GSK plc (GSK) Earnings Call Transcript & Summary

Good morning, everyone. Welcome to day 2 of the 2026 JPMorgan Healthcare Conference, where it's my pleasure to host the GSK fireside session with Tony Wood, the Head of R&D. I'm Zain Ebrahim, European pharma analyst, and we'll move straight into questions. Tony, if that's okay.

Yes, you bet. Let me just say good morning to everyone as well, and thanks for getting out of bed so early. I'm sure not all of you have the advantage of jet lag that I do, so I appreciate it.

So Tony, you've been Head of R&D at GSK for 3.5 years now. How would you reflect on your time as Head of R&D over the last 3.5 years? What are the achievements that you're most pleased with? What are you looking forward to in the future?

Yes. And look, wow, time passes quickly. When I started, I said my first priority would be pipeline execution, and I couldn't be happier with the momentum that we've developed in the pipeline. We got 13 positive Phase III readouts in 2024. That naturally led to 2025, which was a real banner year for us with 5 out of 5 of the approvals we were looking for. And as we look forward, 15 scale launches with a potential of greater than 2 billion peak year sales, which really underpin the growth objectives for the company out to 2031. And sitting behind that a next wave pipeline that is, of around about 25 assets and is increasingly beginning to illustrate the underpinnings of our R&D strategy. I mean I can't help but just emphasize a few of the approvals that we got last year because it was such an important year for us. You'll remember, of course, at the end of the year, we had approvals for depemokimab, now Exdensur as the first ultra-long-acting biologic for respiratory disease. And importantly, remember, a medicine that is effective in reducing the severe exacerbations associated with eosinophilic asthma to a degree of about 70%. And that obviously has a significant impact on hospitalizations. With only 2 administrations a year, that's going to be a really important aspect of us moving the work that we've done with Nucala in clinical remission even further. With Nucala in mind, I was delighted with the COPD label that we got in the middle of last year, particularly with the breadth of the population indication in that heterogeneous disease, and the indication at greater than 150 cells per microliter. Again, Nucala prevents 35% of exacerbations that lead to hospitalization. And for those of you who don't follow the COPD area, it's worthwhile emphasizing, that comes with an enormous burden in the U.S. health care system, the 1.8 million ED visits a year, estimated about 7 billion. And for the patients involved, if you're admitted to hospital with COPD exacerbation, your 5-year survival rate is only 50%. So significant. And then BLENREP, of course, delighted with the fact that we got the secured label that we did there in the third-line setting. Important to emphasize that 30% of myeloma patients are in third-line plus setting and 70% of myeloma patients receive care in the community. And BLENREP is a convenient off-the-shelf option for those folks. Approvals in BluJEpa and PennEVI, I won't go into detail until we can move on in the pipeline. And then with that, having been said, continued momentum into this year. I'm sure a lot of people are interested to hear more about the BEPI data. Obviously, we were delighted to be able to announce that we had met the outcomes for the B-Well study at the top level. I'm not going to be drawn on what those data are. As you might imagine, we want to make sure they're presented in the right way, and you should expect us hear more from us at EASL on that topic. But if you allow me just a few more minutes before I pause and give you -- I promise I'm not going to spend 40 minutes answering your first question. Just a few minutes to stress a few points about chronic hepatitis B and BEPI. I mean, for me, the important thing is the first real advance in 30 years. To put it into context, it's -- hepatitis B has enormous epidemiology, 250 million people living with the disease. It's underdiagnosed. It's poorly treated. Current standard of care achieves a functional cure, which is essentially the removal of surface antigen after you withdraw therapy, only at the level of 1% or 2%. And remember that hepatitis B drives 56% of liver cancers. There are 1.1 million chronic hepatitis B-related deaths every year. And as I've said in the past, we now have a medicine in which we're aiming at a 15% to 20% functional cure in the selected population that we have for B-Well. I'm sure we might come back to that. So I won't belabor the point. One final thing to stress, though, because I do want people to appreciate the consequences of being able to provide a functional cure for hepatitis B. We provided data at -- this is real-world evidence from Drysdale that was published last year, presented at APASL and the Global Hepatology Summit. And what those data show is that if you can achieve functional loss of surface antigen, you get an almost 90% reduction in the risk of hepatocellular carcinoma and a 60% reduction in the risk of all-cause mortality. So enormous epidemiology, a disease with dreadful long-term consequences and now the first effective functional cure and progress in more than 30 years. So really delighted with that. Finally, then just to give you some numbers on what's ahead of us. As I said, 15 reg files already accepted since the start of 2025, 15 scaled assets. We talked about at the end of last year that we started 4 new pivotal studies. I expect more this year, potentially up to 10. So I couldn't be in a happier position with regards to the momentum in the portfolio. And what you can expect from Luke and myself in that context is a continued focus on growth of the company and acceleration of R&D and the use of technology to achieve that. So a bit of a long one, but I wanted to get a few key points across to everybody before we got started.

