Welcome, ladies and gentlemen, to AstraZeneca's Investor Conference Call, ESMO 2024. Before I hand over to AstraZeneca's management team, I'd like to read the safe harbor statement. The company intends to utilize the safe harbor provisions of the United States Private Securities Litigation Reform Act of 1995. Participants on this call may make forward-looking statements with respect to the operation and financial performance of AstraZeneca. Although we believe our expectations are based on reasonable assumptions, by their very nature, forward-looking statements involve risks and uncertainties and may be influenced by factors that could cause actual results to differ materially from those expressed or implied by these forward-looking statements. Any forward-looking statements made on this call reflect the knowledge and information available at the time of this call. The company undertakes no obligation to update forward-looking statements. Please also carefully review the forward-looking statements disclaimer in the slide deck that accompanies this presentation and webinar. There will be an opportunity to ask questions after today's presentations. Please use the Raise a Hand feature to indicate you wish to ask a question, and please remember to unmute your line when invited to speak. And with that, I will now hand you over to AstraZeneca. Good evening, everybody, and, thank you so much for joining us tonight. It's, really a pleasure to have you here all. It's been another, great ESMO. Fantastic to come here and to see the, ESMO grow year on year and, become a true competitor to the ASCO. Of course, we love the ASCO, but it's also nice to see the ESMO, develop the way it has developed in the last few years and show so much good data. This is our forward-looking statement. This is the, agenda. I'll say a few words in introduction, and as you can see here, we're gonna cover quite a number of, topics here. Of course, Susan is gonna give us, an overview of, the World Lung and, this ESMO. We'll talk about Dato-DXd, in particular, the QCS test. We're lucky enough to have tonight with us, Professor Powles, who is the PI for the bladder cancer study we presented, or he presented actually at this ESMO, and Sunil will join him for this session. We have a Q&A that Dave will moderate, and then we'll hear from Matt, who is the VP in charge of early oncology development. We'll hear about ADC, and then Christian will take you through our next generation IO agents. As I said, we are very lucky to have Professor Powles with us tonight, and this is the leadership who is gonna present tonight. Let me start with this slide that you've seen before, at least for those of you who were with us back in May at our Investor Day. It really articulate how we plan to grow to achieve to grow to this $80 billion ambition we have for 2030. But also importantly, what are the technologies that we are working on to grow past 2030? We have a very ambitious goal for 2030, $80 billion, as you know. But it's also important that to remember that we are actually intending to be a growth company for a long time to come, even beyond 2030 and the patent expiries that we will experience. It's an important point to keep in mind because a number of the technologies we work on today, they cost, they impact our R&D budget and our P&L, and for most of them, we get zero value in our share price today. We believe it's important to do this because if we don't, in three or four years, some of you will ask us where our growth rate is post 2030. Hopefully, over time, you will see the value of these investments in ADCs, in radioconjugates, cell therapy, and many others. You'll see the value of these investments in our growth rate post 2030. Between now and 2030, we have minimal impact from patent expiries. Actually, I can't say nil, of course, because those are... Some of them are important products like Lynparza and Farxiga, but overall manageable patent expiries and also some of the IRA impact, and we have growth potential coming from our existing portfolio. There are lots of reasons to believe that products like Tagrisso, Calquence, and HER2, and the others will continue to grow substantially till 2030, and tonight you'll hear one of the reasons we believe Imfinzi can grow substantially, and certainly bladder cancer is one of those vectors of growth. We're also gonna launch new molecular entities, both in oncology and beyond, as you can see on this graph. As I said, we've been investing in quite a number of technologies that we believe will shape the future of medicine and, in particular, the future of cancer. And tonight you'll hear about antibody-drug conjugates and IO bispecifics, but we have many more ADCs, and also we have radioconjugates, we have cell therapy projects, we have T-cell engagers in development, and all of those are making very good progress. And hopefully, over the next couple of years, you'll see more and more data coming out of those programs, and you hopefully, at some point, will give us some value for this program, which some of them are quite exciting actually. In oncology, our ambition is really to cure cancer, and the best way to do this, we believe, is to attack it from different angles and combine products. Combine products, and in fact, potentially combine them together, but combine them over time through a sequence of treatments. For instance, starting treating metastatic breast cancer with a combination of ADCs and IO, and following with a T-cell engager or cell therapy to cure patients. So that's why we are working on those, those all those different targets and mechanisms and technologies. We also want to look at treating patients earlier and in a smarter way. Earlier means diagnosis, and you've heard Dave in the past talk about the work we're doing globally to drive better diagnosis, better screening around the world. So patients are treating diagnosed early and treating early, so we increase the chance of actually curing them. And finally, we are looking also at better targeted medicines, and certainly the work we're doing in biomarkers and testing is a great part of this. Tonight, you'll hear Susan talk about the QCS test, which is a very good example of the work we're doing. We have a team of people that is dedicated to working on some of these computational pathology tests, and QCS, of course, is the first example of, I believe, many more to come. We've had 17 positive readouts since the beginning of last year. 17. Now, we've had also a couple of setbacks because we can't hope to achieve a 100% success rate. We've had a success rate in phase III that has been higher than the industry average, and many of those programs, we believe, actually will continue to drive the growth of our core products. The thing is that there are many of them. Some are large, some are smaller, but collectively, they represents an enormous amount of potential growth to 2030, which we believe is sometimes underestimated, and certainly, NIAGARA and bladder cancer, we think, is one of those. So with this, I want to thank you again, and I'll hand over to, Susan. Thanks, Pascal. And it's very nice to see you all here. So we've had a really great presence at both the World Conference on Lung Cancer and at this ESMO meeting, and it's actually been very inspiring to meet with many of the investigators here and see all the data that we've been presenting. More than 130 abstracts. We've had five presidential plenary presentations. And I just want to say thank you to all of the colleagues across R&D for all the work put into that. So last week in the World Conference on Lung Cancer, we presented, as Pascal's already referred to, the QCS biomarker data, which I think is very exciting. I'll walk you through some more of those slides today. We also presented the NeoCOAST-2 data for Dato-DXd plus Imfinzi in the early stage of non-small cell lung cancer. Again, one of these themes of bringing the combinations together and showing a higher PathCR rate than we've seen historically, including with the AEGEAN dataset. We've also presented some early data in the monotherapy in lung cancer setting in San Diego for volrustomig, our PD-1 CTLA-4, and for rilvogostimig, our PD-1 TIGIT. And again, we'll hear more about that from Christian in a few moments. And then at ESMO, of course, we had the plenary presentation for NIAGARA, which Professor Powles will walk us through, as well as more data for rilvogostimig in combination with chemotherapy from our platform study, the GEMINI study, and also early data from two of our proprietary ADC portfolio, our B7-H4 and our folate receptor alpha. Of course, out of all of these data, there's still been a focus on the TROPION-Lung01 OS data, and I've said a few times that the filing for TROPION-Lung01 is still ongoing, still under review. I don't have a specific update at this time, but I think it's worth just recapping a few points. So of course, TROPION-Lung01 is a positive study for the dual primary endpoint of progression-free survival. And in the progression-free survival in the non-squamous group was a clinically meaningful benefit in that group, but not in the squamous group. So based on these data, by histology and based on other precedents, the FDA accepted a filing in the non-squamous group. Again, you know, that filing was based on non-squamous as a pre-specified stratification factor, but not as part of the pre-specified statistical analysis plan. So that does mean it's a more complex filing than the standard. So what we presented last week was OS data with a hazard ratio of 0.84 in the non-squamous population. And that is a clinically meaningful trend to overall survival improvement at 2.3 months against a tough comparator of docetaxel in this setting. But of course, that magnitude of benefit isn't as great as would make a filing more straightforward. So there are various scenarios from here. So as always, the FDA are gonna be interested in subgroups, interested in the non-squamous subgroups, and they'll be looking at the benefit risk in that group, but also in other subgroups and other populations with benefit. So I hope that summarizes where we are today. I do also want to make sure that we're focused on all of the other data that we've also got at this congress. And I think as Pascal said, that this then aligns with the themes that I presented at the Investor Day, and that we've got highlighted here. So the Niagara data that Professor Powles will talk about in a moment, again, emphasizes the added benefit that you get from IO when you take it into the early stage setting in particular, it seems into the neoadjuvant setting, where you've got the tumor in situ, potentially the neoantigens present. Again, the NeoCOAST-2 data, I think, really emphasize that presence as well. And then again, the ability to combine and attack cancer from multiple angles, I think we've got data now, and it's robust data for both volrustomig and rilvogostimig that are helping build the confidence in these assets as we take them into multiple phase III trials. And I'm very excited to see the proof of principle that we've presented for our in-house ADCs that Matt Hellmann will tell you more about. So now focus a little bit on the QCS biomarker technology, one of the examples of how we've invested in technology and helped to bring this forward. So again, apologies if this is a recap from some of you that listened to the World Conference on Lung Cancer, but I think it's really important to understand some of these elements. So first of all, why is Dato-DXd different from the other TROP2 ADCs? And it's core to this is understanding of good ADC design. It has a very plasma-stable linker, which is enzymatically cleavable in the tumor microenvironment, but stable in the peripheral circulation. The evidence for that is that you've got a five-day half-life with this drug, and that the proportion of the exposure to the drug that is free payload is a small proportion of the overall total ADC. And again, you'll see the same example when Matt Hellmann talks about our own in-house ADCs. That low free payload exposure is also reflected in the data with lower bone marrow