GSK plc ($GSK)

A warm welcome to this GSK call on our agreement to acquire Nuvalent. My name is Constantin Fest, Investor Relations. I'm delighted to have here today with me Luke Miels, CEO; Nina Mojas, President, Global Product Strategy; Tony Wood, Chief Scientific Officer; Julie Brown, our CFO. Also for the Q&A part of this call, we will be joined by David Redfern, President, Corporate Development; as well as Mondher Mahjoubi, our Chief Patient Officer. Please go with me to the next Slide 3 for a disclosure statement. Also note our cautionary statement on Slide 4. With this, please turn to Slide 5, and I will hand over to Luke to start this presentation.

Thanks, Constantin. Good morning, and thanks for joining the call at short notice. Look, I'll start here first. As a reminder, this is the framework that we're using to drive value for patients and shareholders. It's got 3 components: so driving top line growth, accelerating late-stage assets and combining this with simplification. And this deal is a disciplined continuation and acceleration of that strategy. Next slide, please. Now we've been following Jim and the team at Nuvalent and their impressive medicinal chemistry work for some time. The deal is a consideration of USD 10.6 billion. It's accretive to EPS in 2029 and is expected to close in Q3 of 2026. This acquisition critically will bring multiple products, with the 2 lead assets already filed and expected to launch in 2026, immediately impacting top line growth. Next slide, please. So here's the logic for the deal. Simply, it's on strategy and helps us deliver top line growth and accelerate R&D. It accelerates our planned entry into lung via 2 well-defined subpopulations, which have clear unmet needs. Both products work via validated targets and are designed to address these established unmet clinical needs. These are small molecules in oncology. So the deal combined with measures to offset costs is accretive to sales and operating profit in 2027 and EPS in 2029. And through discipline, we are maintaining the 70p dividend and the progressive policy. Next slide, please. Finally, just at the start, I would add, this is our type of deal. The difference is that it's a multiproduct deal versus a single asset, single company deal. [ zidesamtinib ] for ROS1 mutations and neladalkib ALP-1-driven tumors are the types of products we like. And they're an extension of our BD strategy, which you can see on this slide, which is to acquire external innovation that works by established targets that addresses unmet efficacy or tolerability needs. So with that, I'll hand over to you, Nina.

Thank you, Luke. As Luke mentioned, Nuvalent is a company that has been very successful in developing precision medicine assets with its clinical stage assets focused on specific well-defined patient segments in lung cancer. ALK-positive and ROS1 positive mutations affect about 3% to 5% of all non-small cell lung cancer patients with about 80% of patients already diagnosed at Stage II, so metastatic stage. These patients are often younger in their 40s and 50s, unlike typical lung cancer patients who are normally diagnosed in their 70s. These patients are more frequently women and in general, have high rates of central nervous system involvement, so brain metastasis. What's very characteristic about this specific defined patient population is that they are one of the most engaged lung cancer communities, and they are actively shaping their treatment. Established treatments have been transformative for these patients, and they frequently remain on ALK and ROS1 tyrosine kinase inhibitor therapy for years. As an example, treatment duration in first-line non-small cell lung cancer patients treated with PD-1s, majority of cancers of patients in non-small cell lung cancer patients, they are treated for approximately 10 months. For EGFR mutated cancers, patients are treated for up to 2 years. And for ALK-positive first-line patients, current median PFS is longer than 7 years and hopefully further increasing. Next slide, please. Acquisition of Nuvalent is going to strengthen our overall oncology ambition and specifically, it will accelerate our efforts in lung cancer, supporting both development and commercial efforts for our ongoing BridgeRES program in lung cancer. After this acquisition, we are looking at a number of milestones from 2 approvals this year, potential approval for ROS1 naive patients next year, ongoing trial in first-line ALK mutated segment and the ongoing trials for the early-stage HER2 assets. On the other hand, GSK's global footprint will maximize the opportunity for Nuvalent portfolio of precision medicine assets. Next slide, please. We are moving to the ALK-specific segment. Over the past decade, we have seen an improvement in the outcome for ALK-positive non-small cell lung cancer patients as innovation focus on solving unmet needs such as blood-brain barrier penetration to address CNS metastases, which I have mentioned before are very frequent in this patient population and then broader coverage of resistance mutations. The current third-generation ALK inhibitor, lorlatinib, recently had 7-year update of its Crown study, where the median PFS still has not been reached. Despite strong efficacy, unmet need remains as tolerability issues significantly limit potential quality of life and therefore, adoption in first line. These limitations stem from the lack of selectivity for ALK and involvement of [ TRK ] kinase, which Tony will describe in details in a moment. Neladalkib is a next-generation ALK tyrosine kinase inhibitor designed to improve on [ lorlatinib ] profile, sparing Tet engagement with first-line pivotal study ongoing. Next slide, please. So how did this development impact the evolution of treatment practice in ALK-positive non-small cell lung cancer space? Most commonly used medicines to treat ALK-positive lung cancer are second- and third-generation ALK inhibitors. Alectinib continues to have a significant use in first line driven by its acceptable tolerability profile. Despite best-in-disease efficacy, [ lorlatinib ] adoption is limited by its tolerability issues. central nervous system events impacting cognition, speech and mood and then metabolic issues like hypercholesterolemia, hypertriglyceridemia requiring comedications and then significantly weight gain. These issues are attributable to lorlatinib's off-target activity. And I'll now hand over to Tony to talk more about [ neeladelpin's ] differentiation and supporting clinical data.

