Roche Holding AG (ROG) Earnings Call Transcript & Summary

So welcome to our Annual Pharma Day in London. I have to say it's a great pleasure to be here again on site and to meet all of you in person, and I'm looking forward to many lively debates and discussions, which we will have, and several opportunities during the event here with the investor and the analyst community. However, I would go for the agenda and some housekeeping items. However, before we start, I would like to say the following. We, at Roche would like to extend our greatest condolences to all the people in the U.K. for the loss of Queen Elizabeth, and we are with you in these very special days. Moving now to the first slide. Just taking you quickly through the agenda. Our event, as you can see here, will be kicked off by Bill Anderson, our CEO of the Pharmaceuticals division, who will provide his update on the ongoing business transformation, summarize past achievements, but most importantly, also outline our vision and show you -- to send to you the future building blocks, which we need for success as an R&D-driven company, which is driven by the credo, where science will take us. The second session then goes to Teresa Graham. Teresa is our Head of Global Product Strategy. Teresa will take us through the existing portfolio, summarizing the most recent event with regards to our key growth drivers; the most recent launches, which we are very proud of, first-in-class launches for VABYSMO in ophthalmology and Lunsumio in the hematology. And then she also will mention, I think, some of the commercial opportunities, which we might have ahead of us. We then have a combined 30-minute Q&A for the 2 morning sessions. This will be onstage with Bill and Teresa before we go into lunch at 12:00. At lunch, we will have all our key managers here -- our key senior managers present and also I think all our IR officers are around. So I'm looking forward to taking any of your questions. So yes. I think after lunch, we will then start our deep dive into the pipeline. I have to say we have a focus, as always, on the late-stage pipeline. However, we will also flag some earlier projects this time as we are developing certain R&D focus areas. Let me also make a more general comment here on the totality of our pipeline. I think, overall, our pipeline has hit a record high in terms of the depth and the breadth, which we are having now. We have like an industry-leading pipeline, just to point out, with 85 NMEs currently in clinical development, thereof 10 NMEs currently in late stage. And I think what is also interesting to see is the growing portfolio of platform technologies. And as you see, the latest additions to this platform -- to these technologies are either in-house developed technologies, but also, I think, technologies which we acquired for business development. So if we go to the first afternoon session, then this will be given by Levi Garraway, our Chief Medical Officer and Head of Global Product Development. Levi will take us through the oncology pipeline, providing updates of key late-stage products. Amongst these products, there are tiragolumab and giredestrant. We just presented the update on the ESR1 mutant data at ESMO 2 days ago. And he will also take us through some of the more recent additions to the pipeline, to the late-stage pipeline, such like our PI3 kinase inhibitor, inavolisib, or our KRAS inhibitor. The second session for the afternoon will then be held by Paulo Fontoura. Paulo is our -- over there. Paulo is our Head of Clinical Development for Neuroscience, Immunology, Ophthalmology, Infectious and Rare Diseases. And Paulo will provide us with updates on both the neuroscience and the immunology pipeline. There is, I think, topics in there like, for example, the 6-monthly subcutaneous formulation for OCREVUS, which is to come, and also an update on our gene therapy program in Duchenne. These are both projects where we expect the pivotal data actually to come in next year. And finally, for the third and last session, we will end on ophthalmology and providing you an in-depth look on our growing ophthalmology pipeline. This will handle to you over here by Nilesh Mehta. Nilesh is our Global Franchise Head, Ophthalmology. And we will also use this occasion here to -- he will give a presentation and recapitulate some of the designs of our pivotal VABYSMO studies. He will also, I think, point out some recent external real-world data, which are coming in on VABYSMO. And then, of course, he will also go into the pipeline some interesting assets we recently added, like, for example, our IL-6 -- anti-IL6 monoclonal antibody. So after the second round of presentations, we have then a second Q&A. This will be with all speakers onstage. And then, basically, we have the end of the event. But you are, of course, invited to stay longer. We have a buffet reception afterwards. So there's at least another 30 minutes where you can ask any questions, which have not been answered by then. I think at the buffet reception, everyone, again, will be available, I think, with the exception of Bill. I think he has some other appointments. And then also one final comment here from my side. Please let me point out that we have prepared also a short feedback survey. The link to the survey will be shown to the participants in the webinar 10 minutes before the end of the event. Participants in the room will receive a message in the app with a link to the survey. We would very much appreciate if you provide us your feedback as this will always -- we always strive for improvement, and this feedback will help us to -- for setting up future events. And with this, actually, I would like to hand over to you, Bill, and to provide us the update on the strategy.

Thanks, Bruno. Welcome, everyone. Really great to see you all in person. Many of you I haven't had a chance to see in person since 2019, and so it feels really good after so many Zoom calls and sort of such distance. Bruno, I appreciate your comments about the passing of the Queen. And I think, obviously, our condolences go out to all of our British colleagues. But I think it's fair to say that not only was the Queen a pillar of strength and resilience for the British people and for the Commonwealth, but I think she also stood as a real sign of integrity and character and peace in a world where many leaders are not following that model. And so I think for that reason, we all mourn her loss. Okay. Today, you're going to hear a lot of exciting details, a lot of cool science, a lot of, I think, really interesting things about what we're doing at Roche to prepare for the future and to deliver for patients. But in the midst of all of that, I want to make sure that we don't lose sight of the big picture. And so let me just propose 3 key takeaways that I hope you'll see over the course of the day. First is just the tremendous growth we've had in our portfolio, growth, and at the same time, diversification over the last years. We have been able to not only absorb CHF 13 billion in losses in biosimilars, but we've actually been able to grow over that same period 9% by adding CHF 15 billion from new products. In fact, that has basically made us the #1 company in the industry in terms of the value in our portfolio from new products. And I'll show you some evidence of that. We've also got major readouts. We've got momentum. We have major readouts coming for breakthroughs in major neurological diseases like Alzheimer's, like muscular dystrophy. In the next 5 quarters, we're going to see Phase III readouts in both those diseases as well as major potential breakthroughs in adjuvant cancer and in many other diseases. And so I think not only have we come through this era of biosimilars with strength in the top line, but we've got a portfolio that is really well positioned for both the short, mid and long term. Secondly, we've made a once-in-a-generation additional commitment to R&D during the same period. I think that's really remarkable. I don't think going back, say, 5 years, people thought, oh, we can weather the biosimilars. We could grow sales, and we can make a major additional investment in our commitment for the future for patients. And we really feel great about that, almost CHF 3 billion in new annual investment in R&D over that period. We're securing the future, not just for patients, but also for our shareholders. And then third, we've made a major transformation in how work gets done. And that relates to everything from what the individual jobs of employees look like. They've gotten much bigger, much broader and more interesting in fulfilling. That's led to much greater productivity. But we've also implemented digital tools and new techniques all the way from R&D through the whole value chain. And so we'll talk about that as well. I think the bottom line is we believe we're extremely well positioned for the future, better than ever, and we look forward to making that journey together and for having you all as part of that. So let me get into the content. I'll start with a bit of an overview at the high level. I mentioned this fact that we've been able to grow and diversify. And again, I was with the company in 2012 and I have to say I don't think we pictured that we could be this successful. We were committed and we had passion, but we've translated that into continued growth, and again, amazing revitalization of the portfolio. And I would point out that from 2012 to 2017, as a total Roche, we grew 17%. And then we had the biosimilar impact and we grew another 23% in the next 5 years. I feel really good about that. I think that's amazing delivery on our strategy. We set out an ambitious strategy. We've delivered on it. You can see we're much better positioned with respect to the breadth of the portfolio now. AHR went from being over 40% of our total and it's now less than 10%. And you can see really advances in every other area. At the same time, as I said, we managed to grow Pharma 18% in that first 5-year period and 9% in the next 5-year period despite absorbing CHF 13 billion in biosimilar losses. So that, I think, is the power of a strategy that you execute with diligence and persistence. And I think that -- hopefully, you know us for that, that we're looking far out ahead, we absorb the speed bumps along the way and we keep marching. This is a little more detail and what that's meant. And as I mentioned, we've plowed additional CHF 2.9 billion per year in R&D in pharma alone, and this is basically what that's paying for. So you can see on the left, our projects in Phase III and registration, which is kind of the best indicator of what's coming for the midterm sales dynamic. And you can see it goes up and down because -- mostly because projects graduate. Sometimes they fail, but they also graduate into the market and they come off the list. So in the last year, we've graduated 3 into the market. And as you can see, a nice upward trend over time. We've got more coming. If you look down along the bottom in terms of additional indications, we're at 51 right now that are in Phase III and registration, which is an all-time record for us. And then you can see also the number of NMEs in Phase I, Phase II, Phase III and across the whole portfolio. And so I think we're incredibly well positioned. Not all of those are going to be big breakthroughs, but some of those are going to be major breakthroughs and are going to be what's fueling the future of our financials and our ability to invest. Now some people are asking, okay, you've weathered the AH&R storm. What about other biosimilars and other losses of exclusivity? And so we put this together. We hoped it would be helpful. Let me explain this slide because it's a little bit complicated. So on the left, the bar on the far left, is a listing of all of our medicines that potentially could have biosimilars coming in the period between now and 2025. And you can see the total sales. And then this is basically the consensus numbers for 2025 for those products. And you can see that there's sort of a gap, a loss of sales potential in those products of about CHF 7.5 billion. Then the list kind of in the middle there is the list of basically all the in-market products and then some of the pipeline. And you can see that adds up to growth of about CHF 18.6 billion. So essentially, we have about CHF 18 billion or CHF 19 billion of sales projected to offset the CHF 7.5 billion gap, which I think we feel really good about. We have an opportunity even to beat that. But let me point out a couple of other things. So if you look, for example, at the pipeline, a total of CHF 4.2 billion from the consensus numbers, gantenerumab in the consensus is CHF 1.1 billion, okay? So we're really excited about gantenerumab and we're going to be talking about it today. But we are not dependent on gantenerumab. If gantenerumab goes away, then the model goes from CHF 18.6 billion to CHF 17.5 billion, all right? And I think that's important to note. We are not a one-trick pony. We have a rich pipeline, and we have lots of opportunities to benefit patients beyond gantenerumab. Likewise, you can see tiragolumab is in at CHF 0.3 million, so CHF 300 million out of CHF 18.6 billion. So obviously, those are nearer-term readouts. And so they gather -- they garner attention. But this is, I think, the long picture. There's also a lot of potential upside. If you look at the NMEs that really aren't playing any significant role in that CHF 18.6 billion, you can see a number of things, bispecific antibodies; our KRAS molecule, which has a potential to be best-in-class, and we'll be talking about all those in more detail. But I think at the big picture level, we believe there's a lot of room -- or a lot of reason for optimism for strong growth in the immediate years ahead. All right. Let's now back up a bit to the fundamentals, what is our big picture view of the world and how are we going at sort of capturing the opportunities we have to help patients. So I think most of you have been along with us on the journey enough to know, we set out a very bold vision several years ago. We said we're going to deliver twice as many medical advances to patients and -- which, by the way, that's saying something. That's not easy to do. R&D is not getting easier. Advances, fundamental breakthroughs are not easier to come by. But we're committed to doing this and doubling our rate of breakthroughs over the next decade compared to what we did in the past decade. But we said that's not good enough. We also need to deliver at a lower cost to society. We're targeting -- internally, we're targeting half the cost to society. And we're going to do that by having more breakthroughs, some of that can be through price, and we'll talk about our responsible approach to pricing. But a lot of that can be delivered by doing things like targeting diseases early, stopping diseases, curing diseases instead of, for example, requiring lifelong treatment. And so these are the opportunities that we have. We believe that Roche can make the biggest difference in health care. So kind of peeling that back a little and saying, what are the fundamentals? You all know about the big trends in health care. The needs are only getting greater. The fact that health care systems are under strain, budgets are under strain and so we believe there's a compelling need, both for better therapies, but also that economic bottom line. Patients are demanding more information. They're taking more action themselves. Telemedicine is a new reality. And that also means there's a lot of opportunities. And for example, we're using telemedicine to improve the possibilities for patients in the poorest countries in the world to get access to our medicines. Often in poor countries, the only barrier is not things like price. It's the lack of specialists who are qualified to administer the kinds of therapies that we're producing. And so we're using innovative partnerships between developed countries and emerging countries to create sort of virtual specialists in the poorest countries in the world. So this is a whole range of things we can do, ranging from the most developed countries to the poorest. And this is part of our access commitment. And again, I'll talk a little more about that. COVID and the pandemic have really just put all of this -- all of these underlying trends, they've sort of accelerated them and made them even more important. So as I mentioned, our vision of doubling medical advances, of providing more patient benefit, we believe, is exactly what the world needs and to do that at a reasonable cost. In order to achieve that, because when we set out on this vision 3 years ago, we sort of said, why should we believe we can do this? Why is this reasonable? It's not like we weren't trying to work hard in the last decade. So how are we going to double things? And we believe, fundamentally, that large organizations can be made much better. Many of you, and I've talked to many of you over the years about this and have asked you like how is it at your firm, right? You have bureaucracy. Is everyone able to punch at 100% of their weight? Or are people held back by the system, by the processes, by the hierarchy? And so we've made a huge commitment on this. And again, I think this is something that's characteristic of Roche. We are not a flavor-of-the-month company. You've been hearing me -- some of you I've been talking with about this for 6 years when I was CEO of Genentech, we began this journey of transformation. We called it a transformation because we knew we could get much better. We didn't know how to do it. And we said, we're going to have to start by understanding like what are the best companies in the world doing? We didn't find examples of that in life sciences. But we found a handful of leading companies in other industries who are trying really fundamentally new organizational approaches. And so we began that journey about 6 years ago. It has swept across all of Roche and it's making a major, major dividend. And we'll talk a little bit more about that. But I want to share some of the results of this. So for example, one of the first places we started was in our technical operations group. And we've now increased sales volume growth. If you look at that period from the first half of '22 to the first half -- or from 2016 to '22, so that's 6 years of transformation. We've grown sales volume 76%. And over the same period, head count is down 19%, all right? And if you do the math on that, that's a doubling in productivity per employee. And again, that's great for the bottom line. That's great for our ability to deliver for patients, but it's also great for the people because we believe if someone is going to invest their whole life at Roche, we deserve -- or they deserve to make that a good investment of their time and they're not wasting time in a bureaucracy, but they're able to deliver the full impact of their talents. 