Definitely. And I want to double-click on some of those key points in terms of the progress you mentioned on each of the drugs that you called out. But before we go there, just thinking about R&D productivity, given the pipeline progress that you've mentioned, how do you measure R&D productivity? And how has that improved or evolved in the time as your Head of R&D? Which areas do you think, in particular, are improving and where do you see scope for further improvement?

Yes. And look, let's first of all, kind of focus on what R&D productivity is all about. And if I can emphasize the position, Luke and I work very closely together. We have a super working relationship. He's going to be focused, as I mentioned, on pipeline growth and acceleration of R&D and the application of technology. I'll get on to a little bit more in terms of the application of technology in a moment, but I wanted to give you a sense of the momentum that we've got building in our late-stage development pipeline and the progress that we're making there. Our end-to-end success rates, and by that, I mean, from preclinical to approval. They've more than doubled since 2018. That takes us back to the partnership that I had with [ Hal ] initially, our cycle times measured as last patient visit to the first file in a major market have more than halved from '21 to '24, they're upper quartile. And of course, because of all of that, we're upper quartile in volume of launches and R&D spend per launch as well. Now this is an area where there's always a need to continue to improve, and there's much more to come. That is not just around ensuring the acceleration of those 15 opportunities that underpin the growth propositions of the company, as I mentioned, but you'll see an increasing focus to bring more of our business, the proportion of business in specialty, which has been growing and a comprehensive tech program that I expect to impact not only in the sorts of areas that we've been talking about, but in aspects of earlier R&D as well. And I might come on to that in a moment. So I'll pause there.

Yes. Just maybe building on that with -- you mentioned technology. So AI and machine learning, we talk about a lot. How are you using that in your R&D process today?

And look, I mean, obviously, in general terms, there's an opportunity for AI and machine learning to improve efficiency of our business across a range of what we do, think about that as a reduction of time and cost. But really, from my lens, the bigger opportunity we see is in terms of transforming what we do in R&D. And I'm going to pick from that a few different segments. We've focused extensively on solving the key problem, which is attrition at Phase II. For us, that goes further. It goes into not only the choice of targets with improved survival characteristics, but also the connection of those targets to very clearly defined patient populations. So think about the union of target and patient. In addition to that, I'm going to talk a little bit more about target and patient, give you some examples. But before I move on to that, I don't want to give you the impression that that's our sole focus. Obviously, we've put a great deal of effort into AI/ML for what we call the predictive design of molecules that whether it's small molecules or protein therapeutics or oligonucleotides. And also, we're getting increasingly interested in new technologies, for example, through our partnership with flagship. I may come on to that later as well. But -- and then obviously, I already mentioned effectiveness of running clinical studies themselves. But let's just go take a quick look back at that target to patient piece because there are a few things that I want to emphasize in the portfolio. You can look at our programs in COPD. I'm very excited about the breadth of the COPD portfolio we have. Those of you who followed us meet the management on that particular topic will remember that we highlighted the fact that we have detailed COPD models across different dimensions of the disease. One highlight that comes from that is our opportunity, for example, to spot the potential in IL-33 and TSLP as combinations. You'll realize we're moving into steatotic liver disease, which we see as an enormous opportunity as well. And their U.K. Biobank data together with detailed disease phenotyping, we hired Sarah Teichmann a few years ago and what her work is doing is allowing us to layer on top of the causality that comes from genetics, detailed understanding of the character of the disease caused by individual pathways. And that was what sat behind our decision to license Efimosfermin, which sits alongside our program with GSK-990 gives us a really nice complementary approach to both the inflammation and fibrosis that's characteristic of steatotic lung disease. One more, and then I promise I'll pause as well. And you might have -- if you followed recent announcements, the Helix data relationship that we announced yesterday, that again, is similar principle, adding much more detailed phenotyping to [indiscernible] type data. And the Noetik deal that we also announced, that's the case in oncology, where we've built a comprehensive approach under [ Tony Ng's ] leadership with digital biologic twins and organoid capabilities. It's going to allow us to really tease apart in the exciting areas, B7-H3 and B7-H4, who are likely to be the responding patients in that context. So hopefully, that gives you a sense of lots going on, underpinning all of that for us. It's not just about AI/ML, it's about collaborations and partnerships, which allow us to access or together generate data that then inform this multimodal approach that links causality from genetics through phenotype, ultimately to clinical readouts that can be initially descriptive of the progression of the disease and image, for example. And then the last step in all of that eventually will be as we continue to explore these areas, the connection of that through to outcomes that patients care about. And you'll see that play out very much in our SLD portfolio.