Thank you, Nina. On this slide, you can see the evolution of the -- of the ALK family inhibitors. [ neladalkib ] is the fourth-generation inhibitor is designed to pair third-generation ALK potency with strong CNS penetration while retaining activity against both single and compound ALK-resistant mutations. This, alongside its TRK-sparing profile underpins its potential for superior efficacy and tolerability compared to previous generations of inhibitors. Each of these characteristics is critical to maximizing the amount of time a patient can potentially benefit from therapy. Next slide, please. Okay. Taking a more detailed look, we believe that NAA has the potential to be a best-in-class ALK inhibitor as it delivers strong performance against the 3 basic pillars of precision medicine. Firstly, and as you can see on the left-hand side of the slide, it's important to hit the driver of disease hard to control the original tumor. Secondly, and also on the left-hand side of the slide, it's important to address or prevent the emergence of key drivers of disease progression to extend the durability of response. This specifically includes designing molecules, which maintain inhibition and the presence of on-target resistance mutations as well as optimizing for brain penetration to address metastatic disease. This is illustrated on the right hand of the slide and is critical because of the high incidence of brain metastases at diagnosis for all positive non-small cell lung cancer patients. And thirdly, it's important to address selectivity to avoid off-target adverse events, which can be treatment limiting. You can see this in the selectivity index calculated on the left-hand side of the slide. This is critical to minimize discontinuations and dose reductions and to maximize potential therapeutic benefit while removing barriers to adoption. This differentiation is important as it allows NELA to improve on the significant cognitive, psychiatric and metabolic adverse events associated with lorlatinib, which severely impact patient quality of life. Next slide, please. With these design principles in mind, data from the ALCOV-1 study support a best-in-class profile for [ nilazalib ] based on cross-trial comparisons and TKI pretreated patients, NELA produces encouraging response rates, including for intracranial disease when compared to [ lorlatinib ]. As you can see on the left-hand side of the slide, in the second-line plus setting, a median duration of response is not yet reached. And encouragingly, 60% of patients achieved a duration of response greater than 18 months, which is a meaningful improvement compared to 9.6 months for lorlatinib. Based on these data, the FDA has granted NLA priority review with a target PDUFA date of November 27, 2026. In TKI-naive patients, [ NELA ] produces promising response rates compared to lorlatinib and importantly shows a 12-month duration of response for more than 90% of patients, exceeding the historical benchmark from lorlatinib in the Phase III CRM study. A Phase III registrational study for NLA in the first-line setting, [ ALPazR ], is ongoing and recruiting well. Next slide, please. ALK-positive non-small cell lung cancer patients are typically younger non-smokers who are otherwise fit and healthy. This is a highly motivated and informed patient population and as such, a manageable safety and tolerability profile is a key consideration. NLA spares [indiscernible] finding, and this is reflected in its adverse events profile, which shows avoidance of the long-term metabolic and neurological adverse events seen with lorlatinib. Importantly, the liver enzyme elevations seen with NL are typically asymptomatic, transient and manageable through monitoring and dose modifications as is consistent with routine practice for TKIs. Overall, they do not result in increased drug discontinuation rate and compared to other ALK inhibitors. This is also supported by physician feedback, which highlights a more manageable safety profile for compared to the intensive monitoring and caregiver burden associated with lorlaatenib's cognitive impairment, mood disorders and metabolic side effects. Next slide, please. On this slide, you can see the ongoing clinical studies of LLA. ALCOV-1 is a Phase II study in ALK-positive solid tumors with several cohorts. It was designed to have registrational intent of TKI pretreated or positive non-small cell lung cancer patients. In 253 TKI pretreated patients, NLA delivered a 31% overall response rate and a durable response of 64% and 53% of responders estimated to remain in response at 12 and 18 months, respectively. Notably, 78% of patients had received 2 or more prior ALK TKIs, of which 91% had received prior lorlatinib. So a heavily pretreated population for which no approved therapies have demonstrated meaningful activity. This pivotal data was presented last week in Chicago at ASCO. As mentioned, based on new data, NE has been granted breakthrough designation with a planned PDUFA date in November of this year. [ ALKASAR ] an ongoing Phase III registrational trial in the first-line setting using the current standard of care [ alectinib ] as a control arm. This study is enrolling well. Now let me hand over to Nima.

