In real time to figure out what you're asking and be able to ask those in the end of the prepared remarks and the questions we have set for this call. I also want to thank Arvind and the IR team at Amgen for setting up this call so quickly after the call. I know we've already passed April 1st, which probably created a bunch of havoc internally with the legal guys, and thank you for working through this and making this happen. With that, I'm going to pass it over to David for some introductory remarks. David?
Thank you, Ronny, and thanks, everyone, for getting together. Good morning. As we all know, we're in the middle of an unprecedented crisis. The biggest public health challenge in a century. Certainly the biggest public health challenge that any of us have seen in our lifetimes. We felt, we being Adaptive and Amgen, that given our technologies and given our existing relationship, that this was an opportunity to see what we could contribute in the battle against COVID-19. We have an ongoing collaboration for minimal residual disease in hematologic malignancies, and it was really based on that scientific interchange and the knowledge that the companies had one another that led to a rapid series of discussions and the launch of this collaboration. The intent here is really to marry the technologies of both organizations.
Adaptive's world-class expertise in immune profiling, of course, we'll go through that in much more detail in the course of the call, with Amgen's expertise in genetics, immunology, and then antibody engineering and manufacturing. We are focused on making progress as quickly as possible. We are trying to strike the right balance between getting the highest quality antibody or antibodies as therapeutic candidates with the shortest timeline that we can. We will take advantage of frequent interactions with regulatory authorities. The FDA has been quite open about offering frequent touch points, and certainly, we will take advantage of that as we move through the development program.
In addition, many of you have questions around scale-up and manufacturing, and of course, we will be working on plans as we move along to ensure an appropriate supply of the therapeutic, should it prove efficacious, while maintaining the supply on the Amgen side of our other protein and antibody therapeutics. We all know, but it bears repeating, that cancer and heart disease don't go on vacation during an epidemic, and of course, we're committed to maintaining supply of other vital medicines as well. We see these efforts as complementary to some of the other efforts that are going on in the industry. We really don't view this as a competition, but as an all-hands-on-deck moment, where everyone is going to bend their efforts and expertise to coming up with an effective therapeutic.
With that, what I'd like to do is invite Harlan to make a few remarks on his side, and then we'll open up the question part of the webinar.
Thanks, David. Good morning, everyone. My name is Harlan Robins. I'm the Co-Founder and Chief Scientific Officer of Adaptive Biotechnologies. Thank you for your time today. We are looking forward to providing more insights into our partnership with Amgen to identify and develop neutralizing antibodies to COVID-19. Adaptive is built on the premise that the adaptive immune system is nature's most finely tuned diagnostic and therapeutic for most diseases. To translate these natural abilities of the adaptive immune system into clinical products, we spent the past decade building a proprietary immune medicine platform that allows us to characterize key immune cells with scale, precision, and speed to transform the way we detect and treat disease. Now, we are well poised to apply our platform to discover both diagnostics and therapeutic solutions to COVID-19 pandemic.
On the diagnostic side, we are extending our partnership with Microsoft to decode the immune system's response to COVID-19 to complement diagnostic efforts that are focused on the virus itself. On the therapeutic side, we are adding a new dimension to a multi-year working relationship with Amgen that David just mentioned, to leverage our platform to identify neutralizing antibody candidates from the blood of patients who are actively fighting or have recently recovered from COVID-19. The key to neutralizing antibodies is they have to be able to work alone or in small groups, meaning that they have to do the job of what normally hundreds of antibodies do in a typical immune response in the blood. This approach has been shown to be efficacious in some settings, such as Ebola, and is promising for SARS-CoV-2, particularly because it is a slowly mutating virus.
However, to find the truly effective antibody, or what we like to say, the Michael Jordan of antibodies, is quite difficult. The search has to include the ability to scout every single high school basketball player in the entire country at the same time. This is what Adaptive's platform is built to do. Together with Amgen, who's a world leader in antibody development, we see COVID-19 as a long-term problem, unfortunately, and there will be patients in need around the world for the foreseeable future. We are committed to applying our resources and expertise to help discover solutions. On a personal note, Amgen is really a pleasure to work with, and their scientists are quite extraordinary. We think together we'll be able to really progress in this effort. Thanks, and I'll turn it back over to Ronny for questions.
Thanks, Harlan. Before we attack the therapeutic approach, in our chat yesterday, as we got prepared for this call, we talked a little bit about the diagnostic approach, which looks really interesting. Can you just take us quickly for what you're trying to do there and the kind of insight that we potentially end up drawing from your ability to analyze the progress, and the variation between individuals' immune system's response to COVID-19?