And you mentioned a few licensing deals there and business development. So Boston Pharmaceuticals being one this time last year, it was IDRx. And earlier this week, you announced the collaboration with Summit on ivonescimab. So just how should we think about the pace of business development going forward, the appetite for more BD? What are your key focus areas? What are you looking to supplement?

Yes. Look, we're very pleased with the BD we've done. Obviously, Luke and I have been close partners together with David and his team in all of that. And so what you should expect to see is pretty much more of the same. And if I can sort of bucket that in 2 ways. The deals you mentioned, the IDRx deal and the Boston Pharma deal, very much in our sweet spot, deals for which we are looking for assets where there's an opportunity for them to drive near term -- by near term, I mean 31 sales, deals for which we expect to be able to accelerate for both IDRx and indeed, Boston Pharma. I mentioned we have pivotal studies that we started within the fall last year. They both fit in that category. We've gone literally from deal signing to pivotal study starts within 6 months in both of those indications. And I won't go into any more detail about what we see as being exciting in the assets in and of themselves, but they're very well aligned to both our research strategies and our commercial strategies. In general, there's a theme sitting behind them in terms of the pursuit of long-acting agents. And in addition to that, I mentioned data and tech-based collaborations. I won't repeat what I've already said, but it is worthwhile emphasizing at early stages. What a lot of the work that we're doing in understanding patient and target is illustrating for us is really the need to begin to change, if you like, or reprogram the cellular proposition associated with the diseases that we're interested in. And that's why you see a focus in ADCs, which are doing that through depletion of particular cell types, oligonucleotides, which reprogram. I'd point you to the Empirico deal that we signed at the end of last year with EMP-012, which will be a long-acting siRNA in Phase I for broad opportunities in COPD and then the LTZ deal, again at the end of last year, and that's looking to take the sort of principles that have applied to T cell engagement and apply those principles to macrophages. So what you should expect to see is more of the same. In terms of Luke's focus on the growth of the company, I'd just take you back to strategically aligned with our areas of commercial interest for me, technically aligned in R&D and a general focus on long-acting agents.

And just maybe pivoting slightly and building on your team of long-acting agents, the most recent one that we've seen in terms of approval is depemokimab or Exdensur. So early commentary on how that launch is progressing, which patients do you see as the most likely to benefit from a 6-monthly option, 6-monthly health care administered option versus self-administered monthly or every 2-month options?