Thank you, Tony. We are moving now to the ROS1 space. Consistent with the evolution of ALK TKIs, zidesamtinib is strongly differentiated through its significant improvement on safety profile and strong efficacy compared to previous generations of ROS1 TKIs. [ Repotrectinib ] carries notable TRK-driven neurological and metabolic side effects that drive dose discontinuations. [ Pelotrectinib ], a third-generation ROS1 TKI improves on tolerability, but still brings meaningful dizziness, GI tox and side effects like diarrhea, nausea and vomiting. Based on the large clinical data sets generated across over 900 patients, zidesamtinib shows a class-leading safety profile. Its ROS1 selective and Tx-paring design allows for a differentiated safety profile avoiding the CNS and GI side effects historically associated with the class. Ability to stay on treatment for longer has potential to translate into improved median PFS in first line compared to current treatment options. Back to Tony again.

Thanks, Nina. Next slide, please. As with NEL, zidesamtinib has a compelling profile. It's a highly potent Tux-sparing ROS1 inhibitor with broad mutational coverage and improved selectivity profile. Importantly, the [ G2032R ] resistant mutation coverage supports durable activity in heavily pretreated patients and improved outcomes in earlier lines of therapy, including in the TKI-naive population. G2032R is a common resistance mutation arising following [ crizotinib ] and next-generation TKI treatment. Once more, high TRK selectivity allows for CNS activity, which is demonstrated in patients who received more than 1 prior brain penetrant TKI as well as in the naive population. Overall, this profile translates to a low rate of serious or severe CNS events with significant improvements in problematic adverse events associated with earlier generation TKIs, including dizziness and GI toxicity. Next slide, please. Moving on to the clinical setting and looking across-trial comparison to [indiscernible], you can see a substantially improved duration of response in both heavily pretreated and TKI-naive patients as evaluated in the ARS-1 study. Based on these data, the FDA has granted [ ZEDO ] breakthrough designation with a target PDUFA date of September 18, 2026 in TKI pretreated patients. These data support best-in-class efficacy with comparable response rates and response durations exceeding that of palotrectinib in both the TKI pretreated and naive settings. Next slide, please. Finally, this slide shows a summary of the ongoing studies. The ARS-1 Phase II study recruited patients with ROS1 positive solid tumors with a specific cohort focused on ROS1-positive non-small cell lung cancer, including those previously treated with TKI inhibitors as well as TKI-naive patients. As mentioned, clinical efficacy supports a best-in-class profile. And importantly, this extends to safety where demonstrates considerable improvements in rates of dizziness, GI and liver toxicity that is seen with previous generation TKIs. The pretreated indication carries an orphan drug designation with an up-and-coming PDUFA date in September, while filing of a supplementary NDA for the TKI-naive setting is planned in the second half of 2026. I'll now hand over to Julie.

Thank you, Tony. So to summarize, the acquisition of Nuvalent includes 2 launch-ready best-in-class -- potential best-in-class assets with blockbuster potential with PDUFA date later this year and will further strengthen our oncology portfolio, building on our exciting development pipeline, which includes [ Red, Red] , our ADC for lung cancer. I will now cover 3 main areas, and all my commentary will refer to core results at constant rates. So first, the transaction details. We have agreed a purchase price of $124 per share, a 40% premium to yesterday's closing price and a 26% premium to the 30-day VWAP, equating to an aggregate consideration of $10.6 billion or net of cash, GBP 9.4 billion, which is GBP 7.1 billion. We will commence a tender offer for the shares within the next 10 business days and subject to regulatory approval, we anticipate closing the deal in Q3. Second, the financial impact. We expect the acquisition to support GSK's revenue growth from 2027 onwards and be incremental to our existing ambition for sales of more than $40 billion by 2031 and to strengthen sales, operating profit and margin through the [indiscernible] loss of exclusivity period. We expect the deal to be accretive to operating profit in 2027 and earnings per share in 2029 onwards, inclusive of synergies and reprioritization. Assuming completion in Q3, we expect a low single-digit dilution to EPS in full year '26 to 2028, but remain confident in our current 2026 guidance range for earnings per share growth of 7% to 9%. Third, capital allocation and the balance sheet. This transaction is aligned with our investment growth priorities and BD strategy. Our capital allocation priorities remain unchanged. Post the transaction, we will retain our strong investment-grade balance sheet with no impact expected on our credit rating. We remain committed to our dividend of 70p this year and our progressive dividend policy. And lastly, the transaction will be funded from existing and new debt facilities and existing cash resources. And with that, I'll hand back to Luke.