200% increase in productivity in PT. At Genentech, in the U.S., sales growth of 25%, head count lower by 22%. That's a 60% increase in sales per person. Again, this is about helping people invest their lives in things that matter. Pharma International, a 50% increase in sales per employee. In pharma in China, over 200% increase in sales per employee. And in PD, this is one of the areas that's the hardest to tackle because drug development doesn't lend itself to simple solutions when you're developing medicines in 30 or 40 different diseases with all different kinds of requirements. But we're chipping away at it. So far, we've made about a 9% increase in productivity. And I think you should expect more to come there because we're really excited about our opportunity to deliver higher productivity also in product development. What's that meant in terms of bottom line and our ability to invest? And so this is a comparison. I use 2017 as the base year because that was the last year before the major biosimilar impact on AH&R. In fact, that was peak sales year for AH&R at about CHF 20 billion. And basically, what you can see is we've managed to weather that 13 billion in CHF sales of AH&R, CHF 13 billion loss. We've been able to grow sales 9%. But at the same time, we've been able to increase R&D by 32%. And again, this has been a major, major lift for all the leaders and all the people at Roche. But this was our commitment. And we told our employees then, we said, look, this is going to be hard. We're going to be changing everything. We're going to change the way we do business. You're going to have a different relationship with your manager. You're going to see your manager a lot less. We have a lot fewer managers, okay? We hire the best and brightest people in the world and they don't need a lot of management. You need to have clear frameworks, people need to understand the vision and then you need to get out of the way and let them go. And so we've been able to increase that R&D investment in that major substantial way while absorbing the biosimilar loss. And I think we're really proud. This is just numbers. But what that took to deliver, we're really proud of not only what we've learned along the way, but how that positions us for the future. And we're not done with this. I mentioned digital. It's also becoming an increasingly important contributor to how we do things. And I won't go into this in detail right now, but I will just say this is happening in research. Major, major changes in the way we develop, invent new drugs, new pathways. It's happening in clinical development. We're working on virtual clinical trials, a much more streamlined data flows from investigation sites into the databases at Roche. It's having a big impact in the production and supply chain as well as our medicines, and enabling us to do things like truly personalized therapies. Like our cancer vaccines that's our partnership with BioNTech, every vaccine for every patient is different, but it's not autologous. It's not -- in other words, we're not using patient tissue to create a drug, but we are using patient tissue to determine the composition of each vial, which is really extraordinary. And then in the future, we're also looking at things like digitally existed therapeutics. So I think this is a big part of what we're going to be in the future, and it's starting to deliver already. Customer engagement. I mean we have totally flipped this on its head starting in Genentech where we made a major change. We said, hey, the old model -- the old push model where pharma has just got multiple sales forces in doctors' offices and in major institutions like Memorial Sloan or MD Anderson, where there might be, literally over 100 employees from the company, we've taken that way down, reduced that by a factor of 3 to 5 and really put a focus on what does the customer need, not what does the company want to convey. And that's also paying off in big ways. And I think you can see that in the success of our launch products over that period. We're also spending a big investment to rejuvenate our manufacturing sites. This is a global network that's very resilient. It steaded us incredibly well through the pandemic, where we were able to deliver millions of doses of ACTEMRA, of Ronapreve at the same time that we basically kept all our core products online. I mean we did not have any major stock-outs in the whole portfolio other than being short on COVID medicines at the worst of the Delta surge. And so this is something that we believe is a vital part of who we are. We have that capacity that allowed us to do those partnerships and allowed us to step up that COVID production, and we want to now invest for the future to make sure we're ready in -- for whatever comes. I mentioned our funding for R&D, let's talk a little bit about what that looks like. So for example, we have -- as you know, we have a lot of these innovation engines. We think that's really important. A company our size, it can never be monolithic. And so the trick is how do you have a bunch of independent R&D sites, but how do you capture kind of the best of everything. And again, this is a major change. Because if you go back 5 years ago, we had a number of innovation engines, but those innovation engines had, I would say, pretty big walls between them. And they didn't collaborate, they didn't talk. That wasn't even necessarily seen as a virtue, collaboration. We have totally changed that. We brought in new leaders that are both leading scientists in the world, but also highly committed to collaboration. And what we're doing now, and you can see this in things like our big investment in computational biology and analytics; at gRED, which will be used to benefit all the other innovation centers; likewise, in pRED, we have a major investment in human tissue modeling. So this is like organoids and doing preclinical work in organoids instead of animal models so that we have basically lower cost, faster development and also hopefully more accurate. And so that investment will benefit not just pRED, but will be used in gRED, in Chugai, at Spark. And so I think this is a key part of why we believe we can be successful and even more successful in the future. This is also leading to an incredible diversity of approaches. I think we probably have the most platforms of any company. And you can see now probably some of the latest additions I would point to. If you see sort of in the middle on the top, the cyclic peptides. So this is the Chugai center of excellence in cyclic polypeptides. This is a class of drugs that was basically -- well, they're not even -- I can't even call it a class of drugs because they don't exist. This was something that a lot of scientists and chemists hoped would be an important contributor to medicines over the last, say, 50 years, but it was never attained. The potential was never realized because we couldn't make drugs. We couldn't make molecules that hit the targets and had drug-like properties. And that's all about to change, I believe. Chugai, and this is kind of characteristic maybe of Chugai in particular and the Japanese in general. They have an incredible stick-to-it-ness. They're not willing to give up and they've been working on this for well over a decade, and now they've got a whole line of molecules for a number of different disease targets. But the leading one is a pan-KRAS inhibitor and it is now showing drug-like properties, and we're super excited about those. So for the science geeks in the audience, I hope you can appreciate how important this is because to have a cyclic polypeptide that's highly bioavailable and is hitting the target is a really exciting development, and there's many more of these to come. Another one I'd point out in the bottom right, and Levi will say more about this -- or actually, where is it, no, top right. So allogeneic CAR T cells. Many of you know we sort of stayed out of CAR-T for the most part up to this point. We got a lot of questions about this, are you missing out? And we always said, no, we think it's really important to be able to put the medicine in a vial and ship it as opposed to bringing patient tissue in and processing it and hopefully getting it back in time for a patient to actually receive the benefit. We believe now is the time to invest because of the maturing technology around allogeneic CAR-T, which basically doesn't rely on patient tissue. And so we've made a significant investment just this year in a company called Poseida, and look forward to hopefully bringing some promising therapies there. All right. This brings us sort of our -- to our partnering approach. I think you all know that Roche, we -- I'd like to say we have partnering as a key part of our strategy, but we're not partnering dependent, the way some other companies are. We have a massive internal innovation, but we know that 95% of the innovation in life sciences is not happening inside Roche, and we need to tap into that. And so this is sort of a list of some of the more important deals that we've made over the previous years. What I want to highlight is this is -- sort of on the right, you can see this is kind of the evolution. We had a flurry of deal-making activity in 2019 and 2020. This was, in large part, because of our big infusion of new R&D funding. We said, again, we're making a generational investment, a once-in-a-generation move to significantly step up the early research and development. Those investments in 2019 and 2020 will be paying off in '23 and '25 and '27 through the course of the decade. Last year, we exercised, I think, extreme discipline. And again, I hope you know that of Roche is we are not going to chase after whatever the industry is doing. We're patient. We use discipline. Last year, the prices -- the asset prices were crazy. We would go into a deal thinking we'd be willing to pay $100 million and the biotech company would say it's worth $1 billion. And we're like, this isn't going to happen. We did 2 significant deals last year. We did 9 the year before and 8 the year before that, okay? That's discipline. This year, year-to-date, 5, okay? So we're pretty much back on track towards our historical average because asset prices -- I would put it this way: it's not a bargain. This is not a bargain hunter time, but we actually have the ability to find reasonably priced deals, partnerships that make sense and we're doing that. On the right, I think it's a very important chart, and I'll explain why. So it's probably no surprise that you'd see we're the leading investor in R&D and that's borne out on the chart on the right. On the left -- sorry, in the gray bars is the M&A activity, okay? And what you can see -- and this is over 5 years, so it sort of averaged out. And you can see -- I think maybe a way to think about this is the difference between paying it upfront versus paying in arrears, okay? Because when you do M&A, it's not hitting your P&L immediately. It's hitting your P&L over time in the form of amortization. That amortization takes up space on the P&L that is not available for future investment, okay? So we -- as you can see, we have really the smallest proportion of M&A spend of any of the major pharmas. And this means we have the lowest amortization. So when we're spending -- if you look at that total, that's all paying it forward. We're not paying for stuff that's already in the market. We're paying for the stuff that's coming in the future. For long-time investors, I think this is a really important fact. And again, we're not always going to hit on every medicine. But over time, with our discipline, we're going to hit on a lot of big medicines. And I think you'll see that we can -- because we've paid it forward we can then invest for the next generation. We're not paying for the stuff that's bringing in revenues today. I mentioned the major step-up. I think this is an important chart. If you look -- before we started this, in 2017, we were spending 53% of our operating expenses were on R&D and 47% on everything else, okay? Already last year, we were at 60-40 and we're continuing that trend, all right? And we're leading the industry with this. And you could say, okay, great. That's great, Bill. You're spending lots of money, but what's that getting you? And I think over time, again, we've shown that we will deliver innovation. And the chart on the right, I think, is a really interesting one because this one is showing in the blue, what is the NPV of recently approved NMEs. So NMEs approved in the previous 5 years, what is the NPV of those? And you can see we have by far the biggest number on that and that's out of a share of total enterprise NPV. So I think -- well, hopefully, that's useful to you all in terms of the way you look at Roche, but maybe the way you look at some of the other companies and how we stack up in terms of strategically and over time. Let's talk briefly about some of those new focus areas in R&D, and you're going to hear a lot more about these. So I won't take any length of time. Hematology. This is an area of great strength and depth for Roche. And we believe there continue to be some very important opportunities to benefit patients. You can see we have a really broad pipeline. I mean we have in-market molecules, small molecules, antibodies, ADCs like POLIVY, bispecifics like Lunsumio. And as I mentioned, we've now entered the CAR-T space with an allogeneic CAR-T partnership. And again, Levi will talk more about that. Ophthalmology. This is an area where we think there is an opportunity to really change the way patients are treated. Right now, you basically have a few medicines. They're all targeting the same diseases and they have very similar MoAs. Now with VABYSMO, we brought a new MoA to bear. We have excellent data. And as you'll see, as we continue to roll out data over the quarters, VABYSMO is being adopted very rapidly, not just in America, but in many countries in the world. In Japan, in the U.K. and the rest of Europe rolling out very soon. We have an opportunity to change the paradigm to bringing in biomarkers, which we're very active in. And these can be biomarkers that are based on tissue and aqueous samples, but also based on basically optical biomarkers that can advance the treatment choices both in terms of the timing, choice of treatments and frequency. Moving on to Alzheimer's. We're very pleased to have the opportunity to get the data. This is something we've been waiting for. I've been involved in our Alzheimer's programs extensively since 2012, so 10 years. And believe me, we've got a lot of heart and soul tied up in gantenerumab not because we're financially dependent on it, but because the same reason many of you care. Personally, I've lost an aunt to Alzheimer's disease, my mother-in-law and it's time we did something for our parents, our grandparents. And so we've got an opportunity with gantenerumab. We're following that up with Brain Shuttle gantenerumab. So for those of you who aren't aware, Brain Shuttle technology, which has been pioneered in pRED, is something to assist getting -- basically getting the active part of the protein across the blood-brain barrier so that we don't have to dose so much systemically in order to get the necessary concentrations in the brain. We have promising Phase I PK data with our gantenerumab Brain Shuttle, and we're looking forward to bringing that also to other therapy areas like MS. We have other MoAs, for example, anti-tau antibodies, we'll have to see. But we're still hopeful that there could be an impact for tau. And then we're also pursuing with the Diagnostics Division both blood-based and CSF-based diagnostics that could be companion diagnostics or at least lead us to know who is likely to benefit from treatment and who can be screened out. Sustainability. I know many of you are particularly interested in this. This is something that's in our DNA at Roche. It dates back to the founding family's involvement with entities like the World Wildlife Fund. And we've been, as you know, #1 or #2 in the Dow Jones Sustainability Index for many years. But this is something that I'm just going to highlight 1 angle on this. The biggest thing we can do for sustainability, by the way, is deliver that vision of better medicines, earlier treatments and less cost to society. I hope we could all see that, that is the most important contribution Roche can make to sustainability. But there's also how we do it. And so for example, in pricing, I think we have a really good balance of price for funding tomorrow's innovation, but price that ensures patient access. Recent examples of that, I think, in the U.S., these are our average discounts on some of our most important products we've launched in the U.S. compared to the standard of care, usually a standard of care that we've beaten. And you can see average discounts between 20% and 50% on major products. Also price increases, which are important. And we've been at or below inflation in almost every year. And so again, I think it's just part of who we are. We believe we want to deliver great medicines, but we need to do that at reasonable prices that allow for broad access. All right. Let me finally turn to the kind of the near-term outlook in terms of readouts. There's a lot on here, so I'm going to be brief. But the -- some of the things that are remaining this year, 3 adjuvant or neoadjuvant readouts for TECENTRIQ, important to watch. We've got an opportunity, obviously, gantenerumab, to change the world in Alzheimer's. And I'll remind you, not only is gantenerumab potentially a breakthrough in Alzheimer's, but it's a subcu injection, which can be done in the home. That's a big deal. I mean monthly IV infusions for elderly people who may have issues getting into a clinic to get an infusion, I think that's a very important thing. Three major readouts in ophthalmology with VABYSMO and SUSVIMO. So continuing to bring it. And we're -- and that's not the end of the sight on ophthalmology. In 2023, a couple of important tiragolumab readouts, including SKY-01, which is the -- we'll have an additional interim and final readout on OS for SKY-01. TECENTRIQ and ALECENSA, more adjuvant readouts next year, including the ALECENSA readout in adjuvant ALK-positive lung cancer. ALECENSA is the leading therapy for ALK-positive patients and now we'll have the first data in a major adjuvant study in that setting. Moving down, you can see Venclexta plus azacitidine in first-line high risk. This is a myelodysplastic syndrome, which is an important indication and another opportunity for Venclexta to expand its impact. Crova, this is our medicine targeting PNH. It's a complement inhibitor. And these are the first global studies. We had a positive study with crova in China this year, readout. Now we'll have the Phase IIIs and the global studies next year. Two bispecific antibody readouts in second line and later DLBCL. This is a big deal, a sizable indication. We have 2 shots at it. And it may be that we have one is a clear winner, and it may be that we have basically differentiated drugs based on efficacy and safety. So we're looking forward to that. And we believe we'll be the first with these indications. There's competition here. We think we have a good shot at being first. I mentioned muscular dystrophy with our partnership with Sarepta, a big readout late next year will be the full Phase III data, highly anticipated. This could be the biggest breakthrough ever in muscular dystrophy. We have rights to outside the U.S. And I think this is -- if this works, this will probably be the largest gene therapy maybe for a long time. And so we're really excited about that one. OCREVUS, which is our MS leading therapy. We have our first subcu Phase III data next year. That will be the data we can file. And so this will be the opportunity for patients to have a twice-a-year subcu OCREVUS. So I think you can all understand how important that is. And finally, I would highlight TNKase, which is one of our older medicines, we have a study to extend the window. Right now, patients with stroke can only get TNKase if they get it in the first 4 hours. So this would take it from 4.5 to 24 hours, which is more than a doubling of the size of that opportunity. And so really exciting pipeline in the near term. Teresa will talk more about some of the readouts that are beyond that, but I wanted to tee that up. So in summary, as I mentioned, we continued to drive growth and diversification. We've got major, major readouts in the near term. Second, we've made a once-in-a-generation step-up in R&D and that is fueling a high level of innovation across our early R&D engines, which is going to fill the pipeline for the second half of this decade. And finally, we've had a major transformation in how work gets done that's leading to high fulfillment for our people and an acceleration of progress. And so I hope you agree that we're very well positioned for the future, and we look forward to sharing more of that. And now I want to welcome my distinguished colleague, Teresa Graham, who's our Chief Medical Officer and Head of -- our Chief Marketing Officer.