Yeah. Our diagnostic approach is primarily focused on the T cell response, not the B cell response. It's basically co-opting the same machinery, both technologically as well as including the software and machine learning that we've developed with Microsoft to be able to identify the specific T cell receptors, the part of the adaptive immune system that's focused directly on the disease and identify it on a patient-by-patient level. Basically what we're doing is trying to identify what each patient's immune response will look like and to answer a whole variety of questions. First, we think the T cell response comes up quite quickly, so it should be a very good early detection strategy.
Second the scale of your immune response, whether or not you have a really robust T cell response versus not, has a strong chance of being predictive of who's going to do well and who's going to have serious conditions in response to the virus. It could be a way of determining who needs serious care and who doesn't. Because you have immunological memory, it'll persist, and we'll be able to assess later who actually had the virus and who didn't. The whole getting back to work campaign, et cetera. It'll hopefully be a complement to serology. We don't yet know who's going to seroconvert. There's some evidence coming out that maybe certain patients who do particularly well, maybe even asymptomatic, don't even develop an antibody response.
Serology might not pick those people up, but we think the T cell response might be able to cover at least some of those. That's what we're doing on the T cell side for diagnostics.
The other thing you mentioned is this idea of looking at patients over time and being able to potentially predict how long the ability of the body to counter an immune response on second exposure is. This goes to the question of what's the length of immunity people have once they're exposed to the virus?
Yeah. I think that's a super important question, and obviously, we don't know the answer yet. The length that a person is immune after being infected with a virus can range from lifetime immunity to a few months. What's the answer for COVID-19? We will be assessing this as well. The idea would be that, how long do your antibodies stay in your system? Figuring out exactly what the level is to convey immunity is going to be challenging. Because we're going to be sequencing a huge array of B cell receptors or antibodies from each person as part of our partnership with Amgen, our IRB allows us to monitor these people over time so we can continue to sequence their repertoires and see if they're maintaining the cells producing these antibodies over time or not.
Okay.
It's a challenging question, though.
Switching over, this is almost like a perfect segue to the next question, Harlan, which is, we are already having folks bringing convalescent sera up and beginning to use it. What are the strengths of this approach and what are its limitations?
Yeah. The idea behind the convalescent sera approach is that if you can take the plasma from a patient that's recently recovered from COVID-19, they have a whole set of antibodies that that person's immune systems develop to be, in theory, efficacious against the virus. What happens is that there's large efforts across the world where people are extracting large doses of sera from patients. It seems to be working, and I think it's a great strategy. The negatives are that it's not really scalable. From each person that you get their plasma, you might get enough for one or two doses to different patients. You could imagine that's not a very scalable situation.
There's also obviously a lack of quality control because you could imagine that sometimes the sera is great and other times not so much, and there's no real way to assess that. I think this effort's great and will be useful to people and probably save many people's lives. On the other hand, it's not really a scalable solution that can be administered worldwide.
What is the neutralizing antibody approach at?
Yeah. The neutralizing antibody approach, in this case, instead of looking for this whole collection of antibodies, we're trying to find one or a handful of antibodies that can do the job of what the entire sera could do by itself. In order to do this, we have to, as I've mentioned earlier, truly find this special Michael Jordan of antibodies, where they have the capability of doing it by itself. This allows this to be scalable because Amgen can then develop and produce this therapy at scale, and then we can administer it much more broadly.
Okay. Now talking to both of you. As you screen through the past experience we had with the antibody approach, they seem to be effective in various conditions. We obviously have RSV therapy. There was clearly some success with Ebola. MERS don't ever make it through it. In none of those cases, they were perfect. That is, they didn't just abrogate the disease. Can you talk a little bit about the past experience, what we learned from them, and where are we going to improve as a science world this time around on those antibodies?
Sure. I'll start and then pass to Dave. As you mentioned, there's been some success in the past with Ebola, RSV, even HIV. There's been a couple limitations. One is that these viruses are small RNA viruses that mutate quickly. We would expect that you're trying to hit a moving target, so it's going to be a challenge to find really good neutralizing antibodies against those viruses. In particular, the antibodies that have been found seem like they are able to neutralize, but probably aren't the superstars, aren't the very vast neutralizing antibodies. Mostly because we haven't probably had as much effort as we're putting into COVID-19 across the world to find them. That's both hopeful and I consider that quite hopeful. We're getting some clinical efficacy even without necessarily the optimal neutralizing antibodies.