Thanks, Julie. So to summarize, we think this deal is a strong strategic fit for GSK, and it represents a continuation and acceleration of our strategy. We get 2 potential best-in-class products that can launch this year, driving growth and operating profit in 2027. And as you've just heard from Julie, we have applied financial discipline to ensure that we manage the cost of the deal and then we recommitted to the dividend and the progressive policy. So with that, let's turn it over to Q&A.

Thank you. With this, we're ready for Q&A. [Operator Instructions] The first question comes from Matthew Weston.

It's about the split of the value or the potential, I should say, by asset. So I think consensus has $2 billion peak approximately for Nuvalent, but that's heavily skewed to ALK over ROS1. I'd be very interested in your view on the relative contribution of [ Zeta ] versus [ NELLA ]. And then also about the speed of launch. I think [ Zeta ] has an expanded access program with over 500 patients on it. Does that mean that we should expect a rapid launch on approval as we convert those patients to commercial drug?

Thanks, Matthew. So I'll answer the second one and then maybe, [indiscernible], if you want to characterize that without too much color because we'd like you to do your work on your own modeling, Matthew. I appreciate the question. I mean the speed of launch, absolutely, and that's -- if you look at our heritage working with [ Sierra ] and other biotechs that we've acquired, this is a very important component. We're already planning to pivot and execute that launch, and we're in a good place. The access program, the recruitment for the programs, the first-line ones are very strong. So again, it's a targeted population we can wrap it up. And it was something that we're going to do for B7-H3 anyway. So thanks for that question. And Nina, the split of the asset and $2 billion, Matthew, I assume, again, we won't comment on that, but it's obviously not the time frame that we're looking at over the life of the deal.

Yes, I was just going to mention that EUR 2 billion. I'm not sure that's the peak reference to peak sales, which are clearly higher than that. The -- in terms of value split on sales, I would say gross space is probably between 1/4 to 1/3 of the total value.

Next question comes from Sarita Kapila.

Just on [ NELA ] and in the TK naive cohort that you have in your original data, are you actively discussing with the FDA whether a frontline label expansion is possible? Or will we have to wait for the [ ALKAzAR ] Phase III PFS readout for approval in the earlier line setting? And then if I could just squeeze in another one, please. Are you sure that doing a head-to-head in the frontline setting versus [ Alecensa ] is the right comp, particularly when we've seen 7-year PFS data for [ LIBRENA ] from the CAM study?

Thanks, Narita. We certainly have looked at those questions in depth. Tony, did you want to cover the first 2? And then maybe, Mondher, if you can comment on the standard of care. On the second question, Nina, feel free to jump in.

Yes. And look, I'm obviously not going to get into the details of regulatory interactions, and it's a little early in terms of the Phase III study. The points you raised are all ones that we have taken into consideration, and we very much see this molecule as being one which has the credentials that will support a first-line indication.

Just to add actually, today, the standard of care in terms of market penetration is still with almost 45% market share and tells you about how important it is for us to pick the right control arm. Just as a reminder, [ lorlatinib ] Phase III trial, the current data that we presented at ASCO [indiscernible] as a control arm. So very low bar in the third-generation TKI. So the fact that we are choosing alectinib, I think it's the right thing to do. Having said that, I believe the community will certainly look at other setting of the disease and also other options to have head-to-head data with lorlatinib to try to figure out actually the best benefit risk ratio.

Thanks, Nina, do you want to add?

Yes, I just realized that we didn't answer Matthew's question on early access program and the number of patients. Yes, actually, Nuvalent did a great job with early access program for both assets, and there are hundreds of patients in early access program, which gave us confidence for multiple reasons because feedback from the physicians who have hands-on experience with actually all TKIs in this space, and we're able to compare the experience of patients and the physicians was really tremendous benefit. When we launch the drug, the assets, we do expect [indiscernible] access patients to be transitioned to commercial source.

Actually, that's an important point, Sarita. Because there's such broad experience and empirical experience with both of these products, we were able to do a massive amount of due diligence, frankly. So we had over 300 interactions, quantitative, qualitative, in-person interviews, online interviews with physicians, a large proportion of whom have direct experience utilizing these drugs in these subpopulations. So again, that's what's behind the confidence and why we think there's a pathway there. We've also spent a lot of time characterizing the profile of [indiscernible] in particular, and have some insights there that we will disclose at a future date.

