real life data coming from human challenge studies, this might be updated. And subjects will be discharged based on the negative rapid antigen test. This concludes this high-level overview on the regulatory requirements on manufacturing and on clinical developments. A few takeaway messages. The manufacturing needs to be done according to GMP. The adventitious agent testing program can be risk-based, but needs to be discussed with regulators. The nonclinical development will consist of 2 parts: in vitro and in vivo characterization. In in vivo characterization, the choice of animal model is important. You will ensure that there is an applicable regulatory maintenance of the files. And then depending on the biosafety level classification of the agents, 3 or 2, a number of items in your manufacturing and in the conduct of the studies might be changed or adapted. But of course, this will all need to be discussed with regulators on both sides of the Atlantic, FDA and EMA and national regulators in Europe. And now I pass on the work to my colleague, Robin Rogiers, who will give some more details on the clinical development of such a challenge agents.

I'll walk you through the clinical aspects of the virus in consideration for COVID challenge trials. We all know that the novel coronavirus cause a respiratory infection. Typical symptoms includes fever, cough, shortness of breath and myalgia. As the flu season 2019, 2020 was simultaneously ongoing when COVID struck Europe, differentiation between the viruses was initially very hard. As we experienced with COVID disease progresses, more and more clinicians are reporting cases of anosmia, the inability to smell, and associated changes in taste with a reversible nature. Associated with more severe disease seems to be the presence of thromboembolic disease, such as pulmonary embolism. The dissemination of this information between health care workers definitely helps them make fast decisions. But it's still early to get statistically significant relations from large data sets. Due to the limited knowledge on the number of infected subjects, mortality of the infection is still difficult to predict, but estimated to be around 2%. As symptomatology on its own is inefficient to differentiate from other pathogens such as influenza, and case isolation and contact tracing are essential to slow down the spread, definite diagnosis by means of virus detection is required. Several tests are currently deployed but no gold standard is identified so far. Upscale in test capacity is of primordial importance to allow for some less stringent contact rules and restarting some form of social life. In considering diagnostic testing, time is important. In early disease, consider detection of viral loads on respiratory tract samples. The sensitivity of current qualitative PCR is estimated to be a little over 60% and dependent on sampling techniques. Sampling from the lower respiratory tract has the highest sensitivity. No cross reactivity with other viruses have been detected, so the specificity of the test is high. Still, a lot of false negatives might go undetected in early disease. Specificity is also frequently used in hospitals to assist in diagnostics and get some information on severity. This test has a high sensitivity in patients with symptomatic disease for over 48 hours and makes it possible to diagnose patients with the previously false negative PCR. Specificity is lower in the early disease and increases when symptoms progress. Bilateral ground glass opacities located peripheral adjacent to the pleura are the typical radiographic presentations of the disease. But they do not allow to differentiate between other viral and pneumonia, such as in influenza pneumonia or others. Events of the consolidation does seem to correlate with the outcome. As every health care worker is extremely sensitive for COVID, we might become at risk for collateral damage. At an individual patient level, patients with chronic diseases might suffer due to decreased follow-up and other acute diseases are prone to delay diagnostics and increased complication risk. The exact extent is hard to predict and will require large retrospective investigation. Due to regulations such as remote GP consultations and insufficient COVID testing capacity, we've observed less testing capacity for other respiratory diseases as well, such as influenza. As the disease has preventive and therapeutic options, we might suffer from an increased influenza mortality as well. Close contact between different animal species, including humans, increases the chance of pathogens to mutate and overcome host-specific immunological defenses, which makes it possible to spread the disease and causing a pandemic as it did. The virus has been able to affect the world's population due to an incubation period of 5 to 7 days on average, followed by a long symptomatic period with droplets creation. For those with bad outcomes, death sets in on an average of 14 to 19 days, which means the disease has sufficient time to spread before killing its host. Serology testing will become important to gain insights in immunological reactions to the virus, but the correlation between antibody titers and the level of protection is still unclear. Viral particles in stools have also been observed, but only one Chinese research team was able to cultivate live virus from feces, indicating that the virus has a hard time withstanding gastrointestinal passage. It is still very difficult to predict who and how long subjects are to be considered contagious. Since the viral load is highest 24 hours before the start of symptoms and in the first symptomatic days, patients are then considered to be highly contagious. But for how long a patient remains contagious is still an important question. Viral shedding can be observed for a long period of time, but cultivating live, infectious virus appears to decrease significantly after 8 days of symptoms. What about asymptomatic carriers then? [ IPLAN ] published preliminary results of the general public testing they performed. 