Just imagine how good we can be with a really good neutralizing antibody, especially against SARS-CoV-2, which mutates at a much, much slower rate than these small RNA viruses, HIV, and influenza, et cetera.
What I might add, Ronny, is that one of the things we clearly learned, and I would say especially from the Ebola experience, is that the time point at which you intervene has a huge effect on outcome. Administering the therapy earlier in the course of the disease gives it a much greater shot of efficacy as opposed to the very, very late stages of the disease. That will inform how we think about the administration of an antibody-based therapeutic going forward. There are a number of clinical settings where one can imagine it from patients who are very seriously ill to those who are earlier in the course of disease. For example, moderately to seriously ill, but not yet requiring ventilator support. Can you prevent them from traversing to that more serious clinical scenario?
Finally, potentially in the prophylaxis setting, can you in fact prevent infection? This might be quite important as the epidemic evolves, and I think we're taking the view, as Harlan alluded, that this may be a two to three-year event, barring a vaccine intervention sometime earlier than that. What that means is that there would be a clear role for prophylaxis. One could imagine, for example, administering prophylaxis to healthcare workers in areas that are a hotspot where an outbreak is developing to really try to protect frontline healthcare workers. We're certainly thinking about a range of clinical spectrums, but the experience that we have from other antibody therapies against viruses is that generally the earlier you go, the more effective.
Got it. Harlan, you mentioned a couple of times this issue that this virus is a more of a slowly mutating virus. Talking to virologists, at least the points we got was not as clear. The question was whether the virus is a slowly mutating virus or simply a virus that did not have a lot of pressure on it from therapies, and because its infectivity is dependent on the upper respiratory tract, even if there is more effect of the body's immune system in the lower alveoli, you are still infectious on the basic virus. The virus that moves from patient to patient is the same, and thus you're not seeing a lot of mutations, but once you start applying drugs against that virus, you will see the virus mutating a lot faster.
In your mind, do we have clarity on this question, or is it still kind of unknown?
I think it's clear that it mutates much slower. You can look within a person. If you look within a person, even within that person, if you sequence the influenza genomes or the HIV genomes, you get this huge diversity, whereas in this virus, for SARS-CoV-2, you get much, much less diversity. There's a good reason for this. The genome of SARS-CoV-2 actually carries a gene that corrects mutations.
The polymerase.
Yeah. We know that it doesn't mutate at the same rate, or it's almost close to 1,000 x less than the other genomes. It's still an RNA virus. It still will mutate. We've seen their different strains, but it's just vastly less than like a HIV or influenza. What will happen when we put true selection pressure on it? I don't think we really know the answer, but looking with individual people and knowing that everybody's immune system is actually putting that pressure on and seeing that there isn't huge variation is a very good sign.
So one of the arguments for-
I would add, Ronny.
Sorry. I would add, Ronny, that one of the things we're doing with deCODE genetics, which of course, is participating in this collaboration as part of Amgen, is actually generating a catalog of viral mutations along with all of the publicly available. There's an extensive molecular epidemiology study that Kári Stefánsson and his team are conducting in Iceland. We hope that the first publication from that should be out very shortly.
Okay.
That will inform some of our thinking as we move along here as well.
Okay. A lot of the arguments here, the question is, should I go with a single monoclonal antibody or multiple ones? The issue of one of the advantages of monoclonal antibodies is that you attack every piece of the virus. Here you'll be attacking a very specific epitope. How do you avoid this idea of the virus mutating away from you upon applying the pressure?
Well, I can start. I think there's two issues that are relevant. One is since we do have, as David was saying, from groups like deCODE, we have some information about what the different strains look like. Being able to show that you can find an antibody that's neutralizing against the different known strains of the virus, if it can neutralize a variety of strains, that's a good sign that whatever the virus is capable of mutating probably isn't away from that neutralizing antibody. There's no reason why we have to limit to one. We vetted this with the Amgen team yesterday, basically, is what's the capacity to handle a cocktail of two or three where we would avoid this problem even if there was escape. I can pass over to David.
Yeah
Amgen's perspective.
Yeah. That's certainly been part of our thinking from the beginning here, Ronny, that while we would love to have a single antibody, it's possible that a cocktail of a handful or a little less will be required. So as we move along we making plans to be able to handle that.The other question is, while of course, most antibodies everyone anticipates will be directed against the spike protein, we are, of course, going to look at membrane envelope, and the selection pressures may vary among those different targets. These are things also that the deCODE work, and the work that will go forward as part of this collaboration will inform.