Next question comes from Kerry Holford.

One for Julie, please, for the financials in the context of the accretion that you cite from next year for revenues and profits and EPS in '29. You referred to synergies and reprioritization. So just intrigued to hear what that involves both in terms of the deal, but perhaps also internally, anything additional to add there?

Thank you very much, Kerry. So as we mentioned, obviously, we wanted to be very disciplined about the way we approach the deal generally, appreciating that this was going to be on the face of it dilutive to shareholders. So taking each of those in turn, and I've worked very, very closely with Tony and his team on this -- the first one is relating to synergies on the integration. We expect some synergies through SG&A, and then we also expect synergies in R&D, including through the discovery portfolio and integration and also contracting synergies for activities such as CMC and clinical study design. This is exactly what we found with other deals that we've embarked on recently. And then the second thing relates to the just prioritization in the business generally across all areas together within R&D. And as far as we're concerned in R&D, approximately twice a year, we do a very thorough review of the portfolio. We are constantly assessing the internal portfolio and reprioritizing assets to really optimize the return on investment and the probability of technical and regulatory success. So net-net, we believe we've been very responsible in terms of exercising discipline around the dilutive impact of the deal.

Next question comes from Emmanuel Papadakis.

It was just a follow-up on [ Acazar ]. I'm not sure if I -- apologies if I missed it. Is the frontline approval predicated on a PFS readout? Or is there a scenario in which you could get that approval sooner? And maybe a quick question on ROS1 back in the day, the original commercial optimism around that landscape is predicated on better diagnosis, unlocking clinical and commercial opportunity that hasn't really happened. Do you have any greater hopes for a diagnostic transformation, unlocking the patient opportunity there?

Sure. Thanks, Emmanuel. Nina, do you want to cover the second question and then Tony, a quick one ...

The frontline study is based on PFS.

Hear that, Emmanuel? PFS. And Nina, on the ROS1 testing and opportunity.

Again, both ROS and ALK are now standard part of testing is part of the panels that are done on lung cancer tissue samples when -- at the diagnosis. That's probably less of a barrier. But as we have seen, the product profile so far actually has significant limitations.

And Nina, do you mind expanding on this? Because I think the more you look at these products and particularly you look at these subpopulations with these particular mutations, what we found is a very interesting relationship in terms of innovation price, small molecule, you've got a defined population, but you've got these very, very long tails, which the better tolerated profile that you can get, you're more able to sustain these. So the mathematics start to look very, very interesting. But you really have to stop and look at it and just work your way through it, which is what we've spent a long time doing. So, you might expand on that?

So again, to repeat what Luke said, there are a small number of patients, that's definitely true. But if you look at the totality of non-small cell lung cancer in terms of value, it's disproportionately higher because for each patient that you start on ALK or very likely in the future on ROS, you have a multiple duration of therapy for these patients. So CRM data update 7 years PFS rate at 7 years is still 55%. So median has not still been reached. It's very likely that, that tail will continue at 50% or above 50%. And these are metastatic patients. They are not stopping their treatment. It's very difficult for mutation-driven cancer to stop treatment because what we have experienced from all these TKI areas, so EGFR, alcisol, as soon as patient stops therapy, the cancer usually comes back. So in terms of value, they are proportionately significantly bigger than what the number of patients would indicate. I would just add, as Luke mentioned, we spoke to probably about 300 or had interactions with about 300 physicians. The positive thing is that they had handsome experience and they could compare the effects of all these TKIs. I will share some of the comments that were made. One is lorlatinib is a drug that everybody loves to hate. But because of the efficacy, it's frequently the first drug that they will offer to their patients. What comes with the use of lorlatinib is CNS-related side effects. So serious conversations with the patient, with patient cares because they have to be aware of these side effects that go into psychiatric area. So one of the physicians said using lorlatinib, I became both psychiatrists and metabolic experts. One of them mentioned that he became a target for sales reps that sell GLP-1s because the patients gain weight, and these are younger patients, it's very relevant for them. One of them described a patient who was a younger woman. Her income was coming from being an influencer on social media. And after gaining weight, she actually stopped therapy and progressed. So these are all issues that you need to address over the years while you're treating patients. And therefore, it was a very, very clear feedback that tolerability is the main barrier. While we were doing due diligence, just a comment between ROS1 and ALK, while we were doing due diligence, I have to say ROS1 space was probably what physicians refer to as slam dung. There is no way that they would not use this drug over other inhibitors, ROS1 inhibitors. So there, our confidence is actually very high that this is going to be the best in disease, best-in-class drug.