43% of all subjects who tested positive did not recall ever having any symptoms, and probably this is still an underestimation. They find asymptomatic subjects in all age groups. So even in the elderly, this is a possibility. Due to insufficient testing capacity, it is difficult to predict the contribution of asymptomatic carriers to the disease spread, and estimations differ significantly. After listing these viral and disease characteristics, some considerations need to be raised when discussing a COVID challenge trial. As always, subject safety is essential. Subject safety starts with identifying the right population to enter such a trial. During the trial, we need correct and sensitive markers to follow up on subject's health and adequate treatment options. These includes ICU access and trained staff. Next to subject safety, it is equally important to consider the safety of your clinical trial staff, and by extent, society. Effective subjects will need to be adequately quarantined and clinical site staff must have access to personal protective equipment, making it impossible for the virus to infect outside the quarantine. On the same topic, we need to ask ourselves the question, when we can discharge this deliberately infected subject? What is to happen when subjects are still shedding viral material at estimated discharge? Do they pose a risk to their family members or are they no longer considered to be contagious? When thinking about trial population, who not to include is a very important question. Old age, obesity and smoking are considered to be associated with disease progression and prolonged hospitalization. Also certain comorbidities increase the risk for severe illness due to their impact on the immune system. Underlying disease might make a patient prone to more severe complications. Mortality in cardiovascular disease patients is estimated to be up to 10% due to cardiac complications of oxygen deprivation. Of course, patients with a pulmonary disease are at risk since this is the primary target of the virus. COPD patients already suffer alveolar damage, and gas exchange is further impaired due to the consolidation. Asthmatic patients are at higher risk for an exacerbation provoked by the viral infection. Even after survival of serious bilateral pneumonia, permanent lung damage is expected, causing further morbidity and delayed mortality. Stopping rules for this type of trials need to be carefully selected. As always, we prefer concrete and ambiguous stopping rules with parameters quick to obtain and based on scientific and medical rationale. We should consider including stopping rules based on clinical parameters, such as pulmonary function; auscultation; vital signs, including pulse oximetry and the need for treatment. Laboratory findings and PT can help investigators in making their decisions on severity and a follow-up of the evolution. Much is going on about possible off-label treatment options. But it is still early for strong evidence for their efficacy. The fact that there is currently no rescue medication available is a significant hurdle when trying to conduct a COVID challenge trial. Also, the reports on concomitant medication which might harm subjects are confusing, and no scientific consensus has been reached. The use of nonsteroidal anti-inflammatory drugs is controversial, and although most paper don't have arguments for or against their use, the Health Department of France advises against them because of a possible ACE2 receptor effect. Corticosteroids are known to cause immune suppression, and an important component of the COVID pneumonia is immune-induced lung damage. So there might be a place for them in early stages of disease. After the first few days of symptoms, the immunosuppressive effect flips the scale to a less favorable outcome and prolonged viral presence. The FFP2 or FFP3 or U.S. N95 masks can filter out about 95% of particles when used in a correct way. They're the best we have at the moment for high-risk procedures, which creates some form of high results such as intubation or bronchoscopy. A large meta-analysis could not find an important benefit of these masks over the general surgical masks when worn for low-risk procedures, such as the transporting of a patient or blood draws. More and more countries are trying to find a way to reuse their equipment even though the WHO cannot recommend it. Some believe that wearing a surgical mask for the general public might provide a false feeling of security, which will allow them to let go some other principles like social distancing or hand washing. To conclude, as this is a droplet infection, adequate containment of the challenge agent is required. Also, some rescue medication needs to be available and stronger evidence on the place and the value of different testing modalities is essential. As there are still a lot of unknowns about the disease and progression, one needs to be very careful in following up on adverse events and be very attentive for serious adverse reaction.

I'd like to thank Robin and Bruno for their contributions, and I'd like to thank everyone for attending this webinar. And I'd just like to summarize the discussions and some of the information we've reviewed. In terms of the agent, we got to look at type, which one we're going to use for challenge agent because of different attachment, the processes, the route, titer. We've got to consider serious adverse events. Which markers do we measure, both during and prior to infecting people, and what endpoints do we look at? In terms of clinical considerations, SAEs, again, can we measure markers to reduce the number of SAEs or AEs? Are these markers of severity, progression or are they predisposing markers? Containment. How long can we keep people in containment? And what level of distancing do we need within containment? What is the standard of care? Can we use hyperimmune serum to try and treat people who have a sudden rise in markers? What do we do when people do progress? Ethically, can we do this? I think we can. Should we do this? Personally, I think we should at least consider it. Who will benefit? I think society will certainly benefit more than the individuals. We have no real proof that people benefit individually as responses appear to be either poor or not prolonged. Who should indemnify this? How do we indemnify individuals and companies performing this? And what value do we see versus risks being taken? Regulatory wise, given the above, can they help us? How fast can they help us? Who should be involved in this? What is the pathway to acceptance of both challenge agents and studies? How do we manufacture this? Is it an IMP? A non-IMP? Is it IMP-like? Do we ignore those regulations and put something entirely new in place to facilitate some rapid production whilst trying to ensure safety through testing? What testing is required? What uniformity is required? And do we need to standardize? Let me just finalize by saying, what do we do to prepare for the next pandemic? Do we have a think tank? Do we need better community surveillance? Do we need firms looking at challenge agent preparation? Do we need governments to look at challenge agent preparation? Do we keep seedstocks of potential pandemic or epidemic organisms? What do we do? I think there are a lot of questions coming out of this, probably more than answers I can give today. But thank you very much for listening.

