Good morning. If everyone could take a seat, please. My name is Peter Welford. I'm one of the European Pharma and Biotech analysts at Jefferies in London. It's my great pleasure to be hosting the next company in this track for a fireside chat, which is GSK. And here from GSK, it's our pleasure to have Tony Wood, the CSO. If anyone has any questions in the audience, we're happy to take some questions. We'd love a bit of feedback. So if anyone would love to put their hand up and ask a question, by all means we will. But let me kick off, first of all. So Tony, obviously, since you've been in the job now for just over two years at this point, perhaps you can talk a little bit, first of all, about what the notable changes you've done to GSK since taking over.
I guess, particularly as well, if you just think about the therapeutic areas, can you also just talk a little bit about your attitude to oncology? Because we do get the question a lot. Is GSK still in oncology? Is that still a key area for you?
Sure. All right. Remind me if I get to the end and don't comment on that particularly, Peter. So look, I'm pleased with the progress we've made over the past two years. There's always more to do. But for me, it's been characterized by, if you like, a very clear definition of our areas of therapeutic focus, those being vaccines and infectious diseases, including HIV, respiratory and immunology, and oncology. And what we've done is sort of organized the R&D structure around that and driven better partnerships with Luke's commercial organization. And the consequence of all of that is, while we've maintained the performance of our late-stage clinical portfolio with regards to the number of launches and the cost per launch, what we've also done is shifted it towards medicines and vaccines of larger scale, so higher value.
So other bookends in terms of technology, but I'll just pause there and finish up on oncology, if I may. We've talked about it as an emerging area of our portfolio. Those of you who are at the Q3 results will see that it's growing very well. I'm, in particular, pleased with the progress we've made with Ojjaara and indeed Jemperli as we see the approvals in endometrial cancer there. And you'll see more focus on what we might call carefully selected areas of development for Jemperli. Obviously, next year is all about Blenrep. I'm sure you'll come back to that, so I won't embroider it right now, and the ADC portfolio in general. So overall, I'm pleased with the progress we've made. Always more work to be done, but very happy with where I stand and the team of people I have around me.
Yeah. So before we dive into the individual therapeutic areas, can we just step back a minute and just look a little bit about, can you just say a few words about what it is you think that differentiates at this point GSK versus other large-cap pharmas now in R&D?
Yeah. And let me just pick at the high level the therapeutic areas. I'm going to give you a bookending of our focus on the application of technology. And we talk about science and technology as being core to our R&D strategy. But picking each of the areas in general, I mean, we have a very strong vaccines business grounded in a range of platform technologies, be those adjuvants characterized by the most recent performance of Arexvy, for example, through to other platform capabilities, our emerging platforms in mRNA for flu and COVID, and the MAPS platform. And of course, behind all of that, adjuvanted subunit vaccines and Shingrix really is, if you like, the leading asset in that portfolio. Long-acting technologies in HIV, long-acting technologies in respiratory with depemokimab, and increasingly an application of oligonucleotides into our immunology business as well.
For oncology in particular, again, an underpinning in really ADCs building on from the progress in IO. I'll pause there so we can get into some more detail.
So let's go through some of the therapeutic areas. We're going to start with vaccines. Now, let's kick off with Arexvy because that's obviously a topic of conversation. Now, putting aside, I guess, the vaccines, RFK and whatever his views may or may not be of vaccines, we won't ask you your opinions on that. But if you just think a minute about, for Arexvy, the immunity data and the safety data that we have at the moment to hand, how would you sort of put in context for the audience, when we get four-year immunity data, how a decision is going to be made on when a booster is needed for Arexvy and how you can sort of help the ACIP get comfortable with the risk-benefit to potentially enlarge the age group to at least include the 60- to 70-year-olds?