toxicity, and it's just something that's useful for looking at across ADCs in general when we're looking at the profiles. That lower bone marrow toxicity, of course, means is they've got a very combinable drug, so it's combinable with IO. I've shown data for that. It's combinable with platinum chemotherapy, which will help to improve the response rates in the early line, and it's also combinable with Tagrisso. That plasma stable linker is also important when you think about the mechanism of action and therefore also the biomarker. So first thing is that Dato-DXd binds to the TROP2 that's expressed on the cell surface membrane. But not all of the TROP2 that's expressed on the cell surface is internalized. This is actually a protein that gets cleaved, so not all of the TROP2 that's on the surface is gonna go inside the cell. That's a fundamentally important piece. So Dato-DXd is internalized, and then the plasma stable linker is cleaved, and then you release the payload into the cytoplasm, and that's what causes the cytotoxicity. You can also again get a bystander effect because that payload is permeable across the cell membrane and can be delivered to surrounding cells that may not have the same level of TROP2 expression or TROP2 internalization. So those are fundamentally important points. What the biomarker solution does is distinguish between the high TROP2 expression, which is actually quite universally high. It's in the IHC 2+ and 3+ range, if you put it into the context of other IHCs, across both normal tissue and tumors, and regardless of whether they are responding or progressing. It has a variable pattern of expression in the cytoplasm as well, and as I've said already, and I'll say again, that not all of the TROP2 that's on the cell surface is the relevant part for what is then going to be internalized. So what this biomarker can do with computational pathology is quantify both the cytoplasmic and the membrane expression. It can distinguish normal tissue from the cancer cells on the slide because that's what the algorithm has been trained by pathologists to do. Therefore, we see that it's associated with internalization of the ADC. It's associated with cytotoxicity preclinically, in vitro, in vivo, and clinically. Its prevalence is different across different histological subtypes, and I'll show you the data for that again. It's more prevalent in the non-squamous than the squamous, and I think this helps to understand and explain the differences that we've seen in efficacy for Dato-DXd across those histologies. And therefore, it's associated also with clinical efficacy. What is the process and the workflow for QCS? The first thing, of course, is you've got to have an immunohistochemistry assay with a TROP2 antibody, and we've done and we developed this in collaboration and validated it in collaboration with Roche Tissue Diagnostics. The second piece is then you have whole slide imaging, so you get upload of the digitization of that slide. And then you have this automated image analysis. So the first thing that the algorithm, which has been fully supervised and trained by a pathologist, does is differentiate the cancer cells from the normal tissue, and the cancer cells in the bottom left are shown in the green. You can then get pixel-by-pixel measure of the optical density, known as OD, in the membrane, and in the cytoplasm. And then you can calculate the normalized membrane ratio, NMR, for every tumor cell, which is just the ratio of that optical density in the membrane divided by the optical density in the membrane plus the cytoplasm. What we've then done is robustly look at, with statistical analysis, bootstrapping of that and Monte Carlo simulation to choose the cut point based on the TROPION-Lung01 dataset. More than 75% of cells that have got a ratio that's less than 0.56. We started off based on the TROPION-PanTumor01 dataset, and this is the basis of a pattern that we had filed, and you can see you get separation of the Kaplan-Meier curves in the biomarker positive QCS, NMR positive versus negative based on that Kaplan-Meier. We then took that feature of normalized membrane ratio; we took it into the TROPION-Lung01 dataset. It was optimized in the non-squamous, non-AGA patients, and that was done deliberately because this was the largest group of patients within the biomarker evaluable group that we had, and then tested in the full biomarker evaluable population. As I've already said, there's a greater prevalence for the biomarker positive in the non-squamous patient population, 66% overall, 63% in the non-squamous non-AGA. And it's also actually more prevalent in the AGA group, which we've already shown in TROPION-Lung01 and in TROPION-Lung05, have a very good response to Dato-DXd. It's a lower prevalence in the squamous population. And again, you can see just for the overall global G7 numbers at the bottom, the prevalence of these different subgroups, non-squamous, non-AGA, AGA, and squamous, across non-small cell lung cancer overall. So here are the Kaplan-Meier curves we showed at World Conference on Lung Cancer last week. On the left-hand side, you've got the non-squamous, non-AGA biomarker evaluable population, which was 221 patients, and on the right-hand side, we've included the overall biomarker evaluable population. You can see that you get a clinically meaningful improvement in the median PFS from 4.1 to 7.2 months, with a hazard ratio of 0.52, in the biomarker-positive group, and similar effect in the overall biomarker-evaluable population. I think what these data do is give us confidence about taking Dato-DXd into the first line, and together with the data that we've had from TROPION-Lung02, TROPION-Lung04, and the NeoCOAST-2 data that I've referred to, I think it gives us confidence about the safety and combinability of that, and the relevance is that there is a patient population that will have particular benefit in the first line. You see, we've got five first-line Dato-DXd non-small cell lung cancer trials ongoing, and across the totality of our program, this represents a significant share of the potentially greater than $5 billion peak year revenue that we highlighted at the Investor Day. So if you look in the first line, of course, first of all, the subgroup that we saw benefiting in the AVANZAR group in TROPION-Lung01 and in TROPION-Lung05, gave us the opportunity, together with the combinability data that we had from the ORCHARD study, to have the basis for the TROPION-Lung14 study in the first line EGFR mutant segment. There were learnings from the TROPION-Lung01 study that we've taken into this first line setting. So what we've done with AVANZAR, we were enriching for the non-squamous population. We have a large study here, it's very well powered for the non-squamous population, and we've capped the number of squamous patients in AVANZAR. We've also taken that approach for the TROPION-Lung08. We've got the combinations for both AVANZAR, Dato-DXd, and Imfinzi plus platinum chemotherapy, and then TROPION-Lung10, where we've taken rilvogostimig into this setting. And then we're including the QCS biomarker validation into both the AVANZAR and TROPION-Lung10 studies. So one question that was raised after the data that was presented last week is about the real-world implementation of this biomarker. And actually, we've got a great track record with Dave's team of actually commercializing successfully, globally, other companion diagnostics. So as we announced with the partnership with Roche Tissue Diagnostics, the intent is to launch with a centralized testing model. This means that institutions can send three slides or blocks to the central lab for analysis and have a routine turnaround time. This is not different from what was done for many other tests. Of course, for some centers, they're going to want decentralized testing model, and again, there is work that's currently ongoing about how we can move this out and roll this out quickly, where institutions can elect to bring the testing in-house with an approved assay, an approved slide scanner, and an image analysis system. So we have the plans in place to do that, and I'm sure if there are questions that Dave or I can answer those. What I'd like to do now is hand over to Professor Tom Powles, who is the presenter and the PI for the NIAGARA study, to walk through the very exciting data that we have for Imfinzi in the perioperative setting of bladder cancer. Please don't fall off the stage. That's okay. Thank you very much. Thank you. Thank you. I've given a few presentations in the meeting, and my voice is going, so I've got some water with me, and I'll do my best. This is only the second time I've given this presentation. The first time was a large audience. At the end of the presentation, I said it was presented at, in the JCO, which is a famous journal, but not the one that I'm the Editor-in-Chief of, which is Annals of Oncology. It was actually published in The New England Journal of Medicine, which is hard to do. The reason it's The New England Journal of Medicine is it is transformative, and I think my colleagues agree with me. I've had a lot of positive feedback about this trial since we published it. 500,000 people die of bladder cancer every year. Neoadjuvant chemotherapy and cystectomy is the standard of care. Half the patients who have neoadjuvant chemotherapy relapse and die from their cancer. Adjuvant therapy with nivolumab, particularly, has a disease-free survival advantage, but no overall survival advantage. The rationale for perioperative therapy, which is before and after, as you all know, is actually quite strong, particularly we've seen in breast cancer, and indeed in lung cancer. And NIAGARA is the first global trial to explore this perioperative space in bladder cancer. There are other trials following, and I think it's important to say here that the AstraZeneca team were ambitious, went first with this, and the trial is really robust. I've looked at all of the other aspects of the trials. If there are problems with it, a lot of that attention will come to me, and I can't find any major flaws in the study. It was an open label trial, and there may be perhaps inherent biases with an open label trial. I'm unable to find that when I look at the dataset. For example, more patients, well, numerically, or a few more patients in the study arm had surgery than in the control arm, and that would suggest, for example, that the fact it was an open label had didn't affect that, number one, and number two is the combination was well tolerated. We've never done a thousand-patient study before in this setting. We may never do one again, I don't know. Prior to this study, I was involved in a plethora of trials that failed to recruit. When people say to me, "Why didn't you put three or four arms in the trial?" We just weren't there in 2018, 2017, when we were designing this trial. Nivolumab was not approved at that time, and the control arm was entirely justified. The patient population represents what we knew at the time, which was about 40% of patients with T2 disease, the majority with T3 or T4 disease. It includes all types of pathology, and importantly, we were able to bring the creatinine clearance down to 40 by splitting the dose of cisplatin. I didn't talk about that at the plenary session yesterday because of time. That increases the potential population. Only about half patients have creatinine clearance above 60, but it's about three-quarters of patients that are above 40. That potentially opens new avenues for these patients. My vision for this cancer in the future is giving systemic therapy, including drugs like nivolumab, initially, and then only patients having surgery who need it, and this trial is the first step towards that goal. The dual primary endpoints of pathological complete response and event-free survival are established in this field, but it's also important to recognize that PathCR isn't the same in bladder cancer as it does in lung cancer or breast cancer. It's not a validated endpoint, and in fact, this is the most robust data. This is the benchmark PathCR data from this particular study, and other