Thanks. And we can expand on that. Again, we've been spending a lot of time on this before breaking cover. And really been very thoughtful and rigorous in profiling these assets and the team that's developed them.

Next question comes from Sachin.

I'm on dialing, apologies. So 2 quick questions, please. One, just a big picture for you, Luke. This is one of the largest deals [ Glaxo ] has done in a while and is focused on launch assets versus mid-stage, early-stage pipeline for 2930. So just what's driven that in a strategic shift is consistent? And then second one for Nina is a follow-on to prior. So thank you for giving the split on ROS versus ala. I wonder if you could just give us a bit of color on second line versus frontline across both, acknowledging the frontline is bigger. So 2-part question. One, how comfortable are you with consensus on the launch in late line before frontline comes through? Then how deal valuations frontline out whether some perceived risk around that data delivering differentiation?

Thanks, Sachin. So I'll answer the first one, and then we'll go to Nina. Look, I think -- and that's what I tried to outline with that slide in my introduction, we have a methodology to our BD. We like validated targets because essentially the clinical scientific components are heavily derisked. And then we're really relying on our regulatory clinical execution and commercial launch capability, which, again, we have to validate every quarter, but I think it would be fair to say there's a high degree of a strong track record there. What we had with Nuvalent was unusual. Basically, it was a little bit like London B's session, right? Suddenly, 2 or 3 come along at once. And each one of these products, we would have acquired as a separate deal. We just had the attractive element of having them all embedded in a single company. And importantly, this was very much a bottom-up driven deal. Some deals have driven top down. I might come to the office and say, I really like this company, let's go for it. This was very much a deal that was driven by [ Chris Sheldon ], who you all know from his previous time at AstraZeneca. We had the individual involved in designing [ Tagrisso ] on the team. There was a number of other AstraZeneca and Roche EPs. I'm sitting next to Tony, who was deeply involved in [indiscernible] design. And so this built momentum inside the organization over quite a long time. And in the end, we reached the conclusion that this was a really smart deal to do and it was a good use of our shareholders' capital. And again, if you take the components apart, it's more consistent with what we've done before. But also there's a traction in critical mass because essentially, we're pulling forward our entry into lung, but in discrete populations to build our credibility there, which positions us very well for B7-H3. So there's a method to this, but also critically, as we've said, there's financial discipline here to maintain capacity for future accretive deals. But this was just too attractive to let pass by. Nina, over to you.

Yes. And Sachin, I'm not sure I fully understood your question. So I will try and maybe either you need to help me or my colleagues. So first and second line, ROS1, as I think we communicated, second line has been filed. PDUFA date is in September this year. In ROS space, all -- this is really ultra-orphan indication. So a number of patients is very small. All the assets are approved on single-arm studies. Second line -- so the first line is -- filing is expected in the second half of this year, so approval next year. Again, this is extension or additional -- based on the additional data from TKI-naive patients from the ongoing study. As I mentioned, there is a very, very clear feedback that this is going to be easily standard of care in the ROS space. I'm not sure I fully -- was there a question on ALK as well for --

Maybe I'll reframe the question. So 2 parts. One, are you comfortable with consensus on launch in the later lines across both ROS1 and ALK in the next couple of years versus yourself framing upside to peak midterm? And then the second was how much the deal valuation is on frontline A about data [ 2829 ] how do you assess the risk around differentiation delivery?

Yes. So Sachin, pretty confident, again, to a previous question about number of patients on early access program and the extent of experience that physicians community had so far. So I would say, again, not going to guide on the year annual sales, but we are quite confident that consensus is not too optimistic or unrealistic. And then the value of first-line ALK, yes, clearly, first-line ALK carries majority of the value of the deal, no doubt. You are looking at the highest number of patients with the highest duration of therapy. Based on the data that we have seen and the profile of the drug, we actually have very little doubt that this would not be successful. Question is time, which also we actually have a pretty good idea when it's happening because the recruitment in the study is actually going exceptionally well.

Sachin, I would say, look, 3 things to consider in the context of confidence in the duration of response data in the first-line setting. And this is not only true for NLA, but also in the ROS1 setting. If you look at the improved resistance profile, that is important in maintaining response because it prevents the emergence of compound resistant mutations as well as taking care of already existing cases. Obviously, tolerability plays into that as well and in particular, in the context of dose reductions or dose holds. NLA has a very favorable profile in that context and also in the context of when reductions are necessary, what we see is mostly only single stepdown reductions in the case of Nella and its intrinsic properties its PK favor a more reliable outcome in that context. And in case anyone is going to ask me the question, we've looked into the impact of dose reductions and holds on the efficacy, and we can see very little to suggest that there's any significant impact there. So it's a combination of features underpinned by the profile of the molecule, which gives us confidence in duration of response.