Yeah. Before I answer that specifically, Peter, let me just frame sort of the ACIP arena for everyone. COVID created a little bit of a false impression about the speed of approval of vaccines. Typically, ACIP moves slowly, particularly with new vaccines, and Arexvy is a new vaccine. We're just beginning to understand its benefit-risk profile. What you saw in ACIP decisions this year was essentially that and an underscoring of the benefit-risk profile in the current two populations of adults for whom we have both approvals and recommendations. And you should expect to see that picture will continue as more data is brought to bear. The most significant data, I sort of think about it in three different components.
There will be the ongoing vaccine effectiveness data that will come out of the CDC and FDA-based systems that, if you like, is a retrospective look and answers questions associated with the protection of at-risk individuals, for example, from hospitalization. That will continue and will weigh into the decisions that they make. It's worthwhile emphasizing that now we have three years' protection for Arexvy. Of course, ACIP will not make a decision on revaccination probably until 2026, and that's why we've not included any revaccination in next year's forecasts. What we will bring in addition to that, and I'm going to do this kind of in order of where I think data will impact, is the continued accrual of immunogenicity data in adults at risk, particularly we published our own 18+ data recently. I expect that to be adjudicated in the near future.
We will add more to that, and in particular in populations who constitute those who are the larger proportion of hospitalized adults. I'm not going to go into the details associated with that. For immunogenicity, we'll continue to assess boosting immunogenicity. You'll see season three boost data on immunogenicity background at the beginning of next year. Remember that what we know, and this is typical for vaccines, is the time between prime and boost usually results in greater boost. And that's all we are seeing with Arexvy. I expect revaccination will be required on three, four, or five-year schedules. The reason I'm somewhat uncertain about that is I'm sure those of you who've been plotting graphs on the decline in vaccine efficacy can see that in the second and third years, it's plateauing to some extent.
That, I think, has more to do with the nature of the individual seasons and therefore the confidence intervals on the data than it does about the profile of the vaccine. It's very unlikely that for a mucosal infection, we will not need revaccination. This is not chickenpox. You get it repeatedly through life, not a single event. So confident at some point in time revaccination will be required. We will bring more immunogenicity data, but what I want to give you, Peter, is a sense that it will be a range of data that ACIP will collect in order to make that decision.
Then moving on to Shingrix, can you just talk a little bit about there, the life cycle management strategies that you have for Shingrix? You mentioned vaccines that last a long time. I mean, Shingrix is a good example now. We've got data that shows it lasts more than 10 years, if not more. Can you talk a little bit about that and, I guess, particularly as well how you're also leveraging the dementia data for that?
Yeah. Well, I won't comment in detail on the sort of geographical expansion associated with Shingrix. That's a topic for Luke. But as you say, the vaccine continues to show long-lasting efficacy. I think we're actually at 12 years now and counting. The interesting data, of course, is the relationship discovered through real-world evidence studies between vaccination and mild cognitive impairment, perhaps dementia. This is an area for which there is very little science. We are taking our own data that we published in September, which was a careful real-world evidence analysis. And for those of you who don't spend time thinking about epidemiological real-world evidence analyses, let me just describe some of the characteristics that we were careful about. The main factor you often have to deal with in these types of analyses is something called healthy vaccine bias.
People who get vaccinated tend to look after their health more effectively. So we were careful to remove that from a very large U.S. database. We actually removed nearly 400 confounding factors, things that would introduce bias into the analysis. And from that, we found that after three years, that Shingrix outperforms or delays the onset of cognitive impairment relative to either a competitor, Shingles vaccination or PCV vaccination. Again, the importance of that being removing healthy vaccine bias, extending on the order of, I think, 27 months on a three-year analysis. So we have this tantalizing view that with the best we can possibly do with real-world evidence, it seems to be having an impact on the onset of dementia. Starting a prospective study requires careful design and consideration of feasibility.
That is work that we're conducting at the moment, and I expect to be able to say more on that before the end of the year.
That's great, and so I guess moving on then to the other part of your business, flu vaccines, but I guess more specifically the future of flu vaccines for GSK. Can you talk a little bit about what you're doing there with the CureVac collaboration for mRNA? But I guess more broadly, how you see, what is your development strategy for the future of your respiratory vaccine franchise?