Got it.
How well are you able to come up with antibodies that are likely to be fairly resistant to mutation by the virus? How well can you, say, identify sites which look like they're pretty well-conserved between coronavirus mutants? How well can you identify sites where it looks like if the virus were to mutate a way that would have significant implications for its viral fitness?
We need to let viral genetics tell us that answer. This is just comparative genomics. We just line up all the viral sequences, and ask where. The assumption is the immune system inherently puts pressure on certain areas, and the areas that don't mutate, either one of two reasons. Most likely, those are the regions that can't mutate, they're needed for function of the virus. Those are the areas that we would focus on. We also don't want to go into this with the assumption that we're that good at this. We'd like to let the immune system tell us what the neutralizing antibodies are.
We plan to do a quite wide search without making too many assumptions externally about what the immune system latches onto to neutralize the antibody, and hopefully screen very widely and let the immune system tell us what truly is neutralizing.
Yeah. What I would add, Vincent, is that clearly, a mutation that compromises receptor binding is going to be one that the virus can't really tolerate. One can make pretty simple and basic predictions like that. As Harlan indicated, we're really taking the approach here, we're going to let the rapidly emerging data that we're generating teach us as we go along. It's to sort of be our own adaptive immune response in a sense.
Coming back just to this issue of the problem that you have, I guess, Harlan, you already mentioned that the differences between the antibodies is actually going to be critical in our success. If you have to look at the kind of affinities or selection you guys are able to do versus previous efforts by other companies before your technology came around, how much better is it? Are we talking here about something that will be 30% better or something where you can see the differences in terms of the antibody effectiveness being one in 100 kind of thing? 100 times better than all the historical efforts in this space. I'm sure we have some experience from the cancer world and so forth.
We expect to be able to screen a set of antibodies that's probably an order of magnitude or maybe two orders of magnitude larger pool. How far that gets us in terms of.
Better.
We assume that, let's just say all the ability to neutralize comes on some kind of bell curve. The more you can sample, the more likely you are to get farther out and get the super, the really special one. Whether that really special one is twice as good or 20% better or 5x better, I don't think we know the answer to that yet. Certainly we're hoping that being able to search vastly more will get us to that needle in the hay-- like the really special one, we'll be able to find it more likely.
David.
I would also add that even modest improvements may translate to a substantial effect in the clinic. You may not need a log better antibody to actually have something 20% better may actually pull through to a fairly substantive effect in the clinic.
Following up with that, so this is the kind of like generating the leads. How mature are the assays that will be able to tell you in vitro how likely the antigen is going to work in vivo? It feels like before this epidemic, we did not really invest that much in trying to correlate the efficacy of or the value of certain assays to the in vivo setup. Can you just talk a little bit about can we actually pick the right clone or are we just going to have a lot of clones and the selection process is going to throw everybody down?
A lot of the technology has been built up over the last decade, even though I don't think it ended up getting applied in the same way as it will in this case because the areas where we thought past pandemics were going to materialize, they didn't really ever get going in that way. The initial SARS, even Ebola, which was a big scare a few years ago, it just never spread in the way that this is spreading. I think there is some pretty nice technology that's been built up for screening. I think we're going to do vastly better than in the past. Is it going to be perfect? I'm sure not. It's going to be certainly vastly better than the past.
I think we'll be able to narrow down quite effectively to a small set where we can then take into sort of truly live neutralizing assays.
Yeah, I would add that in addition, animal models are coming up quickly. A number of groups are working on this, and we believe that improvements in animal models will become available over the course of this project. Of course, we will make appropriate use of them.
Very good. David, you obviously have the capacity of a large company behind you. I guess the question is, could you take us through the lights of how much you think would be needed in terms of antibody manufactured and for each one of the potential use cases, and what can Amgen do in that respect if you actually had the right clone selected and the right product?
Yeah. We want to be very careful to ensure that we could provide drug supply.
Sure
an important point. We have the ability to scale relatively quickly. How much we will need is, of course, going to depend on the characteristics of the therapeutic itself. What's the IC90? What is the biodistribution? Meaning what sort of dose do you need to really penetrate, say, mucosal surfaces? The pharmacokinetics and pharmacodynamics of the antibody. Those are critical technical features. I think the good news here is that we anticipate that you're only going to need a dose or two because it's really, certainly in the treatment setting, you're going to need coverage for a couple of weeks in all likelihood. We think one or two doses over a few week period should be sufficient.