I can jump in. And Sachin, you may remember this. 10 years ago, I think a few people believe that EGFR market will become what it is today. And actually, second and third generation of drugs have turned this disease into a chronic one. I think we have with ROS1 and the opportunity to turn those metastatic patients to really a patient who can live with their cancer for many, many, many years. And the key word here is really tolerability. I think we know that second, of course, third generation and these assets hit the target hard, have a broad coverage of the different driver mutation, penetrate the CNS, has definitely a very good efficacy in terms of tumor shrinkage, but the key word is tolerability and durability. And I think what we have seen so far in the Phase I/II and the momentum that Nina described around clinical trial and expanded access program from the community of oncologists but also from patients is telling us that this is the right drug for them to stay longer on treatment, even in second line, but definitely more in first line in order to live with this disease more than 10 years. Of course, everyone is excited about the crown. But remember, I mean, half of the patients stop their treatment already. And we know what will happen for this patient when they stop treatment.

Next question comes from [ Victor ].

And so maybe starting with the comment you've made on [ ALCASR ]. So you've mentioned that recruitment was progressing well and that you have good visibility on completion. So can we ask whether there is an interim analysis planned on [ ALCASR ]? And what would be the trigger for that potential interim analysis? And then on [indiscernible] signal, can you walk us through how you think about managing this signal in clinical practice and why you believe it won't ultimately undermine its tolerability advantage? And maybe last one for Julie. You've indicated low single-digit EPS dilution for 2026, but guidance for 7% to 9% remains unchanged. So can you just help us reconcile this?

Great. Thanks, Victor. So Tony, do you want to cover interim? Mondher, you cover clinical consequences and management of [ NELA ], which we've learned a lot about and Julie last on that question.

Yes. And obviously, I'm not going to get into the details of the ALKA study, but yes, an interim is planned and you can get a clue from that if you look at the landmarks from the Phase II Crown study.

Right. And Mondher, we've seen the full data set.

Yes. So maybe a quick reminder that one of the challenges that we have seen with third-generation ALK inhibitors is essentially the off-target side effects and the fact that inhibiting the [ tobomyin-related ] primates, there are 3 of them. And actually, the A and C is the one that basically is driving the whole CNS side effects. And actually, it's really disturbing because not only we have somnolence, we have cognitive effect, we have really even mood confirmation that basically prompt patients and physicians to stop treatment. This is the most critical actually side effects. Of course, there are other side effects, in particular neuropathy that occur in more than 34% of the patients that sometimes actually are grade 3 and 4. Those type of side effects are leading to the discontinuation. What we have seen in the 1,500 patients treated so far, both in clinical trial and expanded access program with [ Nera ] is basically no side effects of this type. The only thing that was noticed is an elevation of transaminase that is really asymptomatic. There was no hepatic failure. There is no fatal events. It's only biochemical asymptomatic elevation of tram that can be monitored very easily. Physicians are oncology because it's not the first time that you have this in tyrosine kinase inhibitors. And the approach is to stop treatment for a while, test until it goes back and then rechallenge with a slight dose reduction. And we have seen that patient can completely be rechallenged without any compromise with the efficacy.

And I would just add that other ALK inhibitors have the same monitoring for liver enzyme elevation already as part of their management and clinical practice and the label. So we don't expect that this is going to come as a surprise to the physicians because that's exactly what they are doing with other ALK inhibitors already.

Great, and Julie.

Yes. Thank you very much for the question. So obviously, the EPS is affected mostly by the interest that comes through together with the [ Avance ] profile in its own right. And we are assuming for '26 that we've got a Q3 completion of the deal. And clearly, when we've guided, we've guided an EPS range of 7% to 9%. So it doesn't mean that we move within that range to manage the dilution, but we stay within the range on a net basis.

Next question comes from [ David Evans ].

Hope you can hear me. So just a question, if you could just clarify or give some more information on the level of discontinuations that you see in the ALK space. I thought your slide said that on lorlatinib, well, only 7% discontinuations versus 5% on NLA. Is that a like-for-like comparison? It doesn't quite seem to stack up with your sort of commentary about wildly different tolerability profile. So just anything in the lorlatinib proposition, if you just flesh out, much less well tolerated in theory than those figures would indicate.

Yes, sure, David. Thanks for your question. I mean these are distinctly different profiles. Tony, do you want to cover that and Nina add any market commentary, physician commentary?

Yes. And the data you've picked are indeed relating to discontinuations associated with tolerance. What's important, though, is to look at the complete picture for treatment discontinuation because as Nina mentioned earlier, with discontinuation and dose reduction, you do see disease return and the data for 2 years, I believe, is a 38% progression rate based on the combination of those things.