Well, thanks very much for listening, everyone. We have had some interesting questions come through whilst we were discussing COV-19, and challenge agents and some of the issues surrounding those. I think one of them came through from [ Shiban Abashigan ], who was discussing whether we take people in early. So I think that's a very interesting question. We normally do a challenge study. Do we take people early into the study? Because we want to see whether they're going to develop any other diseases, any suprainfections or preinfections before they come on to the unit are very important because they may make a difference to your data, they may affect the symptoms. So I think it's important, if we can take people in early, we take them in. Normally, we'll take them even 2 days early, because the majority of upper respiratory tract infections have about a 48-hour incubation period, especially the common cold, PIV, things like that. But if we a got 5- to 7-day incubation period, are we going to take people in 5 days prior to enrolling them and taking them into containment? I think that may be the case, which may make for quite a long period, which we will need to have more data to see whether we can detect virus. Is it being shared in that sort of presymptomatic period or prepatent period? If so, then maybe we can do some antigenic testing or we can do some nucleic acid testing at that point, qRT-PCR, to pick people of who are going to be liable to then develop symptomatic disease. And look, there's some evidence there that these people are actually shedding virus. So if we can use an antigen test, though I think more likely, with a nucleic acid test, as we currently use, we use a matrix test, then we can test people for a range of diseases that may be developing in that time, including SARS-CoV-2. I think the answer to that question is yes, we'll probably bring them in earlier, maybe 5 days. It can be a bit tiring for the subjects coming in. I think probably more worrying is we don't know when we can release them. I don't think fecal shedding out to 27, 47 days has been measured. What has been measured is of that greater consequence because we know that there's been through TCID50 testing, live virus setting, almost none of that has been detected. I know live virus has been detected in feces. It just seems to be genomic material. We're more interested really in the nasal pharyngeal route, how much is being shed there. And again, anything out to 11, 14 days. So I think we need some animal modeling there. We need the ferret modeling. We need to make a challenge agent, which I think all this time is probably going to be, as another question that was asked, a wild-type or attenuated. I think from the discussions I've been hearing recently, I think the wild-type one will give you a greater prognostic value for field studies. So if we're going to have a level of risk, genetics committees will balance risk versus benefit, are we looking -- a trumping issue here, whereby we may have to have a slightly more pathogenic virus to get better translation into community studies rather than use a furin deletion, for example, which may have -- as you've seen in hamsters, may have less pathogenicity? At the moment, I feel people are pushing towards wild-type, for good reasons, I think. In as much as we know, the risk factors, and I showed that earlier, risk factors or likelihood of people progressing into serious disease below the age of 40 is pretty small, 0.001 or 0.002, something like that. Below the age of 40 is pretty small. So these people have very low -- and that's all age -- sorry, that's all ranges of morbidity and comorbidity at that point. So I think if we can include the right populations, I think the risks are low enough, sort of round about the levels of kidney dialysis or, at the upper level, maybe a kidney operation. So I think we do have some crazy sort of premise, rather. We do have some premise to allow people in and give them wild-type. But how long we keep them in, I think, is not known as yet. We were also asked -- someone asked a question about ethics. What premise do they base their decisions to allow these sort of studies on? These are often made up of late populations. They look very, specifically, risk versus benefits. And I think at the moment, the ethics committees will, most likely, the feeling I get, lean towards if we can provide them correct data; we can give the correct risk ratios or acceptable risk ratios; if we can give them protective care within hospitals, if it's ITU, HDUs, et cetera; and good standard of care following on from that. Anyone who becomes symptomatic to any moderate degree, I think that ethics committees, hopefully, will look at this and balance the risks versus benefits. And the benefits really are, can we progress? There's 160 drugs and vaccines in the background. Can we progress the correct one through fast enough to give real benefit to society without having to go through phase -- preclinical Phase I, II, III in a normal manner, but each one of those to find out which is the best. Can we sort them more quickly? And I think human challenge models allow us to sort some candidates very quickly. And I think if any -- if ever there was a time to do that, I think the time is now to have a good challenge model to go through those vaccines especially and find something which we can work on. It could take up to -- another question came through, how long would it take? I mean, this could take, ooh, I mean, someone said to me 6 months the other day. Someone very high up and knowledgeable. But I would say 8 to 12 months to make a challenge agent that was well characterized. And that is probably not including a characterization study. So somewhere between 12 and 16, 18 months, I think, to get something truly prognostic that we can use to actually select the correct drugs and vaccines. If we all push together and we get access to some big hitters, to some of the big BSL-3 manufacturers, then I think we can reduce that to 12 months. So with all those providers in place, I think the future could look rosy in terms of vaccinology. I think for the drugs, we still got some way to go. We're seeing already some of these fall by the wayside. Hydroxychloroquine appears to have a full profile. IL-6 doesn't appear to be the great hope that we hoped for. Remdesivir still didn't have some great evidence behind that. So let's hope that new generations, rather than just repurposed drugs, can come through and help us with this current, very worrying pandemic. But thank you all for listening. And I hope I'll speak to you all soon, either individually at a conference, or another one of these webinars. Thank you.

Thank you. Bye.