Yes. Look, obviously, only just got approval, so it's very early days. The feedback we have from both patients and health care professionals for Exdensur is they really do value the long-acting component of it. I think we've got early feedback that suggests that upwards of 80% of pulmonologists see it as potentially becoming standard of care in the treatment of severe eosinophilic asthma. And I think it's a feature of bio penetration in lung disease that's worthwhile emphasizing in the context of Exdensur. And that is that -- if you look for those who are -- by penetration is low. It's typically less than 30%. In addition to that, compliance is also low. A year in, typically only 35% of patients are complying with their medications. So there's a real recognition that 6-month treatment is the tipping point at which we expect compliance to pick up. And that coupled with the point that I emphasized earlier that you have about a 50% reduction in risk of exacerbation, the 70% reduction in severe exacerbations. It's worthwhile emphasizing in the SWIFT-1 and SWIFT-2 studies for those who were on depemokimab, high 60 percentage of individuals experienced no exacerbations at all. And that kind of brings us back to the point that I made about clinical remission and what we're doing with Nucala. That is essentially moving patients to a position where exacerbations are no longer a threat that is part of the reality of living with the disease. In addition, you can begin to tailor off on steroids, for example. So delighted with the early insights we're getting for depe and I'd say early days at the moment.

And Exdensur, you also have the NIMBLE study reading out. So could you contextualize for us what that study is trying to achieve? What are you looking to see there? And how important that is to support the launch in asthma?

Yes. I'll expand it to a number of different studies, so I can give you the context of what's going on around depemokimab. Clearly, the important registrational studies with SWIFT-1 and SWIFT-2. Then what we had was another set of studies that we read out, which are read out, which I been tiering as the next most important. They were the AGILE studies, and that was about an extension to look for the duration of effect that we've seen with Nucala out to 2 years. And it was also about switching individuals who were on the control arm on to depe and being able to show that you had durability of effect and continued efficacy in the context of the headline data that we got from SWIFT-1 and SWIFT-2. NIMBLE was really an informational study. We have the data in-house and we're analyzing it, but it's not a significant component of our registration package. It was asking a question in terms of switching, but it's worthwhile emphasizing that, again, the feedback that we've had from health care professionals that is emerging is that many of them -- around 50% would consider going straight into naive patients with depemokimab. Right now, we initially looked at as sort of largely a switch market, but that seems to be diminishing. So really, for me, it's all about SWIFT-1 and SWIFT-2. The other studies have a lesser importance in terms of the registration.

And the other studies that you started as well in Phase III in COPD for Exdensur. So you talked about the success of MATINEE and the rollout there has been strong in terms of NBRx share. So just what were the learnings from MATINEE that you've applied to the design of the extensive trials? And when can we see data there?

Yes. And if you remember, when we were running the MATINEE study, a lot of people were asking me if I had so much confidence in IL-5, why wasn't I already starting the Phase III pivotal studies with depemokimab. And that was because I wanted to be able to learn from the broader experience of the MATINEE studies in the patient population to consider characteristics that would allow me to design the subsequent pivotal studies for COPD with a still increased probability of success and how best to describe it? Clinical effectiveness in the protocol design. So this is an area where as a result of what we learned from MATINEE, we've deployed, again, an AI/ML solution to help us look at -- and there are 2 distinct sets of studies that we started for depe. There is the ENDURA-1 and 2, that's the paired Phase III study in the moderate to severe COPD patients where GOLD guidelines currently indicate IL-5. I wanted to learn what I could from the MATINEE experience to make sure that I had the best possible effectiveness of those studies. But importantly for us as well, we also wanted to see if we looked at more detailed characteristics of the patients and their progression, could we identify a less severe population, mild to moderate, where progression propositions into the more severe disease would be a risk for those patients. And could we use depemokimab as an opportunity to help the trajectory for those individuals. That's what's behind vigilant. And so waiting for MATINEE was about securing that overall proposition. I might, if you're okay, use this as an opportunity to talk about the broader COPD portfolio because I'm particularly excited by the assets that we've built there. A common theme across all of our COPD portfolio, again, is in the ultra-long-acting area. And let me just stress a point that I made earlier. ultra-long-acting brings confidence in protection. And if you have a disease in which a hospitalization really does reduce the quality of your life, your prognosis, COPD patients never return to the place they started after an exacerbation. So really important for us. in the ultra-long-acting portfolio. COPD, having said that, is a heterogeneous disease. We were delighted with Nucala to get a label, which covered both bronchitic and emphysemic individuals. And so as we're looking to expand our long-acting respiratory portfolio, we have T-slip. That was a deal that we did a couple of years ago now -- well, actually, I think it was the beginning of last year, initially for low T2 asthma, but we're also looking to position that. As I mentioned earlier into COPD, that sits alongside IL-33. And this is another area where through collaborations with Cambridge University and with the Boston University Center for Regenerative Medicine, where together with the data that we've gleaned from our own clinical studies, we're developing a detailed understanding of the underlying phenotypes stratification, if you like, of COPD patients. All of that is playing into the path by which we will initially establish doses and confirm those responsive patient populations across 33 and T-slip and then move into pivotal studies. You should expect to hear more from us in terms of those pivotal studies in the next 2 to 3 years. But I really am delighted with the position that we have with our long-acting agents. As I mentioned, we're not leaving it there. The deal with Empirico looks at a broader approach. We haven't disclosed the target yet. So I'm going to keep that close to my heart, if you don't mind. And then I can't -- just to complete the picture, the Hengrui deal, which we signed at the end of -- at the beginning of last year, delighted about the -- it's a real lead of that deal. It gives us access to 11 mechanisms. We've split them pretty much 50-50 between oncology and R. And the first example there is the PDE3/4 molecule there. That likely will be deployed against dyspnea, which is pain with breathing. If you look at how the Verona asset is being deployed from GOLD guidelines. So you can see I've got a pretty comprehensive portfolio looking to deal with COPD. And again -- as well as the health care burdens that I mentioned, a disease that impacts 300 million individuals.