studies will be benchmarked off that, for example, the control arm. Some people say the PathCR rate might be a bit lower than seen in other studies. The reality is we've done this correctly in that it's been a central review. We included all the patients in the denominator, and really importantly, we included high-risk patients. Many of the previous studies have gone for mainly T2-only disease. Classic statistical analysis plan, dual primary endpoints, tiny amount of alpha on path CR, most of the alpha on event-free survival, and this is the first interim for event-free survival. OS has alpha allocation as well, and we had 288 OS events for this analysis. It's 1000 patients. The population is well-balanced. The event-free survival hazard ratio is 0.68. It's statistically significant, clinically meaningful. The p-value is obviously low. In the study arm, the median's not been reached. In the control arm, it's 46 months. As I said before, in half the patients, this cancer comes back, and bladder cancer that comes back, patients have died from historically. The benefit is across broad subgroups of patients. This is really reassuring. I don't think any of these subgroups are powered to make definitive statements of, in my opinion. The overall survival showed a statistically significant reduction in the risk of death, 25% reduction in the risk of death. This is actually lower than I was expecting, speaking frankly. You know, we've never seen a survival advantage in this setting with immune therapy before. We've had three goes in the adjuvant setting. This, the PathCR rate, there was about a 10% bounce in the PathCR rate. When you put those two together, and you look at the shape of these curves, it suggests to us that that neoadjuvant period is contributing something significant to this. The benefit, again, across broad subgroups of patients, this is reassuring, and it reinforces the EFS data. It's worth focusing on toxicity. I thought the discussant for the breast cancer presentation was really thoughtful. Marleen talked about how we analyze toxicity for cancer drugs. CTC criteria were designed mainly for chemotherapy, not for immune therapies. This toxicity profile is really driven by that chemotherapy. We forget how toxic chemotherapy is compared to immune therapy, and therefore, you can see here key parameters are really well-balanced in both arms. Grade three or four adverse events, 69% in both arms, 41%, Grade three or four treatment related. Importantly, the neoadjuvant period, that period when the patient's working up to surgery, was not associated with a reduction in the surgical rate, surgical deaths, or complications. So patients were getting to surgery, and indeed, more patients in the study arm had surgery than the control arm. One of my worries when we designed this trial was the other way around might be the immune therapy may cause immune-related toxicity, and many patients may not get to surgery. The opposite is probably true, in that we might have been getting more control of disease under those circumstances, more patients were getting to surgery. Adverse event profile dominated by chemotherapy-related toxicity. I tell my patients there's about a 10% chance of life-changing toxicity associated with the immune period. That ranges from things like diabetes, some people say some hypothyroidism, but there are toxicities associated with the adjuvant and neoadjuvant approaches. It's the first study. It opens a new chapter, in my opinion. I'm very excited about the future because there are future exciting studies in this area. The trial was well conducted, and it was large, and therefore, it's a positive study. It doesn't necessarily mean that all the studies in this space that will follow will also be positive. The conduct of trial is really important in bladder cancer in terms of delivering positive results, and the AstraZeneca team, in my opinion, should be congratulated for that. The benefit across subgroups is reassuring. The pCR rate and the OS signal suggest that neoadjuvant phase is making a major contribution. No new safety signals, no surgical problems, and it does support this as, I would have said, a new standard of care. I would say actually it is the new standard care of care, because I think patients with this disease will want to get the immune therapy up front and get it in, because that's the best way of long-term durable remission. My voice lasted. That was my last talk for the meeting. I'm very happy to answer any questions. Thank you. Perfectly timed vocal cord stamina. It was great to hear Professor Powles really sort of bring this amazing data. As you may recall, last year's ESMO, he had the chance and opportunity to shape first-line kidney cancer care, and I think this data certainly has, he, in his words, "established a new standard of care in muscle-invasive bladder cancer." So we wanted to give you a perspective on muscle-invasive bladder cancer and also a view towards how we feel that this is not only going to be able to hopefully change standard of care, but I think we have the commercial execution in place to be able to bring along this change as well. As Professor Powles noted, muscle invasive bladder cancer is about one in five bladder cancer patients, and if you take a look at G8, that's about 120,000 patients across that region. The standard of care has remained the same for the last 20 years, where patients get neoadjuvant chemotherapy with cisplatin-based treatment followed by surgery, which is quite debilitating, and no real options thereafter. Recently, there's been approvals for adjuvant, but as was noted, that has only lead to improvement in event-free survival and has not led to improvement in overall survival. This is the first study, and that has shown perioperative therapy does improve both event-free survival and overall survival. Now, a couple of points to highlight as Professor Powles noted, the cisplatin-eligible patients is generally considered to be about 50% of the patients, but as was noted in his presentation, given that we included creatinine clearance up to 40 mL per minute, that allows us to look at expanding that patient population even more so. And in his words, that could be up to 70%. And furthermore, these patients are at high risk of recurrence, so there's a significant unmet need for 50% of these patients at the risk of recurrence within five years of follow-up. And given these results, we feel that this addresses unmet need, and also the clinical community is really excited for this data. Overall, the bladder program is quite robust. We have four important phase III studies, and I think based on the NIAGARA results, it gives us high confidence for another study that's underway in the muscle-invasive patient population, which is the VOLGA study. And when we take both the NIAGARA and the VOLGA alone, we believe that this is, in itself, a $1 billion-plus opportunity. So we have a lot of confidence when we think about how we're going to move this forward. We have multiple studies with now durvalumab and Imfinzi across lung cancer, gastric cancer, and now, bladder cancer, that show that this impact and improves early-stage disease outcomes, in this case, in NIAGARA, improving survival. We are very confident that given the well-characterized safety profile, as well as the experience that clinicians have in giving neoadjuvant therapy, that this is going to lead to a substantial change in clinical care paradigm with greater use of perioperative therapy, and finally, we have a lot of experience working in PACIFIC, now coming up in AEGEAN, in ADRIATIC, in working with multidisciplinary teams and working to make sure the healthcare practitioners and the clinicians have confidence in upfront, immediate systemic therapy, so they don't have to wait for surgery, and as Dr. Powles noted in his presentations, by giving treatment earlier, it actually may improve the rates of surgery, and that surgical engagement is really critical. So we feel that there's a role for oncologists, there's a role for surgeons, and also, just as importantly, our teams have a very established connections with the GU oncology community that hopefully supports the discussion of this therapy as it gets approved. So with that, I'll pass it on to Dave. Thank you, Dave. ... Okay, thank you, Sunil. Thank you, Professor Powles. So we'll have a Q&A session right now. If I could request that we focus in on using Professor Powles' time well right now and focus in on BLADDER and NIAGARA, that'd be great. We will then take about fifteen minutes now, go on to the rest of the program, and have another Q&A on the end of this. Luisa, you are first. Good student. Go ahead, please. Oh, Andy trumped me. Sorry, Luisa, you'll be maybe later. All right. Please, Andy. Yeah. Okay, yeah. Thank you very much, Andy. Tony Ren from Macquarie. Actually, if I may, I would just ask one quick question about TROPION-Lung01. Understand that the Understand that the QCS was not part of the trial protocol. So now that you have the data, would you submit that data to the FDA? So if we answer your question, we've broken my rule that we set out at the front end of this, which will then encourage others to do that. So can we come back to that question, please, in the regular Q&A? That'd be great. Thanks. Happy to address it and happy to get to it, but I think that that's what we ought to do. Go on, go ahead. Yeah. Go ahead. Okay, thank you. I have a NIAGARA question, and thank you very much for the presentation. So Professor Powles, you describe NIAGARA as the new standard of care, and hopefully very soon it will be available, but you will have a choice between perioperative and adjuvant regimens. How would you compare NIAGARA to the IO adjuvant trials, so CheckMate 274 and AMBASSADOR? Are there differences there? I think obviously, the overall survival stands out for NIAGARA. But how would you think about maybe choosing between the two, or is it just a no-brainer, it's NIAGARA all the way? I think it's a really good question, and it's the one which we've thought about in the community already. Some people will try and make indirect comparisons between the two trials. The hazard ratios of 0.7, 0.68. These two trials are very different, and the patient population is different. The adjuvant patient population go in based off their pathological stage after a surgical operation. They start therapy twelve weeks after a big operation. The way they measure events for progression-free survival is different from event-free survival. So it's a different patient population, it's a different way of measuring the primary endpoint, and they don't have overall survival. I think we are bringing patients into this trial on a different clinical T stage, T2 or more, number one. Number two, and I think this is important, is that that initial component of starting early and getting the immune therapy into the patients without more toxicity during that neoadjuvant phase is quite attractive. Many patients will never get immune therapy if you do that surgery, because by the time you've done the surgery and you've got the complications and problems associated with that, and the patients come back to clinic, many of them will say: "You know what? It's not for me," or they can't have it. So I think the clinical community will look at this and say, "This is an opportunity to get these drugs." And it is the immune therapy drugs that are associated with these durable remissions in the metastatic disease, not the chemotherapy. It's a way to get these durable remission drugs into these patients early before we start. It's my feeling that the data supports the neoadjuvant approach. The cross-trial comparisons between these two studies is somewhat futile. Irrespective of that, the fact that one of the approaches has overall survival and the other doesn't, suggests that this is the right approach. The AMBASSADOR study, with all its challenges, and there was some censoring on their Kaplan-Meiers, it's worthwhile looking at that, but that has no survival advantage at all. Then finally, and I think this is important in Europe, PD-L1 is used as a restriction for adjuvant