Hey everyone, and welcome to the Immunology Breakout Session. My name is Victor Bultó, I'm the Novartis U.S. President, and I'll be your host today. I'm here with the team, with Ingrid Zhang, who is our Chief Commercial Officer for the International Regions. Thank you, Ingrid. That's my pleasure. With Angelika Jahreis, she's our Development Head for Immunology. Thank you, Angelika, as well. And with Richard Siegel. Richard is our Head of Immunology Research, so here you have the entire RDC continuum ready to answer your questions. Before we do that, I just wanted to stress that this has been a phenomenal year for the immunology team here at Novartis, across a number of dimensions. The first one, with the continued solid performance of Cosentyx, which reinforces our commitment to that $8 billion peak sales guidance. As you all know, we had positive phase III results in PMR, both across primary endpoint and all the secondary endpoints. It's an area of high unmet need, and we're very eager to be submitting registration early next year. Then we had the fantastic news of the FDA approval for Rhapsido, or remibrutinib, in chronic spontaneous urticaria. This is the first BTK inhibitor approved in immunology. We are very, very encouraged by the early signs in the market. The feedback from both patients and physicians is very strong on the basis of a broad, clear label, very fast onset of action, and sustained response. We are very excited with the launch, and it is progressing to plan as we speak. It is very important for us because that launch is the foundation for a number of other indications Angelika's team is working on across other indications like CIndU, which is already in phase III with a readout next year, and hidradenitis suppurativa, where based on the phase II data, we saw biologic-like potential in terms of efficacy. In food allergy, where we top-lined phase II results, Angelika's team is now finalizing the phase III design. Of course, that is only in immunology space. As mentioned earlier today, in multiple sclerosis, also expecting two phase III readouts next year as well. As you can see, we have a lot of runway, potential runway with Rhapsido. Very excited about that molecule. The other piece of great news we had this year was NEPTUNUS-1 and NEPTUNUS-2, the two phase III trials in Sjögren's disease with the ianalumab readout positive. You saw the results. Angelika went through them recently in one of our calls. We are very excited to be the first potential approval in Sjögren's disease, an area with tremendous unmet need for patients, where we will be able to bring a molecule that showed placebo-like safety. That was the first-ever molecule to actually demonstrate efficacy in SDAI and significant trends across a number of secondary endpoints. We're also equally excited, like in Rhapsido, where we see a pipeline and appeal with the ianalumab. As you can see, we went very fast with a number of subsequent indications as well, with phase III trials ongoing in lupus and lupus nephritis with readouts in 2027. There is also a phase II ongoing in systemic sclerosis as well with a 2027 readout. These are the immunology indications, also working on some hematological indications across ITP and wAIHA. All, by the way, both in Rhapsido and ianalumab in areas we know well. We know how to develop the drugs. We have commercial presence as we speak, and therefore substantial future commercial synergies as well. Last but not least, Richard's team is working with Angelika on our CAR-T, YTB323, CD19 in autoimmune reset. There we saw very encouraging data in our lupus trials. That gave us the confidence to start four parallel phase II potentially registrational trials in lupus, systemic sclerosis, myositis, and ANCA vasculitis, with a potential there of an immune reset across a number of different immunological indications. Very excited about that as well. All in all, we are expecting about 13 pivotal readouts before 2030 and three phase II readouts, which gives us tremendous confidence as we continue to build this portfolio in immunology. With that, we'll be very happy to take your questions. Yes, please. Thank you. Can I start with the ianalumab? The involvement of B cells in the indication. In terms of. Oh, sorry. Can you start again? Sorry. My bad. Michael Leuchten from Jefferies. Starting with the ianalumab, the B cell involvement in the different settings, if we sort of go from toughest to most obvious, where does Sjögren's sit and then how does that read into lupus and then SSc? I guess what I'm asking is how do you risk all the other indications? Yeah, let me get started on that with respect to the clinical data, maybe. Then I'll pass it on, Richard, maybe to augment on the scientific data. Sjögren's disease, we've been trying to find therapies, and when I say we, then I mean the larger pharmaceutical companies, biotech, as well as academia, for over 20 years. I really want to stress that these are the first pivotal trials that have read out positive, and they replicate pivotal trials, adequate and well-controlled, that have met the primary endpoint and therefore proven effectiveness of ianalumab in Sjögren's disease. Sjögren's disease certainly has been proven to be a difficult-to-treat disease. What we have seen is that it is a very prototypical B cell-driven disease, similar to lupus, lupus nephritis, where we see infiltration of B cells in the tissue. What is particularly interesting is that we have been doing a mechanistic study, and Richard, I invite you to talk about that a little bit better, where we've seen that we target these B cells specifically. That has led not only to statistically significant differences in the primary endpoint in Sjögren's, but also to a very consistent response across patient, as well as physician-reported outcomes in both of these trials that have shown clinically meaningfulness of that response and have allowed us to really, and I've shown these data in our investor call, to have also an increase in cell barrier flow in those patients who we believe still have a reserve for cell barrier flow. Now, looking at lupus and lupus nephritis, where other B cell therapeutics have already been successful, but also only if you deplete tissue-standing B cells. We have seen in our phase II study in lupus an SRI-4 difference of about 35%-40%. That is at week 24, whereas most lupus studies read out at week 52 or later. That is a really impressive result. We have also been able to reduce the flare in this patient population and half the flares that patients typically see. I think, in summary, our Sjögren's data really do give us and increase the confidence for lupus, lupus nephritis. As well, there are interesting data and important data of B cell involvement in systemic sclerosis. For example, Rituxan is approved in Japan. It is the only country it is approved. That gives you some indication that also in systemic sclerosis, B cell pathology is important. I certainly see it has increased, and I would be particularly very excited about the prospect in lupus because we have tissue-dependent B cell depletion there. In addition, we inhibit through the mechanism of action also BAFF as the growth cytokine for B cells. We have a dual mechanism of action you could think of overlaying to some extent several mechanisms here that have already shown efficacy, and we expect much larger efficacy thereby in lupus, lupus nephritis. Maybe Richard, you want to add? Sure. No, thanks. I think you summarized it really well. In some ways, we were the pioneer in Sjögren's in that we didn't have as much prior evidence from approved drugs and B cell mechanisms in Sjögren's than we do in lupus. I do want to draw everybody's attention to, along with the phase III results, these two mechanistic studies that were reported at ACR, the ultrasound study that demonstrated differences in salivary gland pathology. I think even more significantly, the biopsy study where we showed 85% depletion of B cells in the target tissue in the salivary gland. Those 15% of B cells that were left, we demonstrated transcriptional reduction of BAFF-dependent signaling. That really brings to life the second MOA. A reason for it to be there is that we're not depleting 100% of tissue B cells the way that CAR-T therapy does. We think that's an important reason to believe that ianalumab will be differentiated from a standard B cell depleter CD20-based. There's also a difference between CD20 and BAFF receptor expression. BAFF receptor is expressed farther along B cell differentiation to the plasmablast lineage closer to CD19. Important reasons to believe and the strong phase II results that actually lupus and lupus nephritis will be positively differentiated. Yeah, probably all. Thank you very much. Maybe just a first question on remibrutinib development program. You started the SPMS study. Can you tell us a little bit about how you're thinking about the potential primary progressive study, confirming that you're focusing on disability progression? Any sort of trial consideration that you can tell us about based on your learnings from competitors' readouts? Then just a second one on food allergy. You already have the phase II, so we are waiting for the data, but we know you have positive data in peanut allergy. Can you just talk about the probability of success when moving to phase III in food allergy? Because conceptually, it looks quite high if you already deliver data in peanut. Thank you. Thibault, on the multiple sclerosis indication, I will invite you to ask the question to the neuroscience team. What I can tell you is that what the team is right now doing is trying to consider what is it that we could include or how could we shape that secondary progressive MS trial to also reinforce evidence on primary progressive progression. The team will have a better answer for you on the neuroscience section. Angelika, do you want to take the question on food allergy? Absolutely. Thank you. Remibrutinib, we are developing it across CSU in usable forms of urticaria. We also have very interesting data that we have presented in hidradenitis suppurativa. Food allergy, we are entering a phase III study. We will be soon sharing the data at AAA AI, the phase II study that we have conducted. What I do think is very interesting for food allergy is the fast onset of action that many patients with food allergy will want. You will not only see that this remibrutinib is showing a fast onset of action, but also very good efficacy. It has biologic-like efficacy. We also have fast onset of action. Victor, you already talked about the really clean label that we have seen for remibrutinib now in the U.S., which is, I think, also really important for patients with food allergy who are looking for a safe product to take as a first-line therapy for food allergy to mitigate the potential risk of accidental exposure to a food allergen. We agree with you that the likelihood of success, also based on the biology and similar pathways being involved and food allergy being triggered through same histamine release in mast cells, is high. Maybe, Richard, if you want to add something on that aspect. Yeah. We'll be presenting the data, but you can see that we're very excited about it and that we're initiating a phase III study in food allergy. If you want a preview of what it's likely to look like, you can look at the acalabrutinib academic repurposing study that was presented, very small number of patients, but very strong effect size, I think 70%-80% had significant protection from food-induced symptoms and rapid onset of action. We know just mechanistically that even from phase I studies, remibrutinib, within a day of starting the drug, you can see total blockade of basophil activation, which is the blood-based equivalent of mast cells. I think something to really look forward to at the AAA AI and, like I said, a lot of excitement to share with Angelika's group. Great. Florent? Good morning, Florent Cespedes from Bernstein. Thank you very much for taking my questions. Two quick ones. First, to come back on Rhapsido maybe could you share with us a little bit of the opportunities of the different indications, the size of the markets, the populations, because, of course, food allergy is a huge unmet medical need. HS is more competitive with drugs already approved. Some color on that would be great. My second question is on Cosentyx. Could you maybe remind us what you have in mind kind of to some life cycle management to, let's say, protect the franchise beyond IRA and potentially biosimilars as well? Thank you. Thank you. Thank you very much. Look, on the Rhapsido side, we've quantified the population in CSU that can benefit from this therapy in about 415,000 patients in the U.S. alone. Of course, we would add there the international patients. I mean, just to give you a sense, there's about 1.1 million patients with CSU in the U.S. It's a very large indication, about half the size of psoriasis. Out of these 415,000 patients, what we know, sorry, out of the 1.1 million, 400,000 are not well controlled with the current antihistamine treatment. That is despite very high doses of antihistamines. The idea here, having a clean broad label and oral therapy that actually acts very fast, would be to take that step right when patients need to escalate after antihistamine. There is only about 20% of them who have escalated to a biologic for a number of reasons. We clearly aim at having that position after antihistamines. Based on what we are hearing initially, that is a very realistic goal right now. That is the idea with clearly two biologics in the space that take substantially more time to start having an effect. That is for CSU. In CIndU, there is currently no biologic approved. One of the biologics was never studied. The other one failed a phase III trial. Right now in CIndU we have a wide open space. The overall population is as well around 400,000 patients in the U.S. A similar market there when it comes to CSU. I think the HS market, I know there is a debate on the size of the market. We continue to see that market at about $3 billion-$5 billion market. It would be very interesting to bring an oral therapy into the space, now provided we see a replication of the phase II data where we did see biologic-like efficacy. It is a more competitive space, but we do know in all of these areas, there are two benefits to it. One is the oral component of it, which some patients prefer. The second one is in highly symptomatic diseases, we know fast, rapid onset of action is always a benefit. When it comes to food allergy, look, on the surface, this is a massive market if you think about food allergies, particularly with the potential that BTKis have to actually be allergen agnostic given how downstream they operate. We know there's more than 3 million patients with severe food allergy across the main G6 countries. We have to start understanding, and we have to start the work there, what are the patients that would benefit the most and when. We already know, for example, that Xolair is off to a phenomenal start with food allergy. We are very excited about that potential. When it comes to Cosentyx, the main life cycle management right now, of course, will continue to grow. We see HS will continue to be a key growth driver. It's a market that is growing at double digit. That will continue to grow double digit because today we only have about 20% of the patients properly diagnosed and treated. It's a market that in the U.S., in the international markets, will continue to grow. We believe as well that Cosentyx and PMR will be another strong life cycle management for Cosentyx there before we see IRA in 2028 and LoE in 2029. Sorry, you want to. Yeah, maybe I can just build on because Rhapsido is already approved in the U.S., and we're looking at China approval any time and followed by Europe and Japan in early 2026. That is all coming. I'm personally super excited about CSU because it's a huge market, that equivalent number of diagnosed and not controlled by antihistamines is about 4 million. Those are biologics less than 10%. Xolair, CSU alone internationally is almost $1 billion already opportunity. We are really looking at internationally contribute strongly to that multi-billion estimate forecast. Certainly in HS because the healthcare system is very constrained from a capacity perspective in Europe. Having oral medicine with rapid onset, great efficacy and clean safety is going to be super valuable. I'll just stop here. On the Cosentyx part, I mean, I think we're confident to get to $8 billion by the 2029 time frame. In Europe, actually, the compound pattern is a bit later. We are also looking forward to that. Thank you. Matthew? Thank you, Victor. It's Matthew Weston from UBS. Two, please. The first on Rhapsido, it's obviously an important driver based on Vas's comments this morning for 2026 return to growth in the second half of the year and then going into 2027. I'd just love to understand if you can help us on launch cadence in terms of particularly that look through 2026. When should we expect free drug to go to paid drug? What are the key drivers of that that we can look for? We were promised that with Sotyktu. I realize that's not you, by another competitor in the space, and it never materialized. Why do you think you're going to have an access advantage to really drive the revenue growth for Rhapsido? Yeah. Then a second question, if I can, just on YTB323. A lot of focus on using CAR-T and immunology from lots of players. As a regulator, you're clearly having a massive impact on someone's immune system. Clearly, at the very end of the therapeutic spectrum, I get it. If we're going to bring it more mainstream, is there any way that we can look at more modest immunoablation in immunology, or is that just totally counteracts the mode of action of what we're trying to achieve? Great. I'll take Rhapsido, and then Richard will take the YTB323 question. Matthew, a couple of differences here. One, we are extremely confident on the clinical profile of the drug. If you want to make sure that you're going to get access and pull it through, the first condition is that you have a strong appetite from both patients and physicians to utilize the drug. That's what we're seeing in the first weeks of this launch. That's important. What we're doing is two things. We're not only utilizing free drug, we're utilizing samples as well. When you have a highly symptomatic disease where patients are exceptionally miserable with these antihistamines, there's a tremendous appetite for these patients for a rapid response. The main determinant of the urgency is lack of sleep due to the itch. It's a very pressing disease. What we're seeing right now is a strong appetite from physicians to use these samples and get patients started right away. What Richard noted in terms of the speed of action, it's actually the anecdotal reports we get right now is that some patients call back day one, actually already referring substantial symptom relief. I don't want to generalize that, but these are some of the reports we are hearing. Matthew, our free drug program requires a prior authorization denial. That means the physician will have to send a prior authorization. If they're denied, then these patients will get on free drug. That's important for two reasons. Our Head of Market Access for the U.S. is sitting here in the room, Robert Rubinsky. The more pressure payers will see, the more utilization and prior authorizations they'll see coming their way, the more urgency there'll be to negotiate and get access for this medicine. Right now, we are confident that through the first half of next year, we will have a cadence of payers coming down with access. At that point in time, the first priority will be to make sure that those patients on those free drug programs, which we have identified, will be pulled through into paid fields. That, of course, will pull through that access in the normal demand because our free drug program only kicks in if there's a prior authorization denial. That is what gives us confidence, Matthew, that as we start getting that access, we will be able to pull it through. Of course, we do not want to rush it either because we are very mindful about the level of discounting that that will take. We will keep you updated on that cadence as we go. I will reinforce that the demand, the underlying demand that we are seeing is there and that the satisfaction is there from patients. That gives us the confidence that Rob and the team will get the access and we will pull it through. Maybe I can add one sentence before we go to Richard is that I think the comparison with psoriasis is also not quite adequate in my view here because psoriasis, many, many therapeutic options are available to patients. Whereas where we are with CSU right now is in a patient population that have very few treatment options. I do think the unmet medical need is much, much higher in CSU than it was at the time of [TYK2] entry into a very saturated psoriasis market. Absolutely. Yeah. Yeah. Thanks for your question on YTB323 and B cell immune reset. Although you can argue that the immune reset concept is a dramatic reset of the immune system, the idea is to have new generation of B cells from precursors and really change the immune repertoire. In terms of area under the curve of immunosuppression and immunosafety, you could argue that this type of therapy is actually safer than staying on chronic immunosuppression, which is the standard of care. What we're seeing, and really the reason this is kind of a revolution in immunology therapeutic paradigm, patients recover B cells 60-90 days, which is totally different. Please do not think about the original Kymriah and all the original CAR-T and ALL, which is almost chronic B cell aplasia. That does not happen even in DLBCL and in autoimmunity, all of us. There is some nice data from ACR. We're essentially all seeing the same thing as the original reports from [Erlong] and others of this 60-90-day repopulation, which we think is important. If you hear about two weeks of B cell aplasia, we don't think that may be sufficient. That level of B cell aplasia, that short term, it's the depth that we think differentiates from even our own ianalumab, which we know is not 100%. We think from an immunosafety point of view that it's actually a reasonable proposition for a severe refractory patient. The conditioning that's used, fludarabine and Cytoxan, is an issue, but it's very transient. Just to remind everybody that before 2010, Cytoxan, given in higher doses for a much longer period of time, was the standard of care for lupus nephritis. Every rheumatologist knows how to give Cytoxan. I think that is not going to be a barrier to rheumatologists adopting this therapy as their own if it gets to the market. Great. Thank you, Simon Baker from Rothschild & Co Redburn. I've also got a question on YTB323, but from a slightly different angle. As Vas said, and as we can see from the slide, this is potentially a pipeline in a product. It is a slightly harder product to manufacture and scale than most. Where do you see the potential scalability of this? How is manufacturing moving on going forward, both in terms of the logistical ability to scale out, but also from a cost of goods perspective? Another slightly broader question. How sensitive is your growth outlook in immunology to the current structure of the U.S. market, particularly the PBMs? We hear a lot about, and we talk a lot about the importance of rebate walls when you've got multiple indications for the same products. Are we overstating that, or is the current structure critical to growth? Do you envisage any change to the PBM model in the U.S. over the coming years? Thank you. Great. Thank you. You want to take on YTB323 manufacturing and Angelika, the difference in the process and why that makes a difference as well. Then we will touch on the logistical standpoint. I think there are a few very important considerations. Our manufacturing time for YTB323 is only two days, and we aim for a 14-day window from the vein-to-door time as we manufacture the process. I think also when you look at autoimmune diseases that are typically chronic diseases, the short timeline is not quite as critical as it is in oncology. I can tell you that we have deep experience, and that's what sets us apart at Novartis. We have deep expertise and experience in manufacturing, and we have to date been able to obviously for the clinical trials, but we'll also be ready for the launch to then produce YTB323 on a scalable level. We have that. We have shown that with Kymriah. I think we have an Ingrid, you can talk about the numbers of patients we have treated with Kymriah, and we will take all the learnings. Now, let me go also from the question about manufacturing maybe to the clinical trials. They are progressing very well. We have, as Vas has said, started our four pre-bottle trials that we have