Thanks for joining us here at another of Guggenheim's Healthcare Innovation Conferences. We do this annually, and I'm really pleased to have GSK here with us today. For those of you who followed me a couple years ago, you saw how many times I made the mistake of calling GSK GlaxoSmithKline. I'm really pleased to have Dr. Kaivan Khavandi here with us. He's Senior Vice President, R&D, Head of Respiratory, Immunology and Inflammation. And really, I think this is a great opportunity for us to talk about GSK's approach to immunology and inflammation at this point. Can you maybe just kind of walk us through how you're thinking about the kind of discovery, development across the specialty R&D portfolio? And then, maybe just give us a little bit more information about your role, and how you're approaching the opportunity to accelerate the growth strategy at GSK.
Great. Thanks for having me here, Seamus. I'm head of, as Seamus described, an end-to-end R&D unit that covers all non-oncology and non-vaccines ID product discovery and development. I also lead another organization in GSK that's called Translational and Development Sciences, which is really focused on where we derive the highest human causal confidence to decide what we want to advance into early clinical developments and how to prosecute that in the context of complex human disease. That's a good segue to answer your question, which is, as an industry, I think we've moved away from the chronology of do some biology, then decide if it's relevant to humans, then have a handoff where you put a clinical development hat on and you start to think about trial design and what patient phenotypes you want to study. We're reverse translating, where you start off with those questions.
What mechanisms do you want to deploy into disease state-specific settings? That's really important when you think about the diseases that I'm focused on, which are chronic obstructive pulmonary disease, metabolic dysfunction-associated steatohepatitis, pulmonary hypertension, chronic kidney disease. These are complex, chronic inflammatory diseases which are heterogeneous and typically present with multi-morbid risk. Almost none of those diseases present with an isolated end organ dysfunction that's distinct from a system that's perturbed.
When we think about discovery, we're increasingly thinking about discovery through translation. The acid test for that is when you start to think about target identification, immediately think about what the phase II proof of concept study is. That's a natural force function to accelerate, because sometimes it means, "Actually, I don't need to do five years of biology. I've got what I need, and I can test this clinically." Instead, the most material enabler for the program to progress is identification of a biomarker that identifies the responder population. So doing these things by design and by prediction rather than empirically and retrospectively.
Okay. I am going to jump right out into it. What do you think is the most underappreciated asset in terms of the, where the street is maybe just not paying attention, or that you are particularly excited about within the RI&I portfolio?
I will start off with a disease area, which is chronic obstructive pulmonary disease.
Yeah.
You hear these stats, but I am not sure that they ever really sink in when you think about the fact that 400 million people globally are affected, third biggest killer. It's not a good premature mortality. It follows an increasing degree of morbidity, burden on healthcare systems, and we are really in the infancy of targeted therapy. There are two approved biologics for a prevalent fatal disease. We focused very significantly on obstructive respiratory disorders, including COPD, but also non-CF bronchiectasis, which obviously we have seen now with the first approved therapy is tractable and can generate significant value, but still has huge unmet needs.
We see COPD as an area that is probably underappreciated, and then we have a true portfolio strategy where we have used data, technology, and experience to partition out what are the distinct drivers of risk in different subtypes of COPD, and we have a solution across those. Within that, then if you think about the IL-33 mechanism, we have what we believe to be a best-in-class approach with a long-acting profile, but actually, we really telegraphed what ended up happening with the class by initiating a study in non-CF bronchiectasis.
Which is not a Th2-driven eosinophilic disease. It is a disease that is defined by mucus plugging, airway remodeling, typically in non-smokers. That is currently in phase II, which we will read out alongside with our phase III initiation in COPD. One study of which, very excitingly, we have announced will include a cardiopulmonary outcome study. Patients with COPD have exacerbations. Those exacerbations are prognostically important. They predict cardiovascular all-cause mortality. We are looking forward to generating what would be the pivotal study with the most hard endpoints evidence in support of a product.