And how is your PDE3/4 differentiated from the Verona or Merck asset?

Yes. Let me just take a couple of minutes to sort of describe it. PDE3/4 has been of interest in inhaled therapies since I was in shorts. I kind of been following that mechanism. I've had a few goes at it myself and my days as a medicinal chemist. And I'd say it's not entirely clear yet between the 3 and 4 mechanisms, whether or not the effect is more about bronchodilation. And in fact, if you look at the Verona data, most of their data is really an improvement in lung function rather than a reduction in exacerbations. So what I wanted was not to be too clever, I wanted to exploit the fact that somebody had already shown me what the right profile was, except that, of course, their molecule is nebulized, which gives significant patient compliance issues. You have to sit down and spend time inhaling the medicine. We think that the 3, 4 that we have with Hengrui potentially could be part of a DPI and then it fits beautifully, of course, with our Trelegy franchise. So that's how you'll see us begin to differentiate that asset.

And my last question in the respiratory space is on camlipixant, where we'll have data later this year. So could you talk through latest expectations on timing of that data, the unmet need that you see in refractory chronic cough and what we should be looking to see in terms of placebo do benefit when we see that data later this year?

Yes, sure. So an eagerly anticipated readout from everyone, I think. Let me just start a little bit with the epe of refractory chronic cough to begin with. It's, again, a disease that fits and do I get from a friend help later. It's a disease that very much fits in the poorly treated and therefore, underdiagnosed category. Right now, we see about 28 million patients globally, 10 million with a cough that lasts more than a year. I'll come on and describe the characteristics of that cough in a minute. And those 10 million, we see 3 million in the U.S., 3 million in Europe. There are about 1.8 million individuals who are already currently managed by pulmonologists. And these folks have had a journey of typically seeing 3 health care specialists before they arrive at a pulmonologist. Imagine a life in which you're coughing between [ 500 ] and 950 times a day. Imagine that cough being clustered in such a way that it leads to broken ribs, incontinence, vomiting, depression on top of the stigma associated with sitting on an airplane or in a restaurant next to somebody who's literally coughing all of the time. And so there is huge medical need here. Unsurprisingly, only 3% of patients are satisfied with current treatment options. So that's what really caught our attention about camlipixant. And again, an area that I've been following for the majority of my career in the early days as a medicinal chemist, these [indiscernible] receptors. And P2X3 is overexpressed in the no-dose ganglia of your lungs. And as that overexpression raises, it's that sensitization, which causes the coughing. Now it's an area in which the ability to get a molecule that targeted that without incidentally hitting P2X2 was really a big problem. Because if you hit P2X2, you cause a very unpleasant taste disturbance, and that's the problem that Merck has been handling. We found camlipixant with a molecule that had vastly improved preclinical and clinical selectivity. Merck's incidence of taste disturbance is somewhere in the region of 60%, well less than 6%. And so that, coupled with what we saw in the SOV data, which was a 35% reduction in the coughing frequency across both doses actually with only, as I say, 6% taste disturbance really set this up as a case where we felt not only could we find an important medicine for these people. But we had one whereby the characteristics of the molecule itself would overcome one of the major obstacles that Merck faced with gefapixant, which was functional unblinding of the study because of taste. Now having said that, this is another example where actually being second is pretty beneficial for us because we've been able to learn from the other aspects -- sorry, excuse me, the other aspects of the CRL that Merck got. We've worked closely with the agency in terms of the data processing and the patient-reported outcomes, for example. And we've designed the study as well to take account of some of the degree of placebo effect. There are 2 studies running, KALM-1 and KALM-2. For those of you who are