discussed with health authorities and aligned with them on how to bring this very novel and transformative therapy to patients with, for right now, lupus, lupus nephritis, systemic sclerosis. I like to highlight physicians are very comfortable with that modality because they use stem cell therapy for patients who are refractory, as well as vasculitis and myositis. Richard and his team is also looking currently at the earlier indications with RA and Sjögren's disease. Maybe you want to add a few words, Ingrid, around Kymriah and our ability on the CMC side. Yeah. International, of course, there's a U.S. number. Internationally, we actually serve Kymriah in over 400 centers across 30+ countries. Cumulatively, we're talking about probably tens of thousands of patients. That's a number we're looking at. I would also add Novartis being a very global company, we have a really great strength of being able to manufacture complex modalities with speed, quality, and scale. I think the other extreme, you talk about oncology, is RLT. Vas mentioned 99.9% on-type delivery at a very narrow window. In fact, right now, internationally, we place the order, we guarantee delivery within two weeks, and we're trying to further shrink that time in a very, in some ways, much more difficult given the time frame. Certainly with Zolgensma, you think about, again, a very complex, and we're serving internationally in well over 40 markets. Yeah. Victor? On the PBM landscape, obviously, there's a number of initiatives right now that are forcing or putting pressure on the system as it exists, starting with more transparency on the gross-to-net barriers that are constructed. I don't think that's going to change in the short term, but all these pressures will put some pressure on the system to evolve over time. Of course, that will change again over time. I don't know if it's two years from now or five years from now. We've all said the PBM system will change much sooner than it did in the past. What that means is if you look at our strategy in immunology, already about three years ago, we decided we were going to go into the areas where there's substantial unmet need and where the areas are not that crowded. If you look into Sjögren's disease, CSU, CIndU, if you look into lupus, systemic sclerosis, that's precisely why we wanted to go to areas where clinical differentiation and being one of the first to market or one of the first to markets was going to make a difference. I think that positions us well to also be successful where that changed, where to change. In the meanwhile, of course, it's a benefit for us to have volume in immunology space as we negotiate these new indications, but we are ready to pivot if that changes substantially in the coming years on the basis of this portfolio. Actually, CAR-T is another great example of that, where it will be a medicine that will be quite agnostic to what happens with the PBM space as well. I wanted to point out as well on the CAR-T space, as we try to size this opportunity, there are three factors that we are considering. One is in how many indications does this work? As you know, as you see, and we are being very broad in our approach where we believe this space can work. Of course, the more indications we have, the more volume we have, the more synergies, and the lower the manufacturing costs, so that is an important consideration. The other one you did ask is lymphodepletion and conditioning. The more comfortable the space is going to get, the more comfortable rheumatologists will get. I know we know in China, rheumatologists are conditioning themselves, and they do not have a problem with that. Will that be widespread as we get approvals for these medicines? The third one is what other alternative mechanisms actually emerge, if T cell engages work in these later patients or not. The more this is a differentiated approach and the more we're comfortable with the lymphodepletion, the more this will go into earlier patients given the promise of an immune reset, which is really something very attractive for these patients. These are the factors that we will be considering and will be fully ready from a manufacturing and from a logistics standpoint for that. Yeah. Maybe I'll just add one quick thing there back to maybe Matthew's question also. We're in a very, I think, excellent position. We have CAR-T in multiple indications. You can see also on our pipeline, we have PIT565, a CD19 T cell engager with some interesting properties that might increase efficacy in lupus and RA in early trials. We have ianalumab and remibrutinib. If you go all the way from B cell modulation to immune reset, we have the entire landscape in our pipeline. We want to bring the best therapy to patients and hit the sweet spot between complexity and efficacy. We're in the position to kind of do that in an almost mechanistic agnostic way. I think we have a good position to find the best therapy. We are in research looking, of course, for CAR-T in a pill, but we'll let you know when we get there. Looking forward to that. Thanks for taking my questions. It's Naresh Chouhan from Intron Health. Two questions. One on Cosentyx. Clearly, given the growth you've seen in HS, we've had strong volume and price. How do we think about price mix going forward as kind of HS slows as part of the growth? For Rhapsido and CSU, given the rapid onset of action, do you expect patients to cycle on and off? How are you thinking about stata? Yeah. Do you want to take the Rhapsid o cycle? Yeah. The nature of the CSU disease and the progression? Yeah. I can start out with that. In our clinical trials, we have looked at the disease. The patients that we involved in our phase III trials, for example, all had long-standing disease. Sometimes you think about CSU as being only a short-term intermittent disease. For many patients, this is truly a chronic disease that spans over many, many years. We have seen many patients in our clinical trials who had more than five years of disease duration. This means daily hives. The patients we enrolled in our trials had 50 hives. Now imagine you have one bee sting or one mosquito bite, how that bothers you, but these patients have 50 per day. That has a huge impact, as Victor has said, on the ability to sleep, the quality of life, the ability to keep a job, the ability to even parent with a disease like this. Why am I highlighting this? I just want to ensure that you understand that CSU is, in some patients, a waxing and waning disease. In many patients, it's a very chronic disease. When we have, after our phase II- B, we have stopped the treatment and looked at the relapse rate, and then we started treatment. There are two things I wanted to highlight. One is that when you stop treatment within four weeks, a third of the patients had relapses of the disease and were really happily going back onto treatment. When we retreated, we had the same efficacy again. Because it's an oral drug, there is not a risk for neutralizing antibodies. Patients recovered their response based upon retreatment. Certainly, there may be patients who want to, after a year or two, stop treatment and see if they still have the disease or if they relapse. It's easy for them to go back on, and there's a huge motivation in patients based on the burden of the disease to go back onto therapy. On Cosentyx HS, look, what we continue to see is the market growing at about 14%-15%. I do presume that that's what's going to happen in the coming years, given the amount of patients that are currently not treated or not even diagnosed. We have efforts to continue to grow that market, but these patients are quite stigmatized. Many of them don't even show up to the doctor. It's going to take time for us to activate these patients. We do expect that as new competitors come, they will hopefully invest as well into that development, and we'll see the market continue to grow. In the meanwhile, right now, in a market that is growing at that pace, we have a 50% NBRx share. That means one in two patients are either start or switched to Cosentyx. We do believe that that will give us a good growth profile for Cosentyx in HS. From a pricing perspective, we have very broad access now with Cosentyx and HS, predominantly in earlier lines, which is where the majority of these patients come when you get longer durations of therapy. We do not expect that to change in the short term as well. That is why we are confident on HS continuing to be a growth driver for Cosentyx. Good. Matthew? Then Steve after that. Do you want to take Steve first? No, no. You have the mic. Okay. It's Matthew Weston from UBS again. Two, please, if I can. Just one quick follow-up on Cosentyx HS. The efficacy for all of the HS drugs probably isn't as good as everybody would like based on some of the other immunology agents, which probably means that there's a lot of patients who are dissatisfied and then cycle off. You've put all that effort in. You've already said it's difficult to find the patients. You put all the effort in to get them. Is there anything you can do with Cosentyx to keep them on? Can you raise the dose? Can you talk to physicians about that? That means that you'll make Cosentyx patients stickier. Yeah. The second question is back to YTB323. I think Vas mentioned in his opening comments potentially registrational at phase II. Can you tell us, are you aligned with regulators on what you need to see from the phase II studies in order to file? Great. I'll take the Cosentyx question, and then we'll ask Angelika and Richard to comment on the CAR-T. Right now, we have approved for Cosentyx and HS, at least in the U.S. by FDA, two different dosing regimens, so every four weeks or every two weeks. The majority of patients will start with an every four weeks treatment. If they do not respond, they have the option to go to every two weeks. I'll ask Angelika to mention what we saw in the clinical trials. Right now, initially, that rate of patients that would escalate before discontinuing was 15%. We've been putting efforts to make sure physicians understand that that's an option for their patients before they switch out to another drug. That number is up to 25% now, Matthew. We will continue to make sure that everyone understands that that's an option that they have before they have to switch out. Because you're right, the Hidradenitis Suppurativa Clinical Response scores are about 50% versus psoriasis being 90%. Naturally, you'll see a little bit more turnover of these patients. Actually, we have some data from the clinical trials as well. Maybe you can comment. Yeah. Yes. Certainly, I mean, the data that you've cited are the primary outcomes of 50% high score data. If we look longer term on patients, how are they performing? We have presented our data out to four years. The efficacy actually increases over time. What is particularly important for patients with HS is the pain and the flares that they experience. Many of these young women who come to the physician with hidradenitis suppurativa and their flares, they are in parallel to their menses. We have seen about a 65% reduction in the pain score of those patients who came and entered our trials with moderate to severe pain, and then they experience mild or no pain at week 52. In addition, the flares, and I think that is particularly important, 70% reduction in flare rates over one year. We see that persistence over time. I agree with you, we are not there where we are with psoriasis yet. That is why we have very active research and development efforts ongoing in HS. We are excited. We talked before about the remibrutinib data. Remibrutinib as an oral therapy, fast onset, very safe for these patients with hidradenitis suppurativa who often are young women, overweight smokers. For these, a twice-a-day pill will be a really good treatment option if we can replicate the biologic-like efficacy in the next study. Thank you. Maybe I start talking. Just answer that briefly. I want to make sure we get to Steve's question. Yes. I'll be brief. With respect to YTB323, yes, we have had discussions for all the pivotal trials with the respective health authorities, and we have aligned with FDA on the readouts and the data that is necessary for us that we need to provide for submission if the data are transformative, which we expect that they will be. The current studies are well aligned with that. We expect the first data that we could submit by 2027. Thank you. Steve? We'll take the last question. Thank you very much, Steve Scala from TD Cowen. Two questions. On the first one, I could be wrong. My recollection is that Novartis had a Cosentyx orally delivered formulation and developed many years ago. Is that correct? If so, why was it dropped? I assume bioavailability. Do you continue to pursue that? Also, do you have late-listed patents that you plan to unearth similar to the brilliant strategy you had on Gilenya many years ago? On ianalumab, I think an unanswered question is the subgroups. Now, it's been weeks since the presentation, so I assume Novartis has visibility on the subgroups. What can you share with us at this time? Thank you. Do you want to take the oral? Cosentyx and a pill? Cosentyx pill. I think oral equivalents of biologics have been a goal across the field. There were programs in the past. You've seen how difficult it is across the industry to develop a truly small low molecular weight IL-17 inhibitor. I'm sure you've been following that very, very challenging pharmacologically. Now that we've seen with IL-23 cyclic peptide inhibitors success, I think that's going to probably be highly successful. It's very difficult now to sort of think about entering that space way behind that. I think the cyclic peptide approach, just from a drug discovery development point of view, looks really exciting. If we started a program today, we'd probably be going with that rather than a standard low molecular weight approach. Yeah. Do you want to take a question? I can close by, or I can answer the question about ianalumab Yes, we do have the data in-house, and we are obviously looking at these data with great detail. We plan to submit to the health authorities in the first and second quarter next year. Early first quarter, we expect to submit the data. We have seen consistent trends over most of these SDAI domains. Great. With that, we'll conclude the immunology session. Thank you very much, everyone, for your attention, and have a great rest of the day. Here in the room and as well on the webcast, welcome to the cardiorenal metabolic session. My name is Patrick Horber, and I'm the President for International. I'm here with three colleagues. Starting on the left side with Ruchira Glaser, who is our Business Unit Development Unit Head for CRM. Then Dianne Auclair, who is our U.S. Therapeutic Area Head for CRM. Finally, Shaun Coughlin, who is our Global Head for Biomedical Research as well for CRM. One year later, after our meeting last year, I think we made significant progress across our three strategic pillars: ASCVD, atrial fibrillation, arrhythmia, and as well on renal. If we look into ASCVD, we're developing a number of assets which are targeting different risk factors like Lp(a) and LDL-C, and as well inflammation. In Lp(a), we're progressing with pelacarsen, and we're expecting a data readout of our phase III outcome trial in 2026. In LDL-C, we have Leqvio, which is now a blockbuster. We are continuing to build evidence with Leqvio to support the use in different settings, including as well two outcome trials, ORION-4 and VICTORION-2- PREVENT, which will have the data readout in 2027. Finally, inflammation, where we acquired Tourmaline this year with an IL-6 inhibitor, where we'll see more data of our phase II trial by end of this year. Going to atrial fibrillation, we have abelacimab, which is our lead compound, which is targeting stroke prevention. It provides a benefit-risk profile, which serves patients who are currently undertreated because of a high risk of bleeding with current available medicines. At the same time, we're looking as well broader there into antiarrhythmic molecules with different MOAs, which all of them are in very, very early development. Finally, in renal, we have Fabhalta, which is our anchor drug, which has approvals now in C3G and as well in IgAN. We're developing Fabhalta in four additional renal indications. If we talk specifically about IgAN, we have a product portfolio there with three assets, Fabhalta and Vanrafia, which are approved in the United States and as well in China. The third one is zigakibart, which is currently in phase III development and will have the data readout in 2026. Finally, the acquisition of Regulus Therapeutics, where we got into our pipeline farabursen, which is an oligonucleotide for the treatment of polycystic kidney disease. It's an autosomal dominant disease where we see here the potential for better treatment than the current standard of care. Overall, if we look back these 12 months, I think we have made significant progress and strengthened our pipeline through collaborations and as well acquisitions. If I look from here to 2030, we have seven phase III data readouts, which will definitely be catalyzers in CRM. With that, I would propose that we go into the Q&A. I just saw here in the first row. Thank you, Patrick. It's Matthew Weston from UBS. Two questions, if I can. The first on pelacarsen. And it's about trying to--I'm old enough that I remember the Diovan launch, the Leqvio launch, the Entresto launch. It took five years to get to a billion dollars, but then was a billion dollars a year additionally every year. If I look at pelacarsen, you have kind of two options. You go high Lp(a), normal cardiometabolic price. It probably takes you five years to break even, but a very big area under the curve. Or you take it to kind of ultra high Lp(a) on the cutoff with a much higher specialty cardiology price, maybe the same NPV, but you break even much quicker. I'd just love to understand, is that a reasonable assumption, or you're just going to go for everyone and see what the data brings? I will start with the U.S. Maybe Dianne, you can start, and then I will add on on the international side. I'm old enough to remember all of those launches too. The good thing is when we're old enough to remember them, we can learn from them. I think the reality is cardiovascular launches do have a slower ramp. That is something that we should accept. What I think will be different with pelacarsen is that we do have the ability to leverage the learnings that we have from most recently Leqvio in ASCVD. One of the key learnings is that as we shifted our focus with Leqvio to a post-event patient where the motivation of the patient and the provider were higher, we started to see more momentum pick up. Our intent with pelacarsen is to launch in the early year or years to a focused patient where we believe the motivation is going to be the greatest. That will be consistent with the HORIZON trial looking at patients with early events and family history. I think there's a couple of other things too that will bode differently than what we've seen in LDL-C. There is no standard of care for Lp(a). We wouldn't expect the same level of friction in the system where physicians have to step through a lower-cost generic. I think the other key thing for us to think about with Lp(a) is the genetic component of it should be a motivator. The family history is an emotional motivator for people to take action. Yeah, maybe building on that, I think we made significant experiences over the last decade with our launches. I think as Dianne said, Lp(a) is different than everything else what we had. There is no treatment available today. That's a fact. We have 20% of people who have Lp(a). There is a family history. We will focus clearly on a smaller patient population when we start, and I think that will give us a higher probability as well of success. I think we know as well our customers much better even after the launch of Leqvio. I see that definitely as a significant tailwind, which will give us the opportunity to really make a significant footprint from the beginning, which I think will be critical. Perfect. Thank you. Can I jump in with the second question, which was[abelacimab]? Why don't I pass the mic? Yeah, that would be nice. Thank you. Thank you. Just sticking with pelacarsen, please. If there's kind of a slope like we've seen in the LDL field, how does the readout of HORIZON not lower the bar for fast followers? How do you think about the competitive landscape, and how would you then think to defend that? I think there's a couple of things. One piece, as Dianne was saying, is that in HORIZON, our experience in the trial was that people, whilst the level of entry was 70 mg, people with much higher Lp(a)s were signing up for the trial. Our median Lp(a) level is 108, and 80% were over 90. I think it just signifies that there is an awareness of this genetic risk factor, and people are waiting for that first therapy. I think that the uptake early will be quite different. What we've seen to date are an acceptable safety profile in terms of no changes to our protocol from the IDMC. I think there's another opportunity there in terms of people getting onto drug early and staying on drug because they're satisfied. From a development perspective, the last thing that I would mention is that we do have in our pipeline additional drugs that are ready to go for our early portfolio in terms of long-acting. Yeah. I think the important thing here is that we do believe that we have a comfortable lead. In a market that will be crowded with cardiovascular companies, that lead is going to be important for our ability to establish the standard of care early, but also to start to establish the payer leverage. To the point that Ruchira made, we do have follow-on assets in the pipeline. Anything that we are doing to build the market benefits that too. Next question. Let's go into the fourth row. I think it was the next question. The gentleman on the left side. Oh. Sorry. Yeah, let's go on the fourth row, the gentleman on the left side. I think he raised his hand first. Sorry. Trying to. Thank you very much, Steve Scala from TD Cowen. Three questions, but they're pretty brief. Regarding the recently launched IL-6, Vas said on the Q3 call that Novartis seeks to differentiate via study design. Where can you improve on what Novo is doing? What are they not doing right? Secondly, pelacarsen event accumulation, can you tell us whether it's leading to an H1 or H2 readout? I assume H1 because the original readout was 2026. It would seem if it slipped to 2027, there are real issues. Lastly, ORION-4, somewhere along the line, it slipped from 2026 to 2027 readout. Was that also due to event accumulation? Thank you. Thanks, Steve. Maybe Ruchira? Yeah. Your first question was on IL-6. Could you repeat that one? The differentiation. Oh, differentiation in the CD? Yes. I think you said, is there something that Novo is not doing right? It's not about Novo not designing their program correctly. They have obviously a broad phase III program underway, atherosclerosis and chronic kidney disease, heart failure, and very acute myocardial infarction patients in the first 72 hours. We think that there's an opportunity here to pick a more differentiated patient population, one which combines high inflammatory burden with clinical adoption and practical use. We're looking at that design actively now with our academic collaborators. We're going to engage next year with health authorities globally after the completion of the phase II trial, which we'll complete in December. Once we do those things, we'll be able to share more details about IL-6. I know you asked about ORION-4, so I just to cover that off, that third one, which is that no, there is no event rate is as expected for ORION-4. This is a collaboration with Oxford University. They are managing the trial dynamics, and they have indicated to us that they need more operational time to do data clean out and things like that. The readout in 2027 does not affect our plan, which was to submit both ORION-4 and V2P, our sponsored secondary prevention trial, at the same time. Between the two trials, we have got over 30,000 patients. I think this gives us really robust power, particularly for cardiovascular death endpoints. We would wait for both of those trials to read out, which will come much more similar timeframe at this point. Could you remind me the second question? I didn't grab that. Pelacarsen readout in H1 or H2? If it's H2, it would seem like there's an issue because it was already too up reported. Yeah. I mean, this is an event-driven trial. We are continuing to monitor the event rates carefully. We still believe it could come in first half of 2026. I think regardless of when it comes in, our intent is to submit the file in 2026. Thank you. You, Florent? Yeah. Yeah. Good morning. Good morning. Thank you for