Great. You mentioned IL-33, but the programs that you and the team have had responsibility for include, on market, depemokimab now, and the sort of long-acting TSLP. It is very clear that GSK is taking a concerted view that long-acting assets may in fact be the path forward for immunologic disorders at a minimum. Can you talk a little bit about that, and is that just a commercial angle, or what does it mean in terms of the actual disease state?
We started off with ultra-long-acting, as you said, with Exdensur and asthma. If you think about asthma, we have got to a point in terms of targeted medicine profiles and efficacy effect sizes, which are very, very good. If you look at the Exdensur SWIFT data, over 50% annualized exacerbation reduction. In some regions that might be more pure, uncontrolled populations, efficacy gets up to 80%, 90%. I think we are naturally reaching a ceiling or diminishing return in terms of both the likelihood of having a product that has got greater efficacy in an RCT, but also, frankly, the operating characteristics of the trial endpoints to even demonstrate and tease out those differences, even if the biological truth is there.
We call ultra-long-acting pragmatic innovation, which is in a disease where most patients come off drug by a year, which is the case with asthma, and only a third of the population initiates a biologic, what is the way in which you can most readily transform care? That is to introduce a solution where patients persist. They therefore have sustained benefit from the types of clinical measures we study in trials, which are spirometry, exacerbation reduction, PROs, but actually the community is moving towards concepts of remission, and you can only achieve remission if you stay on the drug long enough.
Yeah.
We think ultra-long-acting in asthma is absolutely critical and where we've obviously invested in terms of the next wave. COPD requires more mechanistic innovation, but of course, all those attributes of the long-acting are important. In COPD, if you have a single exacerbation, your entire disease trajectory could change, and you could die on that admission, of course.
Yeah.
But typically, you have a step change in your disease status that you may never recover. So having that kind of sustained coverage from a long-acting, it doesn't take a big leap to see how persistence translates to efficacy.
Got it. Can you talk a little bit about the differentiation of your long-acting TSLP, and how you move forward with such conviction into asthma? I guess one question is that choice of CRSwNP, so the nasal polyposis endpoint, versus potentially going into other areas like EOE? So interested to know, and also how you're thinking about EOE at the same time.
Our long-acting TSLP, felcorekibart it's called now, came through our acquisition of Aiolos, which was a product that originated from Hengrui, who we know very well as a Chinese partner. Hengrui had established data actually in nasal polyps in phase II, which is part of our collaboration and partnership we had visibility of. Now, we wanted absolutely no uncertainty that this is truly a twice-yearly dose medicine. So we initiated a phase II study, the NAZARE study, in asthma, that evaluated three doses, and looked at a number of pharmacodynamic biomarkers as well as spirometry and PROs.
Part A of that study, which was after the single dose, with six months of follow-up, read out a couple of months ago, and the data was incredibly encouraging, sufficient that we were able to accelerate six pivotal phase III starts, again, in conjunction with the data we had from nasal polyps, and that's the PERSIST program across, as you say, asthma, nasal polyps, and COPD. That provided a very material acceleration of nine months. We're going to share the data next year. What I can tell you now is that there's absolutely no uncertainty that it's a twice-yearly dose medicine. Very encouraging indeed. Of course, we know a lot about the mechanism already. The mechanism's been de-risked, certainly in asthma, where it provides eligibility that's eosinophil agnostic and strong efficacy.
I think in chronic rhinosinusitis with nasal polyps, I would describe it as best in disease efficacy, to answer your question. We know that the synergy and halo effect, particularly across nasal polyps and asthma, is very powerful in terms of uptake, which is greater than the sum of the parts. So you get that true benefit across those two indications. COPD, obviously, the efficacy hasn't been proven yet. The core study was negative. There was a sub-study in the patients with eosinophils over 150 and 300. But this is where GSK's got a real advantage. We effectively have these knowledge graphs internally with huge amounts of proprietary data, so we can integrate those types of observations with prediction based on genetics, genomics, and clinical exposure data, which encouraged us that TSLP truly is a mechanism that's designed to operate in that intermediate Th2 space.