I just got a really big promotion.

And Head of GPS. Teresa?

Thanks, Bill. Thank you. And you took a lot of my talking points, so maybe we can make up a little bit of time here. So as Bill mentioned, my name is Teresa Graham, and I'm the Head of Global Product Strategy. And it's actually really exciting to be back in London in person with all of you and to see so many familiar faces actually 3 dimensionally. I didn't know some of you had legs. So that's actually really nice. So I'm going to start by just introducing a little bit about where our portfolio is at the moment. So in 2022, we had the opportunity to launch 2 new molecular entities, both VABYSMO and Lunsumio. This is important not only because of the benefit that these 2 drugs actually provide to the patients that they treat, but also because that's our second and third first-in-class bispecific molecules following HEMLIBRA. And soon, we'll be launching our fourth with the filing of glofitamab. This really entrenches us as leaders in the bispecific space and continues to give us more information about how this modality can help patients in additional indications. Bill talked about the diversification of our portfolio, but I wanted to take a little bit of a step back here and just really tell you what that means for the near and midterm outlook for Roche. So you can see by half year 2022 where our absolute growth was coming from. And this is actually really important. Our growth is coming from HEMLIBRA, OCREVUS, Evrysdi, PHESGO, TECENTRIQ. These are drugs that we have approved that are in the market today that we are currently commercializing, and really execution and future growth is in our hands. That's where you want to be. You also note that we are increasingly less dependent on Herceptin, MabThera and Avastin, as Bill pointed out. I can confirm guidance, we're looking at about an additional CHF 2.5 billion in erosion this year. But for the most part, we've now got the AH&R biosimilar cliff behind us and we're really looking forward into a new era with a much more diversified portfolio. In 2022, we haven't hit an all-time high of 15 blockbusters annualized revenue in the portfolio with an additional 2 coming in both PHESGO and VABYSMO, which is just sort of on the uptick here. If you look at the right hand of the slide with the half year 2022 sales split, again, this is the diversity that we've been talking about coming for quite some time. Oncology is now 48% of our overall sales, followed by neuroscience at 17% and immunology at 16%, and we will continue to see that diversification happening as all of the new molecules that we'll talk to you about start coming online. So just a really exciting place to be. So now let's actually look at some of our individual treatments. So I'm going to start, not surprisingly with oncology. It is Roche's goal in oncology to be the leaders in bringing a cure to cancer. Our goal is to take groundbreaking science and to turn it into medicines that can actually materially advance patient care. We currently have 20 marketed products in oncology and that depends how you define marketed. But we've got a significant number of oncology products that are currently marketed. Last year alone, we helped 1.25 million patients around the world, which is just staggering. We have 150 clinical trials currently ongoing in oncology, 60 of them are in the late-stage and 24 of them are in early-stage cancers, which is important because that's where we have the best opportunity for cure. So we have a tremendous investment here. And I think as Bill mentioned earlier, one of the things that's so exciting about that is it means we're not overly dependent on any given molecule or any given modality. We really have a lot of opportunity across all of oncology. So let's get started with TECENTRIQ because I think TECENTRIQ actually shows you that strategy in action. So you see on the graph on the left-hand side of the slide that we continue to grow very favorably with TECENTRIQ, double-digit growth, 13% growth so far this year. Our sales breakdown is very much as you would expect. We have our biggest sales where we have differentiated data. The expansion of small cell continues as does HCC as we continue to gain approvals and reimbursement in both the EU and Japan. Much of our near-term growth is driven by adjuvant non-small, we'll actually talk a little bit more about that in a minute. Bill did touch on the positive readout for TECENTRIQ subcut. This is very important for the molecule overall, reducing that treatment time from 30 to 60 minutes down to 3 to 8 minutes. So think about what that means for the office, think about what that means for the patient, particularly as you start moving into earlier lines of therapy, that becomes really, really important. So now let's spend a little bit more time on TECENTRIQ in adjuvant. We were obviously the first-in-class treatment in adjuvant non-small cell, and that launch is going very strong. The U.S., which was our first launch market, where we have the most data, has about 60% market share to date and that is growing at 20% quarter-over-quarter in the last 2 quarters. Testing rates are exceeding our expectations, up from 40% to greater than 80% since TECENTRIQ was approved, which means a lot more patients are getting the opportunity to get into that funnel. And we're approved in more than 55 countries with EU approval achieved June 1, and that means that we're going to start seeing reimbursement come online, and we'll start seeing adjuvant grow even more significantly. But where we're excited about taking TECENTRIQ next is actually into this periadjuvant setting. So what is periadjuvant treatment? So we're all familiar with neoadjuvant treatment, treatment before surgery. And we're all familiar with adjuvant treatment, which is treatment after surgery. That's where IMpower010 sits. Periadjuvant treatment is an area of great interest for the physicians right now where you actually treat both before and after surgery. It's 4 courses before surgery and 16 courses after surgery. And the thought is that will actually allow you to get the best possible outcome for patients. That data, IMpower030 is expected to read out later this year, and it gives us the opportunity to be first-in-class in periadjuvant, which would be a really exciting step forward for TECENTRIQ. And again, this is an area where the subcutaneous formulation would be really meaningful. Now one of the reasons that we're so excited about our investments in early cancer is it actually gives you the best opportunity to have a great outcome for patients. That is similarly true when you talk about targeted therapies. And we've been very intentional about building a leading portfolio in multiple oncogenic drivers within oncology. You can see in non-small cell lung cancer we currently have 8 of the most common modalities represented in our portfolio. Also within targeted therapies, we continue to move into early disease. Bill mentioned the ALINA trial for ALECENSA in adjuvant ALK-positive non-small cell lung cancer. ALECENSA is the standard of care and in ALK-positive treatment today. This would be an important new extension for patients. We continue to think about how all of these molecules could be further enhanced in a pan-tumor environment. So we have the TAPISTRY basket trial, which is currently ongoing, and then our investment in developing the most common mutations. We've talked about KRAS, Levi will talk more about that this afternoon, as well as the PI3 kinase mutated. The PI3 kinase, which is in 17% cancer patients, which are the 2 considerably larger opportunities. I think many of you are aware that commercializing in some of these smaller opportunities can be more difficult. But we're very encouraged that NGS testing rates are actually dramatically increasing globally. And as new kinds of testing starts to come online, hopefully we'll be able to identify these patients with increasing frequency and allowing them to get on therapies which potentially allow them to have the greatest outcome for their treatment. The FMI liquid biopsy was approved. This again is groundbreaking. It's about 30% of patients have insufficient tissue for testing. So it means you would never even know if you have a tumor that could be treated with one of these medicines. The BFAST trial utilizes that liquid biopsy that's underway. And again, we've talked about the fact that we see NGS testing rates increasing dramatically, which is a really positive sign. Now let's talk a little bit about the tiragolumab program. So we've talked a lot about tiragolumab over the course of the last couple of years. And so I'm only going to touch on it relatively lightly today. Because the TIGIT space is so competitive, what I'm about to tell you is pretty much what you're going to hear from us today. So you can send Bruno questions, but chances are we're probably not going to answer them because this is a highly competitive space. So as you know, we made a significant investment in our TIGIT program with one of the most comprehensive clinical development programs out there. And we did this very intentionally taking some smart risks. We wanted to try and get what we thought -- what we believe is a very promising treatment algorithm or treatment -- set of treatments into the patients that had the highest unmet need as quickly as possible. When we saw the results of SKY-01 and with that very early readout, unfortunately, we didn't meet the co-primary PFS endpoint. And the co-primary -- I'm sorry, we did not meet the co-primary PFS endpoint, but the co-primary endpoint of OS was still immature at the time of that reading. But we did see numerical improvements in both and that study does continue until the planned analysis. The regimen is well tolerated, just as we had expected, and no new safety signals were identified. What we saw in SKY-01 makes us continue to believe that the science here is quite promising, and you'll see that we expect to see results in 2023. I want to give you a couple of other updates on the TIGIT program. In SKYSCRAPER-06, we had a planned futility analysis that has happened, and we are now expanding that into a Phase III trial. It is important to note that we remain blinded to that data. We do not have any additional information. In first-line esophageal, this is SKYSCRAPER-08. This is our China-only study. We had a planned interim analysis. That trial will now continue until its final results in 2023. We are also planning to present the data in cervical cancer in half 1 of next year. So lots of activity in the TIGIT space, and we look forward to continue to bringing new readouts as this data continue to mature. So now let's switch over to HER2-positive breast cancer. We see continued growth in our HER2 franchise, specifically in a couple of really key areas. So the global PHESGO launch continues to go exceptionally well. So you see here a 27% conversion rate in launch countries. If you actually take the U.S. out of that number, and for lots of reasons, we all know the U.S. has a slight bias towards IV treatment, that number for global conversion actually goes up to 40%. So you can see that PHESGO is actually gaining quite a bit of traction. Why would that be? It cuts health care costs significantly. Up to 80% of nondrug-related costs are actually addressed with PHESGO conversion. But we also importantly see 85% of patients prefer PHESGO treatment over their prior administration. This is important, again, when you think about continuing to expand more and more into early breast cancer, where that patient experience for patients who are going to be on treatment for a long time, something like PHESGO makes a really big difference for them. We continue to see growth in KADCYLA in the adjuvant setting despite competition. We expect KADCYLA to continue to grow in the low single digits. The majority of that growth and the majority of sales is coming in the adjuvant setting with the expansion of KATHERINE's launch in Europe and other key markets around the world. And we have additional Phase III trials ongoing with both KATE-3 and ASTEFANIA. Moving over into the hormone receptor-positive breast population. I'm not actually going to spend a lot of time on this because Levi is going to do a significant deep dive on both of these programs, giradestrant and our PI3-kinase program. What I will tell you from a commercial perspective, these are really compelling drugs. We have the opportunity to reach a large number of patients. It's 2 best-in-class -- with 2 best-in-class therapies. They have broad development programs and that high potency, that well-tolerated profile and the fact that they are combinable means that we have the opportunity to help a lot of people in some very significant areas of unmet need. So as I said, Levi's going to talk a lot more about both of these going forward, but I think these are 2 areas to watch, particularly when you see how many patients could potentially be impacted by either one of these therapies. So from there, I'm going to switch us over to hematology. Hematology, again, is an area where Roche has a long legacy of leadership. And we are sort of taking 2 approaches here. We both want to build in those areas where we've historically been strong, such as DLBCL, as we expand into areas of significant unmet need like multiple myeloma, which Bill mentioned earlier. So let's start off with POLIVY. POLIVY is obviously the first new treatment in more than 20 years in first-line DLBCL. And we've talked about this a lot, but I think it's important to underscore the unmet need that remains in this market. As effective as R-CHOP is, about 40% of patients still do not respond and those patients that do not respond, that relapse actually have worse outcomes over time. This is a significant market opportunity in first line since we have about 3x more drug-treated patients than we do in the later lines of therapy. And importantly, there's no new competition expected in this area, no new options for patients for about the next 3.5 years. So POLIVY is currently approved in the EU, U.K. and Japan. It's filed now in the U.S. and China. And we've seen strong uptake in those markets that have already launched. Germany, in particular, had it entered into its guidelines and we saw a doubling of sales within a month. I think what's important, as people spend more time with the POLIVY data and really begin to understand what it can mean for their patients, we're seeing more and more physicians become enthusiastic about the opportunity of having POLIVY in their armamentarium and having it as an option for their patients. We continue to believe that this is very compelling data and data that should be changing the standard of care in first-line DLBCL. Moving on to Lunsumio and glofit. So Bill talked about both of these as well. I think what is important to note here is, again, we have 2 best-in-class molecules in Lunsumio and glofitamab. And what's important is our ability to actually tailor who gets what treatment depending on the setting, whether it's the disease that the patient has or the setting in which they're being treated. Lunsumio, as we know, is sort of ideally suited to the community setting in those places where it is currently launched, that is exactly where we see it being used. Glofitamab has that best-in-class efficacy potential with very high CR rates, very durable responses. And particularly in first-line DLBCL, we're seeing very minimal CSR (sic) [ CRS ] observed. So it also may be eligible for treatment in the outpatient setting. Certainly, we have more to discover with these 2 drugs. We'll talk about that in a moment. But I think what's also very unique about both of these 2 is the fixed duration and the very easy-to-understand dosing, which, for physicians, we've heard is extremely important. Again, I'm not going to spend a ton of time on this slide because Levi will cover all of this in his section as well. But again, both of these have the ability to be first-in-class and best-in-class. You can see that we have significant development programs against all of them, both in monotherapy and in combination. And what you can see is that while the first indications that we're going after in third-line DLBCL and third-line follicular represent relatively small patient populations. Those populations get significantly bigger over time, particularly as we get into first-line DLBCL. So 2 molecules again to watch. We think that these could have transformative potential for patients. We don't tend to talk a lot about Venclexta, but we should. It's a really, really remarkable molecule. It is revolutionizing treatment in CLL and AML. We are soon to see the long-awaited CANOVA results, which builds off of the Phase II that we saw in BELLINI on patients with the t11:14 translocation. This is a relatively small patient population. But depending on what data we see with this trial, it could actually open up a lot of opportunity in multiple myeloma. And then Bill also mentioned the first-line trial of the VERONA study in MDS, which is expected to read out next year. So again, I think a lot of really interesting data coming on Venclexta and another one to watch. And then, finally, no conversation about hematology would be complete without talking about HEMLIBRA. HEMLIBRA is, head and shoulders, the #1 prophylaxis treatment for hemophilia A in the U.S. and the EU. You can see that we continue strong double-digit growth in this patient population with 35% share in the U.S. and the EU, 65% of those patients are on greater-than-2-week dosing. We expect further penetration, both by additional countries coming online in our inhibitor and non-inhibitor indications, but also as new data continues to come -- become available, including things like the Phase III HAVEN 7 in infants in 0 to 1 year old. The growth in the rest of the world is really driven by reimbursement and approvals now starting to come online. You can see that the non-inhibitor indication is approved in 98 countries, but only reimbursed in 52. And so that's a tremendous amount of additional upside for HEMLIBRA. Why do we remain still confident in HEMLIBRA? We remain still confident because HEMLIBRA has really redefined the bar for treatment in hemophilia A. We have infrequent subcu injections. Up to 90% of patients are bleed-free at a year. We have an incredibly positive experience with patients on HEMLIBRA. It is not to be underestimated the importance of the safety profile and having a well-characterized safety profile in this patient population. Over 18,000 patients are currently treated with 7 years of safety data that is absolutely fantastic. So we think that the bar that HEMLIBRA has set for patients is just incredibly significant. That having been said, we always like to impress ourselves and step over our own bars, so we continue to think about how we might be able to make a difference for hemophilia patients going forward. We have the gene therapy for Spark and the next-generation bispecific that's coming from Chugai. And our hope is that this may be able to further reduce or eliminate the frequency of administration while maintaining that very high efficacy and safety bar, providing significant ability to improve the quality of life of these patients. So just another drug to watch because while we've had great growth, there's still a lot of growth to come. Now ophthalmology I'm sure is on lots of people's minds. Bill, you mentioned a line of sight earlier, no pun intended. So I'm going to give you a little bit of insight into what we're seeing with ophthalmology at this point. I'm sure that you guys have lots of questions. We do have Nilesh Mehta coming a little bit later this afternoon. So I'm just going to quickly ground us in what the VABYSMO is and what we're seeing in terms of the early launch performance. And then we'll have the opportunity to take a deeper dive into VABYSMO later. So as you know, VABYSMO is the first bispecific that has been approved in ophthalmology. It's simultaneously biased to the Ang-2 and VEGF arms, which potentially allows us to improve vascular stability and retinal inflammation. That is, in fact, what you see when you look at the outcomes from our clinical trials. The updated 2-year data continues to underscore the durability that we see with VABYSMO, and you see from the left-hand part of the screen that, that need for longer-lasting therapies is where the highest level of unmet need is in the ophthalmology community. And we continue to see real-world data that reinforces our belief in VABYSMO. Nilesh is going to talk a little bit later about the TRUCKEE study, which is currently ongoing. We've seen multiple cuts of this data, all of which results which demonstrate consistent efficacy and safety in many different kinds of patients. We are observing evidence of anatomical benefit when we switch from other VEGF therapies, which, again, is very encouraging when you think that in the beginning of a launch, who do you get, you tend to get the most severe patients and that's who we're seeing and we're still seeing great benefit. Importantly, and since this slide was done, we actually have an update. We're now at 130,000 vials that have been shipped with no new safety signals observed, which again gives us increasing confidence in the safety of VABYSMO. So probably top of your mind is how is launch going, and launch is actually going extremely well. We're seeing high initial uptake in high unmet need patients. We're seeing many of our patients switching from EYLEA. We have uptake across both AMD and DME patients with more AMD patients, as you might expect, given the epidemiology of this disease. We're also seeing broad access in reimbursement. We do expect that to accelerate with the receipt of the permanent J code, which we anticipate on October 1. You may not know that Japan is actually the second largest retina market in the world, and they are seeing extremely quick uptake of VABYSMO. Here in the U.K. NICE actually covered VABYSMO just 1 week after MHRA approval, which is quite unheard of. And we are just hearing a lot of excitement from our EU retinal specialists who are awaiting the approval of the VABYSMO. As Bill mentioned, we are continuing to invest in additional indications and formulations, including a prefilled syringe, which for any of you who spent time in a retinal physician's office, you know that they really value the convenience of having that prefilled syringe. So we're excited to see that come on to line as well. We don't want to forget about SUSVIMO, which is also launched now in the U.S. This is the first and only eye implant with a Q6-month drug delivery option. We have always told you that the launch of SUSVIMO was going to be a purposeful launch, and that is very much how we've approached it in the U.S. The focus has been on providing the best possible experience for our surgeons and for our patients. We have over 350 patients who have been -- or I'm sorry, 350 surgeons who have been trained, which is fantastic. We did receive a permanent J code on July 1, which has given us a corresponding increase in scheduled surgeries as we would have expected. We are anticipating EU approval next year, and we continue the platform delivery for portal -- we continue the development for the Port Delivery platform, both in studies in DME and DR, extended dosing studies, and importantly, the next generation of DutaFab bispecifics, which Nilesh will talk about later this afternoon. And finally, neuroscience. OCREVUS remains, far and away, the #1 treatment in the U.S. and the EU5 for MS. You guys are probably tired of hearing me say this, but we are the only drug that has robust, consistent, sustained disability and -- sustained delay and disability progression and the kind of safety that we have -- the kind of safety database that we know people want to see in neurology. And that makes OCREVUS really just a very hard drug to unseat. OCREVUS is the first and only therapy that approved in both RMS and PPMS, and that high persistence compared to other therapies is really important. We're still seeing around 90% of patients remain on therapy over time, which, again, in a chronic disease is quite unheard of. Paulo will talk a lot more about what we're seeing in our MS investments over time, but Bill mentioned the OCREVUS subcu data, which is expected next year. This is really important for OCREVUS. It both allows us to expand the patient population in IV centers where we might be seeing some -- where there might be some capacity constraints, but also potentially expanding into the community setting. The higher-dose OCREVUS studies, again, will help us continue to expand on the efficacy for OCREVUS. And then both fenebrutinib and Brain Shuttle programs are ongoing and have the potential to really revolutionize the care of MS. So lots going on in the MS space. Evrysdi is well on track to becoming the global market leader in SMA. I think it's just always important to level set us in the data that we see with Evrysdi. On the left side of the slide here, you see the RAINBOWFISH data. So this is the data that the FDA approved in less than 2 months old patients. That approval was achieved in Q2. It's important to note the Evrysdi is the only DMT that actually doesn't require baseline monitoring or labs. And that enables patients to get on treatment very, very quickly. And again, this is extremely important, particularly when you're dealing with infants because the faster you treat, the better result you're going to see. Since RAINBOWFISH approval, we've actually seen a number of babies treated within 24 hours of birth, which again is just incredible and gives them the opportunity to have the best possible outcome. We have over 7,000 patients treated worldwide. We're well on our way to becoming market leader in our most significant markets. We have rapid growth that is primarily being driven by switching. About 60% of new patients are coming from switching. About 50% of those are adults. We have high treatment satisfaction, and again, a very high retention rate. That lasting motor function approval is incredibly important and we're continuing to see responses maintained or improved over 3 years. Again, Paulo will talk about this more in the afternoon, but we are continuing to invest in SMA, and in particular, the MANATEE study, which looks at the combination of Evrysdi, along with an anti-myostatin, which could again be a really important advance for these babies. And finally, gantenerumab. Bill mentioned this already. We will finally see this data in Q4 of this year. The world needs a solution for AD, it just does. And we are going to learn a lot with the graduate data when it reads out. These are large, well-powered studies. They are significantly long enough that we should be able to see benefit. They've been well designed from a safety perspective. Probably what I'm most proud of is actually how we have incorporated a patient-centric approach to the design of these trials from the beginning, having the first subcu administration actually embedded in our Phase III development. And as Bill mentioned, the ability to be able to do that at home is really significant and important for these patients. And finally, we are also looking at how we can continue to provide access to patients all around the world for gantenerumab, including the development of an auto-injector, which will really help with the ease of administration over time. So finally, I'm going to close with a quick outlook of what's coming more in the midterm. So Bill mentioned what was happening in '22 and '23. I'm going to cover a little bit of what's happening in 2024. There are some very significant trials that Bill actually called out in 2022 and 2023. The HCC adjuvant trial is definitely one to watch as the head and neck trial is with TECENTRIQ. But as you look into 2024, there are a couple of things that are very significant. So I'm going to start in the line extensions with GAZYVA and lupus nephritis. Lupus nephritis is an area of incredibly high unmet need where we saw some very compelling Phase II data. This is the Phase III reading out and this could really be quite transformative for patients with lupus nephritis, and it could really help us understand the role of GAZYVA in immunological disease. So that is definitely a trial to watch. In terms of our new molecular entities, and again, Levi will talk more about both of these in his section. But the PI3-kinase, we see the first -- we see our first Phase III readout here with our front-line HR-positive trial in metastatic breast cancer, potentially a very significant population in a place where we could really do a lot of good for patients. But then we see the giradestrant first-line trial in ER-positive metastatic best cancer. And this is an interim analysis, but one that could potentially be fileable and that would allow us to actually be the first oral SERD in this patient population. And I think that's going to be a really interesting trial and one to watch because, again, patients in this setting really do need additional options. And giradestrant, for all the reasons that we've talked about previously, could be a really great solution for those patients. So at this point, I'm going to stop. I tried to make up a little bit of time, so we wouldn't be quite so late to lunch. So I'm going to invite Bill back. And I think we can -- how do you want to do this, Bruno, because we're pretty close? Okay.