We and others have done calculations, I'm sure you've seen them out there, if we can get away with, say, something like a 10 mg per kilogram dose, that is feasible and something that you can handle. Obviously, as we select candidates and start to model what that dosing will look like, we'll have to think about production and capacity.
I'm going to push you on that one more, David, which is okay, you got 10 mg per kilogram dose. You can figure out how much a person weighs. You can figure out what you have in terms of capacity. Roughly how many patient doses you would be able to provide, assuming you got the right product with the right affinity and 10 mg per kilogram, call it in the end of the first quarter of 2021. I mean, if we're looking a year out, how much can you make?
I don't know that I want to give a precise answer to that right now because, of course, I would say ballpark, we would be committed to trying to serve as much patient need as possible. It's very hard, I would say, at this early stage, to give you a precise number of doses without understanding some of these parameters that we're talking about.
Like 1 million, 10 million, 100 million. What are we talking about?
I can tell you that our thinking is, what if we had to be able to administer tens of millions of doses?
Okay.
How would we do that? Now, am I promising anything? Of course not, and I don't think anyone in the industry is. We're not thinking that this is 10,000 doses-
Okay
if it's effective.
Everything's got to work. You're not committing to success, obviously. If you think about the product feature, you're looking for tens of millions is what Amgen will attempt. The direction is, if things work properly, this is what we should be trying to achieve.
I would say that is the scale of our thinking now. Imagine a world where you had to provide that. Can we get line of sight to doing it?
Sure. This brings us to the next question. Is this essentially just a bridge to a vaccine? Are you guys thinking, based off our current understanding of the virus, and the infection cycle, that this antibody approach will be a viable approach or would be needed clinically for the foreseeable future beyond the first few years, which is roughly the horizon when we probably start seeing vaccines coming through? Harlan, why don't you take that one?
Hopefully we will have a great vaccine. It's hard to imagine that the first generation, especially with some of the really exciting but sort of yet to be proven sort of RNA-based vaccines, which are likely to hit first. I hope they have efficacy. I think there's a good chance they will, but I don't know that they're going to be truly protective across a broad swath of people. I think there's going to be, so for two issues. One, it is take a while to get vaccine broadly works.
Yep
administering it across the whole world, which is now at the scale we need to be, is also going to take a while, even if it was a perfect vaccine. I don't mean to be super pessimistic, but it's hard to imagine that we're eradicating this in less than a decade scale, not a year or two, even with the development. David might have a different perspective, but Sorry.
Yeah. I would say barring a vaccine, eventually the likelihood, based on everything we know about coronaviruses, is that this will, at some point, transition from being epidemic or pandemic to endemic. Meaning there would be ongoing need for a therapeutic over potentially a many years time horizon, until such time, if there is long-lasting immunity, you've infected a large chunk of people on Earth. That again, as Harlan indicated, fits in the years, not months, timeframe.
All right. Let's hope social distancing will be less than that. Otherwise, I'm going to personally consume most of the alcohol production in California. Just following up a little bit on this issue of pandemic becoming endemic. It feels like an IV is a great product, but if we're going to try to move this away from just being a one-time, let's give everybody an IV and eradicate the disease, but something more endemic, it seems like you might need this to be a subcutaneous injection or something that is much more accessible than an IV, or just hundreds of millions of IVs being administered on a semi-regular basis feels like technically very hard. Is this part of the development plan to take this to a Sub-Q, or is this something that will stay, at least in the first generation, as an IV?
Yeah, maybe I'll handle that. Certainly in the first generation, IV. One thing that I think is an important consideration here in a therapeutic setting, meaning in the setting of an infection, one of the things that you're very likely to want is to get as high a Cmax as you can obtain to ensure biodistribution and penetration into target tissues. Clearly, the intravenous approach would be preferred for that. I would also say that if this becomes an ongoing process and you say, "Hey, do you need, over time, hundreds of millions of doses?" The thing to remember is you don't need them all at once.
Yeah.
That's spread out over some long period of time. What we would envision is a world where you have hotspots coming up, and going down potentially, like a game of Whac-A-Mole. You would be targeting two discrete populations in a city or a region, for instance. In that way, the demand is not going to be all at once like we're facing now, in a pandemic situation. That makes it much more manageable.