taking my question. Florent Cespedes from Bernstein. Two quick ones, please. First, on Leqvio, could you remind us the profile of the patients on primary prevention? Because sometimes if they are a little less sick, there is less events, and sometimes the risk to, let's say, to try to target a too broad population and potentially miss the endpoints of some color, that would be great. My second question is on abelacimab. Could you maybe remind us the positioning of this drug? We understand it's for the most severe sufferers and the ones that do not respond to the existing treatment. Size of the population, potential, it's early days, but the price or the cost of this treatment, given it's not a first line, but for the most severe sufferers. Any color on this one would be great. Thank you. Okay. Let's start maybe Ruchira with the primary prevention. Yeah. For our V1P trial, which is our primary prevention trial for Leqvio, we are targeting a much broader population in this trial. What that means is patients who have all of the traditional atherosclerotic risk factors rather than singularly diabetes are included in the trial. When we look at the baseline characteristics of what that means, 70% of those patients do not have established atherosclerotic cardiovascular disease. This will be a very broad population, which I think I'm very excited about because I think it will be very clinically meaningful to show that LDL reduction really translates to longer lives and free of heart attack and stroke. Maybe I can take that abelacimab question. We see that about 40%-50% of patients today are either undertreated or untreated by the current DOACs, and that's because of drug interactions, bleed risk, contraindications. It is a sizable population that we are going to go after. In the U.S., we see about 9 million patients and growing. We think the growth rate in this space is high because not only of the aging population, but we also see the wearables playing a role in the growth rate over time. I'll go to the second row. Thank you. Thank you very much. Just one on farabursen. When Regulus was in charge of the assets in their development programs, they were planning to start a phase III in the second half of 2025. They were talking about getting an accelerated approval on the base of a kidney volume endpoint at 12 months. As you are starting to plan the trial a little bit later, do you still have the same similar strategy? Are you still looking for accelerated approval? Are you changing the phase III plan compared to Regulus' initial plans? To follow up on abelacimab, can you frame the market opportunity depending on Milvexian data in AFib? Because obviously, if it works, I think the landscape is different versus if there is no Factor XI on the market for AFib. Thank you. Maybe Ruchira? I can start with farabursen. Regulus? Yeah. First of all, I would say that I'm very excited about the Regulus acquisition for farabursen. This is our anti-miR-17 oligo. What we've seen here is, more broadly, a great fit for Novartis. I mean, on the research side, it fits with our platform strategy in xRNA. On the development side, as you point out, the accelerated approval pathway offers us a great opportunity to bring in the phase III trial with a readout potentially in 2029. On the commercial side, it fits in very well with our rare renal commercial footprint, which Dianne could speak more about. In terms of the company's original phase III plan, we looked at this thoroughly and talked extensively with them during the diligence process and obviously since the acquisition. At this point, we are talking with healthcare authorities and regulators. As you pointed out, the height-adjusted total kidney volume and imaging endpoint has been an accepted surrogate by the FDA. We would anticipate, and this is something that we saw great data on in the phase I-B study, where we saw that compared to placebo in the lowest dose, there was almost a complete stopping of growth as measured by height-adjusted total kidney volume. I think that is something that we would certainly pursue. I think was there a question about abelacimab and if Milvexian fails? We did model that scenario in our deal case. We modeled multiple scenarios. Regardless of the outcome, we see a multi-billion dollar opportunity. Maybe just to add on this, this is a positive look. We see significant opportunities in the Asian markets. I think as Dianne was saying, even if Milvexian comes, we still have a market of 40%-50% of patients which are today undertreated or even not treated, which with abelacimab, you would get in there. I think that's what we want to do. You, Matthew. I think you were waiting enough for some time. Sorry. I won't be long. Simon Baker from Rothschild & Co Redburn. Two if I may, please. Going back to pelacarsen, could you give us an idea in the HORIZON study the split between the different entry criteria in terms of previous MI, stroke, and PAD? The reason I ask that is a few KOLs we've spoken to have suggested that the observational epidemiology for the link between Lp(a) and stroke is mixed. MI, it's nailed on, but with stroke, it's mixed. I just want to get an idea of the proportions of those patients in the study. A slightly broader question on obesity when you're thinking about development opportunities. We've seen a dramatic change in the price dynamic within that space in the last few weeks. The expectation is that will lead to an equally dramatic volume shift. The whole price volume thing is a bit alien to pharma, even if it's common in every other consumer category in the world. I just want to know how this has affected your appetite for the category and what you would be looking to develop. Is this now one where niche indications within this huge market are more favored or less favored? Any thoughts would be really useful. Thank you. Okay. Thanks for the question. Maybe let's start with the pelacarsen patient population. I think taking a step back for the reason for your question, the data show in general for stroke across different risk factors that it's just a more heterogeneous endpoint. Sometimes you don't see the translation as crisply in epidemiologic studies. I think the randomized trials where you do the intervention, if you're hitting the right biology, you should actually see an improvement in stroke. Our primary endpoint is a composite that includes stroke. For us to show a benefit in the composite, we want to also enrich for patients who have stroke. I think that the two go hand in hand. It's always going to be a little bit of a minority of the endpoint and patient population that's brought in. By the way, the stroke history is predictive not just for stroke, but for cardiovascular events in general. I would not limit that translation only to stroke. I think the other thing to point out is that that is very different to the approach that has been taken for Amgen, where they are looking at a more coronary heart disease endpoint. Our key secondary endpoint is the same as theirs. Our trial is 90% powered to show an 80% reduction in our primary composite endpoint. Maybe let's start more from a development perspective on obesity, and then I can add something on the commercial. Yeah, sure. Big picture for obesity, we still see tremendous opportunity and unmet need. The contribution to atherosclerotic cardiovascular disease and heart failure and AFib and other core areas for us is very apparent. Managing obesity is central to managing cardiovascular patients now. The GLP-1s have obviously established a huge market, been transformative. The discontinuation rates, I think, are pretty telling. Type 2 diabetics have discontinuation rates around 30% at three years despite the drug being covered. I think the tolerability opportunity is really large. We think the field will evolve. We'd like to evolve with it. There's, I think, an opportunity for differentiated, more tolerable medicines, orthogonal mechanisms, etc. We have quite a number of early programs, very exploratory but novel, to try to contribute in a differentiated way. Maybe just adding on, I think you see that we're really looking to obesity, but we're in very early stages. If we want to get something forward, then it has to be highly differentiated because I think there is a need for higher differentiated and better products. Even with your statement as well that the market is probably becoming bigger because prices are changing, we will not change our overall strategy. We'll stick to that, what we already have communicated in the past. First, Michael, and then going over. Hi. It's Matthew at UBS again. Two quick ones, please. One area where the street seems to be very different from Novartis by 2030 is Leqvio. One thing that I think we've all been surprised with is th ex- U.S. growth. Could you split out, if you had to guess, the percentage of ex-U.S. revenue that you think will contribute to that peak sales number in 2030 to give us an idea of how important the U.S. is in your mind versus probably in our mind? The second question just on abelacimab. In the past, with DOACs and other modulators in the area, physicians have been very focused on having an antidote in case the patient needs some kind of surgical intervention and you want to prevent bleeding. What's the plan for abelacimab? Okay. I take the first one as it's mainly international. Yeah, you're right. I think we're doing extremely well as well in international with Leqvio, mainly driven as well by the Asian markets, China. I think Vas as well mentioned this morning that we were expecting to get public reimbursement, and that would be great. That would open up, of course, the Chinese market with many, many patients there. China will be a significant driver and contributor to the overall top line of Leqvio, which will be as well the case with Japan, where we're doing as well extremely well. We're doing well as well in Europe. I would say Germany is currently the smallest market because we have a more restricted patient population where we will add some additional data over the next one to two years to extend that. With the two outcome trials, we see the opportunity to further drive the international market. Now, what is the split between the U.S. and international? I would say probably here we have a chance to see something like a 50/50 or 60/40 split. 60% in the U.S., 40% international. I think it depends as well on how strong we can penetrate the Chinese and the Japanese market. I think this will be two markets which would drive very strongly the top line from an international perspective. Do you want me to handle the. The abelacimab. The reversal agent. Yeah, I mean, the first thing I would just point out before answering the question is that the Factor XI genetically deficient patients, they don't bleed. They don't bleed through surgery and things like that. We have actually seen through the development of abelacimab growing comfort from investigators about abelacimab. What we saw in our phase II trial, not only did we have a 60% reduction in bleeding compared to Rivaroxaban, but our open label extension study, patients who are still on abelacimab have had procedures, have had surgery, and there has not been an excess in bleeding still. That being said, we understand that as people get comfortable, they still may want to have a reversal agent on the shelf. Clinicians, myself included when I was an interventional cardiologist, were very worried about bleeding and procedures. Our plan, we would have a reversal agent ready in that circumstance. Maybe I could add on that the commercial research that we've already done shows that there is already a segment of the physicians who don't believe that it's needed. I think the rest can come through education as well. To Ruchira's point, we will have a plan. Maybe just building a little bit on that, just recall that LILAC is actually a study against placebo. There is a nice opportunity to see if there is indeed a bleeding signal up and above that in the placebo. Thank you. Please. Thank you. Two questions, please. That's Michael from Jefferies. Patrick, with Giovanni coming in, just wondering, given your long history at Novartis, whether you could talk to whether things have changed and how they may have changed with a new chairman coming in. Back to the research side of things. We get the zigakibart data next year. For me, maybe I'm getting this wrong, biologically, there's a bit of a conflict between the complement side and the erythropoietin baths. How do you think about positioning zigakibart versus the C5s? Okay. I'll start with the first one. Thanks for the question. From a strategic perspective, I think, as you see as well and as you've seen on the presentation this morning, nothing is really changing. Clearly, when there is a change from a person, there are different maybe questions which are coming up and different discussions, great interactions. I have to say, I think Giovanni got in the beginning of the year. We're working very closely with him. He's supporting us. He's fully aligned with the strategy which we have communicated as well this morning. I don't see any major change strategically. Clearly, from a collaboration perspective, I think we see a strong collaboration with the ECN, with all the members, but of course as well with Vas, with our CEO. The second question, I would hand it over first to you. Yeah. I mean, just a reminder, IgA nephropathy, it is a heterogeneous disease. We know that different patients require potentially different individualized treatment options. In that regard, I think we have a great advantage having three of the key mechanisms for IgA nephropathy. You did not mention the foundational therapy, Vanrafia, but that is our endothelium receptor antagonist. It can be seamlessly added on without any adjustment to ACEs and ARBs and with additional benefit shown in SGLT2 patients. I think foundational most likely regardless. Consistent with the recent update to the KDIGO guidelines, the strategy that is being encouraged is to have one therapy like a foundational therapy that helps with the downstream damage and the previous damage to protect the kidney function and the second therapy to get at the source of the damage. This is a place where the [anti-APRILs] have great promise, including zigakibart. Zigakibart is an anti-APRIL. It does not have the BAFF component. It prevents the pathogenic IgA from being formed and creating the complexes that create disease. I think we have a great opportunity with the anti-APRIL class and with zigakibart. Zigakibart specifically, from a data perspective, has the longest data in terms of proteinuria reduction with two-year data showing over 60% proteinuria reduction and stabilization of the EGFR. That phase III trial is on track. If you think about that versus the BAFF combination, there is a theoretical benefit to having the combination potentially, but we have not seen that translate so far in the clinical studies. It is hard to do cross-trial comparisons always. That being said, the proteinuria rates at similar time points have not been different. If anything, numerically slightly worse for the dual. There is, on top of that, the safety considerations for having a BAFF component in terms of infection. We saw with the Vertex drug at their highest dose, which has been not continued, increase in infections and hypogammaglobulinemia. I think that will remain to be seen what the long-term profile is. In terms of the last piece of that that you asked around the positioning of that versus iptacopan and other complement inhibitors, iptacopan really gets at the inflammation that's produced as a result of the pathogenic IgA. When we talk to nephrologists, what we're hearing is that they like to use it for patients that they feel are more inflamed. There are different markers potentially of that, such as hematuria and other signs. That being said, we've seen really good uptake of Fabhalta in IgA nephropathy clinically. There was a very strong interest in studying this more effectively. My last point is that we will be working with our academic community. We've had a strong interest in generating combination data for efficacy and safety to look at the combinations of these foundational therapies with either Fabhalta or zigakibart. Maybe I could add on to this with a commercial view. What gives us confidence in the ability to position all of the drugs appropriately is that we are already seeing today with both Fabhalta and Vanrafia on the market where we do have a specific patient type for each. We are getting the patient types that we're asking for. Fabhalta is the patient that has rapid progression, persistent proteinuria, and glomerular inflammation. Vanrafia is that seamless add-on. The positioning is in line with where we see the benefit to the patient, but also where we see the field going. Zigakibart is also, if you look at our portfolio, does have the biggest value. That positioning is in line. Yeah. Thank you. Florent? Thank you very much. Thank you for taking my questions. Florent Cespedes from Bernstein. Two questions, please. First, on LTP001 and PAH, could you maybe share with us how you anticipate how you will position this product versus the drugs which are already on the market? My second question is more on earlier phase pipeline. I'm really curious. Arrhythmia, inflammation, multiple assets, multiple modalities. That sounds exciting. Could you share with us a little bit more color about these assets? Maybe on arrhythmia, high risk, high reward, do we have to understand that you will maybe target the most severe sufferers? Or will you try to address or to beat existing treatments on broader populations? Some color, and that would be great. Thank you. I'll take first. Sorry. I can do arrhythmia. I'll do LTP001. Do you want me to go first? Yeah. Let's go there. With LTP001, yeah, this is one that I'm actually very excited about. It's our SMURF1 inhibitor for pulmonary arterial hypertension. This field has had nothing for so long, essentially. It was just really tough therapies that would have to be given in critical care units. Now with Merck's sotatercept, I think we've had a real revolution in the care of these patients. The SMURF1 inhibitor pathway aims to balance TGF beta and BMP and so is a parallel pathway to that that sotatercept acts upon. Our preclinical studies have actually been very encouraging in terms of similar, if not slightly better, efficacy against positive control in those studies when you compare them to the other class. The difference is that because it's a parallel pathway, we don't expect to, nor have we seen in phase I, some of the safety issues that affect patients taking sotatercept. That includes bleeding from telangiectasias, includes rises in hemoglobin. Finally, LTP001 is an oral. When we talk to investigators that are super specialized in this field, they are incredibly, strikingly optimistic about LTP001 in terms of the clinical need that it would address. That study is actually doing quite well. We have just started phase II- B and have already enrolled patients. Yeah. You've heard about abelacimab in SPAF, stroke prevention in AFib, and the opportunity there being to treat the patients who are currently untreated or undertreated. We think about the antiarrhythmic space similarly. Maybe it's even more open if you look at how AFib patients are treated despite the fact that maintaining sinus rhythm improves outcomes. Only about 20% of people receive an effort at maintaining sinus rhythm. Huge opportunity there. The lack of treatment is due to the side effects, safety problems with the current standard of care. There's really, I think, an opportunity to disrupt a non-treat paradigm for the whole field with safe antiarrhythmic drugs. The overall strategy is to try to develop atrial selective mechanisms and molecules. Early stage exploratory programs. We do have multiple programs in the clinic now, as promised last year, exploring multiple different mechanisms. Hopefully, we'll have more to say in a couple of years. Thanks, Shaun. Steve? Thank you, Steve Scala from TD Cowen. I'd like to follow up on the obesity question. Last year at this meeting, the company said it was looking at mechanisms for appetite control and sparing lean muscle. With all due respect, it doesn't seem that things have advanced much in the last year, not really much in the way of detail. Can you provide any mechanisms you're exploring? Could there be a candidate in development in 2026, maybe phase I? Just maybe provide a few more details. Thank you. Steve, Shaun. Yeah. Yeah. Thank you. These programs are really early and exploratory. I'm afraid we don't have a lot of information yet. Some are completely novel mechanisms. We're not pushing very hard on lean mass sparing. We are sort of thinking about what that indication might mean. It's not clear what it means clinically. As you know, we had bimagrumab, and it's been out licensed. We are still looking at appetite control mechanisms, energy expenditure mechanisms. Will we have a molecule in the clinic next year? Probably not next year, but hopefully the following. Thank you. Any other questions? Please. Thank you. Jenny Nemes, capital investor shareholder. To follow up on Michael's question on IgAN. Thanks first for sharing the different positionings for your three drugs, which is great. One of the competitors, zigakibart, will have their sibeprenlimab to launch in the market very soon for IgAN. That is also anti-APRIL. I just wonder how will that impact your maybe sales trajectory or the outlook for the potential peak sales, especially considering they will have at-home auto injection. Thank you. Sorry, Dianne. Yeah, sure. All of that is built into our forecasts and what we've already assumed is our peak. We do not see it changing anything that we've already guided. We look forward to having another modality. It is a good thing in a disease like this that has not had any treatments in the past. We think that having more drugs helps to expand the market. We will benefit in time with our own from any of the shaping that they do. I mean, I can add one point to that, which is one thing that's evolved, which I alluded to previously, is that clinicians and the scientific community are realizing that combinations are going to be critical in this heterogeneous disease. The recommendations have already changed in the guidelines for that. I think here we do have an advantage for zigakibart because we will be able to generate evidence for efficacy and safety of combined therapies and combined approaches. I think that's something we're really hearing clinicians asking for. Like you said, Dianne, I think IgA nephropathy is a devastating disease amongst young adults who often are diagnosed and very high rate of end-stage renal disease. Hence the recommendation to go for combination therapies more aggressively now. Yeah. Maybe I can build a little bit more onto that commercially because the fact that we are the only company that has three different mechanisms in this space not only helps to the point that Ruchira made around the evidence generation, but the other commercial factors. It helps with our payer coverage. If these patients are going to require multiple medications through their journey, they also have an opportunity to stay with one company in terms of the patient support that will be needed for these products. That gives them a consistent experience, and we can build loyalty over time. There are a number of benefits to having the portfolio. Thank you. Last question, Steve. Thank you. Just curious, would Novartis consider launching Leqvio and pelacarsen via the DTC program? Maybe I can take that. I think, I mean, you've already seen our announcement for Cosentyx. We continue to monitor and evaluate any innovative ways that we can get medicines to patients. Okay. With that, I would say we close here. Thanks a lot for your attention. Thank you for your prestigious questions. Have a great day with the other teams. Thank you. I'm Shreeram Aradhye. I'm the President for Development and the Chief Medical Officer. And it's my pleasure to introduce my colleagues, Reshema Kemps-Polanco, our Chief Commercial Officer in the U.S. with a long interest and experience in oncology herself, Dushen Chetty, who's been heading up the Oncology Development Unit for the past year, and Shiva Malek, our Head of Research in Oncology. I'm also pleased to introduce you to, Mark, if you wouldn't mind standing up, Mark Rutstein. Mark will take on now and has started as the new Head of Oncology Development. Mark joins us from Daiichi Sankyo. And before that, BMS. Long experience in oncology for many years. And we're delighted to welcome Mark. He won't be speaking today, but this is his first introduction to meet Novartis Management. Mark, thank you for joining us. It's been an exciting year since we last met at M&M in London. Lots has happened in oncology. Great progress on Kisqali, Pluvicto, and