therapy in Europe. The Europeans, we don't love that. We don't love the PD-L1 biomarker. This is a great excuse from a global perspective to ditch the PD-L1 biomarker, and I think we will embrace that strongly. Professor Powles, and I think that on that particular point, it's worth noting that in the U.S., we estimate an epidemiology of about 20,000 muscle-invasive bladder cancer patients. In Europe, it's about 35,000. Japan, it's around 10,000. So I think that it's certainly a sizable portion of patients. And maybe, Professor Powles, on this, as you think about that, you know, are there patients that you would not consider for perioperative? And if so, kind of how big of a population would that be? I think the question you've asked is a really interesting one, and I wish you could come to clinic with me, because these patients are an unusual group. They tend to be in the seventh, eighth, dare I say it, ninth decade of their lives sometimes. Bladder cancer, muscle-invasive bladder cancer is a really lethal disease. The majority of patients who we see almost certainly have MRD at baseline. The vast majority, for example, are ctDNA positive. And so I think the mindset is of patients at the moment, and has historically to some extent been, because neoadjuvant chemotherapy is associated with modest benefits and quite a lot of toxicity, I think the mindset from a urology perspective has been: Just get the operation done, and that's been quite powerful. I think with this new survival signal, with an immune checkpoint inhibitor-... I think there will be more enthusiasm around this approach, and I think it's possible or hopeful that we see increased uptake of neoadjuvant chemotherapy with durvalumab, which has been lower than we would have liked historically. So I think there are two things there. Number one is, I hope the data will encourage this approach rather than waiting. And number two, because the creatinine clearance has gone down to 40 rather than 60, it immediately increases probably 25% of patients. Thank you for that, Powles. Emily? Hi, Emily Field from Barclays. You may have just answered this, but I was gonna ask just about this patient stratification more based on platinum eligibility. It seemed that the trial design did allow for perhaps, you know, greater eligible patients in the trial, and then obviously, some turn it down. So how are you thinking that fifty-fifty split, based on the convincingness of this data, might evolve in the real world in terms of patients going to get this regimen? So what I can't. My ability to predict the future is fairly limited. I get things wrong all the time. What I do know is that when I started as an oncologist in urothelial cancer, surgical communities, particularly in the U.S., were working by themselves. In community practices, most patients were not seeing an oncologist and were not being offered neoadjuvant therapy or indeed adjuvant chemotherapy. That tide has changed quite dramatically over the last five or 10 years, and the introduction of what will be Level One A evidence with a survival advantage in the neoadjuvant setting, I think will encourage more collaborative working, and I think more patients will go down this approach. Hi, Chris Ruitenberg. Can you hear me? Yeah. Yes, I can. Chris Ruitenberg from SEB. Thanks for taking my question. So NIAGARA, you mentioned during your talk, Dr. Powles, that cystectomy is a very difficult procedure. Sounds quite horrible. How real world is this trial then? You said, I mean, 85% of patients got it. What percentage actually get it today? How much work will it take to get patients through it? What's the ceiling for how many you have, and is it important for the use of Imfinzi? And sort of then in the bigger picture, thinking about NILE, what's the opportunity here in light of EV-302? Are there niches where the benefit-risk favors NILE? You know, as NIAGARA illustrates, you've got Imfinzi now in a lot of early settings where the tolerability profile is good, while volrustomig, with the current dosing scheme, seems a bit tough for that kind of setting, but rilvogostimig's tolerability profile seems good enough if that could replace, you know, Imfinzi, if more efficacious. Do you think it could, or do you need to revisit Imfinzi SC to, well, improve quality of life, let's say? If so, where? Yeah, I'll stop with that. So Tom, sometimes we have to sort of go through some layers on the questions there. You got a few. I think that maybe- I think I've got them. You're good. I hope I've got them. Okay. The first question is there's a very long answer, and I've got a really nice 20-minute talk, which I'm not gonna do now because I think Pascal will have a fit. But I am gonna give you the short version. As it currently stands, you wanna have a cystectomy. You don't wanna have radiotherapy. The radiotherapy data, although the radiotherapy doctors talk about it positively, from two randomized phase III trials, they're not real randomized phase III studies because both arms had radiotherapy, but one had chemo rad on top of that. When you look at the PFS data and they report. And again, it comes back to this dataset being the definitive dataset. They report metastasis-free survival, not event-free survival. Everyone's doing it slightly differently in the past. What you see from that on the survival curves is the radiotherapy patients do poorly. Now, the radiotherapy doctors come back and say: "Yeah, that was 'cause bad actors were included." And that's not necessarily the case. We don't know the detail of that. As it currently stands, therefore, with this dataset and these, I think, attractive survival curves at landmarks, this is the best way of maximizing survival, and I think that will be an important statement moving forwards. However, patients don't want to have a cystectomy, and in fact, when patients come back, and in this study, about 40% of patients had a PathCR. When those patients come back and you tell them they've had a PathCR when their bladder is in a pot and there were no cancer cells there, you would expect them to say, "Fabulous, it's been a great success," but they don't. The first thing they say is, "Why have I had my bladder removed? There's no cancer there." And it is possible that this trial is gonna kick off a transformation in this disease, and VOLGA has an important part to play in that transformation. If the EV- durva- tremi or EV- pembro data shows PathCR rates of 50% or 60%, and that's not inconceivable, and the reason why it's not inconceivable is it has a CR rate of 30% in metastatic disease compared to historically 10% for chemotherapy. If that is the case, then what we're gonna start doing is we're gonna stop doing cystectomies, and it's possible we can cure muscle-invasive bladder cancer with systemic therapy and not surgery, and that would be the first step of that transformation. The second question you asked about was NILE and EV-302. I was the PI on the 302, but I was the PI of a number of other combination trials, and I hope I'm equally unbiased in this. I think the 302 dataset is robust. I think the NILE data would have to be outstandingly good to be 302. I also think that while the 901 data with nivolumab and chemotherapy, and dare I say it, even the pembrolizumab data with pembrolizumab and the cisplatin cohort, shows actually the cisplatin subgroup do get some bounce, which bizarrely supports the NIAGARA approach. We didn't include the carboplatin patients in this study. Had we done, we might not have got a good result. I don't think in advanced disease, it's gonna be easy to beat EV-pembro. Your third question was about subcutaneous durvalumab, I believe, and I probably don't know enough about that to give you a really good answer, I'm afraid. We published in Annals of Oncology, not in the JCO. In Annals of Oncology, a randomized phase III from this meeting of subcut nivolumab in renal cancer. I hear very mixed patterns from my colleagues about how important that is. Some people say the pharmacy time doesn't change very much. Some people say to me, "It's gonna make a big change," and many of my U.S. colleagues are not as excited about it as I would expect. Professor Powles, can I ask? I think there's another element that's relevant. Can you talk about VOLGA? So certainly, your NILE commentary was useful, but can you also speak about how you see VOLGA fitting in within this same population as NIAGARA in the competitive context for what's coming next? Drug development in urothelial cancer has been more challenging than we expected. The initial atezolizumab phase I data and some of the melanoma data, and the high TMB and high tumor immune infiltrate suggested that immune therapy in bladder cancer would have the same results as melanoma, and it hasn't had the same results. You know, ipi/nivo has not cured bladder cancer. Because of that, some of the early trials we did, we were overly ambitious. I think the tremelimumab subset in the DANUBE trial, across the board, irrespective of the PD-L1 status, shows about a 10% bounce in pretty much all of the endpoints. The hazard ratio is about 10 points low, the response rate is about 10 points higher. I do think Tremelimumab has something to add in this setting, and the NADIM trial with ipi/nivo, and I'm not suggesting this is melanoma, but the NADIM trial suggested early CTLA-4 may be important. I had the opportunity to take part in different trials. You can imagine, I had lots of opportunities in my life. I decided the VOLGA trial was the most attractive trial in this space for two reasons. Number one is, I think CTLA-4 may have more important long-term effects than the adjuvant enfortumab vedotin that the other trials are employing. So in the EV-pembro trials, you get three neoadjuvant cycles of EV-pembro, and you get five adjuvant cycles of EV-pembro. EV is not the hardest drug in the world to give. It's easier than platinum-based chemotherapy, but many patients during the adjuvant phase will wanna get away from that type of drug, and if we can show similar results with VOLGA, or dare I say it, in the triplet arm better 'cause of the CTLA-4, we'll have two advantages: one, potentially better long-term durability, better tail on the curve associated with CTLA-4, which we believe is the case, number one; and number two, potentially less toxicity associated with decreased exposure to enfortumab vedotin, and that's why I'm taking part in VOLGA. Thank you, Professor Powles. We're gonna go to one question online. Simon Baker at Redburn, you can ask your question, please. Thank you, Dave, for taking the question, and thank you for the presentation. It's been really helpful. One thing, Professor Powles, you mentioned in the presentation on Sunday was around the apparent failure of PD-L1 status to stratify response. I just wondered if you had any thoughts on why that may not, it may not have been a good stratification factor here. We've seen mixed results elsewhere, and I just wonder what other potential stratification approaches have been considered here. And then other question, going back to cystectomy. The overall levels of cystectomy between the two arms were the same. Was there any difference between the proportion of radical and partial in the two arms? Thanks so much. The PD-L1 story in urothelial cancer is a long and complicated and not always happy story. I've taken part, I've led three randomized trials where we've explored the biomarker specifically, and they've all failed with three different drugs, one of which was pembrolizumab, one was durvalumab, one atezolizumab. The PD-1 biomarker in urothelial cancer has never been successful by itself. It's only been successful when the whole trial was successful, and then the PD-L1 subset, there was some enrichment. By luck alone, there's gonna be some enrichment in some of the trials. The immune biomarker in urothelial cancer is prognostic, and the tumor biomarker is maybe predictive, maybe prognostic. It has a very sketchy record, in my opinion, and when you look in more detail, there are more probably useful biomarkers, even from the tissue, and these biomarkers include tumor mutation burden and T effector