Which kind of brings us to the question of, if you're going to go for a prophy setting of sorts, it kind of makes sense, hypothetically at least, to go for a product that will have higher than the natural life of a normal antibody. Try to move away from a biweekly or monthly infusion to something that is every three months or six months. Correct. In your features of the first generation product, are we thinking about an administration every three months, or are we going with native antibodies and a more of a shorter period of administration?
I would think the first generation is certainly to try to get to the clinic as quickly as possible. One would envision a shorter infusion or treatment cycle, if we want to call it that. We are thinking about some of the standard mutations you can make in the Fc domain and other standard engineering techniques that we apply to increase half-life, and that may change the efficacy profile somewhat. Obviously, we're going to be looking at that intensely as we go along. I don't think the initial goal would be to say, "Hey, let's get a Q3 monthly product here." I'm not sure that would actually be required.
Makes sense. I have two more questions, then I'd like to have Vincent come in with a couple of questions of his own. First, everybody seems to be developing IgG. I think you guys are the same. You talked to virologists to discuss IgA or maybe IgM as a preferred antibody type in terms of penetrating the lung epithelium. Can you talk a little bit about, is this feasible? Is it feasible large scale? Is it something you're looking at? Is IgG going to be good enough to prevent infectivity, not just prevent people from getting sicker, by giving a dose IV?
Sure. One of the advantages I think the virologists have here is they've never tried to produce metric tons of a product. It's a little easier to speculate. There are reasons that the industry over 40 years has gravitated towards IgG. It's got the most firm favorable half-life of the isotypes. I think critically, IgA is essentially a dimer of dimers. Of course, IgM is pentamer or hexamer of dimers. They are much more complicated to produce at scale for a variety of technical reasons and simple things such as the glycosylation patterns start to become quite complex in IgA and IgM molecules, meaning that the process development and formulation are more challenging. Are they impossible? No, but they are certainly not a preferred approach, and they certainly would be an approach that requires a much longer timeline.
I think that we, as an industry, will get there towards making these sorts of antibodies, but the timeline that we're shooting for here I think really leads us towards IgG.
Percentage, exposure in the blood versus exposure in the lung epithelium. How much lower is the exposure in the lung epithelium? Is it 1/10% , 1%, [uncertain]? How much can you actually get in there?
I think that's what we're going to be trying to understand. There are many groups around the world looking at that question right now. It's important to remember that the biodistribution into normal lung is likely to be very different than into lung where you've got a fair amount of tissue injury, a leaky vasculature, which may actually help promote biodistribution. This is an important point and one that we're going to be taking a look at with our collaborators.
Okay. last one.
I can take that.
Before Vincent. Vincent, let me just check with one more, and it's the risk of antibody-dependent enhancement. It was discussed in previous models. Some of the physicians are quite concerned, some are not. Can you just give us your perspective of the risk of this with the monoclonal antibody approach?
Yeah. I think a lot of folks here extrapolate from the experience with dengue, where there are basically four flavors that have conserved domains and antibody-dependent enhancement has been an issue. At least what we know about this virus to date, we are less concerned about this. Obviously because we're using the immune system to teach us here in our antibody selection, we hope that we can avoid that issue. Maybe I'll ask Harlan to give his point of view here as well.
Yeah. I agree. I don't see this as a huge risk. We will be screening for neutralization against the virus of interest. It would really be the question of we're asking about a strange off-target effect that we've had in a rare case, but certainly wouldn't be a likely basic concern is in my opinion.
Got it. Vincent, can you jump in?
Yeah. If I could just take that question on biodistribution earlier, just a step further. I guess one of the things we've heard from some virologists is that IgG, mostly that's the antibody of the bloodstream, the extracellular fluids. IgA is sort of where you have more of the mucosal immunity and so forth. I guess folks have kind of wondered what does that mean in terms of the ability to get to, I guess the upper respiratory tract mucosa relative to.
I guess sort of the alveoli of the lower lungs. What's your take there in terms of the ability of IgG to get to sort of the lower lungs and sort of where you get sort of deep pneumonia versus potentially the nasopharynx?
Yeah. It's an open question in the field right now. When you talk about the normal biodistribution of antibodies, remember, that's an intact tissue. Again, certainly in a setting where you've got markedly compromised barriers within, say, lung alveoli, it's a different setting. We also have experience with other antibodies. We're developing tezepelumab for asthma, and the inflammatory response that occurs essentially at the same site that we're discussing here, in many ways for COVID-19. I think there is precedent here as well, certainly in an inflamed state.