Scemblix. Scemblix just got approval in Europe now. As Vas pointed out this morning, an extraordinary example for us of the pace at which we've started to work, getting it approved within three years of having first started the trial. All of the subsequent approvals moving at pace, the most recent one happening three and a half months earlier than we had expected. In my mind, a clear reflection of the quality of the work that is happening across the company. Our overall strategy, prostate cancer and breast cancer remain tumors of focus. Radioligand therapeutics become the platform where we believe we are making the consistent investments across all competencies to create what we believe is a unique expertise in bringing precision radiation, as we have called it, to more patients. Core strategy there, as illustrated with Pluvicto, moving into earlier lines of therapy and bringing that benefit to a larger population. The whole concept of how we can combine radioligand therapeutics with other agents, as well as evaluations of beyond RLT molecules like the CDK2 inhibitor that now enters clinics for breast cancer. Overall, super excited with the progress that we're making. I'll open this up and look forward to your questions. Michael. Thank you. It's Michael Leuchten from Jefferies. Can't let you get away without commenting more on Kisqali versus the Roche data that came out the other day. Just from your perspective, how do you think about the segmentation of the patient pool, stage two, stage three, medium risk, high risk? Where do you see the white space where maybe a next generation oral SERD might be an option as monotherapy? Where is monotherapy definitely not going to play? I guess, how does that fit into your development program for the combo trial that you're going to kick off next year? An important question. I'm going to start, and then I'm going to let each of my colleagues actually comment on their perspective from each of their own areas of expertise. All in all, I think first, we wait to see the data at San Antonio Breast. We believe that hormonal therapy, which this may represent the next advance, is complementary to cyclin inhibition. We believe that Kisqali has demonstrated its role in living up to the promise that inducing sustained cell cycle arrest and reducing distant disease is an important benefit. We have seen that consistent benefit across the entire population in early breast cancer. All in all, wait and see. We see these as complementary mechanisms. We believe the direct impact on Kisqali is, if anything, it's possible that something else becomes the standard hormone therapy pillar. For now, there's a lot of work to be done. In general, maybe, Shiva, why don't we start with the mechanistic thinking? Yeah. I would start first by saying that endocrine therapy, like the next-gen oral SERD s, their mechanism of action actually converges and synergizes nicely with CDK4/6 inhibitors. As you know, estrogen receptor actually regulates cyclin D1 expression. Cyclin D1 is upregulated in breast cancers, including early stage breast cancers. The question of the combination, I think combining with next-gen oral SERD s would certainly be something that we're interested in doing. Of course, we have combination studies ongoing with external partners, including with Olema. I think it's not surprising that they're seeing this level of activity in the early setting. This is where ESR1 is still wild type. There's a greater dependency on in that setting relative to the later stage metastatic setting where we know the biology is actually fundamentally different. Maybe Dushen can comment on the study designs and patient population. Yeah. I think while Shreeram mentioned, we wait to see the data, what is important to highlight is where we are with NATALEE and what are some of the learnings we've had from NATALEE that may be relevant here. When we did the five-year analysis of the NATALEE study, there were a few things that stood out for us in terms of the long-term benefit that patients are deriving. While we might see an initial interim analysis now for the SERD, what we really need to see is longer-term data to understand its potential in preventing recurrence. The whole purpose of administering drugs in an adjuvant setting is to prevent recurrence. About one-third of breast cancer patients will recur. What we saw from the NATALEE five-year data was that even though patients had three years of Kisqali and then two years without therapy, that benefit is sustained. What that tells us is that the hypothesis we had around CDK4/6 inhibition resulting in irreversible senescence is borne out. In fact, further, we also looked at distance disease-free survival, disease recurrence-free survival at the five-year mark. On both of those measures, we saw about a 30% risk reduction. What that tells us is that we're preventing the formation of metastases. I think that's really important. While we have an initial effect on disease-free survival, what we need to see to understand whether this is going to have potential in monotherapy or in combination is what is the potential in terms of preventing metastases and preventing disease recurrence. That is what we'll be looking for as we move forward. Yeah. Just to add to that commercially, I think the question is less about what is the impact on the CDK4/6 class and more about can this truly displace traditional hormone therapy as a new backbone, as an optimized backbone. I think what we've seen is that Kisqali has become the CDK4/6 of choice, whether you're looking at early disease or in the metastatic space, and regardless of combination partner. This question around will an optimized monotherapy displace an innovative combination, I'm not sure that's the right question. It's more about what will it take to displace traditional therapy. One of the things we will be looking for in the data is what is that risk-benefit profile, particularly what is the safety profile in this early disease space. Because the patients tend to be younger, a lot more active. Fitting this into their lifestyle with a favorable side effect profile is very, very important. I think we just have to wait and see. I remain confident on the CDK4/6 class and in particular Kisqali. We have now seen penetration of about 55% class share, which is now well-established standard of care. With that share now has broken towards Kisqali across the broad population. We believe we have more headroom. If you compare it to the metastatic setting where it is over 80% penetration of class share, we believe there is a lot more room there. Matthew? Thank you. It's Matthew Weston from UBS. Two quick follow-ons, please, both on Kisqali and SERD. Vas, I think, in his opening comments suggested that he saw the future potentially as SERD plus CDK4/6. If I look at that commercially, I think the best way of winning there is to own both elements of the combo. Particularly if I think that you've got Kisqali LoE and Kisqali IRA coming, then it would be a great way to protect against both and have much more longevity in your franchise. I think I'm right in saying as a consequence of the Olema-Kisqali supply deal, you ended up with a 20% stake in the company. Can you confirm that? How do you feel about my premise that owning both parts, at least commercially, would be a really successful option? The second quick question, particularly because you suggested you understood why the efficacy was best in the earliest setting, because that is where estrogen plays the strongest role. If we were to see SERDs work in adjuvant but not work in frontline metastatic, do you think that changes your perspective on patients' sequencing in a clinician view? Because you could keep the CDK for a first-line metastatic patient, but you probably would not do that and have the combo in adjuvant. Maybe you start with your answer, Shiva. Right. Right. OK. If I'm understanding, you're asking about the role that the oral SERD s will play in the early adjuvant setting versus the metastatic setting. I'm asking if it doesn't work in if SERDs don't work in frontline metastatic, does that change your view that a doctor won't be influenced about sequencing? Yeah. Maybe that's something Dushen you want to take on. I'll take that. I think we learned from moving Kisqali into earlier lines of therapy that what we're trying to accomplish is to give the person with early breast cancer the best shot. Every time I went to ASCO over the last three years, and I met with people and patients and their treating physicians, and certainly the patients, which is when you see a woman who says that I want to make sure that I've done everything I can to prevent my risk of recurrence because I want to be around for my child to graduate or for my grandchild to be born, whatever it is that was important to them, the idea that they would therefore they would probably want to go for the best option available at the time of their diagnosis at an early setting. We believe that knowing that we have this complementary mechanism of cycline inhibition, CDK4/6 inhibition, in combination with hormone therapy, for which we wait to see whether a different standard has emerged. I do not think that people will say, I'll get you a little bit of something now and then wait for it to use something once you have recurred and had metastasis. I think going back to your first question, the idea that with these data, we have always thought about the fact that we need to have a plan for how Kisqali can be combined, how these two mechanisms of action can be combined. Conceptually, having two agents that are your own, of course, would make sense. Our partnership with Olema is around the clinical trial. I'm not in a place to be able to give you clarity on the content of ownership structures because I actually don't know. Florent? Good morning. Florent Cespedes from Bernstein. Thank you very much for taking my questions. In fact, a broad question on radioligand therapies. This morning, Vas, during his remarks, introduction, highlighted that there is a potential of almost $30 billion sales market potential for this class. Could you maybe elaborate a bit on where you see most of the opportunities beyond prostate cancer, breast, lung, other diseases? If when you use these treatments in earlier phase, what is the risk to have some side effects that would prevent the use in earlier setting or in refractory patients is enough to achieve this kind of almost $30 billion market potential? Thank you. Thank you, Florent. Maybe I'll frame this as how do we see the path to the $30 billion market. I'll ask each of my colleagues to actually share with you how we think about it. We have said that having invested a lot of energy in RLTs and their development and building upon what we have learned from Lutathera and moving Pluvicto starting in the metastatic setting to earlier lines, we know that it takes a lot of effort to build the competencies all the way from the discovery stage to manufacturing to be able to successfully identify, build, develop, and deliver radioligand therapeutics. Each of those pillars has its own need for our own expertise. We believe that we have built that expertise. We live in a stage where I believe that many people outside of the company do not know what they do not know as people are trying to get into the space. Everybody sees the potential. Many of our competitors are coming into the space, which we welcome. I think maybe, Shiva, why do we not start with you with how we think about it in the research space? Vas started his study. We go to Dushen and to Reshema. I'm happy to answer that question. As Vas mentioned, we have over 20 programs right now in research in the radioligand therapy space. We're taking perhaps maybe three approaches. The first, I would say, really excited about targets that have more broader pan-tumor potential. That would be a target like FAP, which we have both the 2286 molecule that Dushen can tell you more about, as well as our 2nd-gen molecule from research. That program targets a protein that's specifically expressed on cancer-associated fibroblasts that are present in a variety of different solid tumors, including pancreatic cancer, lung cancer, breast cancer, so broad tumor potential. The second kind of approach is more of a targeted approach, getting at targets that are more clinically validated. For example, DLL3 RLT, which is currently in phase I. This is a molecule that we acquired through our acquisition of Mariana Oncology last year. The interesting feature of this is, of course, this is a clinically validated target given there is a T cell engager approved in this space. It is a low-expressing target, exquisitely expressed in small cell lung cancers. The Mariana technology was able to really achieve getting nice efficacy with actinium in this space. The third is really thinking about indications that are still radiosensitive, unmet need, and how we might be able to provide better patient benefit through RLT. For example, the HER2 RLT program is one in that setting where we think with the radioligand therapy profile that appears to be safer and potentially opens up different combinations, that may be an area that we could really make a difference. Dushen, why don't you talk about how we are approaching that early evaluation in the clinic and dosimetry and the efforts that we're making and perhaps end with a few programs that you're specifically excited about? Sure. I'm sure everybody here recognizes one of the unique features about an RLT program is that you see what you treat and you treat what you see. I think that's what Shreeram's referring to. It gives us a unique ability when we have a radioligand that's labeled with an imaging isotope like gallium or even a SPECT isotope like indium that you can actually see where the tumor is. You can see the progression of the tumor or how you're able to treat the tumor. I think that's a unique proposition for RLT. In terms of some of the programs where we've applied that, for example, Shiva mentioned the FAP 2286. Because of our ability to see the tumor and to see uptake into tumor, we've studied FAP 2286 across a number of different tumor types. We have seen uptake in 28 different tumor types with FAP 2286. While we are starting with PDAC and non-small cell lung cancer, the potential to expand across other indications is vast. I think that is the potential that we see with the FAP molecule. I think what everybody's waiting for is that what is the data that we're going to present that excites you beyond everything we've done with Pluvicto and Lutathera. I think to set expectations, Dushen, would we say that the early readouts that we expect to give there are in the 2027 time frame? Even in 2026. 2026 for FAP in PDAC. We also expect that with NeoB, there's two breast cancer studies, one in combination with ribociclib and one in combination with capecitabine. We should have early readouts for those in 2026. I think that we approach this with the humility that we have taken on the idea of evaluating these possibilities in early stages. These are early-stage programs. We'll see what the data reveals. Reshema, do you want to close out with the consideration? Sure. What we're seeing commercially, really proud of the work that the teams have done in the U.S. to expand the infrastructure to support RLT. We did this not just for Pluvicto or Lutathera because we imagine the future that is to come with what we believe will be a proliferation of RLTs, not just from Novartis, although we are leading and want to continue that sustained leadership. We also see that other companies now see the promise of RLTs and are now having their own development programs. What we would imagine for the future is that this really becomes a mainstay of cancer treatment, just like we see with chemotherapies, targeted therapies, and other modalities. Today, we have over 700 treating sites, 580 of them continuously ordering. What I'm really pleased to report since being here last year is the scaling that we've seen in the community setting. Last year, when I was here, I talked about what are the things that need to be true in order to see a rapid uptake for PSMAfore. We have indeed seen what we predicted. We see considerable growth coming out of the community setting ahead and faster than the academic setting. We expect this to continue as we get into 2026. Hopefully, with the approval of PSMA edition, we will even see that expansion into the urology setting. We also see now urologists, large clinical practices in urology, starting to set up infrastructure to be able to administer RLTs. Very, very, very promising field. Yeah. I think, Florent, I'll close by saying that the underlying principle remains the same, which is we must demonstrate that an RLT delivers a therapeutic index that is meaningfully different from what a standard of care is otherwise offering. To that end, for example, there's also the entire idea set of activities that are looking at how can we combine RLTs with, for example, DNA damage repair inhibitors, where the ability of potentially further enhancing delivering more efficacy, potentially altering the regimens and radiation delivered, evaluating what are the tumors that will lend themselves better to an alpha emitter versus a beta emitter like lutetium. A lot going on. We see this as continuing to build our competencies in an area where it's not going to be as easy for competitors to catch up with us. Thank you. Simon Baker from Rothschild & Co Redburn. Just continuing on that theme, you've got a number of assets targeting prostate and PSMA, different ligands, different isotopes. Can you just give us perspectives on the characteristics and the tolerability there? Is this more about the isotope and the emission, or is this more about clearance rates in terms of toxicity? Some thoughts there. Just associated with that, given you're now in actinium, what's your perception of actinium supply? From what we hear, it's not perfect, but it's a lot better than it was. Any thoughts on that would be helpful. Then a slightly random question that gets asked a lot around the sector. What's your current perspective on PDX-VEGF bispecifics as a modality? Thank you. OK. Shiva, do you want to start with the first question on how do we think about the PSMA targeting and Dushen about the therapy? Yeah. Maybe I'll start. We spend a lot of time thinking about the choice of radioisotope we use. Lutetium is a beta emitter, has a good linear energy transfer, and we can also get this crossfire effect that we want to see in neighboring cells, whereas actinium is much higher energy linear energy transfer but travels shorter distances. You could envision with an actinium agent, you could get at micrometastatic lesions. You might also want a higher energy agent when you're targeting a low-expressing tumor antigen, for example. We really think about the biology and the tumor type as well as expression of the antigen we're targeting. The safety profile between lutetium and actinium, as you might imagine, would look different. It's really about a risk-benefit analysis of understanding the degree by which we're generating efficacy for patients and what the safety looks like. Dushen can tell you more about the clinical studies. Yeah. We have now two phase IIIs with actinium. The first of those we call PSMA action is in a post-Pluvicto setting, in a post-vision population. We think there the rationale for doing that study is very compelling because there are no approved therapies in that setting. Patients have a median overall survival of about seven months. Offering an alternative with actinium there is very important. The second study we have is in newly diagnosed MCRPC. We also see a potential for advancing therapy in that setting. I think ultimately, our goal there is to have optionality, where we believe that Pluvicto can be more forgiving in terms of the toxicity. With actinium, as long as you select the right patients, it can be very effective. With the higher linear energy transfer that Shiva mentioned, we expect higher efficacy. The other aspect of that study that's important is that we're combining with an ARPI. We've seen from evidence that Louise Emmett generated in an ENZA-p study where you have upregulation of PSMA with an ARPI. We expect that not only the combination with an ARPI in this first-line MCRPC setting is going to be helpful, but having PSMA upregulation is going to contribute to the efficacy of actinium in that setting. Do you want to share our current view on PD1-VEGF? Yeah. We are monitoring the space. We have seen compelling data in solid tumors. We understand how the biology in the tumor microenvironment is changing with the PD1-VEGF. I think our view here is that we are not going to do a me-too. If we go into the bispecific space, we will do something that is differentiated and allows us to add value rather than just following where others have already been. OK. Steve. Three questions. The first may not be appropriate for this panel. But can you clarify, Olema, do you have a right to first refusal on the asset? There seems to be some confusion on this sort. Secondly, does the new Kisqali guidance assume it is selected for IRA in 2028 or not? And then lastly, the next-gen CD4, what makes it next-gen other than the fact that it doesn't have CDK6? I think I'll just take the first question, Steven, that I don't have the answer for you. Maybe we'll save it for one of our other colleagues if they're able to answer it when Ronny is here. Yeah, I don't actually have the answer to it. Reshema, do you want to talk about the Kisqali? Short answer is yes. It is assumed it's baked into the guidance that we've given. For the CDK4, what makes it next-gen, Shiva? Yeah. You definitely hit it spot on that the idea here is really to hit CDK4 as potently and as selectively as possible and really spare CDK6 inhibition, which we know drives a lot of the heme toxicities that have been observed with the dual CDK4/6 inhibitors. The hope there is, right, you can get continuous target coverage and drive greater efficacy and then open opportunities for combinations, both doublet combinations with endocrine therapy but also triplets and other combinations. That is what we are hoping to achieve with our next-gen. Maybe I want to take this as an opportunity to have you elaborate a little bit more on how we are approaching in oncology in our early work, our prior experience in this space, as well. Right. Yeah. So we've learned a lot. Of course, as you know, Kisqali was a homegrown molecule. We call it ribociclib in research. It was developed by the chemists in-house in biomedical research. It was really learnings from that molecule and the structure-based drug design that happened that enabled us to first design a next-gen CDK4 selective molecule. We then, of course, have ECI, our CDK2 selective molecule. That was really grounded on learnings from the Kisqali clinical studies, where we understood that cyclin E upregulation can drive progression and resistance. Really targeting CDK2 that forms a complex with cyclin E1 became important. We have developed that molecule. Thinking about how we can put everything into different flavors of CDK2/4 inhibitors is kind of the next generation that's moving forward, again, all geared from our insights from the original Kisqali work. All right. Michael? It's Michael Leuchten, again, from Jefferies. Questions on or question on Pluvicto. If you take into consideration changes in regulation for the amount of radiation that can go into a patient, where does that fit in terms of the revenue potential? Are patients still not getting the full number of cycles because physicians are worried that they're not going to be able to retreat? Question on just going back to breast cancer, KAT6, is that something that fits into your thinking from a portfolio perspective? OK. Maybe Reshema, do you want to answer the Pluvicto question? And then Shiva? Sure. On Pluvicto, in this earlier setting for PSMAfore, we expect that the duration of therapy to be longer and closer to what we saw in the clinical trial versus what we saw in the vision population. Just a reminder, the vision population, that really was almost you would think of it as salvage therapy, right? Unfortunately, there was a case where maybe the patient progressed or they were too sick or maybe the patient passed away, and so you were not able to finish those doses. We are not seeing that dynamic in this earlier stage. It is still early, as it takes a while for those cohorts to move through to get the six doses. So far, there are no red flags in terms of what we are seeing in terms of the number of doses. We will have a better view on that as this initial cohort gets through in 2026. Dushen, do you want to add something? I just want to add one bullet there. In PSMA edition, 85% of patients completed six cycles of therapy. That was very encouraging for us. Maybe with KAT6, yeah. We are monitoring what's