To answer your question around eosinophilic esophagitis, that's obviously not an indication that we're currently in. We obviously followed closely the data from TEZSPIRE. I think my immediate extrapolation is that it demonstrates proof of concept in a disease that is defined by epithelial injury. Mucosal epithelial injury there. There's quite a lot you can learn from that in terms of the types of combination work that we're doing with TSLP and IL-33. So we'll watch EOE. At the moment, obviously, it's a pretty ambitious program to kick off six pivotal studies simultaneously.
Yeah. Then, in terms of the conviction, can you just talk a little bit about the dosing that you've explored and that we're seeing in the phase III? Is there differentiation between individual disease states as well?
We've chose a single unifying airway dose, which is why it was so important that we had phase II data from Hengrui in nasal polyps, and then we had data in our hands with asthma, and then we were able to perform pretty high-confidence modeling to inform the COPD dose. That's obviously against a background of GSK as the company that's proven the ability to skip actually phase II altogether, which we did with Exdensur, where we went from phase I to phase III for asthma, and then we obviously initiated the ENDURA and VIGILANT phase III studies in COPD. In our asthma study, we tested three doses, as I said, with comprehensive range of pharmacodynamic biomarkers and patient-reported outcomes and spirometry. The data's very good. So as I said, unequivocally, a twice-yearly dose medicine, and confidence in the dose that we've selected across the three indications.
Right. Let's talk about IL-33 in particular. The feedback that we're getting on the tozorakimab data so far has been incredibly positive as it relates to exacerbation reduction in particular, but there are some questions around the sort of clinical, or I guess not the clinical benefits there, but benefits on spirometry, FEV1. So how do you think about that trade-off with IL-33, and/or do you believe that your own has an opportunity for a differentiated result?
It's a great question, and it's interesting because a lot of how the industry's approached biologics and COPD was extrapolated from asthma. In turn, a lot of how the industry's approached IL-33 was extrapolated from IL-5, TSLP, and IL-4/13. It's a fundamentally very different mechanism. Again, I think we had disclosed how we thought about that mechanism before the tozorakimab data in saying that you wouldn't necessarily expect it to have really compelling effects on spirometry. It's a true alarmin, and that term's used loosely. People call TSLP an alarmin. TSLP is more like a master cytokine. It inhibits a lot of the other things that we're interested in.
IL-33 is truly that archetypal alarmin that is responding to this acute, perhaps life-threatening injury with something that it thinks is beneficial, and probably from an evolutionary context, at one point was beneficial to try and get you to survive an acute injury. In the context of chronic sustained inflammatory disease, it has become maladaptive. It is different from those Th2 mechanisms. We were not surprised at all that you get benefits agnostic to baseline eosinophil levels. Again, we see it much more as a prognosticated marker. That is what the data from the genomics show you, that if you have a doubling of your circulating IL-33, you get an 80% increase in mortality. That is what the SPIROMICS data show.
That' s why we see it as a mechanism that will have a compelling benefit on exacerbation reduction, but it also operates in this gray area of cardiopulmonary risk, because IL-33 is co-expressed to the largest degree on lung epithelial cells, but also vascular endothelial cells. So it is expressed on your blood vessels, and that is what encouraged us to announce our plans to move forward with this cardiopulmonary outcomes study.
Interesting. The bronchiectasis opportunity, can you talk a little bit more about that? We are only just becoming educated on that opportunity through the Insmed launch.
Yep.
Maybe you can help us kind of compare and contrast where you think DPP-1 versus IL-33 operates and—
Yes.
—and walk us through that.