Yeah. Well, I think in answering that, obviously, I have to underscore what I mentioned earlier, that we will continue to generate data to support life cycle innovation and label expansion and harmonization ultimately for Arexvy. That will be an important part of our ongoing endeavors. Then if you look at our adult vaccination portfolio in general, there are two areas for which we currently don't have competitive vaccines, and yet they represent significant opportunities, one of those being high-dose flu, and we put that in the context of flu/COVID combinations, and the other being pneumococcal disease. In terms of high-dose flu, we've made good progress in developing our own mRNA platform. I'm sure you'll appreciate that the key factor for high-dose flu and indeed flu/COVID is the overall mRNA load and managing that versus reactogenicity.
What we know from the platform is in its first generation. What we have here is sequence optimization from CureVac's work plus base modification. Though what we get is immunogenicity now across both A and B strains. We ran recently a phase II-B study that we announced that supports that proposition, and with reactogenicity, that is at least equivalent, if not moderately better, to that of Moderna. I expect to be reversioning that platform as we go on, but in the first instance, the major milestones are a flu/COVID phase II proof of concept readout next year and starting a northern hemisphere flu study in the winter season of next year.
Makes sense. So let's move now on to the other bit of the infectious disease portfolio. And perhaps you can just frame for us a little bit GSK in antibiotics, an area a lot of pharma has stepped away from. But you actually have a phase III positive phase III data for gepotidacin. Perhaps you can just talk a little bit about that and how we should position that and more broadly how you're looking at this space.
Yeah. I mean, we've talked about the antibiotic portfolio really as a constellation of three assets, of which JEPO is one. And I think earlier in the year, we commented on an expectation across that portfolio of peak sales of around about GBP 2 billion. JEPO is doing really very well. First new antibiotic in uncomplicated urinary tract infections, actually for more than 20 years. It targets a particular enzyme, actually two DNA gyrases, which are uniquely important to E. coli. So what it provides for urinary tract infection is an interesting proposition that although it is broad spectrum, it's broad spectrum only against the pathogens that cause E. coli and therefore doesn't constitute the resistance risk of, for example, the fluoroquinolones.
And just to give you a sense on that, and we expect, by the way, JEPO approval next year, just to give you a sense in terms of that, our estimates are if we can displace about 25% of second-line use of fluoroquinolones, which is at 25% despite resistance and black box warnings that underpins the £2 billion. I would say it's a portfolio that you should consider in that context and not something that I would point to as the major allocation of capital across the R&D landscape.
Then going on to another infectious disease asset, which gets very little airtime, but is in phase III as well, the Bepirovirsen for chronic Hep B. Can you just talk a little bit about your strategy there, when we can get data, and how we should think about that?
Yeah. I mean, it's had very little airtime this year just because it's been quietly exceeding recruitment rates and accelerating through the B-Well phase III studies. For those of you who don't follow the area, chronic hepatitis B is a significant killer. 300 million individuals infected worldwide. Actually, that's probably a pretty significant underestimate. There are no effective cures for chronic hepatitis B at the moment. The community looks at something called functional cure, which is an expression of the surface antigen from the virus, which you can measure in a blood test, and the DNA, which indicates that you have ongoing viral replication. You need both of those factors to be below level of quantification to count for functional cure. And that needs to be observed for a period of time which exceeds the coverage of the medicine that you're using from a pharmacokinetic standpoint.
Bepirovirsen is the only molecule to achieve that. You get some tiny amount of it with interferon of around about 5%-7%, but interferon therapy in that context is intolerable. We have, excuse me, on the order of, depending on whether you're looking at low or very low or medium surface antigen cover, anything from 15% upwards. Clinically significant is 15%. We're running now two phase III studies looking at patients cut by different levels of surface antigen. They're the B-Well studies that I referred to. They recruited very quickly. We're expecting to see readouts towards the second half of next year. That's because we have to wait for this period of time that I mentioned.