going on with KAT6. Of course, Pfizer has published some interesting data around that target. I think what's unclear is the patient population. KAT6 is amplified in subsets of breast cancer patients. They did not select for that population. We will need to understand what role that plays. There were toxicities observed. Thinking about combinations and the future of that is something that we will have to monitor. OK. Go to the back. Florent, give me one second. Hi, Zain Ebrahim, JP Morgan. Thanks for taking my questions. First question, just to build on the PSMA edition conversation in terms of the data, you saw quite a strong benefit on RPFS. OS is still early but positive trends. Just how you're thinking about the provability of that data. The ESMO discussion was maybe a little bit more lukewarm in terms of the efficacy there. Just your perspectives on that would be helpful. The second question would be on Kisqali and room for growth in metastatic. Your TRx share has been growing, but your NBRx share has been stable. How should we think about headroom for growth in the metastatic setting? Dushen, do you want to take the PSMA edition? Yeah. Yeah, thank you for the question. I’d like to start by saying we respect the discussant’s views. We think it’s important to raise these scientific questions and have a discussion on them. We look at his views with the context of the full data set for PSMA edition and our previous studies with Pluvicto. What he highlighted were three areas that he had questions or concerns around. One was a belief that patients with higher uptake will have better efficacy. When we look at all of the different subgroups for PFS and OS, what we’ve seen is a consistent benefit over the ADT-ARPI comparator arm for all of the different subgroups. Some subgroups do slightly better than other subgroups, as you would expect. All of the subgroups did better than the ARPI-ADT arm. That was the first of his observations. The second was on the safety profile. What we have seen in terms of safety is that there were no new safety findings for the combination. It is an add-on study design. You have Pluvicto plus the backbone therapy, ADT plus ARPI. You would expect a safety profile that reflects the add-on therapy. The third point was around overall survival. He was concerned that we would not show a benefit in overall survival. What we did show at the second interim analysis at the ESMO presentation was at the time with about 47% of information fraction, we showed a hazard ratio of 0.84, despite the fact that there was 16% of patients from the ITT population who had crossed over. About 60% of patients who were eligible for crossover had crossed over. Despite that, we saw this trend in overall survival as the data continued to mature. We expect that at the third interim analysis, the confidence intervals will tighten. What we learned from PSMAfore was that we saw a growing benefit in terms of overall survival as we looked at subsequent interim analyses. Ultimately, it is the data that will speak. We will do that analysis. We expect it is event-driven in 2026, the first half of 2026. Yeah. Maybe I'll go on the Kisqali question. I appreciate the question. It's a really good one. In the metastatic space, I think it's important to understand the dynamics that are happening there first, second, and then line two plus. What we see is consistent growth in the metastatic first-line share. That's exactly where we want Kisqali to be used. What we also see is as emerging data comes in around different patient types and patient groups, in the back end of the disease, you start to see some erosion there on second-line plus. That is okay, honestly, from a strategy perspective, because we do want Kisqali used earlier. Also, if that data is emerging, if it shows that a subgroup of patients are better treated in second and third line on something else, then that makes complete sense. In the first-line setting, we do believe there's more headroom there. As a reminder, in the metastatic space, the way I view it from a strategic standpoint, that's a share-capture strategy. Whereas in the early breast cancer space, that's an expansion strategy. In the share-capture strategy in metastatic, we believe there's more headroom there because we still see patients who are started on palbo as well as BEMA. The question there is, how much more of that share can we capture in terms of those new starts? We believe there's still room there. We are getting up there. I will say we're getting up there. There's still headroom to go in this time period that we've given guidance for Kisqali overall. Thank you, Reshema. Florent? Thank you very much for taking my questions. Florent Cespedes from Bernstein. Two follow-up questions on radioligand therapies. First, what is the difference on your slide here between the 1st-gen alpha-emitting product and the 2nd-gen alpha-emitting product? Is it a question about easier to handle, longer shelf life? Any color on this one would be great. A more broader question on the radioligand long term. How would you believe it is possible to combine these products with existing treatments, given the fact you have risk of having more side effects? Dushen, do you want to take both? Yeah, the first question. I have two different ligands. In the 1st generation, we have actinium PSMA- 617. The ligand there is exactly the same ligand that you have in Pluvicto. The second is a ligand called PSMA- R2. What's different between them is we've seen a difference in uptake to salivary gland. What we're hoping is that we would have a broader therapeutic index with PSMA- R2. We continue to look at data there. It depends on how the data pans out. These were based on dosimetry experiments for a study we did with lutetium on PSMA- R2. If it pans out that there's a broader therapeutic index, it gives us the potential to take it into different settings. That would be how we would apply the two molecules in two different settings. Do you want to comment on the potential for combining RLTs with other agents? Absolutely. We've got actually a number of combinations. In fact, PSMA edition is one example where we've shown combination with backbone therapy in prostate cancer. Also with Lutathera, we have a study ongoing in extensive stage small cell lung cancer, where we're combining with chemoimmunotherapy. We also have in prostate now an ongoing study with an AR degrader. I think that holds great promise because we not only degrade the androgen receptor, but through degradation of the androgen receptor, you upregulate PSMA, so more target for Pluvicto to bind to. Shiva also mentioned DNA damage response inhibitor. We're about to start a study where we're looking at a combination of Pluvicto in the first instance with the DNA damage response inhibitor. If that pans out as we expect through this mechanism of radiosensitization, it will have broad applicability as a combination across all of our RLT portfolio. Shiva, anything to add? Did you want to say? Yeah, no. I think that there's a variety of different ways we're thinking about combinations with radioligand therapies. I would first say that RLTs are actually pretty well tolerated. That's one of the reasons we're so excited about them. We're thinking about both radiosensitization strategies that Dushen described, but also mechanism-dependent combinations, like the combination with the AR antagonist. Thinking about how you might, in an orthogonal way, potentiate the efficacy of the RLT. The overriding principle for now remains the same, which is that the commitment to make sure that we are designing the trials that evaluate how the combination delivers disproportionately greater efficacy without adding to the toxicities. The therapeutic index has to be the fundamental driving principle of what we're doing. Matthew? Thibault? Steve? Thank you. It's Matthew Weston again at UBS. Just following up on the Kisqali question about different lines of therapy, I'm just intrigued. Are we seeing treatment through multiple lines yet? It may well be too early. If you have a patient who's using it in early breast cancer, what's the physician doing when they progress? Are they keeping Kisqali? Are they keeping CDK4/6 but changing to a different agent? Are they having to move beyond CDK4/6? It's a really good question. In terms of Kisqali, we actually haven't seen, it's still really soon to see those patients. Hopefully, we won't see a lot of that. That's the whole goal. Maybe we can speak to the data around what happens in sequencing and how we're thinking about that as these patients progress. I think it will need to be studied systematically. I think we know that it is being used across different lines. We need to have a systematic study that evaluates the efficacy if it's used in early and in late lines. We haven't done that yet. Thank you. Maybe I'll just go to the back, Steve, just behind you. Hi, big picture. I was wondering, do you feel you've missed out by not really having any much ADC over the year now that you've sort of been a couple of years down the line? Second, I was interested in what do you think is your most contrarian take or undervalued oncology asset that you're looking at? It might be the contrarian take as not being in ADCs, but I'd be interested in those two. Thanks. Maybe rather than contrarian, I'll say that we had a pioneering position on doubling down on RLTs as the place where we were going to get focused on delivering precision radiation rather than precision chemo, which is how I might think of an ADC. Having said that, we've also had an interest in recognizing that there is a role for ADCs. We have our own efforts both within and continue to track things happening outside on what is going to be a meaningful ADC that we could bring into the tumors that we choose to focus on. Shiva, do you want to maybe comment a little bit on build on the ADC part? Then I'll go to your last question. Yeah, yeah. First, I would agree with you, Shreeram. This was very intentional to build our radioligand therapy platform, thinking that it's an opportunity to deliver radiation selectively to tumors. We have a chance to learn. We're the leaders in this space. We have the opportunity to learn from the clinical studies and iterate back on that in research. That has been very intentional. At the same time, we're thinking about how we can take the ADC technology, which is really targeted delivery of payloads, small molecule payloads, and has classically been panned cytotoxics that have worked quite well in the solid tumor setting, although they do have some safety challenges. The approach we're taking is our strength internally is really in the chemistry that we've done. We have a long history of really innovation in the small molecule drug discovery space. We are developing what we call biology-matched payloads and delivering them selectively to tumors. That is kind of the next wave that we are bringing forward. Again, really thinking about where panned cytotoxic ADCs may play a role and complement the work we are doing with RLTs. I mean, I think what we have decided is that we get focused on specific tumors. We also are paying attention to areas where there is true unmet need, like PDAC. We aim to deploy against it the modalities that we believe have the greatest potential to make a difference. We are deeply focused on RLT. We remain open, of course, to continuing to evaluate other platforms. This conversation will continue. The most undervalued contrarian asset that you have? You mean that's something that we see as valuable, but you don't? Yeah. I think that the community in general is, in my view, from my perspective, appropriately waiting to see how all of this effort in RLTs is going to actually play out with the next visible, meaningful therapeutic index delivered beyond Pluvicto and Lutathera. We understand that. We believe that that's going to come from one of the many programs that we currently have in the clinic. Having said that, they're all, of course, still in early stage. The key test is going to be, is the data that we deliver demonstrative of a disproportionately better therapeutic index than what's currently available? All right. I think we're at time. Steve, do you want one last question? OK. Can you speak to the data from the competitor that could challenge Scemblix? Presumably, you don't think much of it because you raised guidance this morning. Also, do you think CDK4 alone can beat CDK4/6? Thank you. Sure. Maybe Dushen, you take the first one on Scemblix. Then Shiva, you take the CDK4 one. The data we've seen in the Scemblix competitor, I think, is very early data, small number of patients, and in a later line of therapy. I think really where we're adding value with Scemblix right now is in the first line of therapy. It's going to take a number of years, firstly, to see whether that plays out with the competitor, if it does at all. Right now, it's mainly a small data set in later lines of therapy. Yeah. Can I just take a moment to just comment on that as well as what we're seeing with Scemblix? I would be remiss not to. I agree with what you're saying around that data. I'm not thinking a lot of it at this moment, really, because what we've seen with Scemblix is that it's the market leader, certainly in the third line, where we first launched it, market share leader in the second line, undeniably. We are going for leadership in the first line as well. We are seeing consistent growth in the first line. That's really where we're positioned. Earlier this year, I think we communicated we were at 10%. We were at 16%. Now we're hovering at 22%, approaching having one out of every four patients start on Scemblix. We are newly diagnosed with CML. We continue to see that as a lot of headroom there to grow. Anything that comes in is likely to be used in the back end of the disease. When that happens, it typically does even drive the first line even higher because physicians feel like they have somewhere to go if the patient does progress. Maybe, yeah, if I can address that. First, I will say that Kisqali is the most CDK4 selective of the 4/6 dual inhibitors that's clinically approved, which we believe is one of the reasons why we've seen the overall survival outcomes in the clinical trials. But then the premise is, of course, if you can now just selectively target CDK4, which we know is the critical dependency in positive breast cancer tumors, and spare especially the heme tox that is driven through CDK6 inhibition, we can achieve more durable and constant target coverage. That will drive greater efficacy and greater combination potential. It remains to be seen. Of course, we're going to be doing that clinical experiment. As you know, many CDK4/6 inhibitors have to do intermittent dosing, three weeks on, one week off. The idea is, can we achieve more continuous target coverage with the selective molecule? We'll see what happens. Very good. On that note, Shiva, thank you. Thanks, everyone. Thanks, Dushen. Thanks, Reshema. Thank you all for your attention. See you next time. Good afternoon, everybody. I hope you had an enjoyable lunch. You have been enjoying the day so far. Welcome you to the Neuroscience Group discussion. I'd like to start off by introducing you to the Neuroscience team. Bob Baloh, who leads our neuroscience research. Tracey Dawson, who is representing all of commercial but based in the U.S. I'm Fiona Marshall, Head of Biomedical Research, but a neuroscientist by background. Norman Putzki, who is Head of Neuroscience Development. I think you can hopefully have already sensed from Vas's presentation our level of excitement about the neuroscience area. We, as a company, have been in neuroscience for a long time. We have stayed very committed to the neuroscience field. We have a deep disease area understanding, a deep bench of expertise. We have been frustrated, I think, over previous decades of really understanding some of the causal mechanisms of these diseases. That ranges from the rarer neuromuscular diseases through to the bigger diseases like Alzheimer's. Unfortunately, we have been limited by the tools that we have had available to us, particularly to deliver modifying treatments to the CNS. I really feel that we are at a really key inflection point now where we have targeted delivery of oligonucleotides, both to neuromuscular tissue as well as to CNS, and evolving now capabilities with gene therapy. What we have done is to do a number of strategic partnerships or deals that have brought in new technologies or access to technologies to complement the expertise that we have in-house. You can see really a deep portfolio. Really two key aspects, again, I think Vas referenced this, is maintaining our leadership in disease areas where we already have very established deep expertise, obviously, in this case, multiple sclerosis, and also building on SMA, Zolgensma in SMA, the rarer neuromuscular diseases. That is really one of the triggers for the acquisition of Avidity and our strong sort of belief in the opportunity in the range of muscular dystrophies. Going into areas where we feel there is just huge unmet need, either no treatment at all or very limited treatments. Again, this takes us into these neuromuscular diseases, but with different ways of accessing those, again, ranging from oligonucleotides, in some cases with a transferrin receptor, to deliver to muscle. We also have a lipid conjugate platform, which came from our acquisition of DTx. That is now in the clinic, enrolling patients for the treatment of Charcot-Marie-Tooth disease. The same applies to the areas of neurodegeneration, so Huntington's disease, Alzheimer's disease, ALS, and PD. Again, we're using these same types of approaches, targeting the key genetic drivers of disease, but very much going after them with disease-modifying modalities such as antibody oligonucleotide conjugates or ultimately gene therapy. Despite all of this, we still have fantastic chemistry within Novartis. We really believe that we can, where there are good targets for low molecular weight, then these are very good molecules, especially if they can get into the CNS. Many of the pathways that we're interested in, some of these arise actually in our immunology group, have direct relevance to neuroinflammation. That is really one of the reasons why we were able to move remibrutinib into CNS disorders and multiple sclerosis. We have taken the same strategy in our early portfolio with a number of different mechanisms, like complement inhibition. Again, many of these pathways have common drivers of inflammatory disease from broad autoimmune diseases through to the CNS diseases. Not spending too much more time, I'd very much like to open to questions and refer you to the deep experts on this panel. Thank you. It's Michael Leuchten from Jefferies. Two questions. Multiple sclerosis, as we think about YTB323 going into that space and also remi now going to secondary progressive, do you think there is appetite at the regulator to look at the disease differently and maybe allow more active agents more broadly? Are we stuck with the sort of segmentation of the disease as it's conventionally defined? A commercial question on OAV101. Can you talk about the, so you put a blockbuster number on it. What's the rollout going to look like, initial patient population phasing? Any color would be helpful. Thank you. Okay, thanks, Michael. Let's go to Norman on the MS and how you see the disease landscape evolving with these new types of treatments. I think it's a great question. I've been in this field for like 20 years. We have seen the field evolve with its understanding of disease phenotypes. I think you were alluding to how the phenotypes then are regulated in the regulatory language. I think generally, the field is moving towards a more holistic understanding of the disease based on biology and not as the phenotypes as they're currently described, RMS, SPMS, PPMS. From a regulatory perspective, we have had these conversations with the FDA over more than 15 years in the end. We have seen some changes. When we registered [Kesimpta], for example, that triggered a whole new class label language with the FDA, which was the redefinition of what they understand RMS is, for example. I think with triggers in the community and triggers in clinical trials and data, the FDA will definitely, well, I shouldn't say definitely, but I think the FDA should feel compelled to think about that. We have done a lot of this work also internally in terms of big data approaches. We're working with the Oxford Institute here in the U.K., actually, on large data sets, machine learning. We just published a paper a couple of weeks ago in Nature Medicine to really understand better what drives disability progression in MS specifically. That has actually turned out to be described differently and beyond those phenotypes that people have been using. I would expect the field to move further into that direction. Okay, thanks, Norman. Tracey, can I go to you on the OAV101 question? Absolutely. First of all, we're super excited. We're under review with the agency in the U.S. and expect approval by the end of this year. I would first start with what's the unmet need that we think we can address. We're seeking a broad label for young kids, teens, and adults over the age of two. That's on the basis of the STEER and the STRENGTH study. We believe that one of the core unmet needs that we will be able to address is the frustrations with chronic treatments today. This is a group of patients that did not have the opportunity to benefit from Zolgensma when it was first launched. We're guiding towards this multi-billion dollar opportunity. The way we're going to get to that, not just in the U.S., but globally, is by deep community engagement to activate those young kids, teens, and adults to really make sure that they're aware that they now have, for the first time, the choice of a one-time gene replacement therapy. Everything that we're hearing suggests that there's a very high sense of urgency. There's also a very high sense of excitement within the community. We would expect to be able to give a better update after the first few months of launch. We feel very confident that the things that we have in place to activate that community, as well as the education that we're doing with physicians, will stand us in good stead to achieve that multi-billion dollar opportunity. Yes. Good afternoon. Florent Cespedes from Bernstein. Two questions, please. On myasthenia gravis, could you maybe share with us how you will differentiate the different products? We know that the disease is pretty heterogeneous. If you could give a little bit more color on this, it would be great. On earlier pipeline, could you remind us what is your strategy? What's your view on the next step and your projects in Alzheimer's disease? Thank you. Okay. Norman, do you want to cover the myasthenia gravis question and the different molecules that we have positioning there? Thank you, Florent. You have seen this. The field of myasthenia gravis is incredibly active. I think that reflects the heterogeneous biology and the different needs that patients have. I think with our own pipeline, we currently have three active programs in the clinic. We are actually really excited about the opportunity that we have across the spectrum. When you look at the various programs, we have remibrutinib in an ongoing myasthenia program. We have LNP, oral complement inhibitor. We have YTB323 going on. I think they are all differentiated in terms of their mode of action. Depending on the data, I think they will find its place somewhere in the treatment algorithm. I think without data, it is hard to pre-position that. Maybe I can hone in a little bit on remibrutinib because that is Rhapsido today with CSU. We have the ongoing MS program. It is a really important medicine for us as a company that is evolving. I think it can play a really important role across multiple autoimmune conditions. High efficacy, clean safety profile, if that is something that we can maintain also in the neuroindications, I think this offers a really strong value proposal across myasthenia gravis, across multiple sclerosis. Can I have one thing? Because I think, I don't know if you mentioned it, but I think the other big opportunity is that today, if you look at myasthenia gravis, it's largely treated through IV and subcutaneous treatments. The beauty of remibrutinib and also with iptacopan is that they are orals. These would potentially be the first orals to come to market in myasthenia gravis too. I think for patients, that's a really important point for them because it allows them to potentially not have to be in an infusion center or be tied to a subcutaneous injection. I think that's something that will definitely be very interesting to the community too. Maybe I'll pick up the Alzheimer's one quite quickly. For Alzheimer's disease, again, we see this as a huge unmet need. It