Nina, you want to add anything?

Yes. Again, David, so I think if you look at the discussion that happened at this year's ASCO post Crown 7 years presentation, I believe there was a slide from the discussion, which something along is the Crown is the price. or something along that where the focus of the whole discussion was exactly the tolerability profile and is the benefit that comes with such a long PFS actually worth the adverse events that patients are going to. I would probably say if there is an interest to -- for you to talk to some of the physicians who have used the drugs and hear the experience because that probably tells you more than the numbers in the tables.

Thanks, David. And we can certainly do that for anyone who's interested.

Next question comes from [ Naresh ].

Sorry, can you hear me now? Just interested in the ex U.S. development plan, please. 2/3 of Alecensa sales come from outside the U.S. and China is a big part of that looks like about 1/3 of the sales. There seems to be no Chinese sites in any of the Nuvalent development program as far as I can see. Can you just help us understand the ex-U.S. development plans, please.

There is a plan --

Basically. It is a global plan that just opening Chinese sites.

Next one, please.

Next question comes from -- is a follow-up from Matthew Weston.

Amazing. It's a quick follow-up for Julie. Julie, in your opening comments, you said, I think, that the deal strengthens margins through dolutegravir's LOE. Now I think the previous comments were that margin was going to be flat through the dolutegravir LOE. So does that mean we all need to sharpen our pencils and start taking the margin up? Or this just gives you incremental confidence that you'll be able to keep those margins flat, particularly because it looks like there may be some incremental pressure in HIV from guideline changes and other things.

Okay. Thank you, Matthew. So just in terms of -- just to recap for everybody, in terms of prior margin guidance, we've said that in '26, we'll be above 31%. And then we've said we will be -- have a stable margin through DTG LOE, which is the '28 to '30 period. What we're referring to when we say the deal strengthens is it will improve. Obviously, it's very accretive to sales during this period as the team have talked about. It's high profitability assets. So it brings through benefits to the profit. And therefore, we would expect the margins during that period to be higher than before the deal. As we know, we've managed the dilutive period to be a very short period with accretion quite quickly. So therefore, it does give us a benefit to profit and margin through the DTG years.

Appreciate the other question. We've got time for a couple more if people have them.

Next question comes from Zain.

Zain Ebrahim, JPMorgan. It's another follow-up for Julie, just on the dilution over the next couple of years of low single digit. I think the color you provided earlier was helpful in terms of the expectation for synergies, and it sounds like there's an internal portfolio rationalization goals for CMC costs within Nuvalent that you can potentially unlock to manage at a low single-digit dilution. But just stepping back, I'm still struggling a little bit to get fully get there because I think Nuvalent consensus is about $100 million of revenues for next year. OpEx, they've got about $400 million in terms of spend. So what's the key disconnect that you see? Do you think that maybe consensus top line is a little bit on the conservative side given your bullishness around and what the uptake could be there? Or do you think that maybe there's a portion of your R&D effort rationalization that helps you get there that we might not be appreciating?

Thanks. I mean I think the short answer is it's a combination of those things, but I'm not sure we want to give too much color at this point beyond the commitment. But Julie, feel free to expand if you want.

Yes. I think we've covered it. I mean we wouldn't go into specific projects in the R&D portfolio. And I would emphasize that we're looking at prioritization across the company, not just within R&D, but across the company and driving that hard to be very disciplined, as Luke mentioned. I mean we wouldn't normally also comment on consensus models. But in our view, the R&D cost in consensus models for the target are reasonably high. So I'll just leave it at that.

There's time for one more question, probably also a follow-up and then [indiscernible].

Just a quick one. Anything you could say on the IP for the 2 late-stage assets that you're acquiring here? And just to confirm, you expect both to be exempt from IRA?

Yes. So yes, on IRA because they're orphan and [ 2040s ] plus.

Yes, [indiscernible] 40s.

Thanks, [ Kerri ]. Very grateful for everyone joining at short notice. I'll conclude by saying, again, to reinforce, we think this deal is very much the type of transaction that we do. It's a great fit strategically for us. It's consistent with our strategy in lung cancer and actually forms an accelerant for that. And we're getting a number of very, very attractive assets, which, as you've hopefully heard today, have a deep exposure amongst the clinical community and patient exposure. And I think there's a clear pathway for both of them, the late-stage ones to best-in-class as well as some earlier-stage assets like the HER2 and a stable of preclinical medicinal assets, which are also very creative. And finally, I hope you've seen that we have applied financial discipline here in terms of selecting this transaction, pricing and then integration of it. So thanks again for your very thoughtful questions at short notice and look forward to following up with you.