You come back onstage and then we have the first Q&A session. So also to just to remind the people who joined via the Web, there's 2 possibility that you send us questions via the chat. So if we can -- I can read them aloud. We'll take the first few questions from the room. Simon, please. Do we have a mic? Yes.

Simon Baker from Redburn. Two quick questions, if I may. Firstly, Bill, the drug negotiation provisions of the Inflation Reduction Act mean that certain types of drugs in certain types of development approaches are disincentivized, whether it's small molecules or kicking off in small indications and then expanding. So can you give us some thoughts, although it's still fairly early stage, how this is going to affect what you develop and how you develop it? And then secondly, going back to the productivity slide. You showed how you've capped pharma development head count at 14% since 2016, which is a very impressive achievement. Doubly so, given that the figure was plus 27% on the slide last year. So could you tell us what's changed and what's driven that shift over the last 12 months?

Yes. Sure. In terms of the Inflation Reduction Act, I'd say the good news is that I don't think there's really significant effect on Roche in the -- even -- I mean, I don't know how you want to define long term, but certainly in the next 5 years, we see minimal impact, just sort of the nature of how the provisions are structured and the nature of our portfolio. The bigger impact is on small molecules. Of course, most of our major medicines are large molecules. Where we have our older products, they often would have biosimilar competition before the provisions would kick in. And of course, those provisions don't apply to products with biosimilar competition. So for various reasons, we see really minimal impact in the -- in any reasonable horizon. I mean, I think long term, this is obviously not a good development for innovation in the world. I mean the idea that small molecules would essentially have a 9-year life, which is I'd say that's the bleakest way to look at it, but maybe not wholly unreasonable because we all know government negotiation isn't a real negotiation. The government basically says what the price is. And yes, so I would put it this way. I don't think we plan to do anything differently in our portfolio today. But over time, that would have certainly a disincentive, particularly on small molecules, which would be unfortunate. And I would say we and our colleagues across the industry certainly plan to do everything possible to influence policy in a better direction in the future. In terms of PD, maybe I'll let Levi talk more about that when we're in the afternoon. But I would just say that we're just getting going in terms of understanding how to employ some of these new models for leadership in the realm of product development, which is not only in the in Levi's space, but also in areas like GPS and in process development. And we're really excited about the prospects. But in this case, whereas in the commercial organizations and in PT, we found ourselves often with too many people. In the case of, I think, in late-stage development, it's more of a case of we have a rapidly growing portfolio, which we think we can largely handle with the people we have.

Maybe the next question comes from Matthew.

It's Matthew Weston from Credit Suisse. Two specific product questions, please. Teresa, you called out the adjuvant ALECENSA data, excuse me, but you flagged that the ARIA trial was an interim read. How confident are you that we're actually going to see? Is there some reason why you're calling out an interim next year as the one we should watch? And then secondly, you also flagged subcu OCREVUS. As I understand the commercial benefit to clinicians, there's a very large amount of money made by buy-and-bill in their practice. So is the expectation that subcu OCREVUS will be physician-administered and they will keep that financial benefit? Or is this something that could be done at home?

Two great questions. So the reason that we're flagging the interim is that, that is the next cut of the data. As always, we hope that it's positive and that we're able to bring this treatment to patients as early as possible. This is a highly selective and effective ALK drug. So it's not unreasonable that we could actually see across at interim, but that just happens to be the next predefined data point. In terms of subcu OCREVUS, I think we should look at this as more of a market expansion opportunity. I think for many physicians and many patients, they will continue to appreciate and want to be on IV therapy. But there are some who either might have capacity constraints in their practices or for whom patients just the ability to actually administer at home would be most helpful. And so I think we would expect subcu to actually expand the opportunity for OCREVUS patients over time. The goal is for it to be home administered. It is possible that the first one might be in office, but the goal is for it to be home administered.

Maybe best, Matthew, you asked about the adjuvant readout. Do you have a study about the adjuvant readout?

Yes, we hit up on the first one.

Yes.

Yes, okay. The [ manual said ] originally scheduled for '24 -- this year and then it shifted basically over to next year. We can take a question from Richard Vosser, please.

I'll try a TIGIT one. Maybe on the cervical, TIGIT, you're highlighting data presentation next year. I mean that sounds subtly different to a readout. So have you seen the data for that? And what's it looking like, if you can give us any sort of color? And then on VABYSMO, we saw last week a high-dose EYLEA data, which looked reasonably comparable with VABYSMO. So just your thoughts on how that looks and how that can affect the uptake in commercial launch and potential.

Yes. So I think we're going to let Nilesh cover the EYLEA high-dose data in his section this afternoon just so we can kind of get the full context in. And in terms of cervical cancer, the study is still ongoing. The final readout will be presented next year.

So we hand over to current place? Sachin, you want to continue?

Sachin Jain, Bank of America. I'm going to try another TIGIT one as well as for me even though you said you wouldn't. So SKY-06, just any color you can give on the upsize criteria. Was it RR, PFS, OS? And did you -- this is the one I'm most interested. Did you require any improvements versus the comparator arm to upsize given you've made that Phase III decision? Like what level of confidence you have to move that to Phase III? Second question is on TECENTRIQ adjuvant lung. I think you said in your commentary that adjuvant lung is now the biggest single growth driver. So wonder if you could just clarify that and how much growth you see in that adjuvant lung indication with periadjuvant. Reason for the question is Slide 10, which you cited consensus. TECENTRIQ is the biggest single growth driver. So how much of that can come from the existing indications versus needing some of the adjuvant reads next year? And I shall leave it at that.

Do you want to turn it back on the TIGIT question?

Sure, sure, sure. So SKY-06, which is the non-small cell with chemo combination with tira, TECENTRIQ and chemo in the broader non-small cell lung population. So we had basically a futility analysis planned in, and we had basically a preplanned expansion to a full Phase III contingent on a passing of the futility analysis. And that's all we're saying.

Exactly. And we're blinded to that data.

Yes. We're blinded to the data. Because you can understand if you're taking a Phase II and making it a Phase III, you can't be like looking at the results. So we're blinded to it. There's an IDMC. They said it passed the futility analysis. That triggered the expansion to Phase III.

Maybe I add on from my side on the sales side, I mean, you asked about, Sachin. I think you have seen it in the U.S. that we had a nice reversal, Q1, Q2. This is the dynamic which you see here in sales, it's really driven entirely by the adjuvant non-small cell lung cancer segment and we have been guiding before that this is a multibillion opportunity this setting. And I think definitely additional data here, like the periadjuvant study would be supportive of having hit even some upside here. And you mentioned there's consensus upside till '25. This was a CHF 2.7 billion or something, yes. I think this is largely, I would say, adjuvant estimates and it's neoadjuvant, but then we have the other studies as well. We have the TNBC to come, which was not on the list actually because it's an external study, and the head and neck. So there's upside here from -- depending how these play out. What is the -- we may continue here.

Peter Welford, Jefferies. Two sort of perhaps slightly bigger picture questions. Just first of all, on the manufacturing, can you just talk a little bit about now your utilization at the moment of the sort of biologic facilities? You talked about investing nearly CHF 600 million in gene therapy. Just to understand, is that in anticipation of the Sarepta data? Or is that, I guess, to meet that demand? Or how are you thinking about that? It seems like a lot of money when you basically have relatively little projects at the moment that are in the clinic for gene therapy. And then just another quick one on the utilization productivity then. You talked about the head count going down, particularly in the U.S. Obviously, it's come down sharply. Should we be thinking in the U.S., though, is that trend largely now done? Or is there still room to further reduce head count, do you think, as business models change? And actually, I guess what I'm asking are we kind of at the end of that cycle for productivity of marketing.

Yes. So in terms of the production capacity, so we were running at pretty much 100%. Sometimes the calculation says we're at 103% of capacity, but I'm not quite sure how that works. That's like being -- yes, maybe that's like a football team giving it 110. But the bottom line is we've been at full capacity for a long time. And I'm actually pleased and relieved that in 2022, we are no longer at 100% of capacity. I don't know if it's 85% or 90%, but we're actually taking the opportunity to do some important upgrades and maintenance and things. So that's great in terms of our global biologics network. I think we're well positioned for the future. In the case of a positive gantenerumab readout, we have the capacity lined up and we've got capacity for the pipeline. So I think that's generally good. You asked about the gene therapy investment. That's not primarily about the DMD readout. But really because we have one of the world's largest investments in gene therapy, both at Spark and Philadelphia, but also a significant investment in pRED in gene therapies for a number of diseases. And so that investment, I think, marks our commitment to gene therapy for the future. But it's also, I should note, that it's flexible enough that, that space, if we end up not needing at all for gene therapy, can also be used for some of the more modular production, things like mRNA vaccines or small capacity single-use bioreactors. So we're trying to invest in space and facilities that can be flexible. And if you've been through a biologics plant, you have the giant plants like we have in Vacaville with 25,000-liter stainless steel bioreactors and they're fixed because you can't really move around a 25,000-liter reactor very easily and fixed piping. The plants of the future don't look like that. They're kind of big empty rooms like this, and then you have these modular reactors with single-use liners and flexible tubing that kind of hooks them up. So they're sort of modular plants. So I think it's a good investment. It anticipates our future needs in gene therapy, but it's also flexible enough that if those don't all come through, we can deploy for other things.

And we can take the next question over here. Keyur?

Can you hear me?

Yes.

Great. Keyur Parekh, Goldman Sachs. Bill, one big picture for you and then one kind of specific question. You've spoken very passionately the last 3 or 4 years about the need to kind of increase your investments in R&D. If I look at your R&D spend, I think you've put together a very clear slide, 27% of revenues on R&D and then maybe 3 or 4 other companies in that bucket. Lilly and Astra consensus is forecasting them to grow kind of high single, low double-digit top line. The CHF 11 billion in incremental revenues that you spoke about would be kind of mid-single digits on your CHF 45 billion base. So at what time do you expect that increased R&D investments to translate into increased revenue growth kind of for the group? That's question number one.

So great question. I think there's a couple of factors at play that drive the math, right? I mean one is we're starting from a significantly larger base than the other 2 companies you mentioned. And so a given amount of absolute growth isn't the same relative growth. That's kind of obvious, but it's a factor. Secondly, R&D, it takes a while to mature. You don't invest more in R&D this year and get a revenue payoff next year. So the biggest bump-ups we made were in '19 and '20. And I'd say probably the median time to pay off for that is probably 5 to 7 years, median. So that increased investment funded some Phase III things, like giradestrant, like inavolisib. It funded our Phase III studies in muscular dystrophy, which will read out next year. So you'll see the beginnings of that paying off in 2023, 2024. But I think the sweet spot of the payoff is probably more 2025, '26, '27. And yes, I think that's why we feel really good about our growth prospects in that basically over the course of the decade with steady growth in the early years, but probably accelerating growth in the later years.

And then secondly, gantenerumab, both of you have been very specific. And I think those are the words you use where we will all learn a lot when that study reads out. How would you characterize your level of optimism kind of around the assets, knowing what we know about the space and knowing we're going to see probably a couple of other competitive readouts before we see kind of the GRADUATE 1 and 2 studies? How much of a read-through should we have from those studies on to GRADUATE 1 and 2?

I mean, personally, I think it's probably a coin toss. I think we have every reason to believe that gantenerumab will biologically do what we know it can do, which is to reduce plaque and the question is whether or not that has a meaningful impact on cognition. And I think that is the million-dollar question and the thing that we will learn the most from when that -- when those data read out. And I think that's kind of what we're all waiting for, is to definitively know whether or not reducing amyloid plaque actually can improve cognition over time.

Yes. And you'll get more on this from Paulo in the afternoon. But I think what I would say is I think because we're Roche, we tend to be conservative. We want to deliver something before we proclaim victory. And so sometimes people read that as a lack of confidence in our asset. I think we have great confidence in gantenerumab, in the study designs. These are the largest studies. They're 27-month endpoints. I think we have a great opportunity to show a benefit. And the MoA is it's not conclusive, let's say, based on all the totality of the evidence to date. So that's where we are. I think we're -- you shouldn't read our caution as a lack of enthusiasm for gantenerumab, but rather it just reflects, I think, the scientific consensus about the impact of lowering A beta.

That side is uncertain.

So maybe we take a final question before we head into lunch. One final question.

Ben Yeoh at RBC Asset Management. I was interested in AI. Do you think you're seeing any gains from that machine learning and particularly the advances that deep mine has made in protein folding or where that will go in the long term? Secondly, on the Alzheimer's, were you worried at all some of the originating research around amyloid beta has been shown to be maybe potentially fraudulent or perhaps not as strong as we thought. Obviously, you talk about the totality of data, but that -- did that worry you, I would be interested in your thoughts there. And the last one is, I mean, you spoke a lot about the innovation and things that you're doing. What do you think is most underrated or potentially misunderstood about what you're doing at Roche?

I'm going to let you talk about the underrated and I'll hit the first 2. So you can think about it a bit a minute.

Okay. Excellent.

Yes. The first 2, in terms of the use of AI, it's basically it's happening everywhere. And I think it'd be great to hear more from Levi and Paulo and Nilesh in the afternoon. But I can say I've been around the world twice in the last 8 months to all of our innovation centers. And in each of them -- and now I'm just talking about research and early development because that seemed to be kind of where you were aiming. It's happening. I mean, we're -- for example, in the polypeptide cyclic molecules space, I mean, AI is allowing us to go much more rapidly in terms of creation of molecules and sort of virtual screening. I mean it's probably taken something like -- or allowed us to accelerate the number of molecules, the number of experiments we can run by at least an order of magnitude just in the last few years. And that's just one of many examples. Not particularly concerned about this recent news about potential inconsistencies and early experiments with A beta. Frankly, I think that's a bit of a sideshow because in the meantime there's been clinical results. And we all know clinical results trump these sort of preclinical things every time. And so I think that's -- that added fuel to the anti-A-beta crowd. But the fact is at this point in time, we're looking at clinical results. And we're going to have data on at least 2 of the major Phase III programs in the next few months. And so we'll -- hopefully, 2023 will be the year where we're arguing a lot less about whether A beta works. And hopefully, that will be because we've settled that it does. But in any case, we can maybe stop some of that silliness.

And in terms of where I think maybe we're undervalued or underappreciated, I do continue to think that people don't necessarily understand the depth and breadth of our ophthalmology investments, and I think we'll go into a lot more detail in that this afternoon. There are a lot of patients who are impacted with retinal disease, and we have a very large early stage and late-stage set of development programs that are targeted at helping patients with diseases that we've so not really been able to have much effect on. So in general, I think our ophthalmology portfolio is one that people should take a closer look at. I would say the same thing about our hematology portfolio. I think in general, people do undervalue everything that is in that hematology portfolio. And there is quite a bit in some areas that, again, have significant unmet need where patients stay on therapy over long periods of time, where we have the absolute -- where we have the ability to deliver game-changing results.

And I would add -- I agree with you. And I would add just because there are so many things to like. But the SERD, partly because of sort of the negative market news on that or the competitive news and so actually look forward to what Levi is going to tell you about that. And then inavolisib, which is the PI3K, again, because some of the early molecules haven't been that great and we think we have something that really is great and more from Levi.

Then I'm going to give Bruno one right now and toss in one additional thing. You did mention the TNKase and the stroke extension. I think again, if any of you have a family member or know someone who suffered from a stroke, you know what a debilitating event that is for someone. The ability to expand the treatment window is incredibly important. Right now, many, many patients don't actually get treated because you can't pinpoint the moment when you have a stroke. And so people who might actually have healthy tissue that can't be saved aren't. So the TIMELESS trial has the opportunity to be game-changing for patients all around the world.