also falls into the category where we actually understand a lot of the drivers of the disease. Some of that comes from the genetics. Some of it comes from study of the progression with better imaging capabilities. The two most advanced molecules are VHB937. This is a TREM2 agonist antibody. TREM2 is identified from genetics, both mutations that either predispose you to Alzheimer's or actually can be protective against the background of other genetics. That one is moving forward. We had good target engagement. We have inflammatory biomarkers that we can use as a readout that helped us to set the dose. That is now going forward to phase II study in Alzheimer's. We also think that has potential in other disease areas. The one that's most advanced in actually is ALS. The whole area of microglial biology is one that we are very interested in and continues to build up in terms of level of validation. The other main driver of Alzheimer's disease, which is best correlated with the pathology of the disease, is tau. Tau pathology occurs downstream of beta amyloid. Data coming out of the beta amyloid antibody shows that the best predictor of a response is not the reduction in amyloid. It's actually the tau reductions. Some people, interestingly, in those studies were more resistant to reducing tau. Actually, there was an interesting gender difference as well. Women were less responsive to amyloid and maybe having more tau-driven disease. Our next most advanced molecule is an antisense oligo to tau. That has the advantage as well over tau antibodies that it targets the intracellular model. Behind that, again, we also are very interested in inflammatory pathways that we think are contributing to the disease. We have a number of different molecules that maybe come from our immunology area that have potential to move into Alzheimer's as we go forward. Thank you. It's Matthew Weston from UBS. Two, please, if I can. One a commercial question about, I guess, the intersection of rare disease and mainstream disease that neuroscience has the potential to hit. Because you've got remibrutinib priced for CSU, but you also raised iptacopan as well in myasthenia gravis and taking remi into MG and MS. I know the dose is different in MS. I actually don't know. Is the dose different for MG? How do you think about pricing those molecules to try and bridge therapeutic categories when you're, today at least, those TAs have got very, very different price points for patients? A second question, Huntington's disease. I mean, it's probably one of the biggest areas of unmet need in neuroscience. It's one where Novartis has tried hard. Others have tried hard. Can you remind us about HTT227, why you think it's different, and when we're going to learn more? OK, thank you. Tracey, do you want to take the pricing differences across rare and mainstream? Yeah, I can do that. I mean, I think one of the things that's really interesting is that, yes, you do see very different price differences between what I would call specialty mainstream medicines versus a rare disease setup. One of the things that is really great about where we're coming from with Novartis is because we play in both areas, we have deep expertise. Actually, if you take remibrutinib as an example, you look at how it's, I mean, even playing in both immunology and MS, different doses, different brand strategies. These will be separate BLA, separate violence strategies. We will have separate pricing strategies for them too. Similarly, when we look at myasthenia gravis, when you look at how the interplay might be between remibrutinib and iptacopan, given the price point differences, let's just be honest, there will be a price point difference. What that will probably play out more realistically in is how they're positioned in the treatment algorithm and how they're used. I think more importantly, we need to wait and see what the data tells us first. We are very optimistic, not only in MS, but also in myasthenia gravis, that we have some great opportunities that really address what are very heterogeneous diseases. I think our next milestones would be the data readouts, which I'll throw that one to Norman about just the clinical plans. I was waiting for somebody to hand me the mic, but I. Yeah, I forgot we have got them. I look confused because of that, not because of your question. On the, let me start on maybe on Huntington's disease first. I think that is one of the most exciting programs, particularly at this point in time, since we acquired the license from PTC. You will remember the positive phase II trial that showed a dose-dependent reduction of HTT broadly in the range of what would be expected to lead to disease modification in this horrible condition. That's the experiment that we now need to do to show that we can translate HTT reduction to improve disability outcomes for patients. We had a really productive half a year or so with many regulatory interactions that helped us now to plan and execute the phase III soon in 2026. We are ready to go with a phase III program for patients with Huntington's disease. You can expect a single phase III study with a single dose, a placebo-controlled experiment with a duration, if we look at full study, approximately two to three years based on our understanding of the disease in and by itself. We will make every effort along the way to see if there is what is the potential for accelerated approval. Patients are waiting. I think if we think about unmet need in Huntington's, it is pretty obvious to everybody. We have the phase II data set. We have ongoing open label extension with the two dosages in stage two and stage three patients. We are continuously investigating those data. If they can support accelerated approval, we will also use opportunities during the phase III to align with the regulators, often approach on earlier data readout to also support accelerated approval, plus minus all the emerging data that we will look at in its totality. Okay. T hank you. Thank you, Simon Baker from Rothschild & Co Redburn. I'll stick with the question on Huntington's and then come on to Avidity. There was an interesting paper earlier this year looking at the significance of CAG repeats in HTT, where everything's fine, everything's fine, and then you tip over a limit and all hell breaks loose. Firstly, what's your view on that? Secondly, what implications does that have for trial design in this space? Secondly, moving on to Avidity and Del-zota, exon skipping in DMD. I just wondered what lessons, if any, can be learned by Sarepta's recent experiences, both from a clinical point of view, molecular and regulatory? Thanks so much. OK. I mean, Bob, you've been thinking about the repeats and implications, something we've talked about recently. Why don't you have a go at that? For sure. I mean, the field has been sort of changed by the observation of mutations or GWAS hits, I should say, in DNA damage repair pathway. There are a whole series of those. That then led to this sort of exploration around somatic expansion, expansion of the CAG repeats during a person's lifetime. It is still not clear yet the full dynamics of that because it is extremely hard to measure. You would have to measure within somebody's brain exactly the length of the repeat in every single neuron. Not exactly easy. If you try to interpolate from what data is out there from the studies that you are talking about, it does expand in the brain. It could be that the onset of disease or progression of disease correlates with a phase of rapid expansion. What does that mean? One, that could be an interesting place to target. I think us and many others are interested in the very early preclinical space of how might you do that. I think the other thing it indicates, which is probably obvious, is that in any of these diseases, you're going to want to go earlier regardless. Whether that mechanism is important or not, actually, it's probably the same answer. It does add some emphasis to that. I think it really is an important area to explore. I mean, we see somatic expansion actually in a lot of these repeat diseases. It's been debated for a while. This might be one of the first areas where we see whether that's true. The second question, yeah. Overall lowering of Huntington and Voltaflam, that's maybe say something about the trial design with respect to CAG. Y ou alluded to the biological plausibility of where you can make the biggest difference. I think for us, a starting point to that, and it's definitely early symptomatic patients for the upcoming phase III trial. We haven't disclosed any details yet on the exact inclusion criteria. That is broadly the target population that we will plan to include. I think generally, going as early as you can, I think that is something that we are certainly planning for, also for consecutive considerations in Huntington's disease. Thank you. The other question was around Del-zota, exon skipping, any read across from other trials? I mean, I don't know that a read across is the right way to look at it, at least from my perspective. I think it's more that the story of Avidity and the story of the potential success there is about delivery. It's about getting the drug to the target tissue at the right levels. We've known for a very long time that naked antisense oligonucleotides, PMOs, et cetera, don't get very well into muscle. The degree of splice rescue that you will get from those and then dystrophin restoration is relatively low. In the 1st generation, you would often present it as level or fold above baseline, which baseline, of course, is extremely low. Even if it's 100 fold above, it may not be a lot. It depends on how low the person was. With this new generation of targeted delivery, you're talking about actual percent of normal levels of dystrophin, which was shown by the Del-zota of somewhere in the 30% range so far. I think that's bringing what we think should work, but to another level. I think there, one thing we were, Vas alluded to, and we were really struck by as part of the whole package for the data supporting Avidity's platform was not only the degree of dystrophin restoration, but also actually the CK levels in those patients, which is normally extremely high, was essentially normalized. In the crossover group, all of it came down. That was really an observation that I don't think has been made in that field and suggests that, again, the delivery allowed us to hit the mechanism that we were trying to achieve. I think it's really exciting. We will probably talk again later about how that delivery mechanism goes for the other diseases. Okay. Just get some questions from further back. Thanks. Steve Scala from TD Cowen. Three questions. Novartis seems surprisingly excited about the BTK class overall. Tolebrutinib, we'll see if it gets approved. What label? Fenebrutinib, efficacy looked promising in one of three studies. And it has liver tox. Remibrutinib is in phase III. The jury's out. What do you see in this class that is interesting other than oral? Secondly, does Novartis have a brain shuttle program? If not, do you not believe in it? Lastly, if GLPs, GLP-1s prove beneficial in Alzheimer's disease, which we'll know, I guess, in two weeks, will Novartis just sit this one out? Or is there a strategy to get involved? Thank you. Okay. Okay, let me start off with the BTK compounds because the range of BTK inhibitors are very different in their chemistry and mechanism of action. These are kinase inhibitors. Kinase inhibitors generally, especially ones to get into the brain, have been historically prone with a lot of problems that relates to either selectivity across the kinase class as a whole because of particularly targeting the ATP binding site. They are very conserved. Trying to just hit one of the family is very difficult. Often the types of molecules can be prone to liver toxicity. This has nothing to do with BTK, the protein that is not even expressed in the liver. Again, it goes back to the chemistry. Often when you have a target like this, it takes rounds of work, of molecules, and improving on those till you really can get to the ones that give you the level of kinase inhibition that is required to generate the efficacy. Often the past failures are often related to dose limitation where you're not really blocking sufficiently on the kinase. Now, the way that remibrutinib is differentiated, first of all, it's a covalent inhibitor, which means it binds irreversibly to the target. That also reduces the circulating levels, which reduces off-target toxicity. More importantly, the 1st generation of covalent inhibitors still bound to the classic ATP binding site. They weren't very selective. What our chemistry team did was to understand using structural biology that the inactive form of the protein actually had a different shape. They were able to design molecules that bind to the inactive kinase form that is less conserved across the class. It took the molecule into a different chemistry space and allowed it to be more selective. These molecules are all very different. We do think we have the best in class that allows us now to have a very clean label. We have dosed. Norman, would you like to say how many now with remibrutinib? Yeah. I mean, we have had this conversation over a couple of years. I just continue to be excited. I can add some more data points now. You increasingly add the number of patients. Yeah. We do not see any liver signal. That is why we have a clean label with Remibrutinib in CSU. I think that validates the whole chemistry approach. That is why we are so excited about it because it is allowing us to go into different disease areas like CSU where you need a very clean profile, but it also has the CNS penetration. It allows us to go into some of these other indications. Let us move on. Throw in some. Okay. Yeah, go on, Bob. Because our MS program is progressing really well. Studies are now pretty much fully recruited. We expect the readout next year. As you know, we have taken forward 100 mg b.i.d. into both MS studies and also in the myasthenia gravis program. In terms of the data points that we need to get even more excited, I would say that we now have more than 1,600 patients exposed in the two RMS and myasthenia gravis trials beyond six months treatment. We are in a great position to compare, for example, safety concerns like liver toxicity across programs. We have run a similar program where ofatumumab was compared to teriflunomide in the past. We know what the incidence of liver events in the blind trial would be. We are running a similar program where we know exactly the liver tox that our competitor and comparator drug should actually have. We have not seen any quantitative differences. Of course, we also look into qualitative cases. I think as far as I know, our remibrutinib BTK inhibitor program is the only MS program or BTK inhibitor program in neuro that has not been put on a clinical hold. I think the data is emerging quite nicely. Also, in terms of study methodology, it is an event-driven trial. We monitor that. We have an adequate event rate. We have adequate patient numbers. The trial is designed to deliver Kisqali-like efficacy. Next year, we will see where we are at. OK. Bob, do you want to comment on brain shuttles? Sure. I'm glad you asked. I mean, I think that's been an area of huge activity for us actually over the last year. It's something that we haven't had a chance to talk about in some of the other sessions. The first thing to point out, I suppose, is that you don't just have a brain shuttle and it works for everything, and then you've solved the problem. I mean, I think we're getting very close to understanding things that will move proteins, those biologics, as well as RNA into the brain. That technology is advancing. There are several different transporters that are being looked at. In the last year, because we were really focused on how could we accelerate our own internal efforts for both biologics and RNA, we did three deals in the space. Sironax, BioArctic, and each of those was to partner with some internal assets that we have to see if we could improve their ability to shuttle into the brain. The last one was Arrowhead, which is a TFR-based platform, which also then shuttles in RNA. That is to target synuclein. That lead program actually is going to be, I think, really exciting because when you take an approach to knock down synuclein, one has to really consider the biodistribution. The ability to access the deep brain structures through the IV route is actually potentially going to be critical to get to the right structures in that disease. We have a lot of activity going on in brain shuttles. We're, in fact, super enthusiastic about it. I suppose the last thing, though, is, of course, Avidity is related to this. What we've learned with particularly TFR is that you can tune its properties to better deliver to either muscle or brain or other tissues potentially. That engineering is something that they are obviously very good at. We are very excited to see as that moves forward and how we can leverage that for some of the other efforts we're doing to bring these biologics and RNA into the brain. I often, though, will also say it's not like we're just done with that. You do three deals, and it's over. I think this is going to be an evolving field over the next 10-15 years. It is going to create a potential future where we have access at least to the technology of the blockbusters that oncology and immunology has had for a long time because we can access the tissue with things like biologics, for example. Maybe I'll just do the GLP-1 question. Sure. Yeah, I mean, I'm hopeful that result is going to be positive. I think it would bring huge medical benefit to patients who have a risk of Alzheimer's disease or who are already taking GLP-1s. I think if it is positive, it's going to open up a lot of really interesting components because there are three possible mechanisms by which GLP-1 could be impacting neurodegeneration. One is through a cardiovascular effect. We've known for a long time that cardiovascular disease is a risk factor and a major contributor to Alzheimer's. We've even discussed whether we should actually be looking at some of our own cardiovascular molecules in the context of Alzheimer's. That's a discussion we have. The second area, GLP-1, it does reduce inflammation, particularly in people who have obesity. That will provide further rationale for targeting neuroinflammation as a modifying treatment in Alzheimer's. Again, we have a lot in our portfolio already coming forward. It would increase that confidence in that approach. The other opportunity is that there could be a direct effect of GLP-1 agonist within the CNS. Certainly, GLP-1 receptors are expressed on neurones. I think, again, that opens up a lot of other opportunities for that direct targeting maybe through different types of modalities, low molecular weight molecules that could get in. Whenever we get a breakthrough in these difficult diseases, I think we all have to celebrate and think about how we can really build on that. I'll just go maybe third back, people who haven't had a chance to. Sorry. Hi, Zain Ebrahim, JP Morgan. Thanks for taking my questions. My first question is just to build on the remi question previously in terms of I think you mentioned one of the issues with the previous trials was the selectivity of the BTKs and the fact that they were not able to dose up. It sounds like you are confident in being able to show Kisqali-like efficacy on ARR because of being able to dose up. One of the other issues they had was also the lower-than-expected Aubagio annualized relapse rates, where I think it was quite significantly lower than ASCLEPIOS and OPERA. Just based on the patients you have enrolled, how are you thinking about Aubagio ARR rates? You are confident in Kisqali-like efficacy on ARR. How should we think about the level of efficacy you could show maybe on disability progression would be the first question. The second one is on disability. You started the trial in SPMS, but you haven't started one yet in PPMS for remi. Is that because of the consideration versus YTB323? Or is there something else holding that back from starting? Yeah. Norman, you want to? Yeah. We started the neuro programs with 100 mg b.i.d. That is not the case for all programs we are running in the neuro indication. We could do that because I think that is the only BTK inhibitor in neuro that has not seen dose-dependent toxicity in respect to phase I trials. I think we have seen in other programs with BTK inhibitors that there are different responses depending on the dose and the molecule. I think that just underlines the point that the chemistry makes a difference. I think that is pretty clear on both, I would say, phase II efficacy and then phase I, II, III kind of safety profiles. Our trial is designed to show Kisqali efficacy. We are looking at an event-driven design. We will monitor. At the point where we have sufficient power to detect it, we will read out the trials, which is going to be somewhere in the second half of next year. If the biology pans out, this is what we're going to look at. I don't know if the molecule is going to work. I think it is going to work. We modeled it. We didn't have a phase II. We will determine the exact efficacy there. The trial is also powered to show effect on disability progression at the same time. There is different power for three and six months confirmation. Generally, I think it's important these days to demonstrate to patients that you can suppress the relapse focal activity, which is the mainstay of the disease for many patients still. You want to hit on disability progression as early as you can, I think, in the treatment paradigm. We have seen with now anti-CD20s like Kisqali used a lot that I think on the one hand, early high-effective treatment is a determinant for better disability outcome. The reality is that 50% of patients are still not on high-efficacy drugs. There is still a lot to do across the entire spectrum. I believe that remibrutinib, if the plan pans out as we have defined the TPP for this molecule, can play a really important role going forward. We have just announced that we will do a non-relapsing SPMS trial, which will include a population similar to other programs that you have seen in that BTK inhibitor space. The trial is actually ready to go now. It is really super exciting that we could add it. I think we have changed our mind a little bit on this in the sense that we got a better understanding of what is the biology of BTKi. We have seen evolving better data, I would say, in terms of CNS penetration. Ultimately, with everything that was done in that clinical space now, we feel pretty confident that's something we should do. PPMS, lastly, it's just a small segment of patients. I think running a standalone trial in PPMS now is going to be really challenging. It's going to take really long to do it. Given that this is only 10% of the population, I don't know how much a second or third to market BTKi could still do. The last part is that maybe the regulatory framework is changing anyway. Maybe once we are through all this development with the BTKis, maybe the regulatory framework would also allow us different definitions of the population that is allowed to use the medicine. I think that's for the future to determine. OK. Do you want to go next, Michael? Thank you. Michael from Jefferies. Just going back to Avidity, please, the HARBOR trial in DM1. I noticed the secondary endpoint was changed about a year ago from a more general muscle strength and function endpoint to hand grip strength and then also more of a quantitative composite score as opposed to a more precise one. From a regulatory perspective, can you just talk to why that does not matter or may matter also for the interim readout next year? Thank you. Yeah. Who would like to do that, Norm? Yeah, I think you should. Yeah, go on. The trial endpoints. I mean, the DM1 program is really super exciting. We've been excited about this for many years. I mean, we got to this point after two years of work, ultimately, when we pulled the trigger on the proposed acquisition. We had the opportunity to review the data of the ongoing trial and obviously the previously published data very carefully. I think we have a very high level of confidence that what was shown in the earlier trial can be replicated there as well. The testing strategy, I think, should take into consideration what are the regulatory expectations in terms of showing an effect that is relevant for patients on certain functional markers and also patient reported outcomes. I think in that context, I would interpret the relative relevance of the functional measures that are included in the trial, like upper limb function, strength, walking. I think all of these things ultimately should show something relevant for the patient. You need to look into PROs to