signatures. I think even ctDNA is a better biomarker. It's clear that immune and adaptive signatures are associated with response to immune checkpoint inhibition, and I'm afraid PD-L1 expression by IHC is a gross oversimplification. Not until we start moving away from this biomarker in urothelial cancer can we start a new chapter of biomarker discovery, and that's why I think we do need to move on, and I'm keen to do that quickly, and that's one of the things which I think NIAGARA will do successfully. Thank you. We'll take one more in the room, one online, and then we'll move on to the rest of the program, please. Thank you, Dave. It's Sam Fazeli from Bloomberg Intelligence. Dr. Powles, your patient who just said to you that you told him he's got, or her, that who's got PathCR, "Why did you take my bladder out?" How are you gonna convince them to take adjuvant therapy? Yeah. So, so I think that's a good question. I'm gonna answer two ways, if I may. One is the reality of where we are today, and the second is a fantasy world where I spend a lot of my time. The first, the reality of today is that in urothelial cancer, not in melanoma, with CTLA-4 PD-L1, in urothelial cancer, three months, four months of immune therapy is not currently long enough, in my opinion. And I think that if those patients that had aggressive disease and muscle-invasive disease, and we may use tools like ctDNA in the future as well and get ctDNA clearance, if you've had that pre-surgical response to immune therapy, I think stopping after three months is too short. I think you need the rest of the year. I don't think you need the rest of the year of, you know, I don't think you need the adjuvant period of chemotherapy, but I think at the moment we're not in a position to stop early. And if you had said to me, would I have been happy to do a three-arm trial with neoadjuvant just by itself and the perioperative and the control arm, but we had to reduce the power to, let's say, instead of 500 patients an arm, 300 patients an arm, so there's still 1,000, I wouldn't have been happy with that. It was too risky when we did it, and currently, when we treat metastatic disease, I think we all feel that two or three months of therapy is too short. So that's the answer to the first part of the question. The EV-Pembro data, the 303 and 304 trials, they're ahead, and they will read out first, and when those trials read out and there is a PathCR rate that's high in those trials, I think at that point, the mindset of the doctors is gonna change, and the mindset of the doctors will be one of two things: Can we get a biomarker at the end of surgery associated with low risk, and we need to see more subset data from those two trials and the VOLGA trial that comes forward, and so those three EV-Pembro, EV-Durva-Tremi trials, which I think will be positive, will redefine this space. And I also think finally, there will be some bladder-sparing studies in the future, and they will be specifically designed not to have surgery in them for those patients that have had a PathCR, and then we can explore this topic formally. And I think both EV-Pembro and EV-Durva-Tremi will be explored in that space. Thank you, Professor Powles. Last question for this section online to Alana Lello at Guggenheim. Please, Alana. Alana, if you can hear us, we can't hear you. Okay, well, we'll move on. What I would like to just offer on this as a wrap-up to this particular section is I think that Imfinzi, over the course of the last 12 to 18 months, has demonstrated what we can do with what has been some underappreciated indications in terms of the opportunity that it creates to bring an enormous amount of benefit to patients and drive growth. I think that TOPAZ-1, HIMALAYA, have been very good examples of that. We just had an opportunity to talk about how with Imfinzi moving into the perioperative setting and really moving into a curative intent setting where we understand how to engage with multidisciplinary teams, we understand how to have these conversations. We started it with PACIFIC. We will be doing it with, we're doing it right now with AEGEAN, obviously with ADRIATIC. There's another opportunity to, in a similar sort of curative context, and then of course, we'll have NIAGARA. I'm really confident that these are all opportunities that allow us to be able to propel Imfinzi forward and to continue to stay on the type of path that we're on. Professor Powles, thank you for joining us. I know you have somewhere that you need to be to next. We appreciate it, and we'll turn to our next section. Thank you very, very much. All right. Good evening. I am delighted to have the chance to discuss our new in-house ADCs that we're unveiling at ESMO. So, as you've heard from us before, a core goal for us is to replace traditional chemotherapy with ADCs and radioconjugates. As you can see on your left, this is a profound clinical and commercial opportunity, and to achieve that goal, our first step has been investing in building a topo-based platform that allows us new opportunities with new targets. And I'll talk a little bit about those new assets that we're unveiling here at ESMO. But secondly, we've also been investing in delivering a comprehensive industry-leading ADC toolbox that allows us continued innovation, which includes diversifying our payloads, antibody engineering, as well as proteomics, surface proteomics to enable new targets into the clinic. In thinking of our portfolio today, we now have six wholly owned ADCs that are currently in the clinic, and the two of which we're discussing at ESMO here are shown on the right. One particularly important feature is the proprietary linker payload technology that's embedded in our wholly owned ADCs and is relevant for thinking about the PK profile and overall platform that I'll show on the next slide, and when looking at these PK data, I want to draw your attention to three key things. The first are the orange and blue curves at the very top. This is the PK of the overall antibody, as well as the whole ADC component. What you want from an ideal ADC is that these two curves are overlapping. That represents an ADC that is stable in the peripheral circulation and not falling apart. And that's exactly what we see in both of these assets, both AZD8205 on the left and AZD5335 in the middle. The second component I want to draw your attention to is the gray curve at the bottom. This is the exposure to the free payload, and as you can see, it is much, much lower than the exposure you get with the ADC and the antibody above. And it's particularly more dramatic when you recognize that the y-axis here is on a log curve. And so, as Susan alluded to before, this free payload represents within these molecules about 150,000th of the exposure you get to the overall ADC. This is implications that are important in terms of considering safety, as Susan mentioned before. The third key component is the half-life, and you can see the durability of these curves as they extend across the x-axis, and this is important for delivering the convenience of a Q3-week regimen. It's also important for developing the ability to combine with other regimens. As I transition then to talk a little bit about the data that we've presented here, we've had fantastic feedback from investigators across the clinical space. The first molecule I'll present here is AZD8205. This is our B7-H4-targeting topo ADC. We were pleased to see activity across a range of dose levels, as well as a range of diseases. And you can see that we're particularly enthusiastic about the efficacy that we saw in endometrial cancer. This is an opportunity which may be complementary to opportunities that we have within HER2 in bringing ADCs to this key disease. In terms of safety, we've heard from investigators is that the safety was expected for what you'd see with a TROP2-based class, and that this sort of toxicity is manageable, reversible, and predictable. What we've also heard is the importance of not seeing high rates of dose discontinuations, which are at 4% with this molecule so far, as well as the importance of some toxicities we didn't see. We don't see stomatitis, we don't see ocular toxicity, we don't see neuropathy. Together, that safety profile is enabling combination opportunities, which we're excited to bring into the earliest lines of therapy, and we've begun already with phase II trials in combination with rilvogostimig. We're also excited for the opportunity, as the discussant highlighted, of where QCS might play an additional opportunity in bringing us to an even more strengthened patient selection opportunity. The next asset I wanted to highlight was AZD5335. We have a very encouraging efficacy and safety in this ovarian cancer population. This is an ADC targeting folate receptor, where our opportunity to bring a topo-based ADC to a disease that's so sensitive to topo is incredibly exciting. And what we've been encouraged by includes the activity that we're seeing in FR alpha high population, with an activity of about 55% response rate, as well as for the first time, seeing some activity in the FR alpha low population. This is a place in which there is not currently an approved FR alpha targeting ADC. Much like we described with the previous molecule, we've also had an encouraging safety profile, where the safety has been manageable for clinicians, and they feel very comfortable with that safety profile. But again, there's been some key toxicities we haven't seen, which have included ILD, stomatitis, ocular toxicity, as well as neuropathy. Currently, we're in dose escalations and expansions at key doses, beginning at 1.6 milligrams, 'cause we've seen responses across these doses, and we're excited for key opportunities, both in ovarian cancer and soon to begin in lung cancer. So with that, I'm going to pass the baton over to Christian to talk about our next generation IO. Thank you, Matt. Hello, everybody. So let's go briefly into the bispecifics, the next generation IO. We-- Let me start with this. You probably know this. We discussed this, we presented this to you a few times, but just to reinforce the message that we have two bispecific, they have a unique mechanism of actions because there is a cooperative binding in the presence of both checkpoint inhibitors. Rilvogostimig is a PD-1 TIGIT. There is a very similar affinity for PD-1 and TIGIT, actually slightly higher for TIGIT, so co-binding cells that are co-expressing the two targets. This is a molecule that importantly has an Fc attenuated. This is potentially one explanation for the excellent safety profile we are observing currently with this drug, with very few, severe, immune-related adverse events and overall, a very, very low discontinuation rate, very higher combinability of this asset. volrustomig is a PD-1 CTLA-4 in which actually it has been designed to with much higher affinity for PD-1 binding and less for CTLA-4, to optimize the safety profile of the drug, and fundamentally, this co-binding of PD-1 and CTLA-4 in the same cells hopefully can decrease some of target toxicity. As you see, the phase III program for both these agents is growing very importantly. Each time we are presenting additional phase III studies. Here you have for rilvogostimig the strategies combining with chemotherapy and with our ADCs. We have a phase III study in biliary cancer in combination with standard chemo. We have TROPION-Lung10, as has been presented by Susan, in which we are combining with that to PD-L1 more than 50 non-small cell lung cancer. We have a combination with an HER2 in BTC in frontline, and TROPION-Lung12, that is an early stage study, a stage one study in which we are using very novel study. We are using GRAIL technology to select high-risk patient and the patient post-surgery, randomized to Dato-DXd or placebo. In this program, and this molecule is very complementary to our PD-1 CTLA-4 volrustomig, because the, the-- while the development of rilvogostimig is specifically in those tumors with high PD-L1 expression or PD-L1 positive, rilvogostimig is focusing CTLA-4 sensitive diseases like cervical cancer, mesothelioma, head and neck