Okay. Thanks, Bill and Teresa. And we head into the lunch, and enjoy your lunch. And we will meet back at 12:50 for the second session. [Break]

Okay. Good afternoon, everyone. I'm Levi Garraway. I'm the Chief Medical Officer and Head of Global Product Development at Roche, and it's always a pleasure to have the chance to discuss our pipeline. Before I get into some of the specifics, I'll just -- I'll take a couple of moments to give you just -- there are several strategic pillars in oncology that have guided our investment in recent years and I'll just kind of, at a high level, walk you through some of them. Maybe the best known of these pillars are precision medicine using genetic and molecular data to guide deployment of therapies. And then immunotherapy. And here, we've been, of course, increasingly interested in moving beyond the PD-L1, PD-1 access into new ways of engaging cytotoxic T cells against tumor cells. Now both of those pillars have been empowered by increasingly pursuing rational combinations. As you know, most oncology therapies don't end up as monotherapy. There are regimens that are anchored by best-in-disease therapeutics, and so that's a big drive for us. Teresa briefly mentioned moving more and more into early disease where you have the chance to achieve more cures, durable control. So that's a growing area for us. And then all of this is anchored by an expanding suite of modalities so that as the science progresses, we have new increasingly creative ways to interdict the targets and the disease biology. So on that last point, you've already seen a slide from Bill showing the breadth of modalities. And this is always -- to me, the modalities are a symptom of innovation for R&D organization. So the broader they are, that means that there's innovation going on. And Bill mentioned a couple of these, the cyclic peptides. So this is an area where our scientist colleagues at Chugai, basically, the key innovation here was to be able to generate orally bioavailable cell membrane permeable cyclic peptides. So those are -- each of those are huge feats, and then the ability to scale the technology, to screen these things because what you're doing is you're screening for the ability, for example, to block protein for interaction. So that is a feat in and of itself. And so having that available and now being able to see actual candidates coming through and the first one to actually be in the clinic, and of course, it's a pan-KRAS inhibitor, which could be very interesting. So there's a lot behind this. So this is the kind of -- it's not just, oh, it's a cool modality. You can easily see how there's a lot of unmet need that could be addressed with this kind of approach. And then there is the allogeneic CAR-T therapies. And we passed on making an investment in the autologous setting when we've discussed that. And actually -- in its stead, you can say, well, we've made a fair bit of investment in bispecific antibody therapy and we feel like that was a no-regret decision at the time. But we also appreciated that if there could be some advances that could get beyond some of the encumbrances of the autologous approach that, that might now be worth a different look. And so our strategic collaboration with Poseida Therapeutics gives us now an opportunity to potentially do exactly that. We have rights to several of the products that they generate from that platform, and I'll say a little bit more about those in a second. But let me now turn to sort of, I guess, the meat of the presentation starting in heme malignancies and just maybe a deeper dive into the modality. One of the modalities that we are now first-in-class in, in a couple of these indications, the CD20/CD3 bispecific antibodies. And so you already know about Lunsumio. This is approved in the EU. And so this is, of course, based on the complete response data of 60%, very favorable compared to historical controls. Long duration of response in these heavily pretreated follicular lymphoma patients and a fixed duration treatment with a very favorable toxicity profile. So this all is compatible with the community outpatient setting. So approval in the EU was achieved. Filing in the U.S. has been granted -- priority reviews have been granted. And so now we are -- as has already been mentioned, we have a clinical development program that is -- we have a couple of Phase III readouts upcoming in second-line follicular lymphoma in combination with lenalidomide, in second-line diffuse large cell lymphoma in combination with POLIVY and then a broader set of what we think are intriguing combinations that are in earlier stage. So certainly, Lunsumio has a potential to become an important new option for patients in heme malignancies. And then sort of the companion to Lunsumio in some ways is glofitamab. Just to remind you that the design hypothesis for glofitamab was that it's bivalent against CD20 on the tumor cell and monovalent against CD3. And so the concept there was that, that would generate avidity for the tumor cells that could lead to robust efficacy even in highly aggressive lymphomas, heavily pretreated lymphomas. And that hypothesis has sort of borne out in the form of a complete response rate of almost 40% in the highly refractory population, including a substantial percentage that had progressed even through CAR-T therapy. These complete responses are durable. At ASCO earlier this year, we reported that the median response was exceeding 34 months. So -- and again, this is a fixed duration of treatment, manageable side effect profile. So this has the potential to play an important role in the management of aggressive lymphomas. And here, too, we have a development program that is extensive and growing, including a Phase III trial, the STARGLO study that will read out in second-line DLBCL next year. It's a combination with GemOx and several other exploratory studies as well. So we think that both of these bispecifics have the potential to play an important role in a range of heme malignancies. Now you could say in some ways that the initial success with bispecifics in part has helped to open the possibility of bringing similar advances into different areas, including areas of hem malignancy where, as a company, we haven't necessarily had a major presence. And one area in particular that's of interest to us is multiple myeloma. Multiple myeloma, of course, like every other area of oncology, this is a competitive space. But there's still actually quite a high unmet need in multiple myeloma. The backbone therapies come with real side effects. Some of the emerging regimens, they're cumbersome. Affordability is an issue. There are quadruplet regimens. And so in general, we think that in our portfolio, there are actually several modalities that have the potential to benefit patients. Two of them are bispecifics: there's the small molecule, Venclexta, and then our allogeneic program that we -- through our recent strategic partnership with Poseida. And so in the bispecific space, one of the candidates that we're bullish about is called cevostamab. This is a -- so you have CD3 in one arm and then a target called FcRH5 on the other arm. This is a membrane protein that has near ubiquitous expression on myeloma cells. So this would be a first-in-class opportunity for us. You can see on the right side of the slide that the initial monotherapy data is looking quite encouraging. We have an overall response rate of over 56%, a very good partial response rate of 33%. The safety profile, the CRS in particular, generally combine to early cycles. And so this is an opportunity where we're moving this forward in combination with not only internal partners, but also external candidates that are obvious choices in myeloma as well. So that's one bispecific opportunity. The second one is it's now analogous to glofitamab in that it's a 2:1 format. So bivalent against the target. In this case, the target is called GPRC5D. This is a G-protein-coupled receptor. That is one of the most specific targets get identified in myeloma. And so -- and now you have CD3. So kind of similar mechanistic concept. Here again, we've got over 40 patients treated where the overall response rate looks quite encouraging, over 71%. Very good partial response rate of 52%. So this, too, is an area sort of orthogonal bispecific opportunity that we are exploring again in -- various combination studies are in the planning stages. Now as a company, it's likely that our first actual foray therapeutically into myeloma will occur with Venclexta. And this is through the subset of patients, the 20% subset of patients that have the t(11;14) translocation. And incidentally, the reason why this was a biologically plausible opportunity is because that subset has particularly high expression of Bcl-2, which, of course, is the target of Venclexta. And in general, Bcl-2 is a key survival factor in myeloma. Now so the -- you can see the subgroup analysis of the BELLINI study that led to sort of the clinical plausibility of this hypothesis. So we'll see very soon through the CANOVA study, whether this hypothesis proves correct. If so, we think actually, although the initial indication will be a subset, the translocation subset, but there are other subsets of melanoma that also have very high Bcl-2 expression. So there could be opportunities for expansion there. But while we're on the general topic of Venclexta, I should also just emphasize the data that supported the VERONA Phase III trial in high-risk myelodysplastic syndrome shown here on the right is quite encouraging. The combination with azacitidine and Venclexta, you had an 84% overall response rate and a robust duration of response, which was over a year. So obviously, we are very much looking forward to the readout of the Phase III trial, which is expected next year. Now the last point I'll make on the myeloma front, but it's, of course, relevant to several -- to a broader spectrum of hemo indices, this diving again more deeply into the allogeneic CAR-Ts. So we already mentioned sort of the encumbrances of autologous. And so basically, the concept here, obviously, is that if you now have an allogeneic pool, you have converted cell therapy into something that is a little bit more traditional in the sense of, it's scalable, it's off the shelf, you have lower COGS, you basically don't have delays to treatment. So basically, now you've opened up access to outside of a very restricted sort of health care context. But the reason -- the basis scientifically for our collaboration with Poseida was the fact that the specific expansion of allogeneic stem cell precursors is in this subset of T cells, which are basically the T cell -- stem cell memory T cells or TSCMs. And so these -- this subset of T cells, if you can expand them, the data suggests, including clinical data that you can have a higher proportion of T cells that can be administered. So you have -- you also have a favorable safety profile. You have the opportunity for greater persistence and also the efficacy that potentially could get a boost. And basically on the -- so Poseida has an autologous program, which has already basically shown proof of concept for the method. They've dosed over 100 patients with an autologous CAR-T, but basically the same subset of T cells. And so now, so our initial sort of collaboration involves a BCMA cell therapy target, obviously, well-validated myeloma target. And so this particular modality is already in the clinic for the allogeneic sort of subset and then the one that follows after will be a CD19, CD20. So basically, we're bullish from a scientific rationale about this platform because not only do you have a subset of T cells that potentially could overcome some of the encumbrances, but there's also additional hypotheses about how you might -- mediates or mitigate the host versus graft challenges that may undermine persistence by sort of crispening out some of the mechanisms there. So we think there's a lot of reason to be enthusiastic about this technological approach. So now we move from hemolysis into the solid tumors, and I'll speak about a couple of targeted agents, and I'll start by speaking about giredestrant. It's been an active year for giredestrant, including some disappointing readouts for several of the players, including ours. However, we still feel quite bullish about our program, and there are several reasons for that. And ultimately, this boils down to, obviously, it's a well-validated target, but we believe we have a molecule that has several best-in-class characteristics. And I'll just remind you of a couple of them. So one is the potency. This is the most potent member of the class of search that's been in development, 7- to 15-fold greater potency, depending on who your competitor is. You have a mechanism that is not just about ligand dependent inhibition and degradation, but it's also -- this was published in Cell several years ago, the binding by giredestrant immobilizes prior to degradation, it immobilizes, then sort of prevents engagement of chromatin. And so what that means is you get a profound suppression of the oncogenic transcriptional output. And so you have a ligand independent effect as well as sort of the traditional kind of inhibition effect. The third is the lack -- the combined ability, so the lack of issues of drug-drug interaction. So we know that in hindsight, in particular, that potentially a fatal flaw of some competitive molecules was the fact that you couldn't dose the SERD adequately. So you couldn't dose to the same degree as you could in a monotherapy setting, and you couldn't maintain bioavailability of the partner, for example, the CDK inhibitor when you combine. That is not a liability of our SERD. We can dose fully and maintain -- there are no drug-drug interactions that we've seen thus far with the obvious kind of combiners that we would like to include. And then there's also the safety. We've seen quite a very reasonable safety profile. Bradycardia has basically not been a factor clinically. And so these are all obviously very important best-in-class differentiators. But actually, the most important reason why we are bullish from a best-in-class standpoint is the aggregate clinical data that we've seen. And on the left of this slide is a Phase II trial, it's called the coopERA study, which are basically a neoadjuvant window of opportunity, which was a head-to-head study comparing giredestrant with anastrozole. And then after that, both of those arms were continued within -- in combination with palbociclib. And what you can see in the bar graphs on the right is the fact that giredestrant when compared head-to-head was -- this is a positive study, statistically significant increase in relative reduction of Ki-67 and there was a higher proportion of patients who underwent complete cell cycle arrest as measured by Ki-67 levels. So this is the first example of a randomized study in which an oral SERD has proved superior to an aromatase inhibitor. And obviously, this is in the new adjuvant setting and Ki-67 in ER-positive breast cancer in new adjuvant studies has often been predictive of what you would see ultimately with hormone therapy in the adjuvant setting. So it's relevant to our adjuvant study that's ongoing. So that's the coopERA study. Now the acelERA study was more disappointing sort of in first flush because it failed to meet its primary endpoint in an all-comers population, which, of course, consisted of metastatic -- later-line metastatic ER-positive breast cancer. Many of these patients that have already progressed on earlier lines of endocrine therapy. So in essence, what this sort of taught us is sort of a hypothesis was, oh, well, maybe there's still some residual dependence on the estrogen receptor in this population. And it turned out, in general, that was not true, except in the subset of patients that happened to acquire ESR1 mutations in their tumors over the course of their treatment. And of course, as you know, in ER-positive breast cancer, if you acquire an ESR1 mutation, ESR1 is the gene for the estrogen receptor, those tend to be activating mutations. And what they signify that those tumors are still dependent on the estrogen receptor. And so in this [indiscernible] curve that we're looking at, that's the subset of breast cancer patients that had ESR1 mutations. And you can see in that subset, clear evidence of activity for giredestrant. In fact, a hazard ratio of 0.6. So that tells us even in that metastatic setting, if you have dependence on the estrogen receptor, you have activity with giredestrant. So when you take this data, together with the endocrine -- the earlier stage of the neoadjuvant data that's shown here on the left, you have an aggregate rationale for enthusiasm about the potential for giredestrant to be an effective medicine, including in settings -- including the potential to unseat an aromatase inhibitor in earlier-stage setting. So there's clearly equipoised for the robust development program that we have with the SERD. So that's the giredestrant sort of story. Happy to discuss that more in the Q&A if helpful. And then the other one I'll mention is inavolisib. So Bill, sort of teed this up. This is our PI3-kinase alpha inhibitor. And obviously, there's a marketed PI3-kinase alpha inhibitor. But we think that inavolisib has the chance to be a best-in-disease PI3-kinase inhibitor. And part of the rationale for that is the preclinical data that's shown on the left of this slide. So what you have, first of all, is the potency. 58-fold greater potency for inavolisib than alpelisib. There is greater selectivity, several fold greater selectivity for the various beta and the other isoforms which, of course, in the aggregate are a big sort of issue in terms of off-target toxicity. And so what that adds up to is greater combined ability, the ability to combine full dose inavolisib with full dose CDK inhibitor, full-dose endocrine therapy. And so what that translates to, obviously, in our dose expansion cohort, you could see the favorable sort of clinical activity with this kind of combination. But the other thing that's not shown on here is, there are some patients who have been on this regimen for upwards of 2 years. So the ability to remain on therapy for extended periods of time, this has been an issue with the competitor molecules. So we think that there's a very good chance that we could overcome that. So not only do we obviously have our Phase III trial that will read out in 2024, which Teresa alluded to. But we also, earlier this year, announced a head-to-head study in which we are comparing inavolisib to alpelisib head-to-head in the post-CDK inhibitor setting. So we -- this is all to underscore our confidence that inavolisib could be a best-in-class PI3-kinase inhibitor that creates many opportunities for patient benefits, not just in breast cancer, but ultimately in other, PIK3CA mutant context as well. And then we have our KRAS G12C inhibitor. And so here again, we're behind the competition, but we are persisting because we think there is a very good rationale to believe that we have a best-in-class KRAS G12C inhibitor. Part of that is the preclinical data that you can see in the middle of the slide, which shows that the potency of GDC-6036 is greater than the competitor molecules. And so that sort of already was part of the rationale for our clinical development program, which is shown on the right side of this slide, which includes a Phase III trial in the second-line setting and a variety of combinations with, for example, immunotherapy angiogenesis molecules, other targeted therapies. But here's the clinical data that has us quite intrigued. And so on the left, you can see the waterfall plots from the second-line non-small cell lung cancer context, unconfirmed, admittedly still ORR, but you can see shaping up 53% response rate in non-small cell lung cancer. Colorectal cancer on the right, 35% overall response rate. If you -- you can look up the analogous numbers, obviously, there's usual caveats of cross-trial comparisons, but these compare quite favorably to the competitor molecules. And so we think this provides a reason, not just preclinical, but clinical rationale for the possibility that this could be a best-in-class G12C inhibitor and certainly justifies our clinical development plan. And so the final point that I'll make before I move on in the targeted therapy space is our recent entry into the DNA damage repair space. Obviously, PARP inhibitors exemplify the importance of DNA damage repair as a relevant sort of cancer mechanism and a cancer therapeutic mechanism. But there's a set of next-generation DDR targets that are being explored. One of which is a kinase called ATR. ATR is a key mediator. It basically -- it's activated in response to the DNA replication stress, and it's an important mechanism for repair from all of this recombination disruption. But ATR inhibitors have the ability to disrupt that. And just to cut to the chase, it's still early days, but we already know clinically that there's single-agent activity with the repair compound, which is called camonsertib. And so now a key question is can we dose at -- to adequate levels in combination with PARP inhibitors? And so in collaboration with the team at Repare, that's being explored now. So more to come in this space. But certainly, this is now a foray for Roche into sort of the next generation of DDR. Okay. So with that sort of I guess blitzkrieg view of targeted therapy, I'll move to immunotherapy, starting with Tecentriq in adjuvant lung cancer. What we're showing here is an updated cut of overall survival. And you can see on the left, this is the overall PD-L1 positive population and on the right is the PD-L1 high population. In both cases, the curves are separating in favor of Tecentriq. And already, you can see for the PD-L1 high group, it's looking quite clinically meaningful. And so our follow-up continues here, but this is additional data underscoring the benefit and sort of the rationale for the uptake of this indication that Teresa alluded to in her presentation. We have our tiragolumab slide. I guess basically Teresa sort of teed up, that I'd probably could have taken this slide out. But just, I'll just emphasize the point that the aggregate data sort of justifies that this continues to be an important program to us, and particularly, although we missed the primary PFS endpoint for SKY-01, the numerical differences that we saw in favor of the tiragolumab plus Tecentriq arm, and the fact that we have 2 more shots on goal for OS. And so basically, as you can see, we have a number of readouts, we've already discussed them. We've already discussed some updates. But just to underscore that we remain enthusiastic about the potential for this to be a new important checkpoint blockade approach in -- certainly in lung cancer, and then we'll see what these other indications hold as well. Now another immunotherapy asset in our earlier stage pipeline that we haven't talked as much about is another bispecific antibody, but here, we're cotargeting PD-1 and LAG-3. Now you know that LAG-3 is now a validated checkpoint therapeutically based on BMS' data in melanoma. But in the preclinical data that's shown on this slide, and I'm sorry, it's a little bit small. Basically, what we've seen is that the bispecific antibody seems to have a more profound effect on sort of tumor volume control than the 2 individual monoclonals. So if you have a monoclonal against PD-1 and a monoclonal against LAG-3, they don't work as well as the bispecific antibody. And so we don't yet know why this is. But one possibility is that by having them both on the same molecule, you sort of increase the avidity in the tumor reactive T cell context and you mitigate potentially -- binding is systemically of Tregs, which could bring sort of an immunosuppressive effect. We don't know for sure whether that's the reason, but certainly, these biologics as possible. But here's our Phase I dose escalation data shown here on the right. And it's still early days, but what's intriguing to us is that we are seeing responses in both checkpoint experience as well as checkpoint naive patients. So early days, but already, this data to us justified an expansion of -- we've added several cohorts. We have a cohort in checkpoint experienced melanoma, checkpoint experienced non-small cell lung cancer. We have a study in esophageal cancer. We have exploratory studies with tiragolumab. So stay tuned on this one. We think this could be an interesting one to keep an eye on. Okay. And the last point I'll make in this section, and I'm almost done, is just when we look ahead to immunotherapy, still remaining anchored in the science, but exploring additional approaches to how we can specifically be personalized in immunotherapy. So the ability to identify and then engineer therapeutics against this specific neoantigen makeup of individual tumors, some of these are cancer vaccine efforts, including a collaboration with BioNTech, where at ASCO, we reported -- still early days, but some intriguing data from one of these products in resectable pancreatic cancer. And then we've also have partnerships where we're exploring the ability to engineer T cells against those neoantigens and then put them into either autologous or allogeneic settings as a new cell therapy approach. So there's obviously a lot of hurdles to overcome here, but this is -- these are examples of the ways that we continue to follow the science in areas that seem potentially prime for the coming years to lead to new breakthroughs. Now I mentioned at the beginning, the migration into early disease. This is obviously the importance of the ability to achieve cure and durable control is not lost on us. But we're showing here is that already on this slide, there -- we have 10 ongoing Phase III trials across 5 earlier cancer indications. And there are additional -- at least a couple of dozen additional earlier-stage studies in various early cancer indications that are happening in our portfolio. So as these mature, obviously, we think this gives us opportunity to bring outsized patient benefit in several very important oncology indications. And so I'm going to conclude with just a couple of vignettes from our nonmalignant hematology, one of which has already been alluded to, crovalimab. This is our anti-complement C5 monoclonal antibody, which was designed using a recycling antibody technology from Chugai. And the important thing here is that this allows monthly subcutaneous dosing, which will be a difference from the other C5 antibodies that are available, which have either IV dosing or more frequent sort of patch pump type dosing. The initial data, which Bill mentioned earlier, a positive study in China from the COMMODORE 3 study. So we already have that in-house. Next year, we'll see the global data. One is, looking at patients who might switch to this modality and to this therapy and others looking at naive patients. So that's paroxysmal nocturnal hemoglobinuria. That's the -- it's a mouthful after lunch. But there are several other mouthfuls like atypical hemolytic uremic syndrome, Guillain-Barre syndrome, even sickle cell disease where there's a biological rationale for targeting complement. And so we have studies ongoing in each of those. And then finally, our Spark hemophilia A gene therapy product. So here, the data that we have, we see several favorable features. One is the sort of predictable expression, sustainable expression. You can see that sort of on the right graph, in 16 out of 18 patients that maintain expression. The median follow-up here now is approaching 3 years. And then sort of the middle graph is the reduction in annualized bleed rates, 91.2% reduction. And so there's been a lot of work by our Spark colleagues to optimize the dose to try to -- we want to maximize efficacy at the lowest possible dose and then also to tune the immunomodulatory regimen. We all know very well how important it is to get that right. So -- but with that in hand, we expect to start our Phase III trial with this asset in the first quarter of 2023. So I'll conclude just by reminding you that hopefully you already know that we feel in oncology, we continue to have one of the most intriguing, certainly one of the largest, if not the most interesting portfolios in the industry. And today, I've mentioned to you a few things. One is that we have several ongoing programs that we feel can extend our leadership in the heme malignancy space, including into new areas. We have several precision oncology opportunities that we think offer best-in-disease potential. We have immunotherapy opportunities ongoing that can break new ground in many ways, whether it's new checkpoints, new modalities, et cetera. And generally, in our -- we have a large thrust in early cancer. And then we -- with our modalities as a whole sort of as a symptom of innovation, we think we are exploiting opportunities to take, to attack many new targets in many different ways. And altogether, we think this positions us in oncology to remain leaders for years to come. So with that, I'll stop, and I'll invite Paulo to talk about the neuroscience portfolio.