DPP-1 works on preventing neutrophil elastase activity in your bone marrow, actually. And a great success story from Insmed, great for patients with non-CF bronchiectasis to have an advanced therapy. I think the efficacy was in the region of 19%-20% annualized reduction in exacerbations. Important, but clearly insufficient to provide the management that we aspire to for those patients. Lots of room for other innovation, and obviously you've got the on-target adverse effects in terms of the plantar effects that you see with brensocatib. Then when you think about what have we learnt about IL-33 from COPD.
Still debatable, but if you look across all the evidence, it appears to be most effective in patients that are actively smoking, appears to have a predictive response in patients that have mucus plugging at baseline, appears to have a benefit on mucus plugging, and appears to have benefits that don't require Th2 substrate. Then when you think about bronchiectasis, it's all of those features. Non-smokers, mucus plugged, who don't have Th2 disease. And so you could start to think that actually what AstraZeneca studies by enriching based on the CAT score and productive cough was a bit of a phenocopy of non-CF bronchiectasis.
That encouraged us to start a phase II study with James Chalmers, who was also the PI for the brensocatib studies in NCFB. Huge unmet need. Again, I have described COPD as being nascent. NCFB is even earlier in its journey in terms of drug development. Then of course, if you have a product that is effective across both of those, you can imagine the simplified proposition for prescribers who do not need to necessarily worry so much about complex imaging distinctions between COPD and non-CF bronchiectasis, but they have a medicine that is beneficial and effective across both.
Got it. I guess when are we most likely to see the first IL-33 data from GSK?
We had a study in patients with COPD that had a phase III go decision gated off of very predictive pharmacodynamic biomarkers. The benefits of these diseases is that you can have high predictive properties off of phase I or phase II biomarkers, in this instance, in the patient population. We have now initiated discussions with regulators for phase III starts next year. We will share those data. I think that the data that is more exciting in terms of a novel proof of concept is in non-CF bronchiectasis, and that phase II data will be available by the end of next year.
Good stuff. Combination therapies, combination opportunities. I think there is kind of wide-ranging potential here, not just with the TSLP, but potentially beyond that. Can you talk a little bit about the opportunities that you see there?
Yeah. So both in COPD, but also in MASH and pulmonary hypertension, these multifactorial, complex, heterogeneous diseases are going to require polypharmacy and combination approaches. So we're anchoring very heavily on combination as part of our next wave and how I've been approaching discovery and translation with dedicated efforts around what novel types of predictive methodologies can you provide assurance that when you combine two things, you get additivity, ideally synergy, and absolutely not redundancy. That's really important if you take respiratory as an example, where if you're just combining things that are kind of overlapping to a large extent, you're exposing patients to additional pharmacology without additional benefit.
So we said, "Look, we don't want to do this through trial and error and just empirically find our way through with Brownian motion what ends up being successful." We want to have a data-driven approach, and we benefit from having what I would argue is the largest amount of respiratory data anywhere in the world. So we systematically looked at all available combination approaches today. What we found is that when you combine certain things, you get true redundancy. You get no additivity. One approach which we undertook, for example, was to look at patients that have genetic loss of function of target X, and in rare instances, they may also have genetic loss of function of target Y.
Do you see any difference when you have the combined genetic perturbation? The mechanism combination that stood out that we disclosed a couple of years ago, maybe we shouldn't have done because I think it led to a flurry of biotech activity, and this, of course, was in advance of the IL-33 monotherapy data from peers, was TSLP/IL-33, which is completely plausible that those two mechanisms, master cytokine, TSLP, will inhibit IL-4, IL- 13, IL- 5, and then IL-33, which I just described, an alarmin inhibitor working through the biology of epithelial injury, endothelial dysfunction. But we were able to quantitatively predict that that can offer synergy. What's really exciting is that we have a bispecific solution in our portfolio, but we didn't want to wait for that to be ready for phase II testing.
We've just initiated, actually, in the last few weeks, a study that's looking at co-administration of our two monospecifics in patients with COPD to try and demonstrate that evidence base of additivity or synergy. If successful, we can then leapfrog the bispecific into late-stage development.