And we also, just to remind people, we signed a deal at the end of last year with an siRNA that also targets the pyroversin replicating machinery, if you like, does it orthogonally to Bepirovirsen. And we expect that together those two medicines will produce a broader and deeper impact than we see with Bepirovirsen alone. But even considering the proportion of individuals with surface antigen of less than 3,000, we're looking at, based on our analysis of epidemiology, about 40% of the chronically infected population. So just expect to hear more. It's been quiet because it's been progressing.
That makes sense. Let's go on to respiratory then. So Depemokimab, the long-acting IL-5, basically long-acting Nucala, I guess we could call it. Could you just talk a little bit about, I guess, from an R&D perspective rather than a commercial perspective, how you're developing Depemokimab and how we should think about that relative to Nucala?
Yeah. Well, the first thing to say, we're expecting simultaneous filings for Depemokimab in both severe eosinophilic asthma and chronic rhinosinusitis with nasal polyps. What we did relative to NUCALA was quite an accelerated phase clinical development program. We knew enough about Depemokimab at the end of phase I to be able to predict effective dose, and we went straight from that into phase III. We also know enough about the role of hypereosinophilia in the diseases that I mentioned and others that we were confident to run parallel development programs. So this will be all done for Depemokimab in something like four years, whereas for NUCALA it took seven. The initial indications, as I say, are severe eosinophilic asthma and chronic rhinosinusitis. Just to give you a headline on the eosinophilic asthma indication, there what we're doing is moving treatment focus.
If you like the history of, you start with the bronchodilators in the 1960s onwards. That was about improving lung function through acute intervention, then we've led the field in recognizing that exacerbations are the next wave of treatment in lung disease, particularly in COPD. I'll come on to that in a minute, so what we have for depemokimab in the two SWIFT-1 and SWIFT-2 studies, nearly 70% of the individuals on depemokimab in those studies experienced no exacerbations at all, and this is for patients who had exacerbated at least once in the previous year before recruitment into the study, so it is going to be a very important medicine when you consider that two doses a year protect you to that extent and reduce exacerbations across the broader population by more than 50%. We also have headline data from NUCALA in COPD.
There, the important feature to focus on is the breadth of coverage of different patient types, and what you'll see if you join us for the management event in December is the next wave of our investments in COPD with obviously depemokimab given the data that I've just described there, but IL-33 and TSLP, and all in long-acting formats.
You mentioned NUCALA. You sort of stole the question, but I guess you didn't really address what I was going to ask, which is you said breadth of coverage. So importantly, we've all seen obviously Dupixent is approved for COPD now, and you're therefore potentially the second biologic to market here. So can you just talk a bit about what you mean by breadth of coverage? Why is the NUCALA, MATINEE, and the other study, the AERIFY, different to the Dupixent data that we've seen?
Yeah. So the importance of recognizing Dupixent data is in patients with high eosinophil counts, that's greater than 300, but also carefully selected patients who have a bronchitic phenotype. This is associated with airway reactivity and a phenotype that we know responds well to IL-5. We were careful to include all of the potential patients for COPD. This is the third largest killer in the world. It's a significant consumer of emergency room resources. It's a disease which, if you exacerbate, you do not return to your original level. You're iller than when you started. And actually, about 10% of people who go into ERs with COPD don't come out. So we were very keen on focusing across the entire population, including 30% of the COPD population who have emphysema. This is structural changes in the airway.
What you'll see for Nucala is data which covers both emphysemic and bronchitic patients with high EOs. We were, by the way, the first to show that EOs play a significant role in exacerbating outcomes for COPD. You'll also see a breadth of coverage across the currently accepted different forms of COPD. These are usually determined by, do you have a high EO count? Have you been a smoker? What we have is breadth of coverage across all of those. That's what's exciting the KOL community. The headline exacerbation data, I wouldn't compare directly, but to give you a sense, if you want to dig out numbers for it, look at the metrics and the trial endpoints from our earlier studies, and you'll get a proper sense.