determine this kind of relationship. I think that would be key for the ongoing DM1 program and then also for the FSHD program later. I think that's the task at hand. In terms of the actual clinical relevance of the endpoint, you want to comment? I think, I mean, muscle strength is one thing. I think functional outcomes really matter a lot. I mean, there's a lot to say. I think going back to the first point, I mean, you're hitting the root core pathophysiology of the disease. You see a very strong signal in myotonia, which is one of the main measures. And it's reflective of chloride channel dysfunction. The rest of the events, which then drive muscle degeneration over the lifetime of a person, it's going to take a longer period of time. I think it's quite striking in a way that when you have a trial like this where the average age of a person in the trial was around 50, and you still are seeing a difference, at least in the early signals, in that time period. I mean, I think by knowing that we're really confident that you've hit the root cause biology, that you're seeing a very clear indication of that from what is basically a clinical signal, which is myotonia. Then seeing that play out in the functional outcome measures is something that we're very encouraged by. The PRO that they've used, the DM1 Active, is very driven by motor function and we think is likely very meaningful to them and is something that is a useful scale. Yeah, I think that. It's actually pretty striking that you see in a relatively small trial in a relatively slowly progressive disease early on a signal on a PRO that is related to some of the functional measures. I think that is a really strong signal for confidence that this is going to work out. We're really excited about that. Matthew, if you want to go to you. Hi, Matthew from UBS again. Two, please. One, a very quick follow-up on OAV101, which is just to clarify, the reason that this is multibillion versus normal Zolgensma today is because the prevalent population is bigger than the incident population. Is that kind of the way forward? We have a big bolus of patients we can treat now. Over time, it will shrink back down again because there is the Zolgensma patients treating newborns. We will not reload that patient bucket. Am I thinking of it right? Correct. Yes. If you think about the gene therapy population, you will have a high prevalent population that over time will get depleted. Then effectively, you're left treating the incident population. Yes. Brilliant. Absolutely fantastic. The real question on Avidity. Harry's made clear one to two percentage point margin diminution because of the need to spend on R&D at Avidity. If we look at what Avidity analysts had in, then that's a multiple of what was assumed for the R&D spend at Avidity. It's just a simple question for the scientists that are going to be spending the money. Is it that basically previously people doing our job got it wrong and Avidity was just going to spend a ton more money than we thought they were? Or you're going to do something differently that means you're going to spend a ton more money? I'm glad you uncovered this because it sounds like we have more money than I thought to play with. That's good news to me, actually. I think I'll start with more of a scientific framing. I don't know that I can speak to the financial framing of it. I mean, with this excitement comes a complex novel format. We are able to deliver with this antibody oligonucleotide conjugate. We have to figure out exactly how that works. We have to see what it can do and see its ultimate potential. I mean, I don't think that I can comment on how that would affect the margin. Yeah. How would you? We had to pay Avidity in the first place, which they did not have to pay themselves. That is one difference at the start. I mean, obviously, we want to make a success of this. Whenever we do a deal like this, and again, it is a bit of a tip of the iceberg. These three are just the top, but they actually have a lot of earlier programs and the opportunity as well to take this technology into other disease areas. I think that we want to invest the success and make sure we also do the type of trial we would do as Novartis and also take it globally. That is a different strategy from what a biotech company would have. We'd be happy to spend less if. Yeah, of course. Maybe I will add one more thing because I do think as we bring the programs in-house, we'll get a much clearer picture of exactly all the contributions. The one area where we know there will be an incredible amount of synergies is in commercialization. If you actually think about the skills, the capabilities that we've built today with Zolgensma and SMA, all of these indications stack right on top. It's the same treating physicians, largely the same treating centers. We will be able to expand and leverage a lot of deep capabilities that we already have in rare diseases. That is one good part of the synergization, the synergies that we can get from this deal. Synergization. Hey, we've got one more question. Question from the web, maybe the last one. Oh, OK. From Joe Kim. Are there any mechanisms that can be considered to accelerate the potential approval timeline for the Huntington's program? Maybe we use this as the last one to close. OK. Yeah. Norman? Yeah. I think I talked to this a little earlier. I think we absolutely will exploit every opportunity that there is with evolving data. We work very closely with the regulators. I would say the interactions and the dialogue that we have with the FDA has been really constructive. I think we have a very good idea of what the expectations are. With the phase II, that was controlled two dosages, stage two, three patients with an ongoing open label extension, accumulating data there with some more understanding of how does the Huntington reduction lead to improvements in terms of disability outcomes for patients. I think while we generate all that data, there will be opportunities for at least the exploration of accelerated approval. I think it's pretty obvious that the regulatory agencies and Novartis, we're on the same page when it comes to the unmet needs that we want to address. I think Huntington's disease could be one of those perfect examples, actually, where with evolving science, you can get those medicines to patients faster than in the past. OK. Thank you. I'd like to thank the panel and thank all of you for your excellent questions. All right. Welcome, everybody. This is the final session for today at Novartis's Meet the Management. This is the corporate session. Joining me today are Ronny Gal, our Chief Strategy and Growth Officer; Harry Kirsch, our Chief Financial Officer; Steffen Lang, our President of Operations; and Mukul Mehta, our Incoming Chief Financial Officer. For this session, we'll focus on the overall long-term strategy of the company. As you heard this morning, we believe we're well positioned to continue to build on the strong results we've delivered over the recent years with an attractive growth profile, the upgraded guidance of 5%-6% through 2030, and a robust pipeline with 30 high-value assets advancing in a significant number of late-stage readouts over the coming year. With that, we'll open it up to questions. Florent? Thank you very much. Florent Cespedes from Bernstein. Two questions, please. First, on M&A. We've been pretty active over the last couple of quarters on M&A. Do you still have some, let's say, assets that you'd like to add to your pipeline? Late stage, it seems that you have a pretty full pipeline. Early stage, any color on which areas and which stage assets you would like to add to your portfolio? Second question, a couple of months ago with Paul Hudson, you sent a letter to the European authorities or published a letter about the fact that the drug's value added is not really recognized by the payers. Could you maybe share with us your interactions with some countries in Europe and if they have changed their mindset or if they agree, but they cannot do anything due to the budget constraints? Any color on that would be great. Thank you. Absolutely. Thank you, Florent. First on M&A, I think our strategy remains unchanged. We continue to look for high-value assets in the four core therapeutic areas, assets that fit within our technology platform, not limited by size, but really focused on assets that can improve the growth profile of the company over the medium and long term. The way we look at it is we were very confident, even without additional deals beyond the normal deal inflow, that we can deliver the guidance that we've set forth. We more look at it as any additional deals that can further de-risk the profile and also add to the strength we'd like to build in individual therapeutic areas. I should hand it over to Ronny to provide any more color he would like on the M&A. First, great to see a Bernstein analyst here. Second, the partnering in M&A areas, focus on our four core therapeutic areas. You mentioned the assets either late preclinical or phase I transitions. There are generally multiple mechanisms coming through on each one of our four therapeutic areas on all four. What we are primarily aiming for when we look at those are assets that provide a significant amount of novelty. Either first-in-class assets or later-in-class assets, but with a clear claim what they would be best in class. Those will be generally the focus. We are excited about bringing more neurology assets in. There's a few additional diseases we'd like to target with either RNA or gene therapies. We are excited about new mechanism of activity in cardiovascular. You've kind of seen the mechanism we are focusing on. There are additional risk factors that are generally coming into focus. In immunology, there are additional multiple therapeutic areas, additional multiple mechanisms that are applied to multiple diseases that are generally coming in as people map the progress of the various diseases. There is a big effort around there. Which therapeutic area? In oncology, there are always additional areas coming in. We are looking either for additional targets primarily, essentially cell surface molecules not yet identified, or mechanism-based efforts that are not, with a particular focus on prostate and breast, but also looking for additional therapeutic areas where there might have been a fairly low standard of therapy where patients progress within a year or so on their first-line therapy, where you can actually get a molecule that is going to extend the duration of therapy significantly. Those will be the primary things we are interested in. On Europe, we are clearly in a situation now where as more companies move forward with agreements, I would say for Novartis, we're engaged with the administration, no specific timeline, but continue to have good conversations on a potential agreement with the administration, that this will fundamentally change how the market's in focus. Likely Europe more broadly has to look at new medicines. I think it will make it challenging for companies that have medicines launching within the scope of the agreement to launch if net prices in Europe do not approach the net prices in the U.S., except in the private market where, of course, companies would make their medicines available. Now, we've made that clear, and I've made that clear, but I think the industry overall has made that clear to key policymakers in Europe, as well as at the European Commission. I think there's the beginnings now of understanding that the approach in Europe needs to shift in how medicines are valued, as well as some of the barriers that exist to growth. These are clawbacks and other taxes that are applied to excess growth within markets. I also think that we won't see rapid change. I mean, I think there's an awareness. I think you'll likely see some medicines not launched in the public market. You'll see awareness growing in the public that certain cancers or other diseases are not well treated relative to the standard of care in other parts of the world. That hopefully will start the change that we hope to see on how medicines are valued. Matthew? Thank you, Vas. Matthew Weston at UBS. Two questions from me, please. I think one thing that's been debated a lot today has been your comments on Cosentyx and not raising the peak sales expectations, which was something I think a lot of people anticipated. Can you just clarify a little? Is it that you expect us to see growth slowing in the near future, or is it just that you want to make sure that if we're going to get more excited about remibrutinib and MS, we can't double count two new MS therapies that would have an impact to 2030? And then one for Harry, the Avidity deal, and I guess all the string of M&A that we've seen, share buybacks have been a critical feature of Novartis or critical. They've supplemented earnings growth at Novartis over the last few years. Should we think of the buyback slowing down in 2026, 2027 in light of the Avidity deal, or there's more than enough cash flow to do both? First, on Cosentyx, and just to clarify, the brand is doing extremely well. You see that in the quarters. You see that over the years. That growth looks very strong. I think in the context of two things, we would like to see how things evolve in the next year or so before thinking about any further upgrades. One, we will understand better the BTK inhibitor data, both from our competitor and for Novartis' own medicines. Then, based on that data, be able to see what role BTK inhibition will play at the end of the decade and into the next decade for multiple sclerosis patients. Then factor that in. I think that is a prudent thing to do. The second is the expansion of the B cell class, particularly in the U.S., where we need to see the class share continue to grow for the market to be large enough to hold all of the B cell monoclonal antibodies and the orals. Those are the two things we'll be watching. It's not meant to be any concern about the brand. I mean, the numbers speak for themselves, and the momentum is really strong. I think we just want to be prudent as we have more milestones come out over the course of the next year, and then we'll reevaluate at that point. Harry, in terms of. Yeah. Cash flow, share buyback, M&A. As you know, we have significant EBITDA, $23 billion roughly at the moment. Our cash flow last year, $16 billion, this year, $16 billion after nine months. Clearly on its way to $18 billion-$19 billion. Over the years, two, three years, clearly around the $20 billion range. We have significant cash flow and a strong balance sheet. With that, the Avidity deal does not take away from our ability to also do share buyback and some further bolt-on M&A. The share buyback we have announced in the summer, it is like $10 billion over two and a half years. Not to give you a formula, but it is like $4 billion per year, roughly. That is what we have historically done on top of buying back employee participation shares to not dilute our shareholders anyway as a base. We had a few years where we bought the Roche with $15 billion in like one and one-half, two years. I think what we have now announced is probably more what Mukul and Vas will discuss with the board, but it is more kind of normal level, which we had before the Roche divestment and also see now in this range ± $4 billion a year. Not super high, but also always part of our capital allocation. In the end, it is also not a strict formula. It depends on how much Ronny and Vas find in terms of bolt-on M&A. We see it as, I think, a regular part of our capital allocation program. Yeah. Thank you. Simon, and then I'll go to Steve in the back. Thanks, Vas. Simon Baker from Rothschild & Co Redburn. Two, if I may, please. We've had a huge amount of debate recently about what we pay for drugs. The debate on how we pay for drugs has gone a bit quiet in recent years in terms of novel funding mechanisms. The reason I ask is if your efforts in autoimmune cell therapy are successful, that could be an interesting debating point because you're going to have potentially huge economic advantage over time, but a bigger upfront cost. What does that mean for budgetary constraints? Just where are we with the debate and appetite for looking at different ways of paying for things? Secondly, going back to the slide you talked about this morning, Vas, on the impact of AI, it's always refreshing to see someone talk about the impact of AI on pharma who isn't trying to raise Series A funding and consequently massively overselling the opportunity. I just wonder if you could give us some ideas of where you see the biggest bang for your buck in terms of saved time or money or both by the application of AI. Thank you. Yeah. Thanks, Simon. First on payment models, we do have a lot of experience on these novel payment models between Kymriah and Zolgensma. We've had the opportunity to roll them out really around the world and seen the challenges that are significant challenges for doing this. There is often openness to do them as well, particularly for one-time therapies. I think that with CAR-T in immunology and the whole immune reset portfolio, and given that depending on how big that gets, I do think there's an opening now to take that up again. I think one of the things that's clear is that you need a big enough volume of patients and a big enough cost to make the system really want to take on the complexity. I think for one-offs, it was really hard, I think, in the end because it just isn't worth tracking all the patients and figuring out how it works. I think there is the potential for immunology. I think the other thing I'd say is if we do continue to see the levels of remission or significant reductions in disease activity, I would expect the cost-effectiveness to look really good for these therapies. You'll remember Kymriah had very good Zolgensma and very good cost-effectiveness. That will mean that for most health care systems around the world, there'll be a lot of interest in doing this rather than long-term chronic care or hospitalization for patients that are on the end stage. I think the health economic case will be there. There will be the budgetary topic, but I think we can manage it. I think on AI, and I'll hand it to Steffen to also give some more operational examples, we see three parts of the AI story at Novartis. One is just generally enabling AI to be used by the associates across the world. We're one of the largest users of Copilot. I mean, we have Copilot available. We also have the coding support available for the coders within the company, which isn't a huge population, but an important population. Second, we do a lot of investment, as I noted on the slide this morning, on R&D productivity. That's both upstream and research, so target identification, but more importantly, candidate optimization, supporting candidates through preclinical and then optimizing clinical trials and all of the elements to accelerate the clinical trial process as well. Again, very interesting. I think we'll pay dividends in the 5 -1 0 years. Then, of course, all of the support around that in terms of clinical trial enrollment. When you look at that overall, on average in our industry, it can take three to four years from identifying a target to getting into the clinic and eight years from getting into the clinic to getting to an approval. Can we compress that timeline by two years, four years with AI? I believe that's possible over time. Then similarly, can you get to higher quality medicines, lower POS, better medicines from a production standpoint? All possible as well. I think we have to see over the next five to ten years how that plays out. Where I do think we have a significant opportunity today is the third bucket, which is more just applying AI to the core processes, not super advanced AIs, machine learning, or anything like that, but very established tools to optimize how we work. That is, I think, an opportunity. Maybe Steffen, you can talk about it for manufacturing and elsewhere. Yes. This is for sure one of our key enablers for productivity that we have a huge automation program across all manufacturing sites and the functions supporting manufacturing. We augment this automation with AI where it makes sense. There are more and more examples where already today this generates a lot of savings for us. For example, when we look into biotechnology, drug substance manufacturing, where we use living cells, here we use AI technology to help us to optimize the process conditions and to always deliver the higher yield batch week after week. Across a normal facility, this is easily $10 million-$15 million savings in a given year. Other examples are in IPC in-process controls for aseptic products where in the past, with the human eye, we have done optical control to see are there particulates in the medication or not. This is now done fully automated with camera images and applying the right mathematical models and learnings. This is now done at an accuracy better than the human eye to distinguish between air bubbles and particulates. Also here, nice savings year over year in the facility. This is also very good for our network, our manufacturing network, which is focusing on high tech and automation. As I said, this is every year a larger contribution to our productivity savings. In the operations, we see opportunities within our third-party spend, which we're now really trying to tackle because I think clearly with AI and some of the tools our associates have, we can get after the roughly $13 billion we spend every year of addressable procurement spend. That third bucket is where I think you could see an impact in the five-year horizon, not have to wait for the long run. Steve? Thank you. Steve Scala from TD Cowen. I'd like to follow up on Matthew's question on Cosentyx and maybe the less clear future. I apologize if you mentioned this first one, but visibility on a biosimilar. Does Novartis know that there are biosimilars in development? Secondly, these limited duration academic studies that are looking at kind of treating and stopping, that would certainly maybe curtail Cosentyx's future. That is the first question. Second question is there's this view that companies pursuing deals with Trump have some side agenda, like huge tariff exposure, and/or they want a special voucher, and/or even other things. Novartis presumably would have no special side agenda. Does that mean that it is very likely that you will not sign a deal? Companies all tell us, well, we're in discussions with Washington, but presumably that has always been the case. That is not new. What's new? Thank you. Yeah. I think first on Cosentyx, we do know there are biosimilars in development. We do also expect that they will take advantage of the more abbreviated approach to development. We still do not change our outlook that we do not expect significant biosimilar exposure for Cosentyx in the U.S. before 2032. We think entry perhaps in 2031. That is kind of our current outlook. That is not a driver. As I said to Matthew's question, we are very confident in this brand and how it is performing. We will see next year again how these various elements unfold. We can look at the peak sales guidance from there. I think on the discussions with the administration, there is nothing new that we can report publicly. To your question, there is nothing new. We continue to have very good discussions and continue to advance those discussions. There is no set timeline for the resolution of those discussions. We do not believe we have significant tariff exposure given all the investments that we are making, the various things Steffen and the team have done in the supply chain to get inventory levels to where they are and our overall approach. We also, of course, would like to ensure the company is well positioned with the other policy levers the government has. We will just see how these discussions unfold. Maybe at the back, and then I will come back over to Michael. I cannot see who that is. Hi. It's Ben Yeoh at RBC Global Asset Management. One on AI and one on the board. Just to follow up on the AI question, on this third pillar, like you say, $13 billion in spend, and we're beginning to see impacts now. Is that broadly envisaged within your guidance and thinking about margin? Could there be upside if that goes quicker? If that does happen, are you thinking we just reinvest that because then we can grow and make better decisions? Will some of that potentially come up in margin upside in that in terms of AI? At the very high level, I'd be interested, what's the most kind of constructive challenge that you get from the board? In that strategic thing, what is the board saying about the direction that the company is taking? Thanks. Yeah. Thanks, Ben. First on AI, it's not embedded in our numbers. I think we see it as more of a way to free up more resources to invest in R&D, invest in potential deals, and