Thanks, Levi. Good afternoon, everyone. It's a real pleasure to be here. So my name is Paulo Fontoura and essentially, whatever Levi did not cover, I'm going to cover to a large extent and Nilesh, then is going to follow me with the ophthalmology update. So just to start on neuroscience. As you can see here, and I've been saying this for a few years, I think that we actually have one of the best, one of the largest efforts in neuroscience globally. And as you can see here, this includes several molecules, some of which have been launched and developed, some of which are in Phase III, but a really steady, robust pipeline, very diversified. Several types of treatment modalities, not just small molecules, but also large molecules, gene therapies, antisense and also really addressing both very large medical needs, such as Alzheimer's or Parkinson's, but also more rare diseases in which we think there's a faster path towards development. So really a very, very exciting pipeline. And I'll be trying to provide a few highlights on the key programs here. So starting with Ocrevus, obviously, the flagship program. As Teresa mentioned, Ocrevus has really been a game changer in terms of multiple sclerosis therapeutics, the only drug approved for both relapsing and progressive forms. And the data just keeps getting better year after year. So what are we still doing with it? Well, first of all, as was mentioned, obviously, patient convenience at home is a very key topic, and we think it's a growth opportunity and the ability to get more patients have access at home. Therefore, our subcu program, which allows patients to be dosed every 6 months, would be a big game changer for them. Additionally, and we showed this data before. We know from the analysis of our Phase III program that deeper levels of B-cell depletion leads to higher levels of progression control. And therefore, this is something we're trying to really look at now with our high dose program. This looks at essentially maximizing the effects on progression in visibility for both relapsing and primary progressive patients. And the third arm of this is that as we believe that progression is mostly driven by resident T cells in the CNS, increasing the penetration of Ocrevus into the CNS using our brain shuttle platform, we think it could be another arm that allows us to really control progression in these patients. Now of course, the next target that everyone is really talking about is BTK. And as you know, we have a BTK inhibitor in Phase III development. Now we think ours is possibly the best-in-class molecule here for a variety of different reasons. First of all, it's the most selective one. Secondly, it's a very highly potent one. Thirdly, it's noncovalent binding, so it's reversible, which is important for safety purposes. And fourthly, we already have a very large safety database, including over 13 Phase II trials, over 1,700 patients dosed already, and we know the safety profile of our drug really, really well. Now what we've tried to do with our BTK is actually to do a development program that addresses the real medical needs. So as you can see here, we are going head-to-head against standard of care therapies, both in relapsing and in primary progressive disease. So that's against teriflunomide in relapsing disease and Ocrevus in primary progressive. And the reason for doing this is that we're not just interested in developing another MS therapy, but really in developing the best MS therapy. And BTK might actually be particularly relevant here because it addresses not just B-cell function, but also microglial and macrophage function, which we think are key drivers in the pathophysiology of progression. Turning a little bit over to risdiplam in spinal muscular atrophy. Now of course, the data that has been coming out is super exciting. And I wanted to point out just a few things. First of all, the largest prevalent population in SMA are actually patients with type 2 and type 3. These are the older children that have had progression and visibility for a long time. And for these patients, the thing that matters the most is not just gain in motor function but it's actually stability of clinical outcomes. So not losing any more function. And the data that we keep putting out from our SUNFISH trial which that you can see here on the left-hand side essentially shows that year-on-year, there's not just an increase in motor function, but actually stability. There's no decline in these patients, especially when compared to natural history, which you can see here in the middle box. Increasingly important as well is the -- that the safety profile of these drugs needs to be compatible with long-term dosing. Again, on the right-hand side, you see here the adverse event rates for -- in those SUNFISH patients, those are the ones that were originally on drug and the ones that came through drug after the first year. And you see year-on-year actually a reduction in the adverse event profile. That's possibly due as well to the improvement in overall health motor function. These patients are just getting less sick and therefore, the overall outcome is much better. Of course, we're always very, very excited when we look at data from our RAINBOWFISH trial, and Teresa mentioned this. This is a trial done in babies with SMA who are identified by neonatal genetic screening. This is possibly the biggest paradigm change in terms of how you treat patients with SMA as earlier as possible. And we know that leads to a better clinical outcome. So RAINBOWFISH was decided -- was essentially trying to address that. Could presymptomatic or as early as possible treatment result in an almost normal clinical development? And the data is here, essentially, it shows that, both on CHOP as well as on motor milestones, these babies are developing essentially normally and achieving maximal scores, and again, with the same safety profile. Now of course, this is a big game changer. And the fact that for risdiplam, you do not need to do any sort of testing for antibodies or things like that or you need a special IT. This is much easier for physicians essentially to diagnose and start treatment as soon as possible. And for these babies and moms, every day counts literally. So this is super relevant. So where next? I mean we've had 3 disease-modifying therapies for SMA developed in the past decade. But again, for the most prevalent population of type 2 and type 3 patients, these patients need to regain motor function. And one of the best avenues to do that is just to rebuild muscle. So muscle growth became a very important topic. And one of the best validated mechanisms is that is blocking myostatin. Myostatin is a key negative regulator of muscle growth and therefore, targeting myostatin has become really topical. Now GYM329 is a molecule that came from our Chugai colleagues. We think it's the best-in-class molecule to target the latent myostatin. It does so extremely potently and because it has a sweeping recycling technology. It's once a month dosing. It's a very, very good drug. And we have clinical data -- sorry, preclinical data that you can see here in the middle panel, that shows that in animal models, it leads to muscle buildup and return of function in these animals. So what we did based on this is actually start a trial now called MANATEE that looks at the combination of risdiplam and GYM329. And why does risdiplam make the most sense here in my mind, because you need a drug -- you need a disease-modifying drug that is administered and has systemic effects, not just CNS targeting. So everything being oral and systemically available to all the body essentially corrects the genetic defect in muscle cells, in the whole body and then you can combine that with a myostatin that would lead to muscle growth. Now moving on to Enspryng. So Enspryng, we think, the best-in-class anti-IL-6 receptor antibody. Again, the molecule coming from our Chugai colleagues. Sweeping, recycling technology that leads to very high levels of suppression of IL-6. Now we know that IL-6 is central to several diseases of the nervous system, myasthenia gravis being one of them. And therefore, after our approval in NMOSD, we decided to go to myasthenia gravis as our next big disease. Of course, this is not an uncommon autoimmune condition for which there are several therapies. There's still a tremendous amount of medical need for these patients, especially for patients with generalized myasthenia gravis. And therefore, we started the trial to look at that. At the same time, we have other trials looking at other diseases of the nervous system, which are mediated, we think, by IL-6, such as MOG-AD and autoimmune encephalitis. So you've heard already this morning a lot about our excitement around gene therapy. And for Duchenne muscular dystrophy, gene therapy could be the biggest game changer in the history of this disease. Up until now, this is an invariably fatal condition with very high levels of visibility, and the treatments that are approved right now really provide very, very small benefits. So gene therapy is and will be the big game changer if it works. Now we're very excited about, this one is called delandistrogene, talk about mouthfuls. This is a -- it's a gene therapy that we are developing in partnership with our colleagues at Sarepta. It has, we think, the best mini dystrophin construct with the promoter MHCK7 that essentially drives expression in muscle and cardiac cells. At the same time, it has an AAV vector, which is specialized AAVrh74 for which there's much lower levels of pre-existing antibodies. And therefore, it's much more likely to be administered to a bigger range of patients. Now we're very excited about this one as well because of the emerging clinical data. Now in the middle panel here, we're showing you the data from Study 103. This is the commercial product, essentially the product that is developed in Phase III. And we see here in this trial that has 20 patients, an increase in over 4 points in the NSAA, which is a motor score, the primary endpoint and that's over 1 year. And when you compare that with natural history, which is the bar chart in the middle, you see essentially a delta of about 3, 3.5 points. Now that's not only remarkable, but clinically very, very meaningful. On the right-hand side, you see data from the first study ever done, Study 101. And for these patients, we see already 4 years data. And in those patients, you see essentially a 7-point increase in the NSAA. Now durability in gene therapy is a very important topic, especially as we know these boys grow, muscle cells divide and therefore, there's this worry, will this stay, will this benefit stay in place? And apparently, it does so. Now equally importantly here is that the safety profile that we've observed so far for delandistrogene is very encouraging. As we know and we have seen this, gene therapies are still very young technologies and there's still safety concerns about them. So far, delandistrogene appears to be a safe therapy that can be administered very, very broadly and provides clinical benefit. So based on that, we're actually in the middle now of a very large Phase III program. On the left-hand side, you see our core pivotal trial, which looks at boys between 4 and 8 years old, looking at 1-year duration of follow-up and the NSAA as the primary motor endpoint. This is still, for us, the core package of data that will lead to approval for delandistrogene. On the right-hand side, you see the additional pivotal trial. So we're going to look at both older ambulatory and nonambulatory boys, but also younger below 3 years of age children to really try to get as a broad label as possible and then demonstrate the clinical benefits of this therapy in the widest possible range of these patients. Now switching over, which, I guess, is one of the reasons why you're all here. Gantenerumab, as we've mentioned already several times today and as Bill mentioned, our conservative view here does not represent a lack of commitment or enthusiasm, but it is a very difficult area. So I thought it would be important just perhaps is level set where we are now by starting to describe what is the background as we've come into this big data readout in the end of this year. Well, first of all, gantenerumab is an anti-beta amyloid targeting antibody, targets fibrillary aggregated forms in plaque very, very potently and very safely. We know from our long-term extension data from Scarlet and [indiscernible] that after 2 and 3 years, over 50% and over 80% of patients become amyloid negative by PET. So it clearly does what it's supposed to do. It gets in the brain, removes amyloid. Based on those data, we designed a GRADUATE program. And the GRADUATE program, we believe, is going to provide essentially kind of a conclusive answer to the question that we all care about. Does treatment with gantenerumab over a sufficient duration of time, so we have 27 months at a high enough level? So everyone's titrated up to 1,020 milligrams regardless of APOE genotype in the right patients. So prodromal and mild patients identified by CSF testing and PET imaging. So essentially, does that right experiment demonstrate clinical benefit or does it not? And so we're very, very excited about the outcome here. Obviously, we don't have a crystal ball, but I think we're very confident in the job that has been done. This is a very well-designed program, which will provide clear answers. Now just preempting one of the questions which I know that we'll get is, does the change in AML or Alzheimer's diagnostic criteria over time, make a difference, the inclusion of tau versus not at baseline. We really don't think it matters much. We're showing you here the baseline data for the patients of GRADUATE. Essentially, these are very typical prodromal to mild patients, which are identified with state-of-the-art clinical and biomarker criteria. So again, we think we've done as much as we can here. Now this isn't in here. Of course, we know that Alzheimer's is a disease that develops over years and decades. And therefore, actually, the biggest population out there is the presymptomatic one. And if a drug works in prodromal and mild, I think it's a fair question to say, does it work in presymptomatic patients as well? So based on that, we decided to start this Phase III trial called SKYLINE. This is looking at presymptomatic Alzheimer's patients that we can identify now based on biomarkers, such as CSF testing or PET imaging, and look at them over 4 years to see if we can prevent the onset of dementia. Now of course, it's a very different clinical problem to try and treat presymptomatic patients. These people by definition have no symptoms. So how do you -- have identified them not only for clinical trials but then for future practice? That's why we are developing as well these blood-based biomarker screening tests to allow people to have an easy way to essentially have a high probability of making it into the screening funnel. And so these blood-based biomarker tests essentially allow you to double the screening efficiency if you want. So many, many fewer patients will get into do CSF testing or PET imaging if they have a positive blood-based biomarker test. And for this, this is running in our Elecsys platform, and we've got breakthrough device designation for that as well. Now gantenerumab is obviously the flagship here, but it's not our only program. I'd like to mention very, very briefly 2 other avenues that we're continuing to develop, one of which is the brain shuttle version of gantenerumab. It was already one of the biggest problems we have with large molecules that's only about 0.1% to 0.2% get into the brain. So obviously, technology that would allow up to 50-fold higher CSF and brain parenchyma concentrations would be very, very relevant. Brain shuttle gantenerumab is based on that. It's essentially gantenerumab with a module that targets the transferrin receptor on endothelial cells in the brain and allows the transfer of this antibody. And you see the CSF data here from our first Phase I trial. It shows up to 8-fold higher CSF levels of branch shuttle gan versus the non-brain shuttle version. So we're really excited to see these data. Tau remains, I think, inevitably one of the big targets of interest for us. It is undoubtedly a key in the pathophysiology of Alzheimer's. We're targeting it now with 2 monoclonal antibodies semorinemab and bepranemab. The data are still coming in. It's not been easy and it's not continued to go -- it's not going to be easy, but this is one of the long-term methods we're excited about. Tau is going to be part of a solution for Alzheimer's and for many other neurodegenerative conditions inevitably. So we want to keep doing that. Now switching gears here to the earlier pipeline. So for Parkinson's, which is one of the most common neurodegenerative conditions out there, there really has not been a breakthrough in 2 or 3 decades. So we're still treating patients with Parkinson's essentially with dopa, dopa agonists, and that's really it. So new therapies, disease-modifying therapies could be really a game changer for these patients. Alpha-synuclein is one of the most validated genetic targets in Parkinson's, if not the most. And therefore, we are excited to develop prasinezumab, which is a monoclonal antibody that targets aggregated forms of alpha-synuclein. Now the data I'm showing you here is really the first glimmer of hope for this target and for this molecule. Other companies have tried this and failed. Our data shows at least in the PASADENA trial that in patients with very early Parkinson's off dopaminergic therapy, you do see a separation in the motor score of the UPDRS. Now this separation prolongs itself, not just after 1 year, but after 2 years. We saw it as well in our digital endpoint, which you see here on the right-hand side. This is a new technology based on the Floodlight platform. So that kind of gives us to a little bit more assurance that this is a real signal. But obviously, this is by no means a definite signal. So we're trying to confirm these data in the path of the trial. The path of trial looks at early Parkinson's patients treated with dopaminergic agents, and we're trying to see if we can replicate this signal before moving forward. Just a brief note on Angelman syndrome. Now this follows on our quest really to look at rare neurological diseases with highly validated targets. In this case, Angelman is a rare form of neurodevelopmental disorder. It's a form of autism. These children essentially, after 2 years of age, don't develop any more cognitive function, they have mental retardation, they have seizures, they have sleep problems. Really, really high burden, not just for them, but for their parents, for their caregivers, and there's no therapies here. Now we know that Angelman is caused essentially by a loss of function mutation in the gene called UBE3A on chromosome 15q. And what we found out is that we can actually derepress the parental copy of this gene in the brains of, obviously, animal models, but of these patients as well by using an anti-sensor, essentially blocks, a blocking signal that all of us already have. And therefore, we kind of rescuing this function by using a gene that's already there. It's very exciting new approach. And we're in the middle of our first trial. This is called TANGELO. You see it here on the right-hand side. We have a first portion, which is essentially a dose-ranging multiples and in dose portion in children up to 12 years of age. And we're really going very carefully here because we know and as one of our competitors have this, there are concerns around safety. So far, we really have not seen anything, and we were able to dose escalate to the maximum dose level. And now we've moved from this first multiple ascending dose stage to the long-term extension phase on the extreme right-hand side here. So this part 2 of this trial really will look at the long-term benefits potentially that patients can get by derepression of UBE3A gene. And we're very excited to see that. Hopefully, we'll get data as early as the end of next year, beginning of 2024. The final word on stroke is to pick up where Teresa left. I mean, stroke is one of the real medical needs in neuroscience. It's the fifth global cause of death from any cause. And in neurology, it's the leading cause of disability for all patients, especially for older patients. Then unfortunately, really no advances have occurred in this field in the past 30 years, alteplase, which is still the only approved drug with an FDA label, is a drug from the 1990s. It's a great drug. It has a limitation that can only be given up until 3 hours. Now that precludes the large majority of patients from getting thrombolytics, only 50% of patients can -- the overall stroke patients get alteplase. And now the eligible patients, only 50% of them get it because they don't get to the hospital in time. So tenecteplase or TNKs was really developed to be the best-in-class thrombolytic. It has a higher affinity for thrombin. It has a higher specificity for fibrin as well as faster onset of thrombolysis. So essentially, it allows this drug to be more potent if you want, and you have a faster start, and therefore, to be administered as a single bolus instead of infusion. So on that side, an advancement was made. What has happened in recent years is that more and more trials have been made with this drug, which has not led to an approval, but have generated a lot of evidence that it's equally as good as alteplase, potentially better for a large vessel stroke. But really, the question here is what happens if you don't get to the hospital within that time window. And that's what the TIMELESS study is trying to look at. For patients that have salvageable brain tissue. And now we have advanced imaging that allows us to say for any stroke patient, there's still brain tissue that can be rescued. Can we up until 24 hours still give Tenecteplase safely and have clinical benefit? That's really what the outcome here is. And so we're looking at patients that will be treated both with Tenecteplase as well as mechanical thrombectomy, meaning that against a broad range. And we're very excited to look at these data as they come out early next year because if positive, this could really be the first real breakthrough in this field for decades. Now switching quickly to our immunology efforts. Of course, we're very excited about the medicines we have on the market now, but they are coming to the end of their life cycle, and you saw that. There's still a huge medical need here. So this you're seeing here is really the renewal of our efforts in this field. So right now, the most advanced programs are in nephrology and respiratory. We have earlier programs as well. I briefly mentioned, we also obviously have our dedication to flu, to hepatitis B. But today, I really would want to focus on our emerging pipeline of late-stage molecules for nephrology and for respiratory. The first one is Gazyva in lupus. Lupus nephritis is a disease that affects about 5 million people worldwide. So it's a big medical need. All -- about 25% of them really don't get any stabilization with current therapy. They go on to end-stage kidney disease. And therefore, we need just better therapies here. Now Gazyva is an anti-CD20 antibody that has been engineered to provide much deeper levels of depletion of B cells both in the blood and in tissues. This is relevant as we realized a lot of volume conditions are really driven by locally resident B cells and not by systemically circulating B cells. And therefore, depletion in tissues has become a real target for us. Now we're very encouraged by the Phase II data we generated in the mobility trial, essentially in lupus nephritis patients treated over 54 weeks, we saw essentially a 50% -- sorry, a 20% absolute increase in the percentage of patients that achieved complete renal response. So were essentially stable. And this percentage seemed to increase over time. So now for REGENCY, which is our Phase III trial, we're extending the duration of treatment up to 74 weeks. And based on these data and others, we're also extending to look at other diseases of such as nephrotic syndrome and membranous nephropathy, which are, again, major medical needs in the nephrology field. Our most advanced program in respiratory disease now is the zinpentraxin-alpha program. This builds on our commitment to idiopathic pulmonary fibrosis. Pentraxin is an immune regulatory protein that essentially shifts the differentiation profile of macrophages from profibrotic to proresoltiv, and therefore, we think reduces fibrosis. What we're very excited about is here in the middle panel is the data from our Phase II -- from the Phase II trial that was run and that got breakthrough designation. These data showed for the first time that on top of standard of care therapies, such as Ofev or Esbriet. We still saw a significant improvement or preservation in functional vital capacity, over 2.5 percentage points, over about 50% -- 50 ml of FVC. And at the same time, the stabilization in motor function over 6 months. This is unheard of. And therefore, based on that, we went on to Phase III. Our Phase III trial is a very large trial, around 660 patients, over 1-year treatment, again, on top of either Ofev or Esbriet. And we're really looking forward to seeing these results in 2024. And really the final one that I wanted to talk about is our recent addition to the pipeline, which is our antisense ASO Factor B, a program for IgA nephropathy. Now IgA nephropathy is actually the most common immune-mediated glomerular disease in adults. It is a severe condition. A high percentage of these patients go on to end-stage kidney disease, they need a transplant. And we do know that this is a disease caused by immune deposition of both antibodies and a complement. And therefore, targeting Factor B makes tremendous sense. It essentially blocks that complement cascade. Now we have Phase II data, which is not published yet, but I can share just top line. We've seen a very robust and significant reduction in the decline in glomerular function. And therefore, based on those data, we were very eager to bring this molecule in and start Phase III trials, which should get started as early as the beginning of next year. So with that, I think I'll turn over to Nilesh, tell us about the ophthalmology pipeline. Nilesh?