cancer, where we are running studies post chemo radiotherapy or in combination with chemotherapy, or in those spaces where PD-1 inhibition is not working very well. Non-small cell lung cancer, PD-L1 negative, or less than one, this is EVOLVE-Lung02. Let me share some data with you. In May, we promised you we will share along the year data with our bispecific. You were asking, where are the data supporting your phase III investment decisions or your phase III studies? These last few weeks between lung cancer meeting and this ESMO you have seen you start to see some data sets. This is an important data set because it represent for rilvogostimig the place where you can start to benchmark this drug versus other PD-1 TIGIT. It is a first-line non-small cell lung cancer, PD-L1 positive. There are a little bit of colors in this plot but focus on the fact that there are also points. There are responses. The points represent the PD-L1 expression. You have patients with PD-L1 1-49 and patient with PD-L1 more than fifty or half and half. The activity we have seen in terms of response rate is 29% in PD-L1 1-49, and more than 60% in PD-L1 over 50. If we put this in the context of what you would expect with a PD-1 inhibitor like pembrolizumab, in 1-49 is approximately 15%, and in PD-L1 more than 50 is around 40%. An interesting part also of this data set that we see very durable responses, and in both PD-L1 more than 50 and in 1-49, of course, is better in higher expression. The tolerability is excellent. We have almost no severe grades of immune-related events. Very few patients discontinue the treatment, and at the time of the cutoff of this data analysis, more than half of the patients were ongoing. So 750 is the dose that we are bringing ahead as a in combination with chemotherapy, and the combinability is very good. With regards to the combinability, this is the data we presented at this ESMO. Here you have the in another data set in gastric cancer, where other PD-1 TIGIT has been presented, and I think also this data set is comparing very favorably. This is a setting of first line in which we combine rilvogostimig with FOLFOX, and you see that there is a remarkable response rate, more than 60%, that is reaching more than 80%, if you look at the patient that has a high expression, CPS more than 5 for PD-L1. The median PFS is more than eight months in this data set, that reaches eleven months in PD-L1 positive patients. And again, very good safety profile. Most of the adverse event observed in this study were coming from the chemotherapy regimen. There is 10% grade three adverse event attributed to rilvogostimig, only less than 3% discontinuation for rilvogostimig. So promising data. Let me move now to volrustomig. We updated the volrustomig data at World Lung. This is the study that we ran in combination with chemotherapy in front line in non-small cell lung cancer. Overall, we have seen around 44% response rate. This is a non-squamous setting, a combination with platinum pemetrexed, with a very good DCR rate, disease control rate, around 85%. Now, what is really remarkable and important and differentiated with this data set is the activity in PD-L1 negative, less than one. You have a 34%, 34.5% response rate. That is very high when you think of what the pembro chemo can give you in this space, and a median PFS of seven months. That is exceeding quite a bit what you would expect with a checkpoint inhibitor, PD-1 plus chemotherapy. But it's also important, we start to collect multiple data set with rilvogostimig. We have four pivotal studies ongoing. We are finally start to be confident in a manageable safety profile through a very accurate and prescriptive treatment management guidelines that we give to the investigators. And so, this is this data set is the basis for EVOLVE-Lung02. That is, of course, a very important study in this program, in PD-L1 less than 50, with a primary endpoint, of course, that is will explore initially PD-L1 negative, less than one. I stop here. Hopefully, this give you confidence what is underneath the bispecific program. When we discussed this meeting with external panels, so individual investigators, the profile of this drug, we see very positive feedback on this. Dave, I think it's to you to conclude and go to the questions. Great, thank you, Matt. Thank you, Christian. So as we come into the closing remarks, Pascal started with a view on this, certainly at the enterprise level. When we take a look at the oncology level and what we're building, I think what's pretty important, if we just reflect on WCLC and together here at ESMO, what we see is that if we just come from left to right within HER2, with DB-12, we see a catalyst that unwinds risk and allows us to continue to grow on in HER2. On Imfinzi, with HIMALAYA five-year survival, we talked about NIAGARA. We see opportunities for continued growth coming from here. We've got opportunities with Tagrisso that I think are reinforced by continuing to have a belief in the Tagrisso portfolio against the data that we've seen that's from MARIPOSA and MARIPOSA-2. As we move then to the right here on the launching NMEs and, really the platforms that sit behind that, Matt has taken us through the ADCs and the belief that we've got, certainly in the Daiichi Sankyo partnered ADCs, but also in our own ADCs that we're building of our proprietary ADCs in-house. Christian has walked through on the bispecifics, Rilvogostimig, volrustomig. And so I really think that as we reflect on what has happened just over the course of the last week, is that we've seen both an unwinding of risk within the existing medicines LCM and a building of confidence in the pipeline that's to come. We've had significant news flow throughout all of 2024. Obviously, not everything that comes through is positive, but an awful lot that has come through has been positive. We've got significant news flow coming through for 2025, and I think, and Pascal said this, when we were, together, in May at the investor meeting: by the end of 2025, there's a fair bit of the readouts that we'll need to have confidence to be able to have oncology get to its necessary contribution of that $80 billion. There'll be an awful lot that will have been unwound by the time that we get through some of these really, really important studies that sit within here, and with that, I would like to turn to the Q&A session. I think we will open it up. Tony, we'll come back to you, and we'll let you ask the question now. Sorry that I put you into the brief penalty box on that. Not a problem at all. I was just chomping at the bit to ask the questions about lung cancer. So again, Tony Ren from Macquarie Capital. So, the TROPION-Lung01 application, right? You guys already submitted. I believe QCS was not part of the trial protocol, so now you have pretty good QCS data. Would you resubmit that to the FDA? Susan? Is this microphone on? Is it on? Okay, can you hear me okay? Yes. All right. So, the anticipation is that the QCS data that we've got will need to be validated in a study. So as I've said, you know, the anticipation is we use this in and take it into the first line setting, and we require validation in the first line. We are, of course, you know, sharing data with the FDA. So we share data with the FDA on a continuous basis. We're sharing the QCS data, but we'll anticipate that it's in the first line setting that we're gonna use the QCS data. Thanks. I think, Tony, when we think about kind of the summary overall of this on what QCS allows us, I mean, one, within the second line, it helps to create and explain the biology on some of the histology differences that we see, and I think it's useful in that regard. Then we can incorporate the biomarker, which has already been pre-specified into AVANZAR, and it opens up the potential to unlock outside of the existing kind of set of tumors where we're able to play right now. Okay. Yes, please. Yeah. Richard Parkes from BNP Paribas Exane. I've got a couple of questions for Christian on the bispecifics. Firstly, on volrustomig, I remember two years ago, we were at ESMO, and you presented that randomized phase II study, and that's easy for capital markets to interpret, and maybe we get too complacent when you've got randomized trials. But can you just help us a little bit more now we've got sort of single arm kind of extension with a lower dose in PD-L1 negative? How we benchmark that and how what makes you continue to be confident about the better efficacy? And then can you also talk about the safety profile and exactly what you've seen with those treatment-related deaths and how you're kind of managing that? So that's the first question. The second one is on rilvogostimig. Could you talk a little bit about how important the cooperative binding is to your expectations of differentiation and what's underpinning that? Because obviously, we've seen the power of cooperative binding with bispecifics, with the VEGF-PD-1, so just interested in a bit more detail there. Thanks. Thank you for the question. Let me start with volrustomig. You know, that was an initial data set, and that was presented with initial dose of 1,500. And then we developed, actually, our team in early development developed these multiple different cohorts in which we presented the TROPION-Lung that were running parallel but not randomized versus standard of care. Acknowledge the limitation of a single-arm study. I think the purpose of that cohort was also to optimize the dose, and I think this is what happened because we were able to lock down 750 at the right dose. And this is actually bringing me to the second part of your question. With that starting dose in combination with chemotherapy or as a monotherapy, we then we have a very good target coverage for CTLA-4. Actually, we deliver constant CTLA-4 inhibition in a substantial way on top of PD-1 inhibition. This is probably is the only drug in the space of IO immunotherapeutics at the moment, in which we allow a dose reduction. Not only interruption, also reduction, and this is actually helping a lot the investigator currently to manage the safety profile. Fundamentally, this drug, if you bring down to five hundred, you continue to have a decent CTLA-4 target coverage. But of course, you may reduce some of the adverse event, the CTLA-4 related, specifically liver toxicity. That is the most frequent adverse event, AST/ALT elevation, fundamentally. Asymptomatic, but you have these lab findings. And then you can also reduce further to two fifty, and this is the target coverage for CTLA-4 is becoming minimal, marginal, but remain a very good PD-1 inhibitor. So this is the first strategy in place that we have in the context of the phase III studies, three, four phase III studies that allow investigator to interrupt in case the patient experience toxicity, restart the same dose, or go down one dose and continue the treatment. In addition, of course, because we characterize better the safety profile of the drug, i.e., as a monotherapy and in combination with chemotherapy. Let's not forget that it's not for granted to be able to give a CTLA-4 PD-1 inhibitor with chemotherapy. Was not possible with nivolumab, was very challenging with Durvalumab. You need really to decrease very importantly the dose of CTLA-4. Here we are able to do this because of this strategy. The patients are managing carefully the lab for the liver, and of course we educated very well the sites. The phase III are going well, are enrolling at a pace. So we believe that this strategy with volrustomig, the safety, this is a very active drug. We will share additional data going further, especially data in early setting non-small cell lung cancer. There is an arm that we are running with volrustomig chemotherapy, neoadjuvant and non-small cell lung cancer, part of the NeoCOAST study. You have seen additional arms. We didn't share volrustomig, but we will going forward. This is a very powerful drug in terms of efficacy. Safety is an important aspect that we manage in this way. With rilvogostimig, the cooperative binding is. I mean, the power of bispecific, you mentioned the PD-1 and VEGF, is apparently