Hmm. Interesting. Food allergy. The company's definitely put a stake in the ground and put GSK's money where your mouth is. Maybe you can give us your thoughts on the food allergy opportunity and your enthusiasm there.
Ozureprubart is our long-acting IgG monoclonal antibody, YTE modified, which should enable three-monthly dosing. It could end up being the poster child of what a long-acting monoclonal can offer, because if you think about the disease, the adolescent population, children who really don't want to have a huge burden of managing their food allergy, I think those properties become incredibly important. We wanted to keep it simple in some respects, so it's the same epitope as XOLAIR. That's really important because we've seen actually more potent ligand-neutralizing approaches to XOLAIR that targeted a different epitope in the past that didn't translate to efficacy. More potent, long-acting, same epitope, piggybacking off of that early demonstration of what's achievable with XOLAIR, but improving the profile in two key ways.
One is that about a quarter of the population is ineligible for XOLAIR for food allergy based on either their weight or very high IgE levels. This will overcome that, so it will provide coverage for everyone. For those that are currently eligible for XOLAIR but need to be dosed every two weeks, this will be dosed every three months. We initiated a phase II study that enrolled very quickly, which is always a good sign. Based off of the sample size that's already enrolled, we'll be able to evaluate end of phase II next year and start a phase III study. The plan is to simultaneously then start a phase III pivotal program, both food allergy and spontaneous urticaria, where we've got phase II data from the partner, which was really exciting.
For us, when we were doing the deal, we just wanted to make sure that it had comparable efficacy to XOLAIR, but actually, that phase II had an active arm that had XOLAIR in, and the UAS7 data had a greater benefit than XOLAIR. It's a program actually pretty economical in terms of the upfront, rapid phase II to show us what we needed for dose selection, and next year, we'll start, hopefully, two pivotal indications.
Excellent. Maybe just to wrap us up, we've always too little time with companies. What are you most excited about to kind of see in 2027 from the remaining programs in particular?
I'll just mention it because it's a little bit under the radar because we've already spoken about IL-33 phase III starts, NCFB IL-33. We're obviously in phase III delivery phase for efimosfermin, which we didn't speak about, which is a major program in our portfolio. But we also have a phase II study in alcohol-related liver disease with an siRNA called gatuzosiran, which targets HSD17B13, which is one of the high genetic confidence targets for steatotic liver disease, in particular alcohol-related liver disease. It's a novel target. Of course, it's going to bring risk, which is why it's not on the radar of, I think, the investors. But the study is exciting both to understand the potential of gatuzosiran to advance into phase III in alcohol-related liver disease, which is the primary reason for liver transplantation in—
Right.
—developed countries. But it also will inform efimosfermin, because efimosfermin also has a clear rationale for alcohol-related liver disease. The FGF21 class really stands apart in terms of its benefits on fibrosis improvement. It also has really great effects on metabolism and steatosis. And alcohol-related liver disease is a steatotic liver disease pathology that's principally driven by fibrosis rather than carbohydrate excess and steatosis.
Yeah.
We've got this ability where we can learn from what I think is the first dedicated serious sponsor phase II study in alcohol-related liver disease. We've pioneered that. We've spoken to the FDA about what the trial design looks like, what does a phase III look like. It could inform a phase III start for gatuzosiran in an area with enormous unmet need. I think this could be a mega blockbuster arena as an indication, or it could inform how we approach efimosfermin, which has just started phase II as well, in the ALLSTAR study, or it could inform a combination thereof between the two mechanisms. So alcohol-related liver disease, we've got fantastic amount of talents in the organization with some really leading thought leaders in that space, and we've got the right substrates in late-stage development.
Super interesting. Definitely not one that we had on our radar in a meaningful way. Well, Kaivan, thank you so much for joining us. Thanks to GSK for joining us as well. And yeah, hope you have a great conference. Thanks so much.
Pleasure. Thank you very much.
Thank you.