Thank you, Paulo. So it's a real pleasure to be here with everybody. And it's my role over the next 15 minutes or so to give you an overview of what we're doing in ophthalmology. I'm Nilesh Mehta. I'm based in Basel. I'm the Global Head for Ophthalmology. So let's just start off actually with the unmet need here. In the advent of anti-VEGF therapies, we still know that there's a huge unmet need for patients here, both in terms of in clinical trials where patients are followed up really robustly and treated rigorously. Even then only about half of patients get to really good outcomes for visual function. In the real world, of course, we know that health care systems and patients can't keep up with those treatments and therefore, they get much worse outcomes in the clinical trials. So both ways, what we're looking to do is even when you're treated robustly with anti-VEGF therapies, we know that, that however, much anti-VEGF you give, isn't the full answer for these multifactorial diseases, and therefore, adjacent pathways become really important. And then in terms of real world, as we get to greater durability, we'll be able to treat patients for longer and sustain their visual outcomes that way. So here, I want to just walk us through 2 slides here for Vabysmo trials. So Vabysmo in DME, here, the study design is quite important too. In DME, there were 3 arms in the study. The control arm was a loading phase of aflibercept and then followed by every 8 weeks. And then 2 other ones, faricimab with the loading phase in every 8 weeks, but also faricimab, a loading phase and then what we're calling personalized treatment intervals. And this is really aligned with treat and extend principles that many doctors use in the clinic today. What's important to note over these 2 years, we've now shared these data, the 1-year data in 2021. And this year, the second-year data is the consistency of the BCVA curves over time. Every time point slightly favoring faricimab and the 2-year data in terms of anatomy for both of those arms slightly favoring faricimab and that's been consistent over these 2 years. With AMD, the design is slightly different. So in the first year, again, the control arm was aflibercept and with a loading dose and then every 8 weeks. And then the arm of the study would start off with Q16 dosing -- Q 16-week dosing. But then at weeks 20 and 24, patients could drop down to 12 or 8 and then they maintained that dosing regimen right up to week 60 when patients rolled over into a personalized treatment interval into the second year. And again, you can see here the really consistent results, both in terms of function as well as anatomy on the right-hand side of this graph. And I think we'll see longer-term extensions from that data as we move forward. So we're really excited to be able to share Vabysmo of the Q16-week dosing increasing from year 1 sort of 50% to year 2, 60%. And this is now multiple cycles of Q12 and Q16 dosing. And this is consistent across all 4 studies. So I think this is really interesting and consistent data for the community to take away. You see the disease control in terms of CST. You're also seeing in the safety profile of the product in the Phase III studies, look very much akin to IBTs as doctors know them. And as Teresa mentioned, we now have about 130,000 vials out, and we're not seeing any concerning safety or new signals of safety emerging. Further, the clinical development program continues. So retinal vein occlusion, which is really the third biggest area of retinal diseases. We expect data at the end of this year. And then AVONELLE-X and RHONE-X are going to be really important because these are long-term extension studies that will go on for 4 years, but we'll also be able to see the impact from the comparator arm switched over to faricimab over a long period of time. And so if we talk about real world, and I'd like to talk about the study, which is the Truckee Study. This is not a study that actually supported by Genentech or Roche. This is a ground-up, investigator-led set of studies, they're at 14 sites currently operating. And this data has been shared at a number of conferences this year. It's really early real-world evidence for Vabysmo in clinical practice. And what you're seeing is really even in patients that are quite recalcitrant to anti-VEGF therapies, the early doses of the Vabysmo really showing a really potent impact on the anatomy here with -- on the right-hand side here, you can see previously monthly treated aflibercept patients, treated 1 injection or with Vabysmo and quite a lot of normalization of the retina, which is really encouraging to see. And now you're seeing about 1,000 injections in this study, and again, no concerning safety effects. Susvimo. So if we move to Susvimo, this is the first drug device combination in retina. It's truly miraculous to see over so many years, the development of this program. And obviously, as you know, after the surgical procedure, you only need a refill every 6 months. On the right -- on the left-hand side here, what you're seeing is that this is maintaining vision over 2 years compared to the gold standard which is monthly Lucentis injections. That's never been bettered in retina. And you can see that the durable response doesn't seem to be going down. So you're really still seeing 94% of patients going 6 months or longer with their refill exchanges. This program continues. We're going to see data from the PAGODA and PAVILION studies, which will be in DME and DR, towards the end of the year. And the VELODROME study, which we'll look at 9-month dosing as well. The other thing to note here, I think, is that really, there's no competition in the near term for this product at all or even the midterm even. And so the PORTAL study, which will be real-world evidence, by the time anything we think will emerge in this space, which will be a PORTAL or a long-acting device, this will have such long-term data. I think it will be really challenging for people to catch up with that long-term efficacy and safety data. So it's going to be very important to see how that data continues to roll out. So here with the ophthalmology pipeline, I mean we're really excited that we have Vabysmo and Susvimo being -- have launched in the U.S. and Vabysmo in a number of other markets this year. But there's a huge pipeline following in ophthalmology. And I think this speaks a lot to what Bill was saying about the earlier investments in R&D, allowing us to accelerate some of these programs into clinic. What I'd like to really talk a little bit about is some of the programs that we have disclosed, and I'll touch on some of these in the coming slides. But before I go into that, what I'd really like to do is just frame for you how we're thinking about approaching ophthalmology. So today, where we are in this middle column, today's therapy, we've seen this functional loss, and then we treat patients and they get some of that loss back. We're trying to optimize there with new modes of action -- new modes of biology to the eye, adjacent pathways to anti-VEGF to optimize outcomes. But also in the future, we'd like to go into earlier stages of disease, and that's becoming possible because of the amount of imaging as well as clinical biomarkers that we have. And we can really apply now AI and deep learning to those images to find those tipping points. And in the future, we hope to be able to prevent vision loss, especially in conditions like intermediate AMD or diabetic retinopathy. Further, the signs of evolving some of that with cell-based therapies or genetics, we'll be able to restore vision in parts of the retina, in tissues of the retina to give people who are blind now, the ability to see again, and I'll talk about a couple of those programs as well. Let's start with one of the programs here, which is the IL-6. And the unmet need here is really inflammation in ophthalmology and in retinal conditions. This has been an area of unmet need over many years. I might even say decades, if you go back to the earlier work from Professor Kostas and Mason, and there's been decades in coming. The important piece here is that the only approach to inflammation right now in ophthalmology is steroids, and they come with a plethora of adverse events. So here, we see IL-6. We know that it's involved in inflammation. And you can see in -- on the right-hand side here, how -- in pathologic eyes, how up-regulated that is as a cited. So we have an anti-IL-6 monoclonal antibody with a modified Fc portion that will allow for long acting in the eye, but quite fast systemic clearance. And we have a Phase II study ongoing in diabetic macular edema. That's in combination with anti-VEGF therapies. But this year, we've seen Phase I data in uveitic macular edema, UME. And this is a disease which is predominantly driven by inflammation. And so we think it could be a really interesting area. And we're going to accelerate this from Phase I into Phase III with a study start-up towards the end of the year or early next year. DutaFabs. I think this is something you've all heard of before in some of these meetings, but worth just touching on. We really think of these as the next generation of antibodies. So they are able to be concentrated, their fab fragments to highly concentrated and with high affinity to their binding site. What this allows us to do is have a stable molecule that could also go into the port delivery system, the sustained release. And the first of these that's going into clinic is going to be the VA2 DutaFab. That's in a Phase I study, and it will go from the IBT into the port delivery later this year. So let's move to geographic atrophy. One of the last frontiers of unmet need in retina. It's a multifactorial disease and highly correlated to age. And as you can see from this chart, in fact, I joined Roche partly for the lampalizumab program in GA, and I'm really proud that the team continued to analyze that data and provide data to the community because this is -- this will allow us to expand our knowledge to address this devastating disease. And is approaching Phase II, it's in collaboration with IONIS. ASO factor B, we're seeing emerging evidence that complement is a pathway that is involved in geographic atrophy and ASO factor B targets the alternative complement pathway. The -- both the preclinical and Phase I studies looked very encouraging, and this study is ongoing, and we expect some data in 2024. Then moving on to cell-based approaches. The OpRegen program is via a deal with Lineage Cell Therapeutics. What this does is allow us to introduce retinal pigment epithelium into the retina. And the -- as you can see on the left-hand side, after the introduction of a single delivery, some retinal delivery of this, the retinas really becomes quite more normal than before. So you see a huge amount of both anatomical convergence to norm as well as functional gain here, 7.6 letters, which we thought even with 12 patients was very encouraging. So these cell-based therapies will now really provide the opportunity to restore vision or restore anatomy even in quite advanced diseases. And I'd be remiss if I didn't also mention the tremendous efforts being made across the organization on personalized health care. The -- both in terms of remote vision monitoring, which will allow patients with their chronic conditions to have a much better experience of their diseases and their clinic practice. But also with the amount of data that, as Roche, we're gathering internally in our studies and the ability to apply AI and automated segmentation for OCT, for example. We'll be able to probably have better target selection, enrich patients into clinical studies as well. And so both of these areas are areas that internally and externally, we have lots of collaborations ongoing. And we think that in the future, we'll be able to address patients' disease over time with far more specific management tools. So really, with that, I'd like to say we're in the cusp of really new standards in retinal care. I'd be remiss if being in London and using the analogy of our buses that you wait for 3 hours and then 2 turn up together, we're at the time where really we've waited for 15 years in retina for new innovation. Now we've got 2 Vabysmo and Susvimo that both came together, which I think is quite amazing and using that as a backdrop with our pipeline will be able to hopefully expand outcomes for patients as we move forward. Thank you.

I think with that, I would ask all our speakers back on stage for the final second Q&A session. And I think now we have all the experts on the stage to ask all your high dose earlier questions because I think there're probably some around. Don't be shy. Emmanuel, you're the first one.

Emmanuel Papadakis, Deutsche Bank. Perhaps I'll take the privilege of the high-dose faricimab question, perspectives on Amic data, any nuances you think we should consider in terms of the competitive risk it may or may not represent? Question also on giredestrant. Your degree of confidence is showing in all-comer benefits in persevERA and lidERA based on the clear skewed benefit in ESR1 mutants and the acelERA data you just presented. And then perhaps a final question on bispecifics in myeloma. You're pretty late on GPRC5D. Cevostamab, we saw great data 2 years ago, doesn't seem to have got very far since then. So thoughts on Phase III, how are you going to catch up and differentiate in that landscape?

Do you want to go with the...

Okay. Thanks for the question. Yes. We -- obviously, we've seen the data. I think the headlines are one thing we'd really be interested now to see the fuller data set, hopefully, at AAO, especially with the BCVA curves over time and the CST curves over time. To put that in context, as I sort of mentioned in the studies for faricimab we had quite strict control criterias where you would look at both anatomy, function and clinical examination. And any one of those could trigger a shorter dosing interval. So we'd like to be able to understand this data a little bit further. So once we see the BCVA curves and CST over time, I think that will put more context on that, better.