tracking them, let me just give you the headline on where they are. KALM-1, last patient, last visit at the end of last year, as we said, we're currently transferring data. You'll appreciate when you're using a digital device to count cough, data transfer is something that we need to pay careful attention to. And the second study, KALM-2, which we extended because I was given the opportunity through regulatory interactions to increase the number of high coughers into that population. That now has the last patient first visit. We expect it will complete around about the middle of the year, and we'll report as we do typically for these Phase III studies, on both Phase III studies together with the intention ultimately is to pool the data. So as far as I'm concerned, everything is proceeding as we would expect with respect to KALM-1 and KALM-2. We have the best molecule. We have a really great clinical trial design, and we remain confident in the outcome on the basis of what we've seen from the SOT Phase II study. By the way, we see 15% to 20% improvement in daily coughing frequency as being clinically significant for those of you who are trying to get ahead of when we report the results.

That's very helpful. And maybe now shifting gears to oncology and maybe starting with BLENREP, where you had approval in the U.S. late last year in third line plus to begin with. You've had approval ex U.S. and second-line plus. How is that initial launch going? And maybe broadly, how are you thinking about the development plan going forward based on your latest discussions with the FDA?

Yes. So let's just start with the launch, first of all, obviously, early days. And you'll remember, we got the label first in the U.K. So a couple of things to stress about that U.K. experience. So far, we've got 50% share in the Tier 1 HCPs in the U.K., around about 10% to 15% in the second-line patients. So that's all starting in the way that we'd hoped. You'll remember, Luke said we're going to go slow to go big, and this is really all about ensuring that we help the treating physicians get the right network of care that's required to be able to administer BLENREP. What we hear about the REMS is great feedback in terms of the simplification that we were able to establish there. We've trained several hundred HCPs. By the end of the second quarter, we expect that to be at 7,500. Let me just talk about a few things in terms of third line and the path into second and first line to sort of preempt a few more questions that you have. Worthwhile emphasizing third line plus in the U.S. is 30% of myeloma patients. The market is currently around about $20 billion. It's projected to grow to $40 billion by the end of the decade. Our expectations for BLENREP greater than $3 billion. So I think even with third line at 30%, we're in great shape. Obviously, we want to get to second line. You can think about the major components of second line being the continued following of the DREAMM-7 study. The FDA approved third-line BLENREP on the basis of the third-line patients who are in DREAMM. They progress more rapidly. So as you'd imagine, the data were more mature in that setting. And just to reiterate it for everyone because it's not often we get the opportunity to have data with this sort of impact on patients. There's a halving of the risk of death and a nearly tripling of progression-free survival. That's from 12 months out to 30-odd months. So a really significant impact for these patients for whom actually, there are very few options. Myeloma needs new mechanisms. And BLENREP is really the first off-the-shelf opportunity in that setting. So particularly for community use, where, as I mentioned earlier, about 70% of patients are treated. So you should expect to see us continue to follow the maturation of the DREAMM-7 data. We're running a single-arm study that we're calling DREAMM-15 that we're looking at different combinations. And we're continuing to progress with the DREAMM-10 study, which is in the first line. What's important there is we have reduced loading dose and then increased dose intervals in the first-line setting. That will be an MRD to PFS endpoint. Typically, you look at MRD 9 to 12 months in. And obviously, we'll be continuing to evaluate with the real-world experience of managing the ocular side effects. All of that together to cut a long story short, potentially gives us the opportunity for refiling in second line and first line at the end of 2028.