giving more than what you would expect with a PD-1 and VEGF inhibitor separately. The cooperative binding, in our view, being able to target the PD-1 and TIGIT in the same cell, can engage even more the cells, and can potentially increase the activity that we see with PD-1 inhibition alone. But what is also very relevant is not only this, is the lack of some toxicities that other combination with two different antibodies given together are observing. There were data presented this ESMO from GSK that show a very similar activity in PD-L1 positive non-small cell lung cancer, a very different safety profile. So we believe that and don't forget we believe that this is a very important differentiator point. Don't forget that this drug for us is critically important in our strategy of bringing novel combination, improving on top of bispecific checkpoint inhibitors with this bispecific and on top of chemotherapy with ADC. And we can combine rilvogostimig with a HER2, with Dato, with standard chemotherapy, as I just showed. So it's a very important asset in our strategy and a big enabler for all our portfolio. Thanks, Christian. Oh, yeah, please. Another question, which is about the benchmarking versus other IO agents in the PD-L1 low, I think. Just to come back to that question about the benchmarks. We see with chemo, PD-1 in PD-L1 negative lung cancer with around a response rate, somewhere around 30%. I think it was 32% in KEYNOTE-189. And a PFS of somewhere between four to six months, depending on real-world data and published data. In that context, having seen response rates in this PD-L1 negative population with chemo plus volrustomig ranging from 34% in one cohort, but up to 45% in other cohorts, and PFS in this most recent cohort of seven months and ongoing, with a flattening of the PFS curve, is part of what gives us the encouragement that this is providing new opportunities in a place of key unmet need for lung cancer. Thank you, Matt. Appreciate that. Andrew Berens. Great. Please, Andrew... your microphone's not on yet. Hold on. Try again. Andrew Berens, Leerink Partners. Yeah. This is a question, I guess, for Susan, and about the QCS testing. What do we know about TROP2 trafficking after it delivers the payload? I guess what I'm asking is, what happens after step three in that schematic that you showed? Does it stay in the cytosol, or does it go back out, or is it degraded? And I guess, after you answer that question, I'd like to put that in the context of whether or not having more cytosolic TROP2, we know that that actually means that the cell is actually internalizing more of the receptor. Okay. Can you hear me? Yes. Okay, great. Great questions. So, most cell surface proteins, and I don't think TROP2 is any different, when they internalize, will ultimately cycle around, and they get some reexpression on the cell surface. You also get more production of the cell surface protein. So some of it will be degraded, some of it will be recycled, and you'll get production of new protein on the cell surface. What we've seen in the in vitro data is that the marker that we're looking at is associated with cytotoxicity. You know, you've got greater internalization, you get greater cytotoxicity. So, what we don't see in the cell line data is a difference in sensitivity to the payload between, for example, squamous cell lines and non-squamous cell lines. So the sensitivity to the payload appears to be consistent. So it appears to be unassociated with, you know, this cytosolic expression, which is, you know, associated with that internalization feature. And then we also see that in vivo models. So again, you see it in cell lines, in vivo, and it's associated with what we see clinically. I think that answers your question. I don't know if Mauri Scaltriti from our Translational Medicine team wants to add any more commentary on that? Mauri? Hi, I'm Mauri Scaltriti, Head of Translational Medicine. So yeah, Susan, as Susan mentioned, as many other receptors is internalized and recycled, our cleavable linker, we know it gets cleaved, so we know that when we recycle it, so the payload stays into the cells and elicit the cytotoxicity activity, and now we are also developing measurement assays to measure this in a more quantifiable way, both in preclinical models and in the clinical samples. Thanks, Mauri. I think what's also just important as a take-home through all of this, while it's not exactly and directly related to QCS, is that there's quite a lot of differentiation within TROP2-targeted ADCs. And I think it's important to keep in mind that I think we can't just call this a class, even though they share the same target. I think that we know that the linker, the internalization, the toxicity, all of these things are very different as you look across the various ADCs that we've seen. Yeah, I think that makes the point. I think the question about whether or not the difference between histologies is something that was a random event in TL01. I don't think it is. And the reason why I don't think it is, is the pattern we've seen in the data, the TROPION-PanTumor01 study, in the TROPION-Lung01 study, in the China-based study that we published at ASCO, and in the ICARUS study that the Gustave Roussy Institute published at ASCO. So we've seen it in four different data sets, this difference between histologies. As I said, if you're just exposed to free payload, you wouldn't expect to see any difference between histologies because there isn't any difference in sensitivity to the actual payload that we can see across histologies. It's about the design of the ADC and this linker stability, which is a fundamental differentiating end piece. Thank you, Susan. Okay, question over here on the left. Thank you. Etzer Darout, BMO Capital Markets. So over the last couple of days, we've heard some commentary around the lack of correlation of activity, B7-H4 expression, of ADCs, with activity, and just was wondering around the overall strategy with QCS and how to prospectively look at some of these different targets. You mentioned B7-H4 looking QCS with that and just the overall strategy. Should we think about this as a fundamental, a different way to think about ADC development and maybe even radioconjugates, as you move forward, just given some of this disconnect we're seeing between expression and activity? Yes. The short answer to that is yes, I think this is a fundamental technology which is applicable to mechanisms of action that are targeting cell surface receptors, ADCs, radioconjugates, maybe T-cell engagers, maybe some of the cell therapies as well when you're targeting that. So that is something that we have, that, you know, invested in. And I think, for some targets, there is enough good correlation just with a simple IHC assay, and HER2 is one example. That being said, I do think we can improve on the precision and accuracy of HER2 definition even, and how low is too low across different tumor types may be different, and there's, there may be some help for that. But I think the second point is that you're very rapidly seeing with all of these agents coming through, that you're going to have more than one target within a particular indication, potentially with approved medicines. So in breast cancer, TROP2 and HER2, already a reality. In lung cancer, you see other targets coming. So I think you know, what we have planned for today and for the first-line setting for Dato is a singleplex computational pathology view. But we're already working on what I think the future is going to need to be, which is a multiplex approach and other things. And I think this is one of the reasons why this is going to be a differentiating advantage for the portfolio that we're bringing forwards, is because we have had the thinking about this, we've had the years of investment in understanding this technology, and we have the partnerships in place to help us deliver this, you know, at scale globally, through the work that Dave's team is gonna do. I think also on this, we certainly have seen with targeted therapies, if we just put a commercial lens on this, that with targeted therapies, while there's work and education in the infrastructure component, what we also are able to often do is have uptake happen more quickly, longer durations of therapy. In Europe, we're able to have typically more productive discussions around the value of the medicine because it's a defined population with a very, very, you know, more pronounced enriched benefit within it. So, certainly I think some ADCs will have the opportunity to be able to come to market without a biomarker. Gonzalo, if you can hear us? Yes. Yes. I can hear you. Go ahead. Can you hear me? Great. Yes. Hello, thank you for taking my questions. Gonzalo Artiach, Danske Bank. I have one on TROP2 biomarker testing. How relevant do you expect to be TROP2 testing in your Dato-DXd breast studies? Let's say in TROPION-Breast02. Will you be able to also test TROP2 in biopsies in this, in that study? I'm asking this because early data seems to suggest that triple-negative breast cancer patients could be more sensitive to TROP2 compared to other cancers. So would that mean biomarker testing is less important in order to predict efficacy in triple-negative breast cancer, for example, compared to, let's say, lung studies? Thank you. Susan, would you like to take that? Yeah, sure. Thanks for the question. So, we have looked at the QCS biomarker across more data sets than just TL01. And I do think this is something that replicates in terms of understanding the biology of approach. That being said, I think the prevalence of activity that you've got in a triple-negative breast cancer setting, and best illustrated, if you like, by the BEGONIA data that we've presented with, had an unprecedented 79% response rate and highly durable over the combination of Dato-DXd and Imfinzi. You know, it's not necessarily clear that you're gonna need an enrichment strategy in every setting, so there may be some settings where that isn't required. But absolutely, there were other settings that we've been looking at from the TROPION-P anTumor setting across a range of different tumor types, where I think to compete with the standards of care, particularly in the earlier line of treatment, I think a biomarker approach will be a big advantage and potentially differentiating and improve the probability of success. Thank you, Susan. Let's take another question online, and then I know we're at, just past 9:30 A.M., but we'll take a couple of more questions in the room, as well after this. So online, Mattias Häggblom from Handelsbanken, please, over to you. Thanks so much, Dave. So roughly a year ago, AstraZeneca was exploring Dato-DXd in seven phase III trials, and that was prior to the readout of the TL01 that has triggered so much debate in the investor community and continues to drive debate more or less with every update the company provides. Today, you submitted two of those trials to regulators for review, but still have nine trials in phase III, so a net addition of four. So can you perhaps help us understand what in particular has guided you and your partner's decision to initiate additional studies, despite what some would describe as complex outcome of the initial trial? Is it the evolving understanding of what an improved biomarker test provides? Is it investigative feedback, better understanding of the mechanism of action, or all of the above? Thanks so much. Thanks, Mattias. I appreciate that. I think that before offering to Susan to offer some comments on here, I mean, I think that Susan made some comments in the prepared remarks that are important ones, which is that if we take a look at the five billion-plus opportunity that we laid out at Investor Day associated with Dato, the majority of that is a frontline opportunity that we see to replace IO plus chemotherapy with IO plus ADC. And when we take a look at the NeoCOAST data, the TROPION-Lung02, the TROPION-Lung04 data, all of these are reasons to have building confidence in the evidence base that we've got to be able to move into the frontline setting. I think also with that, we've got