So it's not going to be the 34% that Dupixent has as a headline, but it's going to give broader coverage, which in the population in question is important.
I guess let's move on to another asset then in related, which is chronic cough. Now, perhaps the only sort of late-stage chronic cough asset still in pipeline is Camlipixant. Can you talk a little bit about when we could see the phase III data for this and how you would differentiate Camlipixant with some of the other prior attempts companies have made to get into this market?
Yeah. Let me start with the last question first, and then I'll put the phase III readouts, which are going to be towards the end of next year, into context. So when we acquired Bellus Health, we were very careful to establish a couple of important things about the medicine itself. This is an area of research that's been going on for, gosh, I was looking at these molecules nearly 20 years ago. What was always a problem in the field, the P2X2 and P2X3 receptors, was selectivity over those two things. If you don't achieve it, you have an extreme metallic taste. The previous molecules suffered from that problem. One, it means that you have a lot of discontinuation. And two, it means you can't run a blinded phase III study because it's fairly obvious if you're on active.
The other important thing about the studies themselves is you have to take account of the fact that this is a diagnosis by exception and that you have a range of different coughing frequencies and a high placebo effect. So we've been very careful to do three things. First of all, choose a molecule that has fantastic selectivity propositions. There's a very small dropout rate associated with Camlipixant relative to Gefapixant. Secondly, to make sure that we have placebo run-ins to the study that takes account of the variability that I just described. And we've also worked very hard with the regulator to ensure that the cough counter and the analytical methodology associated with that is going to deal with the original objections they raised with the Merck complete response letter.
What's also true about the population in general is that the higher cough frequency you have, and at the upper end, this is many hundreds, 500-800 coughs per hour, this is a level of coughing frequency that causes broken ribs, ruptured tissues. It's a significant disease and a stigmatizing disease as well. But what we wanted to do is make sure the design of the studies was such that we were properly reflecting the severe end of the coughing frequency where you see the greater effect. So all of that is playing into the design of the CALM-1 and CALM-2 phase III studies.
As I say, I expect they will read out towards the end of next year, but I'm going to use time to make sure that the patient populations in those studies are optimally recruited so that we can properly show the credentials of the molecule.
And then, last few seconds, Blenrep. I wanted to totally make sure we touched on that. You said we should get approval next year. Can you just talk a little bit about how we should think about that? It's often, I still think, sort of overlooked here, given obviously how crowded the Multiple Myeloma marketplace is. But can you just frame for us the data you've seen and how we should think about that?
Yeah. So just very briefly, first of all, the headline is you'll hear about our StatSig OS data from DREAMM-7 at ASH at the beginning of December. That's atypical for myeloma, as are two successful phase III studies, which we have. The headline that we have published is it triples survival relative to Darzalex in the second-line setting. Darzalex is already a pretty good medicine. What we are doing in terms of regulatory submissions, I was careful to time all of that so that we had a very clear efficacy proposition with DREAMM-7, DREAMM-8, and OS and DREAMM-7 so that we can move the conversation on to what exactly the label will look like. And we're busy educating key opinion leaders on managing of the safety effects, which are transient, reversible transient, and entirely manageable with dose scheduling.
The initial impression of the molecule, which was formed by the DREAMM-3 studies, where we knew relatively little about its performance in the myeloma setting, and it was given a high hurdle in DREAMM-3 in the third line plus and in comparison to a doublet of Pomdex relative to Blenrep alone, I think initially set an understanding and expectation, which is not accurate, and so we've been working heavily to make sure that that is properly understood by the key opinion leader community, and as I say, everything going well on the submission landscape at the moment and expecting approval under normal circumstances. I did something unusual. I don't usually talk about submissions until we've had acceptances for submissions, particularly in the U.S., but I felt, given the interest in Blenrep at Q3, it was worthwhile mentioning to you that we had submitted the file.
That's great. I'm conscious we're out of time. We're over time. Thank you very much all for your attendance. Thank you, obviously, Tony, for today's discussion.