That's very clear. And we've seen -- you mentioned lack of options. We have seen some data now from MajesTEC-3, the DARZALEX/TECVAYLI combination, which looks quite compelling data. So what's your perspective on the efficacy there and how the treatment landscape might evolve with the launch of DARZALEX and TECVAYLI and how that could impact BLENREP?

Yes. So look, first of all, as I said, it's great news for myeloma patients. They need new options, and this is another BCMA-based opportunity for them with great efficacy in the triplet that you mentioned. I think though, if I go back to the point that I made that 70% of myeloma patients are treated in the community. And let me sort of just describe the treatment experience with BLENREP. In the first year, you'll get between 5 and 8 infusions. These are 30-minute infusion delivered in an outpatient setting. In the second year, typically looking at 4 infusions, we know that you can stop therapy when grade 2 or 3 ocular events start to appear, and you can do that without impacting efficacy. The MajesTEC data in the TCE setting still have the reality of treatment with TCEs. These are incredibly powerful medicines. So you get hypogammaglobulinemia, which means that you're going to need -- you need IgG infusions. Grade 3 infection rate is still high in that group. If you follow the data closely, you'll see that actually the Kaplan-Meier cross over in the first phases. So clearly an important opportunity, but we see the convenience of BLENREP and the non-life-threatening side effect profile as being something that will give us advantage in the community setting. Actually, it's worth -- it reminds me to say one more thing about the DREAMM-10 first-line study that I forgot to mention. We'll also, of course, be looking to add a quad arm to that to look at those individuals treated in more hospital settings, particularly for those who can tolerate the quad regimen. But I very much see this as great news for patients, but we're operating in a different segment and the simplicity of BLENREP relative to managing the side effects with the TCEs, I think, is something that looks like it will continue to persist.

And thinking about HIV, I think you said you will have an update in the middle of the year about your 6 monthly options. So can you remind us what you're thinking there, what we should expect to see from you at that event in the middle of the year?

Yes. I mean let me just on behalf of David, give you a sort of a broad overview of what's going to happen this year. I think, David, if you don't mind, I'll quote what you said yesterday that this is going to be a year in which you'll hear much more from us on HIV. One of the key pivotal readouts that we're expecting this year that will come with a file is the Q4M PrEP data. We're also going to start a Q4M treatment. And I think what's really important to emphasize here, remember, treatment is by far the majority of the long-acting opportunity. And we are the only ones with an existing licensed Q2M treatment option. What's really important about treatment is the backbone of an integrase inhibitor. Integrase, for those of you who know a little bit about me, my history is in HIV and integrase was a transformational identification of mechanism of action, one that has fantastic durability, but one that's been relatively difficult to find chemical matter for. So we're delighted that we'll be in a position where the Q4M treatment will start this year. We're very excited with VH184, which is the third-generation integrase inhibitor that has not only the characteristics, which we think will support it as a component of Q6M. And that's the point actually which you see a bigger conversion of the market from oral into long-acting. So very comfortable that it has the appropriate characteristics to support that as well as an improved resistance profile. And if you'd ask me, what did I expect to see an improved resistance profile on top of dolutegravir, it would have been, wow, that's a stretch. So we're in great shape because of our partnership with [indiscernible] on that, I'd probably better stop there given that we're out of time.

Thanks, Tony. Yes, we are at the top of the session, but thanks a lot for this at the time. It's been a great discussion. Thanks, everyone, for being here, and enjoy the rest of the conference.

Great. Thank you, Zain.

Thanks, everyone.