Yeah. I was going to say, you guys might have more insight than most into where the antibodies go.
Yeah
The conditions for how this would work. Maybe a second one would simply be, if you think about the different. You alluded earlier to sort of like, when do you need to treat? If you treat earlier versus you treat later. I know we won't know until we actually try this out, but from what you know right now, Tell me, handing capital odds of success in terms of early treatment versus later treatments. What's the point at which you'd say, "At this point, it's slowing down. At this point, it's probably a little bit too late to.
Yeah.
attack the virus versus where is the virus at?
Yeah. To me, in a ventilated patient who's got poor oxygen saturation, that's probably too late now. At what point in the course of the disease is too late, too late? I don't think we know. I think we're actually going to have to understand that probably empirically, in the clinic. I would say one of the challenges that we and the entire field faces right now is that we're going to be doing many things empirically in the weeks and months ahead that we would otherwise prefer to march through scientifically, but we don't have the luxury of that time. Clearly earlier is better. I think that there's no question. Is there some sort of threshold beyond which it's too late, I think is a critical clinical question right now. It's one that we would try to address going forward.
The last one is a corollary here. One other thing we've heard from immunologists we've spoken to is that folks are pretty optimistic about the potential efficacy of neutralizing antibodies and preventing the progression from more mild towards severe disease. I think the one that's a bit more of an open question, and folks sort of fall on different sides of the line, is how effective will this be in actually slowing transmission? Will you simply have a situation where folks still transmit disease, but maybe they're not as likely to then end up in the ICU, for example. What's your sense for the potential on the slowing transmission side?
What I would say is that if you're having an effect on the disease course, you will probably naturally have at least some effect on transmission. Remember, lowering your transmissibility even by 20%, the mathematics of an epidemic tells you that that actually plays through, over time, have a dramatic effect. You don't have to lower transmissibility a lot to ultimately have an effect. I do think a goal would be, again, preventing this progression from earlier stage disease to a more serious disease. If you could do that even in an outpatient setting, for example, by identifying patients who are at higher risk of progressing to serious disease and administering the product and avoiding hospitalization, that clearly helps slow transmission. There are lots of knock-on effects that one can envision here. Again, we're going to have to understand a lot of that empirically.
With 10 minutes to go, I think we're going to have to go to the next question I'm pretty sure everybody wants to know, which is timelines. David and Harlan, I don't think you're going to put a Gantt chart up as part of this discussion, but as specific as you can get in terms of the logic of the pipeline, of the timeline. Arvind, if you could chime in after them and let us know, when do you think Amgen would likely put first release designs around what stats? It would be really appreciated.
Sure. I'm not going to speculate on a timeline given that it's such early days. I would say that we're trying to work on a month, not years kind of timeframe.
Sure.
We're doing many things in parallel that we would do in a typical development program sequentially. Our goal would be to try to get into the clinic on a month's timeline. Again, balancing getting the highest quality antibody against speed, and there's some tension.
Right
inherent in those two attributes. This comes back, I think, also in what informs our thinking here is part of what you started with, which is: If it's more than one antibody required or a cocktail of antibody, it may be all of ours, or it may be a mixture of ours and someone else's antibodies. One might also see a wave of products entering the clinic, where some are less efficacious, and we get better at understanding over time what makes a better antibody. All of those things will affect our thinking as we move into development. To be honest, I think it would be disingenuous to give a timeline when we are moving forward with the research phase right now. Arvind will chime in here in a moment, but I think we'll commit to providing updates when we feel we have meaningful data.
I've given the team a clear mandate here, which is, let's worry about progress, not press releases.
Progress will determine when we disclose things publicly, which will be when we have something meaningful to say.
David, I think that's very well said. We can't speculate on the timelines, Ronny. As we get meaningful updates, of course, we'll provide those progress updates.
I'm going to pressure on this just a little bit and say that, David, if you're not in the clinic by the end of this year, would you be disappointed?
It depends what we ultimately enter the clinic with.
Okay.
Again, as Harlan said, if you can pick the Michael Jordan of antibodies, I'd be happy to wait two or three more months to do that. I think you're likely to have a much greater impact. Again, we will be balancing, optimizing antibody selection with speed as we go along.
Okay. I got a couple of questions on the email, so I'm going to jump into them now. One of the question is, and then we're going to dive into the science here a little bit, but the relative roles of B cell and T cell immunity in COVID-19. I'm sure there's a lot we don't know yet, but to the extent we do, can say certain things. Can you talk a little bit about the relevant role of both of those?