Maybe just to reinforce that a bit, I think in our studies, we had a high bar for putting patients on to a less frequent dosing. And so in other words, a very much of focus on delivering the maximum benefit -- clinical benefit, maximum drying for every patient and a high bar for going on less frequent. I think what you want to ask and probe on the data from competition is, do they have a high bar for achieving or for taking patients onto a less frequent or a low bar for putting patients on less frequent sort of high bar to maximize the percent of patients on less frequent dosing? So I think that's -- to me, that's what really stands out on our data is that the patients in the 16-week arm, in the 12-week arm, they are getting very similar, almost identical outcomes to patients in the other arms. And if you see sort of a tailing off of efficacy in the less frequent dosing arms of another study, that might be because of a compromise on that.

Yes. So maybe I'll take the question -- on the giredestrant question. I think the key point is actually the fact that it happened to be ESR1 mutations in the later line, the key point is the context where there is dependence on the estrogen receptor or signaling in that cancer. And the earlier you go in breast cancer, the greater proportion of patients that have dependence on the estrogen receptor. So basically, in the frontline, untreated metastatic setting and certainly in the early-stage adjuvant setting, this is a near uniform dependence on the estrogen receptor. So basically, that's why -- and the potency that we have it, it irrelevant whether it's mutant or wild type. It's the same phenomenon. So as long as you have dependency on the estrogen receptor, giredestrant has a shot at bringing a benefit. So in the setting of the myeloma, so a general rule in cancer is that it's not about whether you're first to the later-line therapy. It's about whether or not you can achieve the right dose and the right combinations and move forward in therapy. So we think within our portfolio, we now have several opportunities to ultimately generate differentiated combinations that are anchored by a very interesting molecule that can lead to a lot of patient benefit. And there's clearly plenty of unmet need remaining in myeloma that has not been addressed.

Just wanted to mention, we also have one more person here standing by from the team, which is Sascha Fauser, who is our Oncology Head and pRED -- Ophthalmology Head and pRED. So if you have any questions on the early ophthalmology pipeline, he will be the person to take. Wimal, can you please...

Wimal Kapadia from Bernstein. So first, just on zinpentraxin. This is quite a tough-to-treat population. And I would argue some of the Phase II outcomes were not overly compelling, like a 2.3% difference versus placebo and FEC. The lung volume was the same. So I guess this could be a very, very big asset, so there's a level of conviction into that early 2024 readout. I would be curious to hear that. And then maybe just a high-level question on the pipeline. If I take a step back and I look at your portfolio of assets, that readout over the next 2 or 3 years, how does Roche actually think about the risk profile of these assets? So if I'm thinking here prasinezumab in Parkinson's Phase II mixed IPF, heterogeneous population, DMD gene therapy, quite difficult. Fenebrutinib, you've never tested the drug in MS before. There's quite -- it feels like there's quite a few high-risk assets within the portfolio. So are you happy with the risk profile? Or do you think you need a few, maybe more conservative surefire bets within the portfolio?

I guess I'll take this first. Well, I think when you look at those Phase II data, and obviously, there's lots of caveats. It's a small study, it's 180 patients, only 6 months. But it's a breakthrough in the sense that no one has ever or haven't seen that type of reduction in decline in terms of functional vital capacity. So a 2% to 3% over 6 months, possibly translates into something much bigger at 1 year, which is what we're looking at now. And again, this is on top of standard of care for these patients, right? And these are not highly declining patients, at least in this study. I mean there are other Phase II trials out there, which look at much higher deltas. But again, these patients are really like decline super fast. These are normal patients on optimal therapy, and we can still demonstrate the benefit on top of that. And additionally, it's not just on functional vital capacity. You look at the 6 minutes walking test, which is one of the key secondary endpoints and that essentially mirrors how patients function, their endurance to physical stress. And you have stability over 6 months. Again, something that doesn't happen in these patients. IPF is a very severe progressive condition. So I think these data are very encouraging in that sense. Now are they definitive? Of course, not. That's why we're doing a Phase III trial. And so we're extending the duration to 1 year to really see what the benefit is. We're powering that study to really detect a robust difference. So it's from 180 patients to patients 650. And -- but we're still putting on top of standard of care for these patients. Again, at the end of the day, we want something that's clinically meaningful, right? So level of confidence, I'd say it's a very exciting new hypothesis that we're testing right now. And I don't know if you want to take the risk questions this kind of relates to the risk question.

I can start, and then if others want to chime in. So it's a great question. Certainly, something we've talked about a lot. Of course, a cynical approach to reduce risk would be to make nothing but me toos. But of course, that's not who we want to be, that's not who we are. And so in as much as your portfolio is innovative, you will carry some risk in the portfolio. It's sort of just baked into the business that we are. Now having said that, I would argue that there are several programs that we presented today that we actually feel like the risk is quite reasonable. I mean it's not 0, but we think the risk of a SERD, that's ER is a very well-validated target in breast cancer. The risk of PI3 kinase, that's a very well-validated target, important oncogene, not just in breast cancer and many other context. KRAS. There are several programs that we actually think they're reasonable value that they could bring, but also they are reasonable risk. Now there are others that are higher risk. But we -- in all of those cases, what we are trying to do is, say, well, what is the scientific rationale? Is the scientific rationale robust from the standpoint of this therapeutic hypothesis? And if the answer is yes, well, that's usually a bet that we want to consider. We're not going to -- we don't want to skew our portfolio only to those kinds of high-risk bets. We don't have enough of those because that's where the breakthrough advances over time. Some percentage of those will pay off for patients. And that's who we want to be as a company.

Maybe I could just add. So I've been part of this portfolio for 16 years. And I'd say the risk looking ahead is about like the risk looking back. It's -- that's the nature of the game. If you look -- I think in our last 22 molecules that we've had approved, 3 of them had in-class competitors. 19 were all by themselves, which can be lonely and risky, but it's also been our success. So I'm not sure I would do it in any other way, given our commitment to breakthroughs. And as Levi said, I think there's a lot of -- you mentioned DMD and the risk, but now we have data up to 4 years in those first patients, and they're rock solid. I mean, they improved and then they've stayed rock solid. That's not the natural history of muscular dystrophy. I mean this is a rapidly declining disease. And as an example, so the combination of biomarker data and that clinical data, I think, gives us -- again, it's not like, oh, this is in the bank or this is a done deal, but it gives us pretty high level of confidence given that it's a novel space.

Can I ask you to just stick to one question, please, Luisa?

I'm Luisa Hector from Berenberg. Maybe to move to fenebrutinib. I just -- I noted your comments on the number of patients treated and the profile, but how confident are you in the liver enzyme situation that competitors are experiencing? And when might we expect to see the data from the 2 different studies?

Thank you for this question. Yes, I mean, I think these are all -- all these molecules have slightly different physical, chemical properties and profiles, right? The reason why I said I feel very confident in the profile we have is based on the selectivity of this molecule. It's very high. It's the most selective BTK inhibitor out there. The non-covalent binding properties, right? And then, of course, it's big safety database. So we're not -- to your earlier question, not just jumping into Phase III with an unknown molecule. We know this molecule very, very well. And so far, we have seen some liver enzyme elevations that are transient, asymptomatic, no highs' loss cases. And we, of course, have a very stringent safety monitoring program, which means if anything ever happens, we're there. And so far, we haven't seen anything. So I do believe this has to do with the properties of different molecules, not the target itself, right, it's not like BTK inhibition per se causes liver damage, it's how do you go about it, right? And we think fenebrutinib is clearly, well, so far, at least, it seems like to have the best profile out there. The studies are recruiting well. It's still not done. Of course, there's been disruptions recently because of the war and international situation, but we're making all progress to overcome those. So I think we're still on track for a readout, I think, in 2024. I think it's the earliest date.

Okay. With that, Matthew Weston, please.

It's Matthew Weston from Credit Suisse. Bill, it's one for you. I promise I'm not asking for 2023 guidance. But at the very beginning of your presentation, you laid out a question you've got a lot was the next one of your biosimilars and how robust you thought the end portfolio growth was to mitigate that? But next year, depending on COVID, you could lose probably $2 billion of revenue from Ronapreve and the COVID, Actemra, treatment. And I'm just very interested psychologically how the organization is thinking of that. Is it like power through, keep R&D spending and the P&L will sort itself out over the midterm? Or is it, we've got productivity tucked away to partially mitigate that? Or we're going to think about margins over time? I just love to think of how the organization is thinking about it, particularly as we go into '23?

So in those particular instance, you mentioned Matthew, I think the situation is probably a little better than maybe as assumed. So we think we will have Ronapreve sales in 2023. And so it's not a hard stop. Who knows? We might have Ronapreve sales beyond that. I mean, I know that I don't want to be the most unpopular person in the room. And the pandemic conditions are over. Unfortunately, COVID is not quite in the rearview mirror yet. And so let's see. But in any case, we think we'll have some Ronapreve sales based on basically existing commitments that will be delivered in 2023. Likewise, Actemra, the COVID sales have really fallen off. I mean, in Q3, it will probably be the biggest impact on Actemra because last year, we had the delta variant. Remember, that was only a year ago, remember Delta. And so that was actually the biggest sales quarter for Actemra ever was last year Q3. And this year, it's very quiet. So it's not such a 1-year phenomenon. And then meanwhile, the growth engine is strong. I mean we're going to have Polivy in first line launching around the world, including the U.S. next year. We've got Vabysmo launching in around the world this year, and next year will be the big impact on sales because, obviously, this year is kind of like the ramp-up, but next year, the sales starts to hit in a bigger way. So I think that's why we feel good about our growth prospects, despite the -- some headwinds.

Stefan?

Stefan Schneider, Vontobel. Just if gantenerumab works, could you remind me how the testing would work afterwards to identify your patients? So there is biomarker blood testing and there is imaging, which is very expensive, who's going to be tested and what's the sequence?

You want to try? Clinically. So I'll address it from the clinical standpoint, right? So right now, patients are diagnosed, as you said, using either PET imaging or CSF testing. PET imaging is expensive. CSF testing is not just to be clear, and it's available on the Elecsys platform. So it's a very standardized procedure for neurologists. I would imagine that it will be hard to have mass testing because it is an invasive procedure. So there'll have to be a fairly good degree of suspicion that patient has symptoms compatible with Alzheimer's, right? But given that the test is there, it's not a complicated thing. That's why we developed the CSF testing in the first place. Now what we're working on now is blood-based biomarker screening, right? So blood-based testing essentially, and we're pretty close to having it. We have breakthrough device designation, we have the validation of the data. Once that is on the market, I think what that provides is that opening of the funnel for screening. So at an earlier time, if there's a suspicion, it's much easier just to do a blood test and say, well, based on this, I think you should have a CSF test or you should have a PET amyloid imaging, right? So that opens up that screening cascade basically. And it would be hopefully incorporated then into normal clinical prices. Teresa, I don't know.

No, it's spot on.

Go ahead.

I have a question on crovalimab. And next year, in terms of the readouts which coming next year, like where would you rank the importance of the Phase III data of crovalimab? And given that generalized myasthenia gravis is such an important indication for the currently marketed C5 inhibitors, why is this indication not in more of a priority?

Can you repeat the first part of the question?

Just the relative, how would you rank the importance of the Phase III data for crovalimab next year in terms of the readouts that are coming?

And when you say relative importance, so we already have positive Phase III trial data in hand from our COMMODORE 3 study in China. So speaking of risk, we think that one will be following the lower risk into the spectrum we think. And so but in terms of importance value, maybe Teresa, I don't know if you want to take that one?

Sure. I mean I think with PNH in particular, while there are PNH therapies that are approved and in the market, there remains a fairly significant level of unmet need because in many instances, while approved, those drugs have never sought reimbursement and launched. And so I think as we were thinking about crovalimab and how to bring it to market fast, we started with PNH knowing that we could take a very responsible path to actually bringing it to patients, particularly in the PNH indication, as we then continue parallel development and other indications, which are potentially more unique. And then I think Paulo will bounce to you.

Yes. No. So speaking about generalized MG, I think our approach was that both SOLIRIS and the next-generation C5s are already very well entrenched in last-line treatment for generalized myasthenia patients. So for us to go into their rare disease, it wouldn't make a lot of sense and it wouldn't be the most value that we could see out of this. On the contrary, we're thinking about Enspryng because we're trying to address the large majority of patients who don't want to go to a last-line therapy essentially and have the risk associated with that. So that's why it's just a different way to look at that clinical problem. We think that most -- the biggest medical need actually is after you are in first-line steroids or IVIg. You want to go to the next level, you want a therapy that's safe, that's convenient, that controls that disease before you have to go to last line. So it's kind of like, we think it's the best -- it's a best way to use our existing pipeline to have the most impact, basically.

Sachin?

Sachin Jain, Bank of America. Just one back on Gantenerumab and GRADUATE, if I may. So you said a couple of times today, you expect a very clear answer, which is obviously both studies say both failed. There are a lot of gray scenarios. So I just want to be clear, is there any reason you don't think gray happens? Or is there -- and if you could remind us what your intention is within the headline press release to comment on regulatory interactions or filing intention if you do end up in that gray scenario?

Well, of course, everything is gray, right? I agree with you. In science, most things are gray, they're not black and white. But the way we're thinking about it is that, at the end of the day, we're going to have to make a decision whether we think the data is robust enough and convincing enough to take the health authorities and file and try to get approval or not. And that is kind of a black and white scenario, right? You don't have file, either you file or you don't. And so the data at the end has to be that. And what I mean it has to meet that bar, it has to be statistically significant, it has to be clinically meaningful. So the effect on 2 trials have to replicate themselves to be consistent, has to have a good safety profile, obviously. I think there's lots of gray, right? There are scenarios in which we can think about, well, it didn't quite make it, but maybe in the earlier patients, it might. We can look at all of those. But the likeliness is that we're going to try and look for a black and white answer. Do we want to file or not, right? And so I guess based on that, we'll be putting out a press...

I think you said, I mean we've seen lots of mixed outcomes over the years. I can think of very few where the mixed outcome left us uncertain as to whether this is an important medicine or not. And so -- and I think knowing how we're wired, if we think -- well, we're just not the kind of company that would look at a mixed result that's kind of poor and that we don't think is really offering a compelling benefit for patients. But we're going to file it anyway because we think there's some way to make money or -- that's not going to happen. I mean where you're going to have a compelling benefit for patients or we're not going to have a medicine.

Richard?

Richard Vosser from JPMorgan. Maybe on the PI3 kinase alpha. How is the tolerability looking on a blinded basis in terms of the combination? I hear you that it's more selective and safe better tolerability. But what are you seeing in terms of discontinuations and reinforced diarrhea, I think that would be the major concern.

So for the Phase III study, we can't comment sort of in an ongoing way on what the tolerability profile looks like, except that obviously, like with all Phase IIIs, we have IDMC, we have safety -- regular safety reviews, and there's been no red flags during any of those. So we can't comment on the Phase III. All we can comment on is the earlier phase data and certainly in the earlier Phase data, there are some side effects like hyperglycemia that has just baked into it. This is just on target, on mechanism toxicity. But in terms of the -- some of the other challenges and the frequency of challenges that cause early discontinuation, in the earlier data set, it's obviously limited by small numbers, but we're encouraged by the ability of patients to stay on the regimen. So it's not tox-free, but it's manageable in a way that we think could be differentiated.

Maybe the final question. It's for Simon.

Simon Baker from Redburn. Probably another one for Levi. On the slide about allogeneic cell therapies, 2 of the reasons you cited disadvantages, of course, are scalability and no delay to first treatment. You're also involved in individualized cancer vaccines. And last week, Moderna was saying that they're targeting a biopsy to therapy time of about 6 weeks. It's actually slower than cell therapy. So I was just wondering what you're targeting and how much you see that as a problem? And also the general scalability of individualized mRNA cancer vaccine?

Well, so first of all, I'll point out that those are -- at some level, those are kind of apples to oranges because in the cell therapies, for example, these would be heme malignancies, and oftentimes you're dealing with aggressive malignancies, and you really struggle to sit on your hands at all once you have a patient with active disease. So the time to treatment matters very much in that setting. So -- and that's, as you know, in the real world with autologous CAR-T therapy, that can be a fundamental constraint to use. Now in the setting of a cancer vaccine, which, of course, as you mentioned, we have several programs ongoing. The context can be very different. In fact, if you think about it, that sort of scenario is exactly when you don't use a vaccine. It's like you don't use a vaccine, we have floored covered uses to prevent. So actually, a lot of the thinking for the cancer vaccines is you have a patient who basically had definitive treatment, but they're high risk for recurrence. And in that setting, you actually have a bit of time. You've done the resection, you can do at some sequencing, you can find the new antigens. You have time to kind of engineer a more personalized approach while the disease is still not detectable. So that would be the sort of concept behind the cancer vaccine.

Okay. with that, I think we will close the session, and our speakers will still be around. So you will still have the opportunity to wrap them and ask additional questions. I would also like very much to use the occasion to thank all the speakers for their dedication and the time they put in the preparation and also Sascha was online. And then also, of course, to the IR team, I think the IR team has been busily working on all -- preparing all this here. So it's like [indiscernible] Lauren and Anita and Jared, who all contributed over several weeks. And of course, all the pharma teams who contributed and also our colleagues from finance. With that, I would like to close -- Bill, do you have any final comments.

Well, again, we started out this morning talking about how glad we were to have a chance to just interact with you all in person. It's been great. Enjoyed catching up with a number of you at lunch as well. We're super excited about the trajectory of the medicines we have in the countries around the world today, the readouts we have in the next 12 to 18 months and the long-term future with our amazing early R&D centers and the practice-changing medicines that I know they're going to deliver. So anyway, thanks for your confidence in Roche over the years. I know many of you are long-term holders and -- or long-term followers, and we aim to deliver. So great -- again, great to see you all, and wish you a great week.