progress that's very important progress made on the QCS, which allows us to add into AVANZAR. Now we have five studies that are in the frontline setting. We have studies that are in high PD-L1s, we have it in non-squamous, we've got the QCS, and of course, we've also added TL14, which is one of the phase III's that you've mentioned that we've added along with TL15, based on seeing really impressive results in the EGFR population and a huge amount of enthusiasm to combine Dato together with Tagrisso, based upon the FLORA2 data and what we've also seen in TL01. So Susan, is there anything more that you'd like to add to that? ... So I think, you know, again, the goal of the ADC portfolio in general, and Dato is a particular example, is that we can aim to replace some or all parts of chemotherapy regimens in different settings. So what gives us confidence about that is, you know, data, not just the TROPION-Lung01 data, where you're seeing activity. Those are, you know, you've pointed out some of the challenges that we have there, but also the combinability data that we've got with Dato and Imfinzi. I talked about the NeoCOAST-2 data. 34% pathological complete response rate, you know, which if you compare across with Imfinzi plus chemo, you know, clearly shows the potential for something that is greater than what's currently available in that setting, and actually with good tolerability and lower discontinuation rates. That profile is also seen in the TROPION-Lung02 data, in the TROPION-Lung04 data. We've got good combinability, as Dave said, with Tagrisso in a very important segment in non-small cell lung cancer, and we're seeing signals of activity in the TROPION-P anTumor data. You put that together with the biomarker data. I think there's great confidence that we can actually achieve the goal with this drug that we've laid out, and that really is what's supporting, you know, our confidence in moving forward into the phase III trials, which are coming. Thanks, Susan. Thanks, yes, please. Just to repeat what you said, but highlight it, Dave, because those two studies may not receive the attention they deserve. TROPION-Lung14 and TROPION-Lung15 were started because the activity in EGFR mutants is very attractive, and the combination is hyper attractive, so everybody's got very excited about MARIPOSA, which we haven't talked about tonight, and we could talk about it, but this combination of Tagrisso and Dato-DXd is by far much more exciting, quite frankly, both in the second line and the first line, and we have those two studies that are running, and I think we'll unlock a lot of potential and support Tagrisso in the years to come. I think, you know, just in summary on this, and then we can come to a question that's here within this. I think that, you know, Mattias, on this, there's certainly a, as Susan articulated, a more complex filing on TROPION-Lung01, but we see limited read-through from TROPION-Lung01 into the rest of the program. We've obviously applied the learnings from it, but I think that for all the reasons that we've stated, we see the program more broadly, moving along with confidence. Luisa, sorry, please. Thanks. Can you hear me okay? Yes. Not Luis. It's James Gordon from Jefferies. Yeah, a couple more questions just on, well, on Dato and AVANZAR, because I've still been getting questions on this. One of the questions I've had is that AVANZAR next year, so I think you're going to look at non-squamous, non-EGFR, biomarker positive patients. And I think last week you said the OS was similar in this population, but can you talk about how similar it was in the population, which we should extrapolate to AVANZAR? Was it a similar benefit if we look at the patients that you're going to be looking at in AVANZAR, is it similar to that, or does it actually get better when you use the biomarker? I guess that's the first question. The second one was just, does AVANZAR need to show stat sig OS, and is that tougher because it's a first-line trial, and often you get more crossing over when it's a first-line trial that dilutes some of the benefit? And then just the final question, I think, Susan, you said the FDA are looking at subgroups beyond squam status, and it sounds like the subgroups are not the biomarker. So what are the other subgroups that they might look at? In terms of the commentary, I think you're asking about the OS data with the QCS biomarker. Is that what you're asking about? Population that's analogous to AVANZAR. Right. People who are non-squamous, non-EGFR, biomarker positive, does the OS look good there? Okay. So, I think I said last week, and I'll say again, we've looked at the interim analysis of OS using the biomarker. We see similar trends to what we see with the PFS. So yes, you do see an improvement within the biomarker positive, as you know, as you expect. What we want to do is update that with the final OS analysis, and we will show that at the next available congress, likely to be early next year now. So I think that does build confidence in the ability to meet the goal of what we're achieving within the AVANZAR and gives us confidence in the first-line setting. Does AVANZAR need OS? Well, we have co-primaries for PFS and OS for AVANZAR, and again, what we've said is the expectation is that you have a clinically meaningful improvement in PFS, and certainly no detriment in OS is generally what will be looked for. So that's you know the assumptions that we have going there. And in terms of the QCS subgroup, again, I think the expectation is that you know we've set the cut point within the TROPION-Lung01 dataset, which we used to optimize that, and therefore, you need an additional dataset in order to have the full validation of that cut point, and that's why we're anticipating that in the first-line setting. So there are... You know, I've tried to sort of give you as much information as I can without giving a specific update. You're always going to look for the FDA is always going to look for subgroups. There's multiple examples of how they've done that before within the, you know, PROpel dataset, within the CAPI dataset, et cetera, so they will look at different groups, and they will look at the benefit-risk within those groups. All right. I think we have time for a couple more questions here. Okay, and there's one online. And also, if any of the other seven medicines that are $5 billion-plus by 2030 and be of interest, we are eager to speak to them. Louisa. Okay, thank you. Luisa Hector from Berenberg. Really impressive pipeline. So I'm just wondering if you could tell us, if we stick to lung and breast cancer, what proportion of lung and breast cancer patients are on AstraZeneca drugs today, and what would your ambition be by the end of the decade? Thanks, Louisa. I appreciate the question, and Sunil, I'll invite you also to offer commentary on that. Maybe I'll start the other way around. I think by 2030, if you take a look in lung cancer at the bispecifics, the ADCs, Tagrisso, and the combinations that we're looking to have across, you know, multiple bispecifics, multiple ADCs, and then also Tagrisso plus ADCs, I think that 100% of lung cancer patients are eligible. I mean, I think that then the question is, is how do we begin to drive near a hundred? I mean, so we may or may not have, you know, an ALK inhibitor that we're appropriate for this time, but I think you're talking about a very, very large number of eligible patients. And so then the real question becomes, what percent of the share of all of that are we able to gain with our medicines? I've said that we'd like to have one in two of all patients treated for lung cancer with a medicine being treated with an AstraZeneca medicine, and I, I think that that's credible based upon getting share in, in, you know, 80%-90% of the market. Sunil, do you want to talk about breast cancer? Yeah, yeah, no, I'm happy to address both of them. I think 2025 really marks a very pivotal period for us in AstraZeneca. If you think about four important studies, SERENA-6, DESTINY-Breast09, and TROPION-Breast02, they all address first-line treatment of breast cancer across all subtypes. And we have high confidence, of course, in all three studies, and if successful, that addresses every single breast cancer patient across the board. And SERENA-6, of course, is ESR1, SERENA-4 would be all comers. And then if AVANZAR comes in, as was noted, that again gives us an opportunity to be a leader in first-line non-small cell lung cancer. So I think between the programs that we have with Enhertu and its rapid expansion, with what we're bringing with camizestrant, and of course, what we talked about with Dato, we have an opportunity to really reshape and transform first-line treatment in both breast and lung cancer. Thanks, Sunil. Leora, anything you want to add on lung cancer and any other elements that you'd like to incorporate in? Sure. I'm Leora Horn. I'm the Global Clinical Strategy Head for Lung. I think it's also important to remember that in 2024, it's not you get one line of therapy and you're done. Patients go through multiple different lines of therapy, so there's actually an opportunity in the future that patients will see multiple different AstraZeneca medications as we're treating lung cancer as a chronic disease. I think it's a really important point with the combinations, with the sequences. I think also that there's, you know, a lot of interest in maintaining the CNS protection with Tagrisso through multiple lines, and so I think that it's not just simple arithmetic as we get through that. I think it's a very good point, Leora. I'm gonna go to one question online, and then we'll wrap up from there. So, Rajan Sharma with Goldman Sachs, please, question over to you. Hi, thanks for taking my question. Just, on NIAGARA, I just wanted to understand how confident you are in approvability of that trial, given the debate around AEGEAN and the need to show benefit of therapy in each setting of a perioperative trial. Thank you. Thanks, Rajan. I think that, you know, certainly while Professor Powles was here, he spoke to the clinical meaningfulness and relevance of this. Christian, would you speak to kind of our view on the regulatory lens that Rajan is asking about? Thanks for the question. It's in the mind of a lot after the ODAC that we had, that Leora here in the room chaired with AEGEAN. I think there is an important difference in the way we need to see NIAGARA and the situation in lung cancer. The first thing I wanna point out, the position of the agency, FDA specifically, they made very clear during the ODAC that they are willing, and they want to understand better the contribution of component, neoadjuvant and adjuvant, in the next studies that will be run in early settings, non-small cell lung cancer, and probably across indications. The ongoing studies actually very likely will be accepted if they have a positive benefit risk. This is exactly what happened with AEGEAN after ODAC. Actually, AEGEAN was not even a voting question at the ODAC. The second point is, Niagara is in a setting that is different than, for instance, AEGEAN, or the lung cancer studies. There is not in bladder cancer, a study that has been approved with a solely neoadjuvant regimen. So there is not yet evidence that the neoadjuvant only can improve EFS or OS. So this is an important aspect because actually what is approved in bladder cancer is nivolumab adjuvant. That is a regimen that of course, is given post-operatively. The third point, very important, is, Niagara show a statistically significant OS benefit. I don't think this is... This is incredibly relevant for the benefit-risk that the agency, every agency in the world will impact the positive benefit-risk. I think these three factors are quite in my view very, very solid to believe that the NIAGARA regulatory path should be easier than others. Thank you, Christian. With that, I just want to thank everybody for spending the evening with us here and for the great questions that we had. We look forward to seeing you certainly at Q3 or perhaps before then, and I hope you all have a safe trip back from Barcelona. Thank you very much.