Yeah. Very roughly, for sort of the way I would describe to a lay audience what B cells and T cells do, T cells clear pathogens, and B cells prevent reinfection. Normally speaking, B cells or antibodies come up later, and they're not playing a huge role in clearing the infection. Which is why finding a neutralizing antibody is rare, where there you're actually looking for an antibody itself that can clear infection, which is not normally what they do. That's a pretty special antibody. That's how we would think about most vaccines are trying to induce an antibody response that will convey long-term immunity or protection. The T cells are a necessary part because they coordinate the response, the helper T cells. There's obviously some T cell memory as well. Primarily, people would think about antibodies as driving the protection.
David?
Yeah. I don't think I have much to add to that. I think that's our understanding. Our understanding in the immune response to this virus is evolving very rapidly, and there are some strange features of the immune activation that appears to go on here with IL-10 spiking in ways that, for example, haven't been seen with some related viruses. Macrophage and neutrophil engagement that seems a little bit odd. I think we're in the early days of that understanding, but understanding that immunology will be critical to give us insights into how to optimize these therapies.
Arvind, I got a couple of questions from a broader audience. The first one is, if you can comment on this issue of live attenuated virus or a killed virus is the ultimate, what we should be aiming for in terms of getting immunity. Is this just being an immunologist and being so deep in this, it is really what we should be aiming for when you think about a vaccine? That's question one. Go ahead. Why don't you take that one first?
Yeah. I think ideally, yes. The immune system is a system. In order to be most efficacious you need to engage the entire immune system. As David was just saying, even beyond the adaptive immune system, there's parts of the innate immune system that are playing a significant role. Certainly, having a broad, well-coordinated response, we believe, will provide the best overall protection long term. That would be an ideal. However, that's a longer road to tow than sort of these new methods of these RNA vaccines that can be made very quickly. There's the balance there.
There's a trade-off. Yeah.
Yeah.
Yeah, I would say it's early days. The SARS and MERS experience should be humbling to all of us. You could actually generate an immune response in some of the vaccine attempts there, but it was never really neutralizing and trying to understand why that was is very important right now. Are there things about this virus, and there may be, that actually lend itself a little more to vaccine development. That's for, I think, the vaccine experts to comment more than me.
David, I guess I want to follow on this because it's kind of interesting. It sounds like you've looked at the experience of us trying to develop vaccines against other coronaviruses. It sounds like you don't think that there are some issues with developing vaccines for coronaviruses. Can you talk a little bit more about that, just to educate us all about what has happened before and perhaps technically why the coronavirus is being a pain for vaccine development?
Yeah. I think the short answer is we don't know why it's been so challenging. Again, you can generate antibodies, for example, against the spike protein. They ultimately do not seem to prevent infection. Now, I would say there's one big caveat here, which is the effort that's being mounted now is on a scale far beyond what I think we have ever seen before for a coronavirus. The hope is that we can generate insights much more rapidly, and that there are features of this virus that lend itself a bit more to a vaccine. I'm not a vaccinologist, and I would defer to those folks on these sorts of questions.
Harlan, I'm going to give you the last one here, which is a question from the audience to describe a little bit about further how differentiated your approach is. What makes it differentiated versus other ways of taking the Michael Jordan antibody, which I guess is the reason why Amgen is working with you guys. Give us your assessment of where you guys are.
Yeah. I think our advantage is our ability to sequence and screen at scale, so we're able to go after a wider pool of potential antibodies. I know that there's a variety of other groups working in this direction, and a lot of them have their own abilities as well in different areas. That's our particular niche that's kind of proved successful on the T cell front and the same technology which we developed over the last decade on the B cell side works at scale in the same way. I think just from what we know, basically, we think that should be a significant advantage. I don't know enough about it. Most companies aren't super forthcoming about exactly what their techniques and technologies are, so I don't know that I could do a direct comparison across the board of what everybody else is able to do.
Hopefully, there's a variety of successes because this is not one of the areas where we're rooting to win. We're just rooting for a cure.
Vincent, last question for you, if you got something by email that you'd like to ask.
No, I think that's it from my side.
Very good. First, again, this is certainly one situation where we all wish you guys great success in this endeavor. It's very important for all of us. Really appreciate you getting on the phone with us today, and hoping for some good news from you in the coming months. Thank you very much, guys.
Thank you.
Thanks a lot.
Thank you.
Take care.