continue to bolster the long-term profile. We feel confident we'll get to the $40 billion to 40% margin just with the plan we have. This is more just a way to free up resources to be better deployed for better returns for the organization and then ultimately for our shareholders. I think the biggest topic on the mind of our board, rightfully so, is actually when you think 10 years from now, how do you ensure that a company is set up of the potential size of Novartis, which depending on how you run the growth rates is significantly above $80 billion in sales at that point, to be successful at that scale? I think what does that mean for R&D? What does that mean for your commercial setup? What does that mean for a whole series of questions, which I think is a very important question? I mean, when you have a company of that size, but in this sector where you're always having to deal with genericization events, et cetera. I think that's probably the biggest topic on the long-term agenda of our board. Thank you. Michael? Thank you. Question for Harry. Can't let you get away. Just margin structure. If I think about products that may have a little bit more in terms of fixed costs looking forward, is Pluvicto less profitable now again as you roll it into the community setting and then maybe also the urologists? Are there any other products that will have to take a little bit of time before they become meaningfully profitable? Yeah. Pluvicto actually, I mean, Steffen does a wonderful job on his team with the higher volumes, also the cost of goods come down. Now it is a bit lower in the margin structure given there's also royalties attached to it. We also have less marketing and sales, more service model. Overall, Pluvicto does contribute to our margin, maybe not as much as small molecule high priced or biologics, but overall very profitable, very value creating business already. Please go ahead. Thank you. Thibault Boutherin, Morgan Stanley. Just a vast question on your ambition for meeting a digital sales growth beyond 2030. How do you think about the mix between organic and what we'll acquire between now and then? Consensus has sales declining from 2031. How much of the delta is a function of pipeline and new launches contribution that we are missing? How much is a function of potential additional BD between now and then? Yeah. First, I'd say it's not easy to know the difference between all of you because while the view is there's going to be a decline, none of you really model it all that carefully for me to be able to bridge it. It's a little bit harder because you guys just have this drop off, but not actually giving us any of the pipeline. What I would say is what we see is we see our in-market, we take the in-market brands that have protection in the late 2030s, and we see the potential of that. We take our probabilized pipeline. We have a probabilized inflow model, and we have a probabilized BD&L model based on all our historic output and productivity. Based on that, we feel very comfortable that with that, without doing any additional deals per se, we can be in this 4%-6% mid-single-digit growth range. Additional deals that we do would just de-risk that profile. I can't actually give you individual brand level. I mean, when you look at the peak sales, I think we're largely in line with all of you. You guys all have our peak sales on the in-market brands. You've seen that we have a lot of ambition for Rhapsido. We have ambition for ianalumab. These are both relatively, or at least more de-risked than they were, given that you have the positive readouts. We've discussed there could be more upside on Cosentyx or Pluvicto, depending on how the next year rolls out in terms of the trajectory on those brands. On the remaining brands on that second page of peak sales I have given, that really depends on the readouts and the early trajectories before we can give you precise numbers. We clearly think these are all going to be significant medicines with the multi-billion dollar potential. Those are, of course, all phase III assets on that chart generally. Matthew? Thank you. Two quick ones, please. One brand that isn't on your table is Entresto. I would love to know how much Entresto you think you'll sell in 2030, given the longer Japan patent, your experience of cardiovascular drugs in emerging markets. Are we getting that one wrong? The second question is around Kisqali Lifecycle Management because it's a phenomenal brand for Novartis. In the opening comments this morning, you suggested you thought that if SERD had a role, it was likely going to be a rolling combo with CDK4/6. As the company that is currently dominating CDK4/6, I would imagine a fixed dose combination could be a very attractive way forward. I'd just love to understand how you're thinking of an oral SERD strategy, how early you think you may need to act if that's a useful thing, or am I just wrong in terms of whether the strategy would work? First on Entresto, I would say we expect it still to be a multi-billion dollar medicine at the end of the decade. We do not want to get too precise. I think between the potential of hypertension in Japan, hypertension in China, and then just in general, when we look at the stickiness of our cardiovascular portfolio in emerging markets, Diovan, the whole Valsartan portfolio only fell under $1 billion relatively recently. It held Exforge, Diovan, Lotrel portfolio held at $1.5 billion. It still may be even. It's over $1 billion. It's still over a billion. You can imagine then when you look at your Entresto forecast, it's important to consider we'll have protection in two large markets to some extent in China, and certainly in Japan. You have this long tail for such an established brand. Multi-billion, we would expect that period. On Kisqali, we certainly acknowledge the biology and the importance of oral SERDs. We'll have to see again how the data plays out. I would also say, because I think there may be some confusion over the course of today, there's no lack of effort at Novartis to try to find an oral SERD that would fit in our portfolio that makes sense from a whole variety of reasons. It's not from a lack of effort, certainly not, or from our own internal efforts. We will continue those efforts. Yes, we would like to, of course, have an oral SERD that's highly attractive and that makes sense for the company. That said, in the meantime, what can we do given the situation we currently have where our internal programs didn't pan out and some of our early partnerships long ago had safety liabilities that we were working on is generate data with Kisqali in partnership with the available oral SERDs. That's what we try and do so that data is available. We become a partner given our very strong position of leading CDK4/6 in metastatic and in early breast cancer, where clearly logical choice in our view with an oral SERD or an alternative hormone therapy that a physician might use. Steve? Following up on the SERD question, there's some confusion about what your rights are to the Olema compound. Do you have rights of first refusal on that compound, or do you not? Secondly, Harry, since we may not get another chance, what is your biggest concern for Novartis' future? What is your biggest concern for the industry's future? Thank you. Steve, I can't comment on any further details on the Olema collaboration that we have other than to say our goal is to generate data with Kisqali and the Olema asset as well as other assets and Kisqali. Harry? One thing I would add on this one is we do not have any economics on that asset as it currently stands. I think there was some confusion about that before. Yeah, please go ahead. I mean, why wouldn't you just say you don't have rights? I think Ronny. We'll get back to you on this one. Yeah. I was not sure. Your biggest concern questions clearly are set up. Just to comment, overall, the way we have set up this company over years as a team, Pure-Play Pharmaceuticals with two, I think, very attractive spends and a good capital allocation, I think is very good. The replacement power with our internal pipeline, as well as with the default on M&A, is shaping up extremely strong. The only concern is that the sales side continues to underestimate our growth profile. That is OK because we will grow into it. Later on, people will realize it. For the industry, I would just hope that the European countries, or these eight countries in the basket of the MFN, realize that it is time to wake up to basically ensure that patients have access to excellent new medicines. Think about how high the generic prices are and how low the generic penetration is in order to have an innovation-rewarding health care system also in the future. Those would be the two things. Michael? Thank you. On the slides this morning, slide 10, you show the boosted R&D productivity. The median peak sales per projects that you show are up 25% or 22% by 50%. The expected NPV is only up 25%. Just wondering where that big gap's coming from. I would have thought revenues would drive a bigger uptake in the NPV. Maybe I start with a technical one as you think about the strategic one. Yeah. No, some of this is, of course, the peak sales come very late. You have a time value of money thing, which disconnects a bit the peak sales increase from the NPV increase. That is one of the technical explanations, if you will. Yeah. I mean, I think, and also when you--it's time value of money. These are all more mathematical topics. I think for me, the point we were trying to make rather than get is that we've made a significant shift in the company from where we used to develop a number of assets that were even sub-$1 billion asset peak sales. And now our threshold is $2 billion. That's what we want to take forward from an asset standpoint, indication we want to get at least $500 million for an asset that we already believe has that potential. We have shifted that thinking all the way back into research across the entire portfolio. That's actually bearing fruit where we're more selective and focused on much higher value assets. The other thing going on here is, of course, we do have more outcome studies. I think outcome studies from a cost standpoint will also suppress the NPV. The other one is additional indications. The way the assets become bigger is that you actually are testing one indication, testing multiple ones. That drives a higher cost structure. I think Simon. And then who was next? And then rest of you. Matthew. Thank you. Yeah, just two quick ones. Is it possible over time that the impact of MFN on European pricing actually has a positive impact on margin profitability? Secondly, just going back to Olema, I know you're not going to answer it. If I sort of ask it a different way, are the Olema SEC filings that say they have granted you a right of first negotiation a fair representation of the situation? First on the Olema, I think at this point, we probably shouldn't comment any further until we look carefully at all the questions you guys are asking us. What I can say is we are generating data with the asset, and that's what we're focused on at the moment. Now, I think on the question of how European pricing will evolve and how that will impact, I think it's too soon to judge. I think in theory, one could arrive at a place where a private market launches where you have higher margins, of course, because you're closer to the U.S. price, will even at lower volumes lead to perhaps neutral profitability. I'm not sure you'll get more profits because I think the volumes, of course, will be quite low given how these markets are structured. I am also hopeful that in some markets where when you look at the net price and you do a GDP adjustment per the calculation for the five countries, companies that have already made agreements and to what we understand of those agreements, that you can actually bridge the gap and that some of those countries will work to bridge those gaps. I mean, certainly when you look at large markets like Germany and Japan, and for certain categories, bridging those gaps might be doable, in which case those markets will become more attractive as well from an economic investment standpoint. All of that is going to have to work through the system. I think it's not going to be fast. I also think it's important to note that it will be for new launches and then new launches in what we understand at least certain categories. That is when we'll see step-by-step governments starting to assess how they're going to manage this. Then we'll have to see from a budgetary standpoint what they're willing to do. I think the goal of the U.S. administration absolutely is to ensure that these eight countries contribute and then the knock-on effects in other countries to the innovation ecosystem of this industry and then appropriately value the innovations to support future discoveries and future breakthroughs. Matthew, and then I'll go to Naresh. It's actually a manufacturing question. It's all to do with re-onshoring in the U.S. and the announcement earlier this week of the significant investment in North Carolina. I'm just really, the industry has not had a significant footprint of manufacturing in the U.S. for many years. Everyone's now building. It seems that most people are building in North Carolina. I'm just curious as to where the people are going to come from. I know it's a number of years away. Is this an issue where, as an innovative industry, you have to bring people from around the world to bring that expertise into the U.S., or you train them up? Are we going to have an issue that everybody's chasing the same people in a number of years' time to get these facilities up and running? Thank you very much for the question. We already have today a significant manufacturing presence in North Carolina, our Durham facility, which we have up and running. We have no issue attracting talent and also retaining talent. What we need to do and need to take into consideration now is the footprint we add to build that flagship manufacturing cluster there. We will, of course, deploy highly automated technology, the one we use in our other facilities. We will not need a workforce of 10,000 people. Our overall estimation over the next few years is that we will have around 1,000 people there to basically be able to produce all key products end-to-end in the U.S. As such, the profile might be a bit different. Some of it is training. We are very confident based on the legacy experience we have that we will find the talent. Naresh? Thanks. Just another question on MFN. The deals that have been announced so far have had different dynamics by channel. What do you think the fallout is in future payer negotiations that we may not be thinking about in terms of the differing dynamics between cash and commercial and Medicaid, et cetera? Just how do you think this changes things? I think the biggest shift from that standpoint is opening up and perhaps the broadening of the direct-to-patient channel that clearly the administration, but I would say other companies, including Novartis, have opened up on our own. We have a direct-to-patient approach. Other companies have as well where basically netting out the typical rebates paid to PBMs is now being offered direct to the patient. TrumpRx, as a separate platform, is offering "mfn-level pricing" also direct. Now, how that ultimately unfolds remains to be seen. I think for certain categories of products, that might be attractive if you're in a high deductible plan and you'd prefer to just pay out of pocket. I think one open question, of course, is how does that payment apply to your deductible? Can you apply it to your deductible in Medicare or outside of Medicare? Open question. Probably in higher-priced specialty areas, even at that net price, it's going to be unattractive or hard for many patients to use it. I think there'll be relatively low volumes. I'm not sure at this point until we have more data on how these direct-to-patient platforms are used outside of categories like obesity, where I think there are other companies that would be better to comment than us, how big a shift it's going to lead in the payer environment. That volume and how that volume moves is what's going to change how I think PBMs might behave. It's in the back. Hello. Zain Ebr ahim JP Morgan. Thanks for taking my questions. The first question will be on Scemblix, which you raised the guide for today to $4 + billion. Just in terms of thinking about the launch in the U.S. has been really strong. Ex-U.S. looks like it's off to a really strong start as well. Thinking about potential upside to the guide in terms of what the gating factors are to get to Gleevec size or Gleevec plus, particularly with the ex-U.S. rollout being as strong as it looks like it has been. The second question is on ianalumab, where it sounds like you've had a consistent benefit in the subgroups of the U.S. population and also the OUS population. If you could remind us on the timing for OUS filing and approval and how you're thinking about the launch trajectory in the two markets. Yeah, first on Scemblix, the key gating factor now for how much larger it can get beyond the $4 billion is how successful we are in, let's call it, either markets or in physician offices where there's a high loyalty to 1st-gen TKIs, namely generic Gleevec. I think you have segments like that in the U.S. You also have countries and segments. Countries might just be more focused on 1st- and 2nd-gen generic TKI. Our ability to move based on the data that we have that shows superiority to Gleevec in the first-line setting and at least numerically better responses against 2nd-gen TKIs with a strong safety profile. I think the other thing that could help us is over time, following the cohort of patients in the pivotal studies, showing outcomes. I think as we get closer to showing that not only do we show better response rates, but the deep molecular response or major molecular response, do we actually show better outcomes. That better outcomes could sway, I think, more payers outside of the U.S. to switch to Scemblix because then there's a cost-effectiveness analysis as well. Given the pricing difference of Scemblix with historical Gleevec, also the volumes needed to surpass Gleevec are going to be lower. That also is important to take into account. With ianalumab, we would plan to file next year, I think, as we've already stated. I think outside of the U.S., one, it's important to note, depending on how we progress on the agreements with the administration, this would be a brand that would be subject to any agreements that we make. I think in the eight target markets of the administration, the launch here would have to be thoughtfully done, let's put it that way, in terms of being able to launch the medicine outside of the private market. I think more strategically, the question is in countries where there is an unwillingness to move off of systemic steroids as their pricing base case, it's going to be very difficult, of course, to launch the medicine. We would be much more focused, I think, in countries where there is a more attractive pricing environment or, of course, the China market is substantial for Sjögren's. While there is local competition, we still think given our overall presence in immunology, there's a significant opportunity for ianalumab in China. Michael? Thank you. Vas, you made an interesting comment this morning about how IRA might spill into the commercial channels as well. I just wondered in terms of speed and how quickly you think that might happen as prices come down in the government channels. Yeah, and just to clarify, it was less the commercial but more the Medicaid channel. Basically, in the legislation, when a drug goes into IRA negotiation and the price for the so-called MFP price is set in Medicare, that does impact Medicaid best price. Even for patients that are not elderly, you have this rollover effect. For drugs with low Medicaid exposure, not going to make a big difference. For drugs with high Medicaid exposure, you now basically have to provide this lower price in that segment. That would be relatively quickly because it's not volume-weighted. This is going to be a relatively quick impact. Again, it will vary product by product. I think what would happen is you'd have a reset down in the first year. We would grow back up off the base, off of the reset base, primarily driven by the commercial market where we would have not these lower prices. That is kind of the dynamics one would expect in the IRA. A question on the webcast? No? No? OK, Steve? Thanks. Two questions. Would you be surprised if no competitor ever came close to Novartis' success in RLT? Would that surprise you? Secondly, should Kisqali be successfully developed as a fixed-dose combination, how would IRA pricing be impacted? It was said earlier that your new guidance for Kisqali assumes IRA pricing. Can that become maybe not an issue if you developed a fixed-dose combination? Thank you. I guess I'm trying to get the way I framed the answer because the way the question was framed, I would be surprised if a company in the next, I would say, five to eight years approaches Novartis to scale in RLT. Did I get the question right, Steve? Yeah. Because of the investments we've made since 2019 in R&D and capabilities in the manufacturing supply chain, in the ability to deliver the medicines globally, I think we're now 99.8% on time and full, regardless of where we ship. The scale and the learnings we've made in that effort, I think, are substantial. I think that would be difficult to replicate. I wouldn't say it's never replicatable because I think certainly we have very capable peers who could make the investments to do it. It takes experience in the market. We've accumulated a lot of experience in the market, also now reaching community physicians in the U.S., but also navigating the individual complexities of each market. Each market, whether it's Germany, France, China, Japan, has different nuances to how to roll out RLT. That's all accumulated know-how we're building with each launch. I'm looking to my team. To the best of my knowledge, a fixed-dose combination would not avoid IRA. Am I looking at my U.S. team? Is that correct? I think so. I think if it contains the same active ingredient as the line, you can check the language. I think if it says if it contains the same active moiety, you are included. We'll get back to you, Steve. I think it's a fair question. To the best of my knowledge, that wouldn't change the situation for Kisqali. I think for that, is that what your question? Or for the new combination, is that your question? Yeah, yeah. Merck seems to think it would because they're KEYTRUDA. Combination with the to make it a sub-Q. Yeah. Let's say it's an open question. Maybe we'll leave it at that. We need to understand it better. I think last question, Simon. I think we're out of time. Thank you. What better way than to end on a question about tax? If I look over the last 10 years, the Novartis tax rate has been between about 14.5% and just over 16%, which is very, very stable and pretty low. A lot's changed over that. A lot's changing going forward with Pillar 2. How should we think about the tax rate in the medium term? Do you expect it to stay broadly in that range? Is there any significant headwind or tailwind that we should be aware of going forward? Yeah, I would expect it at the moment we have a core tax rate, which anyway is an interesting concept. We do a good job of that. Core tax rate, we try to align better with cash tax rate. While the IFRS reported tax rate is interesting but not very helpful. It has to be correct from an accounting standpoint. The core tax rate is more meaningful also for cash flow forecasts. It is at 16.2% at the moment. It goes up a bit to a 17% range over time. It depends more on geographic. Can be between 15 and 1/2 and 17, actually. Depends on geographic mix. Maybe, long story short, I expect to have continued attractive tax rate. Mukul, what do you think? Yeah, I think that's what in the medium term, Simon, we are forecasting between, as Harry said, 15.5%-17% would be our forecast going into 2030. All right. Thank you, everybody, for the full day. We really appreciate everybody's interest to listen to all the sessions. I also want to thank this room for asking all the great questions for the webcast. Probably leave you with just a sentiment with all of the details that we've gone through. Novartis is highly optimistic about our outlook and our future, whether it's our in-market brands, launches, our pipeline. Probably the most important thing I hope you take away is we have great people. Those people are incredibly talented and dedicated to the long-term success of the company. That is also what gives me great optimism about our future as well. Thank you again for a great day. We'll look forward to catching up with you on Quarter 4 and the full-year results.